Google

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project

to make the world's books discoverable online.

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover.

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the

publisher to a library and finally to you.

Usage guidelines

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to prevent abuse by commercial parties, including placing technical restrictions on automated querying. We also ask that you:

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for personal, non-commercial purposes.

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the use of public domain materials for these purposes and may be able to help.

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find additional materials through Google Book Search. Please do not remove it.

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner anywhere in the world. Copyright infringement liabili^ can be quite severe.

About Google Book Search

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web

at|http: //books .google .com/I

^/,UV

i

\

AN INTEODUCTION

TO THE

GEOLOGY OF CAPE COLONY

AN INTEODUCTION

TO THE

GEOLOGY OF CAPE COLONY

^ lA BY

Arwy EOGEKS, M.A., F.G.S.

DIRECTOR OF THE GEOIX>OICAL SURVEY OF CAPE COLONY

WITH A CHAPTER

ON THE

FOSSIL REPTILES OF THE KABBOO FORMATION

BY

PROF. R. BROOM, M.D., B.Sc, C.M.Z.S.

OF VICTORIA OOLLEGE, 8TBLLEKB08CH

WITH ILLUSTRATIONS AND COLOURED MAP

LONGMANS, GREEN, AND GO.

39 PATERNOSTER ROW, LONDON NEW YORK AND BOMBAY

1905

PREFACE.

A GENERAL account of the Geology of Cape Colony has long been wanted. The best descrip- tion yet published is that of the late Professor <=> A. H. Green, *' A Contribution to the Geology

and Physical Geography of the Cape Colony," >A which appeared in the Quarterly Journal of the

t Geological Society of London for 1888. This

^ essay is not so difficult to obtain as many other

papers published in English or foreign journals, but in some respects it is now known to be in- accurate, and it is of course very incomplete.

In 1895 the Cape Government appointed the Geological Commission for the purpose of or- ganising a Geological Survey of the Colony. The Survey thus established commenced work '^ in 1896, and though its work is still very far

'^ from being complete, even as regards the filling

up of the inadequate maps that are at present the only available ones for the purpose, yet sufficient information has been collected to

I

lo

»>r.

M)

•^ TO- 4.

135 J 03

vi PREFACE

decide many disputed points concerning the fundamental structure of the country, and to enable one to bring the observations of earlier writers on areas that have not been systemat- ically surveyed into harmony with the results obtained. When, therefore, the publishers, on the initiative of Dr. Muir, the Superintendent- General of Education, asked me to undertake the compilation of a geological description of the Colony I agreed to do so, with the consent of the Geological Commission. This work is the first of a series designed by Dr. Muir to promote the study of Natural Science in South Africa.

The chief object of this book is to help students and other people in the Colony to un- derstand the structure of their country and to pursue the subject for tlieniselves. I have, how- ever, taken it for granted that the reader has an elementary knowledge of Geology. There are so many excellent introductory text-books on the principles of the science that it would have been superfluous 'for me to attempt to combine with this description of Cape Geology what has been well done by others.

The following description is necessarily in- complete, for large areas in the (Colony, including the whole of the country north of the Orange

PREFACE vii

River and immense tracts in the north-western, midland, and eastern districts have not yet been surveyed, and nothing more than the broad outlines of their Geology is known. I have naturally devoted most space to those parts of the Colony that are best known geologically.

The earliest comprehensive geological map of the Colony is that of A. G. Bain (1866), who was a self-trained observer of great ability. His map is at once a proof of his grasp of the structure of the country and a most remarkable work for one man to. have accomplished. Other men who were closely concerned in laying the foundations of Cape Geology were Dr. W. G. Atherstone, A. Wyley, G. W. Stow, and E. J. Dunn. A full account of the development of opinion on the more important geological features has been written by Dr. G. S. Corstorphine, under whose direction the Survey was carried on during the first six and a half years of its existence ; it will be found in the Annual Report of the Commission for 1897.

In an appendix I have given the titles and dates of papers referred to in the footnotes and made use of in preparing this book. The numbers in brackets after authors' names in the footnotes refer to the year of publication, but in

viii PREFACE

the case of the Annual Reports of the Geological Commission the number indicates the year on the work of which the Report was written, for the Annual Reports have not appeared regularly. I especially wish to draw attention to the pub- lication of descriptions and figures of Cape fossils in the Annals qf the South African Museum. The plants of the Karroo and Uitenhage for- mations, and many of the Bokkeveld fossils have already been dealt with.

There can be few countries whose geological structure has had such an obviously direct in- fluence upon the form of the present surface as is the case in this Colony. The thick soils and rich vegetation, which in more humid climates may be the chief compensation for the lack of facility for the study of Physical Geology, rarely seriously interfere with geological investigation in Cape Colony, though there are parts of our country that may be compared with any in the world in respect of beauty due to vegetation and form combined. Physical geography can be made a very good means of education, and there are few towns or villages in the Colony where a teacher with a knowledge of the subject cannot find striking examples of many important principles within reach of an afternoon's walk.

PKEFACE IX

Encouragement given to pupils to form collec- tions from the neighbourhood is at once the means of their instruction and pleasure, and discoveries of both scientific and practical value may also be the result.

It may be well to point out here that a geo- logical specimen loses at least the greater part of its value and interest in the absence of a record of the locality whence it came, and also that when a large fossil, e.g., a reptilian skeleton in the Karroo formation, is found, it is better to leave it in the rock till some one who under- stands such things can get it out than to carry away part of it. The partial removal of skeletons has been the cause of great confusion in certain cases, even to the extent of being the cause of two or more generic names for diflferent parts of one species. Should there be no suitable pro- vision for the preservation of fossils in a local museum they should be sent to the public collections, such as the South African Museum, Cape Town, where they will be made good use of. Any available information concerning fossils or rocks can be obtained there. ^

^ It may not be out of place to mention that boxes or other parcels of fossils and other natural history specimens addressed to the Directors of the Public Museums are carried free on the Cape Govern- ment railways.

X PREFACE

I have much pleasure in thanking Professor Broom of Stellenbosch for assistance regarding the names of the reptilian fossils, and for his chapter on the reptiles of the Karroo formation ; Mr. F. L. Kitchin, of H. M. Geological Survey, has kindly given me the correct names of the Uitenhage and Pondoland marine fossils and notes on their relationship to foreign Cretaceous faunas ; and lastly my best thanks are due to my colleague, Mr. E. H. L. Schwarz, who has made many and valuable suggestions during the preparation of this work, and who gave me the photographs reproduced on Plates vi., xix. and XX., and the notes on the Geology of the Ros- mead-Port Elizabeth and Willowmore lines. So much of the field work upon which this account chiefly depends has been done by Mr. Schwarz, and so intimately have we been associated in the Geological Survey of the Colony during the past eight years, that the credit of any advance upon previous views on Cape Geology is very largely due to him. There can be few questions which have suggested themselves during the progress of the Survey that we have not discussed to- gether, usually in the field, and without in the least desiring to make him responsible for views that in the nature of the case are doubtful, and

PREFACE XI

which are certain to be modified, if not altogether rejected, when fuller knowledge is obtained, I wish to acknowledge my great debt to him.

AETHUB W. EOGEES.

Capb Town, 2Qth March^ 1904.

Note on the Map.

The accompanying map has been compiled from various sources. The south and west, from the Olifant's River to Knysna, inland as far as the Boggeveld-Nieuweveld escarpment, the Prieska district, and the Transkei have been taken from the field maps of the Geological Sur- vey. The rest of the map is based upon the previously published maps of A. G. Bain, G. W. Stow, and E. J. Dunn. The portions of Natal and the Transvaal included within this map are taken from C. L. Griesbach and G. A. F. Molengraaff. The Orange River Colony is filled in according to E. J. Dunn, with modifications due to the work of A. C. Seward and T. N. Leslie on the fossil plants of Vereeniging, etc., and to information that has reached the compiler from other sources.

The dolerite intrusions are only very partially represented, as the details of their distribution north and east of the Nieuweveld - Roggeveld escarpment are unknown ; they extend farther north than the limit of this map.

Xll

CONTENTS.

CHAP. PAOB

I, iHTBODUCnON 1

%

II. Thb Pbe-Capbj Bocks of the South and West of the

COIiONY 82

III. The Pbb-Oapb Rocks of the Nobth and Nobth-West. 63

IV. The Oapb System 98

V. The Kabboo System 146

VI. BeptiiiBS of the Kabboo Fobmation .... 228

VII. The Intbusive Dolebites and Allied Bocks . . 245

VIII. The Obetacboub System 281

IX. Volcanic Pipes Youngeb than the Stobmbebg

Volcanoes 331

X. BbCENT OB SUPBBFICIAL DEPOSITS 351

XI. Thb Gbolooioal Histoby of the Colony . 393

XII. Notes on thb Gboloqy of Some of the Kailway Lines 425

Appendix: List of Books and Papebs Bbfebbbd to in

THE Body of the Wobk 445

Xlll

LIST OF ILLUSTRATIONS

PLATO PAOl

Geological Map Frontispiece

Fig. 1. Section drawn about N. 10° E. from Mossel Bay to

the Orange Biver 14

Fio. 2. Section through the Bokkeveld Mountain es- carpment to Galvinia 20

Fig. 3. Diagram to show the three regions in Cape Colony

and adjacent territory 27

Fig. 4. Section from the Pondoland coast to Lusikisiki . 28

Fig. 5. Section through the Worcester Fault ... 29-

I. Waai Kloof, Worcester 85

Fig. 6. Section through the Cango and Zwartebergen, eleven

miles east of Prince Albert Village ... 50

Fig. 7. Section through the Cango from Potgieter s Poort

to the Zwartebergen 50

Fig. 8. Section from the Van Rhyn's Dorp flats to the

plateau above Loeries Fontein .... 61

Fig. 9. Section across the Prieska Division .... 69

Fig. 10. Section through Ezel Band 77

Fig. 11. Section from Piquetberg to the Karroo ... 97

II. Matsiekamma from the N.W. . . . . « .98

Fig. 12. Section through the Warm Bokkeveld and S.W.

corner of the Karroo 100

Fig. 13. Section through the Langebergen in the neighbour- hood of Oudebosch beacon showing the nature of the folding 103

b

xvi LIST OF ILLUSTRATIONS

FLATS PAOB

III. Contorted and overfolded quartzites of the Table Mountain

series 105

FiQ. 14. Fossils from the Bokkeveld beds . . . 124, 125

lY. View in the Cold Bokkeveld showing succession from the Table Mountain sandstone of Schurfteberg to the Witteberg beds of Tafel Berg 127

V. Blink Berg in the Cold Bokkeveld 128

VI. An anticline in the Witteberg beds at Tyger Fontein in

Prince Albert 141

VII. Dwyka conglomerate 149

VIII. Roches mouUmrUea exposed by the removal of the Dwyka conglomerate from the 'Keis quartzites at Jackal's Water, Prieska 156

IX. Near view of one of the glaciated surfaces at Jackal's

Water, Prieska 167

iX. Escarpment of the Dwyka conglomerate near Ibiquas

Biver, Calvinia 161

XI. Dwyka conglomerate with a band of boulders, Witteberg's

Kiver, Laingsburg 167

Fig. 15. Plants from the Ecca beds 187

Fig. 16. Section from the Wittebergen to the Klein Bogge-

veld, from the folded belt to the Karroo basin . . 194

Fig. 17. Plants from the Stromberg series (Molteno beds) 201

XII. A spur of the Drakensbergen near N'quatsha's Nek . . 209

Fig. 18. Skeleton of PareiasaurtLs serridens (Owen), restored 232

Fig. 19. Skeleton of Oudenodon trigoniceps (Broom), restored 237

Fig. 20. A. — Skull of a Therocephalian, Lycosuchus van- derrieti, B. — Skull of a Theriodont, Cynognathtis platyceps. C. — Skull of a Mammal, Dasyurtis maculatus 241

XIII. A dolerite sheet at Paalhuis under the Nieuweveld es-

carpment 248

XIV. The falls of the Tsitsa River in East Griqualand . . 249

Fig. 21. Map of Kentani showing the distribution of

dolerite sheets and " gap " dykes 259

LIST OF ILLUSTRATIONS xvii

FLATS PAGE

XV. Dyke of granophyre (light-ooloured) traversing a thick sheet of dolerite near mouth of Kobonqaba Biver, Eentani 261

XVI. Columnar structure in Dwyka conglomerate produced by

the overlying sheet of dolerite 277

XVII. Surface formed by a dolerite sheet in the Fraserburg Division, near the road between Fraserburg and Williston 279

Fia. 22. Plant from the Uitenhage series (Wood beds) . 288

FiQ. 28. Plant from the Uitenhage series (Wood beds) . 289

Fig. 24. Plant from the Uitenhage series (Wood beds) . 290

Fig. 25. Fossils from the Uitenhage series (Sunday's Biver

beds) 291

XVIII. Cretaceous limestones on the coast 820

Fig. 26. Fossils from the Umzamba beds .... 324 Fig. 27. Sections of the rock-shafts, mines of the Kimberley

area 341

XIX. High-level gravels lying unconformably upon inclined beds of Uitenhage age (Enon type), Paarde Kloof, near Tover Water Poort, Uniondale 854

XX. Gravel- and quartzite-capped terrace and outlier of the* same; north side of the Kouga mountains, near Uniondale 858

XXI. False-bedded limestone near Struys Point, Bredasdorp . 875

CHAPTEE I.

INTRODUCTION.

The backbone of the Cape Colony is the watershed between the rivers that drain into the Atlantic and those which flow south and east into the Indian Ocean. The watershed lies in a general east-north-east direction from the neighbourhood of Ceres and Tulbagh, where two systems of moimtains converge, the Cederbergen and those parallel to them on the west, with a north and south trend, and the Langebergen and parallel ranges on the south, with an east and west trend (see Fig. 3). The watershed is formed by the Klein Kog- geveld, Nieuweveld, Winterbergen, Stormbergen and Drakensbergen, and as a whole it is the highest belt of ground in the country, although certain peaks in the southern and western mountains rise to a greater height than many parts of the watershed. From this main water-parting the surface slopes gradually northward to the Orange Eiver, by which the greater part of the area north of the watershed is drained. Towards the west coast the country which feeds the rivers running directly to the Atlantic^ south of the Orange Biver is consider- ably broken ; the two escarpments of the Eoggeveld and the Bokkeveld Mountain, which eventually become one feature about eighty miles north of Calvinia, bring the

2 GEOLOGY OF CAPE COLONY

level of the surface from some 5,000 feet down to 500 feet above the sea. South of the Bokkeveld Mountain (an important escarpment west of Calvinia which must not be confused with the mountains of the Cold Bokke- veld in Ceres) the Cederberg chain commences, and forms, together with its subsidiary parallel ranges, a broad belt of mountainous country rising to the height of 6,000 feet between the Karroo and the coastal district. The southern drainage slope is also very different in the west and east. In the west there is a sharp drop immediately south of the watershed, and the Great Karroo lies between it and the Zwartebergen, which rise to a height of over 7,000 feet above the sea, and some 5,000 feet above the Karroo. The Zwartebergen, Langebergen, and the minor ranges parallel to them, run nearly east and west, together forming a wide tract of mountainous country which stretches from Tulbagh to the Indian Ocean east of Grahamstown. This belt is traversed by the rivers flowing from the Karroo, generally in deep, steep-sided valleys, which become gorges in the mountain ranges. There are many longi- tudinal valleys in this region much more open and less steeply graded than those of the transverse rivers into which their waters flow. The country between the Zwartebergen and Langebergen, occupied by longitu- dinal valleys, lies somewhat lower on the average than the Great Karroo. South of the Langebergen the surface slopes towards the coast, but it is deeply cut into by rivers, and diversified by mountains such as Aasvogel Berg, Pot Berg, and the mountains of Caledon and Bredasdorp.

INTRODUCTION 3

In the eastern part of the Colony, beyond the Gualana Eiver where the southern mountainous region is cut through by the coast, the descent from the watershed to the coast is more uniform than in the west; it is un- broken by mountain ranges, but is more of the nature of a succession of terraces than a gradual slope. There is no area in the east corresponding to the Great Karroo of the west and midlands ; the rain borne by the south- east winds waters the Eastern Province from the coast to the watershed, but the Great Karroo is deprived of this source of water by the mountains on its southern border.

The geological structure of the Colony is in its main outlines fairly simple ; the country may be looked upon as a shallow basin filled in with nearly horizontally lying rocks, those of the Great Karroo system. The character of the edge of the basin is very different in the north and south, and the basin form is due rather to movements in the earth's crust, which took place after the deposition of the rocks now filling the basin, than to the original shape of the surface on which the rocks were laid down.

Before describing further the structure of the Colony, it will be convenient to give a general account of the various groups of rocks that build it up. The classi- fication of these rocks, which will be used in this book, is as follows : —

GEOLOGY OP CAPE COLONY

Recent and sub-recent deposits :— Sand dunes and consolidated

dunes, calcareous tufa ; alluvial deposits and gravels of low and high levels ; laterite and surface quartzite.

(Cretaceous series of Pondoland. I^day River beds. Uitenhage series - Wood beds.

(Enon conglomerate.

(Volcanic beds. Cave sandstone. Red beds. Molteno beds. rBeds containing Therio- J donts. I Dicynodon beds. IPareiasaurus beds, r Shales and sandstones. i Laingsburg beds, [shales and sandstones. ( Upper shales. J Conglomerates. ( Lower shales.

(Unconformity in north.)

Karroo sjrstem

Beaufort series

Ecca series

Dwyka series

Cape system

( Witteberg series.

} Bokkeveld series.

( Table Mountain series.

In south and west:— In north and north-west : —

Pre-Cape rocks

Ibiquas series.

Matsdp series.

Volcanic rocks of Beer Vley, etc. ?

Cango series.

Malmesbury series.

Griqua Town series. Campbell Rand series.

'Keis series.

Namaqualand schists. Unconformable bases are indicated thus :

The Pre-Cape rocks include a great variety of sedi- ments, of which the original characters have in most cases been greatly changed by the pressure exerted

INTKODUCTION 6

during the earth moveraents that took place before the deposition of the rocks forming the Cape system; the movements subsequent to the Cape system probably affected the Pre-Cape rocks in the south and west of the Colony only. The intrusion of the great masses of igneous material, mostly of an acid type, previously to the formation of the Cape system, brought about con- siderable alteration in the Pre-Cape rocks in the south, west, and north of the Colony. The subdivisions of Pre-Cape rocks and their igneous intrusions will be described in the next chapter, and further details are not necessary at this stage. It is sufficient to note that the ages of these rocks — for we shall find that they include several independent formations separated by great unconformities — are unknown, except that they are older than the Cape system. As yet, no organic remains have been described from the Pre-Cape rocks, and it is therefore impossible to correlate them with the rocks of foreign countries. The Pre-Cape rocks occur in the south-west and north of the Colony, and form vast tracts of country in the north-west (Namaqualand, etc.) and to the north of the Orange Eiver.

The Cape system is composed of sandstones, quart- zites, shales and mudstones, arranged in three series. The lowest or Table Mountain series is chiefly sand- stone, with occasional pebbles of white quartz ; beds of conglomerate are rarely seen ; two thick bands of shaly material are usually met with, one near the top and one near the bottom of the series. The approximate maxi- mum thickness of the series is 5,000 feet. The group

6 GEOLOGY OF CAPE COLONY

forms the great coastal ranges of the Colony, and takes its name from Table Mountain behind Cape Town.

The second group in the Cape system is the Bokkeveld series ; it comprises shales and thin sandstones inter- bedded with thick layers of more or less argillaceous sandstones, which are arranged in a definite order recognisable over wide areas. The maximum thickness of the Bokkeveld series is about 2,500 feet. Towards

the lower part of the series considerable numbers of fossils occur ; they are marine forms, and some of them are identical with or closely related to species which are found in Devonian rocks of America and Europe. They afford the earliest evidence we have for the chronological comparison of the geological history of the Colony with that of other countries. The Bokkeveld series occupies wide areas in the south of the Colony, and takes its name from the Warm and Cold Bokkevelds in Ceres, where it is typically developed. Wherever the base of the series is seen the junction with the underlying Table Mountain series is a conformable one.

The Witteberg series, a group of shales, thin sand- stones and quartzites, about 2,500 feet thick, is the highest division of the Cape system. It contains, so far as is known, very few fossils, and these are of veget- able origin. The series takes its name from the Witte- bergen, south of Matjes Fontein, in the south of the Karroo, and forms several long and high ranges of foot hills north of the Zwartebergen. It lies conformably upon the Bokkeveld series.

The Cape system rests unconformably upon the older rocks wherever the junction between them has been

INTRODUCTION 7

observed. Between Karroo Poort in the west and the Gualana Eiver in the east, the lowest beds of the Karroo formation rest conformably upon the highest of the Cape system. To the north of Karroo Poort, however, the Dwyka series is found to lie upon lower and lower members of the Cape system as it is followed north- wards to the end of the Bokkeveld Mountain, where it rests directly upon the Pre- Cape rocks. Near the mouth of the Gualana Eiver the Cape system disappears be- neath the sea, and where it reappears in Pondoland the two upper members are missing, and the Table Moun- tain series is unconformably overlain by the Dwyka conglomerate.

The Karroo system forms by far the greater part of the surface of Cape Colony ; from the 33rd parallel of latitude northwards to the Orange Eiver, with the ex- ception of the country west of the Prieska division, the rocks belonging to this system form practically the whole surface of the country. Outliers of the Karroo system, including at least the two lower series, have been found south of the main area occupied by it ; they are insignificant in extent, but they are important on account of the evidence they afford of the former southward extension of the Karroo rocks. By far the most interesting outlier is that between Worcester and Eobertson, where the Dwyka and Ecca have been faulted down against the Malmesbury (Pre-Cape) beds. The Dwyka series forms the base of the system, and occurs as a continuous band round the area occupied by the higher beds, The series consists of a varying but

8 GEOLOGY OF CAPE COLONY

usually considerable thickness of conglomerate, which is both overlain and underlain by shales in the south of the Colony ; in the west and north, where the Dwyka rests unconformably upon the older rocks, the lower group of shales is absent. The maximum thickness of the series is over 2,000 feet. The conglomerate is of very great interest on account of its glacial origin.

The Dwyka series is overlain conformably by the Ecca, a group of shales and sandstones containing plant remains belonging to several genera found in many other parts of the world, and these fossils form the second important bench mark for comparing the rocks of the Colony with those of other countries. The thickness of the Ecca beds is about 2,000 feet in the west of the Karroo, and some 2,600 in the south-west and south.

The Beaufort series, distinguished by containing the remains of several forms of reptiles, succeeds the Ecca without any break in the western Karroo, in fact it is often difficult to draw the line between the two series. Shales, mudstones and sandstones, to the thickness of at least 3,000 feet, compose the Beaufort series, which is so named from its occurrence in Beaufort West and Fort Beaufort.

The boundary between the Beaufort and the overlying Stormberg series has never yet been closely defined. The Stormberg beds contain a number of plants and reptiles distinct from those in the underlying rocks, by means of which they can be readily identified. The lower part of the series consists of shales and sand- stones with seams of coal. At the top of the ordinary

INTRODUCTION 9

sedimentary rocks in the Stormberg group there is in places a pecuHar set of beds called the Cave sandstone, with which are associated the lowest of the volcanic rocks of the Stormbergen and Drakensbergen. The thickness of the Stormberg beds, excluding the volcanic rocks, is perhaps about 3,000 feet, and the volcanic beds in some localities must be 4,000 feet thick. The upper- most portion of the series has been removed by denuda- tion, and the volcanic beds now form the highest points of the surface of the Colony, the peaks of the Drakens- berg in East Griqualand. So far as is known at present the Stormberg series only occurs in the higher parts of the country east of Steynsburg; outside our limit it forms the greater part of Basutoland.

One of the chief characteristics of the country occupied by the rocks of the Karroo system is the abundance of dolerite intrusions which are met with in all parts of the system from the Dwyka to the Stormberg series. It is not unlikely that these intrusions belong to one period of igneous activity, which commenced during the deposition of the Stormberg series, and that they were closely connected in origin with the volcanic outbursts that took place towards the close of the Stormberg period.

The rocks belonging to the Cretaceous system in the Colony are divided into two groups, which occur in widely separated localities and in dififerent manners, but the evidence of the fossils is sufficient to prove that one group is considerably older than the other, although both present close affinities to the Cretaceous rocks of other parts of the world.

10 GEOLOGY OF CAPE COLONY

The older, or Uitenhage, series forms several dis- connected areas between Worcester in the west, and Alexandria in the east of the Colony. The lowest part of the series is almost always a conglomerate, usu- ally overlain by shales and sandstones containing the remains of fresh- water and land animals and plants; in the eastern districts the beds of fresh-water origin are in turn overlain by clays, shales and limestcmes with marine fossils, related to forms found in the Lower Cretaceous and Upper Jurassic of foreign countries. The more important areas of the Uitenhage formation are in the divisions of Uitenhage, Knysna, Oudtshoorn and Biversdale. The Uitenhage beds everywhere lie unconformably upon the older rocks, from the Pre-Cape to the Ecca. The unconformity is always very pro- nounced, and proves that the older beds had been intensely folded and had been exposed to denudation for a long period before the Uitenhage beds were deposited. The maximum thickness of the series is probably not less than 2,000 feet, but the top of it is nowhere seen.

The chief outcrop of tlie Pondoland Cretaceous series occupies a narrow strip of country, about ten miles long and half a mile wide, on the Pondoland coast. It is faulted down against the Table Mountain series. The rocks are sandy clays and shelly limestones remarkably rich in fossils, many of which are related to, or identical with, species that are found in the Cretaceous rocks of Southern India. A similarly situated strip of conglom- erate and sandstones is found near the mouth of the Embotyi Eiver, and very probably belongs to the same

INTRODUCTION 11

series, but palaeontological evidence to prove this point has not yet been found. The Embotyi rock is of great interest on account of the boulders of Karroo dolerite imbedded in it.

The Eecent deposits of sufficient importance to be mentioned here are the sand dunes, and the limestone resulting from their consohdation by the deposition of carbonate of hme from solution between their com- ponent grains; these rocks are found on many parts of the coast ; the quartzitic sandstones and conglomer- ates, produced by the cementation of sands and 'gravels of alluvial origin, found over wide areas between Mal- mesbury in the west and the Transkei in the east ; and certain rocks related to laterite. These are all found lying unconformably upon the older rocks in their neighbourhood, generally in thin layers, but in places the limestone derived from dune sand may reach a thickness of 500 feet. So far as the fossils in these rocks have been determined they all belong to species still living in South Africa.

It has already been stated that the structure of the Colony may be likened to a shallow basin occupied by the Karroo formation. The basin extends much farther than • the limits of the Colony, for its northern edge traverses the Transvaal in a north-easterly direction, and practically the whole of the Orange Eiver Colony, Basutoland, and part of Natal, lie within it. On the south-east the edge of the basin is cut into by the Indian Ocean between the Gualana and St. John's Bivers.

12 GEOLOGY OF CAPE COLONY

For the purpose of a more detailed description, the Colony may be divided into three regions : (1) that of the Pre-Cape rocks of the north and west ; (2) the belt of folded rocks belonging to the Cape and Karroo systems, extending from near Van Bhyn*s Dorp to the neighbourhood of the Peninsula, then turning east- wards and finally disappearing beneath the sea near the Gualana Eiver; (3) the region of the plains and plateaux of the interior of the Colony, the area lying within the basin, part of whose edge is formed by the first two regions. This division of the Colony, while convenient for descriptive purposes, brings out strongly the contrast between the northern and southern edges of our basin.

(1) The region of the Pre-Cape rocks in the north and west of the Colony is largely composed of granite and- foliated rocks of igneous origin ; the sedimentary beds invaded by these, together with more recent beds of Pre-Cape age, form, however, great areas in the north and in the south-west.

The nature of the rocks and the structure of the country are less known than those of either of the two other regions, especially in the case of the vast semi- desert country lying west of Prieska. In Prieska, and the country north of the Orange River in that neigh- bourhood the folds into which the rocks have been thrown have a marked effect upon the surface features ; the Doornbergen, for instance, are a range of hills trending north-west along the strike of the rocks com- posing them, and the Ezel Band, lying almost at

INTRODUCTION 13

right angles to the Doornbergen, is found to consist of sedimentary rocks with a corresponding north-easterly strike. The same appears to be the case with the Kaap plateau, the Langebergen, and other ranges in Griqua- land West described by Stow.^ Some of these features are of very great antiquity, older than the Dwyka con- glomerate, which rests in the valleys between the hills. These ranges do not reach a great height above the surrounding low ground, and are different in this respect from the mountains of much later origin that diversify the second and third regions. Stow ^noticed the re- markably rounded form of many of these hill ranges, and attributed them to glacial action, but to glaciation of a much more recent date than can now be admitted ; for since these rounded surfaces have been found passing under the glacial conglomerate at the base of the Karroo formation, we must conclude, in the absence of evidence of recent glaciation, that all the characteristic glacial features observed on the ancient surface were produced during the Dwyka period. The main surface features of the Pre-Cape rocks of Prieska are thus probably due to denudation during Dwyka and Pre-Dwyka times; they have been buried under an unknown thickness of rocks belonging to the KajTOO formation, and have been gradually exposed again by the removal of these over- lying beds. The north end of the section in Fig. 1 illustrates the relationship of the Karroo formation to the underlying rocks of Prieska.

The strike of the Pre-Cape rocks in Griqualand West and the trend of the hills carved out of them is north-

» Stow (73). ^ Ibid., ^. 666.

INTRODUCTION 15

easterly, while the same rocks in Prieska have usually a north-west strike. In the south of the Colony the strike of the Pre-Cape rocks has an intimate connection with the trend of the folds which involved the Cape for- mation and the lower members of the Karroo system, the result of earth movements that did not aflfect the northern area.

West of the Prieska district lie Kenhardt and Little Namaqualand, including the very dry and sandy area called Bushmanland. Beyond stating that there are great tracts of granite and gneiss, the disintegration of which gives rise to the sand covering large parts of Bushmanland, there is little to be said about that country at present owing to lack of knowledge. In Little Namaqualand there is much granite and gneiss continuous with the similar rocks of Bushmanland, and the Namaqualand schists, partly metamorphic rocks of igneous origin. The country is hilly with much sand in the valleys, and the river courses are ill defined, as is usually the case in the dry districts in the north of the Colony. Some outUers of quartzites are stated by Mr. Dunn to belong to the Witteberg series, otherwise there seem to be no rocks later than those of Pre-Cape age in the north-west, outside the limit of the Karroo forma- tion which bounds the region on the south and east.

South of Namaqualand the coast country lying west of the escarpment in the north, and the folded ranges further south that bound the coastal plains on their inland side, falls within the first region, which reaches the shores of False Bay. The southern part of this area is studded with large and small outliers of the

16 GEOLOGY OF CAPE COLONY

Table Mountain series, which must formerly have covered the whole of it, at least as far north as the 3l8t parallel. The greater part of the area consists of slaty rocks with high dips striking some degrees west of north, more or less parallel with the ranges of folded rocks forming the eastern boundary of the region south of the Doom Kiver. Several large masses of granitic rocks intrusive in the slates form important ranges of hills; the chief one is that which extends from St. Helena Bay southwards to Mamre, a distance of some seventy miles; other hills of granite are the Paarl Mountain, and the Paarde Berg-Malmesbury range. These granite hills, and the smaller ones carved out of Malmesbury beds, owe their preservation more to the weather-resisting qualities of the rocks which form them than to their structure, though the parallelism of the trend of the ranges with the general strike of the rocks shows that the structure of the area has de- termined its leading features. It is difficult to discover how far the present surface features are due to denuda- tion eflfected since the removal of the covering of Table Mountain sandstone, but the occurrence of large hills of Pre-Cape rocks near areas of that sandstone, such as the Lion*s Eump near Cape Town, and the slate hills at the south-east end of Eiebeek's Kasteel, point to the protection afforded these slate hills by former extensions of the sandstones of the Lion's Head and Eiebeek Kast- eel now removed by denudation. In the Prieska district we find that the main surface inequalities of the Pre- Cape rocks are older than the Karroo formation that once covered them, but a corresponding relation be-

^

INTRODUCTION 17

tween the surface features of the Pre-Cape rocks in the south-west and the overlying Table Mountain series has not been made out, in fact the evidence so far as it goes, e.gf. the approximately plain surface of granite and slate under the northern boundary of the Peninsula outlier, points to the present surface features in the Cape, Malmesbury and Stellenbosch Divisions being due to denudation since the removal of the bulk of the Table Mountain series.

The southern part of the region is, in marked contrast to the northern portion, a well-populated, fertile land, in which good crops are raised annually and the wine and fruit-growing industries are second to none in the Colony. In the north, except in the as yet small areas watered by artificial irrigation, but little in the way of agriculture is attempted, and cattle and sheep are the mainstay of the farmers.

(2) The second region is the folded belt which runs in a southerly direction from Van Ehyn's Dorp to the neighbourhood of the Peninsula, there turns eastwards, and is continued as far as the mouth of the Gualana Eiver, where it is cut off by the sea. This area is chiefly composed of the three members of the Cape system, the lowest of which, the Table Mountain series, forms the mountain ranges of the Cederbergen, Draken- steins, Langebergen and Zwartebergen, to mention only some of the more important ones, which are such strik- ing features in the south of the Colony. In addition to the Cape formation, the lower parts of the Karroo system,

the Dwyka and Ecca series are involved in the folding,

2

18 GEOLOGY OP CAPE COLONY

and this fact has great significance in that it proves that the folding took place chiefly after the deposition of the Ecca series. The later limit of the period daring which the folds were produced is fixed by the presence of con- siderable areas of comparatively undisturbed beds be- longing to the Uitenhage series lying upon the upturned edges of the folded rocks belonging to all ages from Pre- Cape to Ecca.

The folded belt is flanked on the outside by the Pre- Cape region in which these earth movements produced but little effect, and on the inner side by the almost horizontal strata of the Karroo. The Cape Peninsula and the districts north of it where the Table Mountain sandstone lies nearly flat are on the outer side of the folded belt in the Pre-Cape region.

At its broadest part the folded belt is about 100 miles wide, from the southern part of the Karroo to Cape Agulhas, and its length along the bend is some 600 miles. The most marked character of the region is the presence of many mountain ranges, which are mostly formed by great anticlinal or arch-like ridges of the folded strata. A glance at the map will show that the general trend of these mountains is roughly parallel to the coast; on the western side the Ceder- bergen, Witzenbergen, Cold Bokkeveld Mountains, and other minor ranges, run a little west of north ; while on the south, where the Langebergen, Zwartebergen, and other ranges of less importance, lie nearly east and west, the coast line makes a corresponding change in direction, but towards the east the coast cuts diagonally across the folded belt. In the districts between Ceres

INTRODUCTION 19

and Bredasdorp there is an intermingUng of the east and north trending folds, forming an area where the forces that produced these folds have given rise to a clearly marked diagonal set with a north-easterly course ; the chief ranges due to these north-easterly folds are the great mass extending somewhat irregularly from Cape Hangkhp to the mountains south of Worcester, the Hex Biver Mountains, and the south-west continua- tion of the Babylon's Tower range south of Caledon. The mountain ranges with a north-east trend are tra- versed by a weaker system of north-west folds, and are thereby broken up to a certain extent, especially by the synclines or trough-like folds of Houwhoek and Villiers- dorp. The intricate effects of the contest between the two sets of forces, that which produced the Gederberg (north and south) system of folds, and that which pro- duced the Zwartberg (east and west) system, so far as the Caledon and Bredasdorp districts are concerned, have been described in some detail in a survey publica- tion.^

There is some evidence in favour of the view that the Cederberg system of folds began to be formed rather earlier than the Zwartberg, but probably each reached its greatest development at about the same period, at some time between the deposition of the Ecca and that of the Uitenhage series.

The folding is most intense in the east and west trending portion of the rocks involved. Northwards from the country between Tulbagh and Karroo Poort

» Oeol Comm,, 1898, p. 42, etc. 2*

i^H||^f|||i||vtt^8gk:ock8 have been

JUJlfl^'H^diSlM'aB than to the

I^Mi^l tt W w t^^i^^^'^iii'bood. TKCaifSt'Baaif the Cederberg ••^iWlStSl'W^ flatten out ||!)&tJ!?W!?«4ajHhat on the lati- N^ik^gliSS^V's Dorp village

.ffJE' ■'WW ••''^ W^^ ^ocks which

isturbed farther

At the same

elonging to the

gradually thin

"^' H '^^ItV^ ^^^^ ^^^ ^^^ 'S*^*S*'^*^'jP^^°^ fornoation, the

S • JSsi'i^fili'' '»'""' »° ""■

'^^i*- • S^^«**^^^2'^ ^°^®'^ members 'l^^t^cfr^'^^^em, and finally ^^•£-«Sn*|^{^9||f*^e-Gape age, aa it r%;''S*i^vgl^O%^/£'Acds from Karroo mjja.]i)ca£C'|g^>S^U see later that ^i^ii^iEt:jj|s:4^;^iijgigression, or un- - '^^I&^l"''™*^''^*''"'?' *^ '^^ funda- ^ :f^4W*^^^^*^^tl^'°^ '" enabling us ■::f:jfe2g;p^^|^|^f the geological

^^•P*C§hS= Colony, but at ^^B^j££^^*f|p sufficient to say t^K^i^f^uiilii^chief cause of the tjlp^t^^k^he Cape system 3:^F^ii^«;|Efore and during

WTRODtJCTioN 21

It has been stated that the folded belt disappears under the sea near the Gualana Eiver, and it would be interesting to find out what becomes of it farther east. It is, of course, impossible to discover the exact state of affairs, but a comparison of the structure of the seaboard of Pondoland with that of the Van Ehyn's Dorp end of the folded belt will give us a clue to it.

In Pondoland some of the rocks which form the folded belt in the south of the Colony reappear on the coast near the St. John's Eiver, but are very different in certain respects from their condition west of the Gualana Eiver. They are found to be very slightly folded ; the great anticlines of the south and west have no counterpart there, and the greater part of the Cape formation is altogether absent. The rocks emerge from the ocean with a northerly trend, instead of the east and west strike which they have in the south. At St. John's there is a great block of Table Mountain sand- stone, surrounded on all sides by beds belonging to the Karroo formation faulted down against it, but further north-east towards Natal the Dwyka rests unconform- ably upon the Table Mountain series (see Fig. 4) ; the accounts ^ of the geology of Natal show that the same condition obtains there, and also that the Table Moun- tain sandstone (Palaeozoic sandstone of Anderson) be- comes thinner as it is followed northwards, and finally disappears, so that the Dwyka series rests directly upon rocks of Pre-Cape age. The relation of the Dwyka conglomerate to the Table Mountain sandstone in

^ Griesbach (71), p. 59 and map ; Anderson (01).

GfeOtX)GV 0** CAt>E COLONY

Pondoland is thus just like that of the same two series in the Bokkeveld Mountain north-east of Van Khyn's Dorp.

If we imagine the country between Karroo Poort and the latitude of Van Ehyn*s Dorp to be removed from observation, we have a nearly similar condition of things on each side of the folded belt, extending from Karroo Poort to the Gualana Kiver, but the relatively raised block of the Gates of St. John's has no analogue in the west. The gradual flattening out of the folds north- wards of Karroo Poort has no obvious counterpart in the east of the Colony, simply because the area in which a similar change takes place is under the sea. There is no reasonable doubt that on the sea floor between the Gualana Eiver and St. John's, first the Witteberg and then the Bokkeveld beds disappear, owing to Pre-Dwyka denudation, and that the Dwyka series rests upon lower and lower members of the Cape system, so that in Pondoland it lies directly upon the Table Mountain series, just as it does north of the latitude of Van Khyn's Dorp. It is very probable that, as in the west, the folds become less marked and practically die out altogether in the same area that shows the thinning out of the Cape system, so in the east, the two changes go on together. The comparison of the structure of the northward termination of the folded belt in the west and east of South Africa shows that this end of the continent is built upon a more symmetrical plan than might have been suspected from a mere inspection of the geological map.

The folded belt includes the more thickly populated

INTRODUCTION 23

districts of the Colony outside the Native Territories. Nearly all the various kinds of farming practised in South Africa can be found within this region. The most fertile and valuable land is that situated along the larger rivers flowing through from the Karroo, enriched by the silt brought down by them. The poorest soil is found on the sandstone mountains and near the coast, where the natural vegetation is of the kind known as "sour veld *'. In a region so diversified in climate and rocks as the folded belt, there are naturally many varieties of soil, and we shall have an opportunity of noticing some of these in later chapters.

(3) The limit between the folded belt and the third division of the Colony, the region of the plains and plateaux of the interior, cannot be precisely defined, as the folds die out gradually as one traverses them towards the interior ; the rocks become practically flat at a distance of some twenty miles from the great anti- cline of the Zwartebergen on the south of the Karroo ; on the west of the Karroo the distance between the Cederberg anticline and the nearly flat beds to the east is much less. Near the Nieuweveld and Eoggeveld escarpments there are several small flexures, usually more or less parallel to the axes of the Zwartberg folds, but they have slight effect on the surface features, and do not detract from the plateau character of the country they traverse.

The wide plains of the Great Karroo, and the even more extensive plateaux of the country north of it (often called the Upper Karroo), with sharply defined

24 GEOLOGY OP CAPE COLONY

steep-sided hill^ standing on them, are amongst the chief characteristics of the third region. Eastward of the Great Karroo, approximately bounded by a line drawn between Aberdeen and Jansenville, the structure of the country is essentially the same as that of the western part of the region, but owing to a more general distribution of rain, due to the absence of the coastal ranges which prevent the moisture-laden south-east winds carrying rain to the interior in the west, the eastern portion of the region is better covered with vegetation than the western; the thicker covering of vegetation in the east, which becomes more marked as one approaches the coast, softens the features of the surface, the hill slopes are more rounded and less abrupt, and the distinction between harder and softer rocks is less obvious than in the Karroo.

The sedimentary rocks of the third region lie nearly horizontally, but a careful examination shows that they usually dip at a very low angle towards the central part of the basin. Thus in the western Karroo and Eogge- veld the beds dip east, to the north of the main water- shed the dip is usually south or a little east of south, and to the south of it the beds are inclined slightly to the north or west of north. These sedimentary rocks belong exclusively to the Karroo system, but with them are found intrusive igneous rocks, dykes, sheets and great masses, probably lenticular in shape, of dolerite. The dolerite intrusions are of sufficient importance to have a chapter devoted to them, and at present only the chief facts relating to their distribution will be mentioned. From the western border of Calvinia east-

INTRODUCTION 25

wards to the Indian Ocean, and from the Nieuweveld escarpment northwards to the Orange River, and even far beyond the river, the sedimentary rocks are traversed by sheets and other masses of dolerite to such an extent that it is hardly an exaggeration to say that within the area of some 70,000 square miles one cannot get out of sight of the dolerite hills. This area is but a part, perhaps not much more than half, of the whole range of the dolerite intrusions in South Africa. Though the dolerite is so widely distributed, and varies somewhat in composition and structure, it has an individuality of its own, and can be distinguished from similar rocks in the Colony belonging to earlier periods of igneous activity. It is a remarkable fact that intrusions of this dolerite are extremely rare in the folded belt, and also in those parts of the Karroo basin on the margin of that region. In the Bokkeveld Mountain west of Calvinia, where the Table Mountain series lies almost horizontally, and in the Brandewyn valley (Clanwilliam), where the Cape formation is but slightly folded, dykes of dolerite of the Karroo type occur. In Pondoland also, where the Table Mountain sandstone lies nearly undisturbed, the dolerite has invc^ded it. Throughout the folded belt south of the Karroo not a single intrusion of this nature has been found.

The dolerite intrusions have a very important effect on the surface features of the country, owing to their being less easily weathered than most of the sedimentary rocks associated with them. The steep escarpments of the Nieuweveld and Roggeveld owe their abrupt faces to this rock, for the more easily weathered sedimentary

26 GEOLOGY OP CAPE COLONY

rocks form steep slopes at the bottom of vertical cliffs of the dolerite or sedimentary rocks hardened by contact with it. In the Komsberg, which lies between the Roggeveld and Nieuweveld, there is no dolerite, and although a somewhat similar rdle to that of the dolerite is played by some hard bands of coarse sandstone, the escarpment is less precipitous than either of the other escarpments.

The well-known table-shaped hills scattered broadcast over the interior of the Colony owe their form to a pro- tecting cap of hard rock, either sheets of dolerite or beds of sandstone ; the finest examples of such hills are found amongst those capped by dolerite. Tafel Berg and Spitzkop, two outliers of the Western Nieuweveld, which rise some 3,000 feet above the Gouph and are visible from the railway between Prince Albert Road and Beaufort West, are magnificent hills of this type, and are capped by a dolerite sheet 400 feet thick. Other instances, of smaller size but quite as striking, are the hills called Theebus and Kafifeebus, near the railway between Steynsburg and Eosmead.

A great part of this region is covered with small bushes, but the eastern portion is a grass country. The Great Karroo, Eoggeveld and Nieuweveld are chiefly sheep veld, but the flat land along the rivers is extremely fertile when brought under irrigation. To the north of the main watershed very large areas of alluvial deposits along the rivers,. such as the Zak and Rhenoster, await cultivation. Owing to the cold winter climate of the higher parts of the Eoggeveld and Nieu- weveld the farmers there have to take their flocks to

>*i'^'§.A>i'

5 5 ir«il|i If lultU

I" I I

s.- I I

i* laii-- I; s

2 S €«3sa5ca

w TB. TT Ku «■ •- Ti. Bokkeveld ■S'^^l'^^ 'H'^^'^p^el^y inhabited

|irtfc'|1|lil

tUf^^V^'^Ae three regions

'"^ — - - figure

dotted line

Kooast indicates «.-.-. .-p. -.«.°° °^ ^^^ ""*-

N B< C3 ^nSiBding the small

t'& '^. ■9tif''M ^ H h^oo f ormatioD

iountain sand- d. ii^y^i^S^il^ in the figure

J^fi^^it^^^^Zwatthetg and •^•(t^i^Sii^<^itff^, and the vary- •^*ij|^uC^H^^2he lines corre- *i^^l^^€8^!^K:£I^^^rying intensity (^UsIS^S^r as our infer-

1^n|Si ilSiamined in the

•^g£3{^i||t^early the dififer-

l^ifSpt^^^^of the northern

■S-^§*'$^S-^'H*^I^]^^^^ of the basin

~ =&>l8*»sl^^Mtrtion lies. The "'^8iiiiffifii3''fe P*"^ *° original

=^io'H"il«-S*^ 'yi'ig farther

INTRODUCTION

29

■4S

O

O

o

CD

o

■e

CD

s

GO

(U 0^ ®

••>4 > O 4^

south at the present time than it did at the period of the greatest northward ^ extension of the water in which the Karroo forma- tion was deposited ; but, as we shall see when we de- scribe in detail the lower part of the Karroo system, the present position of the northern edge, although due immediately to the pro- gress of denudation, must lie approximately along a former course of the Karroo shore at a certain period of its existence. The southern edge of the basin, on the other hand, is entirely due to the exposure of the Pre- Karroo rocks by denudation in a folded area. That the Karroo rocks formerly ex- tended far to the south of the Karroo is proved by the occurrence of outliers of the two lowest series in the district between Worcester and Ashton, where they are faulted down against the Malmesbury beds on the north, but lie conformably ^

<M

iH

O

a J

OQ

09

S3 u

OQ 0)

2 o

•v»

3 O

u

d o

o «

GO

u

0)

533

CO

_ *

a a>

O f—* S rt Q

,-4 c4 CO •^' »ft

30 GEOLOGY OF CAPE COLONY

upon the Witteberg series along their southern boundary. The discovery of the true character ^ of the Worcester outlier (see Fig. 5) is perhaps the most important addi- tion to our knowledge of the structure of the Colony made during recent years, for it greatly strengthened the evidence for the conformity of the Dwyka series with the uppermost series of the Cape system, and at the same time afforded a clue to the structure of the Langebergen, which has been found to solve many of the difficulties met with in that range and also in the Zwarte- bergen. There is no direct evidence as to the position of the southern limit of the area in which the Karroo formation was deposited.

The great crumpling of the earth's crust in the south of the Colony was so violent at many places that the rocks are inverted and the older lie above the newer; that is particularly noticeable along both the Zwarte- bergen and Langebergen, and is illustrated in Figs. 1 and 6. The overfolding seems usually to be towards the north, in other words, the folds are bent over north- wards, so that the dip of the strata is towards the south. The country whose southern termination is the third region in our description seems to have served as an immovable block against which the rocks were crumpled on the south, and south-west, and possibly south-east sides. These great movements of the crust, more im- portant to the present structure of the Colony than any others that have affected the southern end of the con- tinent, seem to have been limited to that region. There

J B. H. L. Schwarz, Geol. Comm. for 1896, pp. 27-28.

INTKODUCTION 31

appear to have been no great movements of the same age in the country lying north of the Cape Colony ; the disturbances met with in the rocks which Dr. Molen- graaff calls the Cape formation ^ in the Transvaal are clearly older, for they do not aflfect the Dwyka and Ecca beds. A similar reason must be given for regarding the plications of the Pre-Cape rocks of Prieska and Griqua- land West as of greater age than those belonging to the Zwartberg and Cederberg systems of folding.

There are several other structural features of import- ance which will be better understood by the reader after a closer acquaintance with the character and dis- tribution of the various formations has been made, and they will be especially referred to in the chapter dealing with the history of the development of the Colony.

^I.e.f the Bla.ck Reef, Dolomites, Pretoria beds and Waterberg sandstones.

[Since this was written Dr. MolengraafE has named this group the Transvaal formation to distinguish it from the later Gape system.]

CHAPTER II.

THE PRE-CAPE ROCKS OF THE SOUTH AND WEST OF

THE COLONY.

The various groups of rocks included under this heading have one character in common, they are older than the Cape formation. In the case of three of the groups, Ibiquas, Cango and Malmesbury, their Pre-Cape age is obvious from the fact that they are found directly beneath the Table Mountain series ; but in the case of the northern groups, which are found in a region where the Cape formation was either not deposited or has since been removed by denudation, their age has to be arrived at by reasonings based upon the structural features of the country, for no help in correlating these formations is given by fossils.

The Malmesbury Series.

In the south-western districts sedimentary rocks are in many places met with immediately below the Table Mountain series. These rocks were evidently intensely disturbed, invaded by granite and other igneous rocks, and long exposed to denudation before the deposition of the Table Mountain sandstone. In the immediate neighbourhood of Cape Town the Table Mountain

32

PRE-CAPE ROCKS OF THE SOUTH AND WEST 33

sandstone, which forms all the higher parts of the Peninsula, lies nearly horizontally, and below it are seen slaty rocks dipping at very high angles, with a general north-north-west strike, accompanied by a large intrusion of granite. The slaty rocks are found to extend northwards from the foot of Table Mountain at least as far as Van Ehyn's Dorp, occupying the greater part of the divisions of Malmesbury, Piquet- berg, Paarl, Stellenbosch and Somerset West. This large area of Malmesbury beds is separated by the range traversed by Bain's Kloof, and called the Limiet Berg, Eland's Kloof and Vogel Valley Mountains in different parts of its length, from a rather narrow strip of similar rocks occupying the long depression between Winter Hoek, north of Tulbagh, and Worcester; near the latter town the strip of Malmesbury beds becomes thinner, and extends south-eastwards as far as Swellen- dam as a narrow band overlain to the north or north- east by the Table Mountain series, but cut off on the south or south-west by a fault (the Worcester fault) which has a down-throw of some 10,000 feet near the town of Worcester (see Fig. 5). Inliers of similar rocks have been found at French Hoek, Eland's Kloof (near Villiersdorp), in the Zondag's Kloof east of Stan- ford (Caledon division), and between Elim and Bredas- dorp. Each of these inliers is surrounded by the sandstones of the Table Mountain series. Rocks that can best be placed with the Malmesbury beds occur also in Mossel Bay, George and Port Elizabeth.

The most abundant rock in the series is a sandy clay-slate with imperfectly developed cleavage. Small

34 GEOLOGY OF CAPE COLONY

flakes of white or yellowish mica are frequently sufl&- ciently abundant to give the rock a micaceous appear- ance when broken along the cleavage planes. This mica is easily distinguished from the reddish-brown mica so strongly developed in the clay-slate taken from the immediate neighbourhood of the intrusive masses of granite, and generally only visible under the micro- scope. In certain localities, such as the hills north-east of Moorreesburg and the Tygerberg group, the propor- tion of quartz grains increases so greatly that the rocks may be called impure quartzites, and in other places fairly pure quartzites occur, but they are not often met with. Crystalline limestone or marble forms thick bands in the Malmesbury series near Van Ehyn's Dorp, Piquetberg, Vogel Valley (south of Porterville Eoad Station), at Bakoven's Hoogte between Ashton and Swellendam, in Dassies Hoek near Eobertson, and in small quantity north of Worcester. Many other varieties of rock are met with near the contact with the granite, but these will be mentioned later. Ottrelite- or chlori- toid-schists are found in rather thin bands near the junction of the slates, which have evidently been in- tensely compressed, with the unconformably overlying Table Mountain series in Waai Kloof, near Worcester (Plate I.), and north of the village of Swellendam. In both cases thick quartz-schists occur on one side of the ottrelite-schist, but no granite or other intrusive rock is found in the immediate neighbourhood, and as ottrelite- schist has not been seen near any of the granite areas in the Colony, these are probably two further examples of the production of ottrelite by pressure metamorphism

il

II

11

^1

36 GEOLOGY OF CAPE COLONY

without the concurrence of the influence of igneous rocks.^

Conglomerates are ip,rely met with in the Malmesbury beds. Some conglomerates with quartz pebbles have been described from the neighbourhood of Saron and Honig Berg in the Tulbagh and Piquetberg divisions,^ but it is uncertain whether they really belong to this series. Mr. Schwarz says of the Honig Berg outcrop : ** There are conglomerates between the Table Moun- tain sandstone and the slates (Malmesbury beds), apparently conformable to the former and unconform- able to the latter, but the exposure is too small to say whether these relations hold good in reality '*. He remarks also that the conglomerates resemble those of Oudtshoorn; it is not unlikely that these beds will eventually prove to belong to the Cango series.

The true succession within the Malmesbury series has not been made out. They are nearly always found dipping at very high angles, and as they cover a large area, in places over thirty miles wide across the strike, it is certain that they must be intensely folded, and therefore repeated by folding, so that a much smaller thickness of rock is present than would seem to be the case. The country occupied by these beds is rather flat and has a regular rainfall, and the ground is well covered with soil and vegetation ; in consequence out- crops are not very abundant, and years of detailed work will probably be required before the true structure of

^ Examples of such an occurrence of otfcrelite-schist in the Transvaal are given by Gotz (85), p. 158.

*E. H. L. Schwarz, Geol. Comm. for 1898, pp. 27, 28.

ME-Cape rocks of the south and West S7

the Pre-Cape rocks between the Peninsula and Piquet- berg can be ascertained.

Veins of quartz are abundant in the Malmesbury beds, and at places they have been prospected for gold, with- out gratifying results.

The general strike of the rocks classed in this series is to the west of north in the western part of the Colony, approximately parallel to the trend of the Cederbergen and the other ranges in the west, which were formed chiefly after the deposition of the Ecca series ; but in the south, between Worcester and Swellendam, in Bredasdorp, Mossel Bay, and George, the strike of the Malmesbury beds is on the whole nearly east and west, roughly parallel to the great southern mountain ranges. This change of strike in the Malmesbury beds may per- haps to a very small extent be due to the forces which produced the folds in the overlying rocks ; but as the dip of the lower beds is generally far higher than the dips observed in the unconformably overlying rocks, it is impossible to thus account fully for the change in the direction of strike of the Malmesbury beds as they are followed eastwards. It is certain that these rocks were folded almost as much as we now see them before the deposition of the Cape formation, and the general parallelism between the two systems of folds, older and younger than the Cape formation, points to the repeti- tion of the folding along the same lines at a great interval of time.

The Malmesbury beds have been invaded by igneous rocks of both acid and basic compositions. The acid

38 GEOLOGY OF CAPE COLONY

series, granite, gneiss, and allied rocks, is by far the more important. The masses of granite and gneiss are elongated in form, and lie with their longer axes parallel to the strike of the sedimentary rocks. They form the highest ground in the Pre-Cape area, with the exception of the outliers of Table Mountain sandstone. The largest granite area is that which stretches from St. Helena Bay south- south-east to Klein Dassen Berg, a distance of seventy miles, and the highest points reached by the granite are Kapoc Berg and Contre Berg, both over 1,500 feet above the sea. Saldanha Bay is a deep inlet in this mass of granite. On the western edge of the granite, along the shore near Paternoster, Danger, and Saldanha Bays, large inclusions of slate are frequently seen in the igneous rock, indicating the proximity of the Malmesbury beds ; the edge of the intrusion is probably not far to the west of the present coast line.

Many varieties of granitic rock are found in this great area. The most abundant perhaps is a two mica (ie, with both black and white mica) granite with orthoclase as the chief felspar. Tourmaline is often present in the rock near Darhng. Every gradation between a normal granite and a gneiss, in which the foliation structure can be seen in even a small fragment, can be found ; the massive granite is seen in the interior of the area and the foliated rock near the periphery, but this rule is not without many exceptions. There is no general difference in mineralogical composition between the granite and gneiss; the structural characters which separate the gneiss from the granite seem to have been given to the rock during its consolidation, for the gneiss

PRE-CAPE ROCKS OF THE SOUTH AND WEST 39

does not show evidence of a great amount of crushing or rearrangement of its component minerals after it solidified. The foliation planes lie in the same direction as the strike and cleavage of the sedimentary rocks in the neighbourhood; a similar direction is at places observed in the arrangement of the large porphyritic crystals of orthoclase that are occasionally found in great numbers in the massive granite, which shows no other parallel structure. There is no evidence of a difference in age between the granite and gneiss, and the gradual coming in of the gneissose structure as the area is traversed in various directions points to the whole mass being the product of one period of igneous activity.

Large and small veins or dyke-like bodies of micro- granite and quartz-porphyry with a micro-granitic base are found towards the edge of the area in many places. Near Hoetjes Bay the quartz-porphyries are especially abundant. Near Darling a mass of quartz-porphyry has a well-developed parallel structure, and may be considered to bear the same relation to the massive quartz-porphyry as the gneiss does to the granite.

In the hills to the south and west of Darling there are some remarkable rocks associated with the granite and gneiss. Colourless augite, plagioclase, and sphene are added to the usual constituents of the granite, and the mica is practically absent; the structure is that known as granulitic, the various minerals occurring in grains of a more uniform size than is the case with granite. These rocks often show a parallel structure but have not the foliated or schistose planes seen in the

40 GEOLOGY OF CAPE COLONY

gneiss. The nature and origin of the granulites of Darling are as yet unexplained, as is also their exact relationship to the surrounding granite and gneiss.

A few miles east of the southern end of the great mass of granitic rock just described is the irregularly shaped area of granite on which the town of Malmesbury is built. At the south end of this mass is the rugged mountain called Paarde Berg. The granite area is about twenty miles long and six wide, and lies in the direction of strike of the Malmesbury beds. The rock is much less varied in this area than in the larger mass to the west, and is mainly a rather coarse biotite-granite with porphyritic orthoclase, but fine grained granite composed of the same minerals, and coarse pegmatites are not infrequent. There seems to be no gneiss in this area.

South-east of Paarde Berg is the Paarl Mountain with the well-known group of smooth, naked granite crags on the summit. The most abundant rock in the Paarl Mountain is a biotite-granite. Dykes of quartz- porphyry in continuity with the main mass of granite traverse the surrounding slates along their strike. No gneiss has been observed in this mass.

On the east side of the Berg Eiver between Welling- ton and Paarl is a long, narrow area of granite overlain by the sandstones (Table Mountain series) of the Klein Drakensteins. Both this granite and the Paarl Moun- tain rock have a more northerly direction than the other intrusions, and a corresponding change of strike is noticed in the Malmesbury beds of the neighbourhood.

South of the Paarl and Drakenstein granite areas is

PKE-CAPE ROCKS OF THE SOUTH AND WEST 41

the somewhat irregularly shaped mass of Pniel and Stellenbosch, with which are nearly connected those of French Hoek on the east, and of the Bottelary and Helderberg to the west. Gneiss enters largely into the constitution of those bodies of granitic rock, and, as in the case of the great intrusion on the Saldanha Bay coast, there is no evidence here that the intrusion of the foliated rock was of later or earlier date than the massive granite. In places, such as certain parts of the mountain slopes on the left bank of the Jonker's Hoek stream, the gneiss has been crushed along planes parallel with the direction of the dominant structural lines in the neigbourhood, the cleavage and strike of the slates, and the foliation planes of the gneiss ; the crushing occasionally resulted in the production of a rock more like a gritty schist than a gneiss, but this extreme stage is connected with the uncrushed rock through breccias of different degrees of coarseness. The breccias were evidently formed in their present position by the breaking up of the gneiss, so that large and small subangular fragments of gneiss, and of its larger component minerals, are embedded in a fine- grained matrix. The fine-grained schistose rock is a true mylonite.^

The granites of the Paarl and Stellenbosch districts contain a fair amount of microcline, a variety of felspar which is rare in the Saldanha Bay and Darling area.

^Mylonite is the name giving by. Professor Lapworth to crushed rocks with a parallel structure, in which all traces of the original structure of the parent rock may have disappeared (Lapworth, Intro- ductory Text-hook of Geology y p. 107).

42 GEOLOGY OF CAPE COLONY

although it seems to be the chief felspathic constituent of the granites in the northern and north-western parts of the Colony. On the south-west edge of the Bottelary granite cassiterite or tin-stone occurs in a gneissose muscovite granite together with tourmaline ; wolframite has been found in the same neighbourhood.

Near Somerset West there are two masses of granite ; the smaller one, Schaapen Berg, just east of the village, contains some interesting varieties of rock. The main mass of the intrusion is a biotite-granite with little mus- covite, but the muscovite is very abundant in certain places and the felspar decreases in amount, and may disappear completely, so that the rock becomes a greisen, or quartz-muscovite rock. In other parts tourmaline is extremely abundant, sometimes giving rise to schorl rock, composed of tourmaline and quartz only. At other places andalusite, showing a beautiful pink tint under the microscope, forms a large part of a rock com- posed of quartz, tourmaline, muscovite, andalusite, and apatite.

The granite underlying a great part of the sandstone of Table Mountain and the other mountains of the Peninsula has been described by many previous writers. Professor E. Cohen ^ of Greifswald has described in de- tail the granite and the altered clay-slate near it, from the immediate neighbourhood of Cape Town ; he was the first to record pinite, an alteration product of cor- dierite, in the biotite-granite there.

The contact of the granite and clay-slate at Sea

1 Cohen (74).

PRE-CAPE ROCKS OF THE SOUTH AND WEST 43

Point and in the Platte Klip ravine have long attracted considerable attention. Playfair,^ the enthusiastic dis- ciple of James Hutton,^ edited a description of the two localities written by Basil Hall in 1813. Playfair drew fresh support for Hutton's theory of the relationship of granite to the surrounding sedimentary rocks from Hairs letters and sketches. Clarke Abel ^ a few years later wrote a very accurate account of the same spots, and his conclusions are sounder than those of Hall, who regarded the elevation of the sandstones of the Paainsula as due to the rising up of the granite.

On the beach at Sea Point the junction of the two rocks is an extremely interesting one. The slates have been thoroughly permeated by the fluid granite, and have a shredded structure with granite lying between the slightly bent shreds of slate. Large orthoclase crystals, in every way similar to those in the porphyritic granite, have been formed in the lenticular areas be- tween the laminae of slate.

Small areas of granite intrusive in the Malmesbury beds are known in the south of Caledon, in the Hemel en Aarde and Zondag's Kloof valleys, and again in the western part of Bredasdorp.

1 Playfair (13).

^ Hutton, the leader of the old school of Vulcanists who insisted on the igneous origin of such rocks as granite and basalt, in opposition to the Neptunists, headed by Werner, who regarded these rocks as pre- cipitates from the primeeval ocean, rendered an even greater service to Geology by searching for explanations of geological phenomena in the everyday events on shore and land. His teachings in this respect had evidently been somewhat lost on Dr. Abel, who remarks that the Lion's Head must have been violently torn from Table Mountain.

'Clarke Abel (18).

44 GEOLOGY OF CAPE COLOnY

In the narrow strip of Malmesbury beds north of the Worcester fault there are at least three granitic intrusions, all of which have been considerably affected by earth movements since their intrusion, and to some extent probably by movements during their consolida- tion. There is an abundance of phyllite-gneiss, a rock looking very like a highly micaceous clay-slate with "eyes *' and thin strings of obviously igneous material, composed of quartz, orthoclase and mica. The ortho- clase crystals often form the "eyes" with little other granite material in the same lenticular area. The largest mass of granite forms the high ridge just west of Robertson.

The last granite area in the south of the Colony that must be mentioned is that of George, a mass very variable in composition, at least thirty miles long from east to west, and from four to eight miles wide. It contains both muscovite- and biotite-granites with tourmaline and fluor; gneissose rocks also occur in the district.

The granite has in every case produced considerable mineralogical changes in the surrounding rocks. The result varies considerably in amount and nature, de- pending chiefly upon the character of the rock invaded. Highly quartzitic rocks are the least affected, and the alteration seems to increase with the clay content of the original slate. Up to the present time no metamor- phosed calcareous rocks (except the marbles or crystal- lised limestones) have been noticed in the southern part of the Colony. The clay-slates become highly micaceous near the granite ; sometimes, as in several places east of

PRE-CAPE ROCKS OF THE SOUTH AND WEST 45

the Darling granite and in the George district, they be- come typical mica-schists, rocks which glisten owing to the innumerable flakes of pale mica arranged parallel to one another, the other important constituent is quartz. At Zwart River Bridge, in George, a magni- ficent section of chiastolite-schist, a rock composed of chiastolite, mica, and quartz, can be seen ; the crystals of chiastolite are often over two inches long. The chiastolite-schist is found within a few yards of a remarkably coarse two-mica granite, which also con- tains tourmaline.

Near the Cape Town, the Paarl, Stellenbosch and Somerset West granites the clay-slates become spotted at about 300 yards from the contact; and the spots are found in thin sections of the rock to be clear areas amidst the general mass rendered brownish in colour by the development of minute flakes of red-brown mica. The clear spots are composed of very minute crystalline grains of a mineral which has not been determined. Minute grains of felspar, recognisable by their twinning, have been developed in the spotted rocks, but they are not abundant. -

At several places in the south-western districts igne- ous rocks of more basic composition than granite occur as dykes in the Malmesbury beds and in the granite. The dolerite dykes near Cape Town have been described in detail by Cohen ^ ; they consist of augite, plagioclase and magnetite. These rocks differ in some respects from the average type of dolerite met with in the great

^ Cohen (74|, p^ 1.Q, etc., in the separate copies.

46 GEOLOGY OF CAPE COLONY

central basin of the Colony, the third of the three regions into which we divided the country. Although some of the Karroo dolerites contain no olivine, that mineral is very often present in them, and the rocks generally have an ophitic structure. In the dolerites of the Peninsula and Somerset West there is no olivine, but the felspar is rarely enclosed by the augite, in other words, they are seldom ophitic in structure. These differ- ences are rather slight, especially when it is remembered that they are based upon a comparison between about a dozen representatives of the southern dykes and over a hundred of the Karroo dolerites, taken from an im- mense area. The analyses published by Cohen of one of the Cape Town dykes and of thirteen of the Karroo rocks, show that the former is very similar in compo- sition to the latter. The Karroo dolerite is generally less altered than the southern dykes.

The differences between the two sets of dolerites is so slight, in fact, that they might well be considered to belong to one and the same group of intrusions. The age of the southern dykes is certainly younger than that of the granite and Malmesbury beds ; as they have not been observed traversing the Table Mountain series they are generally looked upon as older than that rock, but it is possible, on the supposition that they belong to the same series as the Karroo dolerites, that they were not able to break through the horizon- tally overlying sandstones after reaching the limit of the granite or slate. The junction of a dolerite dyke and the sandstone has not yet been clearly seen, nor have pebbles of dolerite been found in the sandstone, so

PRE-CAPE ROCKS OF THE SOUTH AND WEST 47

the question of the relative ages of the two rocks is still an open one.^

Some interesting rocks, which may be called dio- rites and quartz-diorites, form rather limited dyke-like masses in the granites of the Malmesbury district. At Klein Paarde Berg there is a broad dyke about a mile long, composed of hornblende, felspar, mica, quartz, magnetite, apatite and zircon. It is a holocrystalline rock, and the hornblende often encloses the felspar crystals, so as to give the rock a partly ophitic struc- ture ; some large crystals of mica (biotite) behave in the same way. Most of the felspar belongs to the oligoclase series of the plagioclases, but there are patches of a very much altered felspar, strongly con- trasted to the clear crystals of plagioclase, which are very probably orthoclase. Quartz is present in consid- erable quantity, filling up the spaces between the other minerals. The rock is little altered as a whole, but some of the mica is replaced by chlorite, and some epidote, derived from the alteration of other consti- tuents, is present. Another variety of diorite in this neighbourhood contains the same minerals as the one just described, but monoclinic pyroxene, with the char- acteristic diallage structure, is present in considerable quantity, forming in thin sections ophitic plates enclos- ing felspar. The pyroxene sometimes forms complicated

^ Since this was written it has come to the notice of the writer that Mr. T. Stewart, M.I.G.E., exhibited a piece of dolerite from a dyke in the sandstone of Table Mountain at a meeting of the S. A. Phil. Soc. in 1895. Lately two such dykes have been mapped by the survey in the T. M. S. of the Peninsula ; compare with the dykes in the same jrock of the Bokkeveld Mountain and Pondoland.

48 GEOLOGY. OF CAPE COLONY

intergrowths with the hornblende and also occurs in the centre of large hornblende crystals ; in such cases one set of prism cleavages is common to both minerals.

In the gneiss of Klein Dassen Berg there is a dioritic dyke intruded parallel with the foliation planes of the gneiss. The rock of this dyke is rather different from the Klein Paarde Berg rock, in that the constituent minerals, plagioclase, hornblende, and quartz form nearly equal -sized grains, and none of them have any proper crystal faces ; the structure is typically granulitic. At Yzer Fontein Point is a large mass of hornblendic rock, coarsely crystalline, with a banded structure ; some thick layers are formed entirely of green hornblende, and others, usually thinner, have a fair proportion of plagioclase in them. These dioritic rocks seem to be confined to the Malmesbury district.

In the George granite there are some dykes of horn- blende - schist, composed of long and rather fibrous crystals of green hornblende, arranged parallel to one another, with a smaller quantity of quartz and plagio- clase grains between them, and a still smaller amount of epidote. This rock is evidently a highly altered basic dyke, but there is as yet little evidence of its original nature.

The Cango Series.

In the Cango district, the country near the northern boundary of Oudtshoorn on the southern flank of the Zwartebergen, there is a group of sedimentary rocks older than the Table Mountain sandstone, and therefore usually classed with the Malmesbury beds. There are, however, so many peculiarities in the Cango rocks

PRE-CAPE ROOKS OF THE SOUTH AND WEST 49

which separate them from the bulk of the Pre-Cape rocks of the Malmesbury and other divisions in the south-west of the Colony, that it is advisable to dis- tinguish them by some other name ; the term Cango conglomerate^ has already been used for a prominent band of rock in the series, and it will be convenient to call the whole group the Cango series.

The series forms a lenticular area about seventy miles in length from east to west, from near Amalienstein (Ladismith) to some few miles east of Meiring's Poort, and at the most about nine miles wide. The Table Mountain series bounds the area on the north, and the southern limit is formed by the conglomerates of the Uitenhage series between Meiring's Poort and Calitzdorp, a distance of fifty miles ; west of Calitzdorp the sand- stones of the Table Mountain series overlie the Cango beds along their southern limit, and farther west again the sandstone is faulted down against them, the fault being so formed that its throw increases and brings the Bokkeveld beds into contact with the Cango; some miles east of Meiring's Poort, also, the Bokkeveld beds are faulted down against the Pre-Cape rocks, and there can be no doubt that this fault, exactly comparable to the Worcester fault, is continued westwards under the covering of Uitenhage beds at least as far as Calitzdorp, and is probably continuous with that already mentioned west of the village (see Figs. 6 and 7).

Along almost the whole length of the northern boundary the Table Mountain series dips at a high angle

^ Oeol. Camm, (98), pp. 7, 68, etc.

4

■»• §';ii'§*a''t'S''^

ma:*'-

i<

PREOAPE ROCKS OF THE SOUTH AND WEST 51

southwards below the Cango beds, and the latter dip at approximately the same angle in a southerly direction. At the south end of the Gamka Poort, where there is one of the very few clean cut sections of the junction of the two formations, there appears to be a conformable passage between the two. At other spots, however, such as the south end of Meiring's Poort, the Table Mountain series dips steeply to the north, and lies unconformably upon the older beds which dip at a still higher angle to the south; the contact of different members of the Cango beds with the base of the Table Mountain series at various points corroborates the evi- dence of the Meiring*s Poort section, so there is no doubt that the junction is an unconformable one. It is very probable that the Table Mountain sandstone was deposited upon the then nearly horizontal Cango beds, which had suffered some denudation, so that the base of the former group rested upon different horizons of the latter series at different localities. During the great earth movements that produced the Zwartebergen the two series were together folded and inverted, so that at places the older beds appear to overlie the younger conformably.

The Cango beds usually have high southerly dips, but in the neighbourhood of Kruis Eiver, west of the road up the Zwartberg Pass, the strike is north-east. The top or bottom of a fold is occasionally seen; this in- dicates that the series is thrown into isoclinal folds, and that the observed great thickness of southerly dipping beds is really due to the repeated folding of a much

smaller thickness of rocks. The true succession of the

4*

52 GEOLOGY OF CAPE COLONY

members of the series is rather uncertain, and the bottom has not been found.

The series consists of conglomerates, quartz-felspar grits, quartzites, slates and limestones, in all a very considerable thickness of rock, not under 10,000 feet. These are accompanied by intrusive rocks of the nature of diabase or altered dolerite (see Fig. 7). The con- glomerates lie next to the Table Mountain sandstone in the western part of the area ; in the central portion the limestone lies in a similar position, elsewhere slates or quartzites are in contact with the sandstone. At the Gamka Poort thick bands of conglomerate are in contact with the Table Mountain series. There are several varieties of conglomerate in the Cango beds, differing chiefly in the nature of their contained pebbles and in the amount of shearing they have undergone. In the west, on the hills north of the Ladismith Boad near Vaartwell, the conglomerate has been sheared to such an extent that the original forms of the pebbles (slaty rocks and vein-quartz) are no longer recognisable, and in many cases the exact limit between pebble and matrix is indefinite. Farther east the conglomerates are more normal in character, but the efifects of shearing are still very evident. In Schoeman's Poort, where excellent sections through the conglomerate are exposed by the roadside, large pebbles or boulders of granite and diabase are seen in it. The occurrence of these is in- teresting, as it proves the Cango beds to be later in age than some rocks — possibly the Malmesbury beds — which were invaded by granite and diabase before they furnished sediments for the building up of the Cango

PRE-CAPE ROCKS OF THE SOUTH AND WEST 63

beds. So far as is known at present there is no uncon- formity at the base of the conglomerates of which there are at least two bands, and although in the Grobbelaar's Valley, and other places farther west, slates are seen on either side of the steeply inclined conglomerate, it is even difficult to decide which is the top and which the bottom of that rock. It may be that the bottom is nowhere seen, and the slates on either flank of the conglomerate overlie the latter.

A remarkable group of beds, formed chiefly of various sized fragments of quartz and felspar, extends for a considerable distance along the strike of the Cango series, half a mile north of the conglomerate between Grobbelaar's Eiver and Matje's Biver. The felspar occurs in fragments of such size and form that in places the rock has the appearance of a porphyritic granite. When examined under the microscope in thin sections the quartz and felspar are seen to be broken crystals, although the crystalline form of the quartz is occasionally seen. The felspar is mostly microcline, but albite is frequently, and orthoclase occasionally, met with. These minerals are enclosed in a ground mass chiefly composed of small grains of quartz and minute flakes of sericite, a pale micaceous mineral; small flakes of brown mica are sometimes found taking the place of the sericite. The mica forms a thin casing round the large grains of quartz and felspar, and the two latter minerals are often seen almost in contact with a very thin film of sericite be- tween them. The sericite occurs in this rock in the same manner as in many gneisses and conglomerates

54 GEOLOGY OF CAPE COLONY

that have been subjected to great pressures in the earth's crust. In some localities the rock shows a distinct schistosity, and in thin sections the large quartz fragments are seen to be elongated in the plane of schistosity, and have patches of interlocking grains of quartz at their two ends, as if the material had been removed from the sides of the fragments and deposited at the ends. The minute sericite flakes lie in one direc- tion, along the planes of schistosity. The quartz-felspar rock of the Cango is very like the so-called porphyroids, and appears to have been a sedimentary rock composed chiefly of fragments of quartz and felspar, in which the micaceous minerals have been developed by pres- sure. In places bedding planes are distinctly seen, and varieties intermediate between the porphjoroid and ordi- nary grits with few felspar fragments have been found between the main band of porphyroid and the southern slope of the Zwartebergen. In the valley from which the Cango caves are entered three beds of conglomerate, a quartz-felspar grit with rounded boulders and pebbles of granite, mica schist, quartzite, crystalline limestone, and vein quartz, are seen in the stream bed below the caves. The transitional varieties and the conglomerates certainly support the conclusion that the porphyroid of the Cango is a sedimentary rock, but whether it was formed by debris derived from a granitic region, or whether it is of the nature of a volcanic tuff is not clear; the abundance of microcline in the porphyroid and the absence of lavas from the district favour the former supposition.

There are many bands of limestone in the Cango

PEE-CAPE KOCKS OF THE SOUTH AND WEST 55

beds, sometimes of great thickness ; they are lenticular in form, but to what extent this is due to folding has not been determined. The chief limestone band is that which is in contact with the sandstones of the Zwarte- bergen near the south end of the Zwartberg Pass. It extends for some fifteen miles eastwards, and in it are the famous Cango Caves.^ The cave, at least that part known in 1897, is nearly 750 yards long, and is probably of still greater extent. The explored portion of this cave lies in a nearly straight line. There can be no doubt that the cave has been formed by the solution of the limestone, aided by the breaking away of the roof and sides and the removal of the debris by running water. The cave has not been sufficiently explored to explain its formation fully, and the level of the floor at various points is not known. The floor itself is at least partly made of debris cemented with calcareous tufa and stalagmite. The walls and roof of the cave, in those parts which have not been disfigured by the smoke of candles, are very beautiful, owing to the number, form, and brilliance of the stalactites attached to them. Other caves, the entrance to which is often on the face of cliffs along the sides of the valleys, await exploration in the Cango district. The band of lime- stone in which the great cave is situated is about 1,800 feet thick, but when traced to the east or west it gradu- ally becomes thinner. The limestone in the Cango beds is crystalline and dark grey in colour and usually con-

* For a description of the cave see G. S. C. Corstorphine, Ann. Rep, (96), p. 34 ; a plan of the cave by H, M, Luttman Johnson accompanies the description.

56 GEOLOGY OF CAPE COLONY

tains some magnesium carbonate, but in some localities it is sufficiently pure to yield good lime. Occasionally oolitic beds are met with, and when examined under the microscope these are found to contain organic remains, although no determinable shell has been seen. These are the only traces of fossils hitherto found in the Cango series.

Slates and fine quartzitic grits form a great part of the series. The slates are irregularly cleaved, and no rock of use for roofing has been found amongst them.

The intrusive rocks in the Cango district are nearly all altered to such an extent that the original minerals composing them have been replaced by others. At present the chief components are the fibrous variety of hornblende called uralite, green hornblende, augite, epidote, chlorite, felspar, quartz, calcite, sericite, magne- tite, apatite, and brown mica. The greater number of the dykes were originally dolerites without olivine, made up principally of augite and felspar; some contained much hornblende which still remains in the rock. The augite has been mostly altered to uralite, but kernels of the former mineral are still left within the patches of fibrous hornblende. The rock has often an ophitic structure, the felspar crystals lying partly or wholly within the patches of fibrous hornblende de- rived from augite. The calcite is sometimes sufficiently abundant in the rock to cause it to effervesce like an impure limestone when a drop of dilute acid is put on it. The calcite is often seen to partly replace the large crystals of felspar, but most of it occurs in the ground

PRE-CAPE ROCKS OF THE SOUTH AND WEST 57

mass of the rock. It is to be looked upon as one pro- duct of decomposition of the lime-soda felspar which once formed a large part of the rock. Epidote is often a very abundant constituent, and is probably derived from the lime-soda felspar. Little of the original felspar remains, although the outlines of that which has been altered to other minerals can usually be found in thin sections, and in the case of porphyritic crystals the pseudomorphs are easily seen by the naked eye.

Dykes of these altered rocks are fairly numerous in all parts of the Cango district ; they are usually only a few feet in width, but are traceable for considerable distances. In the valley of the Nels Eiver in the eastern Cango there are fifteen dykes in the slates within a distance of two miles, all traversing the rocks parallel with or at a small angle to their strike. In the valley of the river which leaves the Cango through Coetzee*s Poort three dykes are seen, the northernmost one is six feet thick, the second over 100 feet, and the southernmost is of much greater size and makes an out- crop nearly a mile in width. This great intrusive mass has been traced for twelve miles along the southern edge of the Cango between Coetzee's and Potgieter*s Poorts, forming rather prominent deep red hills (see Fig. 7). It is a peculiar type of rock, with much horn- blende forming ophitic plates enclosing the felspar, the hornblende is colourless and seems to have been formed from augite. The Gamka River, above the Ladismith Boad, crosses a dyke of peculiar diabase, in which the rather long crystals of felspar form radiating star-shaped bundles. Beyond a marked hardening of the slates or

68 GEOLOGY OF CAPE COLONY

grits in contact with the thicker dykes, there is little alteration in the sedimentary rocks near them.

It has already been said that some rocks resembling parts of the Cango beds occur in the neighbourhood of Saron and Honig Berg, perhaps overlying the Malmes- bury beds unconformably. There is no evidence in the Cango bearing directly upon the correlation of the distant outcrops, as no beds which can be determined as belonging to the Malmesbury series have been found in the district. The presence of granite boulders in the Cango conglomerates may indicate the later age of those conglomerates as compared with the granite intrusions of the so-called Malmesbury beds of the southern part of the Colony, George, and Mossel Bay. The quartz- felspar grits may have had a similar origin. In the absence of more reliable evidence this is of some worth.^ There are some points of resemblance to the Cango beds in the Ibiquas series north-east of Van Ehyn's Dorp to be noticed hereafter. At present it is useless to attempt to compare the ages of the Cango beds and the sedi- mentary rocks of Prieska and Griqualand West. All that can be said is that they are both older than the Cape formation.

The Ibiquas Series.

In the west of Calvinia and east of Van Rhyn's Dorp there is an area of conglomerates, grits, slates, and sandstones lying unconformably below the Table Mountain series of the Bokkeveld Mountain, and so

^ See Corstorphine, Geol. Comm. (98), p. 12.

PRE-CAPE ROCKS OF THE SOUTH AND WEST 59

distinct from the Malmesbury beds of the west and south of Van Ehyn's Dorp, upon which they appear to rest unconformably, that they have been placed in a separate group under the name Ibiquas beds.^ These beds occupy the greater part of the valley of the Doom Eiver If^bind the Stink Fontein Poort, where they can be well seen. The beds are considerably folded, but on the whole they dip eastwards, so that the base of the series lies on the western side of the area. The lower part of the series consists of conglomerates and grits, evidently derived from a granitic area, as granite and quartz-porphyry pebbles are conspicuous amongst the contents of the conglomerates, and the grits contain much quartz and felspar; sometimes these two min- erals are so abundant as to make the rock an arkose. There is thus a resemblance in these rocks to the quartz-felspar grits of the Cango, but the porphyroids of the Cango are as yet unknown in the Ibiquas series. The Ibiquas beds, like the Cango, have not (so far as is known) been invaded by granite, but only contain fragments of that rock in the conglomerates. The Cango beds are not known to rest unconformably upon the Malmesbury series, but there is at least strong presumptive evidence that the Ibiquas lie discordantly upon the latter in Van Khyn*s Dorp. In each case there is a weaker discordance with the overlying Table Mountain seriiss than exists at the junction between the latter and the Malmesbury beds. These points of similarity between the far peparated Ibiquas and Cango

' Oeol. Comm, (00), p. 26, etc.,

60 GEOLOGY OF CAPE COLONY

series suggest that they may belong to one and the same group of rocks, but until fossils are found in them the question cannot be settled.

Above the conglomerates and grits of the lower part of the series lie slates, sandy shales and sandstones, which rarely show distinct cleavage planes, such as are almost always seen in similar rocks in the Malmesbury beds. The shales and sandstones are met with on the steep escarpment of the Bokkeveld Mountain, and in the Doom River Valley. They are rather like the shales and sandstones of the Bokkeveld beds, but the thick groups of sandstone beds, so characteristic of the latter, are not found in the Ibiquas series.

Ripple markings are extremely well preserved in many of the sandstones throughout the series, and point to the deposition of the beds in shallow water. Large tracks and castings of some worm-like animal are occasionally abundant, but these are the only fossils known from the series. The nature of the rocks seems very favourable for the preservation of organic remains, and they are more likely to yield recognisable fossils than any other Pre-Cape rocks in the south and west of the Colony. They are unfortunately situated in a district which is thinly populated and difficult to get at. The thickness of the Ibiquas beds must be very considerable ; on the face of the Bokkeveld escarpment over 1,500 feet of these beds are exposed, but the base is some distance from the foot of the escarpment, and the highest beds visible lie about fifteen miles to the east, where they are covered by the Dwyka conglomerate. Although the beds are partly repeated by folding be-

'%gv^l(^{rai>9,vra]fli^{(^'i^i§r) west 61

feing I tfii t&i Bit^R. 'S^XSZb^S'./^i^^ ^^^ • their whole

III

3f II

62 GEOLOGY OF CAPE COLONY

The only intrusive rocks hitherto found in the Ibiquas beds are dykes of dolerite, evidently belonging to the same group of intrusions that form the sheets and dykes in the country occupied by the Karroo formation to the east and south-east.

The section in Fig. 8 illustrates the structure of the Ibiquas beds in the Doom River Valley. The line of section is so chosen that it runs across the fault on Klomp Boomen, and also through the Dwyka con- glomeratie resting upon the Table Mountain sandstone of the Bokkeveld Mountain on the south-west, and upon the granite on the north-east of the Doom Eiver Valley ; but if the section had been drawn along a line a few miles to the south of that chosen, the conglomer- ate would lie upon the Ibiquas beds.

Up to the present time the Ibiquas beds have not been found in the Malmesbury Division, but lately a group of comparatively unaltered shales and reddish sandstones has been noticed lying unconformably below the Table Mountain sandstone at two places on the Verloren Vley River in Piquetberg. The outcrops are on the farm Witte Drift within a few yards of the highly altered sericitic slates belonging to the Malmesbury beds. Although the actual contact of the shales and sandstones with the slates is obscured by alluvial de- posits there is little room for doubt that the former rest unconformably upon the latter. These shales and sand- stones may be regarded as part of the Ibiquas group, although there is no evidence from fossils to rely upon.

CHAPTEE III. The pre-cape rocks of the north and north-west.

Turning now from the southern and western districts to those lying north of the central basin of the Colony we find that no parallelism can at present be instituted between the rocks of the two areas, and the intervening country, composed probably to a great extent of granite and gneiss, is scarcely known from a geological point of view. The country lying between the Langebergen in the south end of Bushmanland and the Kaaing Bult, between Kenhardt and Prieska, including Bushmanland and the Kenhardt Division, has been traversed by Wyley and Dunn, but very slight accounts could be expected from rapid journeys through it, and they leave the con- nection between the better known rocks in Prieska and in Calvinia and Van Ehyn*s Dorp quite unexplained. The geology of West Griqualand was described by the late G. W. Stow,* and in the map published with his paper the extension of some of the various rock groups south of the Orange Eiver in the Prieska Division is roughly indicated. When the geological survey of the Prieska country was made in 1899 ^ Stow's classification was found to hold good, so the various names used by

1 Stow (73).

^ Oeol. Comm.t (99) ; the whole division has not -yet been mapped.

63

64 GEOLOGY OF CAPE COLONY

him for the West Griqualand groups of rock were ap- plied to the Prieska beds. There are several important points, however, which are not yet clear, and a vast amount of work still awaits the geological explorer in those regions. Stow's paper, one of the most important contributions to Colonial geology yet published, has suf- fered from a want of arrangement of the large array of facts contained in it, but it should be read by all who are interested in the north of the Colony.

Prieska and Griqualand West have an additional interest from the circumstance that some of the rock groups which occur there are very probably continuous with the formations overlying the Witwatersrand beds of the Transvaal (Cape system of Dr. Molengraaff). This part of the subject will be returned to after the structure of the country and the formations have been described.

Granite and gneiss form most of the lower lying part of Griqualand West and Prieska, rarely rising far above the generally sandy ground in hills or ** tors '* as the granitic rocks in Bushmanland and in the south-western districts do. The higher ground is composed of sedi- mentary rocks greatly altered from their original condi- tion both by pressure and by the intrusion of the granitic rocks. The chief hill ranges are : (1) the Campbell Band, or Eaap Plateau, trending south-west through Griqua- land West, and having no continuation in Prieska ; (2) the Asbestos Mountains, parallel to the Campbell Band on the western side, turning through almost a right angle where cut through by the Orange Biver and continued in Prieska by the Doombergen trending south-east ; (3) the ranges of Matsap and the Lange-

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 65

bergen, with a south-south-west trend continued south of the rivers in Ezel Band ; and (4) the Schurfteberg trending south on the north bank of the river, and con- tinued at first in a similar direction but farther south by the Brakbosch Poort range trending south-east, parallel to the Doornbergen, in Prieska. There are many smaller groups of hills parallel to the larger ranges and rising to moderate heights above the granitic plains which surround them.

It was stated in the introductory chapter that the hill ranges are parallel with the strike of the rocks compos- ing them ; the change in direction of the strike of the rocks indicated by the bending of the hill ranges near the Orange Eiver is a fact of the greatest importance in the structure of that part of the Colony.

The sedimentary rocks of these districts are divided up into the following groups from above downwards : —

4. Mats&p series.

3. Griqua Town series.

2. Campbell Band series.

1. 'Keis series.

The 'Keis Series.

The oldest rocks in Prieska are the quartzites and mica-schists of the 'Keis series, which form a long range or rather group of ranges of hills stretching from the Schurfteberg on the north of the river to Jonker Water, ninety miles to the south-south-east, where they dis- appear under the Dwyka conglonaerate. Inliers still farther south prove that they extend a few miles beyond the end of the main mass, but how far they stretch beneath the covering of the Karroo formation is un-

66 GEOLOGY OF CAPE COLONY

known. The dip of these rocks at the north end of the district is at high angles to the north-west, but on Ezel Klauw the dip changes to west and north-west, on Kaboom to west, and farther south still, from Brul Pan to Jonker Water the dip is west-south-west.

The quartzites are remarkably uniform in character, and have not been found to pass into conglomerate beds ; they are light in colour and contain small flakes of mica. By the increase in the amount of mica there is a gradual passage into mica-schists very rich in mica. The more micaceous the mica-schist is the more readily it disintegrates, and it is difficult to obtain fresh speci- mens of the highly micaceous rock, even from the bottom of wells from 40 to 70 feet deep. This is a remarkable fact in such a dry country as Prieska, where those pro- cesses of disintegration which depend upon the presence of moisture are very much reduced. As a consequence of their friable nature the mica-schists occur chiefly in the valleys ; they have in fact determined the positions of the minor valleys in the country occupied by the 'Keis series. The floors of the valleys are almost always deeply covered with sand derived from the rocks in the neighbourhood.

It has been found impossible to distinguish between the planes of bedding and those of schistosity in the mica-schist, and the same is the case with some of the quartzites belonging to the 'Keis series.

At Klein Modderfontein, on the north-east side of the outcrops of the 'Keis series, a rock similar to the highly micaceous schist, vdth the important addition of im- mense numbers of crystals of almandine garnet, occurs,

PRE-CAPE ROCKS: NORTH AND NORTH-WEST 67

interbedded with the usual quartzites of the series. In parts of the garnet rock the mica disappears and the garnets are embedded in quartz, often stained with green copper compounds.

Along the greater part of their course the 'Keis beds are flanked on either side by granite or gneiss, and areas of these rocks also occur in the heart of the series at Kaboom, Brakbosch Poort, and probably other places. At Boschiesman's Berg and Van Wyk's Pan tongue- shaped masses of gneiss project into the series from the great granitic area. These tracts of igneous rock are elongated in the direction of the strike of the 'Keis beds, and the fohation and planes of schistosity of the two rocks are parallel. On Grenaat's Kop there is an inlier of 'Keis beds surrounded by the Dwyka conglomerate, and a comparatively narrow dyke of granite traverses the inlier in a direction at right angles to the strike of the latter. The Grenaat's Kop dyke is the only clear case of intrusion of the granite in the *Keis series seen in the district. In other parts the contact of the igneous and sedimentary rocks has not been seen, owing to the thick covering of sand, and it would be possible to account for many of the facts observed on the supposi- tion that the 'Keis series was deposited upon a floor of granite and that at some subsequent period the rocks were intensely folded, so that on the one hand ridges of gneissose granite were formed projecting into the quartz- ites and schists, and on the other steeply folded synclines of the sediments went down into the granite.

At many places in the granitic areas both east and

west of the ridges of *Keis hills there are isolated len-

5*

68 GEOLOGY OF CAPE COLONY

ticular patches of highly metamorphosed rocks, iiiica< schists, and banded hornstone-like rocks with much epidote and quartz in them, bearing evidence of having been of sedimentary origin, as well as quartzites. These detached outcrops were probably once connected with the main area of the *Keis beds. There are other masses of altered sedimentary rocks situated in the granite areas more like beds in the Griqua Town and Campbell Band series, and it will be more convenient to give a further account of the relation of the granite and gneiss to the 'Keis series after these have been described.

Stow^ described the occurrence of some ** ancient schistose" rocks lying unconformably below the Camp- bell Band series west of Campbell Town ; they are quartzitic rocks with calcareous matter added by in- filtration from the overlying beds. These older rocks are also marked on his map as being found north of Jonker Water in Prieska, but the outcrops at the latter place undoubtedly belong to the 'Keis beds. The ** ancient schistose *' rocks near Campbell Town seem, from Stow's account, to be similar to parts of the *Keis series also, so it is not unUkely that there is direct evi- dence of the unconformable succession of the Campbell Band to the *Keis beds in West Griqualand. In Prieska no evidence on this point has been obtained, as the Campbell Band beds hitherto recognised there are only found at some considerable distance from the older series.

A bed of limestone, presumably interbedded with the quartzites and mica-schists, has been found on the farm

1 Stow (73), p. 619, and PI. XXXIX., Fig. 4.

iiifllll:""

1 ^.

I I 4

I II:

> ill

iS

I

I Jl I lill

70 GEOLOGY OF CAPE COLONY

The Campbell Rand beds consist of quartzites, mica- ceous schists, limestones and cherts. The true base of the series has not been recognised in Prieska, but it is almost invariably the case that the limestones are un- derlain by a considerable amount of quartzite, varying from 200 to 2,000 feet in thickness. At Zeekoe Baard in Prieska the quartzites are apparently conformably underlain by green slates, which have not been seen elsewhere in the series. The lowest beds of the series in the Kaap Plateau are limestones and quartzites, but there does not seem to be such a definite group of quart- zites at or near the base in that district as there is in Prieska.

The quartzites in Prieska are of very much the same nature as those belonging to the 'Keis beds, but mica- schists are much less extensively developed than in the latter series.

The limestones are dark coloured and thoroughly crystalline, usually weathering with a peculiarly rough brown surface, a character that has led to the rock being known as 01iphant*s Klip from its resemblance to an elephant's skin. The limestone often contains a certain percentage of magnesium carbonate, and is therefore a dolomitic hmestone.

On the right bank of the Orange Eiver, opposite Buis Valley, there are some fine vertical cliffs of the lime- stones rising straight out of the water for some distance along the river ; the face of the cliff is indented as if by shallow caves, but there seem to be no caves of any noteworthy extent as there are in the Cango limestones and in the dolomitic limestones of the Transvaal, al-

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 71

though the latter very probably belong to the same series as the Prieska rock.

Thin layers of chert, often somewhat irregular and nodular, are very abundant in the limestones. Although several specimens have been carefully examined under the microscope for traces of organisms that have been found in rocks of this nature in many parts of the world, nothing obviously of organic origin has yet been seen in them. The chert is a very hard rock which breaks into pieces with sharp, splintery edges. The hardness of the chert made it a suitable one for the natives to use as rough cutting and scraping tools, but it seems to have been less used for such purposes than the jasper of the succeeding group of rocks. The beds and nodules of chert stand out from the general surface of the lime- stones in which they lie, owing to the more rapid solution of the limestone, and give rise in places to re- markably jagged and uneven surfaces.

No fossils have been recorded from the Campbell Band beds ; but of late years one has heard so many rumours and statements to the effect that they have been seen in more than one locality in West Griqualand, that the discovery of some recognisable forms may be confidently expected. Any such find will be of very great interest, for without fossils the age of the old rocks in the north can never be satisfactorily determined.

On the farm called Alicedale in Prieska, there is a band of crystalline limestone about fifty feet thick associated with mica-schist, quartzite, and magnetic quartzite, the latter is like some of the rocks belonging to the Griqua Town series ; the beds dip vertically and

72 GEOLOGY OF CAPE COLONY

form a lenticular area surrounded by granite. The mica-schist contains coarse veins of pegmatitic granite with large plates of white mica. The limestone has a band of schistose rock in it with crystals of almandine garnet as much as two inches in diameter. The garnet seems to be a product of metamorphism due to the .proximity of the granite. Whether the mica-schist belongs to the limestone and quart zitic group of the Campbell Band group is not certain.

The limestone of Zeekoe Baard contains thin beds of red jasper, like some of the jaspers of the Griqua Town series, but the occurrence of jasper interbedded with the limestone strata seems to be more frequent to the north of the river than in the Prieska Division.

The maximum thickness of the Campbell Band series is about 7,000 feet in Prieska, but towards the south- eastern part of the Doornbergen it disappears or gets very thin, a fact of which the true explanation has not been ascertained.

Some rather large masses of galena are met with near the base of the Campbell Band beds on the western flank of the Doornbergen near their northern end. Curious veins of white quartz and pink orthoclase are found in the limestones at Zeekoe Baard.

The relationship of the Campbell Band group to the overlying Griqua Town series is best seen between Nauga and Buis Valley, where they have undergone less disturbance than farther to the south-east. The structure of this part of the Doornbergen is broadly a double syncline, and is represented in Fig. 9. The limestones dip under the Griqua Town beds on Kalk

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 73

Fontein, reappear in a narrow anticline on that farm, and pass under the higher beds again to the east, and rise to form the banks of the Orange Kiver on Buis Valley. To the south-east, along the south-western flank of the hills, the beds are frequently overturned, so that the Griqua Town beds dip at high angles under the limestones, and these in their turn under the quart- zites of the base of the Campbell Band group.

The Griqua Town Series.

The Griqua Town series forms the rugged hilly country that stretches sixty-five miles south-eastwards from the Orange Biver at Kameel Puts to Doornberg's Fontein, generally known as the Doornbergen. To the north of the river the series passes north-eastwards in the Asbestos Mountains, and extends far into Bechuana- land, probably reaching the borders of the Transvaal, but nothing definite is yet known of that part of the country.

The series consists of peculiarly heavy green slaty rocks with quartzites and jaspers containing large quantities of magnetite. Much of the rock is banded, the thin layers having slightly different colours of which deep red, bright red, brown and black are the most usual. The black layers are almost entirely composed of minute crystals and grains of magnetite, with a little quartz between the grains; every intermediate stage between almost pure magnetite and pure quartzite can be found ; the quartzites vnth least magnetite in them are met with near the base of the series. The Doorn- bergen, as a whole, contain so much magnetite that a magnetic compass is of very little use in their neigh-

74 GEOLOGY OP CAPE COLONY

bourhood. The jaspers are very fine grained rocks which break with a smooth conchoidal fracture. They are made up of extremely minute crystalline particles of quartz, and are coloured by oxides of iron of various degrees of hydration. The crystalline structure is due to changes that have taken place since the formation of the sediments. The jaspers often contain much magnetite in small grains and crystals. In the lower part of the series near Prieska Poort some highly ferru- ginous rocks with oolitic structure are interbedded with the more usual type of rock. The oolitic beds were probably ferruginous limestones that have been altered to their present condition. The magnetic quartzites and jaspers were probably highly ferruginous rocks when deposited ; the thin layers of various composi- tions continue for considerable distances without ap- preciable variation, and are inexplicable on the assump- tion that the iron was brought into its present position by infiltration.

The Griqua Town beds are the home of the blue crocidolite (a fibrous amphibole related to riebeckite), which is used for various purposes under the name of asbestos ; ^ the alteration product due to the oxidation and slight enrichment by quartz of the amphibole fibres is called griqualandite. There are many stages in the process; sometimes the crocidolite is partly replaced by quartz before any oxidation takes place, and a hard blue mineral results, in other cases the

^ True asbestos is another variety of amphibole ; another mineral, chrysotile, found in veins in serpentine, is often called asbestos, and is used for similar purposes.

PHE-CAPE ROCKS : NORTH AND NORTH-WEST 75

oxidation, made obvious by the yellow-brown colour of the fibres, is in advance of the silicification. The ultimate product is a very hard mineral which takes a fine polish, and has a delicate fibrous structure pre- served in it giving rise to the beautiful chatoyant lustre characteristic of the mineral. The unaltered crocidolite is found in blue-green, heavy, slaty rocks, which are much softer than the jaspers. Thin vein-like layers of crocidolite parallel to the bedding planes are found in the slates, usually in places where the slates are bent, and the layers are thickest in the crests and troughs of the folds, often disappearing altogether when followed along the limbs. The fibres stand perpendi- cular to the surfaces of the layers. The griqualandite occurs only in the jasper slates ; these facts point to the simultaneous conversion of the heavy slates into jasper rocks, and of the crocidolite into griqualandite.

The surface of some of the beds in the series bear well-preserved ripple markings, which are crossed by a sharply defined set of ridges and troughs due* to sub- sequent movements in the rocks.

The Griqua Town beds are often very much folded ; in the Doornbergen they occupy the bottom of a trough- shaped fold running north-west, which is partly over- turned, so that on the south-west flank of the range they dip south-west towards the granitic area between that range and the ridge of *Keis hills on the western border of the division of Prieska. There are many isolated patches of highly magnetic quartzites and white quartzites, which rise above the general surface of the granite and gneiss. One such mass is twenty miles

76 GEOLOGY OP CAPE COLONY

long; it stretches from Zwart Kop Pan to Jackals Water. Whether it belongs to the Griqua Town or Campbell Band group is uncertain, but it and the other similarly situated lenticular masses may be looked upon as pieces of the same rocks that form the Doornbergen, separated from the main area by denudation in an in- tensely folded district, that is further complicated by the intrusion of the granitic rocks, as well as other igneous materials which we shall speak of later.

The thickness of the Griqua Town series is not known, but it must be considerable, although the ap- parent thickness in Prieska is certainly much increased by folding. The top of the group has not yet been found.

The MatsXp Series.

The MatsAp series forms the Ezel Band in Prieska, and the Langebergen and Mats&p hills to the north of the Orange Eiver. It is composed of quartzites and coarse grits with conglomerates at the base. The con- glomerates contain many pebbles of jasper and magnetic rocks probably derived from the Griqua Town beds. The grits usually have a peculiarly mottled colour. The quartzites and grits are distinguishable in even small pieces from both the Campbell Band and 'Keis quartzites. In the Ezel Band the beds dip towards the north-north-west at fairly high angles, and are at least 3,000 feet thick. In the Langebergen they are more folded than in the Ezel Band, but parts of the rock in the latter range also show evidence of having been subjected to great pressure and movements.; some of

■ [■!lS':'jS6Blill!a»»»H''H-WE8T 77 *" Jlitjdfllklltifflboat lost the

S II 1 ■is- 1 a3

78 GEOLOGY OF CAPE COLONY

Much remains to be doxie before the sedimentary rocks of Prieska and West Griqualand can be properly under- stood. •

There is a parallelism between the Campbell Band, Griqua Town, and Matsap series and the Transvaal rocks which Dr. Molengraaff considered to belong to the Cape system,^ an opinion he has lately ^ seen reason to modify in view of the probably greater age of the Griqualand and Prieska beds. It has already been mentioned that the Campbell Band and Griqua Town beds probably extend to the Transvaal border. Prom Dr. Molengraafifs description of the Black Beef, Dolo- mitic and Pretoria series, it seems very probable that they are the same beds as those called the Campbell Band and Griqua Town beds by Stow. Dr. Molen- graafif's account of the Waterberg sandstones in the Palala plateau agrees rather closely with those of the Matsap beds in Prieska^ and West Griqualand, except that the Waterberg sandstones are thought to succeed the Pretoria beds conformably, although usually sepa- rated from them by the great laccolitic intrusion of the Boschveld red granite and its local modifications, the ** newer granite" of the Transvaal, an intrusion that has no exact analogue in Prieska.

The beds in the two countries may be tabulated thus : —

Cape Colony. Transvaal.

Matsap series - - _ . Waterberg sandstones. Griqua Town series - - - Pretoria beds, n u 11 T> ^ I limestones, Dolomite series.

Campbell Rand series • q^^^ites/ Black Reef series.

1 Molengraafi (01). ^ Molengraaff (03).

» Molengraaff (01) ; Qeol, Comm. (99), p. 82 ; Stow (78), p. 682.

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 79

It will be noticed that the 'Keis series finds no place in this comparison, but if that group really lies uncon- formably below the quartzites and limestones of the Campbell Band, a not improbable view, the Transvaal representative of the group must be looked for in the "Primary formation" of Dr. Molengraafif. If, on the other hand, the schistose rocks lying unconformably below the Campbell Band series in West Griqualand prove to be distinct from the 'Keis beds, the latter may have to be regarded as part of the Campbell Band group, though there seems to be but slight evidence in favour of that view at present.

The intrusive igneous rocks of Prieska are of great interest and of varied character, but only a short account of them can be attempted here. By far the most im- portant are the granite and gneiss of the district between the Doornbergen and the western hills of the division, and the similar rocks of the Kaaing Bult to the west of the latter hills. It has already been stated that the foliation planes of the gneiss are in general parallel to the strike of the sedimentary rocks in its neighbourhood. It is probable that the granite and gneiss, the extreme types of each of which are connected by many inter- mediate steps, were intruded amongst the 'Keis, Camp- bell Band, and Griqua Town beds during the production of the greater part of the folds into which these rocks were thrown. The acid igneous rocks as a rule do not show sufficient evidence of having been violently folded after their consolidation to permit the idea being held that they were subjected to the same degree of pres-

80 GEOLOGY OF CAPE COLONY

sure that affected the sedimentary rocks. At the same time their component minerals frequently show optical anomalies due to pressure ; it is not unlikely that the intrusion and solidification of the granite and gneiss occupied a long period, and that we see in the gneiss the earlier and consequently most altered products of the acid magma. Occasionally the gneissose rocks have structures that were produced by pressure and move- ments after their consolidation, such as areas of quartz and felspar mosaic surrounding the larger felspar and quarfcz grains, and the development of thin layers of very minute white mica flakes at the contact of some of the other constituent minerals. Whether any part of the granite is of much later date than the bulk of the intrusions is not yet settled. Some of the very fresh looking granites on the farm Schalk's Puts might certainly be considered younger than the gneiss, but there are so many intermediate varieties that the evi- dence of a considerable difference in age between the extreme types must be clearly made out before that opinion can be accepted.

The chief constituents of the acid intrusions are quartz ; orthoclase, microcline, albite, and an inter- growth of orthoclase or microcline and a plagioclase felspar ; black and white mica, the latter sometimes (e.g. Grenaat*s Kop and Alicedale) in crystals up to ten inches in width, but too frequently bent by the move- ments which the rock has undergone since its solidifica- tion ; hornblende is not often met with ; apatite and iron ores are hot abundant ; garnets occur, especially in certain gneisses, and in the rocks with the same con-

L

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 81

stituents as the granite but with granulitic structure. Tourmaline seems to be absent from the Prieska granites.

Pegmatite or graphic granite, chiefly composed of an intergrowth of microchne and quartz, forms a large mass in the neighbourhood of Saft Sit Pan. Quartz- porphyries are rather restricted in their occurrence; they have been found only within the granite areas, and are not known to traverse the surrounding rocks in the manner of the quartz-porphyries near Paarl Berg.

The granulites of Prieska are abundant and vary greatly in composition. They are fine-grained rocks, usually showing distinct banding on large weathered surfaces, but the banding is often unobservable on a freshly broken surface. They are usually dark in colour, but the more acid or siliceous types are light coloured. In general appearance they look rather like even-grained quartzites. It is only under the micro- scope that the distinctive features of the granulites are seen. The most striking character is the imiformity in size of the grains of the various minerals composing the rocks; another important feature is the almost complete absence of crystalline faces in the minerals, which seem to have separated out in a different manner from that usual in igneous rocks; enclosures of one mineral by another are abundant, but the enclosed mineral is irregularly shaped, usually with a rounded outline. Garnet, which is an important constituent of most of the Prieska granulites, is the only mineral which sometimes shows crystal faces, and it very often

82 GEOLOGY OF CAPE COLONY

contains smiftU grains of more than one of the other minerals composing the rock. All the minerals in the granulites are remarkably fresh and free from altera- tion products. The rocks may be broadly divided into three groups : (1) GranuHtes made up of the same minerals as the granite and gneiss, viz., quartz, felspar (orthoclase and plagioclase), garnet, and biotite. This seems to be a less abundant rock than those belonging to the two other classes to be mentioned, but on account of its being rather closely related to much of the gneiss, into which it passes by the coming in of a pronounced foliation and the increase in size of some of the felspars, it is easy to overlook small outcrops in the gneiss areas. (2) Hornblende-granulites, composed of quartz, ortho- clase, albite, hornblende, biotite, magnetite, garnet, and sphene. The hornblende is a pale bluish-green variety, different from the hornblende of most of the hornblende schists. Garnet is a less abundant mineral constituent than in the next group. (3) Pyroxene-epidote-granulites, composed of plagioclase, augite, epidote, garnet, magne- tite, sphene, and frequently hornblende. The pyroxene is a pale green or bluish-green monoclinic variety, diop- side, and is slightly pleochroic. The abundance of epi- dote, which often forms a large part of the rock, is very remarkable.

The granulites form elongated outcrops in the granite and gneiss, with the longer axes of the areas parallel to the foliation planes of the gneiss ; they have not been found as intrusions in the sedimentary rocks. The nature of their contact with the gneiss has not been made out, as the line of junction of the two rocks is

PRE-CAPE ROOKS: NORTH AND NORTH-WEST 83

almost invariably concealed under the red sandy soil that the granite and gneiss give rise to. The composi- tion of the pyroxene-granulites must be very different from that of any of the gneissose or schistose rocks yet found in Prieska, and it is therefore impossible to consider them as local modifications of any of the latter, as the biotite-granulites may be with regard to the gneiss. The amount of lime and alumina in the pyroxene-granulites must be greater than is usual in igneous rocks containing the same varieties of plagio- clase. The granulites give one the impression of being intrusive, but the question of their origin is quite un- settled.

The homblende-granulites are connected by inter- mediate varieties with some of the hornblende-schists, which form dykes in both the granitic and sedimentary rocks of Prieska. Two main varieties of the horn- blende-schist occur, one contains blue-green hornblende, felspars, and much garnet and quartz ; and the other is made of actinolite, with a very little felspar and quartz. The hornblende-schist dykes in the sedimentary beds are probably highly altered igneous rocks; the blue- green hornblende is at places so abundant that the rock consists of little else.

The blue amphibole called glaucophane forms an important constituent of some of the schistose rocks; the other minerals in the glaucophane - schists are epidote, quartz, orthoclase and microperthite.

There are several varieties of much altered rocks

that originally consisted of augite and felspar, but

which are now usually a mass of minute fibres of

6*

84 GEOLOGY OF CAPE COLONY

hornblende, and small grains of epidote, calcite, quartz and felspar, although the remnants of the original augite which formed ophitic plates can be seen in some specimens ; the outlines of the former crystals of felspar can often be dimly seen under the micro- scope. Up to the present time those rocks have not been traced into the typical hornblende-schist, but from the close resemblance of specimens gathered in one and the same district to the different stages in the Scourie dyke described by Mr. Teall,^ in which the alteration of an augite-plagioclase rock into horn- blende-schist, very like several of the Prieska schists, was proved, it is to be expected that the whole series of changes will be found in one rock-mass in Prieska. These altered augite-plagioclase rocks (dolerites) are always distinguishable in the field from the similar rocks with or without olivine belonging to the dolerite intrusions of late Karroo age which occur in Prieska both in the Karroo formation and in the rocks older than the Dwyka conglomerate. The Pre-Karroo dol- erites are dull-looking and greenish in colour owing to the alteration of their constituents, but the later ones are bluish-black, and when freshly broken the felspar cleavage faces, even within a tenth of an inch of the weathered surface, are bright and unaltered.

There are some dyke rocks at Zwart Kop Pan and Zeekbe Baard that are made up largely of olivine and augite with some basic plagioclase ; the olivine is partly changed into serpentine. These rocks, which have

* British Petrography ^ p. 197, etc.

PEE-CAPE ROCKS : NORTH AND NORTH-WEST 86

rather too much felspar in them to be called augite- picrite, but may be named olivine-gabbro, have lio apparent connection with either the older or newer dolerites, and their age is unknown, but the fresh condition of their minerals points to their being later than the Pre-Karroo dolerites and schists.

Two large masses of serpentine have been found in Prieska, one at Zwart Kop on Blink Fontein, and the other at Zoet Vley. They are almost entirely composed of serpentine with the addition of a small quantity of opaque iron ore and calcite or magnesite. The ser- pentine does not contain unaltered grains of any mineral that it could have been derived from, and the arrange- ment of the fibres is not like that in serpentines derived from olivine, but frequently seems to be due to the de- velopment of fibres parallel to the prism cleavages of a pyroxene, as the fibres often form a square net- work. The serpentine contains veins of chrysotile, a white or pale-green fibrous variety of serpentine which can be used for some of the purposes to which asbestos is put. The serpentine forms dykes or sheets in magnetic quartzites and jaspers probably belonging to the Griqua Town series. The Blink Fontein magnetic rocks are an outlier in the middle of the granite, but those of Zoet Vley occur as an inlier in the Dwyka conglomerate south of the Doornbergen.

Volcanic Books.

In the general description of the Prieska and Hope Town districts, published by the Geological Commission

86 GEOLOGY OF CAPE COLONY

in 1900, mention is made of two groups of amygdaloidal rocks, one of which was called the Beer Vley series ; the other, the Zeekoe Baard amygdaloid, on account of its complicated field relationships, was regarded as intrusive. Since that report was written the rocks have been partially examined under the microscope, and there is reason to modify some of the conclusions based on the field evidence alone.

The Beer Vley group consists of amygdaloidal lavas of an andesitic type, with pseudomorphs of chlorite after hornblende and pyroxene; and more acid lavas, rhyolites with crystals of quartz and felspar lying in a devitrified matrix which has perlitic cracks in it. The amygdales in the Beer Vley rocks are filled with chal- cedony and chlorite, rarely with calcite. Some agglom- erates, evidently composed of fragments of andesites and more acid lavas, have been found interbedded with the lavas. Beyond the fact that these volcanic rocks are older than the Dwyka conglomerate nothing is known as to their age, for they have only been found as inUers in the Dwyka area at Beer Vley, Bidouw Kuil, Jorsten's Berg, and Brak Pan. They are apparently much less altered than the Zeekoe Baard amygdaloids; but the latter, being of a more basic type than the Beer Vley group, contained more minerals that are easily changed.

The Zeekoe Baard amygdaloids are compact dark blue and green rocks with amygdales of calcite, chalce- dony, and chlorite, or a mixture of two or more of these minerals. They occupy a large area in Prieska, and also in Griqualand West. In Prieska they surround the south-west end of Ezel Band (see Fig. 10), and

PRE-OAPE ROCKS : NORTH AND NORTH-WEST 87

form a wide area between the granite and gneiss on the west, and the sedimentary rocks of the Doombergen on the north-east; they also occur as a long strip in the granite area west of Prieska's Poort.

These rocks vary considerably in mineralogical com- position. They are usually very much altered, and in specimens from some of the outcrops hardly any of the original constituents can be recognised; chlorite, epi- dote, calcite, and quartz make up nearly the whole of the rock in many cases, and all these minerals are pro- bably alteration products. In no case has the original dark constituent of the rock been observed, although either hornblende or augite was certainly an important constituent of parts of the rock. At Blink Fontein the rock is less altered than usual, and is there composed of crystals of plagioclase, some of which is andesine, set in a very fine-grained ground mass of probably quartz and felspar, some chlorite and opaque iron oxides are also present. This rock is evidently a less basic one than the amygdaloid at other localities, such as Zeekoe Baard, where it has been largely altered to epidote and calcite. At only one place, near the south-west end of Ezel Band, has a breccia or agglomerate been seen which might belong to this volcanic group, but there is some doubt as to the true relationship of the breccia to the volcanic group and the Matsap beds.

There is a similarity between the Zeekoe Baard amyg- daloid and the amygdaloidal rocks in the Transvaal that are now known to be older than the Black Eeef series. In the Prieska district, however, there is a difficulty in supposing that the amygdaloids are older than the

88 GEOLOGY OF CAPE COLONY

Campbell Band quartzites; for although they lie at the base of that series in several places, yet they are in contact with both higher and lower beds at other locali- ties, and it is more in accordance with the observed facts to regard the amygdaloids as having been poured out at the surface subsequently to the folding and denudation of the Campbell Band and Griqua Town series, but previously to the deposition of the Matsap group.

It is possible that the Prieska amygdaloids may be found to belong to the same group as the Boschveld volcanic rocks of the Transvaal.^

The reasons for classing the 'Keis, Campbell Band, Griqua Town and Mats&p beds as Pre-Cape rocks must now be explained. We have seen that the Matsdp beds are represented by a mere remnant in the Prieska Divi- sion ; that they were much folded before the deposition of the Dwyka conglomerate is proved by the fact that the Dwyka and overlying beds lie horizontally and undisturbed in the same district. The conglomerate lies in the ancient valleys of the Doornbergen, which have to a large extent been re-excavated, so that only outliers of the conglomerate are left as witnesses that the whole range was carved out of solid rock in Pre- Dwyka and Dwyka times. In neighbouring localities the conglomerate rests upon the Griqua Town, Camp- bell Band and 'Keis beds, as well as upon the granites and gneiss, proving that the whole thickness of the sedimentary rocks was removed from certain areas before the conglomerate was formed. The conglo-

1 MolengraAf! (01), p. 62.

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 89

merate has not yet been found lying upon the Matsap beds, but fragments of these occur in it, and there can be no doubt that the greater part of the rocks belonging to the Matsap series formerly present in Prieska were removed by denudation before the deposition of the conglomerate. Taking the thickness of the Matsap beds as 3,000 feet, and that of the Campbell Band and Griqua Town series together as 5,000 feet, and omitting the *Keis altogether as being possibly of the same age as the Campbell Band group, we have a total of 8,000 feet of rock removed from certain parts of the district before the conglomerate was laid down in the same area. It must be remembered that this thickness is a low estimate, and that the whole of the volcanic group is omitted from the argument on account of the uncertainty as to its age.

In this district, therefore, before the Dwyka conglo- merate was deposited, the Campbell Band, Griqua Town and Matsap beds were greatly folded and the greater part of them was removed altogether. All this must have occupied a very long time in a geological sense. In the south of the Colony, as was explained in the Introduction, and as will be described in more detail in later chapters, there was a continuous deposi- tion of sediments (the Cape formation) about 10,000 feet thick, before the conglomerate was laid down con- formably on them. It is obvious that at any rate the upper part of the 8,000 feet of sediments that were removed in the north in Pre-Dwyka times could hot have been formed during the deposition of the beds immediately preceding the Dwyka series in the south ;

90 GEOLOGY OF CAPE COLONY

for the folding and denudation of the northern rocks must have taken place during that period or earlier. It is, of course, difficult to base an argument as to the contemporaneity or otherwise of the beds in the two areas on a comparison of the rate of deposition in the one and that of denudation in the other \ but it is clear that the folding and removal by denudation of the 8,000 feet of sediments in Prieska must have occupied a considerable part of the time during which the 10,000 feet of the Cape formation were formed in the south and west of the Colony. When it is remembered also that 8,000 feet is a small estimate, for the upper parts of both the Griqua Town and Matsap series are un- known, it must be admitted that there is strong reason to regard the Matsap beds as of pre-Cape age, and still more so the Griqua Town series and the underlying rocks.

Namaqualand Schists.

Under this name Mr. Dunn includes the schistose rocks that cover wide areas in the Namaqualand Division ; amongst them are hornblende-schists, epidote-schists and others that are igneous rocks greatly altered from their original condition ; but there are also sedimentary rocks, such as conglomerates, quartzites, limestones and mica schists. Very little is known of these beds.

Near the Orange Eiver there are some quartzites that Mr. Dunn regarded as Witteberg beds ; ^ they lie flat and unconformably upon the Namaqualand schists.

' Geological sketch-map of South Africa (87).

PRE-CAPE ROCKS : NORTH AND NORTH-WEST 91

From the accounts^ of German South- West Africa it seems very likely that those quartzites are the same as those of the Huib and Han-ami plateaux, which are overlain by limestones, and are perhaps the western representatives of the Campbell Band group.

Granite, Gneiss, etc., of the North-West.

A great part of the north-west is occupied by acid, igneous rocks. From the west coast, north of the Bitter Biver, these rocks extend across Little Namaqualand and Bushmanland into Prieska, where they are probably continuous with the gneiss and granite previously de- scribed. The geology of this great tract of country is only known in its barest outlines. The igneous rocks are probably intrusive in the Namaqualand schists. Their southern boundary in Van Bhyn*s Dorp and Galvinia is the line of fault along which the Ibiquas beds are thrown down against them. To the east the boundary is formed by the Dwyka conglomerate.

Amongst the southern Bushmanland granites and gneisses there are rocks of peculiar types ; some well- foliated gneiss at the base of the Langeberg in Calvinia consists chiefly of quartz, plagioclase, enstatite, horn- blende and biotite ; it occurs in bands enclosed in gneiss of a more normal character. Garnetiferous granite and gneiss are abundant in that area. The general strike of the foliation planes is somewhat to the north of east.

The copper ores of Namaqualand are chiefly found in a rock rich in hypersthene ; it is called a greenstone by

» Vop Reichenbach (96), p. 117, etc.

92 GEOLOGY OF CAPE COLONY

Wyley ^ and a dioritic rock by Schenck,^ and appears to form bands in the gneiss. The ores were regarded by Wyley as long ago as 1856 as constituents of the igneous rock concentrated in certain parts of its mass, a view that has again been stated by Schenck. The principal ore is the purple bornite, but the less valuable copper pyrites, chalcopyrite, is abundant in some of the mines, and many other copper-bearing minerals are present in smaller quantities.

» Wyley (56), p. 5 ; and (67), p. 30, etc. « Schenck (01), pp. 64, 66.

CHAPTEE IV.

THE CAPE SYSTEM.

The rocks belonging to the Cape system have only been found in the southern and eastern parts of South Africa ; from Van Ehyn's Dorp in the west, round the coastal districts to the Gualana Eiver, and again northwfirds from the St. John's Eiver into Natal the Cape system plays an important part in the structure of the country. The true succession of these rocks was made out in part by A. G. Bain, but the numerous folds they have been thrown into in the west together with some litho- logical resemblances between parts of the two upper series were responsible for the mistake he made in limit- ing the occurrence of the Witteberg series (the " Car- boniferous** group of Bain) to the eastern province. Moreover it is evident from the gap left in his map between the Kammanassie and Cockscomb Mountains that Bain never had the opportunity of connecting the west and east satisfactorily. This was partly accom- plished by Wyley and Dunn ; but meanwhile a serious error had been introduced by certain observers ^ taking the Bokkeveld beds to be lower in stratigraphical posi- tion than the Table Mountain sandstone, a mistake that

^ Rubidge (58), p. 195, etc. ; Hochstetter (66), p. 31, etc. ; Cohen (87), p. 202, etc.

98

94 GEOLOGY OF CAPE COLONY

led to the identification of the Bokkeveld and Malmes- bury beds on the one hand and of the Table Mountain and Witteberg series on the other. This unfortunate confusion which is not met with in the maps or writings of men who had a considerable personal knowledge of the rocks concerned, such as Bain, Wyley and Dunn, did much to obscure the structure of the Colony. The work of the survey has clearly demonstrated the cor- rectness of Bain's view of the superposition of the Bokkeveld on the Table Mountain series, and the ex- tension of the Witteberg series over wide areas in the south-west, which were indeed made plain by Wyley ^ and Dunn.^ The three members of the Cape system have now been so frequently traversed and mapped between the Cederbergen and Uitenhage by the geolo- gists of the Geological Commission^ that there can no longer be any doubt as to their relationships to one another.

The Table Mountain Series.

This group of rocks forms the most conspicuous features in Cape Colony. Table Mountain itself, rising 3,553 feet above the sea, is visible long before the ship that brings the new-comer to South Africa reaches Table Bay, and on the mountain several characteristics of the series can be seen. The Peninsula mountains, however, are merely small outliers of the main portion of the Table Mountain beds in the Colony.

» Wyley (69). 2 Di^n^ (72, 75, 87).

' Oeoh Comm. (96-99). For a more detailed account of the history of the question see Corstorphine, Qeol. Comm, (97), p. 31, etc.

THE CAPE SYSTEM 95

A description of the distribution of the series will serve also as a description of the main tectonic or struc- tural features of the southern part of the Colony. The broad outline of the structure has been given in the Introduction, but as nearly every important anticline in the south is marked on the surface by a ridge of Table Mountain sandstone a more detailed account will not be out of place here. The position of the main anticUnes mentioned below will be found in Fig. 3, and in the map at the commencement of the volume.

On the seaward side of the folded belt of sedimentary rocks forming the second of the three regions into which the Colony is divided in the Introduction for the purpose of a general description, the Table Mountain sandstone becomes less steeply folded over large areas than any- where within the belt itself. On the west, in the coastal plains of Clanwilliam and Piquetberg, the sandstone lies at low angles ; by its removal the underlying Malmes- bury beds and granite have been laid bare in the divisions of Van Ehyn's Dorp, Piquetberg, Malmesbury, Cape, Paarl and Stellenbosch, and the outliers of the Penin- sula mountains, Eiebeek's Kasteel and Simon's Berg bear testimony to its former extension over that part of the Pre-Cape region of the south-west as a gently undu- lating mass.

A long outlier, faulted down on the north-east side, forms Joosten Berg in the south of the Malmesbury division ; Klapmuts Hill, on the same line of strike, is a similar faulted outlier north-west of Simon's Berg.

To the east of the Peninsula the present coast line passes somewhat irregularly through the marginal part

96 GEOLOGY OF CAPE COLONY

of the folded belt, for although the Table Mountain sandstone is more folded than in the Peninsula or Piquetberg, yet the plications are fewer and much less abrupt than farther inland. The shore at Cape Hangklip, Hermanns, Danger Point and Agulhas, as well as at many intermediate points, is cut out of the slightly bent sandstones. East of Agulhas the coast trends to the north of east and cuts across the folded belt slanting- wise, and the sandstones of Capes St. Blaize, St. Francis and Eecife are highly inclined, for they lie well within the folded belt. There is no direct evidence of the nature of the rocks under the sea floor, but it is prob- able that the Table Mountain sandstone is continued in a slightly bent condition some distance towards the edge of the Agulhas bank. The condition of the sandstone off the south-east coast, if it exist there, is of course quite unknown, but from the close analogy between the structure of Pondoland and Natal, and that of Van Ehyn's Dorp, we may suppose that the Table Mountain series formerly extended in a slightly bent condition right round the outer side of the folded belt.

In the west the first (see Fig. 11) pronounced folds met with form the sandstone mountains on the left side of the Olifant's Eiver valley, where the sandstone is thrown into gentle anticlines trending north-north-west. The valley of the Olifant's Eiver, from its source west of the village of Ceres to a point below Clanwilliam, occupies a syncline in which remnants of the Bokkeveld beds are still preserved at three places. South of the Pikenier's Kloof the western limb of the anticline west of the river has mostly been removed by denudation, and

liiiieii^iii

ffiif-

I tlil

*if fl if'^l^^l'l}!!""^

-4"^"

P

Is

1^1

THE CAPE SYSTEM 99

the other into the Oorlog's Kloof River that hes in a deep precipitous valley about six miles behind the escarpment. The Table Mountain series comes to an end with the Bok- keveld Mountain, although the escarpment is continued some miles farther in the same line by the Ibiquas beds. The sandstone is only some three feet thick at its termina- tion, but gradually increases in thickness southwards, so that at about thirty miles south of its northern limit possibly the whole 5,000 feet, the average thickness of the Table Mountain series, may be present. East of the Olifant's Eiver lies the great anticline of the Ceder- bergen, which trends nearly north-west in its northern portion, but turns nearly north and south at the Trigo- nometrical Station (6,336 feet above the sea) ; in the same neighbourhood the syncline of the Cold Bokkeveld separates the main anticline from that of the Schurfte- berg ^ of which the axis diverges in a south-south-east direction and is inclined southwards, so that the anticline disappears near the Houd den Bek's Biver. The main Cederberg anticline is continued in the Cold -Bokkeveld Mountains and the southern Schurftebergen. From the Schurftebergen the anticline passes round the warm Bokkeveld into the Hex River Range, closely backed by the Olifant's River syncline, so that the Table Mountain series in the block of mountains traversed by Mitchell's Pass is bent into an S-shaped fold (see Fig. 12). This fold becomes wider in the Hex River Mountains, the

^ There are two ranges called Schurftebergen (Bough Mountains) in that part of the Colony. The one here referred to is the more northern range; the other flanks the Warm Bokkeveld on the west and is -the direct continuation of the Cederberg anticline.

7*

iili

^,{iiiW%'S'Bi«tiit^ilwiw|t of that range, -^ •*• B9it«iniRl<]iQe is occupied Ikevetd beds of er Valley. This Wlmcture is re- >ilj*he Keerom and f«?tf?^ergen, the anti- '^*Vm§U\ Qocth fotming 1 Ri'ulXM|&:ffifJoom Berg and J 'S^^Sft'^l^ the Bokkeveld I ]0|^w i^i Coo and the I W'il'fl'fl^ Boathem limb ^^,Q-Jr^'Si€i^ftt!line, rather a ||-S^4Sbl|J|iBH belt than a

•|^tn^H^4n|^, fortuB the com-

iSbeibe^E^ of the Lange-

l||c^^iy^t^>which we shall

d£.=S^«^^i§iintly.

SSii^^^ the Winterhoek

— '•'-"■'-'■"'" limit of the

jC:^^^:<|^ticlines of the

^•K*lSe[^!ll!^iTer area, the

|^i||i^^«^|^*the Klein Berg

^I^^S^^lBivers have been

^vfijci^r^ough the Table

^^C^orBlI^Jseries, and are

^a^'^*ii'i'|i? hy the Pre-Cape

:|&«^ii^c|p'ating the two

^^•^•■|5«^lt**'iious ridges of

^&V^-:^)ill! Zand - Drakens-

THE CAPE SYSTEM 101

tein, and the Witzenberg - Mostert's Hoek Eanges. The former or western one is a simple ridge in its northern part, lying on the Malmesbury beds which are exposed on either side, but south of Slang Hoek its character changes; it widens out considerably, the strike of the sandstones changes and turns eastward and the dip becomes northerly ; the Bokkeveld beds are first met with near Dasbosch Kiver, where the strike of the Table Mountain series again turns through an angle greater than a right angle, and runs south-west to the Bier River Mountains near Villiers Dorp, where a narrow south-west synchne, in which the Bokkeveld beds still remain, separates the mass from the easterly trending range of the Donkerhoek, Boschveld, and Zonder Einde Mountains.

The great block of mountainous country between Rawsonville and Cape Hangklip contains two irregularly shaped depressed areas, in which lie the Bokkeveld beds of the upper part of the Zonder Einde River, and those of the Houwhoek and Palmiet River district. The Groenland and Houwhoek Mountains have a north- west trend, and separate the two depressions. The country between Rawsonville and Cape Hangklip was, as it were, the hottest part of the battle-field where the north-south and east-west fold-producing forces met, and the resulting ridges and depressions trend north- west or north-east. The Boschveld, Groenland, and Houwhoek Mountains are the chief ridges of the north- west group, and the Zonder Einde and Houwhoek- Palmiet River Bokkeveld areas the corresponding depressioQS. Tb© nprth-east group of ridges are tb^

102 GEOLOGY OF OAPE COLONY

Dwars Berg-Bier Eiver, and the Donkerhoek-Paarde Berg ranges, while the corresponding synclines are those of the Villiersdorp and Bot Eiver Valleys. The north- east folds extend eastwards as far as Lady Grey (Eobertson) and as far north as the extremity of the Hex Eiver Eange.

The Zonder Einde Eange, complicated by the north- east folds of the Lady Grey area, is an irregular anticline, and the beds in the northern limb dip down and come up against the Malmesbury beds along the great Worcester fault ; to the south of the range the Zwartberg, better known as the Caledon Mountain, is the only conspicuous, anticline that lies in the wide, synclinal area between it and the less disturbed Table Mountain sandstone ranges that stretch Irom Babylon's Tower to Bredasdorp.

The Worcester fault, with a maximum throw of more than 10,000 feet extends at least seventy miles towards the east, and plays the part of the southern limb of the complex anticline of the Langebergen. The Lange- bergen anticlines, although the mountains are known by other names, such as the Attaquas, Outiniquas, Long Kloof, Zitzikamma, and Eareedouws Mountains in their eastern portions, reach the sea over 300 miles from their commencement at Hex Eiver. At many parts of the Langebergen the beds are overturned, so that the sand- stones are overlain by older rocks on the south side, and underlain by newer beds on the north flank. The structure of the range is shown in the sections Figs. 1 and 13.

To the north of the western part of the Langebergen tjie Table Mountain series disappears under the Bokk§-

icii>JS«?(s§'llj

11 I

= 1 t

■Is

g.s

3°

3 3 t I

104 GEOLOGY OF CAPE COLONY

pear in Anysberg, and are continued in the Zwartberg range 160 miles before the axis of the fold gradually sinks below the Bokkeveld beds near the Zuurberg Poort. About twenty miles west of Ladismith village, the Amalienstein fault is first met with, throwing down the Bokkeveld beds on the south against the Table Mountain series ; the throw increases eastwards, so that near Amalienstein the Bokkeveld beds are in contact with the Cango series. This fault is in many respects like the Worcester fault, and replaces the southern limb of the Zwartberg anticline for a con- siderable distance — over sixty miles.

The Zwartberg anticline has at least as complex a structure as that of the Langebergen, and is also over- folded in many places (see Plate III.), especially be- tween Prince Albert and Klaarstroom ; the overfolding affects both the north and south flanks. On the north the later rocks, from the Bokkeveld to the Dwyka, dip south towards the mountains near Prince Albert (see Fig. 6), and, as was described in the account of the Cango series, the Table Mountain sandstone dips in places below the latter. Where the Gamka Biver traverses the mountains there is a synclinal fold bring- ing in the Bokkeveld beds in the middle of the range, thus dividing it into two distinct anticlinal ridges for some ten miles. The highest point on the range is the peak near Seven Weeks' Poort, 7,627 feet ; the curious tower-shaped peak called Tover Kop is some 4(K) feet lower. Near Klaarstroom the Zwartebergen decrease considerably in width on account of the northern por- tion of the range separating from the southern an4 plunging below the gokkeyel^J beds,

Ntfe|jii|j<||i|i

t^fitkUvaTol^Sta^uk^es of the Table _Jba'^dn^i4<^> toDi ^t|hSNi Ueiring'B Poort

106 GEOLOGY OF CAPE COLONY

Between the Zwartebergen and the Ontiniquas lies the great ridge called the Kammanassie Mountain, a bow-shaped anticline of sandstone with the concavity towards the north ; the east and west ends of the axis pitch in those directions. Between the Kammanassie and the Outiniquas there is a much-folded ridge of sandstone that diverges from the main range near the Montagu Pass, and extends eastwards to form the Eouga Mountains.

The Table Mountain and Bokkeveld series, of which the country between Willowmore and Knysna chiefly consists, have been intensely folded in this region, and the mountain ridges are formed by very sharp isoclines of sandstone.

East of the Willowmore and Uniondale divisions little is yet known of the distribution of the various formations, but it is probable that the Baviaan's Kloof and Kouga Eanges are continued under other names to near the mouth of the Gamtoos Kiver. Farther east and north- east of the Gamtoos River there are several large anti- clinal ridges of Table Mountain sandstone, but their exact limits and characters are not known ; the Eland's Berg and Great Winterhoek Mountains are the chief ones. It is probable that the Cape Recife sandstones are the most easterly part of the Table Mountain series on the coast in the folded belt, and the next appearance of this group near the coast is at St. John's, where it forms the great massive walls on either side of the river, called the Gates of St. John's. This block of rock, cut into two by the river, is separated by faults from the si^rrounding beds, which belong to the Dwyka

THE CAPE SYSTEM 107

and Ecca series. The St. John's sandstone lies hori- zontally. A few miles north-east of St. John's the Table Mountain sandstone is again met with lying horizontally, overlain to the north-west by the Dwyka conglomerate, and on the south-east bounded by the ocean or separated by a fault from a narrow strip of younger rocks (Ecca and Cretaceous) between it and the sea. The difference in level between the sandstone on the coast and that forming the plateau behind the coast is due to the cutting back of the lower terrace by the sea at no very remote period, and certainly not to folds or faults bringing the sandstone down near the coast.

The Table Mountain series is remarkably constant in lithological characters throughout its extent. The maximum thickness is about 5,000 feet, and of this more than 4,000 feet are sandstones or quartzites. The dif- ference between a sandstone and a quartzite is that the component grains are more loosely held together in the former than in the latter, in which the cementing material is quartz. When a sandstone is broken, the fresh face is rough and dull, owing to the fracture passing round or between the grains of sand which form the rock ; a quartzite, on the other hand, has a smoother and brighter face because the fracture passes through the component grains, which are closely joined together by the siliceous cement. It is sometimes found that a large block of sandstone long exposed to the weather becomes a quartzite near the outer surface, owing to the deposition of silica between the grains. Or the other hand, some quartzite^ becojqf^e loose ancl

108 GEOLOGY OF CAPE COLONY

crumbl]^ outside on account of the removal of the cement.

The whitish-grey colour of so much of the sandstone belonging to this series is due to weathering. At a distance of one or two feet from the outside the rock is usually blue, owing to a small quantity of iron in the state of ferrous compounds. The reddish-brown layer so often seen on the broken surface of a large block of sandstone is produced by the oxidation of the ferrous compounds and the formation of a brown hydrated sesquioxide. This is slowly removed from the outer surface, so that a narrow band of light grey or white rock lies between the brown band and the exterior. The red stains so often seen on the sand- stones are deposits of this red oxide of iron.

The sandstone has generally a very rough surface, frequently hollowed out so that it is covered with small and large projections, between which are shallow de- pressions that hold water for some time after rain. Particles of sand collect in these and give the depression a smoother surface than it otherwise would have had, by being moved about in it by strong winds. The gradual lateral growth of the hollows on steeply in- clined surfaces of sandstone may eventually give rise to a perforation, or small arch, by meeting a joint plane or a second depression formed on another surface of the rock.

The sandstone is very much jointed ; and as the processes of weathering naturally go on more easily along joint planes than elsewhere, for the loosened grains are soon rejpoved by the yaiij or wind, th^

THE CAPE SYSTEM 109

large exposed surfaces of sandstone are usually divided up by two or more sets of deep cracks, to which another group is added if the beds are so steeply inclined that the bedding planes make a high angle with the ground. Where these cracks become deeply eroded and are set at close intervals the ground is extraordinarily rough and difficult to traverse. The moderate effects of weathering along joints are familiar to every one who has been to the top of Table Mountain, where there are many curiously shaped knobs and pinnacles due to this cause combined with the unequal weathering of the surface. On the eastern slope of the Cederbergen, below Sneeuw Kop, on which a beacon of the geodetic survey stands, the surface of the hill is extremely cut up by these eroded joints. There are two main sets of joints on that slope, roughly parallel and at right angles to the strike of the beds, and a third group is sometimes developed. Weathering and erosion have gone on to such an extent that the mountain side is covered with an intricate mass of vertical walls and pinnacles of rock from five to forty feet high. Although such a fine development of joint weathering is not often met with, similar features are common on all the folded mountains made of the Table Mountain beds.

A very frequent characteristic of the sandstones of this group is the occurrence of round pebbles of white quartz up to three inches in diameter. They usually occur singly, more rarely in thin layers a few feet long and about an inch thick. The pebbles themselves are rarely more than an inch in diameter. It is rather difficult

110 GEOLOGY OF CAPE COLONY

to explain the frequence of isolated pebbles in the sandstone without recourse to some agency that lifted pebbles from the shore and dropped them in deeper waters. There are several means by which this may be done; in warm latitudes, seaweed torn from the shore and drifted out to sea must often carry out pebbles and bits of rock ; but in cold climates floating ice is a more powerful and usual agency, and may have been the cause of the presence of the pebbles in the Table Mountain sandstone.^

Conglomerates are remarkably scarce in this group, especially when it is remembered that the sandstones are frequently coarse-grained rocks. Hitherto thick conglomerates have only been noticed in the west of the area occupied by the group, at Pikenier's Kloof (Grey's Pass), Baboon Point, and a few other localities in that district; one of the most conspicuous con- stituents of the Baboon Point conglomerate is red jasper, a rock that may have come from the Griqua Town series. The majority of the pebbles are quartzitic rocks of different varieties. Granites and quartz-porphyries have been found in the small outliers of Klapmuts Hill and Joostenberg, as well as at Baboon Point, but they are not abundant. Fragments of slate, strange as it may seem, considering the nature of the underlying rocks, are rare in the sandstones and conglomerates.

In the Peninsula and Stellenbosch areas the base of

^ This was suggested to me by Mr. Dunn in a letter written after reading an account of the glacial conglomerate in this series on the Pakhuis Pass. He had not previously put forward this explanation on account of the lack of other evidence of glacial action in those times.

THE CAPE SYSTEM 111

the Table Mountain series is usually a red micaceous gritty shale. On the north face of Table Mountain this is the first rock met with at the junction with the granite or Malmesbury beds. In many parts of the Langebergen there is a thick band of shaly beds near the base of the series, but the lowest beds are usually quartzites (see Plate I.). On the Montagu Pass the shales near the bottom of the series are exposed in the road cutting, and are found to be a crumpled silky phyllite or schist, in which the silky appearance is due to the development of minute flakes of a micaceous mineral.

In the western mountains a second shale band is found about 1,000 feet below the top of the series. The shales are usually hidden by debris from the sand- stone clififs above them, and it is only on road cuttings and tracks across its outcrop that the rocks forming the shale band can be well seen. The shales are exposed on the Mitchell's Pass Boad, where they are deeply weathered into a red micaceous clay. On the Pakhuis Pass the shale band is exposed along a distance of three and a half miles at the top and on either side ; the rock is here a greenish-brown mudstone, a typical shale in places but generally too thickly bedded to be called a shale. The most interesting point about the Pakhuis section is the occurrence of pebbles up to five inches in diameter scattered irregularly through the shale and mudstone, without any tendency to form beds of conglomerate. Several of the pebbles have been found to be flattened on one or more sides and deeply striated in the manner characteristic of pebbles that

112 GEOLOGY OF CAPE COLONY

have come from a glaciated region. The flattening and striation are produced by the rubbing of the pebble, held by the ice at the bottom of the glacier, upon the floor, rocky or fragmental, over which the glacier moves. The floor and the fragments lying upon it become striated also, and may furnish striated pebbles to beds being deposited off the glaciated land. There is no other agency known by which the typical striated pebbles and boulders are given their peculiar features. The erosion caused by wind-borne sand produces quite different effects, which can be seen in several districts of the Colony. The frequent sliding of debris from a hillside over a smooth rock face may smooth and scratch the rock, but. does not make flattened and striated pebbles. The slickensides on rock on either side of a fault plane may sometimes be mistaken for a glaciated floor, and the evidence for regarding any given striated surface as due to glaciation must be clear and free from suspicion in this respect ; but rock movements cannot give rise to the flattened and well- scratched pebbles that are embedded in a fine-grained mudstone at moderate distances from one another. There are several conglomerates in Cape Colony that have suffered great deformation by earth movements, such as those of the Matsap and Cango series, but their contained pebbles and boulders, although often pulled out of shape and fractured, have never been found to have the characteristics of glaciated pebbles. In the conglomerates at the base of the Uitenhage series, which have at places been considerably disturbed, there are found fractured and indented pebbles, due

THE CAPE SYSTEM 113

to the crushing, or gradual deformation, of one upon another,^ but much searching has failed to discover one that could be mistaken for a glaciated fragment.

The occurrence of flattened and striated pebbles scattered at intervals through a fine-grained laminated rock is very strong evidence that glacial conditions prevailed on the land whence the pebbles came, and that these pebbles were carried away from the land by floating ice and dropped by the melting of the ice on to the mud being deposited at the bottom of the water.

The junction of the shale band on Pakhuis with the underlying sandstones is not seen, but there is no reason to suppose that there is an unconformity at its base.^

The materials of which the pebbles are made include granite, amygdaloidal lavas, quartzites, grits, jasper and vein-quartz. The vein-quartz pebbles are often smooth and almost spherical in shape, like the isolated quartz pebbles in the sandstones and quartzites both above and below the shale band.

The sandstones and quartzites are usually false bedded, and in any natural section of a considerable height examples of false bedding can be found.

No traces of fossils have yet been found in the Table Mountain series, although some of the shales appear to be favourable rocks for the preservation of organic re- mains. It must be remembered, too, that these rocks,

1 Schwarz (03), p. 398 and PL V., Fig. 1.

^ Fuller descriptions of this interesting evidence of glacial auction in the Table Mountain series have been published in Ann. Rep, Geol. Comm, (00), p. 79, and Rogers (03).

8

114 GEOLOGY OF CAPE COLONY

as well as several other formations in the Colony, have not been properly searched for fossils. Any one who thinks of the amount of work done in the north of Devonshire, for example, by two generations of geo- logists before the Morte slates were found to be fos- siliferous, will not be surprised at the apparent absence of organic remains from some of the colonial beds.

The question of the conditions under which the Table Mountain series was deposited has not yet been satis- factorily solved. The rocks are, with the exception of the shale bands, essentially coarse-grained deposits, yet this character is maintained over very wide areas ; from the Peninsula to Algoa Bay, nearly 430 miles in a straight line, and from Cape Point to the north end of the Bokkeveld Mountain, a distance of over 225 miles, the same coarse sandstone with isolated quartz pebbles is met with ; in Pondoland again, 290 miles from Algoa Bay, the sandstone is of identical character with that of the western area, and maintains its character, at least, as far as the Natal border. North of Agulhas the Table Mountain sandstone is seen at intervals for about 100 miles. It is clear, then, that the coarse sandstones that make up the bulk of the series were deposited over an area of at least 43,000 square miles, probably over more than 90,000 square miles, and even then the Pondoland outcrops have been left out of ac- count owing to the uncertainty of the nature of the rock between these and Algoa Bay.

During the denudation of the land that furnished this great bulk of sand, mostly quartz sand, an equal or greater amount of finer-grained material, muddy matter.

THE CAPE SYSTEM 115

must have been produced, but of these fine-grained sedi- ments the only traces in Cape Colony are the shale bands interbedded with the sandstones. The shales belong to definite horizons, or, in other words, were de- posited during a certain part of the period instead of the coarse sand which lies above and below them, but within the area of observation the coarse deposits do not pass laterally into the fine-grained ones. In any wide area of deposition such as that with which we are dealing, it is usual to find a considerable change in the nature of the material deposited, except in the case of oceanic deposits, the organic oozes and red clays which are formed far from land and under circumstances that vary but slightly over immense regions. The sandstones with which we are dealing, however, must have been formed near land, possibly to some extent on the land.

The absence of fossils throughout the series is a signi- ficant fact, although much weight must not be laid upon it until the shales have been better searched than they have been up to the present time.

In some desert regions great thicknesses of sandy

material are accumulated over large areas by the wind

and occasional heavy rains carrying down the debris of

the surrounding mountains and hills into plains that

have become waterless through change of climate. The

rivers that once drained the plains and took away the

sand and mud from the hills, cease to run, and the

occasional heavy downpours are not sufficient to supply

the rivers regularly, but tend to choke up the former

channels and to distribute the gravel, sand and mud

more evenly over the low ground on which temporary

8*

116 GEOLOGY OF CAPE COLONY

lakes are formed during heavy rain. In desert deposits many of the phenomena produced by ordinary deposition under water are noticed, such as false bedding and the alternation of fine and coarse beds, but there are also certain features that are not usually found in ordinary deposits, such as intercalations of layers of soluble salts deposited on the evaporation of the w^ater containing them, the very rounded, almost spherical, form of many of the sand grains, the scarcity of fossils and the absence of marine forms amongst those that do occur, and the presence of sand-etched stones.^

It cannot be said that the Table Mountain series con- tains much evidence of having been formed under desert conditions, although the fact of there being such a great thickness of unfossiliferous sandstone points in that direction.

If the Table Mountain sandstone is regarded as an ordinary coarse deposit formed in either a fresh water basin or the sea, the land from which the material was washed cannot have lain far from the present outcrops of the rock. The only evidence of the closer proximity to land of one part of the sandstone than another is the greater development of conglomerates on the west, in the Piquetberg Division and the OIifant*s Eiver Moun- tains, than elsewhere. There is no such evidence known from the Bokkeveld Mountain, or along the Zwarte- bergen, or the south coast. At present, then, we must

^ The subject of desert conditions in relation to the formation of de- posits is one that has by no means been exhausted by geologists. It is only in recent years that much attention has been paid to it. The best source of information is Professor Walther's book Das Oesetz der WUsten- bildungen, Berlin, 1900, which is also very well illustrated.

THE CAtE SYSTEM 117

conclude that while the nature of the rock renders it probable that the Table Mountain series, so far as ex- posed in the Colony, was formed not far from land, and that consequently the land lay more Or less parallel to the present distribution of the series, the only definite clue to the position of any part of that land is to be found in the conglomerates of the west.

The Table Mountain series furnishes good rough building stone in many places, such as the Cape Peninsula, Hottentot's Holland, and Green Eiver (Nieu- woudtville), where it has not been greatly disturbed by earth-movements. Owing to the quantity of unsuitable stone that has to be removed in quarrying the best beds of rock, it is not used so much as one might expect from the wide distribution of the sandstone. The stone is not easily worked, and is mostly used for foundations. In Cape Town the Huguenot Memorial is partly made of Table Mountain sandstone ; and the new Harbour Board offices are built of the sandstone from a quarry at Grabouw beyond Sir Lowry's Pass. The sandstone from the latter place is more regularly laminated than is usually the case, and good-sized blocks can be ob- tained without much difficulty.

Irregular pockets and fissures in the sandstone are sometimes filled with pjrrolusite, an ore of manganese, but the mineral has not been successfully worked yet. The fissures are usually along fault planes. Some old workings can be seen at the head of Du Toit's Kloof near the Paarl.

Gold has been found in small quantities at many places in the Table Mountain series, but except at Mill-

118 GEOLOGY Of OAtfi COLoNY

wood (Knysna), it has never attracted mnch attentioh. The gold hitherto obtained at Millwood is alluvial) pro- bably derived originally from veins in the Outiniquas Mountains and the country south of them. There is still some doubt as to whether the bed rock at Millwood belongs to the Table Mountain series or to an older group. If the latter proves to be the case, the Millwood beds may belong to the same group that the galena and blende occur in at Maitland Mines, Port Elizabeth.

The Table Mountain series 3aelds a poor, sandy soil, which in spots continually kept damp is black, owing to the presence. of organic matter. Vegetation is abundant where the rainfall is heavy ; a heavier rainfall is re- corded on or near the mountains of the south and west than on the low ground on the coast side or on the inland flank. The most characteristic plants seen on this formation belong to the orders Proteacea, Ericacea and BestionacecB, respectively the sugar-bush tribe, heaths and flowering rushes. The change in the character of the vegetation on passing from the Table Mountain series to another formation is usually very sharply defined. From the Bokkeveld Mountains right round the great sandstone mountains of the folded belt, the same, or similar shrubs and flowers are found. A most striking contrast to any one who is even slightly ac- quainted with the vegetation of the western mountains is seen on passing from the Karroo formation in Pondo- land to the strip of country near the coast formed by the Table Mountain sandstone ; leaving the monotonous grass veld of the interior of Pondoland one meets with the same flowers and small shrubs that are abundantly

THE CAPE SYSTEM 119

found on the western mountains. It is difiScult to under- stand how such a distant outlier can be clothed with the same vegetation as the main area by a process of colonisation and selection by the soil; probably the plants of the Pondoland coastal plateau arrived there when the sandstone was still connected with the western ranges by the more or less rectangular strip, correspond- ing to the bent ranges round the Warm Bokkeveld, that may still exist ofif the south-east coast between the Gualan^i and St. John's Eivers.

Owing to difficulty of access by road and the general poverty of the soil, there are few farms under cultivation on the sandstone areas. The mountain veld is mostly used for grazing. Very rarely one finds a farm, such as Mou ton's Valley on Pique tberg, where many kinds of fruit are grown, wine and tobacco made, and fine plan- tations of oaks laid out on ground that was no better originally than that on hundreds of other mountain farms which are merely grazing veld.

From the old accounts of the Colony it is clear that the mountains of the south were once fairly well covered with forest, now represented by a few isolated patches, as at Groot Vader's Bosch near Swellendam. In the neighbourhood of the Peninsula and Stellenbosch, the oldest settlements in the Colony, the too free cutting down of the timber has been the cause of the almost complete disappearance of the indigenous forest, but farther north and east the chief cause of destruction has been the veld fires lighted for the purpose of allowing young grass and bush to spring up afresh for cattle to graze upon. There can be no doubt that the

120 GEOLOGY OF CAPE COLONY

hindrance of the forest growth is a great evil, except perhaps to the farmers whose cattle graze on some of the mountains. There is a well-supported belief that forest-clad hills receive a heavier rainfall than the same hills deprived of their trees ; but the destruction of forest and bush has a much wider effect than this. Living vegetation and the accumulation of dead twigs and leaves hinder the rapid dispersal of rain water and bind the sandy soil, thus causing a more gradual delivery of the water into the streams, and at the same time allow- ing a greater proportion of it to sink into the ground than is the case in a deforested region. The rivers fed by the mountain streams, therefore, rise less suddenly and maintain their supply of water for a longer period ; and the springs which get their water from the moun- tains are stronger and more constant.

The Cape Government is doing something in the direction of reforesting some of their mountains, but these efforts could be multiplied many times with very great advantage to future generations, even without taking into account the value of the timber, a consider- able asset in a few years after a plantation is made. The only real difficulty in the way of maintaining exten- sive plantations is the reckless burning of the mountain veld, but in that matter a strong current of opinion seems to be setting in the right direction amongst farmers, especially in the Eastern Province, and if that opinion grows and becomes general throughout the districts concerned, there will be very little danger from fire.

The Knysna forest is chiefly on Table Mountain

THE CAPE SYSTEM 121

sandstone, and far to the north-east the St. John's and Egossa forests are on the same formation. Elsewhere the forests are mere remnants preserved in steep kloofs, and they do not spread over large parts of the moun- tain sides.

The Bokkbveld Sebies.

The Bokkeveld series is everywhere found lying directly upon the Table Mountain series, with similar strike and dip, and there are no signs of unconformity between the two. In some localities, such as the small sandstone anticlines in the Warm Bokkeveld and the anticlinal ridge of Jan Niemand's Bosch near Houwhoek, water seems to have percolated freely at the junction of the two formations, the position of which is marked by a layer of crystalline quartz. There are few places where a clean-cut section of the junction can be seen, for the soft beds of the bottom of the Bokkeveld group have generally been worn away by small streams, the beds of which are choked up by debris from the sand- stones when the strata are at all steeply inclined. Where the beds lie nearly flat, as they do north of the Doom Eiver in the Western Karroo, the junction is hidden under the soil. The best section hitherto found is that on the left bank of the Gamka Eiver immediately above its great Poort through the Zwartebergen, and there **the end of the white sandstones and the be- ginning of the blue-black shales of the Bokkeveld is so sudden and exact that one can place a knife between them and say confidently that on one side are the rocks of the Table Mountain series and on the other those of

122 GEOLOGY OF CAPE COLONY

the Bokkeveld *'.^ Other clean-cut sections through the junction may be seen lower down the Gamka (Gouritz) Eiver in the Pogha Hills and near the new road to Cloete's Pass and at the north end of Meiring's Poort,

The Bokkeveld beds are well exposed in the Cold and Warm Bokkevelds, in the Hex Eiver Valley especially between De Dooms and Klein Straat stations, and along the northern flank of the Zwartebergen. They occupy wide areas in the Ladismith Karroo and south of the Langebergen ; but south of the Zwartebergen they have been greatly changed by the movements which gave rise to those mountains, and are much cleaved. They have only been found within the folded belt south and west of the Karroo. No outliers have been met with in the Pre-Cape region of the west and north, and in Pondoland they have been removed by denudation, if they were ever deposited there. There can be little doubt that they once overlay the sandstone of Table Mountain, although the nearest outcrop is at Grabouw, east of Hottentot's Holland, about thirty-six miles in a straight line from Table Mountain.

Where typically developed the Bokkeveld beds consist of shales and sandstone arranged in a definite order, although the details vary from one locality to another. The lowest division consists of shales and thin sand- stones about 300 feet thick and contains many fossils, amongst which trilobites belonging to the genera Phacops and Homalonotus ; brachiopods of the genera LeptocoBlia, Spirifer, Chonetes and Orthothetes ; Orthoceras, Bellerophon,

^Schwarz, Oeol. Comm. (98), p. 36. A detailed measured action through the Bokkeveld beds will be found in that Report.

THE CAtE SVSl:EM 123

Nuculites and crinoids. The shales often contain spherical or elliptical nodules, which are partly filled with red or yellow ochre, sometimes used for making paints with the addition of oil. Another variety of nodule found in the shales is dark coloured jnside, and often contains rather well-preserved fossils.

Some beds of the lowest shale group are coloured black by the amount of carbonaceous matter in them, and in places where the rocks have been intensely crushed these beds are represented by graphitic slate or schist, as on the north of the Pot Berg anticline near Port Beaufort and near Bredasdorp.

This subdivision usually forms a slope below a cliflf or very steep rocky ground formed by the second division, the first or fossiliferous sandstone. The fossiliferous sandstone is a dark-blue rock weathering deep red outside ; at some places the sandstone con- tains many fossils, especially Spirifer and Leptoccslia^ but at other localities the sandstone is not nearly so fos- siliferous. The beds of red-weathering sandstone are separated by blue shales very like those below and above this subdivision. The thickness of the fossil- iferous sandstone reaches 150 feet. This rock can be seen north of the village of Ceres especially on the road up the Gydo Pass, where many fossils have been obtained from it. It is very often seen as an escarp- ment, the steep face of which is directed towards the Table Mountain sandstone. Such an escarpment oc- curs for a long distance, over fifty miles, on the east side of the Cederbergen, where, owing to the steep but constant dip of the beds south of Wupperthal, the

- - "■ •• ^ - ■*■■*•■*•"'

'••"Eg- ■ - -:k

.s»s'-T-

1^ :iE: se ^ s: ;$; ^; :!s;

126 GEOLOGY OF CAPE COLONY

« whole of the Bokkeveld series is exposed within a

short distance. In the view shown in Plate IV., taken on the west side of the Schurfteberg (north) anticline (Cold Bokkeveld), looking south, the escarpment of the fossiliferous sandstone is seen on the right of the road as a low ridge, and also on the horizon. The top of the Table Mountain series is seen on the left of the picture as a long slope with one slight protuberance ; the lowest part of the ridge, at a spot above which some more distant hills appear, is formed by the low- est shales of the Bokkeveld, that also occupy the flat valley in which the road lies; the higher groups of sandstone beds in the Bokkeveld series make ridges on the horizon, but the fourth sandstone is very slightly marked ; the high mountain on the right is the outlier of Witteberg beds named Tafel Berg. Plate V., taken at Eiet Eiver in the Cold Bokkeveld, illustrates the succession on the east side of the Cederberg anticline ; in the foreground is the Table Mountain sandstone dipping east under the Bokkeveld of the high hills (Blink Berg) in the middle of the picture, which are capped by the Witteberg beds. The top of these hills is about 2,000 feet above the bottom of the valley. The four groups of sandstone in the Bokkeveld series appear as kranzes on the face of Blink Berg, and the three lower ones are well seen on the sky-line. The position of the shales below the fossiliferous sandstone is almost invariably marked by a valley along which a road runs. This is the case along the Cederbergen and Cold Bokkeveld Mountains, in the Hex Eiver Valley, in the country north of the Zwartebergen, and

^

H^^ff*

li it

Is ll

1*1

in

ill

SI

J^f^ifJ^Wt^mtj^Ztj^iJj^ 'f- -f- " '*' ^l ^l '^'

»:|!i's'i'.f?ii^("'Y

II

is

Si

■i^iK^S.^'4*sjt?-*«;:si-*.

THE CAPE SYSTEM 129

in much of the country between the Hex Eiver Valley and the Gouritz Eiver Poort. The fossils in the sand- stone are usually in the forin of impressions left by the removal of the calcareous shells. The shells themselves are rarely seen in the rock taken from near the surface of an outcrop, but when the rock from a distance of some feet from a weathered surface is obtained, the calcite shells are often seen in it. The sandstone itself is slightly calcareous, but beds of limestone are of very rare occurrence.^

Above the fossiliferous sandstone is the second group of shales containing fossils, from 100 to 300 feet thick. In the Cold Bokkeveld area the second group of shales is distinguished by the presence of star-fish, but many of the species that occur in the lower group are found here also. Above them is the second sandstone, which weathers into light-coloured outcrops, differing strongly in this respect from the first or fossiliferous sandstone ; it contains few fossils ; Spirifer is occasion- ally abundant. The second sandstone is a thick group with many shale beds, and in the Gamka Poort section reaches a thickness of 400 feet.

The third group of shales is about 350 feet thick, the beds are often micaceous, and have thin quartzites interbedded with them ; they usually contain few fossils, Ntcculites occurs in them at the Gamka Poort. Near the Tunnel Siding on the Hex Eiver line this group

^A bed of limestone was found in the Bokkeveld series in the excavatior of a tunnel in the Hex Biver Valley. See Prosser (79), p. 49. In the Clanwilliam district a nodular lump of limestone crowded with rolled up Trilobites (Phacops and Homalonotus) has been found above the fossiliferous sandstone at Fredericks Dal.

9

130 QfiOLOGY OF CAPE COLONY

of shales yielded Lingula, Nuculites, crinoid stems, a trilobite and GontUaria, and also some badly presented plant stems resembling Lepidodendron. The third sand- stone group (100 feet) with the shales above (300 feet), as well as the fourth sandstone (100 feet) and the over- lying shales (500 feet), have not been found to contain fossils other than badly preserved plant remains. These are not so well defined as the lower groups, and both the shales and sandstones are often very micaceous. The fourth shale group is taken as the uppermost of the Bokkeveld series, and the beds in it often closely resemble those belonging to the Witteberg. The di- vision between these two series is an arbitrary one, and cannot be laid down with certainty in the absence of a clearly exposed succession from below. In the country north of the Zwartebergen, in the Cold Bokkeveld, and in the Hex Eiver-Ladismith Karroo district, there is not much difficulty in fixing upon a boundary which is probably at one and the same horizon throughout; but south of the Langebergen the task is an impossible one, and the limits of the Witteberg beds there as laid down upon the map must be considered as only roughly correct.

Along the northern slope of the Langebergen the Bokkeveld beds are very much cleaved; the cleavage planes have a constant and high inclination to the south, while the dip of the beds is very variable in amount, and in direction is either nearly north or south, the strike of the beds being nearly east and west, par- allel to the cleavage. There is usually no difficulty in distinguishing between the bedding planes and^ cleavage

THE CAPE SYSTEM 131

in this district, for the sandy portions of the rock resist the weather better than the finer grained beds, and stand out more or less prominently on the hill sides. South of the Langebergen, however, especially eaSt and south of the Eobertson Division, the distinction between the two sets of divisional planes is much less marked, partly owing to the strong development of the cleavage, but partly on account of the more uniformly fine-grained nature of the rocks. Few fossils have been found in the Bokkeveld beds south of the Langebergen, probably because those contained in the slates are so much dis> torted by pressure that they are not easily recognisable. In the small synclines of these beds, folded in amongst the Table Mountain series in the Knysna Division, several genera have been obtained ; on the Keurboom*s Biver Orthoceras, PhacopSf Orbiculoideay Leptocalia, Cho- neteSy Spirifer, Ntumlites, Bellerophon, Tentaculites and crinoids have been found. Farther west a few charac- teristic species have been obtained from the Bredasdorp, Caledon and Worcester Divisions, but they are usually greatly distorted.

The distinctly finer grained nature of the Bokkeveld beds south of the Langebergen than to the north of those mountains points to the position of the shore-line of the sea in which they were deposited having crossed South Africa in a general east and west direction to the north of the area now occupied by them. It is not possible to determine the position more closely, for the northern limit of the beds is only seen in the west of Calvinia, and is there an eroded surface of great age;

the denudation which swept away the in-shore portion

9*

132 GEOLOGY OF CAPE COLONY

of the Bokkeveld beds took place in the Pre-Dwyka times, and the greater part of the northern limit is still buried beneath the Karroo formation between the Oorlog's Kloof River west of Calvinia and the sub- merged south-eastern portion of the folded belt ofif the south-east coast.

The marine fossils that occur in the lower half of the Bokkeveld series afiford sufficient evidence that the rocks in which they are imbedded were deposited under the sea; and the frequent occurrence of false-bedding in the sandstones throughout the series points to deposition in shallow water. The bottom of the sea must have been slowly sinking to allow such an^ accumulation of shallow water sediments, although some of the shales may have been formed in deeper water. In the upper part of the Bokkeveld series no marine forms have been noticed ; a few indistinct plants are the only fossils that have been found in them. It is difficult to explain the absence of marine animals if the conditions under which these rocks were formed remained the same as before ; and the absence of marine fossils from the succeeding 2,500 feet of the Witteberg sandstones and shales war- rants the supposition that the conditions which prevailed in the area now called Cape Colony during early Bokke- veld times changed from marine to fluviatile or lacus- trine after the deposition of the third shale group, and remained so throughout the later Bokkeveld and the whole of the Witteberg periods.

The following are the chief fossils from the Bokkeveld beds hitherto described : —

THE CAPE SYSTEM

133

	Falkland Islands.	South America.	North America.	i
WoEM Tube—
SerpulUes sica, Salter . . - -
Ortnoids —
Ophiocrinus stangen, Salter
Lamellibranghs—
PcUoioneilo antiqua, Sharpe
„ subantiqua, E^ed
„ rudisy Sharpe
„ aflf. comtrictay Conrad			*
„ of . fecunday Hall			■X-
Leda inornatay Sharpe . . - -
Grammysia corrugatay Sharpe -
Anodontopsis ^ inidisy Sharpe -
Ortkonotay aflf. undtiliita, Conrad			*
Sanguinolites, sp. - - - - -
Glossitesy aflf. depressusy Hall			*
Cardiomorphaj sp.
Prcecardium'i sp.
Nuculites abbreviatusy Sharpe -
„ africantbSy Salter
,y branneri, Clarke _ - .		*
„ capensisy Keed - - - -
Bysifopteria ? sp. - - - - -
Actinopteria, aflf. boydiy Conrad			*
Modiomorpha bdiniy Sharpe			■J'r
„ aft. pimentatMy Hai'tt and
Rathbun-		*
„ aflf. sellowiy Clarke		*
Gastbropods—
PkuroUmvariay aflf. kayseriy Ulrich		*
Bellerophon qmidrilobatuSy Salter
„ aflf. nwrganianusy Hartt and
Rathbun - - . -		* 1
„ {Bucaniella), aflf. trUobatuSy Sow.				*
^^ „ cf. reissiy Clarke -		*
„ (Plectonotiis), aflf. salteriy Clarke		*
LoxoneTna, sp.	1 1 %3
Tentaculites crotalinns, Salter -	j 1
„ bainiy Reed . - - -	1 1
Littoriria^ bainiy Sharpe - - - -		1
Theca (HyolitJies) submqaalisy Salter -		1
Conularia africa7ia, Sharpe		♦ 1
„ quichiuiy Stein Diann-Doderlein -		*
„ cf. undukUay Conrad		*
., of. acuta^ Roemer		*

134

GEOLOGY OF CAPE COLONY

	||	tth rica.	rth rica.	1
	ii	^•§	*4
Bbacuiopods —
LinguUxy sSL denm. Hall . - - -			♦
Orlnculoidea hain% Moir. and Sharpe	1-	>•<
Stropheodonta, of. conciyina, Morr. and Sharpe
Strophonella, sp. - - - - - Orthothetes suUivanij Morr. and Sharpe		?
	?
Ckonetes fcbUclandicus, Morr. and Sharpe -	*	♦
„ cf. coronaJtvis, Conrad -			♦
„ cf. arcei, Ulrich - - - -
„ aff. setigeTj Hall - - - -
Orthisy sp.
Rhynchonellaf sp.
Rensselairia, sp. a, Reed - - - -
„ sp. ft Heed - -• - -
. i,. sp.?
Trigeria gavdryty Oehlert				•,'• ■-I*
Cryptonella haint, Morr. and Sharpe -
Spirifer (yrbignyi, Morr. and Sharpe -	*
„ pedroanuSj Hartt - - - -		'I-
„ ceresj Reed
„ a, Reed
„ ^, R«ed
Tropidoleptus carinatus^ Conrad Amhocodui umbonataj Conrad - - -			'1-
			»;:
Retzia adrieni, de Vern - - - .
Rhynckospira, cf . dlveti, Ulrich
Leptocoelia flabellites, Conrad ...		♦	i!i
Vitulina pustulosa, Hall - - - -			k*# T-
Tbilobitbs—
Phttcops pupilltis, Tjake - - - - ,, arhutevSy Lake . - - -
„ crista^aUij Woodward
„ africanuSy Salter - - - -
„ ocellusy liftke ....
„ impresstiSy Lake - - - -
„ iCryphcBus) caffevy Salter
Dalmanites lunattiSy Lake - -
Proetus malacuSy Tjake - - . -
Typhloniscus hainiy Salter - - -
HomaUmotus heracheliy Murch. -
„ qtiemuSy Tjake
„ colo88U8y Tjake
Oephalopods—
Orthoeeras, two species ... -

THE CAPE SYSTEM 135

The fossils ^ common to the Bokkeveld beds and the Devonian strata of the Falkland Islands, South and North America and Europe are marked with an asterisk under the columns referring to those countries. In the case of the many species which have close aflBnity to foreign forms (aff.) or are closely comparable to them (cf,), the asterisk refers to the locality of the allied species; many of these may be determined with certainty in the future. The fauna as a whole is more nearly related to that of the Devonian rocks of other countries than to any other, although there seems to be no evidence to correlate the Cape fossiliferous beds with any one part of the Devonian system as developed in Europe or North America. Of the Brachiopods, Mr. Reed writes, "they have a completely Devonian stamp ; and there are none which suggest the presence of Silurian or Carboniferous beds '\^ Imperfect though the list of fossils given above is, it shows that the Bokkeveld fauna is much more closely related to the American Devonian fauna than to that of Europe, and more closely to the South American than to the North, in spite of the fact that the rocks of South America and the Falkland Islands are less well kn6wn than those of North America and Europe.

The country occupied by the Bokkeveld beds north of the Langebergen and in the Worcester and Robertson Divisions south of that range is characterised by strongly

^ For descriptions and figures of the fossils the student must refer to the appendices to Bain (54) by Salter and Sharpe, Woodward (73), and Beed (04). Before long the Trilobites, Lamellibranchs, Gasteropods, Pteropods and some Gephalopods wiU be described and figured in the ^nnftls Qf the Sguth African ^useum^ ^ l^eed, op, ciL^ p, 186,

136 GEOLOGY OF CAPE COLONY

marked escarpments and valleys, so that from the top of a prominent hill in a suitable position the lie of the rocks can be made out over a very wide area. The most accessible of such hills are the Brand Vley Moun- tain near Worcester, Gydo Berg north of Ceres, the high hill near Triangle, in the Hex Eiver Valley, and the top of the hill east of the north entrance to Seven Weeks' Poort. The last-named spot is one of the finest points of vantage in the Colony for the purpose of seeing the structure of a wide area. The folds into which the rocks have been thrown north of the Zwartebergen are distinctly seen, the outcrop of the four groups of sand- stone in the Bokkeveld series make independent escarp- ments or ledges on large ridges, and where repeated by folding the structure is seen clearly. The gradual djdng out of the folds northwards in the Karroo is displayed as if the country were a geological model, and the out- crops of each formation are at once recognised. The sand- stones and quartzites of the Bokkeveld and Witteberg series stand up prominently between the shale bands that have determined the positions of the minor valleys, the soft, easily eroded shales having offered an easier path for the rivers than the more resistant sandstones. The view is limited on the north by the great dolerite-crowned escarpment of the Nieuweveld, seventy miles distant.

South of the Langebergen the structure of the country is not at all obvious until it has been made out in detail, for the Bokkeveld beds have been cut to a level with the outliers of the Uitenhage series ; and although this plain has since been dissected by rivers, the Bokkeveld and Witteberg slates, on account of their uniform chp.r-

THE CAPE SYSTEM 137

acter, have had little effect in determining the positions of the valleys, so that the longitudinal valleys so con- spicuous north of the mountains are not nearly so well developed to the south.

The Bokkeveld beds do not furnish any stone or minerals of much economic value. The sandstones are used for making walls round kraals and camps, and to a small extent for house-building on farms. Their colour is too dark and patchy, and as a rule they are too fissile and difficult to work to •be used when any other building materials are obtainable.

The country occupied by this series is generally well populated, for the soil is rich. The shales break down into good soil, so the positions of the thicker bands of shale are usually marked by lands and gardens, often with a dip slope of the Table Mountain sandstone on the one hand and an escarpment of the Bokkeveld sandstones on the other.

Springs are more numerous along the junction of the Table Mountain sandstone and the Bokkeveld beds than elsewhere in the neighbourhood, and although many of the springs yield **kruit water," i.e., water with the smell of sulphuretted hydrogen, due to the mutual de- composition of pyrites and the organic matter in the shales in the presence of moisture, they are very valuable sources of water. This peculiarity of the water is the cause of so many farms being called ** Stink Fontein," a name that recurs again and again on the Bokkeveld areas as well as on other rocks, such as the Dwyka and Ecca beds, the water from which has frequently th§ same characteristic,

138 GEOLOGY OF CAPE COLONY

The Wittbberg Series.

The Witteberg series consists of sandstones, quart- zites, and shales. The sandstones and quartzites are in thicker groups than those of the Bokkeveld beds, and occasionally contain thin beds of white quartz pebbles, and also isolated pebbles of the same material. The resemblance between the Witteberg quartzites and the Table Mountain beds was the cause of much confusion in the early days of Cape Geology, but it is more apparent than real. The Witteberg quartzites, as a whole, have a more reddish and yellow tint and are more micaceous than the Table Mountain rock, and they are much less massive, shale bands being of comparatively frequent occurrence. The shales are green, dark grey and blue in colour, and they are often very micaceous and sandy, frequently being more properly called thin, irregularly bedded micaceous sandstones than shales. In the Eastern Province there are black carbonaceous shales, which are diflferent from any beds in this series that have been found in the west. The Witteberg beds have so far yielded no remains of animals, and only rather poor specimens of plants which have not been satis- factorily determined for want of good material.

The following genera of plants have been mentioned ^ as having been found in the Witteberg beds : —

1 This list except the last genus is taken from Feistmantel (89), pp. 25 and 26, where references to the original authorities may be found. I have omitted those said to occur at Tulbagh, for a mistake has evi- dently been n^ade in the locality, or it is ii^sufiic;Qntly defineclr

THE CAPE SYSTEM 139

Selaginites Port Alfred.

Lepidodendron Grahamstown, SwellendAm and Riversdale.

Lepido8trohu8 Port Alfred.

Halonica „

Knorria Swellendam.

Sigillaria Port Alfred.

Stigviaria „

Cyclostigma Many places in the west of the Colony.

Little value can be set upon the determinations in the above list, but it is of interest to note that all the genera occur in the carboniferous rocks of Europe, and the Cyclostigma is very like a fossil described by Feistmantel from the Goonoo Goonoo beds (Devonian or Carbonifer- ous) of New South Wales.

By far the most abundant fossil, if it be one, is Spirophyton, but Mr. Seward,^ who has examined some of the specimens collected by the Cape Survey, is of opinion that these markings are not of organic origin.

Spirophyton is found as an impression extending spirally through several inches of rock, with the curved striations radiating from a central depression to a pe- ripheral groove. It is difficult to understand how such a well-defined structure with a sharply marked limit passing spirally through several layers of sediment can be produced by mechanical means, such as the swirling of water through a hole in the sand. No carbonised remains of vegetable matter have been found adhering to the surface of the Spirophyton impressions, but the same is the case with the undoubted plant impressions from the Witteberg and Bokkeveld beds in the west of the Colony. There is a great area of Witteberg beds in

1 Seward (08).

140 GEOLOGY OF CAPE COLONY

the east that has hardly yet been examined for fossils, and as one of the varieties of plant impressions is there found preserved with some coaly matter adhering to the specimens some fresh evidence of the nature of Spiro- phyton may be expected in the future.

Whether a true fossil or not, Spirophyton has been found of great service in enabling the Witteberg beds to be recognised, as it is doubtful whether it occurs in the uppermost Bokkeveld beds, and it has never been found in the Dwyka or later rocks. It is met with in hard quartzites and in shales, the best specimens are those from the quartzites ; the markings are better preserved in quartzite than in the micaceous and sandy shales, although they are more abundant in the latter.

The Witteberg beds have a maximum thickness of about 2,500 feet. They form several important ranges of mountains on the southern border of the Karroo, and their name is taken from the Wittebergen, south of Matjes Fontein. In the west and south of the Colony the mountains composed of the Witteberg beds are remark- ably bare and barren-looking (see Plate VI.). They are less well supplied with rain than the Table Mountain sandstone i;anges, for the latter are generally higher and therefore receive a heavier rainfall. The high percen- tage of quartz sand in the Witteberg beds causes the soils derived from them to be poor and thin. The forma- tion is first met with in the west of the Colony, north of Eland's Vley (Calvinia and Clanwilliam), where the long line of hills called the Zwart Ruggens commences. The northern boundary is a denuded one, and, as is the case with the Bokkeveld boundary a little farther to th^

iii|tiJi;iil;

ii -^

^^-.-^•.-^•.^^ -^v.^. .^. .^. .

142 GEOLOGY OP CAPE COLONY

north, is of great antiquity, being chiefly older than the Dwyka series. Following the Witteberg beds south- wards they become thicker owing to the coming in of higher and higher beds below the Dwyka. Some out- liers, somewhat table-shaped mountains, are found at Bidouw, Gerustheid, and in the angle between the Bosch and Doom Eivers in the north-east of Clanwilliam and south-west of Calvinia. The Zwart Euggens are a long dip slope of the quartzites on the east of the Cederberg and Cold Bokkeveld anticlines. When seen from the Karroo the Zwart Euggens appear to consist entirely of whitish quartzites, for the numerous shale bands are more easily weathered away and can oijly be seen when one enters a ravine or gorge, such as the Tra-Tra or Winkelhaak's (Doom) Eiver valleys, which drain the Cold Bokkeveld. The Zwart Euggens merge into the Bonteberg Eange at Karroo Poort, when the strike of the rocks changes from south to east. The axis, of the Bonteberg anticline is inclined eastwards, so that the Dwyka series sends a tongue west-south-west towards Pienaar's Kloof north of Touw's Eiver Station. The Witteberg beds are continued across Pienaar's Kloof into the Voetpad Berg, and also round the south of the Quarrie Kloof Dwyka outlier into the Wittebergen. In the southern part of the Worcester Division the Witte- berg beds form a V-shaped area ; the two arms of the V meet on the south and are cut off by the Worcester fault to the north, but the western junction is buried beneath the conglomerates of the Uitenhage series ; the apex of the V is at Eoode Berg near the road between Villiersdorp and Worcester. In Eobertson the Witte-

THE CAPE SYSTEM 143

berg beds form an area about twenty-four miles in length, south of the fault ; and they also occur in Swellendam and Kiversdale. To the north-east of Montague they form two synclines connected at the eastern end ; Klein Berg is part of the southern syncline, and the hills near Dobbel Aars Kloof belong to the northern one.

Between the Bonteberg and Matjes Kop these beds cover a considerable area, over forty miles long and twenty wide in places, being thrown into many small folds, and in four of the synclines or troughs outliers of the Dwyka series occur; the Nauga and Coega (or Kouga) hills are in this area. The axis of the main anticline of the Wittebergen disappears eastwards south of Laingsburg, where a long syncline of the Dwyka series lies south of the eastern part of the range. The Witteberg beds pass round the western end of the Dwyka syncline into Eland's Berg, which disappears eastwards in a similar manner to the Wittebergen, but the beds pass round another westerly rising Dwyka syncline into the long range of foot hills north of the Zwartebergen, and extend far to the east, certainly as far as Willow- more; they reappear from under a syncline of the Dwyka series in the Groot Eiver and Klein Winterhoek ranges to the north. East of the Klein Winterhoek Mountains the Witteberg beds form the Zuurbergen, the hills near Commadagga, Botha's hill and the hills south of Grahamstown, and much of the country be- tween Grahamstown and the coast.

The Witteberg country in the Eastern Province is much better covered with vegetation than that in the west, chiefly on account of the greater rainfall, but pos-

144 GEOLOGY OF CAPE COLONY

sibly the eastern rocks are somewhat more argillaceous and less quartzitic than the western, and therefore give rise to better soils. Whether the Witteberg series as a whole becomes finer grained towards the east is still uncertain, for it has not been closely examined in that part of the country.

In the south of the Colony east of Robertson the Witteberg beds are distinctly less quartzose and coarse grained than to the north of the Langebergen ; a similar change takes place in them to that noted in the case of the Bokkeveld series, as they are followed southwards. It has been stated previously that the absence of marine fossils, or rather the remains of animals that are evi- dently related to forms which only live in the sea, from the Witteberg beds must be regarded as evidence that these sediments were not laid down under the sea, but they may have been formed in fresh water. The settle- ment of this question must always be a diflScult task, and the rocks must be known in much greater detail than they are at present before it can be accomplished. False bedding and rippled surfaces are frequently seen in these rocks, which were certainly laid down in shallow water not far from the land.

There can be little doubt that the Witteberg beds once extended over the whole of the southern and western portion of the Colony. The position of the coastline of the land from which the sediments' were derived is as problematical as the position of the Bokkeveld coast line. From the fact that the coarse sediments are found in the northern exposures, it must be concluded that the land lay in that direction,

THE CAPE SYSTEM 145

and it probably lay rather farther south than the Bokkeveld shore.

The Witteberg beds have no economic importance. Many years ago a nugget of gold was found in these rocks at Kragga Poort, near Constable, but nothing further has been found there. The presence of black coaly shales in the Witteberg series on the Kawie Eiver led to prospecting for coal some forty years ago, but without success. A great part of the country occupied by this series is very rugged, owing to the quartzite bands standing out prominently from the general surface. The white quartzites often give rise to great bare stony dip slopes, such as those on the eastern side of the Zwart Euggens west of the Karroo and in the mountains south of Matjes Fontein.

10

CHAPTEE V.

THE KARROO SYSTEM.

The beds belonging to the Karroo system cover the greater part of the Colony ; from a line between Karroo Poort and the Gualana Eiver mouth northwards to the Orange Eiver east of Prieska these are practically the only rocks exposed at the surface, with the exception of the intrusive dolerites. Somewhat monotonous from the repeated occurrence of sandstones, shales, and mud- stones, in all thousands of feet thick, and from the fact that they generally lie at so low an angle that in the absence of considerable changes of level in the surface a comparatively thin group of beds occupies a very wide area, nevertheless they are of great interest from some points of view. Perhaps their chief interest con- sists in the reptilian remains preserved in them, and in the similarity of their fossil plants to those found in the Gondwana system of India, in certain Australian rocks, and in beds in some other parts of the world.

There is at present no very satisfactory classification of the formation, but when its fossils have been more extensively collected with due record of localities, the present subdivision will be strengthened or sufiicient grounds brought forward for a somewhat different one.

At present the system is subdivided as follows, in descending order : —

146

THE KARROO SYSTEM

147

Karroo System"

r Volcanic group Stormberg I Cave sandstone

series

Beaufort series

Ecca series

Dwyka

series

I Red beds

Molteno beds

Upper - - -

Middle .Lower - - * -

Upper beds -

Laingsburg beds - . Lower beds -

Upper shales

Conglomerate .Lower shales

Approximate maximum thickness.

. 4,000^

: : : iZr^"^

- 2,000^

5,000

2.600

600] 1,000 y 2,300 700 J

18,100

The maximum thickness of the Karroo formation is not less than some 14,000 feet, excluding the volcanic beds, although it is of course not certain that the full thickness is now, or ever was, developed in any one locality. This great bulk of sedimentary rocks nowhere contains evidence of marine conditions having prevailed during its deposition; on the contrary, nearly all the fossils known from the Karroo beds were undoubtedly either land or fresh- water forms. The accumulation of so great a thickness of fresh-water beds is a very interesting fact, and we shall return to the subject after describing the various groups of rock in the system.

The Dwyka Series.

Everywhere round the borders of the central basin

a conglomerate with very peculiar characters crops out.

It is usually a blue or greenish rock, compact and fine

grained, made up of small particles of sand, which under

the microscope are seen to be chiefly composed of quartz

and microline, with a smaller quantity of other felspars,

10*

148 GEOLOGY OF CAPE COLONY

epidote, garnet, calcite and other minerals imbedded in mud, using that term for an argillaceous material too fine grained to be more definitely named. This sandy mud contains a vast number of boulders and pebbles of a great variety of rocks, amongst which are conglomerates, quartzites, sandstones, shales, slates, marbles, jaspers, granites, gneisses, diabases, amygdaloidal lavas and serpentines.

These boulders are, as a rule, scattered irregularly through the conglomerate without any arrangement in beds. Plate VII., a photograph of the conglomerate exposed in a ravine near Prieska, gives a good idea of the manner in which the pebbles and boulders occur.

Not only is the great variety in the boulders remark- able, but the shape of a large proportion of them is peculiar. When a rock is broken up by natural causes the fragments are at first angular, their shape and size depending upon the nature of the rock and other con- ditions ; when these angular fragments are rolled along by a stream, or thrown and dragged about on a shore, the corners are worn off, and the boulders become rounded or oval in shape according to the original form of the pieces of rock. Whilst there are many boulders of this description in the Dwyka conglomer- ate, there 'are others distinctly flattened on one or more sides, with scratches of various depths on the flattened surfaces and to a smaller extent on the other parts. The striations in some cases run in one direction only across a flattened surface, but generally two or more groups of striations can be detected, or again, isolated, strongly marked and somewhat curved scratches may

fjispl^|4#

-i

*■

If II

li

•gi .jjF -J' .^' .^' .^' .^. .

III 111

91-:

41

150 GEOLOGY OP CAPE COLONY

be found alone or with the other striations. In all respects these boulders and pebbles are similar in form and in the nature of their striations to the scratched boulders that are found in the moraines of modern glaciers and the ancient boulder clays and moraines of Northern Europe and America, countries that are no longer so extensively covered with ice and snow as they used to be.

If the striated boulders in the Dwyka conglomerate belonged to a less remote geological period no doubt would be cast upon the glacial origin of their peculiari- ties; but as the rock is of Carboniferous or Permian age, an epoch so far back in the earth's history that none of the species then inhabiting the world has sur- vived to the present day, when whole classes of animals and plants now flourishing in every quarter of the globe, such as birds, mammals and flowering plants, were still merely future possibilities, and when not one of the great mountain chains of our present day continents had come into existence, people have been very reluc- tant to accept this explanation. There is a deep-seated prejudice against the idea that glacial conditions could have prevailed so long ago in countries that now enjoy temperate and subtropical climates. This feeling is perhaps no longer so strong as it was in the sixties of last century, when ice-action was first brought forward in explanation of certain features in the Talchir conglo- merates at the base of the Indian Gondwana system, ^ and when Sutherland'-^ showed that the conglomerate

1 H. F. and W. T. Blanford and W. Theobald, Mem. G. S. India, vol. i., 1859, pp. 33-90. ^ Sutherland (68), p. 17, etc.

THE KARROO SYSTEM 151

at the base of the coal-bearing rocks of Natal, which he stated was the same as Bain's ** clay stone-porphyry ** in Cape Colony, was mainly of glacial origin. ^ A paper by Professor Edgeworth David on the evidences of glacial action in Australia in Permo-carboniferous time seems to have brought many European geologists to believe that such climatic conditions prevailed at so early a period.^ The evidence does not merely depend upon the presence of flattened and striated boulders, but the general nature of much of the conglomerate and the form of the floor beneath it in certain areas confirm the glacial theory.

An ordinary conglomerate is more or less bedded, the larger boulders lie together with a small quantity of sand between them, and the pebbles likewise are roughly arranged according to size with a larger propor- tion of sand. The inclusions often touch one anothei^ ; they are not scattered at wide intervals through the fine-grained matrix of the rock. Such conglomerates can be seen in many parts of the Colony. The Table Mountain series in the Olifant's Eiver Mountains, at Baboon Point, and at other places on the west coast contain thick beds of conglomerate with normal characters. The Uitenhage beds in the south-western districts contain numerous instances, and at the base of the series in the Uitenhage Division there is a strongly developed conglomerate of the usual type. Amongst

^ I have omitted all reference to the earlier views as to the origin of the conglomerate. A full historical a.ccount of this matter will be found in Oorstorphine, GeoL Comm. (99), pp. 5-20.

2 Q, J. G. S.y 1896, p. 289.

152 GEOLOGY OF CAPE COLONY

the recent deposits of the southern and western parts of the Colony, both river-formed and beach conglom- erates are not infrequent. In all of these rocks one looks in vain for the characteristic flattening and stria- tions found so abundantly in the Dwyka boulders, and for the occurrence of large isolated blocks in a fine-grained matrix. The reason is that currents or waves that have sufficient power to move large blocks of stone sweep away the pebbles and sand from the same neighbourhood, so that the large stones . come to lie together, while the smaller fragments come to rest in quieter water. When a large block is entirely surrounded by stratified mud or sand, it has been dropped there by some floating body, and of such bodies ice is by far the most important. Practically the only exception to this is the falling of blocks from volcanic explosions into ash or sand, but volcanic agencies had no part in the formation of the Dwyka conglomerate. Drifting trees and masses of vegetation can be called in to account for the presence of isolated blocks of rock in fine-grained beds from which other evidence of glacial action is absent, especially if fossil wood occurs in the same beds ; but such means are out of the question when we have to deal with the repeated occurrence of large blocks in unfossiliferous beds covering wide stretches of country.

Every detailed account of icebergs met with in the Arctic and Antarctic Seas mentions blocks of rock as well as small fragments, sand and mud, contained in the ice and lying upon its surface. The ice that forms along a shore encloses a quantity of pebbles and mud,

THE KAltJlOO SYSTEM 153

and receives additions of both ice and rock-debris from the land side ; when this breaks up and part of it drifts away, the load is carried off and dropped by the melting of the ice. By means, then, of icebergs and drifting floes, it is probable that the boulders and pebbles, as well as some of the matrix of the conglom- erate in the south of the Colony reached their present positions.

It is uncertain to what extent the conglomerate in the north is a true morainal deposit, that is, one formed on land or in very shallow water at the end or bottom of glaciers or ice sheets. The internal character of a moraine may not be very different from that of a sandy clay, into which boulders have been dropped from float- ing ice, and it is difficult to decide which is which in the absence of well-developed bedding planes ; even in morainal areas the sediments deposited in temporary lakes give rise to bedded sands and shales that may be again covered up by typical boulder clay. In the case of recently glaciated regions the original surface forms of many of the deposits can be traced, and lithological changes can often be followed up and assigned to their proper place in the history of the area; but when we have to deal with the results of a glacial period of late Carboniferous age, which have probably been buried under thousands of feet of later sediments, and which are only visible owing to the removal by denudation of these superincumbent rocks, a full explanation of the meaning of each change cannot be expected. In Prieska, the district in which the northern conglomerate has been most fully examined, exposures are by no means plentiful,

154 GEOLOGY OF CAPE COLONY

and the difficulty of arriving at a just conclusion as to its mode of formation is increased by the uncertainty of its thickness at various points, owing to the removal of the overlying beds; otherwise the horizontality of the rocks would make the district a peculiarly favourable one for observations.

With the progress of the geological work in the north much evidence will be collected as to the nature of the conglomerate at different localities, so that it may be possible to delineate areas of true moraines, of glacial lakes, and possibly of conglomerates remade from the moraines as they became submerged during the advance of the water northwards.

It is quite justifiable to regard those portions of the conglomerate resting upon a, striated floor as a terminal moraine formed during the retreat of the ice, or perhaps at an earlier period, that is, as a ground moraine. A considerable portion of the northern conglomerate must be included under this head, but it is uncertain whether the whole of the conglomerate in that region was formed under quite the same conditions.

To the north of latitude 33° the conglomerate rests unconformably upon the underlying rocks, but it is by no means everywhere that one can find a glaciated floor below it. In the divisions of Hope Town and Prieska excellent examples of rounded and striated hillocks {roches moutonn^es) have been found immediately below the conglomerate. Over thirty years ago Dr. Sutherland described somewhat similar appearances in Natal ; ^

1 Sutherland (68), p. 17.

THE KAHROO SYSTEM 155

afterwards Mr. Dunn^ discovered a fine striated floor below the conglomerate at the confluence of the Orange and Vaal Eivers ; in later years Dr. Molengraaff ^ found a similar floor below the conglomerate in Eastern Trans- vaal, and the Cape Survey ^ came across the magnificent roches moutonnies of Prieska and Hope Town.

At Jackal's Water in Prieska the conglomerate lies upon the hard quartzites of the 'Keis series, which crop out in the form of rounded and polished surfaces covered on their northern slopes with nearly parallel groves and scratches of various lengths, Ijang in a north-north-east to south-south-west direction. The southern ends of the hillocks are steeper, rough and unstriated. These two sets of surfaces correspond exactly with the **tail and crag," or *' stoss- and lee-sides *' of the roches moutonndes that are met with in every region where ice has passed over hard rocks. The ice, either in the form of a glacier or a more extensive sheet, in moving over the surface ground down the underlying rock with the aid of the sand and stones contained in it ; the side of a projecting mass of rock exposed to the greatest grinding, naturally that facing the point from which the ice moved, had its surface smoothed, scratched, and polished. Plates VIII. and IX. are views of the quartzite roches moutonndes of Jackal's Water; the second view shows the nature of the surface ; the lines traversing the surface from the lower edge of the picture to the right side are due to bedding planes; other cracks are those formed along

1 Dunn (86), p. 8. «Molengraaff (98), p. 103 ; and (01), pp. 71-74.

3 Rogers and Schwarz (00), pp. 118-120 ; and OeoL Comm, (99), pp. 95, 96.

al- i- as — s

II

|i

is

1£l

|S.

s|l

l-l ill

'^f|iiiifiig?^;|'i

.jj- . j[. . j[. . j[. -jf. -^ :«: . jf . jf -r-

■fl

ft

III It*

ill Hi

Irs 111

Hi il! jJi

4i

SSI

Sf

«

158 GEOLOGY OF CAPE COLONY

joints; the fine striae making a sharp angle with the bedding planes are glacial scratches.

At Vilet's Kuil, in Hope Town, the hard lavas of the Beer Vley volcanic group play a similar rdle to that of the quartzites of Jackal's Water, and the scratches are directed about 10° east of south, the lee-side being on the south. In both of these cases the surface of the older rock retains the roche moutonn4e form for a distance of some 200 feet from the outcrop of the Dwyka con- glomerate. Beyond this limit the rocks have lost their glaciated surfaces owing to weathering since the removal of the overlying conglomerate by denudation. The ground occupied by the conglomerate round the roches moutonn^es, seen in the foreground of the view in Plate VIII., is covered with the characteristically striated boulders; many of these lie upon the surface of the older rocks, exposed in the immediate neighbourhood, having been left there on the removal of the matrix of the conglomerate.

The hard quartzites of the 'Keis series, and the almost equally hard lava of the Beer Vley group, are well fitted to retain the glacial markings for long periods. The reason why such phenomena are not more gener- ally seen in Prieska is partly that many of the rocks lying below the conglomerate disintegrate rather readily, and are consequently not well adapted for preserving their old glaciated surfaces. A great part of the bound- ary between the conglomerate and older rocks passes over coarse granite and gneiss, which break up rapidly under the influence of great differences in temperature, a marked character of the climate in that region. Other

THE KARROO SYSTEM 159

parts of the conglomerate boundary are hidden under sand and other surface accumulations. In addition to these hindrances to the observation of the surface that immediately underlies the conglomerate in the north, it must be remembered that but a very small part of the country has been closely examined, and that the whole tract between the Kaaing Bult and Loeries Fontein, a distance of at least 200 miles along the Dwyka outcrop, has not been touched.

The discovery of glaciated surfaces at the junction of the Orange and Vaal Bivers, in the Eastern Transvaal, in Natal, and in Prieska and Hope Town, in all over a very wide area, is suflBcient to make one expect to find such surfaces below the conglomerate wherever it rests unconformably upon the older rocks. The con- glomerate is unconformable north of Karroo Poort. Be- tween Karroo Poort and the Bosch River in the Tanqua Karroo it rests upon the Witteberg beds ; at two or three spots only along this part of the boundary, sixty miles in length, has the actual contact been seen, and although the surface of the Witteberg quartzites is striated at those places, there are so many slickensided surfaces in the same rocks, produced by the bending, and consequent slipping of one layer over another, after the Dwyka conglomerate was formed, that in the absence of favourable exposures it is impossible to be certain of the glacial origin of the scratches immediately below the conglomerate. The thin-bedded quartzites are ill suited for retaining the striae, if they were ever present. Also at the time when the conglomerate was formed the Witteberg beds had only recently been deposited, and

160 GEOLOGY OF CAPE COLONY

must have been very much softer and less coherent than now after they have been buried under a great load of other rocks, subjected to earth movements, and again exposed to our view.

Between Bosch Eiver and Matjes Fontein on the Oorlog's Kloof Eiver, a distance of fifty miles, the con- glomerate rests upon the Bokkeveld series, gradually coming to lie upon lower and lower beds belonging to that group, till at Matjes Fontein only the lowest band of fossiliferous shales remains between the conglomerate and the Table Mountain sandstone. From the Oorlog*s Kloof Eiver, where the conglomerate and sandstone crop out within a few yards of one another in the river bed, to the escarpment on the south side of the Doom Eiver (Calvinia) Valley, the conglomerate rests directly upon the Table Mountain series, which decreases in thickness from perhaps 5,000 feet to two or three in the interval ; north and east of Uithoek the sandstone no longer intervenes between the conglomerate and the Pre-Cape rocks. The conglomerate there lies upon the Ibiquas series as far north as the fault separating the latter from Bushmanland granite and gneiss. North of the fault the conglomerate rests upon the granite (see Plate X.), and is not in the least affected by the great dislocation, which was therefore in its present state in Dwyka times.

The only place along the western outcrop of the Dwyka conglomerate where actual evidence of the move- ment of ice over a floor of any kind has been seen is at Eland's Vley, near the confluence of the Tanqua and Doom (Clanwilliam) Eivers. On either side of the

i'

il li

-*■:

- ^F^F^f^f

•s'-s'ts

:*:

162 GEOLOGY OF CAPE COLONY

Doorn Biver there is exposed a ** striated pavement/* not of the underlying rock, but of the conglomerate itself, which passes under a further thickness of con- glomerate. The "pavement '* is a flat surface of con- glomerate in which there are numerous boulders up to three feet in diameter ; these are pressed down flush with the general surface of the pavement, and are finely striated in one direction, which is nearly due east ; these boulders may also have another set of their striae which run in different directions, but they have been mostly obliterated by the agency that produced those men- tioned. The matrix of the conglomerate is a tough blue sandy mudstone, and is traversed by numerous furrows which run parallel to the dominant striae on the boulders. There can be little doubt that this sur- face, which is from fifty to eighty feet above the base of the conglomerate, was caused by ice moving across it from west to east. The conglomerate was at the time a stiff, sandy mud containing many pebbles and boulders, which. When at or near the surface, were forced down flush with the latter, and striated and polished by the sand and stones set in the bottom of the ice, that also made the furrows in the mud. After this mass of ice had disappeared, sandy mud with boulders and pebbles, precisely like the conglomerate below, was deposited upon the striated pavement. Many instances of such surfaces have been found in recently glaciated regions ; they are produced wherever a glacier or large sheet of ice moves over a floor of boulder clay or till.

Where the Dwyka conglomerate first appears on the coast of Pondoland near St. John's it is faulted down

THE KARROO SYSTEM 163

against the great block of Table Mountain beds that forms the mountain through which the St. John's Eiver flows just before entering the sea. To the north-east of St. John's, along the western flank of the high plateau of Table Mountain sandstone that borders the coast, the conglomerate rests directly upon the sandstone, as is also the case in Natal ; no part of the Bokkeveld or Witteberg series has been left in those regions between the two formations, which stand in the same relation to one another as in Calvinia, north of the Oorlog's Kloof Biver. The conglomerate in Pondoland has precisely the same general appearance as in Calvinia and the western Karroo ; the colour and nature of the matrix are the same, and in both districts there are large and small inclusions of many varieties of rock, considerable numbers of which are flattened and striated on one or more sides.

The boulders in the Pondoland outcrops are, as usual, derived from many kinds of granite, gneiss, diabase and other igneous rocks, as well as sandstones, quartzites and other sedimentaries ; but the jaspers and banded magnetic rocks from the Griqua Town beds, which form a small but interesting part of the boulders in the west and south, have not been noticed there.

A considerable part of the conglomerate is unbedded ; not only are the pebbles and boulders scattered at ran- dom throughout the rock, but the matrix is without lamination planes. But in any district where the con- glomerate is well exposed over large areas, traces of bedding can be found in the matrix, and it is sometimes

so well laminated that it can be called a shale. In

11*

164 GEOLOGY OF CAPE COLONY

Prieska shaly portions of the rock are found quite close to outcrops in which no lamination can be seen. Within a short distance of the spot at which the photograph reproduced in Plate VII. was taken, there is a patch of shale without any pebbles in it, and in other parts of the district the matrix of the conglomerate is well laminated. In the western and southern Karroo shaly conglomerate is often met with, the pebbles and boulders in it being precisely of the same nature as those in the unbedded conglomerate, and they occur in the same way. These shales must have been deposited in quiet water and the boulders dropped to the bottom from floating ice.

In the Tanqua Karroo a fairly constant band of very large boulders stretches for many miles north and south of Eland's Vley. It is about fifteen feet thick and some of the boulders are from three to four feet in diameter, but most of them are less than half this size ; many are well striated. The rocks above and below the boulder bed differ from it only in the smaller proportion of inclu- sions distributed through them. Another definite boulder bed has been found in the valley of the Witteberg's Eiver south of Laingsburg, and is shown on Plate XI. The largest block seen in the photograph is ten feet across.

Throughout the area in which there is an unconformity below the Dwyka series, the conglomerate lies directly upon the older rocks, except perhaps where small patches of shale occur, such as the one mentioned from Prieska, which may be at the base of the series. In the south of the Karroo, where the Dwyka series lies conformably upon the Witteberg beds, there is always a certain thick- ness of greenish shales between the conglomerate and

THE KAEROO SYSTEM 165

the Witteberg quartzites. These shales pass gradually upwards into the conglomerate, which contains only small pebbles near its base in the southern region. Al- though they are undoubtedly passage beds between the two formations, i.e., they represent the period of transition from the conditions under which the Witte- berg series was formed, to the colder conditions that prevailed later, they are placed for convenience with the conglomerate, and are called the Lower Dwyka shales. They consist of shales and thin quartzitic sandstones, and are in all from 600 to 700 feet thick, measured from the uppermost thick quartzite of the Witteberg group to the lowest bed that is distinctly conglomeratic. Some of the strata are very like the shales of the Witteberg, and others, especially near the top, are of the same nature as the matrix of the conglomerate. The Lower Dwyka shales are well exposed at many places on the north flank of the hills formed by the Witteberg series along the southern edge of the Karroo. They can be well seen south of Matjes Fontein, in the Witteberg's River south of Laingsburg, at the north end of the Buffers River Poort (Leeuw Kloof Poort), and just south of Prince Albert village where the road to the Zwartberg Pass enters the narrow kloof, to mention some of the more accessible localities in the Karroo. East of Prince Albert this horizon has not yet been described, but at Grahamstown, both to the north and south of the town, similar shales 650 feet thick intervene between the Witteberg quartzites and the Dwyka conglomerate, lying conformably to both. There can be little doubt,

166 GEOLOGY OF CAPE COLONY

therefore, that the Lower shales are a definite group of beds present at the base of the Dwyka series wherever it lies conformably upon the Witteberg beds.

In the Witteberg's River the Lower shales have been found to contain impressions of stems resembhng the Phyllotheca stems of the Ecca beds ; these are the only known fossils from the Lower shales.

It is unfortunate that the strip of country immediately north of Karroo Poort occupied by these shales is so ob- scured by gravels and sand that the exact manner of their disappearance has not been determined. On the view of their relationships adopted here the break in the succession should commence at the bottom of the shales.

The Dwyka conglomerate in the south of the Colony is in some respects very different in appearance from that in the north, owing to the earth movements that have affected the former region. Throughout the southern outcrops the conglomerate is a hard blue rock from which the pebbles do not readily break out. When the rock is struck with a hammer the fracture is more likely to pass through a pebble than round it. There is a rough cleavage developed in the matrix, parallel to the strike of the beds, but at various angles to their dip. This causes the conglomerate to weather into lenticular slabs, very characteristic of the rock in the southern parts of the Colony. The slab or tombstone-structure, as the late Professor Green called it, is shown on Plate XI., a view of the steeply dipping conglomerate cut through by the Witteberg's Eiver south of Laingsburg. The appearance resembles more closely that known as

|l'fsft^jig;«

168 GEOLOGY OF CAPE COLONY

*' pillow-structure '* in many basic lavas of Palaeozoic age in Britain than the normal results of cleavaga In the Karroo outside Karroo Poort, where the Dwyka conglomerate has been affected by the pressures that produced the east and west folds (Zwartberg folds), and those that gave rise to the north and south folds (Cederberg folds), the rock has the rough cleavage developed in two directions, and weathers out in pillars, usually tapering upwards. The development of the slab-structure becomes weaker as the conglomerate is followed northwards from Karroo Poort into the region where the folding did not take place, and in Calvinia the rock is of the same nature as in Prieska, a sandy mudstone or shale, according to whether lamination is absent or present. The northern rock breaks up readily, and the pebbles can easily be removed from the matrix. A curious feature in both the northern and southern conglomerates, but more highly developed in the latter, is the regular and close jointing of the enclosed pebbles and Ijoulders. A pebble, four inches long, may be trav- ersed by two or three dozen joints parallel to one another, and quite independent of the original divisional planes, such as those of bedding or foliation, in the pebble. In the north and north-west, where the con- glomerate lies nearly horizontally, the joints are alsc« horizontal, but occasionally vertical ones can be found. In the south the joints, which are parallel in all the pebbles at any one spot, lie more or less parallel with the strike, but not with the bedding planes in the conglomerate. Occasionally one or more of the sections into which the pebbles are divided have shifted rela-

THE KARROO SYSTEM 169

tively to those above and below. The matrix of the conglomerate shows no signs of the continuation of the joints through it. The jointing has been explained, on the supposition that there are faint divisions in the matrix, due to the long-continued action of a moderate pressure and solution deforming the constituent grains along the directions of the supposed planes in the matrix, so that the pebble eventually broke along these planes of deformation.^

At several places in the south and west of the Karroo beds and lenticular patches of white quartzite occur in the conglomerate. Near Matjes Fontein several large lenticles of quartzite lie on the same horizon. They are roughly bedded, and the bedding planes have a similar dip to that of the conglomerate in the immedi- ate neighbourhood. In the Ceres Karroo near Beukes Fontein, there are also several quartzite lenticles like those at Matjes Fontein, but the quartzite is rather yel- lower, and at its periphery it contains boulders. The base of the conglomerate on the left bank of the Doom River at Eland's Vley is very quartzitic in places, doubt- less owing to the large amount of quartz sand derived locally from the Witteberg beds. Boulders of several kinds of rock, diabase, granite, etc., as well as quartzites that may have come from the Witteberg beds, are imbedded in the conglomerate there. The lenticular patches must have had a different origin, for they are considerably above the base of the conglomerate, and they occur where the Lower Dwyka shales intervene

1 Schwarz (03), p. 399, etc. On Plate V., Fig. 2, accompanyiDg this paper a photograph of one of the jointed pebbles is reproduced.

170 GEOLOGY OF CAPE COLONY

between the conglomerate and the Witteberg group. They are surrounded by blue rock of the normal type, and probably represent local patches of sand, but an entirely satisfactory explanation of them has not yet been found.

Some patches of the conglomerate contain more car- bonate of lime than others, and weather out from the rest of the rock in the form of spheroidal and lens-shaped lumps, that occasionaUy pass into masses large enough to be called lenticular beds. In the western Karroo there are many such calcareous beds. The spherical lumps are usually from six to ten inches in diameter ; they seem to be particularly abundant near Laingsburg and in the Tanqua Karroo, but they have been found in many other districts. The carbonate of lime in these concretions has probably reached its present position by a slow process of concentration from the surrounding rock. The matrix of the conglomerate always contains a certain amount of calcite in the form of mud, sand and small limestone pebbles.

The sources of the many varieties of rocks forming boulders in the conglomerate are only partially known. The brown, red, yellow and black banded jaspers and magnetic quartzites are identical in character with rocks belonging to the Griqua Town series in Prieska and Griqualand West. There are two kinds of amygda- loidal lavas widely distributed throughout the conglom- erate in the Colony, a more basic variety like those at

•

Zeekoe Baard, and a more acid rock closely resembling the Beer Vley lavas. Both these types are probably widely distributed in Griqualand West, so that it is im-

THE KAKROO SYSTEM 171

possible to determine the precise source of the boulders. The cherty crystalline limestones of the Campbell Band beds have furnished many fragments to the conglomer- ate in the western Karroo, although they are by no means confined to that region ; the Campbell Band marbles probably supplied most of the calcareous mud so abundant in the matrix. The microline granites and gneisses, of which many varieties occur in the con- glomerate, may be matched by rocks from several known outcrops in Prieska, and similar rocks seem to be abun- dant north of the Orange Biver. The Matsdp beds (purple quart zites, grits and conglomerates), are well represented in the western and southern Karroo, and so are the 'Keis quartzites and mica schists. Serpentines, found in the conglomerate west of Calvinia, are as yet only known in place in the north and north-west of the Colony. There are large numbers of white quartzite and brown sandstone boulders in the conglomerate, but their origin is uncertain ; they may have come from the Table Mountain and Bokkeveld series north of the un- conformity, but no Bokkeveld fossils have yet been found in the conglomerate. Many altered doleritic rocks from the southern conglomerate can be matched from outcrops in Prieska. Several well-marked va- rieties of acid porphyritic and felsitic rocks are met with in the conglomerate, but their source is not yet known.

The bulk of the formations that have supplied the boulders of recognisable origin occur only in the north of the Colony, and have not been met with in the south, although the latter is by far the better known area.

172 GEOLOGY OP CAPE COLONY

The microline granites are the only rocks amongst the Dwyka boulders that resemble at all closely some of the souther^i Pre-Cape rocks, and even they are still more like the northern granites. There can be no doubt that the main source of the boulders lay to the north, a conclusion that is in full accord with the observed direction of the striae on the Jackal's Water and Vilet's Kuil roches moutonnSes, as well as on the striated floor described by Mr. Dunn at the junction of the Orange and Vaal Eivers. It is also in agreement with the general relationship of the conglomerate to the under- lying rocks, for the boulders came from the north where the unconformity is ; there is no clear evidence that any of them had a southern origin, and so far as is known the conglomerate was laid down conformably to the Cape formation throughout the south of the Colony.

The conglomerate is about 1,000 feet thick in the south of the Karroo, but diminishes in thickness north- wards. Where it lies nearly horizontally, as in Prieska, Kenhardt and Calvinia, it covers wide stretches of country, but is of varying thickness, and never more than some 500 feet, if so much. At Kimberley it is represented by a few feet of rock passed through by the shafts outside the diamond pipes.

No fossils have yet been found in the Dwyka conglomerate within the Colony, but outside our area, at Vereeniging, the remains of many varieties of plants occur at a short distance above the conglomerate, and some fragments of plants as well as layers of coal are found in the shaly rocks interbedded with boulder beds

THE KARROO SYSTEM 173

that Mr. Dunn ^ has shown to be representatives of the Dwyka conglomerate.

Lying above the conglomerate in the south and west there are some 500 to 600 feet of shales, sandstones and cherts, called the Upper Dwyka shales, into which the conglomerate passes conformably by the gradual dimi- nution of the number of boulders. The lowest beds are bluish or greenish sandy shales, overlain by thin sandstones, which are in their turn succeeded by a group of black shales weathering white on exposure to the air. The black shales are overlain by fine- grained green beds, with thin beds of limestone and ferruginous rocks, and several layers of chert, grey or black when freshly broken, but with a thin white crust on exposed surfaces. The uppermost of the chert beds, usually from eight to twelve inches thick, is taken as the top of the Dwyka series.

The black shales contain a certain amount of carbonate of lime, often gathered together in the form of nodules, and iron pyrites. These two minerals, and the car- bonaceous matter that gives the black colour to the shales, decompose under the influence of the air, form- ing gypsum (sulphate of lime) and iron oxides, and leave the shales bleached white. These white rocks make very conspicuous features on the southern and western borders of the Karroo, where the vegetation is not sufliciently abundant to hide the colour of the bare hillsides. Thus the black shales near the top of the Dwyka series are known as the ** white band ".

^ Dunn (00), p. 67.

174 GEOLOGY OF CAPE COLONY

The dark colour of these shales has led to their being prospected for coal at many places, but although the percentage of carbonaceous matter rises to 7^ per cent, nothing that can fairly be called coal has been found in them in Cape Colony.

Although the Upper Dwyka shales as a whole appear to change in character in the north of the Colony, espe- cially by the absence of the ferruginous beds and the chert, the black shales persist in the north of Calvinia, Prieska and Hope Town, and probably across the inter- vening country. They exist at Kimberley, where they form part of the rocks called the ** Kimberley shales," and are probably directly continuous with the coal- bearing rocks of Vereeniging that overlie and are inter- bedded with the boulder beds there. Mr. Dunn in 1886 ^ came to the conclusion that these shales, which he showed would certainly be found to extend under the whole of the Karroo, contain coal in some parts of that area ; in 1899 when he found that the Vereeniging coal ^ lay on about the same horizon, *.e., close above the Dwyka conglomerate, he naturally considered his case for the existence of sub-Karroo coal greatly strengthened. Vereeniging, and the other localities, such as Kroonstadt, where coal of inferior quality to that of Vereeniging is said to have been struck, lie far to the north-east of the black shale outcrops south of the Orange Eiver, and nearer the old land on which the plants grew that went to form the coal, if indeed the plants did not live in the immediate neighbourhood of the present coal beds. The

1 Dunn (86). ^Du^n^oo).

THE KARROO SYSTEM 175

fact remains that so far as the black shales have been investigated, they contain smaller quantities of organic matter as they are followed south-west, and there is no reason to suppose that the horizon which has been proved to be without coal at many places south and west of the Karroo basin, as well as along the Hope Town-Calvinia edge, should contain valuable deposits under De Aar or any other spot within the basin where it is hidden from view beneath hundreds or thousands of feet of other beds. The places where the Upper Dwyka shales have been closely examined throughout their whole thickness and have been found to be without coal are the following, taken in order round the Karroo basin from Kimberley : Kimberley, Hope Town, the south of Prieska, Loeries Fontein, several spots west of Calvinia, Blaauw Kranz on the Calvinia transport road^ the Tanqua Valley, outside Karroo Poort, Laingsburg, Prince Albert, north of Botha's Hill, Grahamstown, and again in Pondoland.

It should be remarked also, as will be more fully shown on a subsequent page, that the present position of the shale outcrop in the Prieska-Kimberley region is by no means coincident with the original limit of the group, for an outlier which has been disconnected from the main area of the beds by denudation exists far to the north-west in the Kalahari Desert.

In the banks of the Camdini Eiver near Loeries Fon- tein the black shales are very well exposed, and they are traversed by dykes of dolerite, which has brought about the formation of graphite in minute scales, filling cracks in the immediate vicinity of the igneous rock.

176 GEOLOGY OF CAPE COLONY

The only recorded and determinable fossil from the Upper Dwyka shales in the Colony is Mesosaurtis,^. a small reptile of which only a few specimens have been found. The first specimen came from an unknown local- ity in Griqualand West, and was described by Gervais ^ under the name of Mesosaurus tenuidens ; others were subsequently found in the black shales close above the Dwyka conglomerate in the Kimberley Mine, although these were too imperfect to be named with certainty. Another specimen of the genus has been found in the Upper Dwyka shales west of Calvinia ; and in a very similar rock in southern Bushmanland a fine tail and hind part of the body was discovered a few years ago. All these specimens are impressions left in shale by the removal of the animal's bones.

The occurrence of Gangamopteris cyclopteroides var. attenuata Feistm. and Noeggerathiopsis hislopi Feistm. near Kimberley has been recorded,^ but it is not certain whether these plants came from the Dwyka series or the Ecca.

In the south of the Colony the only organic remains yet met with in the Upper Dwyka shales' are indeter- minable markings that are probably of vegetable origin.

The distribution of the Dwyka series can be seen at a glance on the geological map of the Colony. It fonns a continuous band round the south and west of the Karroo, then turns eastwards and passes through

^ Another genus, Ditrochosauriis, has been described, but its distinc- tion from Mesosau us may be due to an accidental feature in the single specimen known.

2 Gervais (79). =*Moulle (85) ; Feistmantel (89).

THE KARROO SYSTEM 177

Calvinia, Kenhardt, Prieska and Hope Town, where it is crossed by the Orange Eiver, and is continued past Kinaberley into the Orange Eiver Colony and the Transvaal. North-east of that part of the coast where the sea has breached the edge of the Karroo basin the conglomerate appears again in Pondoland, and is continued through Natal to the Eastern Transvaal. Throughout this immense area the conglomerate is probably everywhere present at the base of the Karroo formation, and it has a persistent, though varying dip towards the interior, so that it forms a basin. West- wards from the Gualana Eiver as far as Karroo Poort, and thence to the Tanqua Eiver, the beds often dip at high angles beneath the Karroo, but farther north and east they lie horizontally, or dip at a very low angle towards the south or east. This basin whose edge is defined by the conglomerate is due to folding, but the gentle inclination of the extreme northern portion may be an original feature ; the southern portion has been thrown into its present form by folding, and no evidence of the original southern limit is known.

The outliers of the Dwyka series in the folded belt south of the Karroo are few in number. The chief one is that which forms a semicircular area between Worcester and Lange Vley near Eobertson. It, to- gether with a considerable thickness of Ecca beds, is faulted down against the Malmesbury beds and granite exposed under the Langebergen. The rocks are of precisely the same general character as those along the south of the Karroo, but the black shales have

been converted into graphitic slates, which have been

12

178 GEOLOGY OF CAPE COLONY

unsuccessfully prospected for graphite. In the Wor- cester district as in the Karroo the conglomerate rests' upon the Lower shales, and these again lie conformably upon the Witteberg beds. The Worcester outUer is about forty miles distant from the nearest part of the main Dwyka area in the Karroo, and is particularly interesting because it shows no sign of a change in the nature of the beds or in the relationship between them and the older rocks. These facts, together with the uniform character of the conglomerate anS its mode of occurrence, at least as far east as Grahamstown, warrant the assumption that the area of deposit of the Dwyka series was not limited in a southerly direction within the boundaries of the Colony.

An outlier of the Dwyka beds has been found at the head of the Winkelhaak's Eiver in the Cold Bokkeveld. In the country south of the Karroo there are six other outliers. The largest is that of Quarrie Kloof between Touw's River and Constable stations, and four others lie to the south-east of it ; the sixth is in Dobbel Aar's Kloof, about thirty miles from the Quarrie Kloof outlier. All these patches of Dwyka are boat-shaped synclines preserved from denudation by the fact that they lie in rather deep folds. The rocks composing them do not require special description, for in all respects they resemble the southern Karroo outcrops.

In the Eastern Province the outcrops of the Dwyka series are repeated by folding, as shown on the map at the commencement of the volume, but their distribution is not yet known in detail. In the Albany Division the conglomerate, with the Lower Dwyka shales, occupies

J

THE KAKROO SYSTEM 179

the valley in which Grahamstown is situated, and is well exposed at many places near the town. The series there lies in a syncline ; the Witteberg beds dip under it both to the north (south flank of Botha's Hill) and to the south; it is also well seen north of Botha's Hill near the road to Fort Brown, on the north of the Botha's Hill anticline.

During the past year Dr. Nobbs of the Cape Agricultural Department visited the Kalahari Desert north of Upington and brought back some specimens of the' formation underlying the desert sands near Eenzamheid and the Noro Kei Pan. Those from the former place are spheroidal masses of Dwyka conglomerate, just like the calcareous concretionary masses that occur in thousands in the conglomerate of the western Karroo; and from the Noro Kei Pan came pieces of silicified wood resembling that found in the Ecca beds in many parts of the Colony. These discoveries and Dr. Nobb's statement that grey shales containing the fossil wood are met with in wells near Noro Kei Pan, undoubtedly prove the existence of an outlier of the lower part of the Karroo formation in the Kalahari, more than 100 miles farther north-west than was formerly thought to be the case.

Towards the north the conglomerate has been found at Vryburg.

The Ecca Series.

Lying conformably upon the Upper Dwyka shales

throughout the southern and western Karroo are the

12*

180 GEOLORY OF CAPE COLONY

shales and sandstones called the Ecca beds, a name given them by Atherstone from their occurrence in the Ecca Pass in Albany.

The strata immediately above the uppermost chert bed of the Dwyka in the south and west are usually thin flaky shales, and green shales are found above them, together with thin beds of mottled grey and green sandstone. Some of the shales near the base of the series break up into long roughly prismatic fragments after the manner of the starch of commerce. In the neighbourhood of Patata*s Eiver south of the Klein Eoggeveld hard sandy beds lie immediately above the Dwyka series. The thickness of the lower portion of the Ecca beds in the south and west, in which the shales predominate, is from 1,000 to 1,200 feet, and they are succeeded by some 1,200 feet of strata in which sandstones are the chief feature. These, called the Laingsburg beds from their occurrence near the town of that name, are hard, dark-coloured, fine-grained sandstones and hard shales; they contain Glossopteris, Schizoneura, Phyllotheca and silicified wood. On the weathered surface the sandstones are usually yellow- ish, but some of the finer-grained beds break up into rounded fragments with a thin red crust. The Laings- burg beds have been traced through the country on the south and west of the Klein Eoggeveld, where they form very hilly ground, as far as the left side of the Tanqua Valley; but they become much thinner in that neighbourhood, and apparently disappear, being perhaps represented by shales farther north. The sand- stones in the Laingsburg beds often contain spherical

THE KARKOO SYSTEM 181

nodules of harder material, which stand out prominently on weathered surfaces.

The uppermost portion of the Ecca beds in the southern and western Karroo varies considerably in the proportions of sandstones and shales in different localities. The sandstones are frequently mottled grey and blue. On the Kraai Eiver, near Tuin Plaats, Glossopteris, Gangamopteris and Schizoneura occur in hard shales belonging to these beds.

In the Boggeveld and Hantam region the sandstones that are so conspicuous in the country farther south are but slightly developed, and the whole of the Ecca series becomes an essentially argillaceous group, with only thin beds of sandstone intercalated with the shales ; the thickness of the series diminishes in the same direction, and is probably somewhat over 2,000 feet near Calvinia village. The rocks are well exposed on the Hantam Mountains and on the Koggeveld escarp- ment, of which the former were once a part, having been detached from the main mass by the erosion due to the Oorlog's Kloof Eiver.

Little is known about the Ecca. beds between the Hantam and Prieska where they are probably repre- sented by shales and thin sandstones. They cover wide areas in Hope Town, Britstown, and other districts both to the south and north of the Orange Eiver.

The beds that are called Kimberley shales ^ and Olive shales 2 in that region probably belong in part to the

1 Green (88). « Stow (74).

182 GEOLOGY OF CAPE COLONY

Ecca group, but the demarcation of the latter from the Upper Dwyka shales is not so distinct as in the south of the Colony, and the Kimberley shales are not yet known in detail. Glossopteris, Gangamopteris and Noeggerothiopsis have been described from them, and they also contain silicified wood, resembling that from the Laingsburg beds and other parts of the Ecca series. In describing the geology of the Colony I have tried to refrain from going into details concerning particular views that have been discredited by the fuller knowledge of the country gained during the past decade, but in the case of the late Professor Green's ideas as to the relationship of the Kimberley shales and the Ecca beds a departure must be made from this practice. Green's paper ^ is perhaps the most widely known description of the geology of the Colony, and no other work on the subject approaches it in completeness or lucidity of style in spite of its short length. Its author spent some four months only In the Colony, and much of that time was occupied in an examination of the coal beds of the Stormberg, so that misconceptions regarding the wider questions are hardly to be wondered at. On pages 262-264 he argues that the Kimberley shales are a group of beds lying between the Karroo beds (the Beaufort series of the classification here employed) and the Ecca, and that they lie conformably below the Beaufort beds and unconformably upon the Ecca. In the first place he doubts Dunn's correlation of the conglomerate below the Kimberley shales with the

^ g. J. G. S., 1888.

THE KARROO SYSTEM 183

Dwyka conglomerate of the south; but there can no longer be any doubt on this point, the confirmation of which he admitted would greatly strengthen Dunn's classification of the Kimberley shales with the Ecca beds of the south. The presence of Mesosaurus in the Upper Dwyka shales of Calvinia which are directly continuous with the shales below the Ecca beds in the south, and the presence of Oangamopteris and Glossopteris in the Ecca beds of the Tanqua Valley and of Worcester certainly support the view adopted by Dunn, for these three genera occur in the Kimberley shales. In tracing the Ecca beds from the Prince Albert and Laingsburg districts through the Karroo to Calvinia, a work that has only recently been completed by the geological survey, it became obvious that the sandstones charac- teristic of the series in the south give place to shales north of the Tanqua Valley. It is true, on the other hand, that there are sandstones of considerable thick- ness in the Upper Dwyka shales along the Camdini River west of Loeries Fontein somewhat below the horizon of the black shales that weather white; the sandstones of Hope Town mentioned by Green on page 263 of his paper, and regarded by him as evidence of the occurrence of the typical Ecca beds below the Kimberley shales, almost certainly belong to the upper division of the Dwyka series. The country between Blaney and Kei Road, and the tract between Beaufort West and the Nieuweveld escarpment, which from their general characters led Green to see in them a confirmation of his view that the Kimberley shales lay between the Beaufort beds and the Ecca, are certainly

184 GEOLOGY OF CAPE COLONY

composed of the strata called Karroo bedfe by him and now included in the Beaufort series. They lie well above the true Ecca beds, and are separated by thou- sands of feet of strata from the Dwyka conglomerate, for the Ecca beds themselves lie conformably below them and upon the Dwyka series.

The Kimberley shales must be regarded as the equi- valents of the Upper Dwyka shales and part of the Ecca beds of the south and west, but whether they represent the whole or only a portion of the Ecca group remains to be ascertained, for the stratigraphical details of the country between' the Prieska and Hope Town Divisions and the Nieuweveld have not been worked out.

In Pondoland the Umsikaba beds occur just above the black shales of the Dwyka series. They are of consider- able but unknown thickness, and differ in character from the typical Ecca beds of the west, they consist more of clays and mudstones than of shales and sandstones. Near their base, as seen on the road to Lusikisiki from St. John's and near the Embotyi mouth, they are better laminated than higher up in the group ; the surfaces of the laminae are frequently spotted with circular rusty markings about the size of a shilling, perhaps due to the decomposition of iron pyrites distributed more or less uniformly through them. Above these shales come the clays and mudstones, occasionally sandy, dark blue in colour. On the south of the St. John's fault, along which the Dwyka and Ecca beds are let down against the Table Mountain sandstone, the Umsikaba beds are harder and more like the Ecca of the west than in other parts of Pondoland. At Cape Hermes some thin shales

THE KARROO SYSTEM 185

contain obscure plant remains reminding one of the Schizoneura stems of the west. The Umsikaba beds are found from Libode to Bizana, but have not been fol- lowed south-west of Libode.

The junction with the overlying Iduty wa beds is ap- parently a conformable one, but ill defined, as the passage is very gradual. The Idutywa beds consist of rather loose sandstones weathering to a light yellow colour, interbedded with blue and purple shales. They perhaps correspond to the upper part of the Ecca or the lower portion of the Beaufort series, possibly both. No fossils have yet been found in them.

In the Worcester District the Ecca beds are faulted down against the Pre-Cape rocks between a point some four miles west of the town of Worcester and the Goree Kiver, and again near Eobertson. The beds are green and brown argillaceous sandstones and shales and mud- stones, sometimes coloured green and red. From the sandstones and mudstones exposed in a small quarry near Worcester station specimens of Gangamopteris, GhssopteriSf and Cardiocdrjms have been found ; the last- named genus is not known elsewhere in the Colony although it occurs at Vereeniging; Schizoneura occurs in a quarry west of Worcester.

The list of fossils from the Ecca beds in the Colony is very short, but it is augmented if we go beyond our boundary to Vereeniging, where Mr. Leslie has made large collections which have been described by Mr. Seward \ The following is a list of the plants known from these beds up to the present time : —

» Seward (03), pp. 78-101.

J -J

186 GEOLOGY OF CAPE COLONY

Cape Colony. Yereeniging.

Schizoneura. Schizoneura,

Phyllotheca, - Phyllotheca.

Car diocar pits, Cardiocarpus.

Glossopteru browniatui, Brongn. Gonits.

Gangamopteris cychpteroides var. Glossoptei'is broivniana. Brongn.

attenuaia, Feistm. Gangamopteris cyclopteroides,

Noeggerathiopsis hislopiy Feistm. Feistm.

Sphenopteris,

Psygmophyllum kidstoni, Sew.

Sigillaria hrardi, Brongn.

Bothrodendron leslii^ Sew.

Noeggeraihiopsis hislopi^ Bunb.

This assemblage of plants has a close relationship to the flora of the lower part of the Gondwana system in India, from the Talchir to the Damuda beds.^ The genera Glossopterisy Gangamopteris, Noeggerathiopsis , Schizo- neuray Phyllotheca, and Sphenopteris are common to the two groups of beds ; the Glossopteris flora, as it is called, is also found in the Lower Coal Measures and the New- castle or Upper Coal Measures of New South Wales,^ the Bacchus Marsh sandstones of Victoria,^ the Bowen Eiver formation of Queensland,* the Lower coal bearing rocks of Tasmania, in Brazil and in the Argentine Ee- public (Bajo de Velis beds)^. In Queensland ^ marine beds with numerous fossils of Permo-carboniferous type have been found interbedded with those containing the Glossopteris flora, and in Russia a few characteristic

^ Manual of the Geology of Indiaj 2nd edition, Oldbam. 2 Feistmantel (90). ^ j^ck and Etheridge (92).

4 Jack and Etheridge (92), p. 70.

^ Kurtz, Revista del Museo de la Plata, vi., p. 117. In English in Records^ G. S. J., xxviii., p. 111. « Jack and Etheridge (92), p. 70.

''^MiM ^^ ^t^^t^^St^^.Vis containing

...J I

:*:

ib%^»Aa^i^&|p^j£:a|Lereiore, more 'l'B"4l"^^in the Ecca.

•S' ■»• ■«• :m :W: :^ :^ •»• •*•

188 GEOLOGY OF CAPE COLONY

Glossopteris itself has a very great time range, probably from the Carboniferous to Upper Cretaceous/ but it is the most characteristic genus in the flora named after it; it is usually confined to the lower portion of the long range of beds referred to above.

The mingling of the northern carboniferous genera, Sigillaria, Psygmophyllum and Bothrodendron with the Glossopteris flora at Vereeniging ^ is of considerable in- terest on account of the almost complete absence of the northern forms in India and Australia, although in Brazil the northern and southern genera are again found together. It has been suggested that the cold climate of the south at that time, as evidenced by the glacial conglomerates in Africa, India and Australia at or near the base of the strata containing the Glossopteris flora, will explain the absence of the northern carboniferous plants ; but it must be remembered that there is no reason, so far as South Africa is concerned, to believe that the cold climate was of longer duration than the time represented by the Dwyka series, for no conglomer- ates or isolated blocks of stone have been found in the Ecca or Beaufort beds of Cape Colony ; both Glossopteris and Schizoneura extend upwards into the Beaufort series, and the latter genus occurs in the . Stormberg group. The thickness of the strata above the Dwyka series from the Ecca to Stormberg inclusive is about 12,000 feet, and throughout this great mass of rocks no evidence of glaciation has been seen, so that the northern flora could

1 In desert sandstone of Queensland, Jack and Etheridge (92), p. 528.

2 Seward, Address to Bot. Sect. Brit. Ass, (08), pp. 8-18 ; Ann. 8. A, Museum (03), pp. 99-101.

THE KARROO SYSTEM 189

hardly have been kept out by the severity of the climate. Moreover, the Sigillaria and other northern genera have only been found at Vereeniging, where they are closely associated with glacial boulder beds, and they appear to be absent from the southern Ecca beds.

The Beaufort Series.

In the western districts there is a gradual passage up- wards from the Ecca beds, and those that succeed them contain the remains of Pareiasaurus and other reptiles.

The Beaufort beds get their name from their occur- rence in Beaufort West and the Fort Beaufort Division. They consist of sandstones, shales and mudstones. The sandstones are of two kinds, a rather loose-grained rock that forms thick bands of strata in the Nieuweveld area, often giving rise to plateaux and smaller terraces on the slopes of the Nieuweveld, and a finer-grained rock that is in thinner beds and often weathers with a red crust. The first variety is called " defining sandstone,'* ^ and the second ** intermediate sandstone,** on account of the usual relative positions of the two rocks in the plateau caps and in the slope between the terraces respectively. This difference in position is due to the weather-resisting qualities of the rocks ; the thick sand- stones last longest, and therefore cap the larger terraces, while the intermediate sandstones make smaller ledges on the mountain sides, and shales and mudstones lie between them. The sandstones are often false-bedded, and may have their surfaces ripple-marked. The shales

* Schwarz, Oeol. Conim. (96), p. 15.

190 GEOLOGY OF CAPE COLONY

and mudstones are usually dark-blue or greenish in colour, but thin beds of purple and red shale are not infrequent.

In the more argillaceous beds of the Beaufort series there are concretionary nodules and lenticular layers of blue-black limestone often containing small veins and pipe-like rods of chalcedony, white or pink in colour. The rods are occasionally branched, and seem to be due to the silicification of small roots. Both the limestone and chalcedony are often found permeating the large fossil bones of PareiasauruSy or other reptiles, and are certainly closely connected in their origin with the presence of organic matter, just as is the case with the flints in the chalk of Europe. Very rarely small bivalve shells have been found inside the limestone nodules in the Beaufort beds. In many of the flat areas in the Karroo, where a considerable thickness of shale or mud- stone has been weathered away, the ground is strewn with large numbers of the nodules. On the outside they have a peculiarly roughened surface, from which the veins and other forms of chalcedony stand out promin- ently. The nodules can be seen near any of the railway stations between Groot Fontein and Beaufort West ; they are like the Ecca nodules, but the latter do not contain chalcedony. Both the nodules and the thin lenticular beds have been formed by the concentration of the carbonate of lime, distributed generally through the sediments, since the latter were laid down. Beds of clay-pellet conglomerate are frequently met with at the base of the sandstone bands in this series, and less frequently in the Ecca beds. The clay-pellet conglomer-

THE KARROO SYSTEM 191

ate is a rock with a shale or mudstone matrix containing numerous rounded or flattened lumps of mud rather dif- ferent in colour from the matrix, but otherwise of much the same nature. At places the matrix is more sandy than usual, and the mud-pellets are in consequence more conspicuous, for they weather away more readily than the rest of the rock. The lumps of mud were derived from previously deposited sediment, and were rolled along by the current till they came to rest where they are found. In many rivers which vary in level, either daily on account of the tide, or at irregular periods owing to varying rainfall, mud-pellets may be seen on the muddy or sandy bottom exposed at times of low water. The tidal lagoons of the Eastern Province rivers, and the lower part of the 01ifant*s Eiver (Van Ehyn's Dorp), are good places for the observation of mud-pellets due to daily changes of water-level, and the Orange Eiver, near Prieska, has many sandy stretches along its banks ex- posed during dry seasons and covered with mud-pellets brought down by the last flood. There is no doubt that mud flats, exposed at the surface of a shallow lake or sea, would furnish lumps of mud to the small waves washing their margins, and it is probable that the clay-pellet conglomerates in the Karroo formation were formed in this way. Possibly the deposited silt could become tenacious enough to resist complete disintegra- tion without being exposed to the air, and yielded the pellets to currents that were stronger than usual sweep- ing through the shallower parts of the basin.

Local unconformities affecting the beds over small areas, sometimes only a few yards wide, are very abund-

192 GEOLOGY OF CAPE COLONY

ant in the Beaufort and Ecca beds. The lower lying strata are cut oflf by the upper to the depth of, perhaps, four or five feet, usually less, and the higher beds thicken out to occupy the depression made in the lower. These hollows are usually in shales or mudstone, and the rocks filling the hollows are sandstones or clay-pellet conglomerates. The frequency of these examples of ** contemporaneous erosion and deposit *' point to the deposition of the strata in quite shallow water which from time to time received sudden accessions from rain floods, or possibly also had strong streams developed in it by a constant wind.

The clay-pellet conglomerates in the Beaufort series frequently contain rolled pieces of bone. Pebbles of rock are very rare both in the conglomerates and the other strata, and the few that have been found do not reach a length of two inches.

Coal has been found in thin layers in the Beaufort beds. Behind the Komsberg escarpment on the farm Lange Kuil a nine inch seam of bright coal occurs, but it is unsuitable for burning* under ordinary conditions, as it crumbles immediately one attempts to make a fire with it owing to its large content of water. This coal has a small percentage of ash, 68 per cent. It occurs in beds containing large fragments of bone, probably of Pareiasauncs, Vague reports are sometimes forthcoming of coal near the base of the Nieuweveld, in the highest part of the Gouph, and also in the Pareiasaurus beds further south. Nothing has yet come of these reports, although the country is one that is very easy to prospect in owing to the extensive exposures of the rocks. High

THE KARROO SYSTEM 193

up in the Nieuweveld escarpment at Leeuw Eiver Poort, and also in the Camdeboo, there are some remarkable vertical cracks filled with bright bitumin- ous coal.^ The Leeuw Eiver Poort fissure is over 300 feet deep, and varies in width from twelve feet downwards. The fissure does not maintain a straight course, but at places runs horizontally or at a low angle. It passes through a few thin horizontal seams in a band of sandstone, but the thickest seam is about an inch thick. The coal is remarkably free from ash, an analysis giving only 8 per cent. The fissure seems to have been produced during the intrusion of the dolerite sheets which occur on the Nieuweveld, and the bituminous coal was probably partly squeezed and partly distilled into it at the same time. Although slickensided surfaces in the coal near the edges of the crack prove a slight movement to have taken place after its formation, there is no appreciable vertical displace- ment of the rock outside the fissure. The coal at BulBfers Kloof, Camdeboo, occurs in a similar manner, and no seam worth working has been met with there. Thus jklthough there is coal in the Beaufort beds it has not yet been found in sufficient quantity to work. The re- ports of coal at Tamboer's and Ongeluk's Fonteins in the Gouph, and at Lett's Kraal at Graaff Beinet are based upon the occurrence of carbonised wood in fragments.

The base of the Beaufort beds is the lowest stratum containing the remains of Pareiasaurus or other reptiles given in the list below. Where these are absent, as in

^ Dunn (79) ; Schwarz, Oeol. Comm. (97), p. 24 ; and for a similar occurrence in East Griqualand, Geol, Comm, (08), p. 16.

13

194 GEX)LOGY OF CAPE COLONY

the western part of the Eoggeveld, the Klein Eoggeveld, and, so far as we know, in the country north of Fraserburg and Victoria West, some other means have to be devised for separating this series from the Ecca. In the Moor- denaar's Karroo and Klein Eoggeveld a bed of red weathering sandstone has been taken as the base, and in the Eoggeveld (Fish Eiver Valley) a thick band of sandstone different from any that occurs in the Ecca beds in the same district. The line as laid down on

8. N.

Wlttebergen Klein Roggeveld

Fig. 16. — Section from the Wittebergen to the Klein Roggeveld, from the folded belt to the Karroo basin. Distance about 13 miles. Vertical scale ^ in. to 1,000 feet.

1. Witteberg series.

3. Ecca series.

4. Beaufort series.

the map accompanying this volume is of little real significance except . in the Great Karroo, where the boundary is fixed on palaBontological grounds. The northern portion of the boundary is practically unknown. In the Eastern Province A. H. Green described an un- conformity which may be at the base of the Beaufort beds near Aberdeen/ but there is nothing known in the west corresponding to this unconformity. On a rapid traverse through the Gouph or southern Karroo the re- markable change of dip which takes place at the south

^ Green (83), p. 25 ; (88), p. 261.

THE KARROO SYSTEM 195

of the Klein Eoggeveld (see Fig. 16) and along the same line farther east may be mistaken for an unconformity, but the appearance is due to the sudden cessation of the folds north of the Zwartebergen. Pinchin^ records a marked unconformity north of Port Elizabeth at about the same horizon as that described by Green, but farther east in the Transkei it has not been seen. The true significance of these observations must remain uncertain till the Eastern Province has been connected with the Western by means of systematic mapping.

The Beaufort series can be divided into three groups characterised by various reptilian genera, but at present the classification is not very satisfactory, especially towards the upper limit. No lithological characters distinguishing the three groups have been made out. The chief fossils and some of the localities from which they have been obtained are the following : — '^

Localities. Theriodontia ' —

GynognathuSj Seeley - - Lady Frere.

GomphognathtiSy Seeley - Burghersdorp.

Microgomphodorif Seeley - Aliwal North and Burghersdorp.

DiademodoTiy Seeley - - Aliwal North and Burghersdorp. Stegocephalia —

Bhytidosteus, Ow. - - Beersheba, Orange River Colony.

BatrdchostLchiLS, Br. - - Aliwal North. Anomodontia —

IHcynodon latifronsj Br. - Burghersdorp and Aliwal North.

-»3

o

1 Pinchin (74), pi. iv.

^ I have to thank Professor B. Broom for correcting this list and for giving me the classification of the reptiles. The localities as a rule refer to Divisions and not villages.

'For references to the literature of the Beptiles see Owen, Seeley, Huxley, Broom, Lydekker, in the Appendix.

* See note at end of chapter.

13*

196

GEOLOGY OF CAPE COLONY

( Therooephalia — j^lurosaiiriLSj Ow. GynodracOy Ow. LycosaurtiSj Ow. GynosiLckuSy Ow. Gynochampaa, Ow. Tigrisuch^is, Ow. LycosuchuSf Br. IctidosuchtLS, Br. Ididosaurus, Br. ScymnosawnMf Br. Scylacosaurus^ Br. Scalopomurus, Ow. Gorgonops, Ow. Anomodontia —

Dicynodon, Ow.

.2

•iH

Oudenodon, Ow.

KiMecephaliis, Ow. - Endothiodon, Ow. TheriognathuSf Ow. - Esoterodon, Seeley. - GryptocynodoTif Seeley. Pristerodoriy Huxley - Opisthoctenodorif Br. - Lystromurus, Cope - (= Ptychognathus, Ow.)

Theriodontia— GalesauriiSf Ow.

Procolophonia — ProcolophoTiy Ow.

Lacertilia — Paliguanaj Br. -

Rhynchocephalia — Saurosternon, Huxley

Stegooephalia — Micropholis, Huxley Bothriceps, Huxley -

1

Beaufort West.

Sneeuw Berg, Fort Beaufort.

Kriga Berg, Fort Beaufort.

Sneeuw Berg.

Rhenoster Berg.

Sneeuw Berg.

Aberdeen and East London.

Pearston.

Beaufort West.

Beaufort West ?

Beaufort West ?

Sneeuw Berg.

Fort Beaufort.

{Beaufort West, Fort Beaufort, GraaffBeinet, Cradock. I Beaufort West, Fort Beaufort, 'East London, Sneeuw Berg,

Graaff Reinet. Sneeuw Berg. Beaufort West Sneeuw Berg.

Molteno Pass, Beaufort West. Nieuweveld. East London.

Pearston and Beaufort West. Cradock, Bethulie, Sneeuw Berg.

Rhenoster Berg.

Tarka, Middelburg.

Queens Town.

Sneeuw Berg.

Rhenoster Berg. Orange River Colony.

THE KARROO SYSTEM

197

'^ rFish—

Atherstonia, S. -Woodward

HO

c>

eS O

PalcBoniscicSj Agassiz Lamellibranchs —

Palasamutela, Amalitzky •

Palceanodonta^ Amalitzky Plants—

Schizoneiira

Glossopteris ( Pareiasauria — Pareiasaurus, Ow.

Therocephalia —

TapinocephaluSy Ow. - Titanosuchus, Ow. p§ ■{ DelphinognathuSj Seeley PristerogiiafhiLS, Seeley

Plants— Schizonenra

ee

o

1^

Glossoptei-is

Oolesberg and Fraserburg. Sneeuw Berg.

Graaff Reinet, Bedford. Graaff Reinet.

Sutherland, Beaufort West, Bethulie, Cradock, Pearston, etc.

Beaufort West, Fort Beaufort, Prince Albert.

Gouph. Gouph.

Prince Albert. Gouph.

Beaufort West, Fort Beaufort,

Sutherland. Beaufort West, Fort Beaufort,

Sutherland.

The distribution of these three divisions is only known in its barest outlines. The Lower Beaufort beds form the western part of the Roggeveld Plateau, the whole of the Klein Eoggeveld, the northern part of the Moordenaar's Karroo and Gouph, and they probably stretch from Aberdeen past Somerset East, Bedford, Fort Beaufort to the coast south-west of East London, and are perhaps represented in the Transkei by the Idutywa beds.

The Middle Beaufort beds form the higher portions of the Nieuweveld, the Sneeuwbergen, the country north of the Sneeuwbergen as far as Colesberg and Bethulie, and southwards to East London, where they

198 GEOLOGY OF CAPE COLONY

range into the Transkei and are in part represented by the Idatywa beds. Of the Hmit between the Middle and Lower Divisions between the Nieuweveld and the Orange Eiver nothing is known.

The Upper Beaufort beds crop out below the coal- bearing Molteno beds of the Stormberg series both to the north and south of the Stormberg region, but the details of their distribution are unknown.

The foreign equivalents of the Beaufort series can be given approximately only. In the Panchet beds of the Indian Gondwana system Dicynodon and Ptychosiagum, two Cape genera, have been found, and with them are plants belonging to the Glossopteris flora, especially Glossopteris and Schizoneura ; in the Panchet beds there are also some genera, of which Thinnfeldia is the most important, that in the Colony are found only in the Stormberg group. In New South Wales the Newcastle beds may represent the Beaufort as well as the Ecca beds. Perhaps the naost interesting comparison can be drawn between the Beaufort fauna and flora and those of the Permian formation of Eussia. PalcBomutela and PalcBanodonta are two genera of probably fresh water mollusca that are conunon to the Eussian and South African beds; of the first-named genus four species from the Karroo beds were determined by Amalitzky to be identical with Eussian forms, viz. : P. trigonalis, P. semilunata, P. murchisoni, and P. plana, while seven other species are very closely allied to others from Eussia; of Palceanodonta two species are common to the two formations, P. okensis and subcastor^. Amalitzky has

1 Amalitzky (96), pp. 337-51.

THE KARROO SYSTEM 199

recently ^ found Pareiasaurus and Dicynodon in lenticular beds within the Permian strata on the Dwina Eiver that also contain Glossopteris and Gangamopteris, Both above and below the horizon on which these character- istic Karroo genera occur there are limestones contain- ing marine species of Permian age, belonging to a stage widely developed on the continent of Europe and known as the Zechstein. These discoveries go far towards settling the age of the Beaufort beds

relatively to the European rocks.

•

The Stormberg Series.

In the north-east of the Colony and in Basutoland there is a great area of shales and sandstones capped by volcanic rocks and broadly distinguished from the under- lying Beaufort beds by the presence of a different group of fossil plants. Instead of the Glossopteris, which is so widely distributed through the lower rocks, the genera Thinnfeldia and Tceniopteris now appear. The name Stormberg beds was applied to these upper rocks by Wyley ^ and Huxley,^ and it has been used by all later writers. The series is divided up into the following groups : —

Maximum Thickness. Volcanic beds - - 4,000 feet.

Cave sandstone - - 800 feet.

Red beds - - - 1,400 feet.

^Molteno beds - - - 2,000 feet

Stormberg series

» Amalitzky (00). « Wyley (59), p. 61.

^Huxley (67), p. 6.

200 GEOLOGY OF CAPE COLONY

The Molteno Beds.

The Molteno beds form the lower slopes of the Strom- bergen and Drakensbergen and the country along the foot of the range. The exact position of their base has never been defined, but, as in the case of the Ecca- Beaufort junction in the west, it seems to be a gradual passage. Glossopteris has not yet been found in the Stormberg region, although both to the south and north the genus occurs in shaly beds on a lower hbrizon. Similarly Thinnfeldia, Taniopteris, and Stenopterts are not known from the Beaufort beds. Whether a detailed ex- amination of the passage beds will show a clearly defined junction or an intermingling of the two sets of plants remains to be seen.

The beds consist of shales, mudstones and sandstones. The shales and mudstones are very like those of the Beaufort and Ecca beds, but they do not contain the calcareous concretions so abundant in the lower groups. They are usually grey or greenish in colour, sometimes bluish purple, and in places contain abundant plant re- mains. The localities from which most of the fossil plants hitherto discovered in these beds were obtained are Indwe, Molteno, Cyphergat, Maclear, the Kenigha Eiver, in Mount Fletcher and the Matatiele slopes of the Drakensberg, but as the fossils appear to be more numerous in the Molteno beds than ' in any of the lower beds, it is probable that they will be found to be widely distributed on both sides of the Drakensberg-Storm- berg ridge. The sandstones of the Molteno beds are

•w __	~ ~ ~ " ~ m
i
1 (a 1	i'
i
l t	1 11

"II

.s.:k

202 GEOLOGY OF CAPE COLONY

unlike any that occur in the lower groups of the Karroo system. In general appearance and in the character of the surface to which they give rise they resemble the Table Mountain sandstone more closely than any other in the Colony, but they are coarser in grain and looser in texture than that rock, and do not form such thick masses. In some localities the quartz grains are coated with a later deposit of quartz with more or less perfect crystalline faces which reflect light well and give to much of the rock a glittering appearance in sunlight. Felspar grains are abundant in these sandstones, as they are throughout the sandy beds in the Karroo formation, but the looser texture of the Molteno sandstones has allowed the felspar to weather considerably, and the dull white grains of weathered felspar are always conspicuous constituents of the sandstone. Bounded or spherical nodules, hollowed out in the centre when the hard outer shell has been broken through, are quite a characteristic feature of the Molteno sandstones. The hard shell is due to the addition of hydrated iron oxides to the cement- ing material usually present. The nodules are forrned by the oxidation of pyrites and the deposition of some of the resulting iron compounds in a spherical zone about the lump of decomposed pyrites.

The finer-grained varieties of sandstone are good building stone, easily worked and of a yellow or cream colour. Fencing poles are split from the large sandstone slabs by driving in wedges along straight lines across the slab and breaking it along the rows of holes. Posts up to six feet in length are thus obtained.

The sandstones do not contain so many fossil plants

THE KARROO SYSTEM 203

as the shales, and the fossils are less well preserved than in the latter.

Thin beds of conglomerate occur in connection with the sandstones in the Molteno area, usually with red ferruginous nodules that give the rock a characteristic appearance on the outcrop. This rock is found a short distance above coal seams in several parts of the district and is an aid in the search for coal.

Coal is found in the Molteno beds from the Storm- bergen along the lower slopes of the Drakensbergen in East Griqualand as far as the Natal borde]^, and also on the northern slope of the watershed, although it is only in the Stormberg-Indwe region that any serious work has been done on the seams. ' The whole area has not been surveyed yet, but the work already done in various parts is summarised here.

There seem to be two horizons on which workable coal has been found ; the lower extends from Sterkstroom eastwards to Indwe, beyond Indwe towards Engcobo the coal has been followed but not worked to any extent ; the upper is that to which the Molteno seams belong, and is on a horizon some hundreds of feet higher than the Indwe coal ; its extent is not well known beyond the neighbourhood of Molteno, but it may be represented by some thin coal seams seen in the Gala pass some 300 feet above the Indwe coal. In the Indwe district ^ the base of the Molteno beds is taken at the bottom of a band of bright-coloured felspathic sandstone, which lies upon red, purple, and green shales and mudstones be-

1 Du Toit, Geol. Ctmim. (03).

204 GEOLOGY OF CAPE COLONY

longing to the Beaufort series. The typical Stormberg plants have not been found in these argillaceous beds, which contain bones that have not yet been collected or described from the Indwe district, though it is not im- probable that some of the reptiles from the Albert Division belong to this horizon.

Above the felspathic sandstones lie sandy shales and thin sandstones with a total thickness of some 700 feet, containing Thinnfeldia, Stenopteris, Gallipteridium, TcBniopteris and Schizoneura, Towards the top of this group of argillaceous rocks come the Indwe coal seams. The coal seams are rarely over twelve inches in thick- ness, but at places several occur together, so that in a band of rock composed of coal and shale, six feet thick in all, about four feet of the whole may be coal, which has of course to be picked out from the accompanying shale before it is removed from the collieries. The number of the seams varies within short distances owing to the sandstone, which usually forms the ** roof,'' cutting across one or more of the coal beds, a state of things that was brought about by the erosion of the coal shortly after its deposition, and which is paralleled by thousands of cases of ** contemporaneous erosion and deposit " throughout the Ecca, Beaufort and Stormberg series. The coal is usually laminated and contains very thin layers of silt ; it is a coal that was formed at perhaps a considerable dis- tance from the spot where the plants that furnished the vegetable matter grew, for there is no trace of a land surface on which the coal plants grew, and the alterna- tion of thin layers of coal and silt evidently point to the vegetable matter having been deposited over the floor of

THE KARROO SYSTEM

205

the lake in the same manner as the silt. It is this silt that accounts for the high percentage of ash or incom- busfible matter in the Btormberg coals. ^

The abundant intrusions of dolerite in the form of dykes and sheets, especially the latter, have an injurious influence on the coal. The distance through which this influence makes itself felt varies ; the chief effect is the driving off of the more volatile constituents, and it culminates in the coking of the coal, which is rendered valueless.

The insertion of a few analyses of the coals, taken from the ofl&cial Eeports referred to on a previous page, may be of use in indicating the class of coal to which the Colonial seams belong.

	Molteno (mean).	Cypbergat.	Indwe.	Sterk- stroom.	Matatiele.	Gala.
Moisture Volatile Hy- drocarboTiH - Fixed Carbon Ash Sulphur - Total -	113 10-31 60-89 28-80 •76	i 28-24 50-07 21-69	12-54 63-03 24-42	18 26 51-38 30-36	1-37 .24-68 47-53 25-10 1-33	1-50 9-50 68-51 19-70 -79
101-89	10000	99-99	100-00	100-01	100-00

From the results of numerous experiments it has been concluded that the ratios 1 to 1*5 and 1 to 1'83 represent the weights of Welsh and Stormberg coals required to be burnt in order to do a given amount of work.^ A

^ For detailed information about the coals of this region, see Dunn (78), North (78), Green (83), Galloway (89). a Galloway (89).

206

GEOLOGY OF CAPE COLONY

rock allied to torbanite (oil shale), occurs below a coal seam in Matatiele ^ and in other parts of South Africa ; the following analyses, together with that of the rock from Torbane Hill in Scotland, will show the nature of the substance : —

	NatAl, Upper Umzimkiuu.	Matatiele.	Basutoland.	Torbane Hill.
Moisture - Volatile Hydro- carbons - Sulphur Coke - Ash - Total	1-58 16-30 12-07 7006	1-32 18-16 •89 32-37 47-26	I 3400 16-66 49-34	7010 10-30 . 19-60
10000	100 00	100-00	100 00

The Ked Beds.

The Molteno beds pass upwards conformably into a group of strata that is distinguished from them by its prevaiUng red colour. The name was first. used by Mr. Dunn who described the group in the Stormberg area.^ The Eed beds have been found to extend through East Griqualand, though with varying thickness. PalsBonto- logically they are separable from the Molteno group by almost entirely negative characters, for the comparatively rich flora known from the latter has no representatives in the higher strata so far as our present knowledge goes. Some reptilian bones, as yet undescribed, have

1 Schwarz, Oeol. Comm. (03), pp. 21, 22.

^Duun (78). Other sources of information concerning this and the succeeding group are : Schwarz, Geol. Comm. (02) ; Du Toit, Oeol, Comm. (03).

THE KARROO SYSTEM 207

been found in them, but silicified wood is the only other fossil known from these rocks.

Eed-coloured strata are by no means confined to this subdivision of the Stormberg series ; similarly coloured rocks are found both in the Molteno beds and the Cave sandstone. The Red beds cannot be regarded as of more than local importance, and it is often difficult to decide where the boundary lines between the three groups should be drawn.

The most characteristic rocks of the Eed beds are purple and red mudstones and shales, but red sand- stones and thick beds of yellow and white felspathic sandstones are also present. The thick ** glittering" sandstones of the Molteno beds do not occur in this group. Bands of blue or green mudstones are not un- common. Conglomerates, though rare, are not entirely absent ; the pebbles are of white quartz and quartzite.

Mr. Dunn records 600 feet of Red beds in the Storm- berg area. In Elliot they reach a maximum of 1,400 feet, and in Matatiele they dwindle down to 200 feet. It is obvious that in the case of a group of rocks which cannot b6 very closely defined, different observers are likely to include different strata under one head, but in spite of this there is certainly a thinning out of the Red beds and of the overlying Cave sandstone towards the north-east on the East Griqualand side of the Drakensberg.

The Cave Sandstone.

The Red beds pass upwards into the Cave sandstone, ^s a rule without any sharp line of demarcation. The

208 GEOLOGY OF CAPE COLONY

Cave sandstone is an extraordinarily massive rock with bedding planes feebly developed. The sandstone is largely made up of quartz grains; grains of felspar (mostly microcline) are fairly abundant, and tourma- line, zircon, white mica and hornblende are also present. Generally the rock is white or grey in colour on exposed surfaces, but on a fresh fracture it has a reddish tint. Bands of red sandstone occur in this formation, and are in no way different from the sand- stones of the Eed beds.

In the Stormberg area the Cave sandstone is about 150 feet thick, in Elliot 800 feet, and in Matatiele it decreases again to a maximum of 130 feet. At certain places, as in the north-west of Elliot and in the northern part of Matatiele, the Cave sandstone is not present ; it thins out owing to denudation which took place just before the volcanic outbursts, so that the lavas of the volcanic group rest directly upon the Ked beds.

Fossils are very rare in this rock, the only finds re- corded from the Colony being fragments of reptilian bones. In the Orange Kiver Colony, however, fish (Cleithrolepis and Semionotus) have been described from the Cave sandstone of the Smithfield district.^

The Cave sandstone gives rise to very remarkable features on the slopes of the mountains and on the top of several spurs projecting from the main ridge. It tends to weather into huge pillars and irregularly shaped masses, often with the lower portion hollowed out to form a shallow cave, a characteristic that gave the rock its name. Such rock-shelters were frequented by bush-

^ See note at end of chapter.

.w.'a'f'a'.w.w.'S'

¥

11

I

I ^1

ll

1^*

210 GEOLOGY OP CAPE COLONY

men, whose former presence is indicated by ag9.te chips, fragments of ostrich shells and coarse pottery, and espe- cially by more or less realistic sketches of men and animals done in red and black colours upon the pale surface of the rock.

Above the village of Elliot the hard yellow sandstone forms buttresses and pillars over 300 feet high. The outcrop of the Cave sandstone can easily be distinguished at a distance of many miles by its colour and broken appearance. There is no rock in the country that pro- duces such peculiar features as the Cave sandstone where typically developed (see Plate XII.).

Thb Volcanic Group.

Before the close of the period represented by the Stormberg sedimentary rocks volcanic activity com- menced in the north-eastern part of the Colony. From the neighbourhood of Molteno the volcanic rocks stretch far to the north-east through Basutoland and along the Natal boundary perhaps as far as the Transvaal; but very little information is as yet available on this ques- tion, and it refers to only a small proportion of the whole volcanic district.^

The volcanic rocks form the highest parts of the country in which they occur. The crest of the Drak- ensbergen is carved out of them for a great distance, and the high ridges in Basutoland that are admirably

^The following papers are the chief sources of information on this volcanic group : Cohen (76) ; Dunn (78) ; Churchill, (Natal) (98) ; Schwarz (03) ; Schwarz, Oeol. Comm. (02) ; Du Toit, Geoh Comm. (03).

THE KARROO SYSTEM 211

displayed from many points on the Matatiele border are evidently of the same nature.

On the Colonial border the volcanic rocks rarely reach 3,000 feet in thickness, but in the ridge of the Malutis (Basutoland) north of N'quatsha*s Nek there must be quite 4,000 feet of them, and Mr. Churchill measured a vertical thickness of 4,500 feet on the Mont aux Sources.

By far the greater part of the group is formed by lava streams. Bedded agglomerates and tuffs are quite subordinate features in those districts that have been examined.

In the district of Elliot near the Tembu Pass there is an interesting section showing the following succession of beds from above downwards : — ^

6 Bedded lavas 350 feet.

4 Purple and stratified ash - - - ^ n

3 Cave sandstone 30 „

2 Bedded lavas 50 „

1 Cave sandstone 700

>»

The lavas (No. 2) are very vesicular at the base but become doleritic a few feet from the junction with the underlying sandstone. The sandstones No. 3 pass into the volcanic ash lying above them. The lower lavas probably came from a vent exposed on the farm Mountain Cliff, and they have been traced over a mile between the two parts of the Cave sand- stona The ash beds No. 4 have been traced to a large vent on the farm TuUoch near the Barkly Pass ; towards the east they thin out, and the lavas No. 5

» Du Toit, Geol. Comm, (03).

14*

212 GEOLOGY OF CAPE COLONY

rest directly upon the Cave sandstone which is no longer divided into two portions by the lower group of lavas.

Other thin lenticular beds of ash have been found interbedded with the Cave sandstone in the Elliot Division. During his recent investigation of that area Mr. du Toit came to the conclusion that the earliest volcanic eruptions there took place under water, and that the intercalations of ash beds with the Cave sand- stone represent breaks in the continuous deposition of the latter, during which its usual characters were masked by the abundance of volcanic debris.

In the western. part of the Elliot volcanic area, under the Xalanga Peak, Mr. du Toit found that the lowest lavas rest upon the Cave sandstone for a certain distance and then pass downwards at a slight angle over an apparently eroded surface of that rock till they rest directly upon the Bed beds. To the north-east of this locaUty the same geologist found a band of red sand- stones and shale rather under 50 feet thick, intercalated between the two lower groups of lavas for a distance of some ten miles round the head waters of the Qokama Eiver. The lava below the red sandstone band lies upon the Eed beds. Mr. du Toit considers that this part of the country was disturbed by local earth move- ments at the commencement of the volcanic epoch, and that the lower portion of the Cave sandstone was removed by erosion over a certain area during the deposition of the upper portion of the same rock in other parts of the district. A band of Cave sandstone fifty feet thick occurs above the second group of lavas between the Washbank and Xalanga Peaks.

THE KARROO SYSTEM 213

In the Stormberg area, as described by Mr. Dunn in the report referred to on a previous page, the proportion of ash beds to lavas is somewhat greater than in Elliot, and some of these beds were formed under water before the close of the Cave sandstone stage.

From the most westerly point from which the volcanic group has been described as far as the north-east of Elliot, with the exception of the western part of Elliot mentioned above, the Cave sandstone lies between it and the lower groups of the Stormberg series. In the north-eastern part of Matatiele the volcanic group again rests directly upon the Red beds over a distance of some four miles. From the evidence gathered during his sur- vey of Matatiele Mr. Schwarz came to the conclusion that the Cave sandstone itself was partly volcanic in origin, for he obtained fragments of lava from that rock at Zureka, and the Cave sandstone appeared identical in general character and in the nature of its component minerals with the rock filling some of the volcanic vents in the same district.

In the Matatiele Division the development of true ash beds between the lava flows is very restricted. On the crest of the mountain behind the farm Eyrie there are two bands of sandstone and shale, forty and twenty feet thick respectively, intercalated between thick flows of lava. The section through the upper part of the mountain at this locality is, in downward order: —

Lavas 130 feet.

Sandstones and shales (red) - - - 20

Lavas -70

Shale 40

Lavas - - 630

Cave sandstone . _ - . . 100

214 GEOLOGY OF CAPE COLONY

The red sandstones consist of fragments of altered glass and other rocks of volcanic origin mixed with grains of quartz, microcline and zircon, probably derived from the same source as the materials composing the Cave sandstone. Such beds as these can be regarded as partly of ordinary detrital origin and partly volcanic, although it is of course difficult in the absence of large lumps of lava (bombs) to be certain whether the volcanic material in the rock came directly from a vent or whether it reached its present position through the ordinary agents of denudation. They undoubtedly were deposited under water, and thus support the evidence already quoted to that effect.

So far as we have information about the volcanic group in Natal tuffs are of very rare occurrence there.

The lavas ^ are basaltic in composition and vary very much in outward appearance according to their struc- ture. The glassy varieties are amygdaloidal and usually much altered, a circumstance that makes them less conspicuous than the doleritic lavas (see Plate XII.), for they weather more rapidly and give rise to debris- covered slopes on the mountain sides rather than to krantzes. The mineral components are similar in all the varieties, though the proportions in which they are present differ. The felspar is labradorite or an allied variety as is the case in the dolerite intrusions ; most of the augite is colourless and resembles that of the intrusive dolerites ; olivine is often present either fresh or more or less changed to serpentine ; these

^ For descriptions of the various varieties, see Schwarz, Geol. Contm. (02), pp. 65-9G.

THE KARROO SYSTEM 215

three minerals are the most important constituents ; magnetite is always, and apatite often present in the lavas, and occasionally a green augite. Serpentine, epidote and calcite are the usual alteration products. Glass is found in several varieties of the lavas. Mr. Schwarz lays stress upon the absence of brown mica and original hornblende from the Matatiele lavas, for these two minerals are frequently present in the intru- sive dolerites, though usually in small quantities. On this ground he regards the volcanic rocks as belonging to a distinct phase of igneous activity from the dolerites so abundant throughout the central and eastern parts of the Colony.

The differences between the varieties of lava depend upon the amount of glass present and the relations of the augite and felspar to each other. The glassy lavas are basalts with a greater or less amount of glass and microcrystalline base in which lie more or less well- formed crystals of olivine, felspar and augite. The doleritic lavas may have a very small quantity of residual glass, the felspar is either in fair-sized porphyritic crystals, between which small felspar and augite crystals lie, or in smaller crystals often enclosed by ophitic masses of augite. The last-mentioned type of rock is very similar in structure to the dolerite of the thick sheets and dykes elsewhere in the Colony, and the other variety of doleritic lava is like the dolerite of the smaller sized intrusions, with the exception of the presence of brown mica and hornblende.

The amygdaloidal varieties of lava are almost entirely basalts. The steam holes have in places never been filled

216 GEOLOGY OF CAPE COLONY

in so that weathered out to the rocks they are scoriace- ous. The minerals fiUing these cavities are calcite, chalce- dony, or zeolites, amongst which heulandite, thomsonite, stilbite, and perhaps scolecite, have been recognised ; a green layer of chlorite or delessite sometimes lines the cavities which have been filled in with one of the above- mentioned minerals. The amvfifdales may be more or less spherical in shape or irregL. In certain lavas there are pipe-like amygdales, four or five inches long and often branching upwards. They are found in zones near the base of the flows, separated from the under- lying rock by a few inches of compact or vesicular lava in which the steam holes are of the usual type, and they are approximately perpendicular to the floor.

In the Stormberg district, Elliot and Matatiele, the only parts of this volcanic region that have been closely surveyed, numerous necks of agglomerate and lava have been described.

Mr. Dunn describes Telemachus Kop near Molteno as a crater filled with an agglomerate of many varieties of lava and sedimentary rocks, the latter being highly altered by heat. It is certain that the crater form of this and the few other volcanic pipes which show it is due entirely to erosion and weathering long subsequent to the period of activity. He mentions in his report, or places on his maps of that region, five pipes near Molteno and Jamestown. There are sixteen volcanic necks exposed in the Elliot Division. They are at various distances up to about four miles from the main ridge of the Drakensberg, and are differently situated with regard to the surrounding beds according to their

THE KARROO SYSTEM 217

distance from the main ridge. The necks farthest from the ridge are in the Eed beds, and those nearer to it are surrounded by the Cave sandstone or the lower lavas. They vary in size from fifteen yards in diameter to an area one and a half mile long by a quarter wide (the TuUoch volcano). In some cases lava streams have been traced to a certain vent, but generally denudation has proceeded so far that the original connections have long since been destroyed, and there is consequently little evidence to indicate from which vents the great sheets of lavas, piled up to a thickness of over 2,000 feet in the Washbank peak, came.

Some of the small necks are plugged with dolerite lava, but as a rule the pipes are now filled with a bluish tuff or agglomerate containing fragments of sedimentary rocks and lavas ; these tuffis weather white and some- times look like outliers of the Cave sandstone from a distance. A large neck near the top of the Gat Berg is entirely plugged with dolerite. It is often found that the necks are partly filled by lava and partly by agglomerate.

Dykes of dolerite have traversed some of the Elliot necks, and they occasionally traverse the lava flows. In this area no great fissures through which the lavas may have reached the surface have yet been found, but a survey of the whole breadth of the volcanic band may reveal their presence.

In Matatiele Mr. Schwarz found at least nineteen dis- tinct vents, of which only one lies on the crest of the Drakensberg ; the others are all within seven miles of the highest ridge on the East Griqualand side of it.

218 GEOLOGY OF CAPE COLONY

Whether the volcanoes are confined to the high ridges of volcanic rocks, or whether they are spread broadcast over Basutoland is not yet known.

The largest of the Matatiele pipes is on the farm York, it is about a mile in diameter, and it has been cut in two by a tributary of the Mabele Eiver. The vent is filled partly with amygdaloidal and doleritic lavas, and partly with agglomerate. The dolerite was the first rock to flow from the pipe, and it is still con- nected with a columnar flow of dolerite that lies upon the Cave sandstone. The doleritic rock was succeeded by amygdaloidal lavas, part of which are still preserved in the lava flows, 4,000 ft. thick, near Ongeluk's Nek. Near the volcano the lava contains large masses of sand- stone and shale baked and converted into porcellanite by the heat of the lava. There are some baked shales that Mr. Schwarz regards as having been formed in tempor- ary lakes or streams on the volcano itself, and subse- quently hardened by fresh flows of lava. Brown, gritty soil is preserved between some of the lava streams that issued from this vent, indicating that the volcano, even if it started its activity below the water level, piled up its lava sufficiently to form a land surface. The ag- glomerate is dark blue in colour, and includes large numbers of fragments of lavas and sedimentary rocks ; this material is probably the result of the final explosive outburst of the volcano. Evidence of the long duration of the activity of this vent is given by the old valleys carved out of some of the lava flows and filled in by later ones.

The smallest volcanic neck in this district is only

THE KARROO SYSTEM 219

four yards across, but most of the others are over 100 yards in diameter. The majority are filled with ag- glomerate, of which the matrix is largely composed of quartz grains derived from a sedimentary rock or a granitic one, as both orthoclase and microcline are abun- dant; these are felspars which do not occur in the Drakensberg lavas. Zircon, rutile, hornblende, tourma- line, muscovite and garnet, all minerals that are foreign to the lavas, are also present. With these minerals occur others, plagioclase especially, that are important con- stituents of the lavas, of which both large and small fragments are frequently embedded in the agglomerates. Pieces of charred wood have been found in some of the agglomerates ; they are the remains of trees that grew on the slopes of the volcanoes during periods of quies- cence ; on a renewal of activity, fragments of these trees fell into the crater, and were imbedded in the breccias composed of comminuted volcanic and sedimentary rocks. Taking into consideration the great thickness of lavas in this portion of the Drakensberg, the absence of more normal agglomerates from the necks is certainly remark- able, and gives the vents a character somewhat similar to that of the peculiar pipes of Kimberley, Sutherland and other districts in the Colony, which will be described in a later chapter. Some of the pipes of the Kimberley type, however, contain melilite-basalt, a rock which is entirely unrepresented in the explored parts of the Storm- berg volcanic series ; and the age of the Kimberley type of vent is probably much later than that of the Storm- berg volcanoes. None of these later pipes is known to have given exit to lavas which flowed at the surface.

220 GEOLOGY OF CAPE COLONY

Mr. Schwarz came to the conclusion that a consider- able thickness of the lavas in Matatiele did not issue from the volcanoes, but came from fissures which are now filled with dolerite and are dykes traversing both the sedimentary rocks and the lower lavas. The largest of these dykes is about fifteen miles long and a mile wide at its broadest part. It nms parallel with the main ridge of the Drakensberg from Deer Park to George Moshesh'g country, and on its southern side the amygdaloidal lavas cut through by it are turned upwards in a similar manner to the upturning of sedimentary beds round the walls of a volcanic neck. Along the northern wall of this dyke the lavas are much disturbed and crushed. These are features which have not been noticed in the usual doler- ite dykes in the Colony ; in the latter the molten rock seems to have risen quietly without having to exert a force capable of crushing or disturbing the rocks forming their walls. The formation of the dolerite-fiUed fissure on the Drakensberg ridge was evidently accompanied by explosive action, and through it may have been poured a large part of the lava which now builds up the higher portion of the ridge and a great bulk of rock that has disappeared under the ceaseless attack of the weather.

In no part of the Stormberg volcanic series have there been found great piles of lava and ashes arranged more or less symmetrically about a centre as are the lava streams and tuffs of such volcanoes as Vesuvius and Teneriffe, or the great flows of the Hawaian Islands ; but allowance must be made for the changes wrought by denudation during the very long period, represented in other countries by the Jurassic, Cretaceous, Tertiary

THE KARROO SYSTEM 221

and Quarternary deposits, formations that are but scan- tily developed in South Africa.

The absence of necks of agglomerate or other material from the Transkei beyond a narrow zone lying within a few miles of the mountain crest is certainly signifi- cant ; it points to the existence of a line of weakness more or less coincident with the position of the present ridge of the Drakensberg, along which at least the chief volcanic activity prevailed. Whether this was also the case throughout the region, and whether the lines of vents or fissures of eruption are marked by the im- portant spurs of the Drakensbergen in Basutoland which Mr. J. Orpen ^ found to be made of volcanic rocks, can only be ascertained from an examination of Basutoland.

It is quite possible that some of the larger necks men- tioned on previous pages are the passages through which great quantities of materials were ejected, and that these formed volcanic cones of large size now completely swept away. A general fact which bears on this question, however, is that the great conical volcanoes of the present day consist chiefly of fragmental tuffs which thin out quickly in all directions, though they may cover very wide areas. So far as our information goes the Drakensberg volcanoes were not of this type, for there are but few beds of tuff, and the agglomerates in the necks are largely composed of non-volcanic detritus, a state of things that would hardly obtain were the Drakensberg group strictly comparable with modern

* The first map of the volcanic region, that attached to Professor Cohen's paper (76), was based upon information collected by Mr. Orpen.

222 GEOLOGY OF CAPE COLONY

lava and ash volcanoes, or those of Carboniferous and Devonian age in the British Islands.

It is difficult to ascertain why a region so rich in lavas is so poor in ordinary tuffs ; many of the former rocks were highly vesicular, and therefore contained an abund- ance of water, an important factor in determining the explosive character of volcanic activity ; it is also diffi- cult to understand why so many of the necks should be largely filled with material derived from sedimentary or deep-seated igneous rocks which are very different in nature from the ejected lavas.

The part played by this volcanic episode in the geo- logical history of the country can be more conveniently dealt with in another chapter (chapter xi.), where its relation to previous and subsequent events will be explained.

The following is a list of the fossils hitherto discovered in the Stormberg beds : —

Plants—

Schizoneura krasseri. Sew.

Strohilites,

Thinnfeldia odontopteroideSj Morr.

,, rhomhoidaliSf Ett.

Cladophlebis,

Gallipteridium stor'nihergeme, Sew. Tceniopteris carruthersij Ten. -Woods. Ghiropteris cuneata, Carr.

,, zeilleri, Sew. Baiera stormhergensis, Sew.

„ schencki, Feistm. Ph(£7iicopsi8 elongatus, Morr. Stenopteris elongata, Carr.

THE KARROO SYSTEM 223

Fish 1—

Geratodus kannemeyeri, Seeley. ,, capensisy S.-Woodward.

Dictyopyge ? draperi, S.-Woodward.

Semionotus capensis, S.-Woodward.

Cleithrolepis extoni, S.-Woodward. Reptiles?—

Tritylodon longasvus, Ow. (also thought to be a mammal).

Eiuikelesaurus, Hux. \

Massospondylus, Ow. > Dinosaurs.

Orosaurus, Hux. {Orinosanrus, Lyd.) I

In his discussion of the relations of the Stormberg plants with those of foreign rocks, Mr. Seward ^ came to the conclusion that they are allied to the Ehaetic flora of other parts of the world. This flora had a more general distribution than the earlier one characterised by Glossopteris and Gangamopteris in the southern hemi- sphere, and by Lepidodendron, Sigillaria, and Gordaites in the northern, for it has been found in Europe, Asia, Australia, North and South America and South Africa.

Several of the most striking genera in the Stormberg flora, however, are by no means confined to this series, but occur in either newer or older beds. In the Cape Colony, for instance, species of Tceniopteris, Sphenopteris and Cladophlebis have been found in the Uitenhage series, and Schizoneura in the Beaufort and Ecca beds.

In India the Upper Gondwana beds have yielded many forms that occur in the Stormberg beds. The Panchet beds contain Thinnfeldia odontopteroides and Schizoneura gcmdwanemis, to which some Cape specimens are very

1 See note at end of chapter. ^owen (76), (84) ; Huxley (67).

3 Seward (03).

224 GEOLOGY OF CAPE COLONY

similar. The Panchet fossils, however, are, on the whole, more nearly allied to those from the Beaufort series, as in addition to Glossopteris, Dicynodon and Pty- chosiagvm have been obtained from them. The Eajmahal beds contain plants allied to those of the Stormberg series and also to the Uitenhage flora. The Kota Maleri beds contain a species of Ceratodus like G. capensis from the Stormberg beds of Smithfield, and also Massospondylus.

In Australia the genera Sphenopteris, Thinnfeldia, and TcBniopteris are known from the Hawkesbury-Wiana- matta beds of New South Wales, from the lower ** Trias- Jura " (Burrum) and Ipswich formations of Queensland, and from the Upper Coal-bearing series of Tasmania. The Hawkesbury series also contains Cleithrolepis, Dic- iyopyge, and Atherstonia, the two former being Stormberg and the latter a Beaufort species.

Any attempt to draw close parallels between these distant strata is foredoomed to failure ; but the results of a comparison on broad lines are sufficiently striking, and hold out the prospect of a more detailed correlation in the future when the fossils are better known. The greater part of the correlation of the African, Indian and Australian rocks rests upon the plants which seem to be far less satisfactory than the remains of invertebrates which furnish the means of correlating so many forma- tions in distant parts of the world. The reason for this is twofold ; in the first place fossil plants are too often badly preserved and at the same time the variation amongst individuals of one species is great, so that their determination allows wide latitude of opinion; in the second place the number of species that can be used

THE KARROO SYSTEM 226

for the purpose is comparatively small. In time these diflSculties will be partly overcome, but meanwhile any evidence from better preserved and more highly organised forms of life, such as fish and reptiles, deserves more credit than that from the plants. Unfortunately fish and reptiles are decidedly rare in the formations that may have been contemporaneous with our Karroo system, though the recent discoveries in Eussia referred to in connection with the Beaufort series lead one to expect much more help from the reptiles than we now have. Many species and even genera of reptiles are founded on very fragmentary remains, and too much weight cannot be put on determinations founded on pieces of the skeleton in the absence of the skulls.

The question of the general bearing of the Karroo rocks upon the geological history of the country will be . dealt with in chapter xi.

We may notice here that there is no indication of the sea having invaded the Karroo region during the period of deposition of these rocks. None of the numerous Carboniferous, Permian, or Triassic marine shells known from Europe and Asia have been met with in South Africa. At the same time we must note that there are no deposits of rock salt, gypsum, or other soluble sub- stances which characterise formations deposited in an area where evaporation provides the only escape for the water collected in its hollows. Such beds of soluble salts are well known in the red Permian and Triassic rocks of Europe, and they were formed in a desert country in which the rivers flowed into inland basins

without an outflow to the sea. Similar beds of salts

15

226 GE010(nv OF 'CAPE €C>LON Y

arenow formied iii deaei^ rdgiofta- If iihe Eaft-oo h^mti had b^arventirely iCiit xsfl^frowi' ^he^ meM\ 'aft li/ake; Tchad, tbeiGaapiaO swaid! Airal^alm' now, W6 Bhould fed evidence of iti dn- the deipositfe l^d down; itt' thfe tiiitie. . i " : ' :KFromj.'the: Bcc^ bedS'fco t&^ Sfcotmbeirg there' ate false ijectdad 'rockaj rippk • mairkfitig^ art' the^stttfaceis of I iimmewm» /to»taj bothf shates ■ • and ' saiidBtones; aRd leicaliilnoonformUiafii 'caused^ by the sdoUTlng away of the floor' by oaifrealtft which* 'd^pDBjted•bt1l^l^ d^tritnA' in* the hlottdw so fdttffii^d: - Th«se«iH point t© ttiepi^etalen^e bf shallow 'Wat^i^i in the Karroo* basin ' thtou^hout tlie period: ^ Wh^n thetee^ facts ari^ tak^ti feto • cotisid^rsttioii' with i^e^grdat -thickness'^^jf the 'sediments =conc?e)tt^d they afford clear prodf 4hat d great pstrt of the'Colonly W«yB!'slowty depressed ' dttring a' Very Idn'g pelriod ^x- t^enflmg f torn the Garbdiliferdus^ to tlie JurassicJ.!''

,'v ] : » . J

•'The ^hief rodks'bf'feebn^fc valtiean the Kairrt)0 As- tern are' the coal setois ^of'^ the Mbltenb group, which have been mentioned oh a- previous page. '

' Good ■building stotte is (ibtttiiiedfroiii the Beau'fort beds near Beaufort Westi Port B^eaufbrtj Graaff lEleihet attd Quoetistownj' In general the OBeaufort ^and E(bc4 sandstones -tire too idark in: coloiir and too irregular itf development -to bemused otherwise thai locally, but^the Qiieenstown stoiie has a ihore than local demand' owittg to 5itB better colour,' good working qualities, and'afavour- ablepofeitiicaa with regard to railway transport* ' -^

Infthe Stormberg series there are many places where freestone of good colour has been obtained, but: the eJiisting quarries are far from the railway. .;.,/

THE KARROO SYSTEM 227

Many of the public buildings in East Griqualand are built of sandstones from the Molteno beds.

When more quarries have been opened up for the purpose of supplying the up-country villages with stone there will doubtless be many sources of valuable stone discovered ; at the present time fair samples of most rocks that might be- of ; great* 'Use* are (practically im- possible to obtain.

The calcareous concretions containing clayey matter in" thfe ' Eccra^ and Beatifoi*t beds shbtrld be of value ill cemeiit making, btJt €tt pttesent nothing is being donfe with this limestone. • The e^^nse of • fuel' 'at placed where the limestone- is sufficiently abuhdanlf to \<^ork acboumis f orbits • not^ being lisM in this •Way: • = " : Watetk ahnost every wher^ foutidinmbdelratequan^ tities b/ boring into the Karroo formaition, though the rocks are rarely perimeableto -any e-xtent. The water obtained comes -from the joints which cut thi?ough the strata and allow them to hold water within a few htn- dired feet of the surface: The largest supplies appear to bebbtained behind d^kes of' dolerite, which act as sub- terranean dams in h61ding bjaok the water derived from a higher level. * '' • '

[Since this chapter was written the progress of the survey has raade it certain that CleithrolepiSy SemionotuSy and Ceraiodtis come from the uppermost ^art of the Beaufort series; HortalosaurtiS, e(' Dinosaur, occurs in the Cave san^^t^ne^and I^tochamp${!t, a, c^ococlUe of Jurassio type, has been found in the. Bed beds and Cave sandstone by Mr. du Toit, who has also obtained phyllocarids and wings of orthopterous insects from shales in the Cave sandstone. Nov., 1964.]

. 1 1

16

CHAPTEE VI.

REPTILES OF THE KARROO FORMATION. By R. Bbooic, M.D.

Few groups of fossil reptiles are more worthy of careful study than those found in the Karroo beds of South Africa. The continental conditions which prevailed at the time were favourable to the existence of large num- bers and many varieties of land animals, and the lake deposits which were then being formed were well suited for the excellent preservation of their remains. Not only are the fossil reptiles numerous and well preserved but they are forms of the very greatest interest.

The earliest land vertebrates appear to have arisen in Carboniferous times, in rocks of which period we find the remains of a great variety of Labyrinthodonts, but no undoubted remains of reptiles. The Labjrrinthodonts were peculiarly specialised Amphibians, characterised among other things by having the head hinged to the back bone by two condyles as in the frog, and not by a single knob as in most reptiles and birds. They sur- vived till the close of the Triassic period, and a number of very interesting forms have been met with in the upper Karroo beds of South Africa.

In the age succeeding the Carboniferous — the Permian — true reptiles first made their appearance, and in the

228

EEPTILES OF THE KARROO FORMATION 229

rocks of North America and Europe have been found the remains of a large number of primitive reptiles,

m

some showing afi&nities with the existing Tuatara lizard of New Zealand and others resembling more the ancestral Labyrinthodonts.

In South Africa, as we have a continuous series of beds, probably from the Carboniferous to the Upper Triassic period, we have a much better opportunity of studying the succession of the early reptilian types than is met with in any other part of the world. Some of the American and European types are imlike any that have as yet been found in South Africa, but on the whole the best general idea of the early reptiles can be obtained by the study of the South African forms.

Some conception of the extent of the reptilian fauna of the Permian and Triassic beds of South Africa may be gathered from the fact that at present at least fifty- three genera are known and a hundred species. Much difference of opinion has been expressed with regard to the classification of these forms, but as our knowledge has advanced most of the difficulties have been removed, and it is now found that the very large majority of the species can be conveniently arranged in five distinct, though more or less connected, orders. Of these orders the two lower show marked affinities with the Laby- rinthodonts, and the highest is surprisingly closely related to the lower manmials. The study of these five orders thus not only gives us a very good idea of the Permian and Triassic reptilian fauna, but enables us to see the steps by which the mammals have been derived from their amphibian s^ncestor^.

230 .,..,, €^BOLOCIrV OF. QAFB COU)NYi

The fir^t QTder. to b^ cQi{isid/e;;ecl.lay^sil?eeiv lormedipr fcji€| receptiiou.Qf,.a single g^pus, .Pro^lophonf :Ajk least two well-marked species are known, both lizai;4-Uk? r^ptilQs; ab9ut..tWielve,.tQ/fifte^j> ipcj^e^ i^ temgth.. jln general .proportip»8.:awl< ip »mwy;. points, K>f. atructar^ Procolophon beB.x^ a piar]l^4 i^esemblawe IjQ.th^ existing New Zealand. li?5axd/iSi?/^e?wv?w.;.dti,4iffer6, bptweyejir, :iw having, a much.mcMre priw^v^ jQpnditipn ,of ^bagk^of the skull iai^d o^ tl^e sj^ouldie;:, gii?dj^ apd pi^l^is^ - . . /

The palate. resembles His^^. of, Spkmpdoxti but diffein^an having np t^eth^ on, the palatw^, and in having : a langa number -on the pterygoids ,aiid pcc^yom^^rs./ .The Pwr cqlo^phon. 4iffpr6 frpm th^ iarge. majprity of f reptil6«^' in hftving ,the posterior: part of, the akull iToofed with ;bonej aAd\in this respect.it agraes'withtbe Iiabyrinthodontsi

The vertebra are. pf.ja. Viei:y primittive type^ retaining the passage! fpri the persistent inotOchord., •[

The shoulder gifdle hasion each side-awellrdevelopad scapula, Qoraaoid. and precomcoid,i supported by a pait of large clavicles and ayery Ifiirge modiftninterclaviclej:;

The limbs beau ^conaiderable* resemblaBciei to those- of lizards^ there being; in aa(^h foot 2, 3, 4,. 5, and 3 joints in the five toes, respectively instead' of 2^3^ 3, -3, and < 8 as in mammals. . -i .- .' ^. -.•'. :, \ -::.'. ;;.-.■•.

Abdoiininal tibs, siH'cth as: ajre • foAiLnd> in : Spfmnddbn; the crocodiles, alid naany primitive reptiles, are prefientw i<

The pelvis has the: .anterior elementsr4-the pubes and ischia; broad and ' fiat ■ as in the • Labyrinthbdonts. • : -

Though no other members of this or^er are known

REPTII/BSi -OP THE' HAHUIOO^ iFTOHMATJON .231

either in South Africa or elsewhere there occur in Europe and America one or two genera {e,g,, Sclerosav/rvs, Pariotichm, etc.), which seetn to be intermediate between it and the next order.

Pareiasaxfbia. "

This order was formed for the reception of a genus of very large fossil reptiles, Pareiasaurm^ of which in South Africa there are four or five species known. In North Bussia a spocies is known of even larger dimensions than the South African, and in Central Europe a small allied form with horns. In Scotland another small allied form, also horned, has been found ; and in America there are numerous genera possibly belonging to this order but not very nearly related to Pareiasaurv^,

Pareiasav/rus was a very heavily built animal about eight or ten feet in length and standing ajbout four feet high. It resembles Procolophon in one or two respects, but on the whole is considerably more highly organised.

The skull is very massive, and the surface bones are pitted somewhat after the manner seen in the Labjrrin- thodonts. In fact even in the arrangement of the bones of the upper surface of the skull the resemblance to the earlier types is very marked. The palate, however, differs entirely from that of the Labyrinthodont and agrees in type with that in Procolophon and Spheriodon,

The shoulder girdle resembles that of Procolophon in having well-developed scapulae, coracoids and precora- coids, but differs in having a large acromion process for the attachment of the collar bone, and in retaining the

^i' i0icf;f;ii3f|ilir|^|iv

<e.

•m

EEPTILES OF THE KARROO FORMATION 233

Labjnrinthodont supraclavicle or cleithrum — a splint bone which protects the front of the scapula.

The pelvis bears some little resemblance to the mam- malian type.

No abdominal ribs have as yet been found in Pareia- saurus, and it is highly probable that none existed.

The number of joints in the toes is not yet known for certain. One toe undoubtedly has four joints ; possibly the numbers are 2, 3, 3, 4, 3 respectively, thus belonging to a type intermediate between Procolophon and the Anomodonts, etc.

Thbbocbphalia.

Contemporaneously with Pareiasaurus there existed a large series of other reptiles somewhat allied but belong- ing to a different order. Whereas Pareiasawms was a clumsy slow-moving animal, with uniform teeth only suited for cropping herbage, the other types are for the most part slightly built animals and having teeth differentiated, as in mammals, into incisors, canines and molars. Considerable confusion has hitherto been caused by these early carnivorous types having been placed with the Theriodonts to which they are not very nearly related.

The skull bears considerable resemblance to that of manomals, differing mainly in the structure of the palate and of the lower jaw and its hinge. Each premaxillary bone usually carries five pointed incisors, and in the maxillary there are usually two canines, sometimes three, and a series of small pointed molars. The molars varj^ in number frojn one to ejeyen, The palate is a,

234 / . : QBOLOGY OiP GAPE COLONY

Blight modifieation of that found, in Procol&pkon ieuid PareiasauruSf the iatfernfl.1 nasal opening' being by the side of the canineS) and there is no traee of a secondary palate. On the pterygoid bones there arei asnally a series. of stoaaJl teeth. . .

Of the lower jaw the d^ntary bote only farms a little more thaai the anterior half, the? posterior part being formed by three other large: eleitouentfl as in most rep- tiles* . A , well-developed quadrate, bone is -preseat f6r the aartjculatioJi of the jaw. There is a single occipital condyle. * . /.

The limb bones differ from those of Pareiasaums mainly in being long and slender. There is in the «hoalder girdle no acromion. pBocesa

The best known South African Therooejdialianfl are ^hurosaurusy Ictidosuchml /Lycosuchus .and Titwnosuohm animak varying in size from a cat to a^ hc^se. A very much larger form* TapinocephUlusi, i^ met with. It> was an animal .probably as .large . as .a rhinoceros, but. it is unfortunately vexy iniip^fectly known aad possibly belongs to the Pmeiasomria. . ..

In Russia -a number of Theroo^halians hftve been found, the best known, Jbeipg.i^ewtefio^awms e^^RJwpalQ- don, Eecently very perfect skeletons of a large form, No^tronzewia^ have been tQund in North BtiflsSa. ;

•• ' • I ' . . . - •'.

ANOMdDONTIA (OR DiCYNOBONTIA).

The Anomodontia include a large setdes of fossil forms, cbfebracterised among other things, by having, like the Edentata among mammals, no teeth in the

REPTILES OP THE/ KAiRiROO» iEKDRMATION S^

front of -the jaw.' In. general strnoturei thay (Ureiinti^fr mediate between the Therocephalians and the Then^)!- dontsV but ithey <also\ show fibfli^ »^ffinitie8.'. with. < the Pareiasauriana. • : Im ; si^^ :they.' t var^ fjcom i ramao^hr as small euB a rat to huge hea/vily JDuilt. iforxa& iBoniuewhat larger than a wild boar. .i'\. i ' !. ' .

' The skull :<r66e(mbles ooiisisAei!akly\ that of the Thero- cephalians and jth6 Theriodont^v 'ftiitd isimainly reaxicmrk- able for the. enormous deVelopimentrfof .ithe squajoaosai ^bone. and the large t size of the qtiadtrate. - The palate resembles much: mote lolosely thiat ofiithe Theriodomte than 'the type i met with ih the leariiexi forms. ' r : ;

The 8houj[dei''gir41ef res^mfales veory doaiely th^t of Pareiasaurusy there being usually: >;pre8ent la [distinct ieleithriitn.. Ah ossified stemiim/ or biieast bone is iprob*- •ably- in'^ariably^- present '.r. ;'r •;•: (,i - -..^ ■, •-. \ •»!..•;: ,

The bones of the fore^ limb also resemble /those of 'Pare'iasaurus, the humerus having alwiays a huge deltoid ddga The front foot very clc»ely resembles that of mammails, - the toeis having %8i\3^ S, 8. joints respecr- tively.- . // • i . '/ ! ..• ■'-<•-: i

. The pelvis and the boii^B .of.the hind limb lare strikingly. mammal-like. .n> ^ * .\ '

' The best known .Andmodont' genu« i^ DicynodQn,'<ii which over twenty species \ have: -. beeil : discovered^ some smaller than a cat, othiers possibly nearly as large as Pareiasa/urus. The jaws in front formed a homy beak as in the tortoise, but in addition there were two power- ful tusks, between which the lower jaw .worked. There were no other teeth. In the larger speoiee the head is usually proportionally very large.' In Dicytwden leoniceps

236 GEOLOGY OF CAPE COLONY

the head is narrow ; in DicynocUm tigriceps it is very broad.

Oud&nodon is closely allied to DicynodoUy but differs in having no tusks. A considerable number of species are known varying from less than a foot to probably about six feet in length.

Lystrosawrus (= Ptychognathvs) is an aquatic form of Anomodont. The limbs are very short and ill-adapted for progression on land. The head, though agreeing fairly closely with Dicynodon as regards its essential structure, is remarkably distorted. The beak is long and the back part of the head very short, while the occiput and snout lie in almost parallel planes. The eye and the nose are close together and near the top of the head. The peculiar shape of the skull would enable Lystrosav/rus to lie near the surface of the water with only the eye and nose exposed.

Endothiodon may be taken as the type of a number of genera, closely allied to Dicynodon and Ovdenodon, but differing in having a number of teeth on the maxillary bone and in the lower jaw. Some of the genera are less than a foot in length and have remarkably specialised teeth, while Endothiodon bathystomay the largest form known, was between three and four feet in length. In this large form the maxillary and lower jaw teeth are arranged in three series. The head is of enormous size, with a large parietal crest and a very wide occiput. The vertebrae are short, the ribs well developed and the limb bones very similar to those of Dicynodon. The Endo- thiodonts form a connecting link between the Thero- cephaliang ftnd the Anomodonts, such as Dicynodon

M«

Zm-m -I-

nil "111

ill?

m

fell

.1:1 1 i

Ip 1=1 f?

238 'I GBDLOGYXDir/OAPEI COLONY I i • i i

and Oudenodon, but they are very much more nearly related to the latter.

Theriodon^IA.

The Theriodonts ace medium-sized reptiles remark- able for the strikingly close resemblances which they bear to mammals. Only a few genera are known at present, but fortunately most ,crf the important points of structure have been revealed. The best known genera are Cynognathus, Gomphognathus, Microgomphodon and Galesaurus. The Theriodonts are the carnivora of the upper Karroo rocks as the Therocephalians are of the lower.

The Theriodont skull resembles considerably that of both the Anomodonts and the Therocephalians, and also bears a close affinity to that of the lower mam- mals. The most remarkable features of tfie skull are the presence of two occipital ccmdyles and the develop- ment of a secondary palate.. . The lower jaw is formed almost entirely by the dentary bone, the other bones being of small size. The quadrate is quite rudimentary. The dentition is almost typically mammalian, and not only are the teeth divided into incisorB, canines and molars, but the molars are specialised in different genera into carnivorous and insectivorous tjrpes. The palate is formed as in mammals by secondary plates from the maxillary and palatine bones, the internal nares being carried as far back as in most mammals. The pterygoid bones are of large size as in the Anomodonts and Therocephalians and unlike those of mammals.

REPTILEg OlF THE KAlfeROO 'FORMATION 239

I The ;TOrfebar89 are! itemarkaJble ior having peculiiir flat ovsetliappingi aribs in the lumbar I'egioD. . i »^ ' • "i

In the shoulder girdle the -scapula; icoracdid and-jme^ ooraceid rosemMe miwih/ ihore i those Elements in ^he Anombdbntsithan in th^Therocephalians. • ' i > f"

The pelvis i& much morealnammalian in type^han- that of th«( earlier forms. !; . ; ■'•■": .1 .'! . QytwgTiathusi the ibest Itnown' genu«, is a large wolf-like reptile. The heAd.is about Mxteen inches in length and the. whole' aniimal probably >measU(Ped' about ' six feet. The molar teeth have cusps very similar to those seen in .many carnivorous mammJals; » > ' . GomphognathuSy though vely similar to Cynognathus m geiQeral stmcturev ^ffens in having a broad ^ and fiat head, and. iii having the .molair teeth with flattened GtD wna- iltf probably measures .about four feet. »

Microgomphodon is ia< small form with flattened molars. It is about the. size of. a.meerkat.

' Qaleeamrus is a small carnivoorous' t3rpe, of which only the skull is known. The head is more depressed than' in Cynognai}m&. i

I ■ «

^ , • • '

Relations of the Theriodonts to Mammals.

The study of the fossil reptiles of South Africa has not only revealed some very remarkable types of animal bfey but i has practically resulted in the solution of one of the most) vexed problems of ? biology-— 'the Origin of Miammals. .

. In Procolophon we have a type which, though distinotlj^ mioze closely allied to the ancestors of the lizards^ is

240 GEOLOGY OF CAPE COLONY

probably not very unlike that which formed the com- mon ancestor of the PcLreiasaurians, the mammal-like reptiles and the mammals.

Pareiasatt/nt8, though possibly in one or two respects more primitive than Procolophon, is on the whole dis- tinctly specialised along the line which gives rise to the mammals. The shoulder-girdle and pelvis are strikingly like those of the lower mammals. The well-developed acromion process which forms so marked a feature of the shoulder-blade of mammals appears in Pareiasawrus for the first time.

The Therocephalians in some respects resemble mammals fairly closely. The general arrangement of the face bones and those of the upper surface of the skull generally is surprisingly mammal-like, and the teeth are divided into incisors, canines and molars al- most exactly as in the higher forms.

The Anomodonts though somewhat out of the direct line of mammalian descent, are even more nearly related to the mammals than are the Therocephalians. We here see the secondary palate in its early imperfect con- dition. Most of the bones of the skeleton are so like those of the Monotremes that Owen many years ago suggested the possibility of the ancestors of the Mono- tremes being found among the Anomodonts.

The Theriodonts are most probably descended from Therocephalian ancestors, but they have so far advanced along the mammalian line that they are more closely allied to their mammalian descendants than to the Therocephalians. In the structure of their teeth, palate and limb bones they may be said to be almost mammals.

181

ii#;,iiiJl*i'»*3i|j|iiirK

;.- - •&• •&• L -f- :»: 5|c '^^ ■»• •

Uie Therimionls, ■ ' Australia BimlUr to

242 GEOLOGY OF CAPE COLONY

To whatever point in the structure of the Theriodonts we turn we find the mammalian condition foreshadowed in a most remarkable manner. The two points in which the mammalian skull differs most markedly from the reptilian are (1) the simple nature of the lower jaw, and (2) the presence of two occipital condyles. Both of these peculiarities are explained by the Theriodont condition. The Theriodont jaw differs from that of the Therocephalian and all other reptiles in being formed almost entirely by the dentary, which almost reaches the articulation. The articular is small and to a great extent overlapped by the dentary. The angular, sur- angular and splenial are small rudimentary splint bones. The quadrate on which the articular hinges is a small bone which lies on the front of the downward process of the squamosal. In the mammal the lower jaw is formed entirely by the dentary ; and the quadrate has disappeared as a distinct ossified element, so that the dentary hinges on the squamosal. It will thus be seen that the mammal differs from the Theriodont only in its having lost those elements which already are rudi- mentary in the Theriodont. The quadrate appears to be completely lost in many mammals, e.^., Monotremes, but it is probably represented by the interarticular cartilage in the large majority of forms. The articular element of the jaw is possibly represented by the cartilage found in the condyle during development ; and a small splint bone in the jaw of the very young Omithorhynchus probably represents the angular.

The occipital condyle in the Theriodont is merely a modification of that found in the Anomodonts. In those

EEPTILES OF THE KARROO FORMATION 243

a large single condyle occurs formed by the two exoc- cipitals and the median basioccipital. In the Therio- donts the basioccipital takes less part in the formation of the condyle than the two lateral elements, and hence the condyle appears to be double. In some of the lower mammals a condyle essentially similar in structure occurs, the basioccipital forming part of the joint, but in most of the higher forms the basioccipital takes little or no part, and thus what was originally a single condyle formed by three elements becomes a double condyle formed by the two lateral elements alone.

Though the above view of the origin of mammals seems to have on its side the very strongest palaeonto- logical evidence, various other theories have been pro- posed. Many would derive the mammals directly from Batrachian ancestors through a long Une, of which we know nothing, originating in Devonian times. The quadrate bone of the Batrachians and Beptiles they consider becomes one of the auditory ossicles in the mammal. By others the mammalian tympanic bone is regarded as the homologue of the reptilian quadrate. Neither of these views has the slightest support from palfiBontology.

Other Kbptilian Types.

While the large majority of South African fossil reptiles belong to the phylum which terminates in the mammals there are a few other interesting forms.

A small lizard-like form called Sav/rostemon is believed

to be allied to the New Zealand lizard, Sphenodon^ but

16*

244 GEOLOGY OF CAPE COLONY

it may be a true lizard. Another form known only by the skull, Paliguanay has the quadrate bone free and must thus be classed with the Lacertilia.

Proterosuchus is a moderate-sized reptile with a long narrow pointed skull. Though having affinities with Sphenodon, it also shows a number of resemblances to the Primitive Crocodiles and Dinosaurs, and it would seem to belong to a group which included the conoimon ancestors of Crocodiles, Dinosaurs, Pterosaurs and Birds.

A few Dinosaurs are known, which resemble fairly closely the Triassic Dinosaurs of Europe and America.

[Since the above was written evidence has been obtained which renders it probable that Sawrosternon belongs to the Procolophonia.

The most important recent discovery among the Karroo Reptiles has been that of small crocodiles in the upper Stormberg beds. They belong to a genus which has been named Notochampsa. Though only about two feet in length they are fairly closely allied to certain crocodiles found in the lower Jurassic beds of Europe. Unlike modem crocodiles they have fairly long legs, and were no doubt able to run swiftly.]

CHAPTEE VII.

THE INTRUSIVE DOLERITES AND ALLIED ROCKS.

The dark-coloured, heavy rock, blue-black when freshly broken, and red-brown, black or yellow on weathered surfaces, that occupies such great tracts of country north of a line drawn between Sutherland and East London is generally known by the name of ironstone, or yzer-klip to the people who live near it. It probably got the name from the property it has of ringing like a piece of metal when struck. This rock is composed chiefly of four minerals, plagioclase felspar, augite, olivine and magnetite, in the order of their relative abundance and commencing with the most abundant mineral. There are other constituents, some of which can be found in every piece of the rock examined ; but they are of less importance than those just mentioned, and will be spoken of later.

The mineral composition shows that the rock belongs to the basic group of igneous rocks, and the few chemical analyses that have been made of it show that it has a similar composition to that of dolerites known from other countries. In this book, as in the Eeports of the Geological Commission, the name dolerite is used in the

245

246 GEOI.OGY OF CAPE COLONY

sense adopted by Allport,^ and Teall,^ including rocks composed chiefly of plagioclase and augite. They may or may not contain some glass between the usual con- stituents. The composition varies considerably through- out the country, but in very many localities rocks with obviously different compositions can be seen to belong to one and the same mass. The chief change in com- position is in the amount of silica, which has the effect of altering the proportion of some of the minerals present ; as a general rule the more silica there is the less olivine and augite is seen in the specimen. If strict attention be paid to the mineral and chemical composition of the rocks, those belonging to the great group we are now describing must be given several names. Few of these can be determined without a minute examination of the specimens. The intrusions as a whole can conveniently be called dolerites.

According to the shape of an intrusive mass of igneous rock and its relationship to the surrounding rock it i^ called a dyke, sheet or sill, laccolite, or a boss or batholite.

Dykes are masses of rock filling vertical, or steeply inclined fissures. They may traverse sedimentary or

^ Q. J. G. S., XXX., p. 529.

^British Petrography ^ ch. vii. These rocks generally correspond to the diabase of Kosenbusch and Zirkel, although many examples would belong to the basalt and dolerite of these authors if the question of geological age were left out of account. It may be well to mention for the benefit of those who have no acquaintance with petrography that the naming of igneous rocks is still in a state of confusion or something very like it, and that very many names should not be used without reference to the author whose usage is followed.

INTRUSIVE DOLERITES AND ALLIED ROCKS 247

igneous rocks. The width of a dyke does not as a rule vary greatly, so that when the dyke-rock is more re- sistant than the enclosing beds it has the appearance of a waU. Dolerite dykes are abundant in the Colony, an example of these being shown in Fig. 11.

A sheet or sill is a similar body to a dyke, but it lies approximately parallel to the bedding planes of the sedi- mentary rocks it penetrates. Sheets of dolerite are more abundant in the Colony than any other form of intru- sion ; they are shown in Figs. 1, 2 and 4, and Plates XIII. and XrV. ; they are often connected with dykes that in some cases may be regarded as the channels through which the rock composing the sheets flowed.

A laccolite is of the nature of a sheet that is very thick in proportion to its extent and thins out on every side, forming a thick lenticular mass. A laccolite, moreover, often raises the overlying sedimentary beds into a dome corresponding to its own contour. Certain of the large masses of dolerite in. the east of the Colony are perhaps laccolites, but the arching up of the overlying beds has not been observe.

A boss or batholite is a large deep-seated mass of more or less irregular form and of unknown depth, but no examples of this type of intrusion are found amongst our dolerites. Several of the granite masses in the Pre- Cape rocks belong to this type of intrusion.

It will have been noticed that no mention has been made of lava in connection with the dolerite. The masses here described all consolidated at some distance below the surface of the earth, and can be seen only by the removal of the overlying rocks by denudation. A

^1

^M

SI

II

II

J-

■32

B||

'.j;?.Sf.3jV.Si..5».5. .J. .J. .J. .g. .g. .g. .g.

|f!§|:5^i^.|-:|(^fj

ROOKS 249

11

ii a

fl ?i

II

T1

__^^^ Ij

250 GEOLOGY OF CAPE COLONY

lava is an igneous rock which has flowed from a vent or a fissure over the surface, and though of very varying nature it may have the same chemical and mineral com- position as dolerite. Ancient lava flows that have been deeply buried under sedimentary rocks and are now exposed at the surface by denudation have some charac- ters in common with sheets or sills ; in the case of very ancient sheets of igneous rock lying parallel to the bed- ding of slates or other sedimentary rocks, it is often very difficult or even impossible to decide whether the rock was a lava flow or an intrusive sheet. With rocks that have not undergone much alteration since their forma- tion there is not this difficulty, for lavas are usually rough and slaggy at both their upper and lower surfaces, and the sediment deposited upon them is not hardened at the contact as are the beds above an intrusive sheet. The only serious difficulty in distinguishing between lavas and sills of slightly altered rocks is met with in the case of sills intruded amongst lavas of similar composition. Examples of this are to be found in the volcanic group of the Stormberg series, and there is some doubt as to their true nature. Amongst the hundreds of dolerite sheets that have been examined in the rocks, older than the Stormberg volcanic group, none has been found to have the characters of a lava flow, but there is often conclusive evidence in the hardening of the overlying rock and in the sheet breaking through to a slightly higher or lower horizon that the rock is intru- sive, i.e., that it was injected into its present position in a molten state after the surrounding sedimentary rocks were deposited.

INTKUSIVE DOLEItlTES AND ALLIED ROCKS 251

It was stated in the Introduction that the dolerite intrusions are practically limited to that part of the Colony which was not seriously affected by the earth movements that took place subsequently to the deposi- tion of the Ecca beds. In the west of the Ceres Karroo a nearly straight dyke about thirteen miles long and 100 feet wide runs north and south through Beukes Fontein, traversing the Dwyka conglomerate where that rock dips somewhat steeply to the east. This dyke dies out at each end and gives off no sheets. In the valley of the Brandewyn's Eiver there are two dykes travers- ing the Bokkeveld and Table Mountain series in an area where these beds are slightly folded, and in the neighbourhood of Groen Eiver and the Bokkeveld Mountain escarpment there are also two dykes breaking through beds belonging to the Cape formation, but the beds have there been only very slightly disturbed. No dolerite intrusions have been met with in the great folded belt between the Clanwilliam Mountains and the Gualana Biver. We have to go to Pondoland ^ where the Table Mountain sandstone lies almost horizontally before we again come across dolerite in the Cape forma- tion ; it occurs there as a dyke in the sandstone of the Egossa forest.

A considerable number of dolerite dykes penetrate the Ibiquas series in the west of Calvinia, but it is in the rocks belonging to the Karroo formation that the intrusions attain their greatest development.

' Since this was written Mr. du Toit of the Geological Survey has found two dolerite dykes penetrating the Table Mountain sandstone of the Cape Peninsula, which lies outside the folded belt.

262 GEOLOGY OF CAPE COLONY

In the Tanqua Valley east of Eland's Vley there are several dykes with a north-westerly trend, and some of them are connected with small sheets, but the main area of the dolerite intrusions commences on the north of the Tanqua.

In the Dwyka series between the Langebergen (Cal- vinia) and the Tanqua Valley there is a very extensive sheet which stretches with a few breaks in the northern part of the outcrop for rather over 100 miles, and it is at places 300 feet thick. This sheet and indeed all those in the western part of the country tend to rise towards the south-east, and they traverse higher and higher beds in the same direction. The lowest sheet first appears near the base of the Dwyka conglomerate north of the Oorlog*s Kloof Eiver, but at the south- east extremity on Potkly's Berg East it is in the lower part of the Ecca beds, having passed diagonally through a thickness of about 1,000 feet in the course of some sixty miles. In looking at such a sheet at any one part of its outcrop it appears to have been injected parallel to the bedding planes of the enclosing rock, and it is only by the examination of a very long outcrop that the fact of its breaking across the bedding can be determined ; clearly cut sections are difficult to obtain except on vertical cliffs, and these are not abundant in the case of this sheet. The sheet is crossed by the main road from Ceres to Calvihia at Bosch Kloof, where it forms an outcrop about six miles wide. It forms a considerable part of the upper slopes of the escarpment called Eland's Berg ; the hill is capped by the Upper shales of the Dwyka series which in turn are overlain

INTRUSIVE DOLERITES AND ALLIED ROCKS 253

by a smaller sheet of dolerite, an outlier of an offshoot from the lower one. Outliers of the lowest sheet cap the Guap Mountain and Eiip Bug Eop ; the latter is a very conspicuous conical mountain formed of Dwyka conglomerate standing on the watershed between the Wolf and Oorlog's Kloof Eivers. The offshoot from the lowest sheet in Calvinia is probably connected with the latter near the Drie Fontein Mountain, but the out- crops are apparently separated ; it runs along the foot of the Eoggeveld escarpment as far as the Bhenoster Biver, a distance of fifty miles; but near its point of departure from the lowest sheet there is a second off- shoot at a higher level traceable for over fifty miles on the escarpment as far as Sneeuw Erantz (Boven Plaats) on the Boggeveld. A fourth sheet is connected with the third at Boode Fontein on the edge of the Boggeveld, and in addition to forming the edge of the escarpment for many miles south of Boode Fontein, it covers a wide extent of country to the north round Kreits Berg (Zand Kop), Boep-my-niet, and Hantam, in addition to a great tract to the east. The Boggeveld sheets below the fourth or highest one in this area do not extend into the Sutherland and Beaufort divisions. It is not certain as yet whether the fourth sheet, the one that crowns the Boggeveld at Boode Fontein, is connected on the surface with those north-east of Sutherland. The latter are the continuation of a sheet that forms the summit of the western Nieuweveld, whence it gradually drops to the level of the northern part of the Gouph. This great sheet, traced between points 100 miles apart, is connected at the eastern end, where it is cut through

254 GEOLOGY OF CAPE COLONY

by the Koekeinoer's River, with a steeply inclined sheet or dyke at a lower horizon, which has itself been found to extend over sixty miles 4,0 the west with a continuous outcrop. This inclined sill, which is called the Boode Hoogte sheet, dips at about 30"^ to the north, and is as much as 400 feet thick in places. Like the overling sill, the Boode Hoogte sheet rises towards the Boggeveld; it makes a rapid ascent west of Banks Gaten on the extreme western limit of the Beaufort West Division. The dolerite krantz runs up the left bank of the head-waters of the Dwyka Biver, which for the first six miles of its course has a most remarkable canon-like valley. A tributary has cut o£f a big out-lying portion of the sheet in Alleman's Hoek, and to the west of that locality the dolerite strikes across the plateau behind Komsberg, psksses a few miles to the north of Saltpetre Eop and disap- pears near Jackal's Fontein on the Sutherland main road.

East of Tafel Berg, that -fine flat-topped mountain with such gracefully shaped slopes below the krantz (400 feet) of columnar dolerite, and which can be seen, together with its neighbour, Spitzkop, from the railway beyond Prince Albert Boad, the Nieuweveld summits are formed by outliers of sheets that occupy wide stretches of country behind the escarpment. Some of these sheets appear as continuous outcrops, usually in the form of krantzes or cliffs from 100 to 400 feet high for about twenty miles along the edge of the escarpment, the highest point of which is the peak called Bulthouder's Bank, 6,270 feet above the sea and

INTRUSIVE DOLERITES AND ALLIED ROCKS 255

3,500 above the town of Beaufort West that lies about seven miles to the south-east.^

From one of the prominent peaks near the edge of the escarpment, such as the Tafel Berg just mentioned, or Javander Kop near Steenkamp's Poort, a magnifi- cent view lies before one. To the north range upon range of rough dolerite kopjes, occasionally merging into more important hills, stand upon the plateau that ends abruptly in the Nieuweveld escarpment ; almost at one's feet is the edge of the escarpment, with a dol- erite krantz at the top, and often one or more on the precipitous slope of some 3,000 feet from the summit to the level of the Karroo at the bottom ; to the south stretches the Great Karroo with its low ranges of flat- topped kopjes of shale and thin sandstones, shut in on the horizon by the blue slopes and peaks of the Zwartebergen. The dolerite outcrops do not extend farther than eighteen miles south of the Nieuweveld escarpment, rarely so far.

The Boode Hoogte sheet, which is inclined northwards at a moderate angle, fronts the Great Karroo for nearly fifty miles, and forms the southernmost of the dolerites for a distance of over seventy miles. It may have ex- tended some way farther south than its present outcrops, but as there are no other dykes to the south, that is, no channel whence further sheets could have been supplied, and as there are no outliers of dolerite in that direction,

^A detailed description and map of the sheets and dykes of the eastern Nieuweveld will be found in Geol, Comm. (96), pp. 15-26 ; of the Roggeveld in Oeol. Comm. (00), pp. 50-52 and (03). A map accompanies the latter Report and that of 1896.

256 GEOLOGY OF CAPE COLONY

we must regard the present outcrop of the Eoode Hoogte sheet as near the former southern limit of the intrusions.

In the area between the town of Beaufort West and the west end of the Nieuweveld escarpment there are several thick dykes with a northerly inclination ; one of them runs through the town, and behind it the two town dams have been made by blocking up the exit of streams; this, the Beaufort dyke, has been traced over more than thirty miles and gives rise to a thick sheet at Stoltz Hoek. On the road to Fraserburg up Thee Kloof there is a thick dyke very well exposed for hundreds of feet on the steep sides of the valley; two thin dykes lie parallel to it. The exact posi- tion of the dykes and sheets of the southern edge of the dolerite country is not known east of Beaufort West, but they run between Aberdeen and Graaflf Reinet, thence through the country just south of Bed- ford and Fort Beaufort to a point south-west of East London where they disappear under the sea. North- east of East London they appear in great force in the Komgha Division and throughout the whole of the Transkei, Pondoland and Griqualand East, and they are continued right through Natal.

m

The position of the southern limit of the dolerite intrusions is shown approximately in the small map in Fig. 3. North of this line the dolerites are very widely spread. In the western part of the country, in the drainage basin of the Zak, Hartog's Kloof and Onger's Eivers the dolerite forms the innumerable kopjes and ridges mentioned in the description of the

INTEUSIVE DOLERITES AND ALLIED ROCKS 257

view from one of the Nieuweveld peaks. The most

important ranges of dolerite hills in this part of the

comitry are the Karree Bergen, Slang Bergen, Tulbagh

Mountains, Kat Eop hills and the hills south of Wil-

liston (Amandelboom). Farther to the east, from the

Fraserburg boundary to the Stormberg, the conical

mountains with flat tops of dolerite or with pinnacles,

remnants of former table-shaped summits, are very

frequently met with.

There are some very considerable ranges of mountains

that run more or less parallel to the main watershed in

the Eastern Province and divide the country south of

that watershed into two parts, a northern (Middelburg,

Cradock, Tarka, Queenstown), drained by the main

branch of the Great Fish Biver and the Kei ; and a

southern (Graaff Beinet, Somerset East, Bedford, King

William's Town, etc.), drained by the Sunday's Biver,

tributaries of the Great Fish and Kei, Keiskamma

and Buffalo Eivers. These mountains branch from

the main watershed at the Compass Berg (8,500 feet),

which is the highest point in the Colony, except some

of the peaks of the East Griqualand boundary ; they

are called the Sneeuwbergen, Tandjes Berg, Bank

Berg, Winterbergen and Amatolas in different parts

of their course. They all appear to owe their existence

to the presence of thick sheets of dolerite that have

protected the sedimentary rocks from destruction. The

distribution of these sheets and their relations as parts

of a great system of intrusions have not been worked

out, but there can be no doubt that they connect the

well-known intrusions of Beaufort West and Calvinia

17

258 GEOLOGY OF CAPE COLONY

with those of Kentani ^ and the Native Territories generally.

To the east of King William's Town in the country within forty or fifty miles of the coast the dolerites have much less effect upon the topography of the land than in the western districts, or rather it would be more correct to say that the effects are less obvious, for we no longer find the outcrops marked by krantzes or definite ranges of kopjes. The reason of this is that the country is covered with grass or bush, and the soil accumulates on the slopes as well as on the flat ground instead of being rapidly removed from the slopes to lower levels by rain, as is the case in the Karroo, the high country north of the main watershed, and in the higher parts of Griqua- land East.

There is perhaps more rapid variation in thickness in the Kentani sheets than in those of the Nieuweveld and Roggeveld. The Kentani Division is the only compact tract of country consisting largely of dolerite that has been mapped geologically in the east of the Colony, and a short description of it, illustrated by the accompany- ing plan (Fig. 21), will serve as a typical example of the manner in which the intrusions occur in those parts. The district is bounded by the Gcua and Kei Eivers on the south-west, the Kogha on the north-east, the shore on the south-east, and the main road to Umtata on the north-west.

The sedimentary rocks are shales and sandstones con- taining Oudenodon and belonging to the Beaufort series.

^ A large scale map of the Kentani intrusions has been published in Oeol, Comm. (01).

INTRUSIVE DOLERITES AND ALLIED ROCKS '259

They dip at very low angles to the north-west, and are not folded to an appreciable extent .

The lowest sills are found on the coast where there are two called the Kobonqaba and the Mazeppa Bay sheets respectively from the localities where they are

Fio. 2 . — Map of Kentani showing the distribution of dolerite sheets and **gap" dykes. The area left blank between the Kei and Kogha rivers is made of sandstones and shales of the Karroo formation. Scale 1 in. to 10*6 miles. The vertical scale of the section is much exagger- ated, ^ in. to 1,000 feet. The name Manubi is written across the Manubi sheet. The tiuding station of that name is to the east.

well exposed. The Kobonqaba sheet extends nine miles along the coast and about two and a half miles inland at its broadest part up the Kobonqaba Valley, where it dis- appears underground. Its greatest observable thickness is 300 feet, near the Wheeli Eiver, but the bottom is

nowhere seen. The patch of sedimentary rock on the

17*

260- GEOLOGY OF CAPE COLONY

coast, north-east of the Nxagha River, is part of the overlying beds faulted dov^rn on the north side against the dolerite. An interesting feature in this sill is the occurrence of dykes of a much more siliceous type of rock than the dolerite they traverse. On Plate XV. is shown a thin dyke of light colour traversing the sheet on the shore near the Kobonqaba mouth. The sheet itself is a rather coarse-grained olivine-dolerite, with well- developed ophitic structure, that is, the augite occurs in rather large, irregularly shaped masses into which well- formed crystals of plagioclase felspar project, or they may be entirely enclosed by the augite ; a small quantity of green hornblende is intergrown with the augite and red biotite, magnetite and apatite are present in fair quantities ; the olivine is partly converted into serpen- tine. A very small amount of quartz is also present. This rock is very like that forming many of the Trans- kei and Pondoland sills, and contains more hornblende than is usually seen in the dolerites of the western dis- tricts,, although the same mineral is not entirely absent in the latter. The light-coloured dyke has no olivine or augite in it and very little hornblende, which is at places intergrown with orthoclase felspar ; red mica is abun- dant ; the plagioclase forms zoned crystals, i.e., crystals whose composition changes regularly from the kernel to the outside, thus having corresponding changes in the optical character of the succeeding layers in each crystal that are easily detected under the microscope. In addi- tion to the plagioclase felspar there is much orthoclase in the rock, intergrown with quartz to form micropegmatite. Orthoclase is practically absent from the olivine-dolerites,

___e^^;gi^j|5i&'E«|:«i|f£>

L

Mocks 261

's=..^*y-;g*';

SI

r

IBI

262 GEOLOGY OF CAPE COLONY

and quartz is of very rare occurrence. Apatite, magnetite and zircon are found in the acid rocks, as well as in the dolerites,

A rock similar to the acid dyke near the Kobonqaba mouth is seen a little farther to the north-east in the same sheet, and on the left bank of the Nxagha Eiver, about one and a half miles from its mouth, there is a large dyke of the same nature, but with large plates of titaniferous magnetite, which appear as long needles in a cross-section.

Although these acid rocks, which may be called granophyres (Rosenbusch) on account of the abund- ance of micropegmatitic intergrowths of quartz and orthoclase, are so different from the typical olivine- dolerites, there is strong reason to believe that they were the latest intrusions from the same source that produced the dolerites at a slightly earlier period. Many of the minerals in the acid rock are identical with those in the dolerites, in fact there are no minerals pecuhar to the former, it is chiefly the large proportion of quartz and potash felspar and a corresponding decrease in the augite and lime-soda felspars in the granophyre that distinguish it from the basic rock.

The Mazeppa Bay sheet is exposed along a mile of the coast and has been followed as a thin sheet round the basin of the Kleena Eiver and across the Manubi Eiver, about four miles round the lower part of the Manubi Forest. The Mazeppa Bay sheet may be con- nected with some irregular outcrops of dolerite on the shore between the Manubi and Kleena Eivers. The upper surface of one of these masses of dolorite i^ seen

INTRUSIVE DOLERITES AND ALLIED ROCKS 263

to cut oflf a bed of sandstone obliquely through a vertical thickness of about 4 feet, and then to pass beneath the succeeding bed. There are several large dykes striking inland from the coast between the Kogha and Kobon- qaba Eivers, probably connected with one of the inland sheets.

A thin sheet winds round the divide between the Istamfoona and Umfane Rivers, perhaps an outlying portion of the Kobonqaba sheet.

The Manubi sheet crops out on the right bank of the Kogha, at the junction of the Kabakazi stream, where it is 500 feet thick ; it thins out rapidly to the west, and is represented by thin outliers north and south of the Kabakazi Valley. Near the Manubi trading station the outcrop turns south-west along the top of the escarp- ment on which the forest is situated, and extends some seven miles to a point beyond the Gqunqi station ; north of Gqunqi it is cut into by the stream to a depth of 300 feet, yet the lower surface is not exposed ; it thins out in this direction very rapidly and disappears.

Near Gqunqi there is a short dyke-like mass of grano- phyric rock, rather like the acid dykes in the Kobonqaba sheet ; it traverses both the sedimentary rocks and the Manubi sheet ; the granophyre dyke is a mile long from north to south and several hundred yards wide.

The upper half of the Kologha Valley lies in an ex- tensive sheet, of which only a part is exposed in the Kentani Division, for it is continued across the Kei in the Komgha Division. The main part of the sheet extends eastwards from the Kei below the junction of the Gcua, The clijQfs and slopes on the left bank of the

264 GEOLOGY OF CAPE COLONY

Eei for a distance of four miles rise some 1,200 feet above the river, and two-thirds of the vertical height are composed of the dolerite of the Eologha sheet. On Inver Gcua the sheet dips northwards across the sedi- mentary rocks ; eastwards from this neighbourhood it gives off two thin sheets whose outcrops wind round the north side of the Inver Gcua ridge, and the lower one is continued round the Kombolo and Umnyama Eivers. Another oflf-shoot leaves the main sheet at Kiverstone, and winds round the Eentani escarpment to join the main sheet again south of Kentani ; east of this point the upper sheet separates again and pursues an inde- pendent course as far as the bend of the Eobonqaba Biver at the Columba Mission Station, where it again joins the lower part of the sheet. The lower or main portion of the sheet forms an area of some twenty-five square miles between the Kei and Kologha, and is con- tinued to the north as far as Cat's Pass, where it is cut through by the southern of the gap-dykes which will be mentioned presently. The outcrop has a complicated form owing to the outliers of the overlying shales and sandstones at Nquise, Nxaxo and other places, and the large inliers of the same rocks under the Eentani escarpment. The thickness of this sheet varies greatly ; on the Kei it is as much as 900 feet thick near Mimosa Dale, where both the top and bottom are seen in the cliflfs. On the Kobonqaba Eiver below Nyntughk it is at least 500 feet thick, but about five miles to the north- east, east of Nquise, it thins out completely.

A sheet about 100 feet thick, and apparently uncon- nected with any other sheet, underlies the village of

INTRUSIVE DOLERITES AND ALLIED ROCKS 265

Kentani and the hill of that name about five miles north of the village ; the outcrop appears at the edge of the Kentani escarpment and extends some five miles west- wards on the southern face, and about eight miles north- east round the headwaters of the tributaries of the Kobonqaba.

Near Gentuli and Nqundwyu stations there are two sheets, one is low down near the Eogha Eiver, and the outcrop of the other winds round the slopes about 500 feet higher up. Both these sills are continued in Wil- lowvale, over the left bank of the Kogha.

The last and uppermost sill that needs to be mentioned is the N'Hlambe sheet, which covers a considerable extent of ground in the north-western comer of the Division ; it is cut through by the Gcua Eiver, but is continued far into Butterworth on the east side of that river, and also into Idutywa and Willowvale to the north and north-east. The greatest thickness seen is about 500 feet, near the Gcua, but it is considerably thinner south oi Hughes' beacon. This sheet is cut through by the northern gap-dyke between Tutugha and Gobogobo.

The east and west dykes that traverse the Kentani Division are very remarkable ones; they extend from the Kogha mouth to the Gcua Eiver, and can be fol- lowed across the Kei into Cathcart. From a certain point on the road between the Kei Bridge and Toleni, not far from the Eagle's Nest, a fine view can be ob- tained along the 'valleys weathered out along the course of the dykes ; on the west a long line of valleys with low cols between each pair can be seen on either side of

266 GEOLOGY OF CAPE COLONY

the Kei, and to the east a similar line of valleys stretches for many miles between shghtly higher ground. The dykes are made of a rather coarse rock composed of augite, hornblende, red mica, plagioclase, orthoclase (in micropegmatite), and quartz, with ilmenite, apatite and zircon as accessory constituents. The rock can be called an augite-mica-diorite. The coarse diorite weathers more readily than either the sedimentary rocks or the dolerite through which it passes, consequently the minor streams in its neighbourhood have worked their ways along it rather than through the more resistant rocks, with the result that a series of valleys with low cols between each pair have been formed. These are called '* gap- valleys"^ from the local name of "Transkei Gap" given to the whole series of valleys by the early sur- veyors and residents in the Transkei.

From several spots in the district, such as the N*Debe nek, Gobogobo, Cat's Pass and Lusizi the curious feature can be well seen, and it gives one the impression that a great gouge has been driven along the surface of the plateau and a strip removed. The width of the dykes is at the most about 400 feet.

There are two of these gap-dykes in Kentani, lying parallel and about a mile apart, but they cross, or join and separate again, in the N'Debe Valley. The northern dyke is not continuous on the surface between the Gentuli Eiver and Cat's Pass, but the separate parts are very probably connected underground. The longest valley along the southern dyke is that of the

^ The gap-valleys of the Transkei have been described in detail in tihe Trails. S. A. Ph%l. Soc, Rogers and Schwarz (02).

INTRUSIVE DOLERITES AND ALLIED ROCKS 267

Kabakazi and the lower part of the Kogha, in all about ten miles long.

The intrusion of these dykes was certainly later than that of the dolerites, for they cut through the latter. In its nature and composition the rock forming them is intermediate between the ordinary olivine dolerite and the granophyres mentioned in connection with the Kobonqaba sheet. None of the minerals or structures in the diorites are entirely foreign to the dolerites, and the diorites contain much less quartz and micropeg- matite than the granophyres. Olivine is the only constituent of the dolerites that is absent from the diorites and granophyres.

The gap-dykes must be regarded as a late product of the same molten rock magma that supplied the dolerites ; the more basic portion, represented by the dolerites, was got rid of, and a part of the more siliceous residual matter was extruded after the dolerite sheets had solid- ified; in many places the gap rocks cut through the dolerite as well as the sedimentary rocks, and have solidified as the augite-mica-diorite in the gap-dykes.

A large mass of very acid rock later than the dolerite sheets forms a considerable part of Gonubie Hill in Komgha, it is a microgranite consisting of quartz, orthoclase, and black and white mica. Near Komgha, on the main road to the Draaibosch outspan, there is a large quarry opened up in a thick sheet of dolerite through which run two veins of a granitic rock. The veins are eight inches wide at the most and can be followed downwards as far as the depth of the quarry allows. They are sharply defined and were evidently

268 GEOLOGY OF CAPE COLONY

injected after the dolerite became solid. Under the microscope they are seen to consist of a mixture of quartz and orthoclase with a granophyric (micropeg- matitic) structure, in which lie aggregates of chlorite, pseudomorphs after biotite.

In the west of the Colony the dolerites frequently contain patches of a granophyric intergrowth of quartz and orthoclase ; in many cases these are not in the form of dykes or veins, but occur as constituents of the ophitic dolerites without olivine. In Calvinia and Sutherland some large masses of granophyric rock have been found which are probably of the same nature as the dykes and veins in the Transkei.

In Pondoland and East Griqualand there are some very large masses of dolerite much thicker in proportion to their area than any of the sheets hitherto mentioned. The Tsala hills near Lusikisiki are small examples of these masses, and larger ones are N'tabankulu, Insiswa, Mount Ayliflf, Mount Currey and the Ingeli Mountain. These seem to be thick lenticular or cake-shaped bodies of rock, but their structure is not known in detail. The sedimentary rocks near them do not appear to be dis- turbed, but it is evident that the intrusion of a mass about 1,000 feet thick, such as the Insiswa dolerite, and of no very great horizontal extent, perhaps five miles by two, could not have taken place without the displace- ment of a corresponding volume of sedimentary rock, a disturbance that should leave its effects upon the dip of the beds for some distance from the igneous rock. The only alternative to the displacement of the surrounding rock i^ the absorption of it by the liquid dolerite, but tbi^

INTRUSIVE DOLERITES AND ALLIED ROCKS 269

is a quite untenable supposition on any but a very small scale, which would not explain the phenomena. The dolerites are so uniform throughout the Colony, and inclusions that might be looked upon as remnants of the dissolved sedimentary rocks are so rare, that the molten rock cannot have dissolved the beds it displaces to any considerable extent.

In the higher parts of the Eastern Province thick dykes of dolerite sometimes form more or less circular outcrops. Mr. Dunn found several of these annular dykes hundreds of feet in width enclosing tracts of country some miles in diameter between Windvogel Berg and Queenstown. Mr. Schwarz describes a horse- shoe shaped dyke in Matatiele, and Mr. du Toit found a somewhat irregularly shaped closed dyke round Gala ; the latter dyke coincides in position with a ring-shaped fault, the rock inside the ring has been lifted up re- latively to that outside. A similar feature exists at Indwe.

As a whole the dolerites are of remarkably uniform composition. The constituent that is most variable in amount is olivine. In addition to the plagioclase, augite, olivine and iron ores that form the bulk of the dolerite, biotite is almost always present, sometimes in consider- able quantity, and original hornblende is not seldom met with either independently or in close connection with the augite. The structure varies in one and the same sheet ; the bulk of a thick sheet has an ophitic structure, that is the plagioclase crystals are to a greater or less extent enclosed by the augite, but near the edge of the sheet the augite is granular or forms rather imperfect

270 GEOLOGY OF CAPE COLONY

crystals. In thin dykes and sheets the structure is distinctly porphyritic, crystals of olivine, augite and plagioclase lie in a fine-grained matrix of augite grains and very small plagioclase crystals, often with a con- siderable amount of brown glass. Occasionally an almost pure glass, tachylite, is found at the contact of a sheet or dyke with the surrounding rocks or in the form of thin dykes traversing the dolerite or the sedi- mentary rocks. Tachylite is a black substance with a glassy appearance ; it looks not unlike bright bitu- minous coal, for which it has often been mistaken in this country. The greater specific gravity and hardness of the tachylite, however, distinguish it at once from coal. Porphyritic crystals of augite and plagioclase may occur in the tachylite, and the glass is sometimes converted into an opaque stony material along joints. Both the tachylite and the glassy dykes and sheets owe their peculiarities to rapid cooling. The thick sheets of dolerite naturally took a longer time in cooling than the smaller bodies of molten rock, and consequently the minerals were able to develop more thoroughly in them than in the latter, so the rock as we see it now is coarsely crystalline in the one case and finely crystalline or glassy in the other. The fact that the well-formed crystals of olivine that are often abundant in the coarse dolerites and absent from the fine-grained and glassy dykes points to the fact that the molten rock which forms the latter has been squeezed out of a partly con- solidated dolerite in which the large olivine crystals were retained by the partly formed plagioclase and augite.

INTRUSIVE DOLERITES AND ALLIED ROCKS 271

The questions of the origin of the dolerite intrusions and of the means whereby they were able to force their way between and through the sedimentary rocks are at present beyond our knowledge. The dolerites are quite different in nature from the great intrusions of granite and gneiss that invaded the Pre-Cape rocks, the sources of which were presumably exhausted before the deposi- tion of the Table Mountain sands^pne. The close con- nection of the dolerites in East Griqualand with the vol- canic group at the top of the Stormberg series, in spite of the absence of brown mica and hornblende from the lavas and dykes of Matatiele ^ seems to indicate a common origin of the two groups of rock ; the one consolidated below ground and the other at the surface. Some of the dolerites were certainly intruded after the formation of the Stormberg volcanic and sedimentary rocks, for quite typical members of the intrusions traverse those beds ; it is not assuming too much to suppose that the whole of the dolerite sheets and dykes of the Karroo region belong to one period of igneous activity, so that the later limit to their age is fixed by the occurrence of boulders derived from thick sheets in the Embotyi con- glomerate of the Pondoland coast, probably of Upper Cretaceous age. The Uitenhage conglomerates have hitherto been found only at a considerable distance from dolerite outcrops, so the absence of boulders of that rock from those conglomerates throws no light on the matter. At present it is uncertain whether the intru- sions ceased with the volcanic activity of the latter part

1 Schwarz, Geol Ccmm. (02), p. 66.

272 GEOLOGY OF CAPE COLONY

of the Stormberg period or whether they continued after that period, but the evidence proves that some great intrusions took place after the Stormberg sedimentaries were deposited and before the formation of the Embotyi conglomerates.

During the Stormberg period there must have been an enormous mass of basic rock material lying at an an- known depth beneath the surface of the South African area ready to burst its bonds and rise towards the sur- face when favourable conditions prevailed. What those conditions were is at present a subject for speculation rather than for statement. It may be noted in passing that the mountain building in the south and south-west had probably then reached or passed its maximum, and that the great forces exerted in that process cannot but have influenced the fluid or potentially fluid rock magma. The remarkable freedom from disturbance of the sedi- mentary beds near even the larger sheets and dykes gives one the impression that the igneous rock made its way along channels that were ready to receive it rather than forced a passage through resisting rock. The im- mense areas over which some of the sheets extend with- out very great variation in thickness — the lowest sheet in Calvinia, for instance, certainly extends over an area of 3,000 square miles and probably a third more ^ — prove that the rock must have been in a very fluid condition, and that the enclosing sedimentary beds offered but little resistance to its progress.

It is difficult to form a satisfactory estimate of the

1 Geol. Comm. (00), p. 50.

INTRUSIVE DOLERITES AND ALLIED ROCKS 273

thickness of the rock overlying any particular sheet at the time of its intrusion, but a minimum estimate can be made in the case of the lowest Calvinia sheet, which lies near the top of the Dwyka series in the middle por- tion of its outcrop. It was certainly injected at a time when the Eoggeveld escarpment was not in existence and the rocks now exposed on that escarpment stretched far to the west of their present position. These beds are over 2,000 feet thick, and to this must be added the un- known thickness of the Beaufort and possibly higher beds that have been removed by denudation since that part of the country was exposed to the air. Where the uppermost sedimentary rocks of the Karroo formation are still preserved, as in the Stormberg region, the diflBi- culty of estimating the thickness of the cover at the time of the intrusion is little less than in the country further west, on account of the uncertainty as to the original thickness of the volcanic group and of the exact period of the intrusion during or after the volcanic outburst.

The position of the greatest total thickness of dolerite is at present unknown. At places on the Nieuweveld escarpment there is as much as 800 feet of dolerite in a total of about 3,000 feet of rock exposed in an almost vertical section, and similar proportions of dolerite to sedimentary rock have been noticed on the Eoggeveld cliffs. In a deep bore hole at De Aar over 400 feet of dolerite were traversed within 1,600 feet from the surface ; in the Transkei an even greater proportion of igneous to sedimentary rocks is present in the steep banks of the

Kei and some other rivers, but the total depth of these

18

274 GEOLOGY OF CAPE COLONY

sections is rarely over 1,000 feet. The inclination of the sheets exposed south of the main watershed of the Colony is, on the whole, towards the watershed, but a similar relation has not been made out in the case of the northern sheets, which are not well enough known to allow of a general statement being made as to their behaviour.

The emergence of the Karroo formation from the central portion of the basin probably took place about the close of the Stormberg period or a little earlier ; this emergence seems to have given rise to the east- north-east watershed that is now the main water-parting in the Colony. The intrusion of the dolerite sheets may have added to the height of the surface by arching it upwards, but to what extent cannot yet be decided.

At the contact of the dolerite sheets and dykes with the sedimentary rocks there is generally a noticeable hardening of the latter through a distance varying with the thickness or width of the intrusion.

In the case of sandstones the contact rock is hard and splintery like a quartzite, but, excepting epidote, new min- erals seem rarely to be formed ; the rock becomes harder by the cementing together of the constituent grains by quartz. The epidote gives the green colour to the con- tents of the small cavities found rather abundantly in argillaceous sandstones and mudstones which are trav- ersed by dolerite. Epidote is a silicate of alumina and lime, and is only formed in those sandstones that were originally calcareous. The presence of the amygdale-like bodies of epidote and quartz in the impure argillaceous rocks near dolerite is very characteristic, and has been

INTRUSIVE DOLERITES AND ALLIED ROCKS 275

noted in many districts between Calvinia and the Natal border. Cavities with remarkably smooth surfaces, iden- tical in appearance with the steam-holes in lavas, were formed probably by the conversion of the water held in the then soft sediments into steam, and these spaces were subsequently partially filled by the epidote and other minerals formed by heated water vapour acting on the constituents of the surrounding sediments. The calcareous concretions in the shales are sometimes con- verted into epidote, but the lime-silicate wollastonite has not been noticed in the zone of altered rocks near the dolerite. Shales and mudstones are often changed into homstone, a hard almost glassy-looking rock, which breaks with a conchoidal fracture ; the t3rpical hornstone is only a few inches thjck, and passes gradually into the usual type of rock within about two feet of the dolerite. The hardening effect of the dolerite often extends much farther than any other change in character. A very marked example of this is shown in Plate XVI., a view of the junction of a thick dolerite sheet with the Dwyka conglomerate on the farm Dwas Douw in the Doom Eiver Valley, Calvinia. The rough-looking rock in the upper part of the cliff is the dolerite, and the well- defined columnar rock, forming a vertical krantz fifteen feet high, is the conglomerate. The lower end of the columnar layer is sharply marked, and below it the conglomerate is the usual sandy mudstone containing numerous boulders of many varieties of rocks. The photograph was taken at too great a distance from the krantz to allow the boulders exposed on the joint faces

to be seen. The joints that divide the conglomerate into

18*

276 GEOLOGY OF CAPE COLONY

such regular columns traverse boulders and matrix alike, without deviation.

The larger intrusive sheets of dolerite frequently show a rough columnar structure. Many examples of this can be seen in the sheets which crown the Nieuweveld escarpment in Beaufort West. The sill at the top of Tafel Berg, in Beaufort West, is divided up into columns over 300 feet in length, and from ten to thirty feet in diameter, but they do not traverse the whole thickness of the sill (400 feet).

The country occupied by the dolerite sheets is, as a rule, more fertile than that formed by the sedimentary rocks alone, for the dolerite contains valuable food materials for plants which are set free during the slow decomposition of the rock by the action of the weather and the damp soil. It is only in the eastern portion of the Colony that full advantage can be taken of the valu- able soils derived from the dolerite, for large areas of that rock are there covered with fairly deep soil, and unweathered lumps of dolerite are rarely met with in the soil itself. In the arid central and western districts the soil cannot accumulate rapidly enough to clothe the unweathered rock, for it is not held together suflBciently by grass and other plants to prevent its being washed away by the occasional heavy rains. In the place of the extensive, rich grass-covered plateaux of the east, we find extremely rocky ground sparsely dotted over with small bushes, and yielding grass only for short periods after rain. On Plate XVII. is reproduced a photograph of typical dolerite country behind the Nieuweveld escarp- ment. The innumerable blocks of stone are pieces of

:||;t|!|lil|i|!|(|}itt.i

3t'?W'*^'^*™i'^"B^^I')'i9'?iSS''"''®™'* produced =^DfOg«>^t3gn'^-f§gK*^i*^tIie columnar rock ■?«'|§«:«4ii^Wg?aW»SS''fS'eet high. Dw»a

278 GEOLOGY OP OAf E COLONY

dolerite with a very thin crust of weathered rock ; the blocks are mostly subangular at this spot, but they are often well rounded owing to the strong tendency to spheroidal weathering that is characteristic of the doler- ite. Thousands of square miles in the Upper Karroo are covered with boulders like the foreground in Plate XVII., and the ground is exceedingly troublesome to traverse, either on foot or on horseback, unless one rides a horse born and bred in its neighbourhood.

The colour of the dolerite hills is usually dull red, but extraordinarily vivid crimson and yellow patches are often met with which are due to a lichen growing on the weathered crust of the rock. In certain localities, particularly the krantzes of the dolerite-capped hills in the southern part of the Upper Karroo, the dolerite assumes a blood-red tint when the sun is near the horizon, but this gives place to a duller colour when the sun stands higher.

The most exposed surfaces of large dolerite boulders in the drier regions become coated with a very thin film of deep brown or black material which has often a well-polished appearance. This thin coat seems to be chiefly composed of hydrated oxides of iron derived from the rock immediately beneath it. Dolerite is not the only rock that becomes covered with this dark and shiny film in the dry parts of the Colony. The harder and fine grained portions of the Karroo sandstones behave in the same way, and beyond the limit of the Karroo basin the hard Pre-Cape rocks, both of sedimentary and igneous origin, are often seen to be blackened and polished after long exposure. The implements fash-

ig*;if lifiifi 'fijiiifl'"

I 1

Si

^

1 l1

280 GEOLOGY OF CAPE COLONY

ioned from the jaspery rocks of the Griqua Town beds by Bushmen or Hottentots, which may now be picked up on the surface in Prieska and to the north of the Orange Eiver, usually have their upper surfaces covered with a polished film. It is not known how long a freshly broken rock must be exposed to the sun and air before assuming this character.

The dolerites do not seem to contain any minerals of sufficient value to attract the attention of miners. Copper pyrites is present in small quantities at some localities, and galena fills some very narrow veins in the Eoggeveld sheets, but neither of these has been found in considerable quantity. Dolerite is very dur- able, but it is difficult to work and unsuitable in colour for most building purposes. It is excellent stone for road metal, but its very toughness seems to prevent its general use, for it is difficult to break up. Where roads can be made with the help of heavy rollers it is a very good stone to use.

CHAPTEE VIII.

THE CRETACEOUS SYSTEM.

The Cretaceous rocks in the Colony are divided into two main groups, the Uitenhage series and the Pondo- land Cretaceous series. The two groups have not been found in the same district ; their relative age is deter- mined on the evidence of fossils alone. They both consist of rocks formed near a shore line, and at the base of each group there is a considerable thickness of coarse conglomerate. The Uitenhage beds cover rather wide areas in the folded belt between the Karroo and the coast, resting unconformably upon rocks of all ages between the Pre-Cape and the Ecca beds. The Pondo- land series, on the other hand, occupies two narrow strips on the coast, faulted down against older rocks. The south-western strip is seen to rest unconformably upon beds that probably belong to the Ecca series. Rocks of the same age as the Pondoland beds are found in Natal and Zululand.

The Uitenhage Series.

In the typical area, the valleys of the rivers flowing into Algoa Bay, this series has been subdivided into the following groups : — ^

^ This classification is substantially that of the late Dr. W^. G.

Atherstone.

281

282 GEOLOGY OP CAPE COLONY

Sunday's River beds - Clays, shales and sandy limestones with

marine fossils. Wood beds - - Yellow sanda, shales and limestones with

a few marine shells and numerous

plants. Enon beds - - - Sandstones, marls and conglomerates.

The Enon beds are fonnd at the base of the series throughout the district, but the thickness and nature of the rock differ very much within rather short distances. In the upper part of the Zwartkops River the Enon beds attain a very considerable thickness, as is also the case near Enon ; but near Blue Cliff Station the con- glomerate lying between the sandy and argillaceous rocks of the Uitenhage series, and the surface of the older rocks below, the Bokkeveld beds in this case, is at most oifly a few feet thick, and at places it is entirely absent. •

The Enon beds are here taken to include the Zwart- kops sandstone and variegated marls of Atherstone's classification,^ for the conglomerates are so intimately connected with rocks agreeing with Atherstone's de- scription of these two subdivisions that it is convenient to group the three together. There is indeed much reason to believe that the three subdivisions of the Uitenhage series are to be regarded more as three kinds of deposit formed under different circumstances, but at about the same time, than as successive groups of deposits. In any one spot, such for example as Wolve Kraal on the Sunday's River, the marine Sunday's River beds may be underlain by the Wood beds and

^ Atherstone (57).

THE CRETACEOtJS SYSTEM 283

those again by the Enon, but there is evidence that even in the Uitenhage area rocks like the Enon beds were formed during the deposition of some of the Sun- day's Eiver beds. On the hill west of the native loca- tion at Uitenhage there is a small thickness of grey shale and limestone, containing marine fossils, inter- bedded with red sands and gravels belonging to the Enon type, although to the east of Uitenhage these marine strata are not found interbedded with con- glomerates or sands of the Enon type. The sands and pebble beds west of the native location at Uiten- hage lie against a rather steeply inclined slope of sandstone and quartzite belonging to the Table Mountain series, evidently the shore during a certain stage of the deposition of these rocks. The sands and conglomerates are the deposits formed near the shore, or in most cases probably in steep-sided inlets, drowned valleys in fact, which bordered the sea in which the Sunday's Eiver beds were laid down. The marine beds intercalated with the red beds near the location repre- sent a period of extension or encroachment of the sea on the land-locked inlet in which the red beds were formed.

In the Uitenhage district, then, we find that the Enon beds cannot be regarded as merely the earlier deposits of the Uitenhage period. As far as our know- ledge goes they certainly were the earliest of these deposits, but their formation continued during the lay- ing down of the marine clays and limestones of the Sunday's Eiver beds along the shores of the sea in which the latter were deposited. In the country farther

284 GEOLOGY OF CAPE COLONY

West there is corroborative evidence of this, as we shall see later.

Fragments of wood with a charred appearance, very different from the petrified wood in the Wood beds, occur frequently in the Enon beds, and up to the present time these are almost the only organic remains known from the typical Enon beds in the Uitenhage area.

At Enon, which is situated in a kloof under the Zuurbergen, the conglomerate forms high hills which are curiously carved into crags and caves by the action of the weather on the conglomerate, harder in some places than others. The pebbles, usually about three inches in length and well rounded, were evidently chiefly derived from the Zuurberg quartzites (of Witte- berg age). The matrix in which the pebbles lie is reddish and sandy.

In the upper part of the Zwartkops Valley the conglomerates are very thick, over 1,000 feet, and the same is the case at Hankey in the Gamtoo*s Valley. They are overlain as a whole by the beds called Zwart- kops sandstone and variegated marls by Atherstone, but conglomerate bands are not infrequent in these higher beds. On the right bank of the Zwartkops Kiver below Uitenhage the red clays are worked for brick and tile making. The thickness of conglomerate below these clays and sands is very slight to the south of Uitenhage, where the Humansdorp Eoad leaves the Zwartkops Valley, but the clays and sands contain thin beds of conglomerate. In the clay pits belonging to the Port Elizabeth Brick and Tile Company near Despatch

THE CRETACEOUS SYSTEM 285

Station some bones have recently been found, but they have not been determined.^

In the Bezuidenhout's Eiver Valley from a short distance above Blue Cliff Station to a point some four miles above the railway bridge, the rocks lying below the Wood beds are well exposed at intervals along the river banks. They are reddish yellow sands, red clays and thin sandstones, with occasional pebble beds. Conglomerates like those of Enon are entirely absent from this valley. Near the fortieth milestone on the railway between Uitenhage and Blue Cliff, greenish sandstones very like some that occur in the Bezuidenhout*s Valley, lie against slates belonging to the Bokkeveld series, without the intervention of any conglomerate.

The Wood beds are found overlying the Enon in the northern part of the area, and are especially well seen between Blue Cliff Station and the Witte Eiver below Enon. The valley of the Bezuidenhout*s Eiver below Blue Cliff lies entirely in the Wood beds, and both above and below its confluence with the Sunday's Eiver the rocks are well exposed in the bed of the latter river. The total thickness of the Wood beds in this locality may be as much as 1,000 feet. They consist of various sediments, sands, clays, hard lime- stones and sandstones, and well-laminated shales.

The base of the Wood beds in this valley is taken to be a loose yellow sandstone, seen in a cliff' section

^ Since the above was written Dr. B. Broom has seen these bones, and he informs me that they belong to Dinosaurians, reptiles that were previously unknown from the Uitenhage formation.

286 GEOLOGY OF CAPE COLONY

above Blue Cliff Station. Farther down the valley many large pieces of tree-trunks, one of which is twenty-five feet in length, are preserved in a clayey sandstone. These are probably the trunks of conifers, but no leaves or other parts of the trees have been found with them. Some of the wood evidently lay for some time in the water, for the shells of a small boring mollusc, Oastrochana dominicalis, are found in it in con- siderable numbers. The only other animal remains discovered in these sandstones are oyster shells, and some fragile fragments of large bones, too imperfect to be named. In some hard limestone bands inter- calated with the upper part of the sandstones there are numbers of shells of Psammobia atherstonei. Curi- ously twisted stems, which may have belonged to a cycad, occur in the upper part of the sandstones, as well as stems of Benstedtia.

The chief interest of the Wood beds lies in the well- preserved leaves and other parts of plants that are preserved in the bluish-grey sandy mudstones, clays, and thin limestones between Paltje's Kraal (on Bezui- denhout's Eiver) and the lower portion of the Witte Eiver, including the bed of the Sunday's Eiver near the Dunbrody Mission Station.^ Some of these beds are crowded with the broad fronds of Zamites, a cycad of which several species have been found ; they are accompanied by other cycads, conifers and ferns.^

^ Dunbrody is the Geelhoutboom of the Divisional maps, a name which is used by Atherstone, Tate and other writers.

^ All the plants mentioned in this chapter are named according to Mr. Seward's determinations published in the Annals of the South Africcm Museum^ vol. iv., part 1, 1903. See also Tate (67).

THE CRETACEOUS SYSTEM 287

The following is a list of the plants hitherto found in these beds : —

Ferns— ^

Onychiopm mantdli, Brongn. Gladophlebis hrovmiana^ Dunk.

,, denticulata, Brongn., forma atherstonei (found also

at Herbertsdale). Sphenopteris fittoniy Sew.

„ sp.

Tceniopterisy sp. (found also at Herbertsdale). Cycads —

Zam/Ues recta, Tate. ,, morridi, Tate. ,, africana, Tate. „ rubidgei, Tate. Gycadolepis jenkinsianay Tate. Benstedtia, sp. Carpolithes, sp. Conifers —

Aratuxtrites rogersi, Sew. TdxUes, sp. Brachyphyllum, sp. Gonites, sp. Coniferous wood.

The lowest fossiliferous beds seen on the Witte River contain Onychiopsis mantelli, but the beds containing coniferous wood and reptiUan bones in the Bezuiden- hout's River are probably lower than these. A section taken in an approximately north-east direction along the Bezuidenhout's and Witte Rivers from one side of the Uitenhage deposits to the other is by no means similar towards each end, owing to the much greater development of the conglomerates along the Zuurbergen.

The plant bearing beds pass upwards into bluish rocks containing marine fossils, but the whole of the

i:S-;

L;^-Tii91&<^*aa«^^ggpB>MlB'ody plant beds

Actaonina in the fossil

#

i

iiver Valley and

_ trough lying

^ttbe south-east,

3l;:§Ct with in that

:S5^^^":1^'^*^^**"' -develop-

**i**i-S-

Ski rSi«l'g'< K r§ «I^«>1<

*tlMB£^%i'^iJtKiase, the Bethela- rg) |f;^iif K^lgf MKfi^lt Fan.' ^jBS.l^l94MB»J.Sii9L;CB worked out, and

beds are exposed 'er below Wolve Uitenhage, near laces Bach as the

bads) (from Seward).

in of the fossils in

:.^3'he lowest marine

'er are clays with

-_Ji*l^|i««, Dentaiium and

i^:£Sa!l*are exposed in a

^TtwsoD Bridge, the

-,_-.,^jE§2i^w mentioned some

>^^^ii^j$^t||tersk>iie (ST) sod Stow

l^lzabebh Salt Companj,

■ Sl^g^||^H||l|^i||||m#j|} containing .,^^ttMBIffiffUl«t)FC^|||Bltf;^ eS^rine shells ;

""^JITO^IR B"Sfl®SsJI:S5Shay be older

292 GEOLOGY OP CAPE COLONY

yellow and brown coloured surfaces. The limestones are often crowded with shells, and some layers in the shales are composed almost entirely of the shells of Exogyra imbricata, and others consist largely of Trigonia ventricosa shells. Parts of the skeleton of a reptile re- lated to Plesiosaurus have been obtained from the cliflfe above Picnic Bush.

In the Sunday's RiVer Valley below Addo Station higher marine beds seem to be exposed than are seen anywhere in the Zwartkops Valley. They have yielded a large number of fossils, amongst which Grioceras spinosissimum and Hamites afrioanus are the most in- teresting.^

The following is a list of the more important fossils from the Sunday's Eiver beds, the letters S and Z placed after the names indicate their occurrence in the Sunday's Eiver and Zwartkops Valleys respectively : —

Principal Invbrtebrata from the Uiienhage (Marine) Beds.

Cephalopods —

Bdculites^ sp. Z

Belemnites africanus, Tate S

Grioceras spinosissimum (EEausmann), Neumayr - - S

Hamites africanus, Tate S

Olcostephanvs (Astieria) atherstonei, Sharpe, sp. - - S Z

,, ,, haini, Sharpe, sp. - - - . S

** Amm^onites " »ubanceps, Tate (affinities doubtful) - S Gasteropods —

Actceonina jenhinaiana, Tate ----- S

,, atherstonei, Sharpe S Z

Alaria coroncUa, Tate Z

1 See Krauss (51)» Tate (67), Holub and Neumayr (82), Bain (56) (appendix by Sharpe).

99

THE CRETACEOUS SYSTEM 293

Gasteropoda — continued —

Monodonta haumianniy Neumayr

Natica atherstonei, Sharpe Z

Neritopsia ? turbinata, Sharpe S

Patella caperata, Tate S

Trockus hainij Sharpe Z

Turbo atherstonei, Sharpe - Z

„ hainij Sharpe SZ

Lamellibranchs —

Astarte herzogi, Goldfuss, sp. Z

Imiglandsiana, Tate ----- Z

pinchiniana, Tate S

Avicula hain% Sharpe Z

Gardita nuculoides, Tate ' - S

Geromya papyracea, Sharpe Z

Gorhula ? rochiana, Tate Z

Oucidlcsa jonesi, Tate Z

„ kratuisi, Tate S Z

Gyprina horcherdd, Tate Z

,, rv^fulosa, Sharpe - - - - - - S

Exogyra jonesiana^ Tate Z

„ ivnJbricaJta, Krauss - - - - - S Z

Gastrochcsna dominicalis, Sharpe - - - - S

GervUlia dentata, Krauss S

Lima neglecta, Tate S

„ ohliquissima, Tate ------ S

LiihodomvrS stowianus, Tate ----- S

Modiola atherstonei, Sharpe Z

hainif Sharpe S

rubidgei, Tate S

Mytilus jonesi, Tate ---- .--S-

Parallelodon atherstonei, Sharpe, sp. - - - - S

Pecten projectus, Tate SZ

„ rubidgeanuSf Tate S

Perna atherstonei, Sharpe SZ

Pholadomya doniinicalis, Sharpe - - - - S

Pinna atherstonei, Sharpe SZ

Placunopsis imbricata, Tate S

,, suhjurensis, Tate ----- Z

„ undulata^ Tut^ -»-•»-§

294 GEOLOGY OF CAPE COLONY

LameUibranchs — continued —

Pleuromya hain% Sharpe, sp. ----- S

,, lutraria, Krauss, sp. - - - - Z

Psammobia atherstoneiy Sharpe S

Ptychomya compliccUay Tate, sp. - - - - S

Seebachia bronni, Krauss, sp. Z

Trapezium niveniaiium, Tate, sp. - - - - S

Trigonia hertzogi, Goldfuss, sp. - - - - S Z

,, tatei, Neumayr, S

„ vauj Sharpe SZ

,, ventricosa, Krauss, sp. - - - - Z

,, conocardiiformis, Krauss, sp. - - - S Z Polyzoan —

Berenicea antipodum, Tate S

Worm tubes —

Serpula (several species) SZ

Echinid —

Gidaris pusticliferay Tate - Z

Coral —

hmtruMiy sp. - - -^ Z

The outliers of the Uitenhage series to the west of the division of that name do not contain any deposits similar to the Sunday's Eiver beds so far as is known at present.

In the Gamtoos Eiver Valley (Humansdorp) there are conglomerates and sandstones like those of Enon and the Zwartkops Biver.

In Knysna there are three basin-like areas of quart- zites, sandstones, conglomerates and clay, belonging to the Uitenhage series; the pebbles are mostly, of quartzite derived from the neighbouring hills and mountains made of the Table Mountain series. They occupy deep valleys cut out of the Cape formation, and are themselves cut through by the coast-line. Near the village of Knysna these beds are over 600

THE CRETACEOUS SYSTEM 295

feet thick; the boulders in the conglomerate are often of large size, in places they average a foot in diameter. Along the Bitou River there is a great mass of con- glomerates and loose sandy beds with pockets of lig- nite. The conglomeratic beds in the Bitou basin are peculiar in that the included fragments of rock are angular instead of being well rounded as is usually the case with the pebbles in the Uitenhage conglomerates. The third basin is in the valley of the Pisang River; the beds in it are less conglomeratic and more sandy and clayey than those of the other two areas, and some of the beds are quartzitic owing to the deposition of silica between the grains of the rock. Near Seal Point casts of Trigonia conocardiiformis have been found in the sandstones and conglomerates. This is the only marine fossil yet found in the conglomerates of the Enon type, but as it is a very characteristic member of the fauna of the Sunday's River beds its occurrence is of great interest. It is evident that the water in which the Pisang River beds were deposited must have been salt, or at least so near the sea that the shells of the dead bivalves could be washed back into it by strong tides. But the absence of marine fossils from the bulk of the Knysna conglomerates and sandstones can only be interpreted on the supposition that the rocks were laid down in water sufficiently far removed from arms of the sea to be free of marine inhabitants.

The occurrence, which has been already mentioned, of a bed of marine fossils between the red gravels and sandstones north-west of Uitenhage, proves that the sea 8^t one time ipvade^ the non-marine area, apd tiie

296 GEOLOGY OF CAPE COLONY

Trigonia of the Pisang Biver beds points in the same direction. The Trigonia of Pisang Biver proves also that these rocks were formed at about the same time as the Sunday's Biver beds; whether the latter ever spread far to the west of their present limits must remain an open question, but there can be no doubt that the conglomerates and sands of the Enon type were being laid down in the west while the sea occupied the position of the lower part of the present Sunday's Biver Valley.

Still farther west, in the divisions of Mossel Bay, Biversdale, Bobertson, Swellendam and Worcester there are large areas of conglomerates, sandstones, shales and mudstones, resembling to some extent the Enon beds but containing some varieties of sediments not met with in the Uitenhage Division ; and again in the country between the Langebergen and the Zwartebergen, in the divisions of Willowmore, Uniondale and Oudtshoom, there are large areas of similar rocks that in spite of the absence of fossils must be relegated to the Uitenhage beds.

All these masses of rock occur in a more or less similar manner; they occupy basins partly cut out of the older rocks, but in part due to earth movements subsequent to the Uitenhage period. They extend far below the present level of the rivers traversing them, and are generally elongated in an east and west direc- tion, roughly parallel to the general strike of the older rocks.

The Mossel Bay area is perhaps the most interesting gf these patches pf Uitenhage beds, for it alone hag

THE CRETACEOUS SYSTEM 297

yielded fossils that can be compared with those of the Uitenhage district. It is rather irregular in shape, about fifty miles long from east to west, and at the most fifteen miles wide. The northern boundary is formed by the Langebergen, and tne southern in part by the coast between Mossel Bay and Great Brak Eiver, and, west of Mossel Bay, by the Bokkeveld and Table Mountain series. North of Mossel Bay the George granite and the highly altered Malmesbury beds project far into the area of Uitenhage beds, dividing its eastern end into two tongues which join west of the main road to Robin- son's Pass. The Uitenhage beds thus rest upon granite, Malmesbury beds, Table Mountain sandstone, Bokkeveld and Witteberg beds at different places ; it has been noticed that, to a certain extent, the pebbles and boulders, for the included blocks reach a length of more than eighteen inches in the conglomerates, came from the rocks that are close at hand rather than from those forming the mountains. Thus in the Euitersbosch Valley there is a large propor- tion of granite boulders in the conglomerates which are well exposed round the western end of the George granite. Near Bottle's Kop, that curiously shaped hill of quartz- ite and quartz schist (probably belonging to the Table Mountain series), which is so conspicuous to the north of the Mossel Bay-George road, the conglomerate con- tains many fragments of the quartzitic rock. Along Weyer's River, and generally along the western border of the conglomerate, pebbles derived from the Bokkeveld beds are very abundant. At Cape St. Blaize the con- glomerate is represented only by a very thin layer of breccia, composed of angular fragments of the under-

298 GEOLOGY OF CAPE COLONY

lying Table Mountain sandstone. Along the northern boundary Table Mountain sandstone pebbles are by far the most abundant, and this is also the case in the con- glomerates lying at a considerable distance above the base of the Uitenhage beds, as at Honig Klip Eloof, where there are magnificent sections through a coarse, white conglomerate, composed almost entirely of pebbles and boulders of Table Mountain sandstone and quartz- ite ; the Honig Klip Kloof conglomerates are interbedded with pale, sandy beds, and probably form about a half of the whole thickness, some 500 feet, exposed along the valley. The pebbles in the conglomerates are usually very well rounded ; they must have been rolled about for a long time and reduced to their present form before being buried in the sandy or muddy matrix of the rock.

The beds of conglomerate are by no means confined to the base of the series ; they seem to occur at inter- vals throughout the whole thickness of rock, and are separated by beds of shales, sands or mudstones.

The maximum thickness of the Uitenhage beds in Mossel Bay is rather considerable. They lie compara- tively undisturbed, for the angles of dip are low ; they certainly descend below sea level in places, and the bed of the Gouritz River, both just below the gorge through the Langebergen and to the north of Roode Hoogte lies in sandstone and pebble beds of this series ; they form practically the whole of the hills between Herbertsdale and the watershed north of the Stink Eiver. The tops of these hills are mostly formed by some twenty feet or less of the surface deposits resting unconformably upon the Uitenhage rocks, but as the average height

THE CRETACEOUS SYSTEM 299

of the hills is over 1,000 feet the greatest thickness of the Uitenhage beds is probably rather over that amount.

At a spot about three miles east of the village of Herbertsdale there are some shales containing plant remains. Three species have been recognised amongst them, Gladophlehis denticulata forma atherstoneij which also occurs at Dunbrody in the Wood beds, TaniopteriSf also found at Dunbrody, and Taodtes, The shales are very soft and easily weathered, so that the exposures are very few. The Herbertsdale outcrop has been opened up for prospecting purposes owing to the pres- ence of small fragments of black lignite, which led to the expectation of a workable deposit of coal. No such reward met the searchers, but their work furnished the means of obtaining the three species of plants mentioned above. In a fairly well watered country like the Mossel Bay Division soft shales are usually covered up by soil and vegetation, and in the absence of quarries, pits and cuttings, it is extremely diflScult to get out any fossils there may be in the rock. Although the Herbertsdale plants are almost the only ones yet found west of the Uitenhage district there must be many more awaiting discovery, and any further specimens will be of very great interest.

Many casts of parts of stems have been found in the hard sandstone of Cape St. Blaize, but hitherto none of them has been determined.

The underlying surface of the Cape formation and pre-Cape rocks is probably very uneven. In the Lang Touw Valley below Herbertsdale some sections are exposed, showing the conglomerates and sands of the

300 GEOLOGY OF CAPE COLONY

Uitenhage beds resting against a steep almost cliff-like face of Bokkeveld beds, the north slope of an old valley running east and west. The west end of the George granite is a high ridge reaching a height of perhaps 1,000 feet above the lowest visible portion of the con- glomerates in the Brandwacht Valley to the south, and a less though still considerable height above the con- glomerates between it and the Langebergen.

The sandstones of Cape St. Blaize, lying horizontally and unconformably upon the Table Mountain series, which dips steeply southwards, are much harder than the sandy beds of the Uitenhage series usually are, but not far to the west along the coast the beds are much softer, very like the sandy clays that occur north-east of Heidelberg. The Cape St. Blaize rocks form a narrow outlier lying east and west and are separated by about four miles of rough country of Table Mountain sand- stone from the large area of Uitenhage beds, which are exposed at sea level near Hartenbosch.

The outlier of Uitenhage beds upon which the village of Heidelberg is built is about thirty miles long from east to west, and eight wide at its broadest part near the west end. It stretches from the west side of the Slang Eiver in Swellendam to Assegaai Bosch in Kivers- dale, and both the Duivenhoek's and Kaffir Kuil's Eiver traverse it without exposing the underlying rocks. The total thickness of the beds must be considerably over 1,000 feet, for they have a variable and low but on the whole northerly dip throughout, although owing to want of outcrops it is impossible at present to state bow far the observed dips are due to subsequent move-

THE CRETACEOUS SYSTEM 301

ments and to what extent they are original features. The conglomerates and sands may well have accumu- lated at moderate angles, and sections along the new railway between Heidelberg and Eiversdale show masses of gra,vel piled up very irregularly and lying between sand and clays which are themselves false bedded. Such sections show that the sediments were deposited in water in which strong and varying currents prevailed. Much of the Heidelberg outlier, however, is composed of thin bedded shales and mudstones, which must have been laid down in quiet water, although thin pebble beds are frequently found with these fine-grained sedi- ments. The outlier is certainly basin-shaped, and no connection has been traced with the Mossel Bay beds to the east, or with the Swellendam basin to the west. It is probable that subsequent earth movements have disconnected these basins of Uitenhage beds, aided of course by denudation, which has swept away perhaps the greater parts of the Uitenhage beds originally de- posited in that part of the Colony.

The Heidelberg beds chiefly consist of conglomerates, sands, red and grey mudstones, shales and clays ; near Heidelberg there are some peculiar hard white argilla- ceous beds, which are quarried for foundation stones, and with them some pale siliceous shales crowded with the thin shells of an entomostracan, Estheria anomala^ Eupert-Jones, a fossil that is also found at many other places in the Heidelberg outlier, but hitherto not known from the Uitenhage district, or from any of the other outliers of the Uitenhage series. At Heidelberg village the clays exposed by the excavations for the railway

302 GEOLOGY OF CAPE COLONV

station contain the Estheria anomala and another ento- mostracan genus, probably Cypris ; some badly preserved lamellibranch shells closely resembling the Psammobia athersUmei of the Uitenhage district hdve been found in the same beds. Some fish scales belonging to a ganoid genus, some indeterminable plant remains, and a wing case of a beetle complete the list of fossils from the Heidel- berg outlier. It is certain, however, that a consider- able variety of fossils will be found there in the future. The varied nature of the scanty remains mentioned above show that many classes of organisms were repre- sented in the waters in which the Heidelberg beds were deposited, and only careful searching is required to pro- duce good specimens. The most favourable localities for fossil hunting in that district seem to be the Doom Eiver Valley west of Heidelberg village, the Spiegel River Valley, and the Klein Vette River north-west of Riversdale, but in the course of time new exposures will be opened up along roads, and for various other purposes^ in places where the rock underlying the soil cannot now be seen. With the two villages of Heidelberg and Riversdale to supply people whose curiosity is suffi- ciently aroused to make them look about the neigh- bourhood for fossils there should be a long list of them before many years have passed.

The beds in which the fossils have been found are grey or whitish in colour ; the red clays, sands and marly beds seen to the north of Heidelberg have not proved fossiliferous. It is generally found that red- coloured rocks are not fossiliferous. The red colour is due to the higher state of oxidation of the iron com-

THE CRETACEOUS SYSTEM 303

pounds than is the case in the green, blue and grey rocks ; when much organic matter was present during the deposition of the mud, the red, highly oxidised, iron compounds were reduced to a less oxidised state, and these give a bluish-green, or grey colour to the mud. The amount of organic matter present was of course closely connected with the number of living organisms that might leave traces of their existence in the shape of fossils, hence it is always to be expected that beds that are uniformly red throughout, and therefore to be regarded as having been red when formed, should yield few or no fossils.

On the watershed between the Doom and Klein Doom Eivers the cuttings for the new railway line to Eivers- dale revealed the presence of some limestone bands showing cone-in-cone structure, and a few thin veins of gypsum. The cone-in-cone limestone breaks up in a very curious fashion ; the rock appears to be built up of a number of cone-shaped bodies, closely pressed together, with their axes perpendicular to the bedding planes. The g5rpsum fills narrow cracks and joint planes, and is a product of the mutual decomposition of pyrites and carbonate of lime m the shales.

A very interesting point in the Heidelberg basin is the occurrence of a mass of mehlite-basalt amongst the gravels and sands near the northern boundary of the area on the farm Spiegel Eiver. The rock forms the top of a low hill on a ridge running south from Amandel Bosch Eug, and the outcrop is roughly circular in out- line, with a diameter of not more than 300 feet. The boundary has not been exposed, so that the contact with

304 GEOLOGY OF CAPE COLONY

the Uitenhage beds is invisible. The presence of the conglomerates and sands in the steep kloofs on either side of the ridge, and on the surfc^^ both to the north and south of the melilite-basalt, and the absence of fragments of the latter from the conglomerates, prove that the igneous rock is surrounded on those sides by the conglomerates, and that it was very probably of later age than they. This evidence would be considered sufl&- cient proof of the intrusive nature of the igneous rocks, as regards the Uitenhage beds, if similar intrusions were known elsewhere in those beds, but as this small mass of igneous rock is the only one known in the Uitenhage beds, an actual exposure of the contact would be very welcome. The form of the igneous rock is quite con- sistent with the supposition that it fills a pipe, a more or less cylindrical channel passing vertically downwards like the channels connecting volcanic vents with the source of supply below the surface ; and the nature of the rock itself is not opposed to that idea, for it is a thoroughly glassy rock composed of crystals of olivine up to about a tenth of an inch in length, embedded in a ground-mass of small crystals of melilite, grains of augite, minute crystals of perofskite and magnetite, and brownish glass. Melilite-basalt is not a common rock, far less usual in volcanic districts than the less basic rocks containing felspar, and when the Spiegel Eiver outcrop was found it had not been observed elsewhere in the Colony. Quite recently, however, melilite-basalts have been found in the Sutherland Division in close connection with pipes in the Karroo formation containing some of the rocks and minerals characteristic of the Kimberley diamond

THE CRETACEOUS SYSTEM 305

pipes. In fact, it seems that the melilite-basalt of Spiegel Eiver fills a pipe that is similar in nature to the pipes filled by the *' blue ground " of Kimberley and other parts of South Africa. This resemblance, so much strengthened by the occurrence of melilite-basalt in the Sutherland pipes, will be discussed in a later chapter, where the bearing of the Spiegel Eiver rock upon the age of the Kimberley pipes will be pointed out.

Near the village of Swellendam there is an isolated basin of Uitenhage beds. Its exact limits are not known, as it and the surrounding rocks belonging to the Bokkeveld-Witteberg series are much hidden by gravels and alluvium of a much later age, but it is about twelve miles long and five wide, and extends from the village, the eastern part of which is built on it, to beyond the Buflfeljagt*s Eiver. The rock near the western end seems to be chiefly composed of conglomerates contain- ing pebbles derived from the Malmesbury, Table Moun- tain, Bokkeveld and Witteberg series. At the railway station a bore-hole put down to the depth of 350 feet did not reach the bottom of the conglomerate. Near the lower part of the hole the bore passed through a boulder of micaceous slate seven feet in diameter. There are but few exposures of these beds, but the railway cuttings east of the village show that there are sandy clays interbedded with the conglomerates.

The Swellendam beds have generally a low, north- easterly dip, and the basin-shaped area occupied by them must in part be due to earth movements subsequent to their formation. The west end of the basin must have

a very steep slope, for the slates forming the basin crop

20

306 GEOLOGY OF CAPE COLONY

out immediately west of the river that runs through the village at a much higher level than the railway station, which lies only a few hundred yards across the river, and where the bore hole did not reach the base of the conglomerates at 350 feet.

Two outcrops of red sandstone and conglomerate occur in the bed of the Groot Vader's Bosch Stream and on the hill just south of it, where the main road leaves the valley; these outliers are situated between the Swellendam and Heidelberg basins and point to the former connection of the beds filling them; a minute examination of the district, with particular attention to all excavations and cuttings that may be made, will certainly prove the greater extension of the Uitenhage beds in this area.

In the country south of the Zwartebergen the gravels and other deposits belonging to a comparatively recent period often hide the underljdng rocks, and in some cases the gravels may be mistaken for the Uitenhage conglomerates. With the high level gravels there are often associated compact rocks whose grains are ce- mented together by silica, carbonate of lime, or fer- ruginous matter, and when once a person is well acquainted with these somewhat peculiar rocks he can readily recognise them in even very small fragments ; their presence in a gravel at once distinguishes it from the Uitenhage conglomerates. The high level gravels themselves can usually be distinguished from the Uiten- hage beds by the fact that they cover flat hill tops, often bounded on one or more sides by a low step or krantz, due to the gravels offering more resistance to

THE CRETACEOUS SYSTEM 307

the weather than the underlying rock, whether the latter belong to the Uitenhage beds or the Bokkeveld or Witteberg series. In the Mossel Bay basin excellent sections showing the unconformity of the gravels and surface quartzites to the Uitenhage beds can be seen in the valley of the Nauga Eiver east of Herbertsdale ; a fine example of a similar unconformity in the Willow- more Division is shown on Plate XIX.

West of Swellendam there are two more isolated basins of Uitenhage beds, one stretches from Eobertson to Ash ton, and the other from south of Goudini Eoad Station to beyond Nuy, passing just south of Worcester. The beds exposed in these basins are red conglomerates, containing pebbles from all the rock series from the Malmesbury to the Ecca, which crop out within short distances of the Uitenhage beds. The latter rest upon the older rocks both to the south and north of the Worcester fault, and are apparently unaflFected by the fault, which must consequently have been in the same state in Uitenhage times as it is to-day. The conglom- erates are well exposed on the banks of the Kogman's Kloof Eiver above Ashton Station; on the road to Waai Kloof from Worcester, and in a railway cutting just outside Worcester Station.

Between the Langebergen and Zwartebergen a very

considerable tract of country in the divisions of Oudt-

shoorn and Willowmore is occupied by the sandstones

and conglomerates of this series. The longest area

extends from the west or right bank of the Gamka

Eiver below Calitzdorp to near Tover Water Poort, a

distance of over seventy miles, but near Meiring's

20*

308 GEOLOGY OF CAPE COLONY

Poort the width of the area is very small, under a mile ; south of Coetzee*s Poort the width is over twelve miles. Along the northern edge of the area the con- glomerates lie directly upon the Cango series ; but east of Meiring's Poort they rest upon the Table Mountain Sandstone along the northern edge, and upon the Table Mountain and Bokkeveld series on the south. The 01ifant*s and Gamka Eivers flow for a considerable distance in rocks belonging to this series.

The conglomerates between Coetzee's and Potgieter's Poorts closely resemble those at Enon. They are red rocks, and weather into curiously rugged crags with numerous small caves, and at places two caves on opposite sides of a crag have met, with the result that the crag has a hole through it. These conglomerates were deposited against steep banks formed by the older rocks. The conglomerates as a whole lie at the bottom of the basin, or rather they crop out on its edge, and are probably continuous under the sandstones and shales that occupy a wide area within the basin. Very probably the conglomerates were in part formed near the sides of the valley while the finer grained sediments were being deposited farther away from the hills. Al- though the conglomerates are chiefly found in the peripheral portion of the area they are not confined to it, for near Oudtshoorn thick beds of conglomerate occur at a much higher level than the sandstones on which the town is built. The sandstones are seen between Calitzdorp and Vlakte Plaats, and at the town of Oudtshoorn, where they are much used for building purposes. They are rather soft sandstones, not quartzitic.

THE cretaceous; SYSTEM 309

and are usually greenish in colour. The sandstones and shales contain bits of fossil wood, and near Vlakte Plaats masses of lignite suflSciently large to be dug out and used for fuel have been found, but this lignite, as is the case with similar materials elsewhere in the Uiten- hage beds, near Herbertsdale and in the Sunday's Eiver Valley, is not found in layers that are thick and constant enough to repay systematic working.

No determinable fossils have been found in the Oudt- shoorn and Willowmore basin.^

The depth to which the Uitenhage beds in this basin extend below the surface is not known.

Many small outliers of conglomerates and sandstones belonging to this series occur to the east and south-east of the Oudtshoorn- Willowmore basin, in the valleys of the Olifant's and Baviaan's Kloof Eivers.^ The beds often have considerable dips, and appear to be the rem- nants of deposits that filled up these valleys before the present rivers re-excavated them. The original form of the deposits modified by subsequent earth move- ments and denudation are jointly responsible for the small detached basins that are now observable.

There is still very much to be learnt about the nature and distribution of the Uitenhage beds in the Colony ; the Uitenhage district itself has yielded but a small part of its history, although it has attracted more attention from geologists than any other area in the Colony, excepting perhaps the Cape Peninsula and the Diamond

^ Since this was written Mr. Muller Rex has sent two Dinosaurian teeth from the Oudtshoorn sandstone to the S. A. Museum.

^ A description of these outliers by Mr. Schwarz will be found in Geol. Comm. (03).

310 GEOLOGY OF CAPE COLONY

Fields. At present the limits of the marine beds are not known exactly, nor have any outliers of them been discovered, although it is very Hkely that they exist to the east if not to the west of the Uitenhage area.

At the commencement of the Uitenhage period the southern parts of what is now Cape Colony must have been very mountainous. Great valleys with mountains towering on either side stretched east and west for long distances, and so far had denudation proceeded that all the rock series from the Pre-Cape to the Karroo forma- tion were exposed at the surface. The height of the mountains above the bottom of the valleys was greater than it now is ; allowing for earth-movements subse- quent to the Uitenhage period that have in some cases at any rate brought about the depression of the valleys, the amount of rock removed from the mountain ridges since the beginning of that period must be very con- siderable, since it includes a large part of the material now forming the Uitenhage beds as well as that removed since the close of the period. The rivers, which before that time were able to carry away the mud, sand and pebbles delivered to them by the mountain streams, became unable to cope with their work, and their beds consequently became choked up with debris, at first as a rule of a coarse nature including many large boulders and pebbles together with a large quantity of sand. These accumulations are the conglomerates that lie below the fine grained rocks, the Enon beds of the Uitenhage district and the similar rocks of the outliers to the west, but it is by no means certain that the red conglomerates round the Oudtshoorn-Willowmore basin,

THE CRETACEOUS SYSTEM 311

for example, were formed at precisely the same time as the Enon conglomerate itself. One possible cause of this change of conditions, the change by which the area became one of deposition or accumulation instead of a region in which the destructive agencies had full sway, may have been that the level of the land surface as a whole was reduced with regard to the level of the sea into which the old rivers flowed. Whether such a down- ward movement of the land took place uniformly or whether some parts were depressed more than others is not easy to determine, although the fact that the marine beds have not been found west of Knysna seems to point to an unequal distribution of the change in level. Had the sinking gone on continuously and equally over the whole area we should expect a gradual extension of similar sediments from the sea landwards, i.e., con- glomerates at the bottom, then fine grained rocks of fluviatile origin, and, finally, marine beds on the top. During the uniform and gradual depression of a tract of country, in the course of which the actual grade or inclination of the river valleys would not be altered, those parts of the valleys left above the level of the sea at any one time would naturally be able to carry on their work as they did before the downward movement set in. In the case of the Uitenhage beds, however, the state of affairs is quite different, no such regular spread- ing of the deposits from the marine area is noticeable ; on the contrary the Uitenhage district is the only one where a series of conglomerates, fluviatile sands and muds and marine beds has been observed, and even there the red conglomerates and sands near the native

312 GEOt/xn" OF CAPE COLONY

location at Uitenhage are intercalated with by no means the lowest of the marine beds, showing that a part of the shore of the sea lay round the end of the mountains near Uitenhage some time after the earliest marine beds were formed in the neighbourhood. If the sea ever reached the western outliers of Oudtshoorn, Heidelberg and Swellendam, no trace of its presence has yet been found, and in any case over 1,000 feet of non-marine sediments were piled up before it did so. These filled up the old valleys to the extent of at least 1,000 feet, very probably to a much greater depth, possibly above the level of the lowest passes over the Langebergen and Zwartebergen. If the movement which aUowed the sea to gain access to the Uitenhage district can be shown to have been unequal, so that the lower portions of some of the east and west valleys were raised, the formation of the basins, as well as the gathering in them of such large quantities of conglomerates, sands and shales will be explained.

There is, however, another possible cause which would account for the old rivers receiving more debris than they could carry away, and that is the coming in of a drier climate than had formerly prevailed.^ Under such conditions the supply of rock debris would be as great as, if not greater than, during the preceding moister period, for the hills would be less protected by vegetation, and the breaking up of the naked rock by change of temperature would proceed rapidly. The

^ For an excellent account of the rocks formed under desert condi- tions, such as here spoken of, the student able to read German should peruse Professor J. Walther's Denudation in die Wiiste.

THE CRETACEOtJS SYSTEM 313

occasional rain storms in such a climate sweep down vast quantities of gravel and sand, rounding off the edges of the rock fragments and thus producing pebbles and boulders of the ordinary shapes. The prevalence of unfossiliferous red-coloured conglomerates and sands, especially near the base of the series, in Uitenhage, Oudtshoorn, Heidelberg, Swellendam, Eobertson and Worcester, supports this explanation ; and the irregular piling up of much of the red rocks is evidence in the same direction.

The grey shales and muds of the Wood beds in the Uitenhage Division were probably formed in the waters of a river that had direct communication with the sea, for the oyster shells, the GastrochcBna in the logs of wood, and the Pecten, all found in the Wood beds near Dunbrody, point to the proximity of the sea. The plant-bearing shales near Herbertsdale, and the grey shales with Estheria and the other fossils previously mentioned in the Heidelberg area, have not yielded any proof that the water in which they were laid down was in close proximity to the sea. These beds may have been formed in shallow lakes or lake-like ex- pansions of the river which still drained the country. It cannot be held that the valleys were entirely closed, that they were in a region that had no outlet to the sea; for in such districts the salts that are contained in small quantities in all rocks become concentrated in the water that temporarily or permanently occupies the lowest levels, and form layers of crystalline rock- salt, gypsum and other minerals that are interbedded with the sand and mud carried into the same basins.

314 GEOLOGY OF CAPE COLONY

No traces of such minerals have been met with in the Uitenhage beds/ and their absence is good evidence against the supposition that the isolated basins of the Uitenhage outliers were originally entirely without out- let to the sea.

The description of the outliers on previous pages shows distinctly enough that the deposits vary con- siderably from one basin to another, that although their general nature is very much the same, the order in which they occur is not in the least identical. The position of the outliers also shows that they were formed in separate valleys, in each of which the de- posits were governed by the local conditions. Whether during the later part of the period, represented by beds that have mostly been swept away by denudation, all the outliers were connected, and sediments were spread over the whole of the district in which the outUers occur as well as beyond its limits must be left to the future to decide. It is quite possible that evidence sufficient to settle the question will be forthcoming.

Whether this was the case or not, the absence of transverse valleys in the Langebergen filled with the Uitenhage beds is specially worthy of note, for it shows that the Oudtshoom basin was then quite distinct from the valleys south of the Langebergen, and that the rivers which now traverse that range had no existence in those days. The Uitenhage beds both north and

^The gypsum of the Heidelberg outlier is evidently derived from the shales by the mutual decomposition of some of their components. Since the above was written Mr. Schwarz has found gypsum in some of the Willowmore outliers ; see Oeol. Comm. (03), p. 114.

THE CftETACEOtJS SYSTEM 315

south of the Langebergen extend below the present level of the Gamka-Gouritz River bed, and the dis- locations undergone by the Uitenhage beds in those areas do not seem to be great enough to account for the complete isolation of the beds on either side of the mountains; the sharply defined gorges of the Gouritz Eiver through the Gamka hills and Langebergen seem to have been cut since Uitenhage times, for they con- tain no outlier of the rocks that one would expect to find had they been of pre-Uitenhage age.

Considering generally our present knowledge of the Uitenhage beds, it leads to the conclusion that the depression of the area as a whole, which allowed the sea to encroach upon the previous land surface in the Uitenhage district, was not uniform, but that the grade of some of the valleys was at the same time altered, and that this may have been accompanied by a drier climate.

It is, of course, an interesting problem to decide at what stage in the history of other parts of the world these events in South Africa took place, and the com- parison of the Uitenhage fossils, of which lists have been given on a previous page, with those found else- where afford a means of doing so, although more evidence will be required before the question can be satisfactorily answered.

The plants have recently been examined by Mr. Seward,^ who came to the conclusion that they are related to both Jurassic and Wealden (Lower Cretace-

' Seward (03), pp. 1-46.

316 GEOLOGY OF CAPE COLONY

ous) plants of other countries, but that the relationship as a whole was closer to the Wealden than the Jurassic flora.

Mr. F. L. Kitchin, who has worked at fossils from allied rocks of India, and who is making an examination of the Sunday's Eiver fossils, has kindly given me the following note upon the question of the relationship of the Uitenhage moUuscan fauna . —

** The marine fauna of the Uitenhage series bears abundant evidences of its Cretaceous affinities, and the view, formerly held by some writers, that either a whole or a part of the marine beds is to be brought into parallel with the Oolitic rocks of Europe,^ can no longer be upheld. Sufficiently conclusive is the occurrence of Hamites, Bacvlites, CrioceraSy Olcostephani of the division Astieria, TrigonicR of the section ScabraB, Ptychomya and other bivalve genera which made their first appearance in Lower Cretaceous rocks. Indeed, it is only possible to follow Neumayr and others ^ in maintaining that this fauna is of Neocomian age, although owing to lack of detailed agreement with the faunas of similar age in Europe, a narrower correlation cannot with certainty be established.

** The occurrence of Okostephanus (Astieria) atherstonei and close allies in the Valenginian and Hauterivian of Europe may perhaps give the best indication of the position occupied by these marine beds.

** While the cephalopods of the Uitenhage series supply connecting links to the fauna of the Neocomian with which we are familiar in Europe, certain conspicuous forms amongst the bivalves appear, on the other hand, to

1 Bain (56) ; Tate (67). « Neumayr (82). See also Krauss (47).

THE CRETACEOUS SYSTEM 317

possess no close European allies, but serve to connect the Uitenhage fauna in unmistakable manner with that of the marine beds of the Oomia group in Cutch. Peculiar forms of Trigonia, in particular, play an important r61e in both cases, lending a similar aspect to the faunas developed in these geographically remote regions, at the same time helping to bring these moUuscan faunas into marked contrast with that of the European Neocomian. The characteristic Uitenhage form, Trigonia ventricosay occurs abundantly in the Oomia beds, and is recorded from strata of like age near Coconada on the east side of the Indian peninsula and also in the extra-peninsular district of Hazara.

** On the African continent itself, the only deposits of Neocomian age which may be safely correlated with the Uitenhage beds occur in German East Africa, not distantly remote from the coast-line. Although the cephalopods fail us as a basis of comparison, the occur- rence of Trigonia ventricosa and some other bivalve forms seems to constitute sufficient grounds for the correlation ; the same beds in German East Africa, it is interesting to note, also furnish evidence of their connection with the Oomia group in Cutch by the presence of a species of Trigonia which has not yet been found in South Africa.

*' The fauna of the Belgrano beds in Patagonia may also be considered to display afl&nities to that of the marine Uitenhage strata, more especially by the occur- rence of Trigonia subventricosa, Stanton, which closely resembles the larger form of THgonia ventricosa, and Trigonia heterosculpta, Stanton, which is with little doubt allied to the South African Trigonia vau.**

318 GEOLOGY OF CAPE COLONY

It thus appears that both the flora and fauna of the Uitenhage beds have distinctly Neocomian characters.

The Cretaceous Eocks of Pondoland.

On the coast of Pondoland the Cretaceous rocks occur in two narrow strips faulted down against the Table Mountain series that forms the greater part of the coastal district.

The Umzamha Group}

The larger and more interesting of the two, the Um- zamba group, lies near the Natal boundary, stretching from a point about three miles south-west of the Umtamvuna River, which is the limit between the two Colonies, to near the Umtentu Eiver, a distance of some twelve miles. The greatest width of the strip is not more than about 700 yards, for the Table Moun- tain sandstone crops out in the grass-covered ground at that distance from the shore along part of the coast, elsewhere it approaches the beach more closely and at each end of the Cretaceous outcrops appears on the shore itself. The actual contact of the Umzamba beds with the Table Mountain series has not been observed ; it is everywhere hidden by the sand that forms dunes behind the beach and often covers up the Cretaceous rocks. The Umzamba beds lie horizontally, and even where their outcrops are very close to the nearest out- crop of Table Mountain sandstone, as on the right bank of the Umzamba Eiver about 300 yards from the mouth, they are of the same nature as on the shore,

1 Baily and Garden (56), and Griesbach (71).

THE CRETACEOUS SYSTEM 319

and do not show any tendency to become conglomeratic, as would be expected if the junction were an ordinary one of a beach deposit with a shore. The Table Moun- tain series forms rather high ground close behind the Umzamba beds, rising some 300 feet above them within a short distance. It is very probable that the junction is a faulted one, like the junction of the Embotyi beds farther to the south-east.

The Umzamba beds form a line of low cliffs (see Plate XVIII.) extending about a mile north-eastwards from the sand-spit on the left bank of the Umzamba mouth, and they are also exposed at low tide on the shore between the levels of high and low water, where, however, they are frequently more or less concealed by sand. Between the Umzamba and Umtentu Eivers they are exposed between tide marks only, and do not crop out at the back of the beach below the sand dunes.

The rocks chiefly consist of shelly limestones and hard sandy clays containing much carbonate of lime. These two kinds of rock are interbedded ; the shelly limestones are thinner than the clayey beds, and at the same time oflfer more resistance to the weather and the sea, so that on the low cliffs they appear as projecting shelves or ledges separated by the softer beds. The latter have been deeply worn away by the sea, thus giving rise to Unes of caves, whose floors and roofs are the hard shelly limestones. The native name of the cliffs to the north-east of the Umzamba mouth is Izinhluzabalungu, " houses of the white men," per- haps in reference to the use of the larger caves by a shipwrecked crew.

THE CRETACEOUS SYSTEM 321

The shelly limestones are made up of fragments and perfect specimens of many kinds of shells, mixed with a comparatively small quantity of quartz sand. They may well be compared to the coarse shell sand found upon many parts of the modem South African shore, with the important point of difference that the shells found in them are of quite different kinds from those found on the present beach. Each bed of shelly lime- stone can be followed for a certain distance along the cliff, then it thins out, and another similar bed at a slightly higher or lower level takes its place.

The sandy calcareous clays are blue in colour on fresh unweathered surfaces, and they are so tough that the fossils contained in them are only with diflSculty ex- tracted from the rock, but the outer inch or two of the exposed outcrops are altered to a soft brown clay, from which the fossils are easily obtained by scraping away the decomposed rock with a knife.

The following section measured on the low cliff near the Umzamba mouth illustrates the nature of the succession in these rocks : —

Ft. In.

13 Shelly limestone 0 10

12 Tough sandy clay weathering brown - 1 4

11 Shelly Hmestone 0 6

10 Tough sandy clay 10

9 Shelly limestone 0 4

8 Tough sandy clay 3 6

7 Shelly Umestone 0 10

6 Tough sandy clay 3 0

5 Black impure limestone with many shells 0 6

4 Black shale 10

3 Oyster bed 0 2

2 Fine gravelly conglomerate - - - 0 3 1 (At base) Conglomerate with pebbles im- bedded in broken shells ; many fossils ?

13 3 21 ^"^"

322 GEOLCX^Y 01^ CAPE COLONY

The coarse bed at the base of the section is exposed on the shore at low water on both sides of the mouth of the Umzamba, but the extent of the rock laid bare at low tide varies, much of it being at times buried under the sand thrown upon the beach by the waves. A strong spring tide will uncover a wide area of rock that is usually concealed. This bed contains many interest- ing fossiia Reptiles are represented by Chelonian bones of large size ; the characteristic bony plates of the shell or shield and the shoulder girdle are easily recognised ; another reptile is represented by large jawbones with pointed teeth. Sharks' teeth are rather abundant, and complete the list of vertebrate fossils. The remains of marine invertebrates are plentifully preserved in this bed, the Cephalopods are represented by at least five species of Ammonites, a Nautiltos and a Baculites ; Gas- teropods by Fasciolaria, Avellana, Ghemnitzia and a large thick-shelled species of one of the StrombidsB ; Lamelli- branchs by three species of Pecten, Pectunctdus africanus, Protocardium hillanum, Trigonia elegans, Area natalensis, Gardimn denticulatum and Inoceram/iis, In this lowest bed there are many logs of wood, blackened and partly silicified and often bored into by Teredo, whose shells are still at the end of the holes made by their former inhabitants. Many of these fossils are much water- worn, and their surfaces are in consequence abraded. The more delicate shells are rarely or never found in a perfect condition, and a considerable part of the rock is made up of fragments of various kinds of shells. These facts, together with the presence of pebbles of grits, sandstones and dark -coloured slates, undoubtedly point

THE CRETACEOUS SYSTEM 323

to the bed •having been formed in shallow water, at the bottom of which the pebbles and shells were rolled about until they were covered up by the overlying deposit. The absence of the thin-shelled easily broken fossils, such as Hemiaster and Gassidulus, two echinoderms that are abundant in the overlying fine-grained beds, leads to the same conclusion.

The shelly limestones also contain the stronger shells in a perfect state; some of the weak shells, such as Inoceramus, that break up into small fragments of pe- culiar shape, can be recognised in these beds, but they are only found complete or nearly so in the fine-grained beds. The shelly limestones seem to have been formed in shallow water, for most of the shells were rolled about, broken, and had the projecting points rubbed off their outer surfaces before they came to rest and were buried under the accumulating sediments.

The fine-grained sandy calcareous clays contain strong and delicate shells in an excellent state of preservation. These beds were laid down in quieter water than the shelly limestones, and in consequence the most delicate shells were buried under the sand and mud without being broken. Over thirty species of Foraminifera and Ostra- cods have been found by Mr. Chapman in some small lumps of the rock that were sent to him for examination.

The lowest bed in the section given on a previous

page is the most persistent of the whole series. The

rest of the rocks are separated into many beds by the

thin lenticular shelly limestones in such a way that

two sections measured about a hundred yards apart

would not show precisely the same arrangement of beds.

21*

W-i *i'',l.#jfti®i»i^^

=.—-'S.-f.--^-&-'-S-'S.

S-i**-*-

THE CRETACEOUS SYSTEM 325

This group of rocks was formed near the shore of a sea teeming with life ; the shelly limestones were deposited where strong currents prevailed, for a certain period, over a comparatively small area, which were replaced by quiet water that allowed the fine-grained sandy mud to accumulate. The whole thickness of rock at present exposed is but some thirty feet, and it exhibits this alternation of fine and coarse sediments throughout. The same species of moUusca appear to be distributed through the whole group, but their presence in any one layer depends upon whether they were strong enough to resist the destructive action of the sea during the formation of that bed, for the coarse sediments contain the strong-shelled species only in recognisable condition, while the fine-grained beds contain both the thick and thin shelled species.

The following is a list of the chief species of inverte- brate fossils from the Umzamba beds : —

Cephalopoda —

Anisoceras mgatum, Forbes.

Baculites sulcatusy Baily.

Lytoceras (Gavdryceras) kayeiy Forbes, sp.

{Psevdophyllites) indruy Forbes, sp. Ptizosia {Hauericeras) gardenia Baily, sp.

,, (Hauericeras) revnbday Forbes, sp. Schl(Knbachia souton% Baily, sp. stangeri, Baily, sp. (Prionocyclus) uvibulazif Baily, sp. Gasteropods —

Avellana ampla, Stol. Gerithium detectum, Stol. ' ,, kaffrariunif Griesb.

Euchrysalis gigantea^ Stol. Fasciolaria asdmilis^ StoL „ rigida, Baily, sp.

I

326 GEOLOGY OF CAPE COLONY

Gasteropoda — continued —

Natica muUistriaJta, Baily.

Pollia pondtcherrienstSy Forbes, sp.

Pv/gnellus uncatus^ Forbes, sp.

Scala turbinata, Forbes, sp.

Solarium puMielluTrif Baily. „ wiebdi, Griesb.

Tritontdea trichinipolitensis, Forbes, sp.

TurhoniUa ? undomy Forbes^ sp.

Turritella muUidricUaj Reuss.

Dentalium^ sp. Lamellibranchs —

Area capensis, Griesb. ,, umzambaniensisj Baily.

Asto/rte^ sp.

Cardium denticntatum^ Baily.

Gorbula, sp.

Cytherea arcotensis, Forbes, sp.

GucuUoBa ncUcdenMSy Baily, sp.

Inoceramus expanstis, Baily.

NeUhia quinquecoatcvtay J. Sow., sp.

Nucula, sp.

Ostreay sp.

Pecten amapondermsy Griesb.

Pectunculua africanus, Griesb.

Protocardium hiUanunij J. Sow., sp.

Trigonia elegans, Baily. ,, shepdoneij Griesb.

Teredo, sp. Echinoderms —

Hemiaster forbesiy Baily.

Holaster indictbs, Forbes.

CassiduluSf sp.

Mr. F. L. Kitchin, who has in hand the examination of the invertebrate fossils collected from these beds by the Cape Geological Survey, has kindly furnished me with the following note upon the relationship of the fauna to that of certain be<Js in India 8^nd other gountries,

THE CRETACEOUS SYSTEM 327

"The palsBontological relations of this limited series of Cretaceous strata are comparatively easy of solution. When it is realised that a very restricted time-range is represented, and that there is no evidence of a succession of contrasted faunas, as was formerly believed to be the case, it becomes clear that we are only dealing with a true representative of the Arialoor (Upper Senonian) stage, so well known from its development in the Tri- chinopoli and Pondicherri districts of Southern India. Belationship to the Cretaceous rocks of Southern India was first indicated by Baily, more clearly emphasised by Griesbach, and more recently again by Kossmat, whose writings have thrown such light on the dispersion of the cephalopods of this age and the significance of the evidence yielded by these rocks in South Africa and Southern India. Amongst the more important species connecting this Cretaceous fauna of Pondoland with the Indian Arialoor stage are Puzosia gardeni, Puzosia rembda, Lytoceras kayei, Lytoceras indra, Anisoceras rugatum, Pug- nellus tmcatuSf Pollia pandicherriensiSf etc. Other deposits of Arialoor character in the Pacific region with which these beds in Pondoland show strong palseontological relations, are developed in Japan, Vancouver Island (and California), and Quinquina Island (Chili). The intermingling of essentially Pacific types with other forms having stronger European aflBnities (e.g., species of SchUBribachia) led Eossmat to regard these Cretaceous beds of Pondoland as of special importance in indi- cating the line of dispersal between the North Atlantic and the Indo-Pacific regions during Upper Senonian times,"

328 GEOLOGY OF CAPE COLONY

The Embotyi Group,

Near the mouth of the Embotyi Biver, about seven- teen miles north-east of St. John's, there is a group of conglomerates and green sandstones stretching about four miles south-west from Waterfall Bluflf. At the south-western end of the outcrops the beds lie nearly horizontally, and behind them are shales and sandstones probably belonging to the Ecca series, which have a rather high dip to the south-east. The junction of the two groups has not been seen, but it is very probably an unconformity. Farther to the north-east the Embotyi beds dip at moderate angles to the north-east, north and south, showing that they have been considerably dis- turbed since their deposition. At the north-eastern end of the exposures the Embotyi beds rest against a clifif of Table Mountain sandstone, and the slickensides still visible on part of the cliff, together with other evidence in the conglomerate itself, prove that the Embotyi beds have been faulted down against the older rock. The line of fault runs westwards from Waterfall Bluff, and about two miles from the latter separates the Table Mountain sandstone from the Ecca beds. Waterfall Bluflf is a vertical cliflf, some 300 ft. high, whose base is washed by the sea ; the streams from the country behind the cliflf fall over it, hence its name. The westward pro- longation of the line of cliflfs coincides with the foot of the escarpment on which the Egossa Forest stands.

The finer-grained portions of the beds, which appear on the shore near the. mouth of the Umgwegwane Eiver, are green shales and sandstones containing fragments of blackened wood, the only organic remains hitherto found

THE CKETACEOUS SYSTEM 329

in the group. Further search in these rocks is Ukely to be rewarded by the discovery of plant remains that can- not fail to be of great interest, and it is to be hoped that the search will be made before long.

The conglomerates towards the south-west end of the outcrops are pebbly rocks with water- worn fragments of dark grits and mudstones, certainly derived from the un- derlying Karroo beds. North-east of the Umgwegwane River the conglomerate becomes extremely coarse, and bedding planes are often difficult to find. Near the conical green hill on the Waterfall Bluff side of the river, and between that hill and the Bluff, immense blocks of coarse and fine-grained dolerites are found interbedded in a matrix of smaller boulders of similar material and of dark grits, mudstones and shales Uke those in the conglomerate farther south-west. Some of the dolerite blocks measure twenty feet in length. This conglomerate is the most tumultuous looking rock in the Colony ; magnificent exposures of it can be seen on the seaward face of the green hill, and near Waterfall Bluff. The irregular spaces between the boulders are sometimes filled with radiating bunches of brown calcite. The fine- grained portion of the rock is greenish and very similar to the sandstones near the Umgwegwane mouth.

The occurrence of the dolerite boulders in the Em- botyi rock is of great interest, as it proves that the dolerites had been injected into the Karroo formation before the deposition of the conglomerates, and were exposed at the surface during their accumulation. The similarity in situation of the Embotyi group to that of the Umzamba beds, which crop out at a distance of some

330 GEOLOGY OF CAPE COLONY

twenty-four miles to the north-east, and the fact that they are both faulted down against the Table Mountain sandstone, thus belonging to an earlier age than the chief disturbances that have affected this part of the Colony since the close of the Karroo period, make it pro- bable that the Embotyi group belong to the same series as the Umzamba beds. They may be regarded as the basal portion of the Pondoland Cretaceous rocks, and as bearing the same relation to the marine Umzamba beds as the Enon type of the Uitenhage series does to the Sunday's Eiver beds.

There is nothing to be said concerning the economic value of the Pondoland Cretaceous series. The Um- zamba beds make a very pretty bit of coast with its line of cliffs hollowed out into numbers of caves overhung by Strelitzia and other plants that are only found in the eastern parts of the Colony. The Embotyi beds occur in what is perhaps the most beautiful place in the Colony. The Egossa Forest forms a fine background, rising some 1,200 feet above the sea ; below it are low hills covered with tall grass and large bushes and trees, and between the hills wind the Embot5ri, Umgwegwane, and another river, widening out into broad lagoons just before they reach the sea. The writer was on that shore one after- noon when a thunderstorm passed over the forest, while the sun still lighted up the white breakers of the Indian Ocean and not a breath of wind disturbed the lagoons, from which were reflected the subtropical trees and bush growing on their banks. The scene was certainly the most beautiful one it has ever been his fortune to look upon,

CHAPTEE IX.

VOLCANIC PIPES YOUNGER THAN THE STORMBERG

VOLCANOES.

In many parts of the Colony there are remarkable pipes, channels through which materials were thrown from the lower region of the earth's crust to the exterior, and now filled with substances of dififerent kinds, sometimes clearly of volcanic nature, but often of such peculiar character that their volcanic origin is not obvious and can only be surmised from the manner in which the rocks occur.

The first of these pipes to be discovered was the Jager's Fontein Mine, in 1870, but those at Du Toit's Pan, Bult Fontein, Colesberg Kopje (Kimberley Mine), and De Beers were found soon afterwards.^ These discoveries were entirely due to the finding of dia- monds, which had been met with by chance near the Orange Eiver three years previously. It was, of course, some time after the diamond mines were opened that their nature was understood.^ The earliest search for diamonds was carried on in the alluvial deposits or ** Eiver diggings " on the Orange and Vaal Bivers ; the later or *'Dry diggings" in the volcanic pipes, which

^For an interesting and fully illustrated account of the early discoveries and of the whole history of the diamond mines and their working see The Diamond Mines of South Africa by Mr. Gardner F. WiUiams, 1902.

'Oohen (72), pp. 857-62. This paper, or letter, contains the first suggestion of the volcanic nature of the pipes.

m

332 GEOLOGY OF CAPE COLONY

have been the source of so great an industry in South Africa, followed upon the discoveries mentioned above.

Several other pipes are known in West Griqualand, but not much detailed information about them is available. South of the Orange Eiver two vents near Hanover are marked on Mr. Dunn's map (3rd edition, 1887), and four near Fraserburg ; others exist near Carnarvon,^ but no accounts of most of these have been published. Lately nearly thirty vents have been mapped in the Sutherland Division. One other, the neck on the farm Spiegel River in Riversdale, is known ; in some respects this one is of very great interest, as it affords more evi- dence of the later origin of the whole class of vents than is obtainable farther north, and it is at one end of the group in a petrological sense as it is filled with an igneous rock resembling a well-known but scarce variety of dyke-rock in foreign countries and in East Central Africa. The Saltpetre Kop (Sutherland) vents stand at the other end of the petrological series in being almost entirely filled with fragments of sedi- mentary rocks.

There are many intermediate conditions between the two extreme types to be found amongst the compara- tively few vents that have hitherto been examined from a geological point of view, and when a fuller series is known every gradation will doubtless be recognised.

We shall commence the description of the pipes with an .account of those filled with rocks of the purely igneous type and proceed in the order of their departure from this type without regard to their geographical

^ Dunn, Geological Sketch Map of Cape Colony (73) ; (74) pp. 64-60.

VOLCANIC PIPES 333

positions. At the end of the description the reasons for considering the whole group as belonging to one period of volcanic activity will be given together with other points of general interest.

On the farm Spiegel Eiver in the Riversdale Division there is a most remarkable mass of melilite-basalt ex- posed at the top of a hill composed of conglomerates and sandy beds belonging to the Uitenhage series. The outcrop is about 300 feet in diameter from east to west and rather less in the other direction, but the exaxjt junction with the surrounding rock is difficult to find on account of the debris covering the slopes on which it should be exposed. The grey-black igneous rock is in places roughly columnar, but the columns are very feebly developed ; they slant towards the east. The only feasible explanation of the occurrence is that the melilite-basalt fills a volcanic neck. The want of good exposures and the crumbly nature of the con- glomerates prevent the observation of the dip of these beds at the contact. The beds are seen at several places within 200-300 yards of the vent but they present no points of difference from their nature at a greater distance from the spot. No other neck or intrusion has yet been found in the Uitenhage beds, and till lately no other occurrence of melihte-basalt had been observed in South Africa.^ The rock is composed of

^The peculiar rock described by Cohen (Tschermak's Min. u Petr, Mitth.f Bd. xiv., Heft 2) as a melilite-augite rock is quite different from any of the rocks mentioned in this chapter. It is composed of melilite and augite, without any olivine, perofskite or iron ores, and contains native copper. It came from the Zoutpansberg District, Transvaal. It has been regarded as a rock altered by use in the hearth of a furnace.

334 GEOLOGY OF CAPE COLONY

a ground mass of glass in which there are minute crystals of perofskite and magnetite, irregular grains of augite, immense numbers of melilite crystals showing the usual characters of that mineral, and fairly large well-formed crystals of olivine. It is in a remarkably fresh state for so basic a rock.^

The vents and semicircular dyke on the Commonage near Sutherland village are mostly filled with rocks of a thoroughly igneous character ; tuflfs, or rocks made up of small fragments of various kinds, including lava and minerals derived from it, are found in three or four of the seven necks, but with them are the igneous rocks ; in the case of three of the pipes the igneous rock is melilite-basalt with more glass and biotite and less augite and melilite than the Spiegel River rock con- tains; serpentine, calcite and zeolites, the products of alteration of the other constituents are abundant.^ The curved dyke is composed of a similar rock. The tuflfs in the vents in which the melilite-basalt occurs are light blue sandy rocks containing biotite, ilmenite, serpentine and perofskite in addition to the debris derived from sedimentary beds. The other vents on the Commonage are filled with a dark-coloured amygdaloidal basic glass, and in some cases blocks of sandstone and shale with smaller fragments of the same rocks are imbedded in a matrix evidently composed mainly of altered glass of the nature of the glassy lava in these vents. Serpentine pseudomorphs after olivine are the only large crystalline

^ An analysis by Mr. Lems, one of the Cape Government analysts, is given in Geol. Comm. (08).

' Description of these rocks will be found in Qeol, Comtn, (03).

VOLCANIC PIPES 335

constituents of this lava, and they appear to have come from another rock, a fine-grained highly altered material which still adheres to the serpentine psendomorphs ; augite and magnetite are the other constituents that have been recognised, and they are in very minute grains and crystals. The steam holes in this lava are filled with calcite, analcite, natrolite and other zeolites, but silica, which in the form of quartz or chalcedony is frequently found in the steam holes of the ancient Zeekoe Baard lavas and those of the Stormberg series, has not been found in the Sutherland Commonage amygdaloids.

At Matjes Fonteiuy a farm nine miles south-east of Sutherland, there is a pipe partially filled with melilite- basalt of rather peculiar characters ^ and partly with a gritty breccia containing large fragments of granite, dolerite of the Karroo type, quartzite and other sedi- mentary rocks, mica, ilmenite and hornblende. The three latter constituents are identical in nature with the same minerals in the Silver Dam pipe to be mentioned presently. The melilite-basalt of this outcrop is com- posed of olivine, melilite, perofskite, biotite, magnetite, calcite and serpentinous fibres, probably derived from a glassy ground mass. Excepting the presence of calcite and the serpentine fibres the rock is remarkably fresh, and differs in several respects from the other njelilite- basalts. It shows a marked flow-structure.

In the remaining pipes there is no large, body of igneous rock corresponding to the melilite-basalts and

^ See Geol. Comm. (08) and Rogers and Du Toit (Oi).

•• -jj. ••'••isnis'a- -

N isi" 'f -

••II i.gj^.« -==

;4^ -5|..5.;*.;s::s:;*::ffl:^_._

•^:*^

i:#^'

K.'fi'.*a*3l:'«!'^&^:^^^ ,p^(rij „/ the

i^^^r liave a retem- ba^:^^e tilue-Krotuid 0^

^i^^-S^^^p»^:«&'-h occnr in these

VOLCANIC PIPES 337

heavy basic rocks are also the most conspicuous frag- ments in the breccia, and there is no doubt that they were derived from the same source that the boulders came from. The less conspicuous constituents of the breccia, only determinable under the microscope, are perofskite, serpentine pseudomorphs after olivine, grains Vi quartz anrargiUaceous matier derived from sedi- mentary rocks and calcite. The harder variety of breccia contains less serpentine and more sand and clay than the softer, but all the minerals mentioned above occur in both kinds.

Saltpetre Kop is a very prominent hill in the Suther- land Division rising about 1,000 feet above the general level of the high plateau on which it stands. It is composed of breccia and tuff, filling a vent about 1,000 yards long by 600 wide ; the vent traverses the Beaufort beds which are turned upwards for a considerable dis- tance on all sides ; the dip of the Beaufort beds is extremely slight in the surrounding district, but at points about a mile and a quarter from the neck the strata have a distinct dip away from it and the inclina- tion increases as the neck is approached, so that near the breccia the beds are nearly vertical.^ Bound aboui this large neck are nineteen others of smaller size and forty-six dykes, mostly filled with fine tuffs or breccias. In the case of one dyke the rock has been found to be largely composed of one of the less basic plagioclase felspars, and is evidently an igneous rock of somewhat peculiar character, but it has been greatly altered by the

* A fuller description and plans of the Saltpetre Kop area will be found in Geol. Comm, (03) and Rogers and Du Toit (04).

22

338 GEOLOGY OF CAPE COLONY

substitution of calcite, hydrated ferric oxides, and silica for some of its original components. The breccias and tuflfs vary greatly, but they all consist mainly of frag- ments of sedimentary rocks set in a matrix of similar substances finely comminuted ; but in addition to these constituents there are pieces of granite, gneiss, mica schist and Karroo dolerite, and also mica, hornblende and ilmenite, identical in character with the similar minerals in the Silver Dam breccia. Parts of the brec- cias and tuffs are strongly impregnated with carbonates of lime and magnesia, barium sulphate, hydrated oxides of iron and silica. This has happened chiefly in the smaller pipes and in the peripheral portion of the large vent ; a similar process has caused the hardening of the shales and sandstones at their contact with the vents and dykes. The carbonates, sulphates, oxides of iron and silica were probably carried to their present position by water ascending the channels of eruption after the period of violent activity had closed ; their deposition may be regarded as analogous to the effects of the ** solfataric " stage of recent volcanic areas.

The smaller necks in the Saltpetre Kop area do not materially affect the regularity of the qu8l.-qua-versal dip about the central vent.

In no other vent of the kind we are dealing with in this chapter is the outward dip or up-turning of the surrounding strata so strongly marked as in the case of the central neck of the Saltpetre Kop group. Wherever the strata in immediate contact with one of the pipes are exposed, and have been examined with attention, they have been found to dip away from the contact, as

VOLCANIC PIPES 339

though the ascent of the materials filling the pipes had bent the edges of the strata upwards. This has been noted at some of the Sutherland Commonage vents, at Balmoral (Batel Fontein), at Matjes Fontein, Schiet Fontein and at Kimberley. This feature seems to be peculiar to these vents, for where notice has been taken of the dip of the strata near the pipes of volcanoes of the more usual types the strata have been found to be inclined towards the pipe as though dragged downwards by the settling in of the contents after the activity of the volcanoes ceased.^

On the farms De Vrede, PortugaPs Kiver, and Blaauw Blommetjes Keep, in the Sutherland Division there are breccia-filled pipes and dykes. The Blaauw Blommetjes Keep pipe gives off a sheet-like extension of the breccia, which distinctly traverses a thick sheet of dolerite, and thereby proves that the production of the vent was posterior to the consolidation of the dolerite, a strong confirmation of the evidence afforded by the fragments of coarsely crystalline dolerite found in the breccias of many of the necks of this class.

At Balmoral (Ratel Fontein), in the Fraserburg Divi- sion, there is a circular depression in the ground about 300 feet wide and ten to twenty feet deep, surrounded by the truncated edges of the Beaufort beds dipping away from the depression. The depression is caused by the weathering away of a soft breccia which fills a pipe.

^ It is naturally only in long extinct volcanoes that observations on the dip of the sedimentary strata, below the pile of the volcanic debris forming the cone or mountain, can be made. Several sections through such strata are given in Sir A. Geikie's Ancient Volcanoes of Great Britcnn,

22*

340 GEOLOGY OF CAPE COLONY

The breccia is a blue muddy rock containing fragments of sandstone and shale, dolerite, biotite, garnet and ilmenite. This pipe is remarkably well exposed, and the nature of the contact and the up-turning of the edges of the sedimentary rocks through which the pipe passes can be more satisfactorily seen than at any other locality yet described.

At Schiet Fontein and other farms near Carnarvon and near Hanover similar pipes are known ; they have been briefly described by Mr. Dunn ^ but no details have yet been published concerning them.

To the north of the Orange Eiver, in the Cape Colony, the Orange Eiver Colony, and the Transvaal, there are many of these volcanic pipes. Several of them, includ- ing those at Kimberley, are surrounded at the surface by rocks belonging to the lower stages of Karroo forma- tion, but farther to the north and west, where these strata have been removed by denudation, the pipes crop out through the Pre-Cape rocks. At Kimberley the mines are being worked far below the base of the Karroo formation, as the accompanying sections ^ show. The quartzites, amygdaloidal rocks (** melaphyres ") and quartz-porphyries passed through by the rock shafts, from which access to the mine (the pipe filled with blue-ground) is gained by horizontal tunnels, belong to the Pre-Cape formations ; but the dolerite, or diabase as it is usually called by French and German writers, is part of the great intrusions of late Karroo age de- scribed in chapter vii.

»Dunn(74), pp. 54-60.

^ I am indebted to Mr. Gardner F. Williams for these sections.

r

H ■

'J^

T' -*•.■-*•-#•

342 GEOLOGY OF CAPE COLONY

The "blue-ground" or kimberlite (Carvill Lewis) which fills these pipes is a serpentinous breccia containing many kinds of minerals. The chief varieties are olivine or serpentine pseudomorphs after that mineral, biotite, chrome-diopside, enstatite, smaragdite, garnet, perofskite, magnetite, ilmenite, chromite, picotite, apatite, epidote, orthite, tremolite, tourmaline, rutile, and diamond. Calcite, various zeolites, chalcedony, and talc are also present, but they must be looked upon as having been introduced after the volcanic activity ceased, or as alter- ation products of the other constituents. It is impossible to be certain which constituents are part of the blue- ground as distinguished from the fragments contained in it, but there is reason to regard the olivine, magne- tite, ilmenite, and perofskite as belonging to a formerly molten magma which carried up with it, during the explosion that established the pipes, part of the olivine, the pyroxenes, garnet, smaragdite, diamond and several other minerals that were derived from deep-seated rocks other than the then molten lava. This view was strongly supported by Professor Bonney,^ who gives con- vincing evidence in favour of it as regards diamond and other constituents of an eclogite from the Newlands Mine. Descriptions of the Newlands Mine show that the blue-ground occurs in an irregularly shaped pipe and as dykes and sheet-like extensions in the surround- ing rocks.^ At the De Beers Mine a hard variety of blue-ground, called snake-rock, which occurs in the form of a dyke in the softer blue, extends as a dyke through

1 Geol. Mag. (99), pp 309-21. 2Graichen(0a).

VOLCANIC PIPES 343

the country rock outside the pipe itself, just as at Salt- petre Kop, Blaauw Blommetjes Keep, and De Vrede the breccias form dykes in the Beaufort beds. The shape of the pipes appears to vary at different depths, but on this and many other points of great interest concerning the occurrence of the breccias no complete or detailed information is yet available. The composi- tion of the breccias is by no means constant in the different mines or in one and the same pipe. A striking instance of this fact is the abundance of enstatite in the rocks from De Beers Mine described by the earlier writers, while in many specimens from deeper levels it is certainly a rare constituent.

In some of the pipes in the Kimberley area large masses of sedimentary rocks have been found embedded in the blue-ground ; some of these contain fossils. The sandstone fragments with Atherstonia, a fish, have prob- ably come from the Beaufort beds, and some reptihan remains which have been found in the Premier Mine may have had a similar origin. These fragments prob- ably dropped into the pipes from the wall at a higher level than that at which they were found. The large logs of charred wood sometimes met with in the blue- ground may have fallen into the vents from the surface after the explosions had taken place. Had the logs been fossilised wood derived from the Karroo strata we should expect to find them in the same strata as the fossil wood in those beds, viz., in the form of silicified wood, in which silica replaces the woody tissue and fills the cells.

There have been many views held as to the real

344 GEOLOGY OF CAPE COLONY

nature of the blue-ground,^ but the best supported is certainly that in which it is regarded as a breccia derived from an igneous rock of ultra-basic composition. The late Professor Carvill Lewis ^ considered that the " abun- dance of calcite as a decomposition product, the high magnesia and low alkali, the presence of biotite, and more especially of perofskite," indicate the former pre- sence of nepheline or melilite, and that the rock may have been a melilite-basalt. The discovery of melilite- basalt in similar pipes in Sutherland is certainly striking in view of this opinion. The same author considered the blue-ground to be a true igneous lava, and not a mud or ash, but this view is difficult to reconcile with many of the facts, as the editor of his papers points out. It seems more justifiable to regard the contents of the Kimberley pipes as breccias derived from the explosive disintegration of a body of lava of ultra-basic composi- tion ; another effect of the explosion was to break up masses of rock (which may be called eclogite), composed of pyroxenes, olivine, ilmenite, biotite and garnets, to mention the more abundant minerals only, and to throw the minerals thus obtained up the channels opened by the explosion, mingled with the lava in a solid or plastic state. Whether the eclogites were originally altogether distinct from the molten lava, or whether they were an

^ A general summary of these views will be found in chap. xvi. of Mr. Gardner Williams' book, cited on a previous page.

^ The Genesis and Matrix of the Diamond^ edited after the death of the author by Professor T. G. Bonney, London, 1897. This contains the best account of the rocks and minerals of the Kimberley pipes in the English language. For other references see under Bonney, De Launay, Story-Maskelyne and Flight, Lacrpix and Cohen in the appendix.

VOLCANIC Mf ES 346

early product of the same magma is another matter, which is difl&cult to decide with the help of the evi- dence at present available. The abundance, however, of perofskite in the blue-ground, and its absence from the eclogites hitherto described from the pipes, seems to indicate separate origins for the two rocks.

Another feature of importance is the occurrence of well-rounded boulders of several of the rocks enclosed by the blue-ground. Professor Bonney described an eclogite boulder from the Newlands Mine, and came to the con- clusion that it had been picked up in that form by the blue-ground when the explosions took place. At Salt- petre Kop and other vents in Sutherland, similarly shaped boulders of hard quartzite, eclogite and granite occur. At Balmoral a garnet rock and dolerite of the Karroo type are found in this form. The dolerite boulders could hardly have been obtained in that shape from a con- glomerate, as there are no known conglomerates, con- taining dolerite boulders, of earlier age than the pipes. It is possible that the quartzite boulders of the Saltpetre Kop agglomerates were derived from the Dwyka con- glomerate, but a similar explanation cannot be held to account for the eclogites. These rocks have not been observed in the Dwyka conglomerate, and their extra- ordinary abundance in the Silver Dam breccia, as well as in some of the northern pipes, is inconsistent with the paucity of boulders of granite and certain other rocks in the breccias, for those rocks are very frequently seen in the Dwyka conglomerate. In addition to this, the eclogites, or minerals deiived from them, are quite as characteristic of the breccias from the pipes in

346 GEOLOGY OP CAPE COLONY

regions where the Dwyka conglomerate is but thinly developed or entirely absent, as in districts where that conglomerate almost certainly underlies the sur- face.

From the foregoing descriptions of the breccias and other rocks filling the pipes and fissures we see that, though they dijfifer widely, yet there are usually connect- ing links between them to be noticed. Had the Spiegel Eiver melilite-basalt been the only example of this type of rock known in the Colony it would have been ex- tremely rash to postulate any connection between it and the contents of the long known Kimberley pipes ; but the association of the melilite-basalts of Sutherland Conmionage and Matjes Fontein in the same pipes with breccias containing some of the characteristic minerals of kimberlite, and their occurrence close to the Silver Dam vent, which is filled with an agglomerate still more like typical kimberlite, render the supposition much less improbable. In the remarkable agglomerates of the Saltpetre Kop group of pipes we find that though the bulk of the rocks are composed of the debris of sedi- mentary beds, yet there are also numerous fragments of the biotite, ilmenite and hornblende characteristic of the Silver Dam breccia. The occurrence of kimberlite in the form of dykes and sheets, as well as in the pipes, is analogous to the agglomerate dykes and sheet of Salt- petre Kop and Blaauw Blommetjes Keep, though such phenomena are distinctly unusual. It is interesting to notice that similar dykes of a rock apparently indis- tinguishable from kimberlite have been found at two places in North America, at Syracuse (New York) and

VOLCANIC PIPES S47

in Kentucky.^ In the Cape Colony only one other grit or detrital dyke has been found. It is a remarkably regular outcrop of a gritty rock composed of grains of quartz, felspar, garnet, epidote and other minerals, and it ex- tends for a long distance through the Witteberg beds near Elands Vley, west of the Tanqua Karroo. It has a width of about eight feet and is said by the local farmers to be clearly traceable for thirty miles across the country. Its age is unknown, but as it is so unlike any other geological feature in the country it may be mentioned here in connection with the phenomena which bear the nearest analogy to it. The remarkable characters of the contents of certain pipes, such as those of Silver Dam and Balmoral, are closely similar to those of Kimberley, and the fact is sufiBcient to support the view that these vents were established by similar means, and at about the same time.

The age of each vent can only be determined by observing the strata which it traverses and by finding rocks of known age in the breccias. It is obvious that a vent is younger than the beds passed through, and younger than the rocks contained in it in the form of fragments or boulders. The usual evidence of the date of the extrusion of volcanic rocks is entirely wanting here. The -only satisfactory answer to such a question is the interbedding of tuffs or lavas with contemporaneously formed sediments. In the case of the Stormberg volcanic beds, for instance, they have

^ Descriptions of these rocks and their occurrence and references to the original papers will he found in Professor Bonney*s edition of Garvill Lewis's papers.

348 GfiOLOGY Of CAtE COLONY

been found intercalated between the ordinary sediments of the upper division of the Stormberg series, and their age is thereby satisfactorily settled, though we do not know how long the volcanic activity prevailed. In the cases of the Kimberley, Sutherland, Fraserburg and other similar pipes, excluding the Spiegel Biver neck, we know that they were formed after the intrusion of the Karroo dolerites, for they either pass through sheets of dolerite or contain fragments of that rock evidently torn from sheets or dykes. The dolerite intrusions as we saw in the last chapter, probably belong to the Stormberg period, therefore the pipes were probably produced later than that period. This is as much as can definitely be stated with regard to the age of those vents. If, now, the general resemblance of the Spiegel Biver melilite-basalt to the somewhat similar rocks of Sutherland be considered as evidence of their close connection in origin, or, in other words, of their belong- ing to one and the same phase of volcanic activity in the Colony, as in my opinion it may be, then the earlier limit of the age of these pipes is advanced from Post- Stormberg to Uitenhage or Post-Uitenhage times. It is worth while mentioning the fact that the other known African rocks containing melilite and having a distinct, though perhaps not very close, resemblance to the Co- lonial melilite-basalts occur in East Africa at Doenyo Ngai, Makinga Hill and Mount Elgon.^ At the present

^ Short descriptions of these rocks are given in Zirkel (94) and Eosenbusch (96) p. 1,276. G. T. Prior (03) describes the Mount Eigon rocks. The others are described by Miigge (86) and Lenk, but I have not had access to these two papers.

VOLCANIC PIPES 349

time no great importance can be attached to the resem- blance between rocks so far removed from one another, especially as the examples in the Colony show no indications of the surface features consequent on their eruption, while the East African rocks are of quite recent date.

We may sum up this account of these peculiar vents which are distributed widely over South Africa by saying that at some period after the close of Stormberg times (probably after the commencement of the Uitenhage period) great explosions took place which drilled holes of various sizes through great thicknesses of rock, and that although some of these holes were filled with lavas of basic composition, the majority are occupied by agglomerates, breccias or tuflfs. These fragmental rocks are composed of material derived from the molten magma which was intimately connected with the imme- diate cause of the explosions, mingled with other matter torn from deep-seated rocks or from the strata through which the pipes were opened.

The occurrence of diamond as a constituent of some of the breccias has been the cause of a far wider interest in the pipes than would otherwise have been the case. For many years the diamond was thought to have been derived from the crystallisation of the carbon originally contained in the carbonaceous shales surrounding the pipes, but the presence of the mineral in the blue- ground at levels far below the shales, and its occurrence near Pretoria in kimberlite filhng a pipe in the Pretoria series, which lies below any known carbonaceous rocks,^

1 Molengraafi (98) p. 128.

350 GEOLOGY OF CAPE COLONY

finally disposed of that theory. The presence of dia- mond in the form of good crystals in the garnetiferous eclogites^ affords strong support for the view that it crys- tallised out from solution in an ultra-basic rock-magma, which sometimes gave rise to eclogites. Whether any other variety of rock magma enabled the mineral to form remains to be proved. Hitherto eclogite containing dia- mond has only been found in breccias of the kimberlite type, no outcrops of eclogites or other ultra-basic rocks containing the mineral have yet been found.

^ Bonney (99) pp. 309-321.

OHAPTEK X.

RECENT OR SUPERFICIAL DEPOSITS.

In many parts of the Colony there are accumulations of sand, gravel, alluvium, limestones, quartzites, and fer- ruginous rocks that belong to a comparatively recent order of things, and in some cases are to-day in process of formation. There is invariably a marked uncon- formity between those rocks and the strata upon which they rest, although it is not always easy to find a suit- able exposure of the junction.

For the purpose of description the superficial deposits may be divided up into the following groups, but they were not so distinct in origin, and, as we shall point out later, some groups grade into others : —

1. Older gravels, alluvial deposits and quartzites.

2. Newer gravels and alluvial deposits.

3. Laterites.

4. Blown sands.

5. Limestones of the coast belt.

6. Limestones of the interior.

7. Baised beaches.

8. Vley and pan deposits.

1. Throughout the folded region and to the west of its western portion there are many signs that the country was to a certain extent reduced to a plain at

851

352 GEOLOGY OF CAPE COLONY

a period when the rivers flowed at levels of some 600 — 1,000 feet above their present beds. In the Euggens of Galedon, Swellendam, Bredasdorp, and Mossel Bay, a great tract of hilly country carved out of rocks chiefly belonging to the Gape and Uitenhage formations, the hill-tops reach a more or less common level from 800 to 1,200 feet above the sea. The summits are com- posed of gravels, alluvium, and quartzites of a peculiar nature, and are frequently table-shaped. If one looks over the Ruggens from any prominent point in Swell- endam or Riversdale the conviction that these isolated patches were once continuous, and that they formed a gently undulating surface connected with the terrace that is at places a very conspicuous feature along the lower slopes of the Langebergen, is immediately borne in upon one.

The terrace on the mountain-side north of Zuurbraak is separated by the deep valley of the Buflfeljagt's Eiver from the gravel-capped plateau south of that place ; the gravels are coarse and contain many pebbles and boulders of Table Mountain sandstone that must have come from the Langebergen, although the ground on which they lie is now quite cut oflf from the mountains by the deep valley. There is no doubt that the terrace and the plateau were once continuous, and that the pebbles were brought from the Langebergen by the mountain streams that now feed the Buflfeljagt's Eiver. In this case the rocks underlying the plateau are mainly Bokke- veld slates, but on the west a tongue of the Swellendam basin Uitenhage beds enters into its composition with- out altering the character of the plateau, in spite of the

RECENT OR SUPERFICIAL DEPOSITS 353

fact that the Uitenhage beds are more easily eroded than the Bokkeveld.

To the east of the Gouritz Eiver the road from Herbertsdale to Hagel Kraal lies on a terrace stretch- ing far to the south .of the Langebergen, cut out of the Table Mountain, Bokkeveld and Uitenhage beds in different parts. This terrace is considerably cut up by eastern tributaries of the G-ouritz Biver, but it is not divided into a terrace and a plateau as is the case with the old Zuurbraak terrace ; the unity of the whole is still preserved. The Nouga Eiver has exposed admir- able sections showing the terrace gravels lying uncon- formably upon the Uitenhage beds, which there have a moderate northerly dip.

To the north of the Langebergen both the terraces and the plateaux are well represented. The former can be seen from the roads to Oudtshoorn from Mossel Bay where they leave the mountains at Saffraan Eiver (Eobinson Pass) and Doom Eiver (Montagu Pass). The best example of a plateau in this district is the Tafel Berg, between the Waterval and Bok Kraal Eivers south of Buffels Fontein, a wide table-shaped area that does not deserve the name of Berg. It is covered with gravels derived from the Langebergen, from which it is now separated by the Waterval Eiver.

In the Oudtshoorn-Uniondale-Willowmore area there

is a great development of high level gravels (see Plates

XIX. and XX.). The watershed between the 01ifant*8

and Baviaan's Kloof Eivers is on one of them. The

present rivers for the most part run in deeply eroded

valleys cut down through the plateau gravels. The

23

W&r^itMvWy^irface of the

Hf''«^.R^«M«)^8lriii-covered terraces 'fi'^^Sf ir^' ftWfS*^ as they are to

and Baviaan's

old

iSettt^kOimably upon inclined " ~S|loof, new Tover Water

'iMi'Sj^'^lml^i^^^^S^'^^ ^f>^n cat to a

iud its afflaents at

|3^ed some 200 feet

se of the terrace

Similar featuree

.^'^■kU. 8oc. 8.A. (M).

RECENT OR SUPERFICIAL DEPOSITS 355

occur along the northern flank of the range at least as far as Prince Albert, where there are also some fine table-shaped and gravel-capped hills lying considerably to the north of the mountains.

In the neighbourhood of Grahamstown the gravel and quartzite terraces south of Botha's Hill and the curious Sugar Loaf Hill nearer the town are parts of a slightly undulating plain that has been cut into by the Blaauw Krantz Eiver. The underlying rocks belong to the Witteberg and Dwyka series.

In the country north-east of the Gualana Eiver, where the coast is formed by the Karroo formation, there are extensive plateau-hke terraces bordering the coast, deeply cut into by the rivers flowing from the Storm- berg and Drakensberg. At a few spots on the plateau that lies about 2,000 feet above the sea there are remains of deposits analogous to the old alluvium and quartzites of the country to the south-west. Kentani Hill is a conspicuous example of these. At the present time, however, little is known of the extent of these rocks.

In the Western Karroo a fairly well developed terrace is visible along the foot of the Zwart Euggens, the dry mountain ridge of Witteberg beds that limits the Ceres and Tanqua Karroos. In the Tanqua Valley a corre- sponding terrace covered with gravel derived from the Klein Eoggeveld forms a conspicuous feature on the south side of the valley.

In all these cases the gravels are coarser near the

mountains than farther away from them. Pebbles and

boulders derived from the Table Mountain sandstone

are by far the most conspicuous constituents in the high

23*

356 GEOLOGY OF CAPE COLONY

level gravels of the southern coastal region and in those of the country between the Langebergen and Zwarte- bergen. The boulders are sometimes of great size, four or five feet in diameter, and they have their edges rounded off ; the smaller fragments are more rounded and are like the watervirorn pebbles to be found in the modern stream beds. These fragments are embedded in a matrix that varies very greatly ; in the bulk of the rock the matrix is a sandy material, but slightly hardened, from which the pebbles may be easily broken out ; in other cases the matrix is deeply coloured by hydrated iron oxide, and the rock is in consequence reddish brown. Such ferruginous gravels are well de- veloped near Genadendal in Caledon and at the foot of the Zwart Buggens in the western Karroo. Near the village of Napier there is a conspicuous kopje formed of a dark, highly ferruginous conglomerate, which probably belongs to the same group of gravels that are developed to the west of the village, at a consider- able height above the bed of the Elands Kloof Eiver. The ferruginous cement has in many cases hardened the gravel to such an extent that the rock breaks across pebbles and matrix alike when struck with a hammer. • There is a gradual passage laterally from these ferru- ginous gravels to the fine-grained ferruginous rocks that lie farther from the mountains, and which often con- tain a few angular or subangular pieces of white vein quartz derived from the slaty Bokkeveld or Witteberg beds underlying them. Magnificent examples of these hardened alluvial deposits are to be found in many parts of the Euggens, forming rough-looking caps on the

RECENT OR SUPERFICIAL DEtOSlTS 357

higher hills, such as Klaas Kaffir's Heuvel near the road from Swellendam to Bredasdorp.

The ferruginous rock is often directly underlain by white and yellowish clays, bleached by the slow removal of tlje iron they once contained which is now concen- trated in the overlying rock. In some cases the traces of cleavage and joint planes are to be seen in the bleached material, which must then be regarded as a product of weathering in situ of slates ; but most of the clays ap- pear to be alluvial deposits formed by the rivers when they were at a relatively higher level than at present.

The ferruginous material is closely related to the lat- erites that occur at lower levels in many parts of the south-west, and which will be described on a later page.

Another very widespread variety of the gravels is due to the deposition of silica in the matrix subsequently to the formation of the gravel. All stages between a rather incoherent conglomerate and an extremely hard rock from which it is practically impossible to detach the contained pebbles can readily be found in one and the same patch of rock. The deposition of silica is most ad- vanced on the upper surface of the mass, the lowest part of which is often a loose gravel. By the diminution in size of the pebbles and their gradual disappearance as the outcrops are followed away from the mountains the quartzitic gravels pass into the typical ** surface quartzite *' so widely distributed throughout the western and southern parts of the Colony (see PI. XX.).

As a rule the surface quartzites have certain pecu- liarities that enable one to recognise the smallest chip without difficulty ; their fracture is smoother, more con-

J^ i|ifffl:|

,^LONY

^Si^f^Js;!:.*^

RECENT OR SUPERFICIAL DEPOSITS 359

choidal, and less splintery than that of the quartzites of the older formations ; small quantities of argillaceous matter, yellow or grey in colour, are present in the siliceous matrix enclosing the grains of quartz sand that are often visible without the aid of a magnifying glass. The quartzites generally enclose many small irregularly shaped cavities, which are sometimes lined with minute crj'stals of quartz, or with the chalcedonic form of silica. The original quartz grains in the rocks are at places converted into bipyramidal crystals by the addition of new quartz in crystalline continuity with the quartz of the grain. By the mutual interlocking of the new quartz added to all the sand grains in the originally sandy portion of the rock, the loose sands have become intensely hard quartzites in which the original grains are no longer recognisable without the use of a microscope and thin sections of the rock, when the outlines of some of the grains can be seen within the new growth of quartz ; the quartz deposited round any one sand grain interlocks closely with that round the neighbouring grains. Good examples of these quartzites may be seen in any of the south-western divi- sions. They often appear above the soil as rounded polished surfaces, due to the weathering out of the rock along irregularly disposed vertical joints, which leave a massive lump of rock in their interstices. On the hill- top near the road from Swellendam to the bridge over the Buffeljagt's Eiver the quartzite has been quarried for building purposes ; the bridge piers are made of it. As a rule, however, the rock is too intractable and too variable within short distances to be worth quarrying

360 GEOLOGY OF CAPE COLONY

although it is certainly a very durable stone. Near Grahamstown the surface quart zites appear in the Sugar Loaf Hill and on the terrace to the north of it mentioned on a previous page. The hard quartzite is at most ten feet thick, but the underlying soft clayey material, into which the quartzite passes without any definite break, is at places as much as forty feet thick.

The top of Kentani Hill, the only conspicuous eleva- tion above the general surface of the plateau that stretches northwards from the Kentani escarpment, is formed by a hard quartzite, vitreous in parts, but usually with a rough pitted surface. The quartzite, which is only a few feet thick, passes downwards into variously coloured clays from thirty to forty feet thick.

A similar siliceous rock from a farm about nine miles south of Komgha village contains the silicified seeds of Chara, small spherical bodies with ribs passing spirally round them, and silicified shells of Limncsa. This is the only surface quartzite in the Colony known to contain recognisable fossils, but at present nothing is known of its extent.

On the Cape Flats there are several outcrops of sur- face quartzite, some of which contain plant remains that have not been determined. One well-known out- crop is near the main road to Stellenbosch about ten miles from Cape Town, and there are several others in its vicinity. The Cape Flats quartzites are usually whiter and more uniform in grain than the similar rocks in other parts of the Colony, v The white colour is due to the almost complete absence of clay and ferruginous colouring matter; the quartzite passes downwards into

RECENT Oft RUPEllPIClAL DEPOSITS 361

a sandstone and that again into loose sand, which is identical with the white sand that occurs under the sur- face soil over a great part of the Flats.

In the Malmesbury, Piquetberg, Clanwilliam and Van Ehyn's Dorp Divisions surface quartzites are met with in many places on the coast side of the Olifant's Eiver Mountains and the other ranges in connection with them. The quartzites are underlain by sandy clay or gravel into which they grade. By an increase in the amount of ferruginous colouring matter they become very similar to the laterites, and on the Van Ehyn's Dorp coast they pass into coarse conglomerates con- taining the shells of living species of marine forms, raised beaches which lie from 50 to 100 feet above the high-water mark. This summary of the distribution and features of the high level gravels and associated rocks shows that throughout the southern, western and south-eastern portions of the Colony there are gravels and alluvial deposits, altered to some extent by the deposition of silica and other cementing substances between the grains, lying high above the levels at which similar accumulations are being formed at the present day. The deep channels through which the rivers now flow, and the consequent cutting up of the former plains whose existence is evidenced by the numerous flat-topped hills capped by the deposits laid down before the deep valleys were eroded, show that the country as a whole is now at a relatively higher level than it was during the formation of the plains. The rocks underlying the remnants of the old plains, now exposed in the river valleys, are of various natures and in part intensely

362 GEOLOGY OF CAPE COLONY

folded. The more resistant of these, chiefly the Table Mountain series, still project above the general surface of the plains in the great anticlinal ridges that have so frequently come under our notice. The terraces cut into these ridges show that the great anticlines of quartzites and sandstone were being attacked, and had to a slight extent been reduced to the level of the plains, at the time when the country began to rise and give renewed downward eroding power to the rivers. Previously to this period of elevation the country as a whole must have stood for long ages at about the same level, unless, indeed, some slight down- ward movements broke the quietude. The rivers were thus enabled to erode their valleys laterally after they had reached their base levels, ie., when the slope of their valleys was such that they could carry away all the debris furnished them but were unable to deepen their channels.

Towards the close of the period of great lateral erosion large areas south of the Langebergen, west of the Ceder- berg group of ranges, and between the Langebergen and Zwartebergen, were reduced to gently undulating sur- faces, across which the rivers flowed with many bends in their courses, and they were bordered by low-lying land covered with gravel near the mountains and sand or loam farther away from them. Probably there were many damp and swampy patches, or even shallow lakes, such as w^ould be called vleys in this country, on the low land, and in these places the changes may have com- menced that resulted in the formation of the surface quartzites.

RECENT OR SUPERFICIAL DEPOSITS 363

From an examination of specimens sent to Europe by Dr. Passarge, Professor Kalkowsky ^ came to the con- clusion that a certain kind of rock, which seems very similar to our surface quartzites, was formed by the silicification of an aluminous sandy mud deposited in salt pans in the Kalahari. The addition of the silica is attributed by him to the action of salts and organic sub- stances in solution upon the silicious remains of diatoms and other plants, although the diatoms were not actually found in the rocks examined. This explanation may apply to the Gape surface quartzites, for it fits in with the sporadic occurrence of the rock.

2. The newer gravels and alluvial deposits.

At various levels between the high level deposits just described and the beds of the present rivers in the southern, eastern and western parts of the Colony there are more or less well-marked terraces covered with gravels and alluvium. Several such terraces can be seen along the Breede Eiver below Swellendam. It is often difficult to separate the higher of these from the high gravel plateaux, and hard ferruginous rocks and even quartzites may be found on them, but they may often be distinguished from the plateau gravels by the finding of pieces of the quartzitic or ferruginous gravels amongst their pebbles. The Breede Eiver terraces have gravels containing such pebbles and boulders derived from the older deposits, originally of a similar nature.

In some parts of the Swellendam, Eiversdale and Mossel Bay Divisions, the gravels met with far from

^ Kalkowsky (01), p. 55, etc.

364 GfiOLOGY OF CAtE COLONY

the mountainous ground often contain large pebbles derived from the conglomerates belonging to the Uiten- hage series. These pebbles were well rounded, and were probably in much the same condition as they are to-day, before they reached their present position. The same is the case in other districts, such as Oudtshoorn, where the Uitenhage conglomerates occur. The abundance of these derived pebbles in positions where an explanation of their presence would be very difficult on the supposi- tion that they were brought directly from the original source of the rocks of which they are made, is at places very striking.

Near the mouths of many of the rivers of the south and south-east coasts there are sandy deposits which extend to a considerable depth below the beds of the rivers. At the Bitou Eiver ^ the green sands containing many marine shells, including large numbers of Crypto- don glohosus, which is now comparatively rare in the adja- cent sea, were pierced to a depth of forty-seven feet below the river without their base being found. The shells hitherto found in these sands and in similar deposits in other places all belong to existing species. At East London the sandy mud in the estuary of the Buffalo Eiver has been found to be over 120 feet thick. The considerable depth below sea level to which these estuarine deposits extend may point to a subsidence of the coast, but it is perhaps more likely that the scour of the river and tide combined are sufficient to account for the excavation of the estuaries. This certainly

^ Schwarz, Oeol. Comm, (99), p. 61.

RECENT OR SUPERFICIAL DEPOSITS 365

seems to be the case with the short but deep estuary of the Kaaiman*s Eiver, near George, where there is a rapid fall of the bed below the old road drift.

The alluvium along the great rivers draining the Great Karroo is often extensive and of considerable depth. It occurs chiefly behind mountain ridges through which the rivers have cut their way more slowly than in the softer ground now occupied by the alluvial deposits. A very good example is found in the Olifant's Eiver (Oudtshoorn) ; this river rises south of Antonie's Berg in Willowmore, but it receives very important tributaries in the Traka, Meiring's Poort, Grobbelaar's and Kammanassie Eivers before it joins the Gamka in the middle of the Eoode Berg mass of Table Mountain sandstone. The junction of these two rivers makes a great Y-shaped gorge, with vertical walls some 600 feet high, in the heart of the mountains. Before entering the gorge the Olifant's Eiver runs for some eighty miles over flat country, and this tract is very rich in alluvium, especially the lower part of it below the town of Oudtshoorn. Underlying the allu- vium there are rocks belonging to the Uitenhage group, which are soft and easily eroded compared with the Table Mountain sandstone. The mountains have acted as a check to the downward cutting of the river, that has consequently widened its valley behind them and deposited the alluvium to which the Oudtshoorn Division owes its wealth. These accumulations are gathered from nearly all the rock systems in the Colony, from the Pre- Cape rocks of the Cango to the Uitenhage beds of their immediate vicinity. The Gamka has formed a similar

366 GEOLOGY OF CAPE COLONY

but smaller alluvial tract between Sand Berg (or Paarde Berg) and the Eoode Berg gorge, and others occur lower down its course.

Another tributary of the Gamka, the Buflfel's Eiver, has cut a wide alluvial plain behind the Klein Zwartberg, which it enters at Leeuw Kloof Poort.

Great tracts of alluvium are found along the rivers which flow northward from the main watershed to the Orange Eiver. The great Fish, Ehenoster, and Zak Eivers in Sutherland, Fraserburg, and Calvinia, are especially rich in alluvial deposits derived from the Beaufort beds and the dolerite north of the watershed. Where water can be easily brought on to these lands they are extremely fertile. Tontelbosch Kolk in Cal- vinia, a farm on the banks of the Ehenoster, is perhaps the finest grain farm in the Colony. The fall of these tributaries of the Orange is very slight compared with that of the rivers south of the main watershed ; their valleys are more open, and towards their lower ends tend to disappear in the pans or '* vloers,*' the flat alluvial ground quickly flooded during storms but baked hard and white a few hours later, that are a character- istic feature of the arid country south of the great river. In the western Karroo the rivers draining the Eoggeveld escarpment receive a sudden check on leaving the Karroo formation and entering the region of the Witteberg beds, which are of a harder consistency. In the Bosch Eiver Valley on Witte Vlakte a well has been sunk 140 feet through alluvium without reaching solid rock ; this river has deserted its former channel, now marked by a very conspicuous poort in the beds west of the Poortje

KEOENT OR SUPERFICIAL DEPOSITS 367

pan, and has turned southwards to enter the Draai KraaFs Eiver several miles from its former point of junction.

The rivers of the west coast, from the Great Berg to the Olifant's, have considerable tracts of alluvium along the lower forty miles or so of their valleys. The Rerg Eiver alluvium extends to a depth considerably below sea level at many spots where wells give information bearing on the question.

Very little is known of the fossil contents of these river deposits, many of which are of quite recent origin and therefore probably contain only the remains of living or lately extinct animals. The imperfect head of a gigantic buflfalo,^ Buhalus haini, Seeley, measuring eight feet six and a half inches between the horn-tips, although these are broken and therefore shorter than they were originally, is preserved in the South African Museum, and seems to be the only known example of an extinct mammal from the river deposits. It came from the Modder Eiver, forty feet below the surface.

3. In many parts of the southern and western coast districts there are layers of ferruginous rock resting either immediately upon the slates, granite, or other rock of the vicinity, or with the intervention of a few feet of sandy clay. The underlying rock is usually con- siderably weathered, and sometimes bleached by the loss of its colouring matter, which seems to have been transferred to the ferruginous layer. The latter varies very greatly within short distances. It is usually a

1 Seeley (01), p. 199.

368 GEOLOGY OF CAPE COLONY

hard lumpy-looking rock, with innumerable small and irregular channels lined with a red-brown or yellow material. In places the hydrated sesquioxide of iron, limonite, is so free from sand and clay that it might be used as an ore ; but generally there is a large quantity ^ of clay, sand, and subangular fragments of vein quartz and other rocks that do not decompose under the in- fluence of the weather, cemented together by the iron oxide.

Along the edges of the Cape Flats near the high ground of the Peninsula and the Tyger Berg the laterite, or ironstone as it is usually called, is found a few feet below the surface. Farther inland, in the Malmesbury, Paarl, Caledon and other Divisions near the coast, where there is no general covering of sand as on the Cape Flats, the laterite lies just below the soil, or is exposed at the surface, over considerable areas of flat and slightly inclined ground. It is rarely or never found in its typical form on steep slopes, al- though even in such situations the subsoil is in places partly cemented into a fairly hard substance by ferru- ginous matter, thus making an approach to the laterite of the lower ground.

The formation of the laterite is due to the concentra- tion of the iron oxide near the surface in the decomposed rock or subsoil, occasionally in sandy soil that has been brought to its present position J)y water. The nature of the clay that accompanies the laterite in many places, especially where it lies upon clay slates, has not yet been ascertained.

The high-lying lateritic rocks are closely connected

RECENT OR SUPERFICIAL DEPOSITS 369

with the older gravels and alluvial deposits, and are now represented by mere remnants, but the low-lying ones are to-day in process of formation.

Very similar looking laterites appear to have been derived from rocks of diverse natures, such as granite and slate; even the Table Mountain sandstone of the west coast, Clanwilliam and Van Rhyn*s Dorp, is in places covered with a ferruginous cemented material grading on the one hand into the raised beaches of that coast and on the other into the surface quartzites. Near Strand Fontein, a few miles south of the 01ifant*s River mouth, the almost flow-Uke appearance of the remains of the dark limonitic quartzite lying on the Table Mountain sandstone and filling up the open joints at various levels from that of the high tide to 200 feet above it has given rise to the idea amongst the people in the neighbourhood that it is lava. This somewhat re- markable variety of the lateritic rocks is certainly due to the deposition of the hydrated iron oxide, leached out from the underlying sandstones, between the sand grains which reached their present position through the agencies of wind and water.

4. Extensive areas in various parts of the Colony are covered to a more or less considerable depth by sand. These deposits of sand may be roughly divided into two groups ; those formed inland and those near the coast.

The inland sands are chiefly developed in the north- west; the Namaqualand, Calvinia, Kenhardt and Prieska Divisions contain large tracts of sand, and the same is the case with the great dry country formed by the

Colonial portion of the Kalahari Desert. Little is known

24

Sto geologV of cape colony

from a geological point of view of much of this count^> especially of the Kalahari region. The sand occurs in the form of well-defined ridges in the Kalahari and in the more arid parts of the country south of the Orange River,

In Bushmanland (parts of Namaqualand, Calvinia and Kenhardt) the sand is derived from the minerals composing the gneissose granite that occupies such wide areas there. Quartz and felspar are the chief constituents, and by the breaking up of the granite under the influence of the great diurnal change of temperature, one of the climatic features of that region, the minerals are set free to be carried about by the wind and rain. The sand is pink owing to the abund- ance of red felspar, and also to the iron oxide derived from the ferruginous constituents of the igneous rocks, biotite, hornblende, hypersthene and magnetite.

In Prieska the granite, gneiss and mica schist areas are usually covered with deep sand ; the more compact rocks, the quartzites of the *Keis group and the Griqua Town and Campbell Band beds, disintegrate less rapidly than the rocks just mentioned, and do not yield so much sand.

There is much sand in the valley of the Orange Eiver, where it forms extensive dunes in favourably situated spots. This sand is blown from the river banks at times of low water.

In the district between the 01ifant*s Eiver mouth and the Berg River, as far inland as Piquetberg and the Olifant's River Mountains, there is a great quantity of sand. The country is known locally as the Sand Veld.

RECENT OR kSUPERFICIAL DEPOSITS 371

The underlying rock is chiefly Table Mountain sand- stone, although the southern part of the area is probably underlain by the Malmesbury beds. The whole area is characterised by a remarkable scarcity of running water and even of definite stream beds, although the southern part at least has a fairly heavy rainfall ; the northern portion is much drier, but the absence of stream beds is due to the rapidity of absorption of the water by the ground and not to the lack of rain. From the Berg Eiver to the Olifant's, a distance of some seventy-five miles in a straight line, there are only five stream beds to be found ; the Zout, Verloren Vley, Lange Vley, and Jackal's Eivers and the Zand Leegte. The Zand Leegte is a very well-marked valley about twenty miles long, commencing near Eonaqua's Berg and terminating on the coast at Strand Fontein. The lower part of the valley is almost a gorge, some 180 feet deep, and at places only a few yards wide at the bottom, cut out of the hard Table Mountain sandstone. No water has been known to flow down this valley during the period covered by tradition in the district, perhaps 150 years, although a severe thunderstorm sometimes — about once in fifteen years — makes a stream of short duration in its upper part. The valley is being filled in with sand chiefly brought there by the wind. It is decidedly a striking proof that the district is drier now than it was at no very remote period, for there is no doubt that the valley was cut by a stream, and it was made since the advent of the still-living species of moUusca ; for at the mouth of the gorge a raised beach lies about 100 feet

above sea level, and appears to have stretched across the

24*

372 GEOLOGY OF CAPE COLONY

ravine ; the raised beach contains the shells of moUusca of the same species as those found on the modern beach in addition to the water-worn pebbles and boulders that make it a conspicuous feature on the top of the cliffs.

The evidence afforded by the Zand Leegte will ex- plain the development of the Sand Veld. The tops of the sandstone hills still project above the sand, but the old valleys, that were carved out by rivers before the climate became as dry as it now is, are almost entirely filled up by the sand derived mainly from the sandstone hills and from the mountains built of the same rock to the east of the Sand Veld. Where exposed to constant sifting by the wind the sand is white or very light- coloured, but throughout the greater part of the area it is reddish. The red colour is certainly due to oxide of iron, but the source of the iron is not so evident. In sinking wells it is found that the lower layers of sand are paler in colour than those near the surface ; it may be that the rain water, with the aid of organic com- pounds taken up during its passage through the soil, dissolves the iron oxides deep under the surface and brings them in solution to the top where it leaves them as thin films round the sand grains on evaporation. But it is possible that the very fine red dust brought into that part of the country by the strong east winds will account for the red colour of the surface sand. The fertility of the Sand Veld is remarkable, considering the general appearance and nature of the soil, good grain crops being obtained when average winter rains fall ; it is probable that the wind-borne dust adds the necessary constituents to the otherwise extremely poor soil.

The Sand Veld sand passes somewhat abruptly into

RECENT OR SUPERFICIAL DEPOSITS 373

the dunes that line the west coast. The proximity of the coast makes itself noticeable by the increase of calcareous matter in the sand ; the carbonate of lime is derived from marine shells which are pounded to dust on the shore and then blown inland.

Patches of sand dunes of greater size than usual are found south of Saldanha Bay, on the shores of False Bay whence the sand has invaded the Cape Flats, near the Bot Eiver mouth, at Cape Agulhas, Cape Barracouta and Cape Kecife. These are calcareous sands composed of a mixture of broken shells and fragments of minerals, chiefly quartz. The strong winds and constant supplies of fresh sand, as well as the facility with which the dune sand is moved, account for the diflSculty of getting vege- tation to gain and maintain a footing on these sand areas, which are a source of danger to the farms behind them.

5. The calcareous sands of the coast belt pass into limestone by the solution of carbonate of lime from parts of the mass, and its deposition near the surface when the water evaporates. In almost any part of the south-coast dunes a thin hard crust can be found cover- ing sand which has been protected from the wind for some time ; it may be less than a quarter of an inch thick, and is easily broken. By the long-continued de- position of the carbonate of lime the sand dunes are converted into hard rock through a distance of many feet from the surface, and where repeatedly wetted and dried, as happens when the sea has encroached upon old dunes, the rock becomes intensely hard and weathers with a peculiarly jagged surface. At Hoetjes

374 GEOLOGY OF CAPE COLONY

Bay, an inlet of Saldanha Bay, the limestone derived from hardened dunes has been quarried for building stone, and furnished the material of which the General Post OfiBce and South African Museum are mainly constructed. In the large quarry at Hoetjes Bay the gradual hardening of the stone from the deepest portion exposed towards the exterior is well seen. This lime- stone contains a smaller proportion of quartz sand than usual, about 12 per cent., but the composition varies considerably according to the amount of carbonate de- posited between the original grains of the rock, and also according to the proportions of broken shell and particles of non-calcareous minerals in the original sand.

False bedding is a very marked feature in many sand dunes, being perhaps better developed in wind- borne accumulations than in sediments deposited under water. Magnificent examples of this structure can be seen in several cliflf sections through the hardened dunes on the south coast between Cape Agulhas and Mossel Bay, and again to the east of Algoa Bay.^ Plate XXI. is from a photograph of a cliflf near Struys Point on the Bredasdorp coast.

In addition to the usually fragmentary remains of marine shells the dune limestones contain many fossils of animals that lived upon land, and these are in a much more perfect condition than the former. Snail shells, especially a large species of Helix that is com- monly found living near the coast, are abundant in the limestones of Saldanha Bay and the south coast.

•

1 Atherstone (58).

a l&>^p|i<iif>f<ii>i!t#>

Ir

376 GEOLOGY OF CAPE COLONY

Mammalian remains are frequently found, and they include species such as the elephant, rhinoceros and eland, that are no longer living in the neighbourhood. Hitherto no extinct forms have been discovered in any of the coast limestones.

In the Bredasdorp Division there is a prominent range of dune limestone hills stretching from near the village to Cape Infanta.^ In their western part the hills lie some twelve miles from the coast, and are separated from it by a tract of low ground ; near the coast the country again becomes hilly owing to the modern dunes. The inland range must be of considerable antiquity, and it is now being destroyed by the weather and rivers without receiving any fresh material to compensate for this loss. These old dunes were formed at a time when the coast was at a lower level than now, during the period represented by raised beaches in several parts of the Colony.

The dune limestones are in places rather easily dis- integrated, and weather very unequally, hence shallow caves are of frequent occurrence in them. At Cape Infanta there is a fairly large cave with a small en- trance on the cliff; the roof is hung with stalactites, long tapering tubes of calcite deposited from the water percolating through the overlying limestone, and the floor is formed by a mixture of sand and bat-guano. The origin of the cave was probably due to a stream that no longer exists. Thi3 cavern is perhaps the largest (some 160 feet long and 20 feet high in parts)

^ For a more detailed description of these and allied rocks see Rogers and Schwarz (97), p. 427, etc.

EECENT OR SUPERFICIAL DEPOSITS 377

yet found in the dune limestones. Other caves of considerable depth, such as the Kellers near Danger Point, have streams of water still flowing through them.

6. On the coast side of the Langebergen there is frequently ^ thin layer of whitish impure limestone immediately below the soil, and a similar rock covers wide areas in the western portion of Malmesbury and Piquetberg. It is possible that some of this represents dune limestones that have disappeared, or it may be due to the slow accumulation of shell fragments blown inland from the coast. The calcareous layer is especially well developed between the Kafl&r Kuils and Gouritz Rivers in Riversdale. There is a particular variety of the limestone seen in the soil about a foot below the surface that is now in process of formation. This is a nodular rock, rather compact, and it contains numerous sand grains and other particles derived from the soil. The calcareous matter collects together in certain spots and forms irregularly shaped lenticular lumps; neigh- bouring masses coalesce and produce layers. The bulk of the clayey material in the soil seems to be pushed aside by the calcite, but the sand grains remain behind. This rock is well shown in some of the railway cuttings berween Heidelberg and Eiversdale. It is similar to the ** Kankar ** of India.

The springs that come from the Bokkeveld series and from the Karroo beds frequently deposit a white tufaceous limestone which forms irregular layers in their neighbourhood, filling up the joints of the exposed rocks and cementing together the particles of soil. The

378 GEOLOGY OF CAPE COLONY

springs are usually weak and in the course of time block up the channels through which they flow by the deposition of the tufa. Thus many patches of lime- stone occur without any sign of water. The farmers are aware of the connection between the tufaceous limestone and spring water, and are often successful in opening up springs by removing the surface and follow- ing up any traces of water that may appear. It is quite clear, however, that the process of filling up of the joints through which the water flows may have gone on so far that the attempts to release the water will be unsuccessful; or, again, the water may have found another exit at a lower level.

The sediments of the Karroo formation contain a fair proportion of carbonate of lime, and the dolerite which is so abundant in the form of intrusions in these beds contains about 10 per cent, of calcium oxide ; this, on the decomposition of the dolerite, is chiefly converted into carbonate of lime. From these two sources the impure limestone that is so widely spread between the main watershed of the Colony and the Orange Kiver has chiefly been derived. Every heavy rain that carries the products of decomposition from their place of origin to the flat ground, and especially to the shallow pans, brings with it some carbonate of lime which it leaves behind on evaporation. To this source must be added the slow creep of water towards the surface by capillary attraction and the influence of plants.

The thick calcareous tufas that are found in the Orange Eiver Valley, as in the neigbourhood of Hope Town, were probably deposited in pans that have been

RECENT OR SUPERFICIAL DEPOSITS 379

cut through by the Orange Eiver ; but at present very little is definitely known of the nature and extent of these old tufas.

7. At many places on the coast there are beaches of rolled pebbles, sand and shells at various heights above the present day shore. These deposits frequently rest upon a more or less extensive shelf cut into the sloping land behind the shore.

The most northern raised beaches yet found in the Colony are on the coast between the 01ifant*s Eiver mouth and Thorn Bay. The coast is formed by a range of cliffs about 100 feet high, composed of the Malmesbury beds to the north of Strand Fontein and of Table Mountain sandstone to the south. South of the Zand Leegte the cliffs are remarkably fine, and they are broken into many small inlets and rocky points by the attacks of the Atlantic waves. The Table Moun- tain sandstone dips eastwards at about 35 °, and is cut flat on the top of the cliffs. The old beach deposits lie on this flat surface, and consist of water-worn boulders mixed with sand. The beach has been cemented into a hard conglomerate by the deposition of iron oxides and siliceous matter in places, and in these conglom- erates shells or fragments of them are scarcely to be found ; but in other parts of the beach at the same level, where this process has not gone so far, shells belonging to species still living on the west coast are abundant, and the rock is a loose shelly conglomerate. Transitions from the latter to the former condition of the beach are to be found, and as the amount of change increases the shells decrease in quantity ; they are dis-

380 GEOLOGY OF CAPE COLONY

solved without being accurately replaced by the cement- ing material.

On the peninsula to the west of the south end of Saldanha Bay there are shelly limestones with abundant shells of living marine forms lying from ten to twenty feet above high water. These limestones pass inland into the hard dune limestone with land shells. They are an old beach formed when the land stood somewhat lower than at present. It is a curious fact that the dune limestone passes below sea level in Saldanha Bay and on the south coast ; this rock on careful examination is always distinguishable from the calcareous beach de- posits, and its occurrence below sea level in the same districts as the raised beaches points to a slight sinking of the land since the beaches were formed and elevated.

In the Cape Peninsula there are a few patches of supposed beach deposits at a height of from 50 to 100 feet above the sea. They contain the remains of living species of moUusca. It is a curious fact that no shell- bearing sands or other recent marine rocks have been dis- covered below the Cape Flats, but they certainly should be there if the correct interpretation has been found for the deposits just mentioned from the Peninsula.

In the neighbourhood of Hermanns there is a very well-marked rock shelf between the Klein Eiver Moun- tains and the coast about fifty feet above the sea. It is a wave-cut terrace of Table Mountain sandstone, covered in places with dune limestone. Similar terraces are to be found near Danger Point, Zout Anys Berg and Pot Berg. At Cape Infanta there is a raised beach at the base of the dune-limestone, which there forms high

RECENT OR SUPERFICIAL DEPOSITS 381

cliflfs. The beach conglomerate is about 100 feet above the sea. On the shores of Algoa Bay there is a well- developed terrace cut through the Uitenhage beds, it slopes gradually tov^ards the sea from a height of about 400 feet above high tide in its inland portion to 200 feet where it is concealed under the blown sand of the coast. The shelf is covered in places with shelly conglomer- ates containing the remains of moUusca still living off the South African coast. A characteristic shell in this deposit is a very large Pectunculus. At lower levels nearer the sea there are patches of old beaches which contain shells belonging to living species.^ Many of these raised beaches contain numerous species of shells, and the careful collection and determination of these from the different deposits is certain to jdeld interesting results.

Near the mouth of the Buffalo Eiver there is a layer of earthy clay 200 feet above sea level containing re- mains of recent shells ; it was regarded by its discoverer, Mr. McKay, as a marine deposit, and he found a frag- ment of native pottery in one of the shell layers.^ This fragment of pottery is the only recorded evidence of human occupation of the country at the period of these raised beaches and allied deposits, but before it can be accepted as good some corroborative facts should be brought to light elsewhere.

Although the evidence bearing on the question of a

^ The best account of these deposits is still that of G. W. Stow (71), pp. 515-22. A list of species found in the low level beaches near Port Elizabeth has lately been published by J. P. Johnson (08), pp. 9-11.

2 Quoted by Huxley, Scientific Memoirs^ vol. iii., p. 300 ; also QeoU Magazine, 1868, p. 201.

382 GEOLOGY OF CAPE COLONY

recent .change in level of the whole coast line is so widely distributed much remains to be done before it can be fully understood. So far as it goes it is in accordance with the presence of the river-cut high-level plains now deeply channelled by the existing streams. There is good reason to believe that while these plains were being made the higher raised beaches were also in process of formation. In the Swellendam Euggens, for instance, the old gravel and alluvial plateau that slopes gradually towards the coast and is trenched by tributaries of the Breede Eiver terminates at the foot of the Bredas- dorp limestone hills, which we have seen were once calcareous sand dunes. These are continued into the limestone that overlies a pebbly beach deposit at Cape Infanta, now being cut back by the sea. At the time when the inland plateau was being cut the dunes that now form the limestone range were being piled up by the wind, and the coast was indented by a broad bay between Cape Infanta and Bredasdorp village. The eastern corner of the bay extended farther seawards than the present position of Infanta, for the high cliffs made of Table Mountain sandstone in their lower part and of the beach deposit and limestones in the upper half must have been undergoing destruction ever since the raised beach was removed from the reach of the sea.

In the Algoa Bay region the high-level gravels of the Zwartkops Heights were probably formed at the same time as the terraces covered with surface quartzites and allied deposits near Grahamstown, and the wide rock terrace traversed by the main road from Port Elizabeth

RECENT OR SUPERFICIAL DEPOSITS 383

to Humansdorp. The upward movement of the land which raised the Zwartkops Heights beach to its present level also brought about the renewal of the downward erosive power of the rivers inland, so that they trenched the gravel and quartzite plateau of Grahamstown.

8. Near their mouths many South African rivers ex- pand into wide shallow lagoons. The larger rivers, such as the Berg, Breede, Gouritz, Kei and St. John's, which maintain open channels to the sea throughout the year, have comparatively small lagoons or none at all, although some of them, such as the Berg, give rise to shallow vleys beyond their banks in times of flood. The smaller streams whose mouths are more or less regularly choked up by sand bars terminate in vleys of various dimensions. * The formation of a wide vley in place of a sharply defined channel is easily understood ; the water flowing into the lagoon cannot escape quickly, but filters slowly through the sand bar ; it therefore stands above the sea level, and owing to its constant movement it laps against the usually soft sandy banks and gradually washes them away, depositing the debris in the deeper portions of the channel. The absence of an open mouth prevents the tide from assisting to keep the channel clear. The mud brought down by the river mingles with the sand blown or washed by rain into the vley and makes a sandy loam, which tends to form a flat surface somewhat above sea level, so that should the mouth become open for a long period the river will flow through a flat alluvial tract just before entering the sea. Such may be the origin of the flats at the mouths of the Zwart Kops and of the Great and Klein Brak Eivers in

384 GEOLOGY OP CAPE COLONY

Mossel Bay. The same feature is seen at the Kowie mouth, although in this case the channel is maintained by the walls built for the harbour. The Bot and Klein Eivers in Caledon and Bredasdorp have large vleys, which are only open after the winter rains. Many large lagoons, such as Zoetendal and Salt Biver vleys in Bredasdorp, have quite small rivers flowing into them, and are very rarely open to the sea. Zoetendal vley is fed by two rivers, and near the mouths of one of them a small stream flows to the sea at certain times by a longer route than would be afforded by the vley if it were open to the sea near Northumberland Point. In the Transkei and Fondoland very many small streams rarely bring down enough water to break through their sand bars, and in time they will form corresponding alluvial tracts with small channels traversing them. The comparatively recent elevation of the coast that enabled these rivers to cut deep valleys through the coastal plateau has not been of sufficient duration to allow them to silt up their lagoons.

A vley is sometimes formed along the course of a river just behind a ridge of rock that is with difficulty cut through by the stream. The softer rock behind the obstruction allows the river to cut out a wide plain, and by the unequal distribution of debris over the plain the bed of the stream may be raised slightly above the level of the plain, causing the latter to be flooded at times. A process of this sort has taken place in the valley of the Bosch Eiver where it approaches the Bokkeveld hills west of Witte Vlakte. An extensive vley or pan, on the farm named Poortje,is the result, and the river has found

RECENT OR SIlPERP^lOlAL DEPOSITS 385

an easier course to the south, where it joins the Draai KraaFs Eiver.

Very extensive vleys, which rarely have any water in them, are formed along the rivers entering the Orange Eiver from the south. Not much is known as yet about these great " vloers,'* but they are probably due to the flooding produced by blown sand obstructing the rivers, which tend to distribute their silt over wide areas and thus to level up their valleys, that have a very gradual fall.

The water that gathers in these river vleys is some- what brackish from the salts derived from the surface soil in their drainage basins, but these vleys do not seem to contain salt deposits of any value as a source of that commodity.

There is another class of pan, not obviously connected with the river vleys, whose origin is more difficult to account for. There are two subclasses of these; the first consists of the pans near the coast, and the second of those lying far inland.

The pans on the coast are usually at a low level,

separated from the sea by a belt of sand dunes. There

are several of these on the west coast south of the Oli-

fant's Eiver. Eain water collects in them, and owing to

there being sufficient clayey matter or limestone round

them the water does not drain away but evaporates

slowly, leaving a thin crust of salts, mostly composed of

sodium chloride or common salt. Usually the thin

crust is not sufficiently free from sand to be used for

domestic purposes, so shallow trenches are dug in the

floors of the pans during the dry season and a deposit

25

386 GEOLOGY OP CAPE COLONY

of salt three or four inches thick is formed in them after the rains. The salt is probably collected by the rain water in its course through the surrounding sandy soil, which receives it gradually from the sea in the form of spray or attached to the grains of sand blown from the shore.

On the coast of Bredasdorp there are several produc- tive pans. Some of them are within a short distance of pans which contain fresh or nearly fresh water only, yet no difference in the conditions of the salt and fresh vleys is observable. This fact is difficult to explain on the supposition that the salt is washed into the pans from the surrounding soil ; but at no distant period the low-lying parts of the Bredasdorp coast must have been under the sea, or at any rate liable to inundations of salt-water at high tide during storms, and it is possible that the salt derived from this source is still inexhausted in spots where, owing to a slightly lower level or to the presence of more favourable surface deposits, a larger quantity of the sea water evaporated than elsewhere.

Perhaps the richest pan in the Colony is that on the Zwartkops heights north of the river of that name. The pan is surrounded by the shelly beach deposits de- scribed on page 381, and is underlain by the Sunday's Eiver beds. An enormous quantity of salt is taken yearly from this pan, yet it shows no sign of exhaustion. The salt must come from the rocks close at hand, and as no beds of rock salt are known to exist in the Sunday's Eiver beds, or indeed in any other formation in the Colony, it seems certain that it is derived from the recent beach deposits.

RECENT OR SUPERFICIAL DEPOSITS 387

The inland salt pans are mostly found near the Orange River, where vleys are abundant. A great stretch of country, extending from the north of Calvinia through the Divisions of Kenhardt, Fraserburg, Carnarvon, Prieska, Hope Town and Kimberley, is particularly rich in more or less circular pans of various sizes, from a few yards to a few miles in diameter. This tract is some- times called the Panne-veld, and coincides roughly with the outcrop of the Dwyka conglomerate, a rock that is less permeable than either the sandstones and shales that lie south and east.

The salt in these inland pans must be derived from the surrounding rocks. There is no evidence that the interior of the Colony has been under the sea, or indeed under water of any kind, since the close of the Karroo ' period, and that water was probably not salt. There is very little information as to the composition of the salt from these pans, but sulphates (of calcium and magne- sium) as well as chlorides are present.

In a pan at Klip Fontein's Berg, in Clanwilliam, the depression, from which common salt is gathered for domestic purposes, is surrounded by a thick layer of carbonate and sulphate of lime. The sulphate of lime (gypsum), occurs in small and large crystals embedded in a calcareous, sandy mud, and it forms the larger part of the deposit. The material is well stratified, and the layers are thin.

The coast pans owe their existence to the barrier of dune sand blown up from the shore, but an adequate explanation of the inland pans has not yet been given.

25*

388 GEOLOGY OF CAPE COLONY

The Eecent Deposits and the Human Occupation

OF THE Country.

Except in the case of the piece of pottery found in the high-level marine clay of East London no human remains have been found in any of the higher raised beaches, or in the quartzites or other deposits on the higher plateaux. The surface quartzites indeed furnished the favourite material of which the aborigines made their rough knives and other implements with a more or less sharp edge, proving that this rock was available at an early period of man's occupation of the country.

By far the greater number of stone implements are found either upon the surface or at a small depth within the soil, and specimens that in Europe would be regarded as of Palaeolithic type, i.e., roughly fashioned without ground or carefully chipped edges, may well have been in general use during the early years of the European settlement. Although the use of the round stones with holes through them as make-weights to digging sticks amongst the Bushmen is recorded by Burchell ^ and other travellers, the use of stone axes or weapons of the nature of the stone " celts " found in Europe does not appear to have been seen.

In the south-western districts from the Peninsula to the Olifant's Eiver stone implements with a pear-shaped or oval outline are not infrequently met with. Any hard close-grained stone was used for their manufac- ture, but the surface quartzite seems to have been the most abundant and suitable stone. Amongst many

^ Burchell, Travels in South Africa^ vol. ii., p. 29.

RECENT OR SUPERFICIAL DEPOSITS 389

specimens collected in any place in that area or in the Karroo a few made of the chert from the Upper Dwyka shales are nearly always found. In the north of the Colony the jaspers of the Griqua Town beds and the cherts of the Campbell Band series seem to have been most widely used.

In the coast districts the stone implements are often found lying with the remains of edible marine shells and fragments of coarse pottery at various heights above the sea. The abundance of shells in such ** middens " is liable to make a casual observer think the deposit is a raised beach. In the Transkei and Pondoland the mak- ing of these middens by Kaffirs can be seen on most days, but especially at spring tides. The natives collect the shells, carry them to a convenient spot close to the shore, and there remove the edible portions which they take back to their kraals in baskets or cloths, leaving the shells behind. In this way astonishingly large piles of more or less broken shells accumulate in course of time.

In the inland area the implements are chiefly met with near streams or springs, on flat-topped kopjes in the Karroo, and near the caves in the mountains. The presence of small fragments of stones unlike any that crop out in the immediate neighbourhood is the sign that more or less well-fashioned cutting or scraping implements and stones used for rubbing or digging may be expected.

In the drier parts of the Colony the surface of an implement that lies uppermost is generally coated with the thin varnish-like glaze that forms on exposed rocks under the influence of the weather in such places. It

390 GEOLOGY OF CAPE COLONY

is not known how long a piece of rock must be ex- posed before it gets this glaze, so its occurrence does not enable us to settle the minimum age of an imple- ment.

In all the occurrences above mentioned, with the exception of Mr. McKay*s pottery which must at present be regarded with some scepticism, there is nothing in the position of the chipped stones to indicate their great antiquity. A few years ago, however, Mr. Peringuey, the Assistant-Director of the South African Museum, found a large series of rudely shaped stones in certain gravels at a considerable height above the present levels of the valley bottoms in the Stellenbosch district. De- tails of these finds have not yet been published, but there is no doubt that the stones were fashioned by human hands, or that they occurred several feet from the sur- face in old river gravels that must have been laid down at a period far removed from the present according to human reckoning, but less ancient than that of the high level plateau gravels and quartzites. The implements vary in size, but they are remarkably large on the aver- age ; one is as much as fourteen inches long. They are more or less symmetrically formed, with one end more pointed than the other. Many of them were evidently made by chipping water- worn boulders of suitable shape. They are all made of compact quartzite or hard sand- stone, probably from the Table Mountain series. This interesting discovery opens up a wide field for investiga- tion, and the pursuit of it will assuredly give us some definite knowledge of the earlier phases of man's occupa- tion of South Africa.

RECENT OR SUPERFICIAL DEPOSITS 391

Hot Springs and Their Eelation to the Structure

OF THE Country.

Springs from which water issues at temperatures con- siderably above that of the air ^ are rather numerous in Cape Colony. Some of these yield water of much the same composition as ordinary spring water ; the Brand Vley, 01ifant*s Eiver (Clanwilliam), and Montague springs are of this kind. The water from the Caledon springs contains much ferrous carbonate, and the Warm Water Berg spring water has a smaller quantity of the same salt in it. Sulphuretted hydrogen is a constituent of the Malmesbary, Cradock, and Graafif Eeinet mineral waters.

The majority of these springs rise in Table Mountain sandstone areas, but their distribution is not connected obviously with the great dislocations or folds visible in that formation ; there is no spring situated on or near the largest strike faults, those of Worcester and the Cango, nor does one occur in the more intensely folded portions of the east and west ranges south of the

1 Detailed information on the contents of the water from some of these springs will be found in Krauss (43), Gumprecht (51), Noble (93), and Daniel 1 (95), and in prospectuses issued by their present proprietors. A systematic examination of the waters, not only for their saline con- stituents but also for the gases containing rare and radio-active sub- stances would be of great interest. The temperatures of some of the springs are the following : —

Brand Vley 145 ° F. Malmesbury 88 ° F.

Caledon 120 ° F. Cradock 86 ° F.

Olifant's River (Gudtshoorn) 114 ° F. Koega 79 ° F.

Montague 112 ° F. These figures are taken from the papers cited ; Dr. B. Marloth of Cape Town kindly gave me corroborative inforraation concerning many of them,

392 GEOLOGY OF CAPE COLONY

Karroo. The Olifant's Eiver (Clanwilliam) hot-bath is on the eastern limb of the gentle anticline that forms the Cardouw Mountain; the hottest spring, that of Brand Vley, is near the locality where the dip of the Table Mountain sandstone south of Worcester changes from north to east ; the Caledon, Warm Water Berg, Montague, and Olifant's Eiver (Oudtshoorn) springs issue from the sandstone on the flank or at the end of anticlines.

The Malmesbury spring flows from a mass of granite, and those of Cradock, Graaff Keinet (cold) and Aliwal North from the nearly horizontal Karroo formation in the great interior basin. It is remarkable that the Malmesbury and Karroo mineral springs contain sul- phuretted hydrogen, while the others do not. This gas, in small quantities, is given off by many of the ordinary springs in the Karroo, and is probably derived from the decomposition of pyrites. Whether the gas in the hot springs has a similar source is of course not known.

The probable reason of the high temperature of the springs is that the water comes from great depths. So far as one can judge from the surface geology none of the springs is in any way connected with volcanic action. Many of the older travellers took the dark slaggy-looking deposits of hydrated ferric oxide at Cale- don for lava, but the dark rock is derived from the ferrous carbonate in the water by oxidation on contact with the air. In the western Karroo there are several cold springs at the foot of the Zwart Ruggens that leave a similar deposit of limonite, but there is hardly sufficient iron in the water to make it taste unpleasant,

CHAPTEE XI.

THE GEOLOGICAL HISTORY OF THE COLONY.

Even under the most favourable circumstances it is a difficult task to decipher the records of the past in such a way as to make clear the evolution of a country from the earliest times to the present day. By the *' earliest times " we mean the period at which the lowest or oldest of the sedimentary rocks, recognisable as such, were deposited. All over the globe these ancient rocks have been found to possess characters that cannot be looked upon as original, but which must be regarded as having been produced by metamorphism due to great pressure, heat, the action of percolating water, or all three combined.

In the Cape Colony many of the rock groups classed under the heading " Pre-Cape rocks " have been altered by these agencies, and no attempt can be made at present to unravel the history recorded by them. We have no idea, for instance, where the land lay from which the sediments were brought to build up the quartzites, slates and schists of the 'Keis or Malmesbury series. Since their deposition they have been intensely folded, invaded by enormous masses of granite, and then subjected to long periods of denudation. The cases of the Campbell Band, Griqua Town, and Matsdp series

m

394 GEOLOGY OF CAPE COLONY

are not much better; but we know that some of the rocks belonging to the two former came to rest again as pebbles in the Matsdp beds. There is reason to believe that the Ibiquas and Cango series derived some of their materials from the older Malmesbury beds and from the granite intrusive in the latter.

At the commencement of the Cape period, i.e., about Lower Devonian times, we may imagine that a great tract of land lay west and north of the position of the southern part of the Colony, for the materials comprising the Table Mountain series become somewhat coarser in those directions. That land furnished the enormous amount of sand, almost entirely of quartz grains, that now is the Table Mountain sandstone. This sandstone, which is roughly in the form of a broad belt about 500 miles long and 100 wide, was deposited in shallow water; denudation and earth movements have played a greater part in defining its present boundaries than original deposition. During its formation the floor must have been gradually sinking to allow of the accumulation of 5,000 feet of sediment which through- out bears evidence of deposition in shallow water. The shale bands may possibly indicate deeper water con- ditions, but not necessarily so ; the striated pebbles in the Pakhuis shales and mudstone prove that glacial con- ditions prevailed for a time during that remote period, and that the ice which floated away from the shore carried with it these flattened and scratched pebbles, and dropped them in the mud being deposited at some distance from the shore. The fact that the series is thinner near Nieuwoudtville, at the extreme northern

THE GEOLOGICAL HISTORY OP THE COLONY 395

end of the area occupied by the overlying Bokkeveld beds, than farther south, points to the subsidence which allowed of the accumulation of the sandstones having gradually proceeded northwards. This means that de- position began in the south earlier than in the north, so that the bottom of the series in the Bokkeveld Moun- tain area was formed later than the lowest beds in the Worcester or Ceres Divisions.

We cannot regard the Table Mountain series as a marine formation, it is probably a fluviatile deposit laid down near the source of origin of the materials composing it. The great thickness of sediment, and the evidence throughout that it was laid down in shallow water, prove that the area occupied by it underwent slow but steady depression, which continued for a long period after the peculiar conditions under which it was formed came to an end. This depression in the southern part of the Colony must have gone on till some time during the deposition of the Karoo formation, perhaps till late in the Beaufort period ; it was brought to a close by the earth movements which produced the northern and western mountain ranges.

The northern limit of the depressed area cannot be defined, but it probably lay to the north of the thirtieth parallel.

During the Bokkeveld period the waters of a southern ocean that lay south and west of the Colony, and which spread at least as far as the position of the Falkland Islands and the South American Continent, gained access to the area where the Table Mountain series had been deposited. The presence of plant remains in

396 GEOLOGY OF CAPE COLONY

the Bokkeveld beds, along with the marine shells, shows that the land on which the plants grew was not far off. In the account of the Bokkeveld beds in chapter iv., the generally noticed increase of sandstone towards the north and west was explained on the supposition that the sediments wete chiefly derived from land lying north and west of the districts where the Bokkeveld beds occur.

Marine conditions prevailed in the southern part of Africa till the middle of the Bokkeveld period, when open connection with the sea seems to have been cut off, for the muds, shales and sandstones of the upper Bokkeveld and the Witteberg series contain no other than plant remains. The cause and manner of this loss of connection with the ocean cannot be explained, as the evidence which might solve the problem lies below the waters of the Atlantic. The abundance of sandstones in the Witteberg beds, with their occasional white quartz pebbles, often in some respects closely resembling the Table Mountain sandstone, point to a recurrence of the conditions under which the latter was formed, though the frequence of thick shale bands proves that much of the finer grained sediment came to rest within the Colonial area in Witteberg times, while in the earlier period of the Table Mountain sandstone much less of the clays and silt, which must have been produced during the destruction of the rocBs that furnished all the sand now forming the Table Mountain sandstone, remained in the same area.

Plants are the only fossils hitherto discovered in the Witteberg beds, and they are usually found in frag-

THE GEOLOGICAL HISTORY OP THE COLONY 397

ments, bits of stems without leaves or other organs, and these fragments probably drifted far before becom- ing waterlogged. In the Eastern Province some beds are largely made up of compressed coaly-looking stems. Current bedding and ripple marks are very usual phe- nomena in the Witteberg series. In the south of the Colony the Witteberg period was brought to an end by the deposition of the green shales and mudstones of the Lower Dwyka beds, and no physical break or uncon- formity separates the two groups of rock. Deposition must have gone on continuously in the south of the Colony while the great change of climate took place that caused the glaciation of the country to the north of the Karroo.

While the deposition of sediments of various kinds went on uninterruptedly in the southern districts from the period of the Table Mountain series till far on in that of the Karroo formation, a rising of the floor began in the country north of the thirty-third parallel at some time during the Bokkeveld or Witteberg periods ; for both in the west and east of the Colony north of that parallel of latitude an unconformity separates the lowest beds of the Dwyka series from the Cape formation. This rising of the land relatively to the water level must have taken place very gradually, as there is no strong discordance between the newer and older rocks. The Witteberg and Bokkeveld beds become gradually thinner and thinner northwards owing to the removal of a greater thickness of the beds in that direction during Pre-Dwyka times.

It is clear that in the country immediately north of

398 GEOIXXIY OF CAPE COLONY

Karroo Poort, where the only beds usually met with in the southern districts that are missing are the Lower Dwyka shales, the exposure of the Witteberg series must have been of very short duration. Farther north their exposure to the agencies of denudation began at an earlier time, so that more and more of the Witteberg and Bokkeveld rocks were washed away before the Dwyka conglomerate was laid down upon the renmants. It is obvious that deposition and denudation on a large scale cannot go on at the same time in one and the same district, so that at Matjes Fontein on the Oorlog's Kloof Eiver, where only the lowest of the Bokkeveld beds remain between the Dwyka conglomerate and the Table Mountain sandstone, and where some 2,000 feet of the Bokkeveld beds, if the series was ever so com- plete there as farther south, are missing, the removal of the rest of the group must have taken place during the formation of the Witteberg beds in the south. We can be certain, therefore, that the Witteberg beds were never deposited in the area just north of Matjes Fontein (Oorlog's Kloof Eiver).

The northward extending depression, which allowed first the Table Mountain sandstone and then the marine beds of the Bokkeveld series to be deposited north of the thirty-third parallel, gave way to the opposite movement of upheaval at some time during the deposition of the upper part of the Bokkeveld or lower part of the Witte- berg group.

It is possible that this change of direction in the vertical movement of the land was coincident with the beginning of the change in geographical conditions

THE GEOLOGICAL HISTORY OF THE COLONY 399

which eventually brought about the cutting ofif of the Colonial area from the ocean in the middle of the Bokkeveld period.

The shore line at the commencement of the Dwyka period lay in an approximately east and west direction through the neighbourhood of Karroo Poort, and the shales and muds which were deposited near it are very like the more argillaceous sediments of the Witteberg series ; they contain none of the fossil plants found in the latter, but a few plants of a similar nature to some of those found in the Ecca beds have been obtained from them. This shore line appears to have gradually crept northwards, but it did not gain much upon the land area to the north before the conditions set in that caused a general glaciation of that land.

We have seen in a previous chapter that there can be no doubt of the fact that South Africa north of the thirty-third parallel was in part, at least, covered with snow and ice, and that the Dwyka conglomerate is made of the mud, sand, pebbles and boulders derived from the glaciated country.

In the northern parts of the Colony, as well as in the eastern districts of the Transvaal and western portion of the Orange Eiver Colony, the Dwyka conglomerate has to a certain extent the character of a morainal de- posit. It lies upon a well-striated rock surface, and is mostly unbedded ; it is a sandy mud or clay with large blocks and smaller fragments of various kinds of rock scattered through it. The occasional patches of con- glomerate with a shaly matrix in the north can well be looked upon as having been formed in small glacial

400 GEOLCXiY OP CAPE COLONY

lakes within the morainal area, i.e., the area which belonged to the land rather than to the water.

Evidence of the movement of solid ice over a sm"face of earlier deposited conglomerate occurs as far south as Eland's Vley in the western Karroo. In Natal, 2° N. of that latitude, the conglomerate rests upon a glaciated surface of the Table Mountain series. It seems likely that the conglomerate to the south of Eland's Vley also rests upon a glaciated surface of the Bokkeveld or Witteberg beds, but this has not yet been proved.

The Dwyka conglomerate in the south is certainly much thicker on the average than it is north of the Karroo, and a gradual diminution in thickness has been noticed in passing northwards along the western border of that country from Karroo Poort to Calvinia. This is in perfect concord with the fact that the transgression, or gradual extension of the water area, and consequently of the shore line, took a northerly direction as shown by the increasing gap in the succession below the Dwyka series. There are no representatives of the Lower Dwyka shales in the north, and a considerable thickness of the southern conglomerate must have been deposited before the northern conglomerate began to be laid down. The few feet of conglomerate at Kimberley, for instance, were probably formed during the deposition of the uppermost part of the southern conglomerate.

The conglomerate in the south of the Colony was probably formed entirely under water ; into the sand and mud there being deposited the pebbles and boulders, many of them well scratched, were dropped by the floating ice that drifted southwards from the shore.

THE GEOLOGICAL HISTORY OF THE COLONY 401

No remains of animals or plants have been found in the Dwyka conglomerate, so the question of the nature of the water in which it was deposited is unsettled. The absence of marine shells is certainly presumptive evi- dence against the water having been a part of the ocean, for it is well known that a cold climate is by no means unfavourable to marine life at the present day. Many genera of molluscs and crustaceans are represented by unusually large species in arctic and antarctic regions. In any case the absence of fossils is difficult to explain, but considering also that only land or fresh water forms have been found in the beds underlying and overlying the conglomerate it is more probable that the water in the Dwyka basin was fresh than salt. The absence throughout the Karroo formation of deposits of rock salt, gypsum or other substances that accumulate in inland basins with no outlet is good evidence that the basin in which the rocks were formed was kept fresh by the continual escape of the water.

We may picture to ourselves a great inland water

basin, with one or more outlets to the ocean towards

the south, and covering what is now the southern part

of the Cape Colony, at the commencement of the

Dwyka period. The southern mountain ranges were

not yet in existence, the rocks which afterwards built

them up were lying horizontally below the surface of

the lake. The nearest land lay to the north; the

southern portion of it consisted of the then recently

exposed Witteberg deposits, north of this area there

were belts composed of the Bokkeveld and Table

Mountain series, while still farther north lay a hilly

26

402 GEOLOGY OF CAPE COLONY

country composed of the Pre-Cape rocks. This country gradually became snow-clad, and glaciers and perhaps eventually a sheet of ice, of too great size to be called a glacier, slowly moved from various directions towards the lake, carrying with them mud, sand, pebbles and large blocks, derived from the surrounding land. Most of these materials reached the bottom of the great lake, but it is more than likely that parts of the unbedded conglomerate in Prieska and elsewhere in the northern districts are the remains of moraines that lay between the ice and the floor in the lower parts of the land, or that were piled up in front of the ice. Meanwhile the floor of the lake sank, so that at least 1,000 feet of conglomerate accumulated over the southern part of the Colony ; the water stretched farther and farther north as time went on, so that at the close of the glacial period shales were being deposited at least as far north as the Kalahari Desert.

The thousand feet of mud and stones which must extend over thousands of square miles under the southern part of the Karroo, and formerly spread as far south as Worcester, and very probably farther than the present southern limit of the continent, represent the products of denudation of a large tract of country during a long time. The wide distribution of the striated blocks and pebbles, which are found wherever the outcrops are sufl&ciently good to allow one to obtain the contained boulders, shows that the glaciation was no merely local phenomenon, to be likened to the very limited snow and ice covered areas within tropical Africa at the present day, but that it

THE GEOLOGICAL HISTORY OF THE COLONY 403

was a wide-spread glaciation, extending over a large part of the continent north of the Karroo area. The source of the Dwyka boulders has been described in an earlier chapter, and we found that though the source of many is at present unknown, yet a sufl&cient number have been recognised as having come from the Pre-Cape rocks north of the Karroo to show that the main source of the Dwyka series, so far as the Colonial area is concerned, lay to the north; the evidence hitherto noticed of the movement of the ice in the northern districts is to the same effect, i.e.,

m

that the ice moved southwards from those districts. Whether land to the south ajso contributed ice-borne debris is unknown, but at least at a certain stage of the period another source lay to the west, as shown by the striated pavements in the western Karroo.

The evidence for the glacial origin of the boulder beds at the base of the Gondwana system, and in the lower part of the beds containing the Glossopteris flora in Australia, does not seem to one who has not seen it himself to be so strong as that for the glacial origin of the Dwyka conglomerate and the scratched surfaces below it in the northern parts of Cape Colony, but the testimony of so many geologists who have seen the Indian and Australian rocks, and who are agreed that the striated boulders found in them owe their form to glacial action and their position to carriage by ice, cannot easily be set aside. The very fact of the occurrence of such phenomena at the base of the beds containing the Glossopteris flora in those far dis- tant lands, in a precisely analogous position to that

26*

404 GEOLOGY OF CAPE COLONY

of the Dwyka conglomerate in South Africa, is itself presumptive evidence that all the peculiar character- istics had a common cause, and no agency save glacial conditions can be put forward to explain the appear- ances in the Dwyka series described in a previous chapter. The explanation of this glaciation is not yet clear. Penck has examined the supposition of an altered position of the earth's axis with one pole in the middle of the Indian Ocean and the other in North- West Mexico, but the recorded direction of the move- ment of the ice in South Africa, India and Australia are not in agreement with such a position of the South Pole, and there is no evidence of corresponding glacial conditions in the American Continent; in addition to this astronomers seem to be agreed that such a change in position of the axis of rotation (some 66°) is quite out of the question, at any rate since the birth of the moon, which would take us back to a period far more ancient than the one we are now dealing with.

It is difficult to find a sufficient cause of the glacia- tion in the supposed existence of a tract of very high mountainous country to the north of Cape Colony, for the ice certainly reached the shores of the water in which the southern conglomerate was laid down, and unless we have good reason to believe that this water- level was at a great height above sea-level, which is not probable, the ice must have passed into the Karroo water at a level that was not very much above that of the sea. In addition to the objection to the existence of very high land north of Cape Colony, the widespread distribution of the boulder beds in Africa, India and

TbE GEOLOGICAL HISTORY OF THE COLONY 405

Australia, demands, as Penck points out in the paper referred to above, a correspondingly widespread cause, and the existence of a greatly elevated country of such extent is at least improbable. Whatever may have been the cause of the cold climate, the fact of its having pre- vailed is as certain to the mind of a geologist who has seen the Dwyka conglomerate, and the underlying striated floors in the north of the Colony, as the former presence of man is to the person who picks up potsherds on a sandhill or sees figures of men and beasts rudely painted on the wall of a cave.

Although it can hardly be maintained that the fossil evidence in South Africa, India and Australia is as yet sufficiently strong to prove the contemporaneity of the boulder beds in the three continents, for that would re- quire much longer series of fossiliferous strata in those localities than have been found, yet so far as the facts go they undoubtedly give us very good grounds for as- suming that the boulder beds were formed at about the same period. There seems to have been in late Palaeo- zoic times a great mass of land, whose boundaries are very imperfectly known, but which included part of Africa to the north of the Colony, a part of Australia, and a part of India, and which stretched across the Indian Ocean ; on this land glacial conditions prevailed during a certain period. The flora and fauna of the land during and subsequent to the cold period was quite different to those which spread over the European and North American areas at the same time, for only a very few of the typical Karroo and Gondwana forms have been found in those regions. Some of the products of

406 GEOLOGY Ot OAtfe COLONY

the denudation of this ancient continent — Gondwana- land — accumulated in great fresh water lakes, of which the Karroo area of South Africa is one. It is useless at the present time attempting to fill in the details of the history of the sediments derived from Gondwana- land ; to discover, for instance, how many fresh water basins existed, and to what extent they communicated with each other and with the ocean. In South Africa all the fossils yet found in these sediments lived upon land or were fresh-water forms, no distinctly marine animals are amongst them. In New South Wales, on the other hand, a striated boulder bed has been found in strata containing marine fossils of Upper Carbon- iferous types. Whether any evidence of an encroach- ment of the ocean upon the Karroo lake exists in South Africa remains to be discovered.

On the African portion of Gondwanaland at first grew Glossopteris and its associates mentioned on a pre- vious page ; and soon there appeared the remarkable reptiles, of which Pareiasaurus was one of the earlier and larger forms. Pareiasaurus and Dicynodon were certainly vegetable feeders, but carnivorous beasts were by no means wanting,' a glance at the formidable teeth in such an animal as Titanosuchus is sufficient to convince any one that their possessors lived upon their fellows and did not graze on the Glossopteris and other plants that covered the ground. The bones now found in the Karroo belonged to bodies that were carried down by rivers or drifted from the shores of the lake.

The Karroo area, and with it probably the whole of the folded belt, must have sunk to allow the accumula-

THE GEOLOGICAL HISTORY OF THE COLONY 407

tion of the thousands of feet of shallow water deposits that we see in the Karroo formation. Occasionally perhaps wide stretches of mud or sand lay exposed for a time above the water, to be submerged again and buried under similar sediment. Such flat islands can now be recognised where the slight unconformities in the Ecca and Beaufort series mentioned in chapter v. are found.

The duration of this slow depression was unequal in different parts of the Colony; it was less along the

B

southern and south-western area, where the Cape formation was thickest, than to the north. Probably while the upper part of the Beaufort series was being laid down, the folding began that eventually produced the great southern mountains. It is not yet known exactly when this process began, or when it reached its maximum, but there is little doubt that it was in progress during the later portion of the Karroo period. The numberless places along the southern edge of the Karroo where the lower Karroo beds can be seen rest- ing conformably upon and involved in the folds that affect the Witteberg, Bokkeveld and Table Mountain series as well as the occurrence of the Dwyka and Ecca beds at Worcester, and the outliers on the northern edge and in the heart of the folded belt, prove con- clusively that the main part of the disturbances took place after the Ecca beds were deposited. The Uiten- hage beds, lying comparatively undisturbed upon the contorted strata belonging to the Cape formation, and in places upon the Pre-Cape rocks, give the clearest evidence for believing not only that the earth-movements

408 ({EOLOGY OP OAPE COLONY

responsible for the mountain chains had done their work before these beds were formed, but also that a tremen- dous amount of rock had been removed from the folded belt before that time. We have seen in earlier chapters that the Dwyka and Ecca beds belong to the later part of the Palaeozoic era, to the period for convenience called Permo-Carboniferous, and that the Uitenhage beds are of early Cretaceous age. It was during the interval between those roughly defined periods that the mountain building in Cape Colony went on. In other countries this interval is represented by the Triassic and Jurassic systems, but in South Africa the only beds that can be referred to either of these are the Beaufort and Stormberg series, and they belong to the Trias rather than to the later stage.

The southern folding seems to have been produced by a thrust from the south towards the north, for the folds, where not symmetrical, tend to turn over towards the north. The mii:ior ranges, such as the Caledon Mountain, Warm Water Berg and Touw*s Berg are symmetrical, both limbs of the anticlines are equally inclined, and the same is the case with Anysberg, the western end of the Table Mountain sandstone ridge of the Zwartebergen ; but in the high ranges, the main portion of the Zwartebergen and the Langebergen, the folds lean over northwards, so that both limbs of any one fold dip southwards. This structure seems to indicate that the region of the Great Karroo acted as an immovable block against which the strata of the folded belt were crumpled and turned over. The over- thrust faults in the Dwyka series near Laingsburg are

THE GEOLOGICAL HLSTORV OP ThE COLONY 409

also directed towards the north, as though the pressure had to be reheved by the sUding of blocks of beds over the fractured edges of the next block to the north. It is on the south flanks of the most crumpled ranges that the great strike faults of Worcester and the Cango occur, and their downthrows are very considerable, reaching at least 10,000 feet at Worcester. The western folds are not nearly so intense as the southern, and may have begun at an earlier date. The easternmost of these anticlines, that which forms the Cederbergen, is also the greatest, and it is fairly symmetrical ; no consider- able folds lie parallel to it on the east; to the west, however, there are several parallel folds decreasing rapidly in amplitude towards the coast.

The neighbourhood of Worcester, where the Uitenhage conglomerates lie upon the Ecca beds and the Pre-Cape rocks on either side of the great fault, affords a grand object-lesson in denudation. To the north of the fault the conglomerates lie directly upon the Malmesbury beds ; to the south, part of the Ecca, the Dwyka series and the whole of the Cape formation intervene between the two. The thickness of the intervening rock is not less than 10,000 feet. Between the fault and the mountains to the north of it over 10,000 feet of rock must have been removed during the interval (Jura-Trias) spoken of above. Nowhere else in the Colony is the evidence of this denudation so clear as at Worcester, but with it before us we can believe that a similar amount of rock was removed from the Pre-Cape areas of Mossel Bay and the Cango, which are now partly overlain by the Uitenhage conglomerates. It must not be forgotten,

410 ^EOLCKiY OF CAPE COLONY

however, that the Worcester conglomerates may be somewhat later in age than the similar rocks at Enon and Uitenhage, but the difiference is certainly small.

When describing the dolerite intrusions of the Karroo we noticed that they seem to have avoided the folded belt ; they occur to the north of it and on its extreme limits, where the intensity of folding is much less than in the major portion of the belt ; we noticed also that this peculiarity in the distribution of the dolerite pointed to the folds having been in existence or in progress when the dolerite was intruded. Now the dolerite is probably of late Stormberg age, for the points of resemblance to the dolerites which are found in the volcanic beds are so numerous, and at the same time of more importance than the differences between them, that it seems that both the general mass of dolerites in the central and eastern parts of the Colony and the distinctly volcanic rocks belong to one series and reached their present position at about the same time, the end of the Stormberg period. The only other direct evidence of the age of the dolerites at present known is the occurrence of the rock as boulders in the Embotyi conglomerates, which we must regard as of Cretaceous age but probably younger than the Uitenhage beds. This fixes a later limit to the age of the intrusions. If the Embotyi beds should eventually prove to be of Uitenhage age the limit will be correspondingly set back. But the first argument, concerning the connection of the dolerites and volcanic beds, certainly supports the assumption that the intrusions took place at the close of the Stormberg period, and this helps us to determine

THE GEOLOGICAL HISTORY OF THE COLONY 411

the date of the folding in the southern mountainous region.

Whether closer limits can be set to the period of folding than the Ecca and Uitenhage periods remains to be found out in the future. It is possible that the unconformity near Aberdeen, between the Ecca and Beaufort beds, described by the late Professor A. H. Green may* be more than a local phenomenon, and if so it may lend material aid to the solution of the ques- tion, but so far as our knowledge of other parts of the Colony goes there is no physical break at that horizon. It may be that all traces of the unconformity which probably existed within the Karroo formation some- where to the south of the main Colonial watershed iiave been removed by denudation. The uprising of the folded belt exposed the southern parts of the Colony to the air and to all the destructive agencies, such as change of temperature, wind, rain and streams, that this entailed. There were then formed the great longitudinal depressions between the Zwartebergen and Langebergen, and the other more or less east and west ranges in the south. To this period probably belongs also the first rough shaping of the western coastal districts, the removal of the upper parts of the Cape formation from Malmesbury, Piquetberg and neighbouring districts, and the Olifant's River Valley (Clanwilliam). While this was going on the upper parts of the Karroo formation were being laid down in the north-east, possibly also far to the north and north-west of the existing boundary of the Stormberg series. The time represented by these rocks witnessed

412 (iEOLOGY OF CAPE COLONY

a great change in the inhabitants of the land round the Karroo lake. Glossopteris and many of its fellow plants of the earlier period died out and were replaced by a new vegetation, of which Thinnfeldia, Taniopteris, Baiera and Callipteridium are the best-known members. The fauna likewise changed, Pareiasaurusy Dicynodon and their allies disappeared to make way for more highly organised reptiles.

Below the Molteno beds there is no direct evidence of a diminution in size of the water basin in which the Karoo formation was deposited, but the coarse sandstones in the Molteno beds and the overlying strata, the coal seams and the occasional thin conglomerates in the Molteno beds all point to the proximity of land during their deposition. It is not yet possible to define the position of the neighbouring land, but it is probable that part of it lay to the south-east of the Drakensberg ridge.^

The present main watershed of the Colony was prob- ablj^ produced during the Stormberg period by the rising of a low tract of country from the Karroo area, trending in a north-easterly direction. The water which fell on this land drained off towards the north and south, giving rise to the chief rivers draining what are now the Great and Upper Karroos. It is as yet difficult to account for the appearance of this land, for there is now no sign of an anticlinal ridge corresponding in direction with the main watershed. On the other hand, the structure of that area is that of a very gentle syncline. It is possible

1 Schwarz (03).

THE GEOLOGICAL HISTORY OF THE COLONY 413

that the synclinal structure was given to the country at a somewhat later period. The intrusion of the dolerite sheets, which on the average reach a thickness of per- haps 1,000 feet over a wide area in the central districts of the Colony, must have produced some effect at the surface, and may have been the cause of the emergence of the watershed.^ The varying position of the base of the Stormberg volcanic series proves that the Cave sand- stone was subjected to denudation before the volcanic outbursts commenced, but there is also evidence in the interbedding of the two groups of rock that the denuda- tion was local, and that the Cave sandstone continued to be formed after the earliest activity of the volca- noes. The outpouring of the immense thickness of lava, described in. a previous chapter, put a stop to the deposition of ordinary sediments, and the conditions under which sandstones, shales and other sediments are formed seem never to have prevailed again in the interior of the Colony. The present state of our know- ledge of the volcanic series is too imperfect to allow a satisfactory statement of the effects due to the volcanic episode to be made ; but it may be taken as certain that one result was to add a great volcanic pile to the north- eastern end of the newly emerged land.

There seems no escape from the conclusion that denu- dation has proceeded uninterruptedly from the close of the Stormberg period (Ehaetic) to the present day in the interior of the Colony. No deposits of later age,

^ For discussions on the origin of the watershed see Schwarz, The Volcanoes of Griqualand East (03), and Rogers, The Geological History of the Gouritz River System (03).

414 GEOLOGY OF CAPE COLONY

other than river gravels, alluvium and sand, have been found north of the folded belt.

During the Jurassic period the valleys in the folded belt were greatly enlarged and deepened, so that the Pre-Cape rocks became exposed in several areas south of the Zwartebergen, where the chief rivers appear to have had east and west courses. The rivers running south from the main Colonial watershed have left no trace of their passage through the mountain ranges in pre-Uitenhage times, though the valleys excavated in the folded belt before the deposition of the Uitenhage beds were deeper, relatively to the ranges, than the modern ones in the same districts. The water flowing southwards from the main watershed probably drained away to the sea in an easterly direction. This great period of denudation received a partial check in early Cretaceous times, so that the longitudinal valleys in the folded belt became filled with conglomerates, sandstones and shales, now represented by the outliers of Uitenhage beds described in chapter viii. The cause of this may have been twofold, first, the sinking of the land, and, secondly, the coming in of a drier climate. That the former cause played an important part is evident from the fact that the marine Sunday's Eiver beds occupy an old valley between Port Elizabeth and the Zuurbergen ; and the second of the two causes is indicated by the nature of much of the conglomerates and sands, the Enon type of the Uitenhage beds. It is not improbable that the Uitenhage beds eventually covered the whole of the folded belt, with the exception of parts of the mountain ranges; there is reason to believe that in

THE GEOLOGICAL HISTORY OF THE COLONY 415

places the lower passes in those ranges were buried under the gravels and other rocks of the Uitenhage series. As we saw in chapter viii., there is no evidence to show how far west of the Zwartkops Valley the marine beds extended.

Towards the end of the Uitenhage period we may sup- pose that a low belt of land stretched north-east through the middle of the Colony, ending in a great mass of volcanic rocks, and that to the south of this land there were ridges of mountainous ground projecting above a shallow sea, or through gravel and sand deposited by local streams in a flat country only partially under water. Whether these sediments, in whatever way they were formed, spread north of the Zwartebergen will perhaps never be known, but it is possible that they did so, and that the streams flowing southwards from the main watershed eventually delivered their loads of silt into the same area instead of reaching the open sea to the south-east. It appears to be probable, however, that the rivers ran southwards across the newly deposited Uitenhage beds when the uplift oc- curred which put an end to the deposition of those beds in the folded belt.

The course of events north of the watershed, in the country drained by the Orange Eiver, is much more difl&cult to decipher, and at present too little is known of the details of its geology to allow one to attempt the task. The mountain building which produced the southern and western ranges did not affect the north of the Colony, and no equivalents of the Uitenhage beds are known to exist in that region. Probably the

416 GEOLOGY OF CAPE COLONY

Orange Biver cominenced its work at the same time as the streams flowing south from the main watershed.

At some time after the deposition of the Uitenhage beds earth-movements took place in their area, and the effects of these are seen in the inch'nation of the Uiten- hage strata, and in the faults that traverse them. These movements had partly the same direction as the earlier ones that produced the mountains, but there is no evidence yet discovered that proves the new move- ments to have always followed the older very closely. For example, the Worcester fault, and the southern boundary fault of the Cango district do not appreci- ably affect th'fe Uitenhage beds, though the latter have slight dips in the neighbourhood of those great disloca- tions. Along parts of the faults the Uitenhage beds lie comparatively undisturbed on the old surface on both sides of the fault without any indication of faulting along the same line. In Uniondale and Willowmore, on the contrary, conglomerates of the Enon type have been let down against the Cape formation along faults that are parallel with the strike of the folds produced in the latter formation in Pre-Uitenhage times ; the downthrow is always to the south, as in the case of the older faults. The effect of the faulting and folding of the Uitenhage beds must have been to accentuate former longitudinal valleys, if they were in existence, or to give rise to them. The extent to which the dislocations were car- ried was, however, insufficient to disarrange the already established southward courses of the rivers draining the Karroo. These rivers gradually cut down their val- leys through the Uitenhage beds, so that they reached

THE GEOLOGICAL HISTORY OF THE COLONY 417

the underlying sharply folded Cape formation, a process that still continues ; but, as we shall see, there have been periods of diminished downward erosion during which the rivers widened their valleys and cut exten- sive plains instead of deepening their channels. The river systems south of the main watershed thus de- veloped on a country whose structure has no direct relationship to the origin of the main rivers, and the deep gorges of the transverse streams, such as the Gamka and Gouritz, were sawn through by the rivers cutting their way downwards through soft and hard rocks alike as they were exposed. It is not improb- able that the earth-movements of post-Uitenhage age deepened the depression between the Zwartebergen and Langebergen, but the movements were greater in some places than others, and were not sufficiently regular in direction and extent to deflect the chief transverse stream into valleys parallel to the mountain ranges.

In the marine beds of the Uitenhage series we have the inshore deposits of an ocean that stretched from India to South Africa, but its general form is very im- perfectly known. So far as South Africa is concerned that ocean probably only touched the country and never spread over what is now the interior of the Colony. The next inroad of the open sea is recorded in the Umzamba beds of the south-east coast. The fossils in these rocks are most closely related to Indian forms, and indicate that the beds were laid down at a later stage of the Cretaceous period than the Sunday's Eiver

beds. The fact that the Umzamba and the Embotyi

27

418 GEOLOGY OF CAPE COLONY

beds are faulted down against the Table Mountain series shows that they once extended farther inland than their remnants are found to-day, but as they are distinctly in-shore deposits, as opposed to those formed under deep-sea conditions, they probably never stretched far inland beyond their outcrops in Pondoland. We have as yet no sign of a passage from the Uitenhage into the Umzamba series, but negative evidence on such a point is worthless under the circumstances ; we cannot, therefore, say whether the ocean retreated and returned, or whether a conformable group of beds, from the Uitenhage to the Umzamba series, once existed in or near the south-east limit of the Colony.

The boundary faults of the Pondoland Cretaceous rocks were evidently formed in post-Cretaceous times, and they appear to have no connection with the earth- movements that affected the Uitenhage beds of the west. In direction (north-east) they agree more closely with the line of volcanic vents in the Drakensberg, so far as the latter is known, than with the nearly east and west, or east-south-east flexures into which the Uitenhage beds were thrown. The Pondoland faults are approxi- mately parallel to the coast, and were probably closely connected with the formation of that part of the South African coast line. So far as our information carries us at present this is the only part of the Colonial coast defined by faults.

At some time subsequently to the deposition of the Uitenhage beds volcanic explosions took place at various spots from Spiegel River in the south to Griqualand West in the north, and the chief products of this third

THE GEOLOGICAL HISTORY OF THE COLONY 419

phase ^ of volcanic activity in the Colony were melilite- basalts and peculiar breccias ; the latter include the blue-ground of the Kimberley and other pipes. These volcanoes seem to be distributed sporadically without any relationship to the earlier established structural lines in the Colony, and they do not appear to have thrown out any considerable quantity of lavas or ashes.

Eeturning now to the southern rivers, which we described as cutting down their valleys through the Uitenhage beds to the underlying rocks, we must en- deavour to trace the events which have taken place since they began to cut through the partially buried mountain ranges. Throughout the southern districts of the Colony there is abundant evidence bearing on this portion of its history ; this evidence is given by the gravel and alluvial terraces lying high above the bed of the modern rivers. The absence of contoured maps and a close knowledge of the height of the terraces greatly increase the difficulties of the problem, and it would be useless to attempt more than a sketch of the main con- clusions to be derived from the facts at present known. The oldest, or highest, well-developed terraces lie over 1,000 feet above the modern river beds, and there may be still more elevated terraces. When the rivers from the Karroo flowed at levels about 1,000 feet higher than at present their downward cutting powers were checked, and they, together with their tributaries, planed off the country to a more or less common level, producing a

1 The other two gave rise to (1) the Pre-Gape voloanic rocks of Prieska and Griqualand West, and (2) the Stormberg volcanic rocks. The former, however, may represent more than one period of activity.

27*

420 GEOLOGY OF CAPE COLONY

slightly undulating plain, from which rose the long mountain ranges and the smaller anticlines, such as the Caledon Mountain and Warm Water Berg, The dis- tinct terraces forming conspicuous features on the slopes of the Langebergen and Zwartebergen, where the hard folded quart zites are cut to a nearly flat surface, were made during this period of lateral erosion. The cause was widespread, for we find its effects from the Transkei to the Zwart Euggens west of the Karroo. An obvious possible reason for the cutting of these terraces is that the sea stood higher relatively to the land than is the case to-day, but whether the rivers had previously cut their channels down to sea level and so were unable to further deepen their valleys, or whether the country as a whole sank and therefore checked the deepening of the valleys, is not certain. The former is perhaps the more likely, for no fiUed-up channels near the coast have been found, and they might be expected had the valleys been eroded to a greater depth than could be maintained under the new conditions. Local plains might well have been formed behind the larger blocks of mountains, just as we see wide alluvial flats in the course of the Buffel's Eiver behind the Leeuw Kloof Poort and the plain cut by the Olifant's Eiver before entering its gorge in the Gamka Hills. The terraces on the flanks of the mountains, with the outlying table-shaped fragments of the plains that formerly connected them, are so exten- sively developed, both to the north and south of the Zwartebergen, that they cannot be explained by a local cause like that which is sufficient to account for the alluvial flats of the Olifant's Eiver. It is probable

ttifi GEOLOGICAL HLSTORY OP THE COLONY 42l

that the high-level plateau in the country south of the Langebergen was formed at the same time as the ter- races we have been discussing, although it lies at a lower level, for then, as now, the rivers must have had a fall towards the coast, and each gorge through the great ranges was perhaps more steeply graded than the valley-bottom above or below it.

The rising of the country relatively to the sea-level renewed the downward cutting powers of the rivers and restricted the stream erosion within narrow limits, so that great parts of the old plains were permanently abandoned. The change in the drainage system thus effected was considerable in certain areas ; the water- shed between the Olifant's and Baviaan's Kloof Eivers now hes on a high level gravel plateau, and before the platform was cut the watershed may have been far from its present position.

At the present time the Dwyka and Gamka traverse the Zwartebergen together by the Gamka Poort, and at a point thirty miles below that gorge the Buffel's Eiver joins them to form the Gouritz Eiver ; to do this the Buffers Eiver turns sharply to the east, away from what one would suppose its proper course to have been ; the depression on the crest of the Langebergen, called Garcia's Pass, lies directly in the supposed normal course of the river, so it is not unlikely that its upper part was captured by a western tributary of the Gouritz. This must, however, have happened before the high- level plain was cut, for the summit of Garcia's Pass lies higher than the terrace on the north flank of the Langebergen.

422 GEOLOGY OF CAPE COLONY

There are other terraces at lower levels than those mentioned above, but to bring all these into order and to place them in chronological sequence is at present impossible on account of the lack of detailed information as to their relative heights and distribution.

The raised beaches and the limestone formed from sand dunes now lying at considerable heights above the shore and at some distance inland, must belong to one of the periods of plain-cutting ; there is as yet insuffi- cient evidence to correlate these phenomena definitely, but it may well be that the higher shore terraces, such as that covered by the marine gravels of the Zwartkops Heights, and the old beach underlying the limestone of Cape Infanta, were removed out of reach of the waves when the inland plateaux, 1,000 feet above the modern river beds, were elevated.

The numerous S-curves of the southern river valleys with precipitous sides, often several hundred feet high, are relics of the time when the streams meandered slowly across nearly flat plains ; on the fall of the streams being increased by the rise of the land their downward cutting power was renewed and they deepened the valleys in which they flowed, so that in many cases the S-bends were retained and deepened to the extent we now see. One of the most remarkable of these is the S-shaped gorge in the Klein Zwartberg occupied by the Buffers Eiver; another has been cut by the Gamka between the Boode Berg and the Pogha Hills ; from the Eastern Province the extremely sinuous and deep valleys of the Great Fish, Kei and Bashee Eivers are analogous examples.

THE GEOLOGICAL HISTORY OP THE COLONY 423

Hitherto we have only given the evidence for elevation of the land relatively to the water in connection with the superficial deposits. There is, however, some in^ dication of recent depression ; the great depths of the estuarine shelly sands and muds near the mouths of some of the rivers, especially the Zwartkops and Buffalo Rivers which are the only ones that have been explored in this sense, may be due to depression. On the west coast, Saldanha Bay, an almost land-locked basin in granite, appears to be a drowned valley. There is no well-defined valley entering the bay, though the thick superficial sandy deposits that stretch south-east of the bay may conceal an old river channel. At many places in Saldanha Bay the dune hmestone containing the re- mains of land snails passes below sea-level^ as is also the case near Struys Point and the mouth of the Duiven Hoeks River on the south coast. At Paternoster, north of Saldanha Bay, a well sunk at a spot about twenty feet above sea-level revealed the presence of ninety feet of sandy limestone and sand containing land shells, tortoise bones, and broken marine shells, evidently an accumulation formed on the land behind the beach, and not below tide-level. These facts all point to a recent depression.

Throughout this account of the changes of level which have affected the Colony the expressions ** upUft " and ** depression,*' or equivalent terms have been used. It is, however, one of the obscure problems of geology to find out whether apparent uplifts and depressions are due to the movement of the land or to that of the sur- rounding ocean. Where the strata concerned are bent,

424 GEOLOGY OP CAPE COLONY

the changes must at least in part be due to the move- ments of the rocky crust of the earth; but where a widespread alteration of relative level has taken place, such as that which caused the abandonment of the high- level plateaux by the streams which once flowed across them, the question is not easy to decide. Should it be found that terraces or raised beaches that were once on the same level are now at different heights above the sea, then earth movements must have played a part in bringing about the change. The evidence to decide even this detail with regard to the Cape terraces and beaches has not yet been collected, and it is not such a simple matter as it may appear. The recent deposits as a whole are remarkably deficient in organic remains, though up to the present time they have not been systematically searched ; and it is only possible to de- termine the contemporaneity of detached portions of terraces and beaches by a study of their fossils. A thorough investigation of the facts bearing upon the past changes in level in the Colony will add much to the materials for the decision of the problem.

CHAPTEE XII.

NOTES ON THE GEOLOGY OP SOME OF THE RAILWAY

LINES.

In this chapter it is proposed to point out the chief points of interest to be seen along the railway lines, or rather those portions of them about which I have definite information.

The Westeen Main Line to the Orange Eiver.

From Cape Town the line passes through the Cape Flats as far as Durban Eoad Station. On either side of the line blown sands overlying ironstone clays and sands stretch for several miles. As a result of the systematic planting pursued by the Government the drifting sands in this area have been checked, and wattle thickets cover what used to be one of the heaviest parts of the country for travellers before the railway was made. The only outcrops of hard rock in the Cape Flats are the surface-quartzites, patches of which lie close to the line about ten miles from Cape Town. At Durban Eoad the line enters a country which is still flat, but the Malmesbury beds and the intrusive granite lie close under the surface soil. To the north-west are the Tyger Berg and the hills near Durbanville formed of slaty and quartzitic beds be-

425

426 GEOLOGY OF CAtE COLONY

longing to -the Malmesbury group ; to the south-east lie Eanon Kop and Bottelary, composed of granite, while farther oflF rise the great Table Mountain sand- stone mountains of Stellenbosch and Hottentot's Hol- land. Between Mulder's Vley and Klapmuts the line traverses a north-westerly fault, on the south-west side of which outliers of the Table Mountain series have been let down against the Malmesbury beds and form Klapmuts Hill and Joosten Berg.

At Paarl the railway turns northwards and runs parallel to the Klein Drakensteins (Table Mountain sandstone) and the ranges north of them, down the val- ley of the Berg River, with the great granite Mountains of Paarl and Paarde Berg to the west. The isolated mountain lying about eight miles west of the railway between Hermon and Porterville Eoad is Eiebeek's Kasteel, an outlier of the Table Mountain sandstone. The mountain ridge extending north from the Draken- steins is part of the western limit of the folded belt. The Cape formation east of the range is considerably folded and faulted, but to the west it is but slightly folded and over large areas it has been removed by denudation. The country so far described belongs to the Pre-Cape region. Just beyond Porterville Eoad the line crosses the mountains by the New Kloof and enters the great depression drained by the Klein Berg and Breede Eivers, and in which lie the agricultural districts of Tulbagh and Worcester. The watershed between the two rivers is a scarcely noticable rise near Ceres Eoad Station, but to the north the water flows through the New Kloof, while to the south the Breede

GEOLOGY OF SOME OF THE RAILWAY LINES 42?

Eiver, after a long journey through comparatively flat ground, traverses a high range of Witteberg hills on its way to the sea at Port Beaufort. East of the New Kloof the line runs south to Worcester, where it turns north-east to traverse the Hex Eiver Mountains; for some thirty-five miles it lies between two ranges of mountains, the Witzenbergen and their southern con- tinuations on the east, and the Drakenstein Slang Hoek mass on the west. The Table Mountain sandstone of both these is seen to lie apparently horizontally when viewed from the railway, and obviously once stretched across the valley, covering the Malmesbury beds forming the low ground. When these mountains are examined closely, however, the strata are found to dip at various angles away from the valley, and to present their edges towards it. South of the Breede Eiver Station the sandstone to the south-west of the line takes on a different dip, towards the valley instead of away from it. This becomes more and more marked towards the corner of the great mountain mass at Brand Vley, and is one of the phenomena connected with the Worcester fault. The wide area of gravels along the Breede Eiver and the Uitenhage conglomerates (Enon type) to the east conceal large portions of the underlying Malmes- bury, Cape and Karroo formations near the fault. The railway passes over the fault twice, once about four miles west of Worcester Station and again two miles north-east of the station. It makes no feature at the surface, and its presence is only indicated along this part of its course by outcrops of Ecca beds in contact with the Malmesbury series. From Worcester some

428 GEOLOGY OF CAPE COLONY

interesting excursions can be made. To the south the greater part of the succession from the Table Mountain sandstone to the Ecca beds is well exposed east and north-east of the mountain between Brand Vley and Stettin's Berg ; at Waai Kloof, twelve miles east from the town, the unconformable junction of the Table Mountain sandstone with the ottrelite schists and quartzites of the Malmesbury group can be seen ; while to the north there are numerous exposures of slates, schists, together with gneiss, and other igneous rocks of Pre - Cape age. The railway crosses the second range of mountains by the Hex Eiver Valley, which is situated just to the east of the bend or angle formed by the meeting of the north and east trending ranges. On emerging from the Poort the Hex Eiver Mountains lie to the north and are admirably displayed to a traveller by train. The railway is carried up the left side of the valley, along the V-shaped synclinal area of Bokkeveld beds. Between De Dooms and Touw's Kiver some of the richest localities for Bokke- veld fossils are passed, and for the greater part of the way the line is laid on the lower divisions of the Bokke- veld beds which alone contain marine fossils. The best localities for searching for fossils are near De Dooms, the quarries at Tunnel Siding, and Klein Straat. A short distance beyond Klein Straat a fault with down- throw to the north is crossed ; it bounds the northern face of the eastern spur of the Hex Eiver Mountains, and along it the Witteberg beds are brought down against the Table Mountain sandstone. Near Klein Straat an isolated anticline of Table Mountain sand-

GEOLOGY OF SOME OF THE RAILWAY LINES 429

stone, Baviaan's Berg can be seen to the south ; it rises from a rather flat country, the Touw's Vlakte, cut out of Bokkeveld beds. The Witteberg series forms the high hills to the north and east. Some of the remark- ably sharp folds in the Witteberg beds can be seen from the railway three miles on the up side of Touw's Eiver Station. The prominent bands of rock are quartzites, and the intervening shales, darker in colour than the quartzites, have weathered away more rapidly, leaving the quartzites standing out on the sides of the hills. Six miles on the down side of Touw's Kiver the line enters the synclinal outlier of Dwyka conglomerate of Quarrie Kloof. The Witteberg quartzites dip under the outlier and form the bare precipitous hills to the north and south ; the conglomerate crops out near the railway line in irregular lumpy masses showing the rough cleavage or slab-structure which is characteristic of that rock along the south of the Karroo. After journeying some ten miles on the conglomerate we cross to the Witteberg beds again, but this formation is finally left near Pieter Meintjes, where we enter the main area of the Dwyka series. The dark cliffs to the north of the line between this station and Matjes Fontein show the rather feebly developed stratification planes in the conglomerate, and the kopjes nearer the railway are good examples of the usual aspect of the conglom- erate south of the Karroo. The included boulders are often large enough to be seen from the passing train, and the slab- structure producing the characteristic pil- low-form of the exposed surfaces is prominent. About half a mile south of the line at Matjes Fontein there

430 GEOLOGY OF CAPE COLONY

are three white quartzite kopjes formed by lenticular deposits of that rock within the conglomerate. Matjes Fontein is a good centre for an examination of the conglomerate. Numerous striated boulders may be ob- tained from the rock in the hills north of the station, where good exposures are numberless ; the variety of rocks forming the boulders is also very great in this locality. To the east-north-east of the village the suc- cession from the conglomerate to the bottom of the Ecca series is exposed on the steep sides of a high hill. South of Matjes Fontein and for some twelve miles on the way to Laingsburg the steep, bare dip-slopes of the uppermost quartzites of the Witteberg formation bound the view ; at places high up on these mountains (the Wittebergen) the strata appear to be lying horizontally upon the steeply dipping beds of the lower slopes, an appearance due to the sharp bending of the beds and the removal of the outer part of the bend by denudation ; on ascending the range from Matjes Fontein its struc- ture becomes obvious. The line leaves the Dwyka series about nine miles from Laingsburg and enters the great area of Ecca beds, the sandstones and shales of which are exposed in the railway cuttings and on the bare hills on either side of the line. Near Laingsburg the most prominent ranges of hills are formed by the middle por- tion of the Ecca series, called the Laingsburg beds. Just before reaching the station the line crosses one of the rivers which drain the Karroo region, the Buffers Eiver from the Moordenaar*s Karroo. This river, which usually has only isolated pools of water in its bed, passes through the Zwartebergen by means of a great gorge

GEOLOGY OF SOME OF THE RAILWAY LINES 431

about six miles in length with vertical sides. It is well worth making a journey from Laingsburg to within the great poort in order to see it. Laingsburg is a con- venient place for excursions to the Zwartebergen and the Dwyka and Ecca beds. A walk from the village up the Buffel's Eiver towards the Moordenaar's Karroo illustrates admirably the passage from the folded belt to the little-disturbed interior basin; the strata are thrown into extremely sharp folds and are overthrust towards the north ^ at places near the village ; farther up the river, which traverses the beds at right angles to their strike, the folds die out rapidly and at a distance of about eight miles the strata lie nearly horizontally, a condition that is maintained for hundreds of miles northwards, with the exception of a few small monoclinal folds south of the Komsberg.

From Laingsburg the train passes along the northern- most portion of the folded belt for some forty miles, but before Prince Albert Boad is reached the folds are no longer seen and the strata everywhere lie at very low angles. The Great Karroo is entered at Laingsburg ; the almost bare hills of shale and thin sandstones and the scanty vegetation, consisting of small bushes which only look green after good rains, are characteristic of thousands of square miles from Karroo Poort in the west to Somerset East, and from the main watershed

^ About one and a half miles from the village on the road to Zout Kloof there is a clearly exposed section showing the chert band re- peated three times by overthrust faults, and near by the lower part of the Upper Dwyka shales are thrust over the higher portion containing the white band.

432 GEOLOGY OF CAPE COLONY

on the north to the Zwartebergen in the south. The thorn trees along the river beds are the only green things usually visible from the train in this area. The various forms of kopjes due to the action of rain and wind on thin sandstones and shales dipping at different angles are well displayed along the line from Laings- burg to Beaufort West. On the up side of Prince Albert Boad the ridge, or hog back, type is the usual one, owing to the inclined position of the strata, but when we reach the almost horizontal beds, low table- shaped hills with steep sides are the predominating forms. The table-shape is due entirely to the weather acting on horizontal beds, the thin but hard sandstones check the destructive process and give rise to flat caps to the hills. These flat-topped hills are very different in nature from the somewhat similarly shaped hills seen along the line from Swellendam to Biversdale ; the latter are parts of a stream-cut plain isolated by the renewed activity of the streams owing to the elevation of the whole country. Southwards from the railway the great range of the Zwartebergen towers 5,000 feet above the lower portions of the Karroo. The various ranges of foothills can be distinguished from certain points on the line on favourable days. One great gash in the range several miles east of its highest point (Seven Weeks' Poort Mountain) marks the passage of the Gamka, whose two chief feeders are bridged at Bloed Biver Siding (Dwyka) and near Fraserburg Boad. The traveller will rarely have the opportunity of seeing any water in either of these river beds, and he may be sceptical as to the power of their temporary streams to

GEOLOGY OF SOME OF THE RAILWAY LINES 433

cut such a gap in a mountain range. After heavy rain the sand and gravel of rivers like these are pushed or carried forward long distances, and the force of the current is immense ; then it must be remembered that the age of this river system is very great, even in a geological sense, and that the vigour of the streams has been renewed more than once by elevation of the country. A smaller but sharply defined gap west of the Gamka Poort is the Seven Weeks' Poort, close under the highest point of the range. A third gorge, to the east of the Gamka Poort, is Meiring*s Poort. The Buffers Eiver Poort is not well seen from the railway line. To the north the view is closed by the great cliffs of the Nieuweveld escarpment, capped by massive and roughly columnar sheets of intrusive dolerite. The highest point is Bulthouders Bank, 6,270 feet above the sea. As Beaufort West (2,850 feet) is approached the details of the structure of the cliffs become more and more obvious, and a second thick sheet consider- ably lower than the uppermost one, can be distinguished. It caps the high plateau projecting far from the moun- tain west of the town of Beaufort. In reahty there are three thick sheets near the town, but the highest one cannot be distinguished from the second till Beaufort West is left. For many miles along the line beyond Prince Albert Boad two mountains at the western end of the Nieuweveld cliffs are very conspicuous, one is Tafel Berg, a flat-topped mountain crowned by a columnar sheet of dolerite 400 feet thick, and rising 3,000 feet from the ground at its base, and the second

is the pointed Spitz Kop, slightly lower than Tafel Berg

28

434 GEOLOGY OF OAPE COLONY

but capped by a remnant of the same sheet. These are outliers of the sheet at the top of the extreme western part of the NieuwevelcL Just beyond the Beaufort Station the line crosses a thick dolerite dyke inclined northwards ; it has been cut through to allow the rail- way to pass ; to the east a corresponding section is visible at the end of the wall of the town dam. The Beaufort dyke, as it is called, has been traced for several miles each side of the town, and on the west it appears to have supplied the second of the dolerite sheets men- tioned above.

A few miles south-west of Prince Albert Eoad the line passes over the boundary between the Ecca and Beaufort beds ; no conspicuous feature marks its posi- tion, but north of it the remains of Pareiasaurus and other reptiles are found. The Dicynodon b^ds are passed over beyond Beaufort West.

After traversing the wide alluvial flats beyond Beau- fort the line ascends the main Colonial watershed, which is crossed near Biesjes Poort. Along this section dolerite sheets are the most conspicuous features in the country ; the great variety in the shapes of mountain sides and kopjes is due to the progress of denudation in a rock mass of horizontal strata with sheets and dykes of dolerite. The reddish or deep brown boulders, often many feet in diameter and covered with a thin varnish of black oxides of iron on their most exposed surfaces, are portions of the dolerite separated from their parent outcrops by the weather.

From the watershed to the Orange Eiver, both on the Kimberley and Johannesburg lines, the train runs across

GEOLOGY OF SOME OF THE RAILWAY LINES 435

wide flats where outcrops are hardly to be found, with flat-topped dolerite or sandstone-capped hills in the dis- tance, then it approaches groups of these hills and winds its way between them. The geology of this part of the Colony as far east as Stormberg Junction has not been examined in detail. The most remarkable features near the railway in this district are the two flat-topped hills, Theebus and Coffeebus, between Eosmead Junction and Steynsburg ; they owe their form to the presence of dolerite caps which have protected the softer sedi- mentary beds below.

From Eosmead the Port Elizabeth line follows the valley of the Great Fish Eiver as far as Commadagga, a distance of over a hundred miles, lying upon the Karroo formation all the way. At Commadagga the Dwyka series is traversed ; the similarity in character of that rock at Matjes Fontein and in the Eastern Province can be noticed. The survey of the country through which this line passes has not yet been made. One of the points of interest in travelling across the Karroo formation a second time many miles to the east of the western main line is the difference in vegeta- tion connected with the different climates in the two regions, and the distinction is still more marked when a comparison is made with the country traversed by the East London line below Queenstown, where grass veld predominates over bush. Below Commadagga the Witte- berg beds are seen on either side of the line as far as Sand Flats, but near Alicedale, a syncline of the Dwyka, the western end of the Grahamstown syncline is trav- ersed. At Sand Flats the railway enters the area of

28*

436 GEOLCKIY OP CAPE (U)LONY

the Uitenhage beds and remains in it as far as the Zwartkops bridge, from that place to Port Elizabeth the low ground near the line is formed of superficial deposits, chiefly raised beaches. From the bridge over Sunday's River may be seen the light-coloured cliflfs of the Sun- day's Biver beds containing marine fossils.

The line from Alicedale to Grahamstown and thence to Port Alfred lies on the Witteberg beds for the greater part of the distance, but near Grahamstown it traverses the Dwyka conglomerate for a few miles. In that neighbourhood a well-preserved terrace, north of the line, can be seen from the railway. The bridge over the Blaauw Krantz Eiver, a tributary of the Kowie, is built just to the north of a rather fine gorge through the folded Witteberg quartzites ; this gorge is analogous to the far greater poorts of the Buflfel's, Gamka and other rivers in the Zwartebergen and Langebergen, through which the Great Karroo is drained.

The East London line descends the southern flank of the main watershed at Bushman's Hoek, and an excellent view of the almost precipitous face of the escarpment is obtained from the train. To the east of this region the Stormberg series is well developed, but near the railway only the Molteno beds are seen ; the spoil heaps at the entrance to drives and pits near Molteno and Cyphergat mark the coal mines. The most striking features of the Stormberg series, the Cave sand- stone and volcanic beds, are not seen near this line. The Karroo formation with its intrusions of dolerite extends to the coast at East London, but no detailed surveys have yet been made in that part of the Colony.

GEOLOGY OF SOME OF THE RAILWAY LINES 437

KosMBAD TO Poet Elizabeth and Oudtshoorn via

Klipplaat.^

Leaving Eosmead (4,044 feet), the railway passes over high veld covered with grass and small bush ; dolerite sheets cover the tops of the hills, and lines of kopjes mark the courses of the dolerite dykes. Passing Middelburg, the line approaches the escarpment of the Sneeuwberg, and is taken over the edge a little to the east of the highest point in the range, Compass Berg (8,600 feet), and indeed the highest in the Colony, with the exception of some peaks on the Drakensberg. The escarpment is formed of a sheet of dolerite capping the Karroo rocks ; it is at the edge of the plain that slopes to the Orange Eiver, and the edge forms the main watershed dividing the streams flowing north and south. Graaff Eeinet (2,463 feet), lies on the lowest slopes of the escarpment, the precipitous part of which may be reckoned at 1,000 feet. After winding down the face of the cliff, the line is taken along the banks of a stream, and the sides of the hills are steep and heavily charged with dolerite sheets, which give the gorge a wild and forbidding appearance. The dolerite, both when capping the hill tops, or exposed on a level with the river, is coarsely columnar, and gives rise to fantastically shaped pillars. The prickly pear has taken possession of the veld, and renders much of it useless. Springs come to the surface all along the river, and there is ample water for irrigation, but there is very little soil on which to use it, as only very narrow patches of alluvium occur ;

^ This section was given me by Mr. E. H. L. Schwarz.

438 GEOLOGY OF CAPE COLONY

nearing Graaflf Beinet, however, these patches become larger and are covered with lucerne lands.

At Graaff Beinet the line leaves the mountains and the dolerite. On the east of the town are the Tandjes Bergen, the capping sheet of dolerite looking from a distance like the teeth of a saw, and on the west is a fine conical hill, Spander's Kop, with a crown of sandstone which has been hardened by dolerite and forms a vertical cliff all round ; the dolerite now forms only an inconsiderable heap of boulders on top of the sandstone ; to the north, however, the full thickness of the same dolerite sheet can be seen, and the celebrated Valley of Desolation is cut in it.

Leaving Graafif Beinet, the line runs over a wide plain formed of a peculiar variety of Karroo sandstone and shale, the surface of the ground being sandy and littered with small fragments of siHcified wood, chert and limestone, till Klipplaat is reached. One branch goes down to Uitenhage and Port Elizabeth, passing the Dwyka conglomerate at Mount Stewart, and the Witteberg beds between there and Barroe, and thence it descends into the low-lying coast country made up of the various members of the Uitenhage series, Enon conglomerate. Wood bed, etc., the older rocks occasion- ally appearing at the surface.

At Uitenhage the marine beds of the Uitenhage series occur with many fossils, and the plateau that reaches the coast is here seen. Leaving Uitenhage the line follows the Zwartkops Biver; on the left are clififs cut in the marine beds and levelled at the top. Great beach deposits lie on the plateau about here, and

GEOLOGY OF SOME OF THE RAILWAY LINES 439

nearer the coast the shell beds contain a large species of Pectunculus, Some distance away in among the beach deposits on top of the plateau is the celebrated Zwartkops Salt Pan. On the right there are ridges of red and variegated marls which are used for tile making, and away to the west the corresponding cliffs, level topped and beach covered, are seen. At their foot lies the Bethelsdorp Salt Pan.

The other branch of the line leaves Klipplaat to go to Willowmore and Oudtshoorn. The line approaches the hills at a very acute angle and before reaching them passes between kopjes made of Ecca (mottled) on the north and Dwyka conglomerate on the south. The pillowy and pinnacled features of the latter are well shown, but the ** White band '* that lies on the top of the conglomerate is badly exposed and is only noticeable from the white chert that occurs in it.

The line then enters the Witteberg hills at Swanepoel's Poort. The Witteberg quartzites are bent into acute folds, but the tops of the hills have been cut more or less level and in places great open grassy flats occur between them. The folds repeatedly bring the Dwyka shales and the conglomerate to the level of the Plessis Eiver, along which the embankment is carried, and the axis of the folds being east and west the valleys are likewise in that direction. At Waai Kraal there is a very wide syncline filled with Dwyka conglomerate and the shales immediately above and below it. The line then passes through a poort and enters a flat country covered with deep red soil derived from the weathering of the Bokkeveld beds. The line steadily rises an(J

440 GEOLOGY OF CAPE COLONY

approaches the ridges of hills formed of folded Bokke- veld beds, the sandstones of which look very much like those of the Witteberg, but the amount of clay-slate and shale between them is greater. After passing through a small poort in these hills the train runs into Willowmore.

From Willowmore there is a long stretch of country formed of folded Bokkeveld beds and then the line turns round and makes straight for a narrow slit in the mountains. These mountains are the^astern end of the Zwartebergen, and are composed of Table Moun- tain sandstone ; the tremendous folding and crumpling observable in the Zwartberg Pass and Meiring*s Poort have died out, though even here the beds stand vert- ical. The slit is Tover Water's Poort, through which runs the Traka Eiver. On the south side of the mountains there is again a tract of Bokkeveld hills to the east, but to the west the Enon conglomerate occupies a consid- erable area. On the outcrop of the junction of the Table Mountain sandstone and the Bokkeveld to the east there is a hot spring. The line is carried over the Bokkeveld beds past Uniondale Koad and past the bend along the Olifant's Eiver. On the Oudtshoom side of the bend there are high krantzes of red Enon conglomerate, which rock, however, soon crosses the river and the overlying white Enon forms the centre of the valley. To the north are the Zwartebergen with a very characteristic shelf or old river plateau high up on the mountain side ; to the south, in the distance, are the Kammanassie Mountains, also made of the Table Mountain sandstone, and between them and the line

GEOLOGY OF SOME OF THE RAILWAY LINES 441

are first, kopjes of the Bokkeveld beds, then red Enon and then white Enon conglomerate. The same fea- tures extend past Vlakte Plaats, where the line enters a tract of Bokkeveld and skirts a peculiar inlier of Table Mountain sandstone; then it passes by the mouth of Meiring's Poort, but between the poort itself and the line there are high hills of Cango beds cut to a level top and capped with river gravel. To the south the end of the Kammanassie Mountains can be seen where the village of Dysseldorp stands. Thence to Oudtshoorn one passes through red Enon, white Enon and finally the sandy beds above the last.

The Eende Kuil and Hopefield Lines.

Leaving the main line at Kraaifontein the Malmes- bury branch traverses undulating country cut out of the Malmesbury beds as far as the town of that name, where granite is met with. At Klipheuvel, a faulted outlier of Table Mountain sandstone, the continuation of Joostengerg is crossed.

At Kalabas Kraal the Hopefield railway branches ofif from the Malmesbury line. The great granite masses of Paarde Berg and Dassen Berg form considerable hills to the east and west of the line near Kalabas Kraal. The Dassen Berg mass is followed as far as Darling where the line turns northwards through the Zwartland, a flat grain country of little geological interest, as far as Hopefield.

From Malmesbury the Eende Kuil line skirts the eastern edge of the Zwartland, and some good sections of the sericitic slates of the Malmesbury series are ex-

442 GEOLOGY OF CAPE COLONY

posed in the cuttings. The structure of the wide extent of Malmesbury beds which stretch from the west flank of the 01ifant*s Biver and Cardouw's Mountains to Piquetberg and the Saldanha Bay granite is not under- stood. These beds are intensely folded and consist of phyllites with occasional interbedded layers of grits and quartzites. From the Berg Biver Bridge to Eende Kuil fine views of the escarpments of Piquetberg and the 01ifant*s Biver Mountains can be seen from the train.

Caledon Line.

Leaving the main line at Durban Boad the railway passes the granite of Kanon Kop and Papagaai's Berg and the Helderberg outlier of Table Mountain sand- stone ; it skirts the eastern part of the Cape Flats. Near Somerset West a small but interesting granite mass is passed about two miles on the down side of the station ; this granite contains much tourmaline, andalu- site, and other accessory minerals. Beyond Sir Lowry's Pass the line ascends the steep sandstone escarpment of Hottentot's Holland and enters the Grabouw-Houw Hoek area of Bokkeveld beds, a more or less quad- rangular sunken tract defined by north-west and north- east lines of folding. The Bot Biver Valley is gained by the Houw Hoek Pass and from that river to Caledon the line runs over the Bokkeveld beds. The rugged mountain of Table Mountain sandstone near Caledon is an anticline, to the south the sandstone again rises from below the Bokkeveld beds in the Babylon's Tower Bange, and to the north in the Zond^r Jlinde Mountains,

GEOLOGY OF SOME OF THE KAILWAY LINES 443

The Eiversdale Line (Cape Central Eailway).

This line leaves the Government Eailway at Wor- cester. The Ecca beds are seen near Worcester Station, but for some distance to the east nothing but river gravels and occasional banks cut into the Uitenhage conglomerates (Enon type) which are not easily dis- tinguishable from river gravels from the train, can be seen from the railway. Near Nuy Siding the Ecca beds are again seen, and at Lange Vley the line passes on to the Dwyka series. From Vink Eiver to Eobert- son the Witteberg, Bokkeveld, and Table Mountain sandstone are traversed. The high hill to the north of the railway between Vink Eiver and Eobertson is a great mass of granite intrusive in the Pre-Cape rocks (Mal- mesbury series) north of the Worcester fault which makes a bend round it. The Cape and Karroo forma- tions abut against the fault in this region, having been folded in a north-easterly direction on the down-throw side. Between Eobertson and Ashton the railway crosses an outlier of the Enon conglomerates which cover the great fault in this neighbourhood. From Ashton to Swellendam Bokkeveld and Witteberg beds are seen, the latter form the conspicuous hills with thick groups of quartzite beds. Near Swellendam an ill-de- fined Uitenhage outlier is crossed, and yet another is entered at Slang Eiver ; from there to Eiversdale excel- lent sections through the clays, shales, and conglomer- ates of the Uitenhage beds are exposed in the cuttings ; fossils have been obtained from several of those cuttings. East of Swellendam numerous extensive gravel plat-p

444 GEOLOGY OF CAPE COLONY

eaux lying high above the rivers are seen, and small outliers of them form the table-shaped hills characteristic of the Buggens — the hilly country between the Zonder Einde and Langebergen on the north and the coast mountains on the south.

APPENDIX.

List of books and papers referred to in the body of the work : —

Abel, C. Narrative of a Journey in the Interior of Chiiia, chap. xi.

London, 1818. Anialitzky, W. ** A Comparison of the Permian Freshwater

Lamellibranchiata from Russia with those from the Karoo

System of South Africa," Quart, Journ. Geol. Soc., p. 337.,

1895.

Sur les Fouilles de 1899 de Debris de Vert^hrh dans Us Depots

Permiens de la Russie du Nord, Warsaw, 1900.

Anderson, W. First Report of the Geological Survey of Natal and-

Zululand. Pietermaritzburg, 1901. Atherstone, W. G. " Geology of Uitenhage,'* The Eastern Province

Mmithly Magazine, vol. i., pp. 518 and 580. Grahamstown,

1857.

**A Geological Tour from Grahamstown to the Kasouga,"

Cape Monthly Magazine, 1st series, vol. iv., pp. 273-282 and 328-334. Cape Town, 1858.

Baily, W. H. " Description of some Cretaceous Fossils from South Africa Collected by Capt. Garden of the 45th Regiment," Quart. Journ. Geol. Soc, xi., p. 454, 1855.

Bain, A. G. " On the Geology of Southern Africa,*' Tram. Geol. Soc, 2nd series, vol. vii., p. 175, 1856; with appendices : by D. Sharpe on the ^^ Fossils from the Secondary Rocks of Sunday River and Zwartkop River " ; by D. Sharpe and J. W. Salter on "Palseozoic Fossils from South Africa" ; by D. Sharpe, J. D. Hooker and Sir P. Egerton on ''Some Fossils from the Karoo Desert and Its Vicinity ".

Blanford, H. F. and W. T., and Theobald, W. "On the Geo- logical Structure and Relations of the Talcheer Coal Field, in the District of Cuttack," Memoirs of Geol. Survey of India, vol. i., p. 33, 1859.

445

446 GEOLOGY OF CAPE COLONY

Bonney, T. G. " The Parent-rock of the Diamond in South Africa/* Geol. Mag., p. 309, 1899.

Broom, R. **0n an almost Perfect Skeleton of Pareiasaurus strrideiu% Owen," and five other papers, in Ann, of tJie South African Museum, vol. iv., pt. ii., 1903. Also many papers in Trans. S. A. Phil. »S'or., vols. xi.-xv., Records of the Albany Museum, and Report of the South African Association for the Advaiicennient of Scierwe, vol. i.

Chapman, F. '* Foraminifera and Ostracoda from the Cretaceous of East Pondoland, South Africa," Ann. of the South African Museum, vol. iv., pt. v., 1904.

Cohen, E. Letters to the Editor, Neues Jahrbuch filr Min., etc., p. 857, 1872 ; and p. 62, 1873.

" Geognostisch-petrographische Skizzen aus Sild-Afrika,"

Neiies Jahrbuch fur Mia., etc., p. 460, 1874 ; and Beilage- band, v., p. 195, 1887.

Corstorphine, G. S. C. See Geol. Commission, Anniuil Reports.

Daniell, G. W. B. " The Mineral Waters of Caledon," South African

Medical Journal, vol. ii., p. 242, 1895. Dunn, E. J. Report on a Gold Prospecting Expedition in Nam-aqua-

land, Parliamentary Report, G. 21. Cape Town, 1872.

" On the Mode of Occurrence of Diamonds in South Africa,"

Quart, Jonrn, Geol. Soc., xxx., p. 54, 1874.

Report on Gamdeboo aiid. Nieuweveldt Coal, Parliamentary Re- port, G. 37. Cape Town, 1879.

Report on the Stormberg Coal Fields, Parliamentary Report, G.

8. Cape Town, 1878.

Report on a Supposed Extensive Deposit of Goal Underlying the

Central Districts of the Colony, Parliamentary Report, G. 8. Cape Town, 1886.

'* Notes on the Dwyka Coal Measures at Vereeniging, Trans- vaal," Trans. S, A. Phil. Soc, xi., p. 67, 1900.

Geological Sketch Map of Cape Colony. London, 1872.

Geological Sketch Map of South Africa. London, 1875.

Melbourne, 1887.

Feistmantel, O. " Uebersichtliche Darstellung der Geologisch- Palaeontologischen Verhaltnisse Siid-Afrikas" (1. Theil), Abh, der kimigl, bohrn. Gesellschaft der Wiss., vii. Folge, 3 Band, 1889.

APPENDIX 447

Geological Commissiony Annual Report ofy for 1896. Cape Town, 1897.

For 1897. Cape Town, 1898.

For 1898. Cape Town, 1900.

For 1899. Cape Town, 1900.

For 1900. Cape Town, 1901.

For 1901. Cape Town, 1902.

For 1902. Cape Town, 1903.

For 19a3. Cape Town, 1904.

Gervais, P. Description du MesosaurtLS tenuidens, Reptile fossile de UAfriqiie australe. Montpellier, 1865.

Graichen, W. "Die Diamentminen Sud-Afrikas," Zeitschrift fiir prakt. Geologiej p. 448, 1903.

Green, A. H. Report on the Goal Fields of the Cape Colony, Parlia- mentary Report) 1883.

"A Contribution to the Geology and Physical Geography

of the Cape Colony," Qtuirt, Journ, GeoL Soc,, xliv., p. 239, 1888.

Griesbach, C. L. "On the Geology of Natal in South Africa,"

Qiiart. Journ, Geol. Soc., xxvii., p. 53, 1871. Gumprecht, T. E. Die Mineralquelle auf dem Festlande von Africa,

etc. Berlin, 1851.

Hochstetter, F. v. " Beitrage zur Geologie des Caplandes," Reiser der Osterreichischen Fregatte Novnra um die Erde, Geol. Theil, vol., ii., p. 19. Vienna, 1866.

Holub, E., and Neumayr, M. "Ueber einige Fossilien aus der Uitenhage-Formation in Stid-Airika," Dmk, der Math.-natur- wissenschaftlichen cUisse der Kaiserlichen Akad. der Wissen- schaften, Vienna, 1881.

Huxley, T. H. " On Saurostemon Bainii and Pristerodon McEayi, two new fossil Lacertilian Reptiles from South Africa," GeoL Mag., vol. v., p. 201, 1868. This paper contains sections and notes by Mr. McKay illustrating the Geology of East London. Other papers by Huxley dealing with Karroo Reptiles will be found in vols, xv., xxiii., xxvi., of the Quart. Journ. Geol. Soc.

Jack, R. L., and Etheridge, R. The Geology and Palceontology of Queensland and New Guinea. London, 1892.

448 GEOLOGY OF CAPE COLONY

Johnson, J. P. ** Notes on Sections at Shark River and the Greek, Algoa Bay," Tram, GeoL Soc. S,A., vi., p. 9, 1903.

Jones, T. B. ''On the Enon Conglomerate of the Cape of Good Hope and its Fossil Esthoriae," Geol. Mag., p. 350, 1901.

Kalkowsky, E. " Die Verkieselung der Gesteine in der nordlichen

Kalahari," Abh. der naiurwiss. Gesellschaft ^^Isis" in Dresden,

1901. Kossmat, F. ''On the Importance of the Cretaceous Rocks of

Southern India in Estimating the Geographical Conditions

during late Cretaceous Times," Trans, in Records Geol, Surv.

of Lvdia, xzviii., p. 39, 1895. Krauss, F. "Ueber einige Petrefacten aus der untem Ereide des

Kaplandes," Nova Acta Acad, Gaes, Leop,-Gar, Nat. Gur,, xxii.,

p. 441, 1847.

" Ueber die Quellen des Siidlichen Afrika's," Neues JahrbueJi

fur Min., etc., p. 150, 1843.

Lacroix, A. ** Note sur les Mineraux et les Roches du Gisement diamantifere de Monastery (Etat libre d'Orange) et sur ceux du Griqualand," Bull. Soc, Franc, Min,, xxi., pp. 21-29, 1898.

Lake, P. " The Trilobites of the Bokkeveld Beds," A7m, of the South African Museum, vol. iv., pt. iv., 1904.

Launay, L. de. Les IHarnants du Gap, Paris, 1897.

Les Riclvesse MinP.rales de VAfrique. Paris, 1903.

Lewis, CarvilL Papers ayid Notes on the Genesis and Matrix of the Diamond, edited by Professor T. G. Bonney. London, 1897.

Lydekker, R. Catalogue of the Fossil Eeptilia in the British Museum. London, 1888-1890.

Molengraaff, G. A. F. " G^ologie de la R^publique Sud-Africaine du Transvaal," Bulletin de la Soci^t^ Giologique de France, 4th series, vol. L, p. 13, 1901.

"The Glacial Origin of the Dwyka Conglomerate," Trans.

Geol, Soc. S, A., vol. iv., pt. i., p. 103, 1895.

Molyneux, W. Report on the Geology of the Karoo and Stomiberg,

Parliamentary Report, G. 71. Cape Town, 1881. MouUe, A. " M^moire sur la G^ologie g^n^rale et sur les mines de

diamants de I'Afrique du Sud," Annales des Mines, 8th series,

vol. vii., p. 193, 1885. Miigge, O. "Ueber einige Gesteine des Massui-Landes," Neues

Jahrbuch fur Min., etc., Beilage-band iv., p. 603, 1886,

APPENDIX 449

Keuxnayr, M., and Holub, E. "Ueber einige Fossilien aus der Uitenhage-Formation in Stid-Afrika," Denk. der MatK-natur- wissenschaftlichen cltisse der Kaiserlichen Acad, der Wisseri" schaftm, Vienna, 1881.

Newton, R. B. ''On the Occurrence of Alectryonia ungulata in S. E. Africa, with a Notice of Previous Researches on the Cretaceous Conchology of Southern Africa," Joum. of Con- chology, p. 136, 1895.

Noble, J. Official Handbook of the Cape and South Africa, Cape Town, 1893.

North, F. W. Report on the Goal Fields of (he Stormbergen, Parlia- mentary Report, G. 47. Cape Town, 1878.

Osborne, C. F., and Bain, T. Report on Gold Discoveries in (he Knysna Division^ Parliamentary Report, G. 46. Cape Town, 1886.

Owen, R^ Descriptive and Illustrative Gatalogue of the Fossil Reptilia of South Africa in the Gollection of the British Museum, London, 1876. Also many papers in the Quart. Joum. GeoL Soc. between 1845 and 1887.

Penck, A. '* Die Eiszeiten Australiens," Zeitsch, Gesellsch. /. Erdk,^ xzxv., p. 239. Berlin.

Pinchin, R. "A Short Description of the Greology of Part of the Eastern Province of the Colony of the Cape of Good Hope,'* Quart. Joum. Geol, Soc.y vol. zzzi., p. 106, 1875.

Playfair. *^ Account of the Structure of Table Mountain, and Other Parts of the Peninsula of the Cape, from Observations by Capt Basil Hall, R.N.," Trans, Roy. Soc. Edin., vol. vii., p. 269.

Prior, G. T. *' Contributions to the Petrology of British East Africa," etc., Minerahgical Magazine^ xiii., p. 228, 1903.

Prosser, W. ''Limestones of the Colony," TramA, 8, A. PhiL Soc.y vol. i., pt. ii., p. 47, 1879.

Reed, F. R. C. "Brachiopoda from the Bokkeveld Beds," Ann, of the Sovih African Museum, vol. iv., pt. iiL^ 1903.

^'Mollufloa from the Bokkeveld Beds," Ann. of (he South

African Museum, voL iv., pt. vL, 1904.

Reiohenbach, E. Fr. Stromer v. Die Geologie der Deutschen Schutz- gehiete in Afrika, Munich and Leipzig, 1896.

Rogers, A. W. *' On a Glacial Conglomerate in the Table Mountain Sandstone,'* Trans. S. A. Phil. Soc., vol. xi., p. 236, 1902.

29

450 GEOLOGY OF CAPE COLONY

Bogen, A. W. "The Geological History of the Gomritz River

System," Tram. t>. A. Phil. 8oc., vol. xiv., p. 375, 1903. and Sohwarz, £. H. L. ** Notes on the Recent Limestones

on Parts of the South and West Coasts of Cape Colony,"

Tram. S. A. Phil. Soc., vol. x., p. 427, 1899. "The Orange River Ground-Moraine," Tram. S. A.

Phil. Soc., vpl. zi., p. 113, 1900. "The Transkei Gap," Tram. S. A. Phil. Soc, vol. xiv..

p. 60, 1903. — '• — and du Toit, A. L. " The Volcanic Pipes of Sutherland and

Their Relationship to Other Vents in South Africa," Trans.

S. A. Phil, Soc., voL xv., 1904. Roaenbusoh, H. Mikroskojmche Phytiographie der Mamgengedeine.

Stuttgart, 1896. Rubidge, R. N. " On Some Points in the Greology of South Africa,"

Quart. Jourv. Geol. Soc., xv., p. 196, 1859. Schenok, A. "Die geologische Entwickelung Si&dafrikas,*' Peter-

mann's Mittheilmigen, vol. xxxiv., p. 225, 1888. " Ueber die Kupferzlagerstatte von Ookiep in Klein Nama-

land," Zeitschrift der deutsch geol, GegelUchafi, p. 64, 1901. Schwarz, E. H. L. "The Volcanoes of Griqualand East," Trans.

S. A. Phil. Soc., vol. xiv., 1903.

"An Unrecognised Agent in the Deformation of Rocks," Trans.

S. A, Phil, Soc., vol. xiv., 1903.

"High-level Gravels of Cape Colony and the Problem of

Karroo Gold," Tram. S, A. Phil. Soc., vol. xv., 1904.

and Rogers, A. W. " Notes on the Recent Limestones on

Parts of the South and West Coasts of Cape Colony," Trans. S. A. Phil. Soc, vol. X., p. 427, 1899;

"The Orange River Ground-Moraine," Tram. S. A.

Phil. Soc, vol. xi., p. 113, 1900. "The Transkei Gap," Tram. S. A. Phil. Soc, vol. xiv..

p. 66, 1903. Seeley, H. G. "Researches on the Structure, Organisation and Classification of the Fossil ReptiUa" (pts. i. -ix.), Phil. Tram. Roy. Soc, vols. 178-186, 1887-1895.

" On Bubalus Bainii," Qeol. Mag., p. 199, 1891.

"The Mesosauria of South Africa," Quart. Joum. Geol. Soc.^

xlviil, p. 586, 1892. Papers in Quart. Joum. Geol, Soc, vols, xxxiv., xlviii., and Ann. and Mag. Nat, Hist,, xv., 1895,

APPENDIX 451

Story-Maskelyne, N., and Flight, W. **0n the Character of the Diamantiferous Rook of South Africa," Qiiart, Joum, GeoL Socy XXX., pp. 406-416, 1874.

Stow, G. W. ** On Some Points in South African Geology," Qiiart, Joum, GeoL Soc., xxvii., p. 497, 1871.

" Geological Notes upon Griqualand West," Quart. Joum, GeoL

Soc., XXX., p. 581, 1874.

Sutherland, P. C. On the Geology of Natal {South Africa). Durban, 1868.

" Notes on an Ancient Boulder-Clay of Natal," Quart. Joiim.

GeoL Soc.j xxvi., p. 614, 1870.

Tate, R. "On Some Secondary Fossils from South Africa," Quart. Joum. GeoL Soc., xxiii., p. 139, 1867.

Du Toit, A. L., and Rogers, A. W. " The Volcanic Pipes of Suther- land and Their Relationship to Other Vents in South Africa," Trans. S. A. Phil. Soc., vol. xv., 1904.

Williams, G. F. The Diamond Mines of Souih Africa, etc. New York, 1902.

Wyley, A. Report upon the Maitland Mines near Port Elizabeth, Parliamentary Report, G. 5. Cape Town, 1856.

Report upoH the Mineral and Geological Structure of South Nama-

qualand and the Adjoining Mineral Districts, Parliamentary Report, G. 36. Cape Town, 1857.

Report of the Geological Surveyor upon a Journey Made hy Him,

mainly during the Year 1858, in Two Directions O/cross the Colony, and Its Results, with an appendix containing a detailed account of the journey. Parliamentary Report, G. 54. Cape Town, 1859. Zirkel, F. Lehrhuch der Petrographie, 3 vols. Leipzig, 1893.

29

INDEX.

Aberdeen, 24, 194, 256.

ActcBonina, 288, 292. .

Actinopteriaj 138.

Addo, 292.

JElurosaurus, 193, 284.

Agulhas, 96.

Alariay 292.

Algoa Bay, 381, 282.

Alicedale (Ptieska), 71, 80.

Alleman's Hoek (Beaufort West),

254. Alluvium, 363-367. Amalienstein, 49. — or Cango Fault, 104. Amandel Bosch Bug, 303. Amatolas, 257. Amboccelia, 134. Ammonites, 292, 322. Amygdaloidal lavas of Zeekoe Baard, 170.

Beer Vley, 170.

Stormberg series, 214.

Anisocerasy 325.

Annular dykes, 269.

Anodontopsis, 133.

Anomodontia, 234.

Anysber?, 104.

AraticariteSj 287.

Asbestos Mountains, 64, 69.

Ashton, 34, 307.

Assegaai Bosch (Biversdale), 300.

Astarte, 293, 326.

Astieriay 316.

Atherstonia, 197, 224, 343.

Avellana, 322-325.

Avicula, 293.

Baboon Point, 110. Babylon's Tower, 19, 102. Baculites, 292, 316, 322, 325. Baiera, 222. Bakoven's Hoogte, 34. Balmoral (Fraserburg), 339.

Bank Berg, 257.

Banks Gaten (Beaufort West), 254.

Barkly Pass, 211.

Basutoland, 221.

BatracosuchtLSy 195.

Baviaan's Kloof (Willowmore), 363.

Mountains, 106, 353, 354.

Beaufort dykp, the, 256.

— series, 189, 226, 256, 337. Bedford, 197, 256.

Beer Vley volcanic rocks, 86. Belemnites, 292. Bellerophoriy 122, 131, 133. Benstedtia, 286» 287. Bereniceaf 294. Bethelsdorp salt pan, 290. Beukes Fontein (Ceres Karroo),

169, 251. Bezuidenhout's valley, 285, 287. Bidouw (Clanwilliam), 142.

— Kuil (Hope Town), 86. Bier-Biver Mountains, 101. Bizana, 185.

Blaauw Blonuxkekjes Keep (Suther- land), 339.

— Kranz (Calvinia), 175.

Black shales in the Dwyka series,

173. Blink Berg (Geres), 126.

— Fontein (Prieska), 85, 87. Blown Sand, 369-373.

Blue Cliff, 282.

— -ground, 336, 342. Bokkeveld Mountain, 25, 60, 62, 97.

— period, 395, 396.

— series, 49, 121-137. Bonteberg, 142. Bosch Kloof, 252. Boschiesman's Berg, 67. Bosch veld Mountains, 101. Botha's Hill (Albany), 143, 175,

179, 358. Bothriceps, 196,

463

454

INDEX

Bothrodendran, 186, 188.

Bottle's Kop, 297.

Boven Plaats (Sutherland), 253.

Brachyphyllum, 287.

Br&kbosch Poort HilU, 65, 67.

Brak Pan (Hope Town), 86.

Brand Vley (Worcester), 891.

Brandwacht valley, 800.

Bredasdorp, 38, 131, 352, 376.

Britstown, 181.

Brul Pan (Prieska), 66.

Bubaltu, 367.

Building stone, 117, 226, 276.

Buis Valley, 70, 72, 73.

Bulthouder's Bank (Beaufort

West), 254. Bult FoDtein Mine, 331. Bushmanland, 63, 370. Butterwortn, 265. Byssopteria, 133.

Gala, 203, 269.

Calcareous concretions in Beaufort

beds, 190. Dwyka beds, 170, 179.

— tufa, 377-379. Galedon, 131, 352, 368.

— hot spring, 391.

— Mountain, 102. Galitzdorp, 49, 307. Callipteridium, 204, 222. Calvinia, 58, 131, 268, 272, 366, 370. Gamdeboo, 193.

Gampbell Rand, 64, 69. series, 68, 171, 370.

— Town, 68. Gango caves, 54-55.

— fault, 49, 409, 416.

— series, 36, 48. Gape Agulhas, 373.

— Baixacouta, 373.

— flats, 360, 368, 373.

— formation, 5, 93. .

— Hangklip, 19, 96, 101. -1- Hermes, 184.

— Infanta, 376, 380, 382.

— Recife, 96, 106, 373.

— St. Blaize, 96, 297, 299, 300. Francis, 96.

— Town, 32. Cardioccurpus, 185, 186. Cardiomorpha^ 133. Cardita, 293.

Cardium, 322, 326. Carnarvon, 332. Carpolithes. 287. Cassidulus, 323, 326. Gat's Pass, 264. Gave sandstone, 207. Gederbergen, 17-23, 99, 251. Ceratodus, 223, 224, 227. Geres, 123.

— Karroo, 251, 355. Cerithium, 325. Ceromya, 293. Chalcedony, 190, 216. Chara, 360. Chetnnitzia, 322, 326. Chert in Dwyka series, 173. Chiastolile-schist, 45. ChiropteriSj 222. ChaneUs, 122, 131, 134. Cidaria^ 294.

Cladophlebis, 222, 223, 287, 290,

299. Clay-pellet conglomerate, 190. Cleithrolepis, 208, 223, 224. 227. Goal (Beaufort beds), 192.

— (Molteno beds), 203.

— (sub-Karroo), 174.

Goetzee's Poort (Oudtshoorn), 67,

308. Cold Bokkeveld, 122, 126, 129, 142.

Mountains, 18.

Colesberg, 197.

Columba Mission Station, 264. G3mmada9[ga, 143. Compass Berg, 257. CoDglomerate (Cango), 52, 56.

— (Dwyka), 147-179.

— (Enon and Uitenhage), 282-285,

294-309.

— (Ibiquas), 59.

— (Table Mountain series), 110-

113.

— (Embotyi), 328-330. Coniferous wood, 286. Conitesy 186, 287. Constable, 145, 178. Contemporaneous erosion and de- posit, 192, 204.

Conularia, 130, 133.

Coo, 100.

Copper ores of Namaqualand, 91.

Corbula, 293, 326.

Cordaites, 223.

INDEX

455

Cradock hot spring, 891. Cretaceous system, 9, 281, 318, Crinoids, 130, 131; Crioceras, 292, 316. Crooidolite, 74. Cryptocynodon, 196. CryptodoUj 364. Cryptonelia^ 134. CiictdlcBa, 293, 326. Gycadolepis, 287. Cyclostignia, 139. Cynochampsa^ 196. Cynodracot 196.

Cynognathus, 195, 288, 239, 241. CynosuchuSf 196. Cyphergat, 200. Cyprina, 293. Cypris, 302. Cytherea, 326.

DalmaniteSy 134.

Danger Point, 96, 377.

DasyuruSt 241.

De Aar, 273.

De Beer's Mine, 331, 336, 342.

De Dooms, 122.

Deer Park (Matatie'e), 220.

DelphinognathiiSy 197.

Dentalium, 290, 326.

Denudation in pre-Uitenhage

times, 310. Despatch (Uitenhage), 284. DeuterosauruSy 284. De Vrede (Sutherland), 339. Diabase, 52, 56-58. Diademodony 195. Diamond, 349. Dictyopyge, 224. Dicynodan, 196, 198, 224, 236, 236

406. Dicynodontia, 234. Diorite, 47-48. DitrockosauruSj 176. Dobbel Aar's Kloof, 148, 178. Dolerite, 24 26, 45, 62, 245. Donkerhoek Mountains, 101. Doombergen, 12, 64, 69, 72^75. Doomberg's Fontein (Prieska), 73. Doom River valley (Galvinia), 60. Drakensberg, 210. Drakensteins, 17, 97. Drie Fontein, 253*. Dry Diggings. 331.

Dunbrody, 286,313. Dune-limestone, 273. Du Toit's Kloof, 117.

— Pan Mine, 331. Dwars Berg, 102.

Dwas Douw (Calvinia), 275. Dwyka series, 147 if., 397-405.

Eagle's Nest (Butte^worth), 265. Bast London, 197, 256, 381, 388,

390. Ecca series, 179, 328. Eclogite, 336, 344. Eenzamheid (Kalahari), 179. Egossa Forest, 251, 328. Eland's Berg (Calvinia), 252.

(Laingsburg), 143.

(Uitenhage), 106.

— Kloof, 33.

— Vley (Clanwilliam and Cal-

vinia), 140, 160, 162, 164. 169,

252. Elim, 33. ■ Elliot, 208.

Bmbotyi series, 271, 318, 328. Endothiodan, 196, 236. Enon beds, 282, 284. Esoterodofiy 196. Estheriay 301, 313. EuskelesauruSy 223. Exogyray 292, 293. Eyrie, 213. Ezel Klauw (Prieska), 66.

— Kop Vlakte, 61.

— Rand, 12, 65, 76, 86.

Fasciolariay 322, 326. Fault (Bushmanl'and and ^Van Rhyn's Dorp), 61.

— (Cango), 49, 104.

— (Pondoland), 184.

— (Worcester), 29, 30, 33, 49, 102,

142, 172, 177, 186, 307. 409,

416. Ferruginous gravels, etc., 367. Fissure eruptions, 220. .

Folded belt, 17. Foraminifera, 323. Forests, 119-121. Fort Beaufort, 197, 226, 266. Fraserbuig, 194, 332, 366. French Hoek, 33, 97,

456

INDEX

Galena in Campbell Rand bed-*, 72. Oalesanrust 196, 288. Oamka Poort, 51, 104, 121, 129. Gamtoo*B YaUey, 284, 294. Oangamopkris, 176, 181-186, 199,

228. Gaicia's Pass, 421. Qastrochcena, 286, 288, 818. Gat Beig (Elliot), 217. Gates of St. John's, 22, 106. Geelhoutboom (Uitenhage), 286. Genadendal, 856. Gentuli, 265, 266. George, 88, 48.

— granite, 297, 800.

— Moshesh's country, 220. Gerustheid (Clanwilliam), 142. Qervillia, 298.

Gift Berg, 97.

Glacial oonglomerate in Table Mountain series, 111-118.

Glaciated floor below Dwyka oon- glomerate, 154-160.

Glaucophane-schist, 88.

Glittering sandstone, 202.

Glossopteris, 180-188, 197-199, 228- 224, 406.

Olossites, 188.

Gneiss (Bushmanland), 61.

— (Prieska), 64. Gobogobo, 265.

Gold in Table Mountain series, 117-118.

Witteberg beds, 145.

GomphognathuSt 195, 238. Gondwanaland, 405, 406. Gonubie Hill, 267. Ocrgonops, 196. Goudlni Boad, 307. Gouph, 194, 197. Gouritz River Poort, 129. Government Salt Pan (Uitenhage),

290. Gqunqi, 263. Graafi Beinet, 226, 256.

mineral spring, 891.

Grabouw, 117, 122. Grahamstown, 143, 355, 382. Orammysia, 133. Granites (in north and noith-west),

61, 64, 79-81, 91.

— (in south and south-west), 38,

41-45.

Granophyre, 260, 268, 268.

Granulites, 81, 886.

Grass Ridge, 290.

Great Karroo, 28.

Great Winterhoek Range, 106.

Grenaat's Eop, 67, 80.

Grey*s Pass, 110.

GriquaJand East, 256.

Griqualandite, 74.

Griqualand West, 68.

Griqua Town series, 69, 78, 88, 170,

870. Grobbelaar*s valley (Gango), 53. Groenland Mountains, 101. Groot River Range, 143.

— Vader's Bosch, 806.

Guap Mountain (Galvinia), 258.

Gydo Pass, 128.

Gypsum, 178, 314, 808, 387.

Hagel Kraal (Mossel Bay), 353. HamiteSf 292, 816. Hangklip, 97. Hankey, 284. Hanover, 832.

Hantam (Galvinia), 181, 258. Hartenbosch (Mossel Bay), 300. Heidelberg, 300, 313.

— beds, 301. Helix, 874. Hemiastery 323, 326. Herbertsdale, 298, 299, 313, 353. Hermanns, 96. . Heulandite, 216.

Hex River Mountains, 19, 99, 102.

valley, 126,

High level gravels, 806. Hoetjes Bay, 873. Holaster, 326. HonialonottiSf 122, 184. Honig Berg, 36, 58.

— Klip Kloof, 298. Hope- Town, 154, 175, 181. Hornblende-granulite, 88. schist, 48.

Hot springs, 391, 892. Houwhoek, 19, 101, 121. Humansdorp, 294.

Ibiquas series, 58, 99, 251.

Ice, movement of, in Dwyka times,

162, 172. Ictidosaurus, 196.

INDEX

457

IctidostichtiSt 196, 234.

Idutywa beds, 185, 197, 198, 266.

Indwe, 200, 269.

Ingeli Mountain, 268.

Inland surface limestone, 377-379.

Inoceramus, 322, 328, 326.

Insiswa, 268.

Inver Gcua, 264.

Ironstone gravels, 367.

Isastrcea, 294.

Izinhluzabalungu, 319.

Jackal's Fontein (Sutherland), 254.

— Water (Prieska), 76, 156. Jager's Fontein Mine, 331. Jamestown, 216.

Jan Niemand's Bosch, 121. Jansenville, 24. Javander Kop, 256. Jointed pebbles in the Dwyka con- glomerate, 168. Jonker's Hoek, 97. Jonker Water (Prieska), 65, 68. Joosten Berg, 96, 110. Jorsten's Berg (Hope Town), 86«

Kaaing Bult, 63. Kaap Plateau, 13, 64, 69, 70. Kaboom (Prieska), 66, 67, 69. Kalahari, 175, 179, 363, 370. Kalk Fontein (Prieska), 72. Kameel Puts (Prieska), 73. Kammanassie Mountain, 106. Kareedouws Mountains, 102. Karree Bergen, 257. Karroo basin, 8, 11, 23.

— formation, 7, 146.

— Poort, 175. Keerom Berg, 100. 'Keis series, 67, 370. Keizie, 100. Kenhardt, 15, 63, 370. Kentani, 258, 264.

— Hill, 355, 360. Kimberley, 175, 340.

— Mine, 331.

— pipes, 219, 304, 339.

— shales, 174, 181, 184. Kimberlite, 342, 346. Klaarstroom, 104.

Klaas Kaffir's Heuvel, 357. Klapmuts Hill, 95, 110. Klein Berg, 143.

Klein Dassen Berg, 48.

— Modderfontein (Prieska), 66.

— Paarde Berg, 48.

— Boggeveld, 194, 180, 197.

— Straat, 122.

— Winterhoek Bange, 143. Klip Fontein's Berg pan, 387.

— Rug Kop (Calvinia), 253. Klomp Boomen (Calvinia), 61. Knysna, 106, 118, 294.

Kobe Mountain, 97- Kogman's Kloof, 307. Komgha, 256, 360. Komsberg, 26, 192, 254. Kouga Hills (Worcester), 143.

— Mountains, 106, 354. Kragga Poort, 145. Kreits Berg (Calvinia). 263. Kwardouw Mountains, 100.

Ladismith Karroo, 103, 122. Lady Grey, 102. Lagoons, 883. Laingsburg, 175, 354.

— beds, 180.

Langebergen (Calvinia), 61, 63, 91, 252.

— (Griqualand West); 13, 66, 76.

— (South Coast), 17, 100, 102, 352. Lange Kuil (Sutherland), 192.

— Vley (Robertson), 177. Laterite, 368.

Lavas of the Stormberg series, 214.

Leda, 133.

Leeuw Kloof Poort, 366.

— River Poort, 193. Lepidodendroid plants, 130, 223. Leptoccelia, 123, 131, 134. Libode, 185.

Lignite in Uitenhage beds, 309.

Lima, 293.

Limestone (Cango), 54-56.

— (Cretaceous), 291, 319.

— (Karroo), 190, 227.

— (Pre-Cape), 34, 55, 70. Limiet Berg, 33. Litnncea^ 360. Lingula, 130, 134. Lion's Rump, 16. Lithodomtis, 293. Littorina, 133. Loeries Fontein, 175. Long Kloof Mountains, 102.

458

INDEX

Lower Dwyka shales, fossils in, 166. Loxonema^ 183. Lusiksiki, 184, 268. Lusizi, 26S. LycosauruSt 196. Lycosuchus, 196, 234, 241. LystrosaurtiSf 196, 286* Lytoceras, 325.

Maclear, 200.

Main watershed, formation of, 274. Maitland Mines, 118. Malmesbury, 38, 368, 391.

— series, 32, 68, 297. Maluti Mountains, 211. Massospondylusy 223, 224. Matatiele, 200, 208, 210, 213, 269. Matjes Fontein, 140, 169.

(Calvinia), 160, 898.

(Sutherland), 336, 339.

— Kop (Worcester), 143. Mats4p Hills, 64.

— series, 76, 88, 171. Matsiekamma, 97. Mazeppa Bay, 259, 262. Meiring's Poort, 49-61, 308. Melilite-basalt, 303, 333-336, 344,

346. MesosaurtiSf 176, 183. Metamorphism due no dolerite, 274

granite, 44.

Mica-diorite of the Transkei, 266.

Microgomphodony 196, 238.

Microgranite of Komgha, 267.

MicropholiSt 196.

Millwood, 117, 118.

Mimosa Dale (Kentani), 264.

Mitchell's Pass, 99.

Modiola, 293.

Modiomarphaj 133.

Molteno beds, 200.

Monodonta, 293.

Montagu Pass, 106.

Montague, 391.

Moordenaar's Karroo, 194, 197.

Moorreesburg, 34.

Mossel Bay, 33, 296, 352.

Mostert's Hoek Range, 101.

Mount Ayliff, 268.

— Currey, 268. Mountain Building, 407-408.

— Cliff, (Elliot), 211. MytiltiSy 293,

Namaqualand, 16, 370.

— schists, 16, 90. Napier, 366. Nardouw Berg, 97. Natal, 21. Natica, 293, 326. Nauga (Prieska), 72.

— Hills (Worcester), 143. Nautilus, 322.

N'Debe Nek, 266.

Neithia, 326.

Neritopsis, 293.

Newlands Mine, 342, 346.

N'Hlambe, 265.

Nieuweveld, 26, 189, 197, 264, 276.

Noeggerathiopsis, 176, 182, 186.

Noro Kei Pan (Kalahari), 179.

Nostronzetuia, 234.

Notochampsay 227, 244.

Nquise, 264.

Nqundwyu, 266.

N'tabankulu, 268.

Ntumla, 290, 326.

Nuculites, 133.

Nuy, 307.

Nxaxo, 264.

Nyntugha, 264.

Oil shale, 206. OlcostephanuSf 292, 316. Olive shale group, 181. Onychiopsis, 287, 289. OphiocnnuSf 133. Opisthoctenodon, 196. Orange River valley, 370. Orbiculoidea, 131, 134. Orinosaurus, 223. OrosauruSy 223. Orthis, 134.

Orthocerasy 122, 131, 134. Orthonotay 133. Orthothetes, 122, 134. Ostracods, 323. Ostreay 288, 313, 326. Ottrelite-schist, 34. Oudenodan, 196, 236, 268. Oudtshoom^ 296, 307, 308, 366. Outiniquas Mountains, 1(M2, 118.

Paarde Berg (Ladismith), 103.

(Malmesbury), 16.

Paarl, 33, 368. — Mountain, 16,

INDEX

459

Pakhuis Pass, 111.

Pakeanodanta, 197, 198.

PalcBomutela, 197, 198.

Palcemeilo, 123, 129-183.

PalcBoniscus, 197.

Paliguana, 244.

Paltje's Kraal (Uitenhage), 286.

Panne-veld, 387.

Pans, 385-387.

Parallelodofif 293.

PareiasaurttSy 189-193, 197, 199,

231-234, 240, 406. PariotichuSf 231. Patella, 293. Paternoster, 423. Pecten, 288, 290, 313, 322, 326. Pectunculus, 322, 326, 381. Peninsula, the, 33, 94. Pema, 293. Petrified wood, 286. Phacops, 122, 131. PhoenicopsiSf 222. Pholadamya, 293. Phyllite-gneiss, 44. Phyllotheca, 166, 180, 186. Pienaar's Kloof, 142. Pikenier's Kloof. 96, 110. Pinna, 293. Pipe-amygdules, 216. Piquetberg, 33, 34, 96. Placunopsis, 293. PlesiosauruSj 292. Pleuromya, 294. Pleurotomaria, 133. Pogha Hills, 103, 122. Pollia, 326.

Pondoland, 21, 96, 162, 256. Poortje pan, 366, 384. Porphyroid, 54. Port Beaufort, 123. — Elizabeth, 33. Pot Berg, 123. Potgieter's Poort (Oudtshoom), 57,

308. Potkly's Berg East, 252. PrcBca/rdium, 138. Premier Mine, 343. Pre-Gape region and rooks, 4, 12,

32-92. Prieska, 12, 63, 163, 370. Prince Albert, 104, 175, 355. Pristerodon, 196. Pristerognathus, 197.

Procoloplum, 196, 230, 234, 289, 240. Proetus, 134. Proterosuchiis, 244. Protocardiumy 822, 326. Psammobia, 286, 288, 294, 302. Psygmophyllum, 186, 188, 198. Ptychognathus, 196, 236. Ptyckcymya, 294, 316. Ptychosiagum, 198, 224. PugnelluSy 326. Puzosia, 325. Pyrolusite in Table Mountain series,

117. Pjrroxene-granulite, 82.

Quarrie Kloof (Worcester), 178. * Quartzite-lenticles in Dwyka con- glomerate, 169. Queenstown, 226, 268.

Raised beaches, 379-383.

Kawsonville, 101.

Rawson bridge, 290.

Recent alterations in level, 419-424.

— deposits, 11, 351-390. Red beds, 206. Rensselceria, 134.

Reptiles in the Karroo formation, 228-244.

Uitenage formation, 292, 309.

Umzamba formation, 322.

Retzia, 134. Bhopalodon, 234. Rhynchonella, 134. Rhynchospira, 134. Rhytidostetis, 195. Riebeek Kasteel, 16. River-diggings, 831. River Bavian's Kloof, 309.

— Berg, 367, 383.

— Bitou, 295, 364.

— Bot, 102, 873, 384.

— Brandewyn, 25, 251.

— Breede, 363, 383.

— Buffalo, 354, 366, 381.

— Buffeljagts, 805, 352.

— Camdini (Calvinia), 175.

— Goega (Uitenhage), 290.

— Dasbosch, 101.

— Doom (Galvinia and Van Rhyn's

Dorp), 59, 61.

(Geres and Glanwilliam), 142.

(George), 363.

460

INDEX

Biver Doom (Heidelberg), 802-803.

— Draai Kraal's (Galvinia), 367,

385.

— Duivenhoek's, 800.

— Dwyka, 264.

— Embotyi, 184.

— Fish (Boggeveld), 194.

— Gamka, 57, 807, 808, 366.

— Gamtoos, 106.

— Qcua, 258, 268.

— Goree, 185.

— Gouritz, 298, 353, 382.

— Great Brak, 297, 388.

— Grobbelaar's, 53.

— Groen (Calvin ia), 251.

— Gualana, 12, 21, 356.

— Hartog's Kloof, 266.

— Hex, 122.

— Houd den Bek's, 99.

— Istamfoona, 263.

— Kaaiman's, 366.

— Kabakazi, 263, 267.

— Kaffir Kuirs, 300.

— Kei, 258, 263, 273, 383.

— Kenigha, 200.

— Kleena, 262.

— Klein, 384.

Brak, 383.

Doom, 303.

Vette, 302.

— Kobonqaba, 259, 263.

— Koekemoer's, 254.

— Kogha, 268, 263, 265, 267.

— Kologha, 263.

— Kombolo, 264.

— Kowie, 145, 384.

— Kraai ^Sutherland), 181.

— Kruis (Gango), 51.

— Lang Touw, 299.

— Mabele, 218.

— Manubi, 262.

— Matje's (Cango), 63.

— Modder, 367.

— Nauga (orNouga) (Mossel Bay),

307, 353.

— Nels (Gango), 67.

— Nxagha, 262.

— OlifanVs (Glanwilliam), 96, 391. (Oudtshoorn), 308, 365, 392.

— Onger's, 256.

— Oorlog's Kloof, 99, 131, 160, 181,

252, 363.

— Orange, 331. •

Biver Palmiet, 101.

— Patata's, 180.

— Pisang, 295, 296.

— Portugal's, 339.

— Bhenoster (Galvinia), 253.

— Saffraan, 353.

— St. John's, 383.

— Slang (Swellendam), 300.

— Spiegel, 332, 383.

— Stink, 298.

— Sunday's, 288, 290. - Tanqua, 175, 180.

— Touw's, 178.

— Tra-Tra, 142.

— Troe Troe, 97.

— Umfane, 263.

— Umgwegwane, 328.

— Umnyama, 264.

— Umtamvuna, 318.

— Umtentu, 318.

— Vaal, 165, 881.

— Waterval (Biversdale), 363.

— Weyer's, 297.

— Wheeli, 269.

— White (or Witte), 285-287.

— Winkelhaak's, 142, 178.

— Witteberg'g, 164, 166.

— Wolf, 253.

— Zak, 256.

— Zondag's (Galedon), 83.

— Zwart Kops, 282, 283. Biversdale, 296, 352, 377. Biverstone (Kentani), 264. Bobertson, 131, 185, 296, 307, 313. Bobinson's Pass, 297.

Boches Moutonn^es, 172. Boep-my-niet Mountains, 253. Boggeveld, 25, 181, 194, 197, 253,

273. Boode Berg (Oudtshoorn), 103. (Worcester), 142.

— Fontein (Galvinia), 253.

— Hoogte sheet, 254, 255. (Biversdale), 298.

— Zand Mountains, 97, 100. Buggens, 352, 356, 382. Buitersbosch, 297.

Saft Sit Pan (Prieska), 81. St. John's, 21, 106. Saldanha Bay, 373, 380, 423. Salt-pans, 385-387. Saltpetre Kop, 254, 337, 346.

INDEX

461

Salt Vlei (Port Elizabeth), 291. Sand dunes, 872, 373.

— Veld, 370-372. Sanguinolit0Sf 133. Saron, 36, 58.

SaurosterncM, 196, 243, 244, Scala, 326.

Schalk's Puts (Prieska), 80. Schiet Fontein (Carnarvon), 340. Schizaneura, 180-186, 188, 197, 198,

204, 222, 223. Schloenbachia^ 325. Schoeman's Poort, 52. Schurffceberg (Ceres), 99, 126.

— (Griqualand West), 66. SclerosauruSy 231. Scolecite, 216. Scylacosaurus, 196. ScymnosauricSt 196.

Seal Point, 295.

Seebachia^ 294.

Semicmotus, 208, 223, 227.

Serpukiy 294.

SerpuliteSj 133.

Seven Weeks' Poort Mountain, 104.

Shale bands in Table Mountain series, 110.

Sigillaria, 186-188, 223.

Silicified wood, 179, 207, 343.

Silver Dam (Sutherland), 336, 336.

Simon's Berg, 96.

Slab-structure in Dwyka con- glomerate, 166.

Slang Bergen (Fraserburg), 257.

— Hoek, 101. Sneeuwbergen, 197, 257. Sneeuw Kop (Cederberg), 109.

— Erantz (Sutherland), 253. Solarium^ 326.

Somerset East, 197.

— West, 33.

Sphsnopteris, 186, 223-224, 287. Spirifer, 122, 129, 131, 134, SpirophytoUt 139.

Spitzkop (Beaufort West), 26, 254.

S-shaped gorges, 422.

Steenkamps Poort, 255.

Stellenbosoh, 83.

Stenopteris, 200, 204, 222.

Stilbite, 216.

Stink Fontein Poort, 69.

Stoltz Hoek (Beaufort West), 266.

Stone implements, 388-390.

Stormberg series, 199.

Strand Fontein (Van Rhyn's Dorp),

369-372. Striated boulder-pavement, 160.

— pavements in Dwyka conglom-

erate, 162. StrobUiteSf 222. Stropheodonta^ 134. Strophonellay 134. Struys Point, 374. Sub-Karroo coal, 174. Sugar Loaf Hill (Grahamstown),

355, 360. Sunday's Biver beds, 282, 29a Superficial deposits, 351. Surface deposits, 298.

— quartzites, 357. Sutherland, 268, 332, 334, 366.

— pipes, 304.

Swellendam, 33, 34, 296, 305, 313, 362, 369.

Table Mountain, 94.

series, 6, 26, 49, 94-121, 318,

328. glacial conglomerate in,

111-113. soils and vegetation on,

118-121. Tachylite, 270. Tcmiopteris, 199, 204, 222-223, 224,

287, 299. Tafel Berg (Beaufort West), 26, 264, 276.

(Riversdale), 353.

Tandjes Berg, 257. Tanqua Karroo, 164, 356.

— valley, 252. TapinocephaltiSj 234. TaxUes, 287, 299. Telemachus Kop, 216. Tembu Pass, 211. Tentaculites, 131, 133. Teredo, 322, 326.

Terrace and plateau cutting, 419. Terraces, 351-363. Thecay 133. Thee Kloof, 266. Theriodontia, 238,. 239. TheriognathuSy 196. Theroceph&lia, 233. Thinnfeldia, 197, 198, 204, 222, 224.

462

INDEX

Thomsonite, 216. Thorn Bay, 879. Tigri8uchu8, 196. Titanofuchm, 284, 406. Toleni, 265. Tontelbosch Kolk, 366. Touw'8 River Station, 142. Touw Vlakte, 108. Tover Kop, 104.

— Water Poort, 307. Transkei, 256.

— gap-dykes, 264, 265. Trapeziuniy 294.

Trigonia, 184, 222, 292-296, 316,

817, 826. TriUmidea, 326. Tritylodon, 228. Trochtu, 298. TropidoleptuSt 184. Tsala hills, 268. Tuff dykes, 887. Tuin Plaats (Sutherland), 181. Tulbagh, 83.

Tulloch (Elliot), 211, 217. Turbo, 288, 298. Turhonilla, 826. Turritella, 826. Tutugha, 265.

Twenty Four River Mountains, 97. Tygerberg, 84. Typhloniscu»t 184.

Uitenhage, 288, 818.

— conglomerate, 271.

— series, 49, 281, 414, 417. Umsikaba beds, 184. Umzamba beds, 818. Uniondale, 106, 296. Upper Dwyka shales, 173*

— Karroo, 28.

Vaartwell, 52.

Van Rhyn's Dorp, 20, 33, 34, 68, 96.

— Wyk's Pan (Prieska), 67. Vereeniging, 172, 186. Verloren Vley (Piquetberg), 62. Victoria West, 194.

Vilet's Kuil (Hope Town), 168. Villiersdorp, 19, 101, 102. Vitulina, 134.

Vlakte Plaats (Oudtshoom), 308. Vleys, 383-387.

Vloers, 366, 885. Voetpad Berg, 142. Vogel Valley, 84.

Mountains, 83.

Volcanic periods, 413.

— pipes, 813.

— series of Beer Vley, 86.

Stormberg, 210.

Zeekoe Baard, 86.

Viyburg, 179.

Waai Kloof (Worcester), 84, 307. Wagenboom Berg, 100. Warm Bokkeveld, 121, 122.

— Water Berg, 103.

hot spring, 391.

Washbank peak, 212, 217.

Water, boring for, in the Karroo,

227. Waterfall Bluff (Pondoland), 328. Watershed, main Colonial, 1-3,

412. White band, 173. WiUiston, 267. WiUowmore, 106, 296, 307. Windvogel Berg, 269. Winterbergen, 267. Winterhoek, 33, 100. Witte Drift (Piquetberg), 62.

— Vlakte (Calvinia), 366. Wittebergen, 140. Witteberg period, 396, 896.

— series, 188-146, 396, 396. Witzenbergen, 18, 101.

Wolve Kraal (Uitenhage), 282, 290. Wood beds, 282, 285, 813. Worcester, 38, 84, 142, 177, 178, 186, 296, 307, 313.

— fault, 29, 30, 33, 49, 102, 142,

177, 186, 307, 409, 416. Wupperthal, 123.

Xalanga Peak, 212.

York (Matatiele), 218. Yzer Fontein Point, 48.

Zamites, 286-288, 291. Zand Kop (Galvinla), 263.

— Leegte, 371.

Zeekoe Baard (Prieska), 70, 72, 84.

volcanic groups, 86, 336.

Zitzikamma, 102.

INDEX

463

Zoet Vlei (Prieska), 85.

Zoetendal Yley, 384.

Zonder Einde Mountains, 101,

102. Zuurberg Poort, 104. Zuurbergen, 148, 284. Zuurbraak, 352, 853. Zwart Kop (Prieska), 85. Pan (Prieska), 76, 84.

Zwart Kop Kuggens, 142, 855, 856. Zwartberg (Caledon), 102.

— folds, 18, 408.

— Pass, 51.

Zwartebergen, 17-28, 51, 108, 104,

854. Zwartkops Heights, 882.

— salt-pan, 886.

— vaUey, 284, 290.

THB ABBBDEBN UNIVERSITY PRESS UMITED

% Classtto Catalogue

OF WORKS IN

GENERAL LITERATURE

PUBLISHED BY

LONGMANS, GREEN, & CO.,

39 PATERNOSTER ROW, LONDON, E.C,

91 AND 93 FIFTH AVENUE, NEW YORK, and 32 HORNBY ROAD, BOMBAY.

CONTENTS.

Badminton Library (The) Biography, Personal Memoirs

Children's Books

Classical Literature, Trans LATiONS, Etc. .

Cookery, Domestic Manage MENT, Etc.

Evolut.on, Anthropology, Etc

Fiction, Humour, Etc.

Fine Arts (The) and Music

Fur, Fea ther and Fin Series

History, Politics, Polity, Po LiTiCAL Memoirs, Etc. .

Language, History and Science

OF

Logic. Rhetoric, Psychology

PAOB 12

8 31

22

36 20 25 36 14

1

19 16

Mental, Moral and Political Philosophy . . . .

Miscellaneous and Critical Works

Poetry and the Drama

Political Economy and Eco- nomics

Popular Science

Religion, The Science op

Silver Library (The)

Sport and Pastime .

s tonyhurs t philosophical Series

Travel and Adventure, the Colonies, Etc.

Works of Reference

PAOB

16

38 23

20 29 21 33 12

19

10 31

History, Politics, Polity, Political Memoirs, etc.

ABBOTT. —A HISTORY OF

GREECE. By Evelyn Abbott, M.A., LL.D.

Part I. — From the Earliest Times to the Ionian Revolt. Crown 8vo, lOs. 6^.

Part II. — From the Ionian Revolt to the Thirty Years' Peace, 500445 B.C. Crown 8vo, lOs. 6d.

Part III. — From the Peace of 445 B.C. to the Fall of the Thirty at Athens in 403 B.C. Crown 8vo, 105. 6d.

ABBOTT. — TOMMY CORN-

STALK : being some Account of the Less Notable Features of the South African War from the Point of View of the Australian Ranks. By J. H. M. Abbott. Crown 8vo, 55. net.

ACLAND and RANSOME.— A

HANDBOOK IN OUTLINE OF THE POLITICAL HISTORY OF ENGLAND TO 1896. Chronologically Arranged. By the Right Hon. A. H. Dyke Acland, and Cyril Ransome, M.A. Crown 8vo, 6s.

AIRY.— CHARLES II. By Osmund

Airy, LL.D., M.A. With Photogravure Portrait. Crown 8vo.

ALLGOOD.— CHINA WAR, 1860:

LETTERS AND JOURNALS. By Major- General G. Allgood, C.B , formerly Lieut. G. Allgood, 1st Division China Field Force. With Maps, Plans and Illustrations. Demy 4to, 125. 6d. net.

LONGMANS AND CO, S STANDARD AND GENERAL H^OR/CS.

History, Politics, Polity, Political Memoirs, etc— continued.

ANNUAL REGISTER (The). A

Review of Public Bvents at Home hnd Abroad, for the year 1902. 8vo, 18s.

Volumes of the ANNUAL REGIS- TER for the vears 1863-1901 can still be had. 185. each.

ARNOLD. — INTRODUCTORY

LBCTURBS ON MODBRN HISTORY. By Thomas Arnold. D.D., formerly Head Master of Rugby School. 8vo. 7s. 6d.

ASHLEY (W. J.). ENGLISH ECONOMIC HIS-

TORY AND THBORY. Part I.. 55 ; Part 11., 105. Qd.

Crown 8vo,

SURVEYS, HISTORIC AND BCONOMIC. Crown 8vo. 95. net

BAGWELL.— IRELAND UNDER

THB TUDORS. By Richard Baowbll, LL.D. (3 vols.) Vols. I. and II., from the first invasion of the Northmen to the year 1578. 8vo, 32s. Vol. 111.. 1578-1603. 8vo.

185.

BAILLIE.— THE ORIENTAL CLUB AND HANOVER SQUARE. By Alexander P. Baillie. With 6 Photo- gravure Portraits and 8 Full-page Illustra- tions. Crown 4to, 25s. net.

BELMORE.— THE HISTORY OF

TvVO ULSTER MANORS, AND OF THEIR OWNERS. By the Earl op Bbl- MORB, P.C., G.C.M.G. (H.M.L., Countv Tyrone), formerly Governor of New South Wales. Re-issue Revised and Enlarged. With Portrait. 8vo, 55. net.

BESANT.— THE HISTORY OF

LONDON. By Sir Walter Bbsant. With 74 Illustrations. Crown 8vo, l5. 9d, Or bound as a School Prize Book, gilt edges, 25.6^.

BRIGHT. — A HISTORY OF

ENGLAND. By the Rev. J. Franck Bright, D.D.

Period I. MEDiiCVAL Monarchy :

a.d. 449-1485. Crown 8vo, 45. 6d.

Period II. Personal Monarchy.

1485-1688. Crown 8vo, 55.

Period III. ConstitutiOxNAL Mon- archy. 1689-1837. Crown 8vo, 75. 6d.

Period IV. The Growth of De-

hocracy. 1837-1880. Crown 8vo, 65.

Period V. Imperial Reaction :

Victoria, 1880-1901. Crown 8vo, 45. 6d.

BRUCE.— THE FORWARD

POLICY AND ITS RESULTS : or, Thirty- Five Years' Work amongst the Tribes on our North- Western Frontier of India. By Richard Isaac Bruce, CLE. With 28 Illustrations luid a Map. 8vo, 155. net.

BUCKLE.— HISTORY OF CIVILI-

SATION IN ENGLAND. By Henry Thomas Buckle.

Cabinet Edition, 8vo, 245.

3 vols. Crown

• Silver Library ' Edition, 3 vols. Crown 8vo, 105. 6d.

BURKE.— A HISTORY OF SPAIN,

FROM THE EARLIEST TIMES TO THE DEATH OF FERDINAND THE CATHO- LIC. By Ulick Ralph Burke, M.A. Edited by Martin A. S. Hume. With 6 Maps. 2 vols. Crown 8vo, 165. net.

CASSERLY.— THE LAND OF

THE BOXERS; or, China under the Allies. By Captain Gordon Cassbrly. With 15 Illustrations and a Plan. 8vo, 105. 6d. net.

CHESNEY.— INDIAN POLITY : a

View of the System of Administration in India. By General Sir George Chbsnby, K.C.B. With Map showing all the Admini- strative Divisions of British India. 8vo,

215.

CHURCHILL (Winston Spbncbr,

M.P.).

THE RIVER WAR: an Historical Account of the Reconquest of the Soudan. Edited by Colonel F. Rhodes, D.S.O. With Photogravure Portrait of Viscount Kitchener ot Khartoum, and 22 Maps and Plans. 8vo, 105. 6<i. net.

THE STORY OF THE MALA- KAND FIELD FORCE, 1897. With 6 Maps suid Plans. Crown 8vo, 35. Bd,

LONDON TO LADYSMITH VIA PRETORIA. Crown 8vo, 65.

IAN HAMILTON'S MARCH. With Portrait of Major-General Sir Ian Hamil- ton and 10 Maps and Plans. Crown

\ 8vo, 65.

I -

LONGMANS AND CoJ's STANDARD AND GENERAL IVORJCS.

History, Politics, Polity, Political Memoirs, etc.— continued,

FROUDE (James A.)'-confinued.

THE SPANI-H STORY OF THE

Cr. 8vo,

dORBETT (JuLiAv S.)

DRAKE AND THE TUDOR NAVY, with a History of the Rise of England as a Maritime Power. With Portraits, Illustrations and Maps. 2 vols. Crown 8vo, 16s.

THE SUCCESSORS OF DRAKE

With 4 Portraits (2 Photogravures) and 12 Maps and Plans. 8vo, 21s.

ENGLAND IN THE MEDITER- RANEAN: a Study of the Rise and In- fluence of British Power within the Straits, 1603-1713. 2 vols. 8vo, 24s. net.

CREIGHTON (M., D.D., late Lord

Bishop of London).

A HISTORY OF THE PAPACY FROM THE GREAT SCHISM TO THE SACK OF ROME. 1378-1527. 6 vols. Cr. 8vo, 5s. net each.

QUEEN ELIZABETH. With Portrait. Crown 8vo, 5s. net.

HISTORICAL ESSAYS AND RE- VIEWS. Edited by LouiSB Crbighton. Crown 8vo, 5s. net.

HISTORICAL LECTURES AND ADDRESSES. Edited by LouiSB Crbighton. Crown 8vo, 5s. net.

DALE.— THE PRINCIPLES OF ENGLISH CONSTITUTIONAL HIS- TORY. By Lucy Dale, late Scholar of Somerville College, Oxford. Crown 8vo, 6s

DE TOCQUEVILLE. — DEMO- CRACY IN AMERICA. By Alexis db TocQUEViLLE. Translated by Hbnrv Rbbve, C.B.. D.C.L. 2 vols. Crown 8vo, 16s.

FALKINER.— STUDIES IN IRISH

HISTORY AND BIOGRAPHY, mainly of the Eighteenth Century. By C. Littos Falkiner. 8vo, 12s. Qd. net.

FREEMAN.— THE HISTORICAL

GEOGRAPHY OF EUROPE. By Edward A. Frekman, D.C.L., LL.D. Third Edition. Edited by J. B. Bury, M.A., D.Litt., LL.D., Regius Professor of Modem History in the University of Cambridge. Svo, 12s. Sd.

Atlas to the above. With 65 Maps in colour. Svo, 6s. SeU

FROUDE (Jambs A.).

THE HISTORY OF ENGLAND,

from the Fall of Wolsey to the Defeat of the Spanish Armada. 12 vols. Crown Svo, 3s. 6d. each.

THE DIVORCE OF CATHERINE OF A R AGON. Crown^Svo* 3s. W.

ARMADA, and other Essays. 8s. 6^.

THE ENGLISH IN IRELAND IN

THE EIGHTEENTH CENTURY. 8 vols. Crown Svo, 10s. Sd.

ENGLISH SEAMEN IN THE SIXTEENTH CENTURY.

Cabinet Edition. Crown Svo, 65.

Illustrated Edition. With 5 Photo-

fravure Plates and 16 other Illustrations, .arge Crown Svo, gilt top, 6s. net.

' Silver Library ' Edition Cr. Svo, 8s. 6^.

THE COUNCIL OF TRENT,

Crown Svo, Ss. 6d,

SHORT STUDIES ON GREAT SUBJECTS.

Cabinet Edition. 4 vols. 45

* Silver Library * Edition 4 vols. Crown Svo, 8s. 6d. each.

C^SAR : a Sketch. Cr. Svo, 3s. 6rf

SELECTIONS FROM THE WRIT- INGS OF JAMES ANTHONY FROUDE. Edited by P. S. Allbn, M.A. Crown Svo, 3s. 6d.

GARDINER (Samuel Rawson, D.C.L., LL.D.).

HISTORY OF ENGLAND, from

the Accession of James I. to the Outbreak of the Civil War, 1608-1642. With 7 Maps. 10 vols. Crown Svo, 55. net each.

A HISTORY OF THE GREAT CIVIL WAR. 1642-1649. With 54 Maps and Plans. 4 vols. Cr. Svo, 55. net each.

A HISTORY OF THE COMMON- WEALTH AND THE PROTECTOR- ATE. 1649-1656. 4 vols. Crown Svo, 6s. net each.

THE STUDENT'S HISTORY OF

ENGLAND. With 378 lllustradoM.

Crown Svo, gilt top, 12s.

Also in Three volumes t price 45. each.

WHAT GUNPOWDER PLOT WAS. With 8 lUuotratioM. Crown Svo,

5s.

•I 1

LONGMANS AND CO, S STANDARD AND GENERAL WORKS.

History, Politics, Polity, Political Memoirs, etc. — continued.

GARDINER (Samuel Rawson,

D.C.L., L.L.D.) — continued.

CROMWELL'S PLACE IN HIS- TORY. Pounded on Six Lectures de- livered in the University of Oxford. Crown 8vo, 3s. Qd.

OLIVER CROMWELL. With Frontispiece. Crown 8vo, 5s. net.

GERMAN EMPEROR'S

(THe^SPBf^CHHS : being a Selection from the Speeches, Edicts, Letters and Telegrams of the Emperor William II. Translated by Louis Elkind, M.D., 8vo, 12s. %d. net.

GERMAN EMPIRE (THE)

OP TO-DAY s Outlines of its Formation and Development. By 'Vkritas'. Crown 8vo, 6s. net.

GRAHAM.— ROMAN AFRICA:

An Outline of the History of the Roman Occupation^ of North Africa, based chiefly upon Inscriptions and Monumental Remains in that Country. By Alexander Graham, F.S.A., F.R.I. B.A. With 30 reproductions of Original Drawings by the Author, and 2 Maps. 8vo, 1 6s. net.

GREVILLE.— A JOURNAL OF

THE REIGNS OF KING GEORGE IV., KING WILLIAM IV., AND QUEEN VICTORIA. By Charles C. F Greville, formerly Clerk of the Council. 8 vols. Crown 8vo, 3s. 6rf. each.

GROSS.— THE SOURCES AND

LITERATURE OF ENGLISH HISTORY, FROM THE EARLIEST TIMES TO ABOUT 1485. By Charles Gross, Ph.D. 8vo, 18s. net.

HAMILTON.— HISTORICAL RE-

CORD OF THE 14th (KING'S) HUS- SARS, from A.D. 1715 to a.d. 1900. By Colonel Henry Blackburne Haniilton, M.A., Christ Church, Oxford; late Com- manding the Regiment. With 15 Coloured Plates, 35 Portraits, etc., in Photogravure, and 10 Maps and Plans. Crown 4to, gilt edges, 42s. net.

H A R T.— A C T U A L GOVERN- MENT, AS APPLIED UNDER AMERI- CAN CONDITIONS. By Albert Bush- NELL Hart, LL.D., Professor of History in Harvard University. With 17 Maps and diagrams. Crown 8vo, 7s. 6 '. net.

H I L L.— THREE FRENCHMEN IN BENGAL; or. The Commercial Ruin of the French Settlements in 1757. By S. C. Hill, B.A , B.Sc, Officer in charge of the Records of the Government of India. With 4 Maps. 8vo, 7s. 6df. net.

HARVARD HISTORICAL STUDIES.

THE SUPPRESSION OF THE AFRICAN SLAVE TRADE TO THE UNITED STATES OF AMERICA. 1638- 1870. By W. E. B. Du Bois, Ph.D. 8vo, 7s. 6rf.

THE CONTEST OVER THE RA- TIFICATON OF THE FEDERAL CON- STITUTION IN MASSACHUSETTS. By S. B. Harding, A.M. 8vo, 6s.

A CRITICAL STUDY OF NUL- LIFICATION IN SOUTH CAROLINA. By D. F. Houston, A.M. 8vo, 6s.

ft

NOMINATIONS FOR ELECTIVE OFFICE IN THE UNITED STATES. By Frederick W. Dallingbr, A.M. 8vo, 7s. %d.

A BIBLIOGRAPHY OF BRITISH MUNICIPAL HISTORY, INCLUDING GILDS AND PARLIAMENTARY RE- PRESENTATION. By Charles Gross, Ph.D. 8vo. 12s.

THE LIBERTY AND FREE SOIL PARTIES IN THE NORTH WEST. By Theodore C. Smith, Ph.D. 8vo,

7s. %d.

THE PROVINCIAL GOVERNOR IN THE ENGLISH COLONIES OF NORTH AMERICA. By Evarts Bou- TELL Greene. 8vo, 7s. 6a.

THE COUNTY PALATINE OF DURHAM: a Study in Constitutional History, By Gaillard Thomas Lapsley, Ph.D. 8vo, 10s. %d.

THE ANGLICAN EPISCOPATE AND THE AMERICAN COLONIES. By Arthur Lyon Cross, Ph.D., In- structor in History in the University of Michigan. 8vo, 10s. %d.

HISTORIC TOWNS. — Edited

by E. A. Freeman, D.C.L., and Rev. William Hunt, M.A. With Maps and Plans. Crown 8vo, 3s. %d. each.

Bristol. By Rev. W.

Hunt. Carlisle. By Mandell

Creighton, D.D. Cinque Ports. By

Montagu Burrows. Colchester, By Rev.

E. L. Cutts. Exeter. By E. A.

Freeman. London. By Rev. W.

J. Loftie.

Oxford. By Rev. C. W. Boase.

Winchester. ByG.AV. Kitchin, D.D.

York. By Rev. James Raine.

New York. By Theo- dore Roosevelt.

Boston (U.S.) By Henry Cabot Lodge.

LOrTGMANS AND CO. S STANDARD AND GENERAL WORKS.

History, Politics, Polity, Political iViemoirs, etc. — continued.

HUNTER (Sir William Wilson).

A HISTORY OF BRITISH INDIA. Vol. I. — Introductory to the Overthrow of the English in the Spice Archipelago, 1623. With 4 Maps 8vo, 18s. Vol. 11.— To the Union oF the Old and New Companies under the Earl of Godolphin's Award, 1708. 8vo, 16$.

THE INDIA OF THE QUEEN,

and other Essays. Edited by Lady Hun- ter. With an Introduction by Francis Henry Skrine, Indian Civil Service (Retired). 8vo, 95. net.

INGRAM.— A CRITICAL EXAMI-

nation of IRISH HISTORY. From the Elizabethan Conquest to the Legislative Union of 1800. By T. Dunbar Ingram, LL.D. 2 Tols. 8vo, 245.

JOYCE (P. W.).

A SHORT HISTORY OF IRE- LAND, from the Earliest Times to 1603. Crown 8vo, 105. 6rf.

A SOCIAL HISTORY OF ANCIENT IRELAND: Treating of the Government, Military System and Law ; Religion, Learning and Art ; Trades, Industries and Commerce ; Manners, Custonris and Domestic Life of the Ancient Irish People. With 361 Illustrations. 2 vols. 8vo, 21 5. net.

KAYE and MALLESON. — HIS- TORY OF THE INDIAN MUTINY, 1857- 1858. By Sir John W. Kaye and Colonel G. B Malleson. With Analytical Index and Maps and Plans. 6 vols. Crown 8vo, 35. %d. each.

LANG (Andrew). THE MYSTERY OF MARY

STUART. With Photogravure Plate and 15 other Illustrations. Crown 8vo, 6s. M. net.

JAMES THE SIXTH AND THE GOWRIE MYSTERY. With Cowrie's Coat of Arms in colour, 2 Photogravure Portraits and other Illustrations. 8vo, 125. 6^. net.

PRINCE CHARLES EDWARD STUART, THE YOUNG CHEVALIER. With Photogravure Frontispiece. Crown 8vo, 75. 6./. net.

THE VALET'S TRAGEDY, AND OTHER STUDIES IN SECRET HIS- TORY. With 3 Illustrations. 8vo, 12s.6i^. net

LAURIE. — HISTORICAL SUR

VEY OF PRE-CHRISTIAN EDUCA- TION. By S. S. Laurie, A.M., LL.D. Crown 8vo, 7s. 6^/.

LECKY (William Edward Hart- pole).

HISTORY OF ENGLAND IN THE EIGHTEENTH CENTURY.

Library Edition. 8 vols. 8vo. Vols. I. and II., 17001760, 36s.: Vols. III. and IV., 1780-1784, 36s.; Vols. V. and VI., 1784-1793, 36s.; Vols. VII. and VIII., 1793- 1800. 365.

Cabinet Edition. England. 7 vols. Crown 8vo, 5s. net each. Ireland. 5 vols. Crown 8vo, 5s. net each.

LEADERS OF PUBLIC OPINION IN IRELAND: FLOOD-GRATTAN— O'CONNELL. 2 vols. 8vo, 25s. net.

HISTORY OF EUROPEAN MORALS FROM AUGUSTUS TO CHARLEMAGNE. 2 vols. Crown 8vo, 10s. net.

A SURVEY OF ENGLISH ETHICS : Being the First Chapter of the * History of European Morals *. Edited with Introduction and Notes, by W. A. Hirst. Crown, 8vo, 3s. 6^.

HISTORY OF THE RISE AND INFLUENCE OF THE SPIRIT OF RATIONALISM IN EUROPE. 2 vols. Crown 8vo, 10s. net.

DEMOCRACY AND LIBERTY.

Library Edition. 2 vols. 8vo, 365.

Cabinet Edition. 2 vols. Crown 8vo, lOs. net.

LIEVEN.- LETTERS OF DORO- THEA, PRINCESS LIEVEN, DURING HER RESIDENCE IN LONDON, 1812- 1834. Edited by Lionel G. Robinson. With 2 Photogravure Portraits. 8vo, Us. net.

LOWELL. — GOVERNMENTS AND PARTIES IN CONTINENTAL EUROPE. By A. Lawrence Lowell. 2 vols. 8vo, 21s.

LUMSDEN'S HORSE,

RECORDS OF. — Edited by H. H. S. Pearse. With a Map and numerous Por- traits and Illustrations in the Text. 4to, 21s. net.

LYNCH.— THE WAR OF THE CIVILISATIONS: BEING A RECORD OF 'A FOREIGN DEVIL'S' EXPERI- ENCES WITH THE ALLIES IN CHINA. By George Lynch, Special Correspondent of the 'Sphere,' etc. With Portrait and 21 Illustrations. Crown 8vo, 6s. net.

LONGMANS AND CO, S STANDARD AND GENERAL WORKS.

WORKS OF 0 vols. 8vo,

MACAU LAY (Lord)— o»/my/^rf.

CRITICAL AND HISTORICAL ESSAYS, WITH LAYS OF ANCIENT ROME, etc., in 1 Volume.

Popular Edition. Crown 8vo, 2s. Sd.

'Silver Library' Edition. With Portrait and 4 Illustrations to the * Lays \ Crown 8vo, 3s. 6d.

History, Politics, Polity, Political Memoirs, etc. — contintied,

MACAU LAY (Lord).

THE LIFE AND LORD MACAU LAY.

' Edinburgh * Edition. 6s. each.

Vols. I.-IV. History of England.

Vols. V.-VII. Essays, Biographies, Indian Pbnal Code, Contributions to Knight's 'Quarterly Magazine '.

Vol. VI n. Speeches, Lays of Ancient Rome, Miscellaneous Poems.

Vols. IX. and X. The Life and Letters op Lord Macaulay. By Sir G. O. Trevelyan, Bart.

Popular Edition.

Essays with Lays of Ancient Rome. etc. Crown 8vo, 2s. 6^.

History of England. 2 vols.

Crown 8vo, 5s.

Miscellaneous Writings,

Speeches and Pokms. Cr. 8vo, 2s. 6^. The Life and Letters of Lord

Macaui^y. By Sir G. O. Trevelyan, Crown 8vo, 2s,

AND HISTORICAL 1 vol. Cr. 8vo, 2 vols. Cr.

CRITICAL

ESSAYS. StudenVs Edition.

6s. ' Trevelyan * Edition.

8vo, 9s.

Cabinet Edition. A vols. Post 8vo,

24s. ' Edinburgh ' Edition. 3 vols. 8vo,

6s. each. Library Edition. 3 vols. 8vo, 365.

ESSAYS, which may be had sepa- rately. Sewed, 6^. each ; cloth. Is. each.

Bart.

THE WORKS.

6rf.

'Albany' Edition. With 12 Por- traits. 12 vols. Lar}{e Crown 8vo. 3s. %d, each.

Vols. I. -VI. History OF England, from the Accession of James the Second.

Vols. VII.-X. Essays and Bio- graphies.

Vols. XI.-XII. Speeches, Lays of Ancient Rome, etc., and Index.

Addison and-Walpole.

C r o k e r's BoswelPs Johnson.

Hallatti's Constitu- tional History.

Warren Hastings.

The Earl of Chatham (Two Essays).

Frederick the Great.

Ranke and Gladstone.

Lord Bacon.

Lord Clive.

Lord Byron and The Comic Dramatists of the Restoration.

Cabinet Edition. 8vo, £4 16s.

16 vols Post

HISTORY OF ENGLAND FROM

THE ACCESSION OF JAMES THE SECOND.

Popular Edition. 2 vols. Cr. Svo,

5s. Student's Edition 2 vols. Cr. Svo,

12s. People's Edition. 4 vols. Cr. Svo,

16s, ' Albany ' Edition. With 6 Port aits.

6 vols. Large Crown Svo, 3s. 6^. each. Cabinet Edition. . 8 vols. Post Svo,

48$. ' Edinburgh ' Edition. 4 vols. Svo,

6s. each. Library Edition. 5 vols. 8vo, £4.

MISCELLANEOUS WRITINGS,

SPEECHES AND POEMS. Popular Edition. Crown Svo, 2s. 6rf. Cabinet Edition. 4 vols. Post Svo,

24s,

SELECTIONS FROM THE WRITINGS OF LORD MACAULAY. Edited, with Occasional Notes, by the Right Hon. Sir G. O. Trevelyan, Bart. Crown 8vo, 6s.

MACKINNON (James, Ph.D.).

THE HISTORY OF EDWARD

THE THIRD. Svo, 18s.

THE GROWTH AND DECLINE OF THE FRENCH MONARCHY. Svo.

21s. net.

MALLET. — MALLET DU PAN AND THE FRENCH RBVOLUTION. By Bernard Mallet. With Photogravure Portrait. Svo, 12s. M. net.

MAY.— THE CONSTITUTIONAL HISTORY OF ENGLAND since the Accession of George III. 1760-1870. By Sir Thomas Brskinr May, K.C.B. (Lord Farabereuflh). 3 vels. Crewn Svo, lis.

LONGMANS AND CO. S STANDARD AND GENERAL I^ORA'S.

1

History, Politics, Polity, Political Memoirs, etc. — continued.

MERIVALE (Charles, D.D).

HISTORY OF THE ROMANS UNDER THE EMPIRE. 8 vols. Crown 8vo, 3s. 6d. each.

THE FALL OF THE ROMAN

REPUBLIC : a Short History of the Last Century of the Commonwealth. 12mo, 7s. Si.

GENERAL HISTORY OF ROME.

Tt from the Foundation of the City to the Pall h. of Augustulus, B.C. 753-A.D. 476. With 5 ^ Maps. Crown Bvo, 7s. 6d.

MONTAGUE.— THE ELEMENTS

OF ENGLISH CONSTITUTIONAL HIS- TORY. By F. C. MoNTAOUE, M.A. Crown ^«8vo, 3s. Sd.

MORA N.— T HE THEORY

AND PRACTICE OF THE ENGLISH ► GOVERNMENT. By Thomas Francis MoRAN, Ph.D., Professor of History and Economics in Purdue University, U.S. Cr. 8vo, 55. net.

PEARS. —THE DESTRUCTION OF THE GREEK EMPIRE AND THE STORY OF THE CAPTURE OF CON- STANTINOPLE BY THE TURKS. By Edwin Pbars, LL.B. With 3 Maps and 4 Illustrations. 8vo, 18s. net.

POWELL and TREVELYAN.—

THE PEASANTS' RISING AND THE LOLLARDS: a Collection of Unpublished Documents. Edited by Edoar Powell and G. M. Trbvblyan. 8vo, 6s. net.

RANDOLPH.— THE LAW AND

POLICY OF ANNEXATION, with special Reference to the Philippines : together with Observations on the Status of Cuba. By Carman F. Randolph. 8vo, 9s. net.

RANKIN (Reginald).

THE MARQUIS D'ARGENSON ; AND RICHARD THE SECOND. 8vo, lOs. Sd. net.

A SUBALTERN'S LETTERS TO HIS WIFE. (The Boer War.) Crown Svo, 35. %d.

RANSOME.— THE RISE OF CON- STITUTIONAL GOVERNMENT IN ENG- LAND. By Cyril Ransomb, M.A. Crown 8vo, 6s.

SCOTT.— PORTRAITURES OF

JULIUS C>eSAR: a Monograph. By Frank Jesup Scott. With 38 Plates and 49 Figures in the Text. Imperial 8vo, 21s. net.

SEEBOHM (Frederic, LL.D.,

F.S.A.).

THE ENGLISH VILLAGE COM- MUNITY. With 13 Maps and Plates. 8vo, 165.

SEEBOHM (Frederic, LL.D.,

F.S.A.) — continued.

TRIBAL CUSTOM IN ANGLO- SAXON LAW: being an Essay supple- mental to (1) 'the English Village Com- munity,' (2) • The Tribal System in Wales '. Svo, IBs.

SETON-KARR.— THE CALL TO

ARMS, 1900-1901 ; or a Review of the Im- perial Yeomanry Movement, and some sub- jects connected therewith. By Sir Henry Seton-Karr, M.P. With a Frontispiece by R. Caio.v-Woodville. Crown Svo, Ss. net.

SHEPPARD.— THE OLD ROYAL PALACE OF WHITEHALL. By Edgar Sheppard. D.D., Sub- Dean of H.M. Chapels

■Royal, Sub-Almoner to the King. With 6 Photogravure Plates and 33 other Illustra- tions. Medium Svo, 21s. net.

SMITH.— CARTHAGE AND THE

CARTHAGINIANS. By R. Bosworth Smith, M.A. With Maps, Plans, etc. Crown 8vo, 3s. 6d.

STEPHENS. — A HISTORY OF

THE FRENCH REVOLUTION. By H. Morse Stephens. Svo. Vols. I. and II. ISs. each.

STERNBERG.— MY EXPERI- ENCES OF THE BOER WAR. By Ad- albert Count Sternberg. With Preface by Lieut-Col. G. F. R. Henderson. Crown Svo, 5s. net.

STU BBS.— HISTORY OF THE

UNIVERSITY OF DUBLIN. By J. W. Stubbs. Svo, 12s. 6rf.

STUBBS.— HISTORICAL INTRO- DUCTIONS TO THE * ROLLS SERIES '. By William Stubbs, D.D., formerly Bishop of Oxford. Svo, 12s. 6rf. net.

SUTHERLAND.— THE HISTORY

OF AUSTRALIA AND NEW ZEALAND, from 1606-1900. By Alexander Suther- land, M.A., and George Sutherland, M.A. Crown Svo, 2s. 6rf.

TAYLOR.— A STUDENT S MAN- UAL OF THE HISTORY OF INDIA. By Colonel Meadows Taylor, C.S.I. Crown Svo, 7s. 6i.

THOMSON.— CHINA AND THE

POWERS : a Narrative of the Outbreak of 1900. By H. C. Thomson. With 2 Maps and 29 Illustrations. Svo, 10s. 6d. net.

TODD.— PARLIAMENTARY GOVERNMENT IN THE BRITISH COL- ONIES. By Alpheus Todd, LL.D. Svo, 30s. net.

TREVELYAN.— THE AMERICAN REVOLUTION. By Sir G. O. Trevelyan, Bart. Part I., Svo, I3s. Sd. net. Part II., 2 vols. Svo, 2Is. net.

8

LONGMANS AND CO. S STANDARD AND GENERAL l^ORKS,

History, Politics, Polity, Political Memoirs, etc. — continued.

TREVELYAN. — ENGLAND IN

THE AGE OF WYCLiFFE. By George Macaulay Trevkuyan. 8vo, 15$.

WAKEMAN AND HASSALL.-

BSSAYS INTRODUCTORY TO THE STUDY OF ENGLISH CONSTITU- TIONAL HISTORY. Edited by Henry Ofplby Wakbman, M.A., and Arthur Has- SALU M.A. Crown 8vo, 6s.

WALPOLE. — (Sir Spencer,

HISTORY OF ENGLAND FROM THE CONCLUSION OF THE GREAT WAR IN 1815 TO 1858. 6 vols. Crown 8vo, 6s. each.

THE HISTORY OF TWENTY- FIVE YEARS (1856-1881). Vols. I. and II. 8vo. [In the Press.

W I L L O U G H BY.— POLITICAL

THEORIES OF THE ANCIENT WORLD. By WestelW. WiLLOUOHBY, Ph.D. Crown 8vo, 6s. net.

W I L L S O N. — LEDGER AND

SWORD; or, The Honourable Company of Merchants of England Trading to the Bast Indies (1599-1874). By Beckles Willson. With numerous Portraits and Illustrations. 2 vols. 8vo, 21s. net.

WYLIE (James Hamilton, M.A.).

HISTORY OF ENGLAND UNDER

HENRY IV. 4 vols. Crown 8vo. Vol. L, 1399-1404, 10s. &i. Vol. II., 1405-1406, 15s. (out of print). Vol. III., 1407-1411, 15s. Vol. IV., 1411-1413, 21s.

THE COUNCIL OF CONSTANCE TO THE DEATH OF JOHN HUS. Crown 8vo, 65. net.

Blos^raphy, Personal Memoirs, etc.

BACON.— THE LETTERS AND LIFE OF FRANCIS BACON, INCLUD- ING ALL HIS OCCASIONAL WORKS. Edited by Jambs Spbouing. 7 vols. 8vo,

B A G E H O T.— BIOGRAPHICAL

STUDIES. By Walter Bagehot. Crown 8vo, 3s. M. B A I N.— AUTOBIOG .v APH Y. By Alexander Bain, LL.D., Emeritus Pro- fessor of Logic and English, University of Aberdeen. With Supplementary Chapter. 8vo. [In the Press.

BLOUNT. — THE MEMOIRS OF

SIR EDWARD BLOUNT, K.C.B., ETC. Edited by Stuart J. Reid. With 3 Photogravure Plates. 8vo, 10s. Sd. net.

BOW EN.— EDWARD BOWEN :

|A MEMOIR. Bv the Rev. the Hon. W. E.

BowEN. With Appendices, 3 Photogravure

^Portraits and 2 other Illustrations. 8vo,

^12s. 6d. net.

CARLYLE. — THOMAS CAR-

LYLB : A History of his Life. By James Anthony Froude.

1795-1835. 2 vols. Crown 8vo, 7s. 1834-1881. 2 vols. Crown 8vo, 7s.

COLVILLE.— DUCHESS SARAH:

being the Social History of the Times of Sarah Jennings, Duchess of Marlborough, Compiled and arranged by one of her descen- dants (Mrs. Arthur Colvillb). With 10 Photogravure Plates and 2 other Illustra- tions. 8vo, I8s. net.

CROZIER. — MY INNER LIFE: being a Chapter in Personal Evolution and Autobiography. By John Bbattie Crozibr, LL.D. 8vo, 14s.

DANTE.— THE LIFE AND

WORKS OF DANTE ALLIGHIERI : being an Introduction to the Study of the ' Divina Commedia '. By the Rev. J. F. Hogan, D.D. With Portrait. 8vo, 12s. Sd.

DANTON.— LIFE OF DANTON,

By A. H. Beeslv. With Portraits. Crown 8vo, 6s.

DE BODE.— THE BARONESS DE BODE, 4775-1803. By William S. Childe-Pemberton. With 4 Photogravure Portraits and other Illustrations. 8vo, gilt top, 12s. &d. net.

ERASMUS.

LIFE AND LETTERS OF ERASMUS. By James Anthony Froude. Crown 8vo, 3s. 6^.

THE EPISTLES OF ERASMUS.

From his Earliest Letters to his Fifty-first Year, arranged in Order of Time. English Translations, with a Commentary. By Francis Morgan Nichols. 8vo, ISs. net.

FARADAY.— FARADAY AS A

DISCOVERER. By JohnTyndall. Crown 8vo, 3s. 6d.

F^NELON; his Friends and his Enemies, 1651-1715. By E. K. Sanders. With Portrait. 8vo, 10s. Sd.

FOX.— THE EARLY HISTORY OF

CHARLES JAMES FOX. By the Right Hon. Sir G. O. Trevblyan, Bart. Crown 8vo, 3s. Qd.

FROUDE. - MY RELATIONS WITH CARLYLE. By James Anthony Froude. Together with a Letter Trom the late Sir Jahes Stephen, Bart., K.C.S.I., dated December, 1886. Svo. 2s. net.

!

LONGMANS AND CO. S STANDARD AND GENERAL WORKS.

Biography, Personal Memoirs, etc. — continued.

GREY. — MEMOIR OF SIR

GEORGE GREY, BART., G.C.B., 1799-1882. By Mandbll Creighton, D.D., late Lord Bishop of London. With 3 Portrait«. Crown 8vo, 6s. net.

HAMILTON.— LIFE OF SIR

WILLIAM HAMILTON. By R. P. Graves, 8vo. 3 vols. 155. each. Addendum. 8vo, 6i. sewed.

HARROW SCHOOL REGISTER

(THE), 1801-1900. Edited by M. G. Dauolish, 8vo, 10s. net.

HAVELOCK.— MEMOIRS OF SIR

HENRY HAVELOCK, K.C.B. By John Clark Marshman. Crown 8vo, 3s. 6^.

HAWEIS.— MY MUSICAL LIFE.

By the Rev. H. R. Hawbis. With Portrait of Richard Wagner and 3 Illustrations. Crowo 8vo, 6s. net.

HIGGINS.— THE BERNARDS OF

ABINGTON AND NETHER WINCHEN- DON : A Family History. By Mrs. Napier HiGoiNS. 2 vols. 8vo, 21s. net.

HILEY.— MEMORIES OF HALF A CENTURY. By Richard W. Hilev, D.D., Vicar of Wighill, near Tadcaster, Yorks. 8vo, 15i.

HUNTER.— THE LIFE OF SIR WILLIAM WILSON HUNTER, K.C.S.I., M.A.. LL.D. By Francis Henry Skrine, F.S.S. With 6 Portraits (2 Photogravures) and 4 other Illustrations. 8vo, 16s. net.

JACKSON.— STONEWALL JACK- SON AND THE AMERICAN CIVIL WAR. By Lieut.-Col. G. F. R. Henderson. With 2 Portraits and 33 Maps and Plans. 2 vols. Crown 8vo, 16s. net.

KIELMANSEGGE. — DIARY OF

A JOURNEY TO ENGLAND IN THE YEARS 1761-1762. By Count Frederick KiELMANSBGGE. With 4 Illustrations. Cr. 8vo, 5s. net.

LUTHER.— LIFE OF LUTHER. By Julius Kostlin. With 62 Illustrations and 4 Facsimiles of MSS. Cr. 8vo, 3s. 6^.

MACAULAY.— THE LIFE AND

LETTERS OF LORD MACAULAY. By the Right Hon. Sir G. O. Trevelvan, Bart.

Popular Edition. 1 vol. Cr. 8vo,

2s. 6</. StudcnVs Edition. 1 vol. Cr. 8vo, 65.

Cabinet Edition. 2 vols. Post 8vo,

12s. ' Edinburgh * Edition. 2 vols. 8vo,

6s. each. Library Edition. 2 vols. 8vo, 865.

MARBOT.— THE MEMOIRS OF

THE BARON DE MARBOT. 2 vols. Cr. 8vo, 74,

MAX MOLLER (F.).

THE LIFE AND LETTERS OF THE RIGHT HON. FRIEDRICH MAX MULLER. Edited by his Wife. With Photogravure Portraits and other Illus- trations. 2 vols. 8vo, 32s. net.

MY AUTOBIOGRAPHY: a Frag- ment. With 6 Portraits. 8vo, 125. 6d.

AULD LANG SYNE. Second Series. 8vo, 10s. 6^.

CHIPS FROM A WORKSHOP. Vol. II. Essays. Crown 8vo, 5s.

GERMAN

Biographical

MORRIS. — THE LIFE OF

WILLIAM MORRIS. By J. W. Mackail. With 2 Portraits and 8 other Illustrations by E. H. New, etc. 2 vols. Large Crown 8vo, 10s. net.

ON THE BANKS OF THE

SEINE. Bv A. M. F., Author of * Foreign Courts and Foreign Homes '. Cr. 8vo, 6s.

PAGET.— MEMOIRS AND

LETTERS OF SIR JAMBS PAGET. Edited by Stephen Paget, one of his sons. With Portrait. 8vo, 6s. net.

RkMAKRtSHNA : HIS LIFE

AND SAYINGS. By the Right Hon. F. Max Mt^LLER. Crown 8vo, 5s.

RICH.— MARY RICH, COUNTESS

OF WARWICK (1625-1678): Her Family and Friends. By C. Fell Smith. With 7 Photogravure Portraits and 9 other Illus- trations. 8vo, gilt top, 18s. net.

ROCHESTER, AND OTHER

LITERARY RAKES OP THE COURT OP CHARLES II., WITH SOME AC- COUNT OP THBIR SURROUNDINGS.

By the Author of * The Life of Sir Kenelm Digby,' 'The Life of a Prig,* etc. With 15 Portraits. 8vo, 16s.

ROMANES. — THE LIFE

AND LETTERS OF GEORGE JOHN ROMANES, M A., LL.D., F.R.S. Written and Edited by his Wife. With Portrait and 2 Illustrations. Crown 8vo, 5s. net.

RUSSELL. — SWALLOWFIELD AND ITS OWNERS. By Constance Ladv Russell, of Swallo\* field Park. With 15 Photogravure Portraits and 36 other Illustrations. 4to, gilt edges, 42s. net.

SEEBOHM.-THE OXFORD RE- FORMERS—JOHN COLET, ERASMUS AND THOMAS MORE: a History of their Fellow- Work. By Fred^ic Sbbbqhm, 8vo.l44»

lO LONGMANS AND CO.'s STANDARD AND GENERAL WORKS.

Bio^aphy, Personal Memoirs, etc. — continued.

SHAKESPEARE. — OUTLINES

OF THE LIFE OF SHAKESPEARE. By J.O. Halmwbll-Pmillippb. With Illua- trationa and Facsimiles. 2 vols. Royal 8vo,

215.

TALES OF MY FATHER.- By

A. M. F. Crown 8vo, 6s.

TALLENTYRE.-THE WOMEN

OF THE SALONS, and other French Por- traits. By S. G. Tallbntyrb. With II Phot<^ravure Portraits. 8vo, ICs. 6rf. net.

THOMSON. — EIGHTY YEARS'

REMINISCENCES. By Colonel J. An- 8TRUTHER THOMSON. With 29 Portraits and ot! er Illustrations. 2 vols. 8vo, 21 5. net.

VERNEY.— MEMOIRS OF THE VERNEY FAMILY DURING THE SEVENTEENTH CENTURY. Compiled from the Papers and Illustrated by the Portraits at Claydon House, Bucks. By Frances Parthenope Vernev and Margaret M. Vernev. Abridged and Cheaper Edition. With 24 Portraits. 2 vols. Crown 8vo.

VICTORIA, QUEEN, 1819-190L By Richard R. Holmes, M.V.O., F.S.A. With Photogravure Portrait. Crown 8vo, gilt top, 55. net.

WAL POLE. —SOME UNPUB- LISHED LETTERS OF HORACE WAL- POLE. Edited by Sir Spencer Walpolb. K.C.B. With 2 Portraits. Cr. 8vo, 45. 6<i. net.

WELLINGTON.— LIFE OF THE

DUKE OF WELLINGTON. By the Rev. G. R. Gleig, M.A. Crown 8vo, 3s. 6d.

WILKINS (W. H.).

A QUEEN OF TEARS: Caroline Matilda, Queen of Denmark and Princess of England. Sister of George III. With 47 Portraits and other Illustrations. 2 vols. 8vo, 36s.

THE LOVE OF AN UN- CROWNED QUEbN : Sophie Dorothea Consort of George I., and her Correspon- dence with Philip Christopher, Count Kdnigsmarck. With 24 Portraits and Illustrations. 8vo, 12s. ^. net.

CAROLINE THE ILLUSTRIOUS, Queen-Consort of George II. and sometime Queen- Recent : a Study of Her Life and Time. With 42 Portraits and other Illus- trations. 8vo, 125. 6t2. net.

Travel and Adventure, the Colonies, etc.

ARNOLD.— SEAS AND LANDS. By Sir Edwin Arnold. With 71 Illustra- tions. Crown 8vo, 3s. 6<^.

BAKER (Sir S. W.). EIGHT YEARS IN CEYLON.

With 6 Illustrations. Crown 8vo, 3s. 6<^.

THE RIFLE AND THE HOUND

IN CEYLON. With 6 Illustrations. Cr. Svo. 3s. Bt/.

BALL (John).

THE ALPINE GUIDE. Recon- structed and Revised on behalf of the Alpine Club by W. A B. Coolidge.

Vol. I., The Western Alps :

the Alpine Region, South of the Rhone Valley, from the Col de Tenda to the Simplon Pass. With 9 New and Revised Maps. Crown 8vo, 12s. net.

HINTS AND NOTES, PRAC- TICAL AND SCIENTIFIC, FOR TRA- VELLERS IN THE ALPS : being a Revision of the General Introduction to the * Alpine Guide '. Crown 8vo, 3s. net.

BENT.-THE RUINHD CITIES

OF MASHONALAND : being a Record of Excavation and Exploration in 1891. By J. Theodore Bbnt. With 117 Illustrations. Crown 8vo, 3s. 6rf.

BRASSEY (The Late Lady).

A VOYAGE IN THE 'SUN- BEAM ' : OUR HOME ON THE OCEAN FOR ELEVEN MONTHS.

Cabinet Edition, With Map and 66 Illustrations. Crown 8vo, gilt edges, 7s. 6J.

' Silver Library ' Edition, With 66 Illustrations. Crown 8vo, 3s. 6^.

Popular Edition. With 60 Illustra- tions. 4to, 6d. sewed ; Is. cloth.

School Edition. With 37 Illustra- tions. Fcp., 2s. cloth, or 8s. white parch- ment.

SUNSHINE AND STORM IN

THE EAST

Popular Edition. With 103 Illus- trations. 4to, 6^. sewed ; Is. cloth.

IN THE TRADES, THE TROPICS AND THE • ROARING FORTIES '.

Cabinet Edition. With Map and 220 Illustrations. Cr. Svo, gilt edges, 75. 6d.

COCKERELL. — TRAVELS IN

SOUTHERN EUROPE AND THE LE- VANT, 1810-1817. By C. R. Cockerell. Architect, R.A. Edited by his Son, Samuel Pepvs Cockerell. With Portrait. Svo, IDs, 6d, net.

LONGMANS AND CO*S STANPAAJ> AND GENERAL WORl{$, 11

Travel and Adventure, the Colonies, etc. — continued.

\\

FOUNTAIN (Paul). THE GREAT DESERTS AND FORESTS OF NORTH AMERICA. 8vo, 9s. 6d. net.

THE GREAT MOUNTAINS AND FORESTS OF SOUTH AMERICA. With Portrait and 7 Illustrations. 8vo,105. 6^. neA

THE GREAT NORTH-WEST AND THE GREAT LAKE REGION OF NORTH AMERICA. 8vo, 10s. 6^. net.

FROUDE (Jambs A.). OCEANA: or England and her Colonies. With 9 Illustrations. Crown 8vo, 3s. &;.

THE ENGLISH IN THE WEST INDIES : or, the Bow of Ulysses. With 9 Illustrations. Crown 8vo, 2s. boards ; 2s. M. cloth.

GROVE.— SEVENTY-ONE DAYS' CAMPING IN MOROCCO. By Lady Grove. With Photogravure Portrait and 32 Illustrations from Photographs. 8vo, 7s. ^. net.

HAGGARD.— A WINTER PIL- GRIMAGE: Being an Account of Travels through Palestine, Italy and the Island of Cyprus, undertaken in the year 1900. By H. Rider Haggard. With 31 Illustrations from Photographs.

HARDWICK.— AN IVORY

TRADER IN NORTH KENIA : the Re- cord of an Expedition to the Country North of Mount Kenia in East Equatorial Africa, with an account of the Nomads of Galla- Land. By A. Arkell-Hardwick, F.R.G.S. With 23 Illustrations from Photographs, and a Map. 8vo, 12s. 6<:/. net.

HOWITT.— VISITS TO REMARK- ABLE PLACES. Old Halls, Battle-Fields. Scenes, Illustrative of Striking Passages in English History and Poetry. By William HowiTT. With 80 Illustrations. Crown 8vo, 3s. M.

KNIGHT (E. R). SOUTH AFRICA AFTER THE WAR. With 17 Illustrations. 8vo, 10s. %d. net.

WITH THE ROYAL TOUR: a

Narrative of the Recent Tour of the Duke and Duchess of Cornwall and York through Greater Britain. With 16 Illustrations and a Map. Crown 8vo, 5s. net. THE CRUISE OF THE * ALERTE ' : the Narrative of a Search for Treasure on the Desert Island of Trini- dad. With 2 Maps and 23 Illustrations. Crown 8vo, 3s. 6</.

WHERE THREE EMPIRES MEET: a Narrative of Recent Travel in Kashmir, Western Tibet, Baltistan, Ladak, Gilgit, and the adjoining Countries. With a Map and 54 Illustrations. Crown 8vo, 35. <k<.

KNIGHT (E. ^ .)-^ontinued. THE *FALCON' ON THE BALTIC: a Voyage from London to Copenhagen in a Three Tonner. With 10 Full-page Illustrations. Cr. 8vo, 3s. %d.

LEES AND CLUTTERBUCK.

— B.C. 1887: A RAMBLE IN BRITISH COLUMBIA. By J. A Lees and W. J. ClUtterbuck. With Map and 75 Illustra- tions. Crown 8vo, 3s. 6^.

LYNCH.— ARMENIA: Travels and Studies. By H. F. B. Lynch. With 197 Illustrations (some in tints) reproduced from Photographs and Sketches by the Author, 16 Maps and Plans, a Bibliography, and a Map of Armenia and adjacent countries. 2 vols. Medium 8vo, gilt top, 42s. net.

NANSEN.— THE FIRST CROSS- ING OF GREENLAND. By Fridtjof Nansen. With 143 Illustrations and a Map. Crown 8vo, 3s. 6d.

R I C E.— OCCASIONAL ESSAYS

ON NATIVE SOUTH INDIAN LIFE. By Stanley P. Rice, Indian Civil Service. 8vo, 10s. 6^.

SMITH. — CLIMBING IN THE

BRITISH ISLES. By W. P. Haskett, Smith. With Illustrations and Numerous Plans.

Part I. England. 16mo. 3s. net. Part II. Wales and Ireland.

16mo, 3s. net.

SPENDER.— TWO WINTERS IN

NORWAY : being an Account of Two Holi- days spent on Snow-shoes and in Sleigh Driving, and including an Expedition to the Lapps. By A. Edmund Spender. With 40 Illustrations from Photographs. 8vo, lOs. 6^. net.

STEPHEN.— THE PLAY-

GROUND OF EUROPE (The Alps). By Sir Leslie Stephen, K.C.B. With 4 Illustrations. Crown 8vo, 3s. 6<i.

STUTFIELD AND COLLIE.—

CLIMBS AND EXPLORATIONS IN THE CANADIAN ROCKIES. By Hugh E. M. SruTPiELD and J. Norman Collie, F.R.S. With 2 Maps, 24 Full-page Illustrations, and 56 Half-page Illustrations. 8vo, 12s. 6d. net.

SVERDRUP. — NEW LAND:

being a Record of the Voyage of the Fram to the Arctic Regions. 1898-1»02. By Captain Otto Sverdrup. With Maps and Illustra- tions. 8vo.

THREE IN NORWAY. By Two

of Them. With a Map and 59 Illustrations. Crown 8vo, 2s. boards, 2s. 6d. cloth.

TYNDALL (John).

THE GLACIERS OF THE ALPS.

With 61 Illustrations. Cr. 8vo, 6s. 6i. net.

HOURS OF EXERCISE IN THE

ALPS. With 7 Illustrations. Crown 8vo,

6s. 6f<. net.

12 LONCMAlfS AND CO. S STANDARD AND GENERAL JVOR/CS,

Sport and Pastime.

THE BADMINTON LIBRARY.

Edited by His Grace the (Eighth) DUKE OF BEAUFORT, K.G.,

and A. B. T. WATSON.

ARCHERY. By C. J. Longman and

Col. H. Walrono. With Contributions by Miss Lbgh, Viscount Dillon, etc. With 2 Maps, 23 Plates and 172 Illustrations in the Text. Crown 8vo, cloth, 6s. net: half- bound, with gilt top, 9s. net.

ATHLETICS. By Montague

Shearman. With Chapters on Athletics at School by W. Bbachbr Thomas; Athletic Sports in America by C. H. Shbrrill; a Contribution on Paper-chasing by W. Rye, and an Introduction by Sir Richard Webster (Lord Alvbrstgne). With 12 Plates and 37 Illustrations in the Text. Cr. 8vo, cloth, 6s. net ; half bound, with gilt top, 9s. net.

BIG GAME SHOOTING. By

Clivb Phillipps-Wgllby.

Vol. I. Africa and America.

With Contributions by Sir Samuel W. Baker, W. C. Oswbll, F. C. Selous, etc. With 20 Plates and 57 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half- bound, with gilt top, 9s. net.

Vol. II. Europe, Asia, and the

ARCTIC REGIONS. With Contributions by Lieut.-Colonel R. Heber Percy, Major Algernon C. Heber Percy, etc. With 17 Plates and 56 Illustrations in the Text. Crown 8vo, cloth 6s. net ; half bound, with gilt top, 9s. net.

BILLIARDS. By Major W. Broad-

POOT, R.E. With Contributions by A. H. Bgyd, Sydenham Dixon, W. J. Pgrd, etc. With 11 Plates. 19 Illustrations in the Text, and numerous Diagrams. Crown 8vo, cloth, 6s. net ; halF-bound, with gilt top, 9s. net.

COURSING AND FALCONRY.

By Harding Cox, Charles Richardson, and the Hon. Gerald Lascelles. With 20 Plates and 55 Illustrations in the Text. Crown 8vo, cl )th 6s. net ; half-bound, with gilt top, 98. net.

CRICKET. By A. G. Steel and the Hon R. H. Lyttelton. With Contributions by Andrew Lang, W. G. Grace, P. Gale, etc. With 13 Plates and 52 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half- bound, with gilt top, 9s. net.

CYCLING. By the Earl of Albe- marle and G. Lacy Hillier. With 19 Plates and 44 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

DANCING. By Mrs. Lilly Grove-

With Contributions by Miss Middlbton* The Hon. Mrs. Armytage, etc. With Musical Examples, and 38 Full-page Plates and 93 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

DRIVING. By His Grace the

(Eighth) Duke op Beaufort, K.G. With Contributions by A. B. T. Watson, the Earl op Onslow, etc. With 12 Plates and 54 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

FENCING, BOXING, AND WRESTLING. By Walter H. Pollo k, F. C. Grove, C. Prevost, E. B. Mitchell, and Walter Armstrong. With 18 Plates and 24 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half- bound, with gilt top, 9s. net.

FISHING. By H. Cholmondeley- Pbnnell.

Vol. I. Salmon and Trout. With

Contributions by H. R. Francis, Major John P. Traherne, etc. With 9 Plates and numerous Illustrations of Tackle, etc. Crown 8vo, cloth, 65. net : half-bound, with gilt top, 9s. net.

Vol. II. Pike and other Coarse

Fish. With Contributions by the Mar- quis OF Exeter, William Senior. G. Christopher Davis, etc. With 7 Plates and numerous Illustrations of Tackle, etc. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

FOOTBALL. History, by Mon- tague Shearman ; The Association Game, by W. J. Oakley and G. O. Smith ; The Rugby Union Game, by Frank Mitchell. With other Contributions by R. E. Mac- naghten, M. C. Kemp, J. E. Vincent, Walter Camp and A. Sutherland. With 19 Plates and 35 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

GOLF. By Horace G. Hutchinson.

With Contributions by the Rt. Hon. A. J. Balfour, M.P., Sir Walter Simpson, Bart.. Andrew Lang, etc. With 34 Plates and 56 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 95. net.

LONGMANS AND CO»S STANDARD AND GENERAL WORKS. 1 3

Sport and Pastime — continued. THE BADMINTON \A^^\^S— continued.

Edited by His Grace the (Eighth) DUKE OF BEAUFORT, K.G.,

and A. E. T. WATSON.

HUNTING. By His Grace the (Eighth) Duke op Beaufort, K.G., and Mowbray Morris. With Contributions by the Earl of Suffolk and Berkshire, Rev. E. W. L. Davies, G. H. Longman, etc. With 5 Plates and 54 Illustrations in the Text. Crown 8vo, cloth, 65. net ; half- bound, with gilt top, 9s. net.

MOTORS AND MOTOR-DRIV- ING. By Alfred C. Harmsworth, the Marquis db Chassbloup - Laubat, the Hon. John Scott-Moniaou, R. J. Me- crbdy, the Hon. C. S. Rolls, Sir David Salomons, Bart., etc. With 13 Plates and 136 Illustrations in the Text. Crown 8vo, cloth, 95. net ; half-bound, 125. net. A Cloth Box for use when Motor- ing, 25. net.

MOUNTAINEERING. By C. T. Dent. With Contributions by the Right Hon. J. Brycb, M.P., Sir Martin Conway, D. W. Freshfield, C. E. Matthews, etc. With 13 Plates and 91 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

POETRY OF SPORT (THE).—

Selected by Hedley Peek. With a Chapter on Classical Allusions to Sport by Andrew Lang, and a Special Preface to the BAD- MINTON LIBRARY by A. E. T. Watson. With 32 Plates and 74 Illustrations in the Text. Crown Svo, cloth, 6s. net ; half-bound, with gilt top, 95. net.

RACING AND STEEPLE-CHAS- ING. By the Earl of Suffolk and Berkshire, W. G. Craven, the Hon. F. Lawley, Arthur Coventry, and A. E. T. Watson. With Frontispiece and 56 Illus- trations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

RIDING AND POLO. By Captain Robert Weir, J. Moray Brown, T. F. Dale, The Late Duke of Beaufort, The Earl of Suffolk and Berkshire, etc. With 18 Plates and 41 Illustrations in the Text. Crown 8vo, cloth, 6$. net ; half-bound, with gilt top, 95. net.

ROWING. By R. P. P. Rowe and C. M. Pitman. With Chapters on Steering by C. P. Serocold and F. C. Begg ; Metro- politan Rowing by S. Le Blanc Smith ; and on PUNTING by P. W. Squire. With 75 Illustrations. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

SEA FISHING. By John Bicker- dyke, Sir H. W. Gore-Booth, Alfred C. Harmsworth, and W. Senior. With 22 Full-page Plates and 175 Illustrations in the Text. Crown 8vo, cloth, 65. net ; half- bound, with gilt top, 95. net.

SHOOTING.

Vol. I. Field and Covert. By Lord Walsingham and Sir Ralph Payne-Gallwey, Bart. With Contribu- tions by the Hon Gerald Lascellbs and A. J. Stuart- Wortley. With 11 Plates and 95 Illustrations in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

Vol. II. Moor and Marsh. By Lord Walsingham and Sir Ralph Payne-Gallwey, Bart. With Contribu- tions by Lord Lovat and Lord Charles Lennox Kerr. With 8 Plates and 57 Illustrations in the text. Crown 8vo, cloth, 65. net ; half-bound, with gilt top, 95. net.

SKATING, CURLING, TOBOG- GANING. By J. M. Heathcote, C. G. Tebbutt, T. Maxwell Witham, Rev. John Kerr, Ormond Hake, Henry A. Buck, etc. With 12 Plates and 272 Illustra- tions in the Text. Crown 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

SWIMMING. By Archibald Sin-

CLAIR and William Henry, Hon. Sees, of the Life-Saving Society. With 13 Plates and 112 Illustrations in the Text. Cr. 8vo, cloth, 6s. net ; half-bound, with gilt top, 9s. net.

TENNIS, LAWN TENNIS, RACKETS AND FIVES. By J. M. and C. G. Heathcote, E. O. Plbydbli^Bouvbrib, and A. C. Ainger. With Contributions by the Hon. A. Lyttelton, W. C. Marshall, Miss L. DoD, etc. With 14 Plates and 65 Illustrations in the Text. Crown Svo, cloth, 65. net ; half-bound, with gilt top, 9s. net.

YACHTING.

Vol. I. Cruising, Construction of yachts, yacht racing rules, fitting-out, etc. By Sir Edward Sullivan. Bart., The Earl of Pembroke, Lord Brassey, K.C.B., C. E. Seth-Smith, C.B., G. L. Watson, R. T. Pritchbtt, E. F. Knight, etc. With 21 Plates and 93 Illustrations in the Text. Crown Svo, cloth, 6s. net; half-bound, with gilt top, 9s. net.

Vol. II. Yacht Clubs, Yachting

IN AMERICA AND THF. COLONIES, YACHT

RACING, etc. By R. T. Pritchett, The Marquis of Duffbrin and Ava, K.P., The Earl of Onslow, James McFerran, etc. With 35 Plates and 160 Illustrations in the Text. Crown Svo, cloth, 95. net; half-bound, with gilt top, 9s. net.

14 LONGMANS AND CO. S STANDARD AND GENEkAL WOkKS.

Sport and Pastime — continued, FUR, FEATHER, AND FIN SERIES.

Edited by A. E. T. Watson.

Crown 8vo, price &. each Volume, cloth.

*^* Tke Volumes are also issued half-bound in Leather, with gilt top.

Price 7s. 6d, net each.

THE PARTRIDGE. Natural His- tory, by the Rev. H. A. Macphbrson ; Shooting, by A. J. Stuart - Wortlby ; Cookery, by Gborgb Saintsbury. With 11 Illustrations and various Diagrams. Crown 8vo, 5s.

THE GROUSE, Natural History, by the Rev. H. A. Macphbrson; Shooting, by A. J. Stuart-Wortlby; Cookery, by Gborob Saintsbury. With 13 Illustraiions and various Diagrams. Crown 8vo, Ss.

THE PHEASANT. Natural History, by the Rev. H. A. MacphiiRson ; Shooting, by A. J. Stuart-Wortlby ; Cookery, by Alexandbr Innbs Shand. With 10 Illus- trations and various Diagrams. Crown 8vo, 5s.

THE HARE. Natural History, by the Rev. H. A. Macphbrson ; Shootmg, by the Hon. Gerald Lascbllbs ; Coursing, by Charles Richardson ; Hunting, by J. S. Gibbons and G. H. Longman; Cookery, by Col. Kbnnby Herbert. With 9 Illustra- tions. Crown 8\ o, 5s.

THE RABBIT. By James Edmund Hartino. Cookery, by Alexander Innbs Shand. With 10 Illustrations. Crown 8vo, 5s.

SNIPE AND WOODCOCK. By L. H. De Vismb Shaw. With Chapters on Snipe and Woodcock in Ireland by Richard J. Usshbr. Cookery, by Alexander Innbs Shand. With 8 Illustrations. Crown 8vo, 5s.

RED DEER. Natural History, by the Rev. H. A. Macphbrson ; Deer Stalk- ing, by Cameron op Lochibl: Stag Hunt- ing, by Viscount Bbrington ; Cookery by Albxandbr Innbs Shand. With 10 Illus- trations. Crown 8vo, 5s.

THE SALMON. By the Hon. A. E. Gathornb-Hardy. With Chapters on the Law of Salmon Fishing by Claud Douglas Pennant; Cookery, by Alexander Innbs Shand. With 8 Illustrations. Crown 8vo, 5s.

THE TROUT. By the Marquess

op Granby. With Chapters on the Breed- ing of Trout by Col. H. Custancb; and Cookery, by Alexander Innes Shand. With 12 Illustrations. Crown 8vo, 5s.

PIKE AND PERCH. By William Senior ('Redspinner,' Editor of the 'Field'). With Chapters by John Bickbroykc and W. H. Pope : Cookery, by Alexander Innbs Shand. With 12 Illustrations. Crown 8vo, 5s.

ALVERSTONE AND ALCOCK.

SURREY CRICKET : its History and Asso- ciations. Edited by the Right Hon. Lord Alverstone, L.C.J., President, and C. W. Alcock, Secretary, of the Surrey County Cricket Club. With 48 Illustrations. 8vo, 5s. net.

BICKERDYKE.— DAYS OF MY

LIFE ON WATER, FRESH AND SALT; and other Papers. By John Bickbrdykb. With Photo-etching Frontispiece and 8 Full- page Illustrations. Crown 8vo, 8s. 6^.

BLACKBURNE.— MR. BLACK-

BURNE'S GAMES AT CHESS. Selected. Annotated and Arranged by Himself. Edited, with a Biographical Sketch and a brief His- tory of Blindfold Chess, by P. Anderson Graham. With Portrait of Mr. Blackbume. 8vo, 7s. 6^. net.

ELLIS.— CHESS SPARKS ; or,

Short and Bright Games of Chess. Collected and Arranged by J. H. Ellis, M.A. 8vo, 4s. M.

LONGMANS AND CO. S STANDARD ANp GENERAL WORKS. 1 5

Sport and Pastime — continued.

FOR D.— THE THEORY AND PRACTICE OF ARCHERY. By Horace Ford. New Edition, thoroughly Revised and Re-written by W. Butt, M.A. With a Preface by C. J. Longman, M.A. 8vo, 145.

FRANCIS.-^A BOOK ON

ANGLING : or. Treatise on the Art of Ftsh- injg in every Branch ; including full Illustrated List of Salmon Flies. By Francis Francis. With Portrait and Coloured Plates. Crown 8vo, 155.

FREMANTLE.— THE BOOK OF

THE RIFLE. By the Hon.T. F.Fremantle, V.D., Major. 1st Bucks V.R.C. With 54 Places and 107 Diagrams in the Text. 8vo. 12s. ^. net.

GATHORNE-HARD Y.—

AUTUMNS IN ARGYLESHIRE WITH ROD AND GUN. By the Hon. A. E. Gathornb-Hardy. With 8 Illustrations by Archibald Thorburn. 8vo, 65. net.

G R A H A M.— COUNTRY PAS- TIMES FOR BOYS. By P. Anderson Graham. With 252 Illustrations from Draw- ings and Photographs. Crown 8vo, gilt edges, Ss. net.

HUTCHINSON.— THE BOOK OF

GOLF AND GOLFERS. By Horace G. Hutchinson. With 71 Portraits from Photographs. Large crown 8vo, gilt top, 75. 6./. net.

LANG.— ANGLING SKETCHES.— By Andrew Lang. With 20 Illustrations. Crown 8vo, 35. M.

L I L L I E.— CROQUET UP TO

DATE. Containing the Ideas and Teachings of the Leading Players and Champions. By Arthur Lillie. With 19 Illustrations (15 Portraits), and numerous Diagrams. 8vo, IO5. 6d. net.

L O N G M A N.— CHESS OPEN- INGS. By Frederick W. Longman. Fcp. 8vo, 25. 6d.

MACKENZIE.— NOTES FOR

H UNTING MEN. By Captoin Cortlandt Gordon Mackenzie. Crown 8vo, 2$. %d. net.

MADDEN.— THE DIARY OF MASTER WILLIAM SILENCE: a Study of Shakespeare and of Elizabethan Sport. By the Right Hon. D. H. Madden. Vice-Chan- cellor of the University of Dublin. 8vo, gilt top. 165.

MASKELYNE. — SHARPS AND

PLATS : a Complete Revelation of the Secrets of Cheating at Games of Chance and Skill. By John Nevil Maskelyne. of the Egyptian Hall. With 62 Illustrations. Crown 8vo, 65.

MILLAIS (John Guillb).

r THE WH^D-FOWLER IN SCOT- LAND. With a Frontispiece in Photo- gravure by Sir J. E. MiLLAiP, Bart., P.R.A., 8 Photogravure Plates. 2 Coloured Plates and 50 Illustrations from the Author's Drawings and from Photographs. Royal [4to, gilt top, 90s. net.

ITHE NATURAL HISTORY OF THE BRITISH SURFACE-FEEDING DUCKS. With 6 Photogravures and 66

^^ Plates (41 in Colours) from Drawings by

^^the Author, Archibald Thorburn, and from Photographs. Royal 4to, cloth, gilt

■- top, £6 6s. net.

MODERN BRIDGE. By * Slam '.

With a Reprint of the Laws of Brdge, as f^dopted by the Portland and Turf Clubs. 18mo, gilt edges, 3s. 6^. net

PARK.— THE GAME OF GOLF. -«By William Park, Jun., Champion Golfer,

1887-89. With 17 Plates and 26 Illustrations

in the Text. Crown 8vo, 7s. 6d.

PAYNE^GALLWEY (Sir Ralph,

Bart.).

THE CROSS BOW : Medieval and Modern : Military and Sporting ; its Con- struction, History and Management, with a Treatise on the Balista and Catapult of the Ancients. With 220 Illustrations. Royal 4to, £3 3s. net.

LETTERS TO YOUNG SHOOTERS (First Series). On the Choice and use of a Gun. With 41 Illustrations. Crown 8vo, 75. 6^.

LETTERS TO YOUNG SHOOTERS (Second Series). On the Production, Preservation and K lling of Game. With Directions in Shooting Wood-Pigeons and Breaking-in Retrievers. With Portrait and 103 Illustrations. Crown 8vo, 125. 6<i.

LETTERS TO YOUNG SHOOTERS (Third Series). Compris- ing a Short Natural History of the Wild- fowl that are Rare or Common to the British Islands, with complete directions in Shooting Witd''owl on the Coast and Inland. With 200 Illustrations. Crowq 8vo, 18s

1 6 LONGMANS AND CO.'s STANDARD AND GENERAL IVORKS.

Sport and Pastime — continued.

POLE.— THE THEORY OF THE

MODERN SCIENTIFIC GAME OF WHIST. By William Pole, F.R.S. Fcp, 8vo, gilt edges, 2s. net.

PROCTOR.— HOW TO PLAY

WHIST: WITH THE LAWS AND ETIQUETTE OF WHIST. By Richard A. Proctor. Crown 8vo, gilt edges, 3s. net

RONALDS. — THE FLY-

FISHER'S ENTOMOLOGY. By Alfred Ronalds. With 20 coloured Plates. 8vo. Us.

SOMERVILLE.- SLIPPER'S

A B C OF FOX-HUNTING. By E. (E. SoMBRViLLB, M.F.H., Joint Author of ' Some Experiences of an Irish R.M.,* etc. With Illustrations in Colour by the Author. 4to, boards, 10s. %d. net.

T H O M A S-S T A N F O R D.— A

RIVER OF NORWAY: being the Notes and Reflections of an Angler. By Charles TnoMAS-SrANPORD. With 10 Photc^ravure Plates, I Map and 1 Plan. 8vo, 9s. net.

THOiyiPSON, CANNAN AND

DONERAILB.— COMBINED HAND-IN- HAND FIGURE SKATING. By NoR- CLIPPB G. Thompson, F. Laura Cannan and Viscount Doneraile, Members of the Skating Club. 16mo.

WARNER.— CRICKET ACROSS

THE SEAS : being an Account of the Tour of Lord Hawke's Team in New Zealand and Australia. By P. F. Warner. With 32 I llustrations from Photographs. Crown 8vo, 5s. net.

«

Mental, Moral and Political Philosophy.

LOGIC, RHETORIC^ PSYCHOLOGY, ETHICS, ETC.

ABBOTT.— THE ELEMENTS OF LOGIC. By T. K. Abbott, B.D. 12mo, 3s.

ARISTOTLE,

THE ETHICS: Greek Text, Illus- trated with Essay and Notes. By Sir Alexander Grant, Bart. 2 vols. 8vo, 32s.

AN INTRODUCTION TO ARIS- TOTLE'S ETHICS. Books I.-IV. (Book X. c. vi.-ix. in an Appendix). With a continuous Analysis and Notes. By the Rev. E. Moore, D.D. Crown 8vo, lOs. 6d.

BACON (Francis).

COMPLETE WORKS. Edited by R. L. Ellis, James Spbdding and D. D. Heath. 7 vols. Bvo, £3 13s. 6d.

LETTERS AND LIFE, including

all his occasional Works. Edited by Jambs Spedding. 7 vols. 8vo, £4 4s.

THE ESSAYS: with Annotations. By Richard Whately D.D. 8vo, lOs. 6d.

BACON (Francis) — continued.

THE ESSAYS: with Notes. By P. Storr and C. H. Gibson. Cr. 8vo, Ss.ed.

THE ESSAYS : with Introduction, Notes, and Index. By B. A. Abbott, D.D. 2 vols. Fcp. 8vo, 6s. The Text and Index only, without Introduction and Notes, in One Volume. Fcp. 8vo, 2s. 6d.

BAIN (Alexander).

MENTAL AND MORAL SCIENCE : a Compendium of Psychology and Ethics. Crown 8vo, 10s. Sd.

Or separately, .

Part I. Psychology and His- tory OF Philosophy. Crown 8vo, 6s. Sd.

Part II. Theory of Ethics and

Ethical Systems. Crown 8vo, 45. 6d.

LOGIC. Part I. Deduction. Cr. 8vo, 4s. Part II. Induction. Cr, 8vo, 6s. Qd,

LONGMANS AND CO, S STANDARD AND GENERAL WORKS,

17

Mental, Moral and Political Philosophy — continued,

HODGSON (Shad WORTH H.y^

BAIN (Alexander) — continued.

THE SENSES AND THE IN- TELLECT. 8vo, 15s.

THE EMOTIONS AND THE WILL. 8vo, 15s.

PRACTICAL ESSAYS. Cr. 8vo, 2s.

DISSERTATIONS ON LEADING PHILOSOPHICAL TOPICS. 8vo, 7s. 6£/. net.

BALDWIN.— A COLLEGE MAN- UAL OF rhetoric. By Chakles Sears Baldwin, A.M., Ph.D. Crown 8vo, 4s. 6(/.

BROOKS.— THE ELEMENTS OF

MIND : being an Examination into the Nature of the First Division of the Ele- mentary Substances of Life. By H. Jamyn Brooks. 8vo, 10s. ^. net.

B ROUGH.— THE STUDY OF

MENTAL SCIENCE: Five Lectures on the Uses and Characteristics of Logic and Psychology. By J. Brough, LL.D. Crown 8vo, 25. net.

CROZIER (John Beattie), CIVILISATION AND PRO

GRESS: being the Outlines of a New System of Political, Religious and Social Philosophy. 8vo, 14s.

HISTORY OF INTELLECTUAL DEVELOPMENT: on the Lines of Modern Evolution.

Vol. 1. 8vo, 145.

Vol. II. (In preparation.)

Vol. III. 8vo, 10s. Gel.

F I T E. — AN INTRODUCTORY

STUDY OF ETHICS. By Warner Fite. Crown 8vo, 6s. 6d.

GREEN (Thomas Hill).— THE

WORKS OF. Edited by R. L. Nettleship.

Vols. I. and II. Philosophical Works. 8vo, 16s. each.

Vol. III. Miscellanies. With Index to the three Volumes, and Memoir. 8vo, 21s.

LECTURES ON THE PRIN- CIPLES OF POLITICAL OBLIGA- TION. With Preface by Bernard Bosanquet. 8vo, 5s.

GURNHILL.— THE MORALS OF

SUICIDE. By the Rev. J. Gurnhill, B.A. Vol. 1., Crown 8vo, 5s. net. Vol. II., Crown 8vo, 5>s. net.

TIME AND SPACE: a Metaphy-

' sical Essay. 8vo, 16s.

I THE THEORY OF PRACTICE:

an Ethical Inquiry. 2 vols. 8vo, 24s.

THE PHILOSOPHY OF FLECTION. 2 vols. 8vo. 21s.

RE-

THE METAPHYSIC OF EX- PERIENCE. Book I. General Analysis of Experience; Boole II. Positive Science ; Book III. Analysis of Conscious Action; Book IV. The Real Universe. 4 vols. 8vo, 365. net.

HUME.— THE PHILOSOPHICAL WORKS OF DAVID HUME. Edited by T. H. Green and T. H. Grose. 4 vols. 8vo, 28s. Or separately, Essays. 2 vols. 14s. Treatise of Human Nature. 2 vols. 14s.

JAMES (William, M.D., LL.D.). THE WILL TO BELIEVE, and

Other Essays in Popular Philosophy. Crown 8vo, 7s. 6d.

THE VARIETIES OF RELI- GIOUS EXPERIENCE: a Study in Human Nature. Being the Gifford Lectures on Natural Religion delivered at Edinburgh in 1901-1902. 8vo, 12s. net.

TALKS TO TEACHERS ON PSY- CHOLOGY, AND TO STUDENTS ON SOME OF LIFE'S IDEALS. Crown 8vo, 4s. 6d.

JUSTINIAN.— THE INSTITUTES

OF JUSTINIAN: Latin Text, chiefly that of Huschke, with English Introduction, Translation, Notes and Summary. By Thomas C. Sanuars, M.A. 8.'o, I85.

KANT (Immanuel).

CRITIQUE OF PRACTICAL REASON, AND OTHER WORKS ON THE THEORY OF ETHICS. Trans- lated by T. K. Abbott, B.D. With Memoir. 8vo, 12s. 6d.

FUNDAMENTAL PRINCIPLES OF THE METAPHYSIC OF ETHICS. Translated by T. K. Abbott, B.D. Crown 8vo, 3s.

INTRODUCTION TO LOGIC, AND HIS ESSAY ON THE MIS- TAKEN SUBTILTY OF THE FOUR FIGURES. Translated by T. K. Abbott. 8vo, Ss.

1 8 LONGMANS AND CO.'s STANDARD AND GENERAL IVORKS.

Mental, Moral and Political Philosophy — continued.

KELLY.— GOVERNMENT OR HUMAN EVOLUTION. By Bdmond Kblly, M.A., P.G.S. Vol. I. Justice. Cr. 8vo, 75. M. net. Vol. II. Collectivism and i Individualism. Crown 8vo, 105. %d. net.

K I L L I C K.—HANDBOOK TO

MILL'S SYSTEM OF LOGIC. By Rev. A. H. KiLUCK, M.A. Crown 8vo, as. %d.

LADD (Gborgb Trumbull).

PHILOSOPHY OF CONDUCT: a

Treatise of the Pacts, Principles and Ideals I of Ethics. 8vo, 2l5.

ELEMENTS OF PHYSIO-! LOGICAL PSYCHOLOGY. 8vo, 2l5. 1

OUTLINES OF DESCRIPTIVE,

PSYCHOLOGY : a Text-Book of Mental | Science for Collies and Normal Schools. ' 8vo, 12s. I

OUTLINES OF PHYSIO-I LOGICAL PSYCHOLOGY. 8vo, 12s.

PRIMER OF PSYCHOLOGY. Cr. ! 8vo, 55 6(i.

LECKY (William Edward Hart-'

POLE).

THE MAP OF LIFE: Conduct, and Character. Crown 8vo, Ss. net.

HISTORY OF EUROPEAN MORALS FROM AUGUSTUS TO CHARLEMAGNE. 2 vols. Crown 8vo, 105. net.

A SURVEY OF ENGLISH ETHICS : being the first chapter of W. B. H. Lecky's ' History of European Morals'. Edited, with Introduction and Notes, by W. A. Hirst. Crown 8vo, 35. %d.

HISTORY OF THE RISE AND INFLUENCE OF THE SPIRIT OF RATIONALISM IN EUROPE. 2 vols. Crown 8vo, lOs. net.

DEMOCRACY AND LIBERTY. Library Edition. 2 vols. 8vo, 36s. Cabinet Edition. 2 vols. Cr. 8vo, 105. net.

LUTOSLAWSKL— THE ORIGIN

AND GROWTH OF PLATO'S LOGIC. With an Account of Plato's Style and of the Chronology of his Writings. By Wincenty LUTOSLAWSKI. 8vo, 21s.

MAX MOLLER (F.).

THE SCIENCE OF THOUGHT. 8vo, 21s.

THE SIX SYSTEMS OF INDIAN

PHILOSOPHY. Crown Svo, 7s. 6^/. net.

THREE LECTURKS ON THE

VEDANTA PHILOSOPHY. Cr. Svo, 5s.

MILL (John Stuart).

A SYSTEM OF LOGIC. Cr. Svo, 3s. &/.

ON LIBERTY. Cr. Svo, U. Ad.

CONSIDERATIONS ON REPRE- SENTATIVE GOVERNMENT. Crown 8vo, 25.

UTILITARIANISM. Svo, 2s. 6rf.

EXAMINATION OF SIR WIL- LIAM HA.MILTON'S PHILOSOPHY. 8vo, 16s.

NATURE, THE UTILITY OF RELIGION AND THEISM. Three Essays. 8vo, Ss.

MONCK.— AN INTRODUCTION TO LOGIC. By William Henry S. Monck, M.A. Crown Svo, 5s.

M Y E R S.— HUMAN PERSON- ALITY AND ITS SURVIVAL OF BODILY DEATH. By Frederic W. H. Mvbrs. 2 vols. Svo, 425. net.

PIERCE.— STUDIES IN AUDI- TORY AND VISUAL SPACE PERCEP- TION : Essays on Experimental Psychology. By A. H. Pierce. Crown Svo, 6s. 6^. net.

RICHMOND.— THE MIND OF A CHILD. By Ennis Richmond. Cr. Svo, 3s. 6^. net.

ROMANES.— MIND AND MO- TION AND MONISM. By George John Romanes. Crown Svo, 45. 6<f.

SULLY (James). ^

AN ESSAY ON LAUGHTER : its Forms, its Cause, its Development and its Value. Svo, 125. %d. net

THE HUMAN MIND; a Text- book of Psychology. 2 vols. Svo, 2l5.

OUTLINES OF PSYCHOLOGY.

Crown Svo, 95.

THE TEACHER'S HANDBOOK OF PSYCHOLOGY. Crown Svo, 65. 6rf.

STUDIES OF CHILDHOOD. Svo,

12s. %d. net.

CHILDREN'S WAYS: being Selections from the Author's ' Studies of Childhood'. With 25 Illustrations. Crown Svo, 4s. %d.

SUTHERLAND.— THE ORIGIN

AND GROWTH OF THE MORAL IN- STINCT. By Alexander Sutherland, M.A. 2 vols. Svo, 2Ss.

S W I N B U R N E.— P I C T U R E

LOGIC: an Attempt to Popularise the Science of Reasoning. Bv Alfred James Swinburne, M.A. W'tb ^ Wo9<lcuts, Crown Svo, 25. 6rf,

LONGMANS AND CO.*S STANDARD AND GENERAL IVORKS, I9

Mental, Moral and Political Philosophy — continued, T H O M A S.— INTUITIVE SUG- 1 ZELLER (Dr. Edward).

GESTION. By J. W. Thomas, Author of ' Spiritual Law in the Natural World,' etc. Crown 8vo, 3s. %d. net.

WEBB.— THK VEIL OF ISIS: a Series of Essays on Idealism. By Thomas E. Webb, LL.D., Q.C. 8vo, 10s. 6d.

WEBER.— HISTORY OF PHIL- OSOPHY. By Alfred Weber, Professor in the University of Strasburg. Translated by Frank Thilly, Ph.D. 8vo, 16s.

WHATELY (Archbishop).

With Anno-

BACON'S ESSAYS.

tations. Svo, lOs. %d.

ELEMENTS OF LOGIC. 8vo, 4s. %d.

Crown

ELEMENTS OF RHETORIC. Cr.

8vo, 45. %d.

THE STOICS, EPICUREANS, AND SCEPTICS. Translated by the Rev. O. J. Reichbl, M.A. Crown Svo, 15s.

OUTLINES OF THE HISTORY OF GREEK PHILOSOPHY. Translated by Sarah F. Allbynb and Evelyn Ab- bott, M.A., LL.D. Crown 8vo, 10s. 6<^.

PLATO AND THE OLDER ACADEMY. Translated by Sarah F. Alleynb and Alfred Goodwin, B.A. Crown 8vo, 18s.

SOCRATES AND THE SO- CRATIC SCHOOLS. Translated by the Rev. O. J. Rbichel, M.A. Crown Svo 10s. %d.

ARISTOTLE AND THE EARLIER PERIPATETICS. Translated by B. F. C. C08TELLOE, M.A., and J. H. Muirhead, M.A. 2 vols. Crown Svo, 24s.

STONYHURST PHILOSOPHICAL SERIES.

A MANUAL OF POLITICAL

ECONOMY. By C. S. Devas, M.A. Crown Svo, 7s. 6d.

FIRST PRINCIPLES OF KNOW- LEDGE. By John Rickaby, S.J. Crown Svo, 5s.

GENERAL METAPHYSICS. By John Rickaby, S.J. Crown Svo, 5s.

LOGIC. By Richard F. Clarke,

S.J. Crown Svo, 5s.

MORAL PHILOSOPHY (ETHICS AND NATURAL LAW). By Joseph Rickaby, S.J. Crown Svo, 5s.

NATURAL THEOLOGY. By Ber- nard Boedder, S.J. Crown Svo, 6s. 6^.

PSYCHOLOGY. By Michael

Maher, S.J., D.Litt., M.A. (Lond.). Crown Svo, 6s. %d.

History and Science of Lan^ua^^e, etc.

DAVIDSON.— LEADING AND

IMPORTANT ENGLISH WORDS: Ex- plained and Exemplified. By William L. Davidson, M.A. Fcp. Svo, 3s. 6rf.

GRAHAM. — ENGLISH SYNO- NYMS, Classified and Explained: with Practical Exercises. By G. F. Graham. Fcp. Svo, 6s.

MAX MULLER (F.). THE SCIENCE OF LANGUAGE.

2 vols. Crown Svo. lOs.

BIOGRAPHIES OF WORDS, AND THE HOME OF THE ARYAS. Crown Svo, 5s.

MAX MtJLLER (F.)— continued.

CHIPS FROM A GERMAN WORKSHOP. Vol. III. ESSAYS ON LANGUAGE AND LITERATURE. Crown Svo, 5s.

LAST ESSAYS. First Series. Essays on Language, Folk-lore and other Subjects. Crown Svo, 5s.

ROGET.— THESAURUS OF ENG- LISH WORDS AND PHRASES. Clas- sified and Arranged so as to Facilitate the Expression of Ideas and assist in Literary Composition. By Pbter Mark Roobt, M.D., F.R.S. With full Index. Crown Svo, 9s. net.

20 LONGMANS AND CO.*S STANDARD AND GENERAL WORKS.

Political Economy, Economics, etc.

HIS- Crown 8vo, Part

ASHLEY (W. J.).

ENGLISH ECONOMIC TORY AND THEORY I., 5s. Part II., lOs. M.

SURVEYS, HISTORIC AND ECONOMIC. Crown 8vo, 95. net.

THE ADJUSTMENT OF

WAGES : a Study on the Coal and Iron Industries of Great Britain and the United States. With 4 Maps. 8vo, 12s. M. net.

BRITISH INDUSTRIES : a Series of General Reviews for Business Men and Students. By various Authors. Edited by W. J. Ashley, Crown Bvo, 5s. 6.7. net.

BAGEHOT. — ECONOMIC

STUDIES. By Walter Bagehot. Crown 8vo, 3s. %d.

BARNETT.— P RACTICABLE

SOCIALISM: Essays on Social Reform. By Samuel A. and Henrietta Barnett. Crown 8vo, 6s.

DEVAS.— A MANUAL OF POLI- TICAL ECONOMY. By C. S. Devas, M.A. Crown 8vo, 7s. 6f/. (Stonyhurst Philosophical Series.)

DEWEY.— FINANCIAL HISTORY OF THE UNITED STATES. By Davis Rich Dewey. Crown 8vo, 7s. 6ii. net.

LESLIE.— ESSAYS ON POLI- TICAL ECONOMY. By T. E. Cliffe Leslie, Hon. LL.D., Dubl. 8vo, 10s. 6d.

MACLEOD (Henry Dunning). BIMETALLISM. 8vo, 55. net. THE ELEMENTS OF BANKING.

Crown 8vo, 3s. 6d.

THE THEORY AND PRACTICE OF BANKING. Vol. I. 8vo, 12s. Vol. II. 14s.

MACLEOD (Henry Dunning)—

continued.

THE THEORY OF CREDIT.

8vj. In 1 Vol. 30s. net; or separately, Vol. I., 10s. net. Vol. II., Part I., 10s. net. Vol. II., Part II. 10s. net.

INDIAN CURRENCY. 8vo, 25. 6J.

net.

MILL.— POLITICAL ECONOMY.

By John Stuart Mill. Po.'ulnr Edition. Crown 8vo, 3s. 6rf. Library Edition. 2 vols. 8vo, 30s.

MULHALL.— INDUSTRIES AND WEALTH OF NATIONS. By Mich all G. Mulhall, F.S.S. With 32 Diagrams. Crown 8vo, 8s. 6rf.

S Y M E S.— POLITICAL ECON- OMY : a Short Text-book of Political Econ- omy. By J. E. SvMES, M.A. Cr. 8vo, 2s. 6d.

T O Y N B E E.— LECTURES ON

THE INDUSTRIAL REVOLUTION OF OF THE 18TH CENTURY IN ENGLAND. By Arnold Tovnbee. 8vo, 10s. 6d.

WEBB.— LONDON EDUCATION.

By Sidney Webb. Crown 8vo, 2s. 6d. net.

WEBB (Sidney and Beatrice).

THE HISTORY OF TRADE

UNIONISM. With Map and Biblio- graphy. 8vo, 7s. 6d. ner.

INDUSTRIAL DEMOCRACY: a

Study in Trade Unionism. 2 vols. 8vo 12s. net.

PROBLEMS OF MODERN IN- DUSTRY. 8no, 5s. net.

THE HISTORY OF LIQUOR LICENSING IN ENGLAND PRINCI- PALLY FROM 1700 TO 1830. Crown 8vo, 2s. 6rf. net.

Evolution, Anthropology, etc.

ANNANDALE AND ROBIN- SON.—fasciculi MALAYENSES: An- thropological and Zoological Results of an Expedition to Perak and the Siamese .Malay States, 1901-2. Undertaken by Nelson Annandale and Herbert C. Robinson. With Plates and Illustrations in the Text. ANTHROPOLOGY, Part I., 4to, 15s. net. ZOOLOGY, Part I., 4to, 30s. net; Part II. 4 to, 20s. net.

AVEBURY.— THE ORIGIN OF

CIVILISATION, and the Primitive Con- dition of Man. By the Right Hon. Lord AvEBURV. With 6 Plates and 20 Illustra- tions. 8vo, 18s.

CLODD (Edward).

THE STORY OF CREATION: a Plain Account of Evolution. With 77 Illustrations. Crown 8vo, 3s. 6rf.

A PRIMER OF EVOLUTION :

being a Popular Abridged Edition of 'The Story of Creation'. With Illustrations. Fcp. 8vo, Is. 6d.

DOUBTS ABOUT DARWIN- ISM. By a Semi-Darwinian. Cr. 8vo, 3s. 6d.

KELLER.— QUERIES IN ETHNO- GRAPHY. By Albert Galloway Keller. Ph.D. Fcp. 8vo, 2s. net.

LONGMANS AND CO.*S STANDARD AND GENERAL WOkKS. 2i

Evolution, Anthropology, etc. — continued.

LANG AND ATKINSON.—

SOCIAL ORIGINS. By Andrew Lang, M.A., LL.D.; and PRIMAL LAW. By J. J. Atkinson. 8vo, 10s. %d. net.

PACKARD.— LAMARCK, THE

FOUNDER OF EVOLUTION: his Life and Work, with Translations of his Writings on Organic Evolution. By Alpheus S. Packard, M.D., LL.D. With 10 Portrait and other Illustrations. Large Crown 8vo, 9s. net.

ROMANES (George John). ESSAYS. Edited by C. Lloyd MORGAN. Crown 8vo, 5s. net.

AN EXAxMINATION OF WEIS- MANNISM. Crown 8vo, 6s.

ROMANES (George John) -to«^

DARWIN, AND AFTER DAR- WIN : an Exposition of the Darwinian Theory, and a Discussion on Post-Dar- winian Questions.

Part I. The Darwinian Theory.

With Portrait of Darwin and 125 Illustra- tions. Crown 8vo, 10s. %d.

Part II. Post-Darwinian Ques- tions: Heredity and Utility. With Por- trait of the Author and 5 Illustrations. Crown 8vo, 10s. 6rf.

Part III. Post-Darwinian Ques- tions : Isolation and Physiological Selection. Crown 8vo, 5s.

The Science of Religion, etc.

B A L F O U R.— THE FOUNDA- TIONS OF BELIEF: being Notes Intro- ductory to the Study of Theology. By the Right Hon. Arthur James Balfour. Crown 8vo, 6s. net.

BARING-GOULD.— THE ORIGIN

AND DEVELOPMENT OF RELIGIOUS BELIEF. By the Rev. S. Barino-Gould. '2 vols. Crown 8vo, 3s. 6^. each.

C A M P B E L L.— RELIGION IN

GREEK LITERATURE. By the Rev. LEWIS CAMPBELL, M.A., LL.D. 8vo, 15s.

JAMES.— THE VARIETIES OF

RELIGIOUS EXPERIENCE: a Study in Human Nature. Being the Gifford Lectures on Natural Religion delivered at Edinburgh in 1901-1902. By William Jaues, LL.D., etc. 8vo, 12s. net.

LANG (Andrew). MAGIC AND RELIGION. 8vo,

lOs. 6^.

CUSTOM AND MYTH: Studies

of Early Usage and Belief. With 15 Illus- trations. Crown 8vo, 3s. 6^.

MYTH, RITUAL, AND RE- LIGION. 2 vols. Crown 8vo, 7s.

MODERN MYTHOLOGY : a Reply to Professor Max Miiller. 8vo, 9s.

THE MAKING OF RELIGION.

Crown 8vo, 5s. net.

MAX MtJLLER

Hon. F.).

(The Right

THE SILESIAN HORSEHERD

(• DAS PFERDEBU RLA ') : Questions o? the Hour answered by F. Max MCller. With a Preface by J. Estlin Carpenter. Crown 8vo, 5s.

CHIPS FROM A GERMAN WORKSHOP. Vol. IV.. Essa>s on Mythology and Folklore. Crown 8vo, 5s.

THE SIX SYSTEMS OF INDIAN PHILOSOPHY. Crown 8vo, 7s. M. net.

CONTRIBUTIONS TO THE SCIENCE OF MYTHOLOGY. 2 vols. 8vo, 32s.

THE ORIGIN AND GROWTH OF RELIGION, as illustrated by the Religions of India. The Hibbert Lectures, delivered at the Chapter House, West- minster Abbey, in 1878. Crown 8vo, 55.

INTRODUCTION TO THE SCIENCE OF RELIGION : Four Lec- tures delivered at the Royal Institution. Crown 8vo, 5s.

NATURAL RELIGION. The Gifford Lectures, delivered before the University of Glasgow in 1888. Crown 8vo, 5s.

ii LOffGMANS AND Co's STANDARD AND CENTRAL WORKS,

The Science of Religrion, etc. — continued.

MAX MULLER (The Right Hon.

F.) — coMtinued.

PHYSICAL RELIGION. The Gifford Lectures, delivered before the University of Glasgow in 1890. Crown 8vo, 5s.

ANTHROPOLOGICAL RE- LIGION. The Gifford Lectures, delivered before the University of Glasgow in 1891. Crown 8vo, 5s.

THEOSOPHY. OR PSYCHO- LOGICAL RELIGION. The Gifford Lectures, delivered before the University of Glasgow in 1892. Crown 8vo, 5s.

THREE LKCTURES ON THE VEDANTA PHILOSOPHY, delivered at the Royal Institution in March, 1894. Crown 8vo, 5s.

LAST ESSAYS. Second Series- Essays on the Science of Religion. Crown 8vo, 5s.

OAKESMITH.— THE RELIGION

OP PLUTARCH : a Pagan Creed of Apos- tolic Times. An Essay. By John Oake- SMITH, D.Litt., M.A. Crown Svo, 5s. net.

WPOD-MARTIN (W. G.).

TRACES OF THE ELDER FAITHS OP IRELAND: a Folk-lore Sketch. A Hand-book of Irish Pre- Christian Traditions. With 192 Illus- trations. 2 vols. Svo, SOs. net.

PAGAN IRELAND: an Archaeo- logical Sketch. A Handbook of Irish Pre- Christian Antiquities. With 512 Illustra- tions. Svo, 15s.

Classical Literature, Translations, etc.

ABBOTT.— HELLENICA. A Col-

lection of Essays on Greek Poetry, Philo- sophy, History, and Religion. Edited by Evelyn Abbott, M.A., LL.D. Crown Svo, ls.6d.

iESCHYLUS.— EUMENIDES OF

>ESCHYLUS. With Metrical English Translation. By J. F. Davies. Svo, 7s.

ARISTOPHANES. — THE

ACHARNIANS OF ARISTOPHANES, translated into English Verse. By R. Y. Tyrrell. Crown Svo, Is.

BECKER (W. A.). Translated by the Rev. F. Metcalfe, B.D.

GALLUS : or, Roman Scenes in the Time of Augustus. With Notes and Excursuses. With 26 Illustrations. Cr. Svo, 3s. Sd.

CHARICLES: or, Illustrations of I the Private Life of the Ancient Greeks. I With Notes and Excursuses. With 26 Illustrations. Crown Svo, 3s. 6d. '

C A M P B E L L.— RELIGION IN |

GREEK LITERATURE. By the Rev. ' Lewis Campbell, M.A., LL.D., Emeritus Professor of Greek, University of St. Andrews. Svo, 15s.

CI C E R O— ClCEl^O'S CORRE- SPONDENCE. By R. Y. Tyrrell. Vols. 1., II., III., Svo, each 12s. Vol. IV.. 15s. Vol. v., 14s. Vol. VI., 12s. Vol. VII. Index, 7s. 6rf.

HARVARD STUDIES IN

CLASSICAL PHILOLOGY. Edited by a Committee of the Classical Instructors of Harvard University. Vols. XI., 1900; XII., 1901 ; XIII., 1902; XIV., 1903. Svo, 6s. Sd. net each.

HIME.— LUCIAN, THE SYRIAN

Satirist. By Lieut.-Col. Henry W. L. Hime, (late) Royal Artillery. Svo, 5s. net.

HOMER.— THE ODYSSEY OF

HOMER. Done into English Verse. By William Morris. Crown Svo, 5s. net.

HORACE.— THE WORKS OF

HORACE, RENDERED INTO ENGLISH PROSE. With Life, Introduction and Notes. By William Coutts, M.A. Crown Svo, 5s. net.

LANG.— HOMER AND THE EPIC.

By Andrew Lang. Crown Svo, 9s. net.

LUCIAN. — TRANSLATIONS FROM LUCIAN. By Augusta M. Camp- bell Davidson, M.A., Edin. Crown Svo, 5s. net.

LONGMANS AND CO, S STANDARD AND GENERAL WORKS.

23

Classical Literature, Translations, etc. — continued.

OGILVIE.— HORAE LATINAE :

Studies in Synonyms and Syntax. By the late Robert Ogilvie, M.A., LL.D., H.M. Chief Inspector of Schools for Scotland. Edited by Alexander Souter, M.A. With a Memoir by Joseph Ooilvie. M.A., LL.D. 8vo, 125. 6^. net.

R I C H.— A DICTIONARY OF ROMAN AND GREEK ANTIQUITIES. By A. Rich, B.A. With 2,000 Woodcuts. Crown 8vo, 6s. net.

SOPHOCLES.— Translated into English Verse. By Robert Whitblaw, M.A., Assistant Master in Rugby School. Crown 8vo, 8s. 6rf.

THEOPHRASTUS. — THE

CHARACTERS OF THEOPHRASTUS : a Translation, with Introduction. By Chas. E. Bennett and William A. Hammond, Professors in Cornell University. Fcp. 8vo, 2s. M. net.

TYRRELL.— DUBLIN TRANS- LATIONS INTO GREEK AND LATIN VERSE. Edited by R. Y. Tyrrell. 8vo, 6s.

VIRGIL.

THE POEMS OF VIRGIL. Translated into English Prose by John Conington. Crown 8vo, 6s.

THE ^NEID OF VIRGIL.

Translated into English Verse by John Conington. Crown 8vo, 6s.

THE ^NEIDS OF VIRGIL. Done into English Verse. By William Morris. Crown 8vo, 5s. net.

THE ^NEID OF VIRGIL, freely translated into English Blank Verse. By W. J. Thornhill. Crown 8vo, 6s. net.

THE ^NEID OF VIRGIL. Translated into English Verse by James Rhoades.

Books I. -VI. Crown 8vo, 55.

Books VII.-XII. Crown 8vo, 55.

THE ECLOGUKS AND GEOR- GICS OF VIRGIL. Translated into English Prose by J. W. Mackail, Fellow of Balliol College, Oxford. 16mo, 5s.

WILKINS.— THE (5ROWTH OF

THE HOMERIC POEMS. By G. Wil- KiNs. 8vo, 6s.

Poetry and the Drama.

ARNOLD.— THE LIGHT OF THE , D A B N E Y. — THE MUSICAL

BASIS OP VERSE : a Scientific Study of the Principles of Poetic Composition. By J. P. Dabnby. Crown 8vo, 6s. 6^. net.

WORLD : or, The Great Consummation. By Sir Edwin Arnold. With 14 Illustrations after Holman Hunt. Crown 8vo, 5^. net.

BELL (Mrs. Hugh).

CHAMBER COMEDIES: a Col- lection of Plays and Monologues for the Drawing Room. Crown 8vo, 5s. net.

FAIRYTALE PLAYS, AND HOW TO ACT THE.M. With 91 Diagrams and 52 Illustrations. Crown 8vo, 3s. net.

NURSERY COMEDIES: Twelve Tiny Plays for Children. Fcp. 8vo, Is. 6^.

RUMPELSTILTZKIN : a Fairy

Play in Five Scenes (Characters, 7 Male ; 1 Female). From ' Fairy Tale Plays and How to Act Them '. With Illustrations, Diagrams and Music. Cr. 8vo, sewed, 6^/.

COCHRANE. — COLLECTED

VERSES. By Alfred Cochrane, Author of 'The Kestrel's Nest, and other Verses,' ' Leviore Plectro,' etc. With a Frontispiece by H.J. Ford. Fcp. 8vo, 5s. net.

GRAVES.— CLYT^MNESTRA : A TRAGEDY. By Arnold F. Graves. With a Preface by Robert V. Tyrrell, Litt.D. Crown 8vo, 5s. net.

HITHER AND THITHER :

Songs and Verses. By the Author of • Times and Days,' etc. Fcp. 8vo, 5s.

INGELOW (Jean). POETICAL WORKS. Complete

in One Volume. Crown 8vo, gilt top, 6s.

net.

LYRICAL AND OTHER POEMS.

Selected from the Writings of Jean Ingelow. Fcp. 8vo, 2s. 6^. cloth plain, 3s. cloth (|ilt,

24

LONGMANS AND CO, S STANDARD AND GENERAL WORKS,

Poetry and the Drama — continued.

KENDALL.— POEMS OF HENRY MORRIS (William).

CLARENCE KENDALL. With Memoir by Frederick C. Kendall. Crown 8vo,6s.

-continued.

LANG (Andrew). GRASS OF PARNASSUS. Fcp.

8vo. 2s. 6^. net.

THE BLUE POETRY BOOK.

Edited by Andrew Lang. With 100 Illustrations. Crown 8vo, gilt edges, 65.

LECKY.— POEMS. By William

Edward Hartpolb Lbcky. Pep. 8vo, 5s.

LYTTON (The Earl of), (Owen Meredith).

THE WANDERER. Crown 8vo, lOs. &i.

LUCILE. Crown 8vo, IO5. 6</.

SELECTED POEMS. Crown 8vo, lOs. 6^.

MACAULAY. — LAYS OF

ANCIENT ROME, WITH * IVRY ' AND 'THE ARMADA*. By Lord Macaulay.

Illustrated by G. Scharf. Fcp. 4to,

10s. 6rf. Bijou

Edition. 18mo, 2s. 6<^ gilt top.

Popular

THE ODYSSEY OF HOMER.

Done into English Verse. Crown 8vo, 5s.

Edition. Fcp. 4to, %d. sewed, Is cloth. Illustrated by J. R. Weguelin. Crown 8vo, 3s. net.

Annotated Edition. Fcp. Svo, \s.

sewed, Is. %d. cloth.

MacDONALD.— A BOOK OF

STRIFE, IN THE FORM OF THE DIARY OF AN OLD SOUL : Poems. By George MacDonald, LL.D. 18mo, 6s.

MORRIS (William). '

POETICAL WORKS. Library i

Edition. Complete in 11 volumes. Crown Svo, price Ss. net each.

THE EARTHLY PARADISE.

4 vols. Crown Svo, 5s. net each.

THE LIFE AND DEATH OF

JASON. Crnvn Svo, 5s. net.

THE DEFENCE OF GUENE-

VERE, and other Poems. Crown Svo, 5i. net.

THE STORY OF SIGURD THE VOLSUNG, AND THE FALL OF THE NIBLUNGS. Crown Svo, 5s. net.

POEMS BY THE WAY, AND I LOVE IS ENOUGH. Crown Svo, 5s. net.

net.

THE .CNEIDS OF VIRGIL. Done into English Verse. Crown Svo, 5s. net.

THE TALE OF BEOWULF, SOMETIME KING OF THE FOLK OF THE WEDERGEATS. Translated by William Morris and A. J. Wvatt. Cr. Svo, 5s. net.

Certain of the Poetical Works may also be had in the followiug Editions : —

THE EARTHLY PARADISE.

Popular Edition. 5 vols. 12mOy 25s. ; or 5s. each, sold separately.

The Same in Ten Parts, 255. ; or 2s. %d. each sold separately.

Cheap Edition^ in 1 vol. Crown Svo, 6s. net.

POEMS BY THE WAY. Square

Crown Svo, 6s.

THE DEFENCE OF GUENE- VERE AND OTHER POEMS. Cheaper Impression. Fcp. Svo, Is. 6^. net.

*** For Mr. William Morris's other Works see pp. 27, 28, 37 and 40.

MORS ET VICTORIA. Cr. Svo,

5s. net.

*^* This is a drama in three acts, the scene of which is laid in France shortly after the massacre of St. Bartholomew.

MORTE ARTHUR.— An AUitera-

tive Poem of the Fourteenth Century. Edited from the Thornton MS., with Intro- duction, Notes and Glossary. By Mary Macleod Banks, Fcp. Svo, 3s. Qd.

NES BIT.— LAYS AND LE- GENDS. By E. Nesbit (Mrs. Hubert Bland). First Series. Crown Svo. 3s. 6rf. Second Series. With Portrait. Cr. Svo, 5s.

RILEY. —OLD FASHIONED

ROSES : Poems. By Jambs Whitcomb Riley. 12mo, gilt top, 55.

ROMANES.— A SELECTION

FROM THE POEMS OF GEORGE JOHN RO.MANES, M.A., LL.D., F.R.S. With an Introduction by T. Herbert Warren, Pre- sident of Magdalen College, OxforH. Oowi^ Svo, 4s. %d.

LONGMANS AND CO, S STANDARD AND GENERAL WORKS,

25

Poetry and the Drama — continued.

SAVAGE-ARMSTRONG.— BAL- LADS OF DOWN. By 0. F. Savage- Armstrong, M.A., D.Litt. Crown 8vo, 75. W.

SHAKESPEARE.

BOWDLER'S FAMILY SHAKE- SPEARE. With 36 Woodcuts. 1 vol. 8vo, 14s. Or in 6 vols. Fcp. 8vo, 21s.

THE SHAKESPEARE BIRTH- DAY BOOK. By .Mary F. Dunbar. 32mo, Is. %d.

STEVENSON.— A CHILD'S GAR- DEN OF VERSES. By Robert Louis Stevenson. Fcp. Svo, gilt top, 5s.

TREVELYAN.— CECILIA GON

ZAGA : a Drama. By R. C. Trevelyan. Fcp. Svo, 2s. 6<f. net.

WAGNER.— THE NIBELUNGEN RING. Done into English Verse by Reginald Rankin, B.A., of the Inner Temple, Barrister-at-Law.

Vol. I. Rhine Gold, The Valkyrie. Fcp. Svo, gilt top, 4s. 6^.

Vol. II. Siegfried, The Twilight of

the Gods. Fcp. Svo, gilt top, 45. 6^.

Fiction, Humour, etc.

(Reprinted

ANSTEY (P.). VOCES POPULI.

from ' Punch '.)

First Series. With 20 Illustrations by J. Bernard Partridge. Crown Svo, gilt top, Ss. net.

Second Series. With 25 Illustra- tions by J. Bernard Partridge. Crown Svo, gilt top, 3s. net.

THE MAN FROM BLANKLEY'S,

and other Sketches. (Reprinted from 'Punch'.) With 25 Illustrations by J. Bernard Partridge. Crown Svo, gilt top, 3s. net.

BAILEY (H. C). MY LADY OF ORANGE : a

Romance of the Netherlands in the Days of Alva. With S Illustrations. Crown Svo, 6s.

KARL OF ERBACH : a Tale of the Thirty Years' War. Crown Svo, 6s.

THE MASTER OF GRAY : a Tale of the Days of Mary Queen of Scots. Crown Svo, 6s.

BEACONSFIELD (The Earl of).

NOVELS AND TALES. Com- plete in U vols. Crown Svo, Is. 6</. each, or in sets, 11 vols., gilt top, 15s. net.

Contarini Fleming ; The Rise of Iskan- der.

Sybil.

Henrietta Temple.

Venctia.

Vivian Grey.

The Young Duke ;

Count Alarcos : a

Tragedy. Alroy ; Ixion in

Heaven ; The In-

femal Marriage ; ' Conin^sby.

Popanilla. Lothair.

Tancred. Endymion.

NOVELS AND TALES. HUGHENDEN EDITION. Portrai s and 11 Vignettes. Crown Svo, 42s.

THE

With 2 1 1 vols.

BOTTOME.— LIFE, THE INTER- PRETER. By Phyllis Bottome. Crown Svo, 6s.

CHURCHILL.— SAVROLA: a

Tale of the Revolution in Laurania. By Winston Spencer Churchill, M.P. Cr. Svo, 6$.

CONVERSE. -LONG WILL: a

Tale of Wat Tyler and the Peasant Rising in the Reign of Richard II. By Florence Converse. With 6 Illustrations by Garth Jone& Crown Svo, 6s.

DAVENPORT.-BY THE RAM- PARTS OF JEZREEL: a Romance of Jehu, King of Israel. By Arnold Daven- port. With Frontispiece by Lancelot Speed. Crown Svo, 6s.

DOUGALL.— BEGGARS ALL. By

L. Dougall. Crown Svo, 3s. 6f/.

DOYLE (Sir A. Conan).

MICAH CLARKE : a Tale of Mon- mouth's Rebellion. With 10 Illustrations. Crown Svo, 3s. 6i/.

THE REFUGEES, a Tale of the Huguenots. With 25 Illustrations. Cr. Svo, 3s. 6d.

THE STARK MUNRO LETTERS.

Crown Svo, 3s. ^d.

THE CAPTAIN OF THE POLE- STAR, and other Tales. Cr. Svo, 3s. 6<<.

36

LONGMANS AND CO, S STANDARD AND GENERAL WORKS,

Fiction. Humour, etc. — continued,

HAGGARD (H. ^\x>B.vi)— continued.

DUNBAR.— THE SONS O' COR-

MAC : Irish Legends. By Alois Dunbar. With 8 Illustrations by Miss Luxhorb. Cr. 8vo.

FARRAR (F. W., late Dean op Cantbrbury).

DARKNESS AND DAWN: or. Scenes in the Days of Nero. An Historic Tale. Crown 8vo, gilt top, 65. net.

GATHERING CLOUDS: a Tale of the Days of St. Chrysostom. Crown 8vo, gilt top, 6s. net.

FOWLER (Edith H.).

THE YOUNG PRETENDERS. A Story of Child Life. With 12 Illustra- tions by Sir Phiup Burnb-Jones, Bart. Crown 8vo, 6s.

THE PROFESSOR'S CHIL- DREN. With 24 Illustrations by Bthbl Katb Burgess. Crown 8vo, 6s.

FRANCIS (M. E.)

CHRISTIAN THAL: a Story of Musical Life. Crown Svo, 6$.

FIANDER'S WIDOW. Crown 8vo, 65.

YEOMAN FLEETWOOD. With Frontispiece. Crown 8vo, Ss. net.

PASTORALS OF DORSET. With 8 Illustrations. Crown Svo, 6s.

THE MANOR FARM. With Frontispiece by Claud C. du Pre Cooper. Crown 8vo, 6s.

FROUDE.— THE TWO CHIEFS OF DUNBOY: an Irish Romance of the Last Century. By Jambs A. Froudb. Crown Svo, 3s. &i.

HAGGARD SIDE, THE: being

Essays in Fiction. By the Author of ' Times and Days,' 'Auto da F^,* &c. Crown Svo, 5s.

HAGGARD (H. Rider).

ALLAN QUATERMAIN. With 31 Illustrations. Crown Svo, 35. %d.

ALLAN'S WIFE. With 34 Illus- trations. Crown Svo, 3s. 6t^.

BEATRICE. With Frontispiece and Vignette, Crown Svo, 8s. &/.

BLACK HEART AND WHITE HEART, AND OTHER STORIES. With 33 Illustrations. Crown Svo, 3s. &i.

CLEOPATRA. With 29 Illustra- tions. Crown Svo, 3s. 6(2.

COLONEL QUARITCH, V.C.

With Frontispiece and Vignette. Crown Svo, 3s. 64.

DAWN. With

Crown Svo, 3s. ^.

16 Illustrations.

DR. THERNE. Cr. Svo, 35. %d.

ERIC BRIGHTEYES. With 51 Illustrations. Crown Svo, 3s. 6<2.

HEART OF THE WORLD. With

15 Illustrations. Crown Svo, 3s. 6<i.

JOAN FIASTE. With 20 Illustra- tions. Crown Svo, 3s. 6(2.

LYSBETH. With 26 Illustrations. Crown Svo, 6s.

MAIWA'S REVENGE. Crown Svo,

Is. 6(2.

MONTEZUMA'S With 24 Illustrations.

DAUGHTER.

Crown Svo, 3s. 6(2.

MR. MEESON'S WILL. With 16 Illustrations. Crown Svo, 3s. 6(2.

NADA THE LILY. With 23 II-

lustrations. Crown Svo, Ss. 6(2.

PEARL-MAIDEN : a Tale of the

Fall of Jerusalem. With 16 Illustrations. Crown Svo, 6s.

SHE. With 32 Illustrations. Cr. Svo, 3s. 6(2.

STELLA FREGELIUS : a Talc of

Three Destinies. Crown Svo, 6s.

SWALLOW: a Tale of the Great Trek. With 8 Illustrations. Crown Svo, 3s. 6(2.

THE PEOPLE OF THE MIST.

With 16 Illustrations. Crown Svo, 3s. 6J.

THE WITCH'S HEAD. With 16 Illustrations. Crown Svo, 3s. 6(2.

HAGGARD AND LANG.— THE

WORLD'S DESIRE. By H. Ridbr Hag- gard and Andrew Lano. With 27 Illus- trations. Crown Svo, 3s. 6(2.

r

1

LoifGMArfs And co,*s stAn1>aMd aI^d general tvokKS. 27

Fiction, Humour, etc. — continued.

HARTE.— IN THE CARQUINEZ By Bret Harte. Crown 8vo,

WOODS

35. 6^;.

HOPE.— THE HEART OF PRIN- CESS OSRA. By Anthony Hope. With 9 Illustrations. Crown 8vo, 3s. %d.

HOWARD.— THE FAILURE OF SUCCESS. By Lady Mabel Howard. Crown 8vo, 6s.

HUTCHINSON.— A FRIEND OF NELSON. By Horace G. Hutchinson. Crown Svo, 6s.

JEROME.— SKETCHES IN LA- VENDER: BLUE AND GREEN. By Jerome K. Jerome, Author of ' Three Men in a Boat,* etc. Crown Svo, 3s. 6rf.

JOYCE.— OLD CELTIC RO- MANCES. Twelve of th» most beautiful of the Ancient Irish Romantic Tales. Trans- lated from the Gaelic. By P. W. Joyce, LL.D. Crown Svo, 3s. 6rf.

LANG (Andrew).

A MONK OF FIFE: a Story of the Days of Joan of Arc. With 13 Illus- trations by Selwyn Image. Crown Svo, 3s. %d.

THE DISENTANGLERS. With 7 Full-page Illustrations by H. J. Ford. Crown Svo, 6s.

LYALL (Edna). THE HINDERERS. Crown Svo,

THE AUTOBIOGRAPHY OF A SLANDER. Fcp. Svo, Is. sewed.

Presentation Edition. With 20 Illus- trations by Lancelot Speed. Crown Svo, 2s. %d. net.

DOREEN. The Story of a Singer. Crown Svo, 6s.

WAYFARING MEN. Cr. Svo, 65.

HOPE THE HERMIT: a Ro- mance of Borrowdale. Crown Svo, 6s.

MARCHMONT.— IN THE NAME OF A WOMAN : a Romance. By Arthur W. Marchmont. With 8 lUustratioos. Crown Svo, 6s.

MASON AND LANG.-PARSON

KELLY. By A. E. W. Mason and Andrew Lang. Crown Svo, 3s. Qd.

MAX MOLLER. — DEUTSCHE

LIEBE (GERMAN LOVE): Fragments from the Papers of an Alien. Collected by F. Max M&ller. . Translated from the German by G. A. M. Cr. Svo, gilt top, 5s.

MELVILLE (G. J. Whyte).

The Gladiators. The Interpreter. Good for Nothing. The Queen's Maries.

Holmby House. Kate Coventry. Digby Grand. General Bounce.

Crown Svo, Is. 6^. each.

MERRI MAN.— FLOTSAM: a

Story of the Indian Mutiny. By Henry Seton Merriman. ""■•- '' Vignette by H. G Zs.Qd.

With Frontispiece and Masse Y. Crown Svo,

MORRIS (William).

THE SUNDERING FLOOD. Cr.

Svo, 7s. %d.

THE WATER OF THE WON- DROUS ISLES. Crown Svo, 7s. 6d.

THE WELL AT THE WORLD'S END. 2 vols. Svo, 2Ss.

THE WOOD BEYOND THE WORLD. Crown Svo, 6s. net.

THE STORY OF THE GLITTER- ING PLAIN, which has also been called The Land of the Living Men, or the Acre of the Undying. Square post Svo, 5s. net.

THE ROOTS OF THE MOUN- TAINS, wherein is told somewhat of the Lives of the Men of Burgdale, their Friends, their Neighbours, their Foemen, and their Fellows-in-Arms. Written in Prose and Verse. Square crown Svo, Ss.

A TALE OF THE HOUSE OF THE WOLFINGS, and all the Kindreds of the Mark. Written in Prose and Verse. Square crown Svo, 6s.

A DREAM OF JOHN BALL, AND A KING'S LBSSON. 16ino, 2s. net.

2& LO.VCMANS AND CoJs STANDARD AND GENERAL iVO^h'S,

Fiction, Humour, etc. — continued.

MORRIS (William)— foM//M«<'</.

NEWS FROM NOWHERE*, or,

An Epoch of Rest. Being some Chapters

from an Utopian Romance. Post 8vo, l5. %d.

THE STORY OF GRETTIR THE STRONG. Translated from the Icelandic by BiRiKR MagnC'SBOn and William Morris. Crown 8vo, 55. net.

THREE NORTHERN LOVE STORIES, AND OTHER TALES. Translated from the Icelandic by BirIkr MagnOsson and William Morris. Cr. 8vo, 65. net.

♦»* For Mr. William Morris's other Works, see pp. 24, 37 and 48.

NEWMAN (Cardinal). LOSS AND GAIN : the Story of a

Convert. Crown 8vo, 35. 6^.

CALLISTA : a Tale of the Third Century. Crown 8vo, 35. 6rf.

PHILLIPPS-WOLLEY. SNAP :

a Legend of the Lone Mountain. B^ C.

Phillipps-Wollev. Crown 8vo, 35. 6rf.

With 13 Illustrations.

PORTMAN. — STATION

STUDIES : being the Jottings of an African Official. By Lionel Portman. Crown 8vo, 55. net.

SEWELL (Elizabeth M.).

A Glimpse of the World. Laneton Parsonage. Margaret Percivsfl. Katharine Ashton. The Earl's Daughter. The Experience of Life.

Amy Herbert. Cleve Hall. Gertrude. Home Life. After -Life. Ursula. Ivors.

Crown 8vo, cloth plain, l5. %d. each. Cloth extra, gilt edges. 25. 6^. each.

SHEEHAN.— LUKE DELMEGE.

By the Rev. P. A. Shebhan, D.D., Author of • My New Curate '. Crown 8vo, 65.

SOMERVILLE (E. CE.) AND

ROSS (Martin).

SOME EXPERIENCES OF AN

IRISH R.M. With 31 Illustrations by E. CE. SoMEKViLLE. Crown 8vo, 65.

ALL ON THE IRISH SHORE: Irish Sketches. With 10 Illustrations by E. CE. SoMERViLLE. Crown 8vo, 65.

THE REAL CHARLOTTE. Cr.

8vo, 3s. %d.

THE SILVER FOX. Crown 8vo,

35. %d.

AN IRISH COUSIN. Cr. 8vo, 6s.

STEVENSON (Robert Louis).

THK STkANGE case OF DR. JEKYLL AND MR. HYDE. Fcp. 8vo, l5. sewed, l5. 6^. cloth.

THE STRANGE CASE OF DR. JEKYLL AND MR. HYDE; WITH OTHER FABLES. Crown 8vo, bound in buckram, with gilt top, 55. net.

' Silver Library ' Edition. Cr. Svo, Zs.%d.

MORE NEW ARA'BlAN-NIGHTS —THE DYNAMITER. By Robert Louis Stevenson and Fanny van de Grjft Stevenson. Crown Svo, 3s. %d.

THE WRONG BOX. By Robert Louis Stevenson and Lloyd Osbournb. Crown Svo, 35. 6d.

SUTTNER— LAY DOWN YOUR

ARMS (Die Waff en NiecUr): The Auto- biography of Martha von Tilling. By Bertha von Suttner. Translated by T. Holmes. Crown Svo, l5. 6(/.

TROLLOPE (Anthony).

THE

15. 6(i.

WARDEN. Crown Svo,

BARCH ESTER TOWERS. Crown

Svo, l5. %d.

VAUGHAN.— OLD HENDRIK'S

TALES. By Captain ArthurO.Vaughan, With 12 Full-page Illustrations by J. A. Shepherd. Crown Svo.

WALFORD (L. B.).

STAY-AT-HOMES. Cr. Svo, 6s.

CHARLOTTE. Crown Svo, 6s.

ONE OF OURSELVES. Crown 8vo, 65.

THE INTRUDERS. Cr. 8vo, 25. 6rf.

LEDDY MARGET. Cr. Svo, 2s. 6rf.

IVA KILDARE: a Matrimonial

Problem. Crown Svo, 25. 6^.

MR. SMITH : a Part of his Life. Crown Svo, 2s. 6rf.

THE BABY'S GRANDMOTHER. Crown Svo, 25. %d.

\

LONGMANS AND CO. S STANDARD AND GENERAL WORKS. ig

Fiction, Humour, etc. — continued.

WALFORD (L. B.)— continued. COUSINS. Crown 8vo, 2s. ^d.

TROUBLESOME DAUGHTERS.

Crown 8vo, 2s. 6^. PAULINE. Crown 8vo, 2s. 6rf.

DICK NETHERBY. Crown 8vo, 2s. %d.

THE HISTORY OF A WEEK. Crown 8vo, 2s. %d.

A STIFF-NECKED GENERA- TION. Crown 8vo, 2s. 6rf.

NAN, and other Stories. Cr. Svo, 2s. %d.

THE MISCHIEF OF MONICA.

Crown Svo, 2s. 6rf.

THE ONE GOOD GUEST. Cr.

Svo, 2s. 64.

'PLOUGHED' and other Stories. Crown Svo, 2s. %d.

THE MATCHMAKER. Cr. Svo, 2s. W.

WARD.— ONE POOR SCRUPLE. By Mrs. Wilfrid Ward. Crown Svo, 6s.

WEYMAN (Stanley).

THE HOUSE OF THE WOLF. With Frontispiece and Vignette. Crown Svo, 3s. 6^.

A GENTLEMAN OF FRANCE.

With Frontispiece and Vignette. Crown Svo, 6s.

THE RED COCKADE. With

Frontispiece and Vignette. Crown Svo, 6s.

SHREWSBURY. With 24 Illus- trations by Claude A. Shbppbrson. Crown Svo, 6s.

SOPHIA. With Frontispiece. Cr. Svo, 6s.

THE LONG NIGHT. A Story of

Geneva in 1602. Crown Svo, 6s.

YEATS (S. Levett). THE CHEVALIER D'AURIAC-

Crown Svo, 3s. 6rf.

THE TRAITOR'S WAY. Crown Svo, 6s.

Y O X A L L. - THE ROMMANY

STONE. By J. H. Yoxall, M.P. Crown Svo, 6s.

Popular Science (Natural History, etc.).

FURNEAUX (W.).

THE OUTDOOR WORLD: or,

The Young Collector's Handbook. With IS Plates (16 of which are coloured), and 549 Illustrations in the Text. Crown Svo, gilt edges, 6s. net.

BUTTERFLIES AND MOTHS

(British). With 12 Coloured Plates and 241 Illustrations in the Text. Crown Svo, gilt edges, 6s. net.

LIFE IN PONDS AN D STREAMS. With S Coloured Plates and 331 Illustrations in the Text. Crown Svo, gilt edges, 6.S. net.

THE SEA SHORE. With 8 Coloured Plates and 300 Illustrations in the Text. Crown Svo, %s. net.

I HARTWIG (George).

1 THE SEA AND ITS LIVING WONDERS. With 12 Plates and 303 Woodcuts. Svo, gilt top, 7s. net.

THE TROPICAL WORLD. With 8 Plates and 172 Woodcuts. Svo, gilt top, 7s. net.

THE POLAR WORLD. With 3 Maps, S Plates and S5 Woodcuts. Svo, gilt top, 7s. net.

THE SUBTERRANEAN WORLD.

With 3 Maps and 80 Woodcuts. Svo, gilt top, 7s. net.

30 LONGMANS AND CO.^S STANDARD AND GENERAL WORkS,

Popular Science (Natural History, etc.) — continued,

HELMHOLTZ.- POPULAR LEG- PROCTOR (Richard A.)— co«//««^'t/.

TURES ON SCIENTIFIC SUBJECTS.

By Hermann von Helhholtz. With 68 Woodcuts. 2 vols. Crown 8vo. 35. %d. each.

HOFFMANN.-ALPINE FLORA:

For Tourists and Amateur Botanists. With Text descriptive of the most widely distri- buted and attractive Alpine Plants. By Julius Hoffmann. Translated by B. S. Barton (Mrs. A Gepp). With 40 Plates containing 250 Coloured Figures from Water- Colour Sketches by Hermann Pribsb. 8vo, 7s. 6<<. net.

HUDSON (W. H.).

HAMPSHIRE DAYS. With 11 Plates and 96 Illustrations in the Text from Drawings by Bryan Hook, etc. 8vo, 10s. %d. net.

BIRDS AND

8vo, 6s. net.

MAN. Large crown

NATURE IN DOWNLAND. With 12 Plates and 14 Illustrations in the Text by A. D. McCormick. 8vo, 10s. 6^. net.

BRITISH BIRDS. With a Chap- ter on Structure and Classification by Frank B. Beddard, F.R.S. With 16 Plates (8 of which are Coloured), and over 100 Illustrations in the Text. Crown 8vo, gilt edges, 6s. net.

MILLAIS.— THE NATURAL HIS- TORY OF THE BRITISH SURFACE FEEDING-DUCKS. By John Guille MiLLAis, F.Z.S., etc. With 6 Photogravures and 66 Plates (4l in Colours) from Drawings by the Author, Archibald Thorburn, and from Photographs. Royal 4to, £6 6s.

PROCTOR (Richard A.).

LIGHT SCI ENCE FOR LEISU R ti HOURS. Familiar Essays on Scientific Sufatjects. Crown 8vo, 3s. 6d.

ROUGH WAYS MADE SMOOTH. Familiar Essays on Scientific Subjects. Crown 8vo, 8s. 6d.

PLEASANT WAYS IN SCIENCE.

Crown 8vo, 3s. 6t/.

NATURE STUDIES. By R. A. Proctor, Grant Allen, A. Wilson, T. Foster and E. Clodd. Crown 8vo, 3s. 6rf.

LEISURE READINGS. By R. A. Proctor, E. Clodd, A. Wilson, T. Foster and A. C. Ranyard. Crown 8vo. 3s. 6rf.

**♦ For Mr. Proctor's other books see pp. 16 atui 35, and Messrs. Longmans & Co.'s Catalogue of Scientific Works.

STANLEY.— A FAMILIAR HIS- TORY OF BIRDS. By E. Stanley, D.D., formerly Bishop of Norwich. With 160 Illustrations. Cr. 8vo, 3s. 6rf.

WOOD (Rev. J. G.).

HOMES WITHOUT HANDS : a

Description of the Habitations of Animals, classed according to their Principle of Construction. With 140 Illustrations. 8vo, gilt top, 7s. net.

INSECTS AT HOME: a Popular

Account of British Insects, their Structure, Habits and Transformations. With 700 Illustrations. 8vo, gilt top, 7s. net.

INSECTS ABROAD: a Popular

Account of Foreign Insects, their Struc- ture, Habits and Transformations. With 600 Illustrations. 8vo, 7s. net.

OUT OF DOORS: a Selection of Original Articles on Practical Natural History. With U Illustrations. Crown 8vo, 3s. %d.

PETLAND REVISITED. With 33

Illustrations. Crown 8vo, Ss. 6d.

STRANGE DWELLINGS: a De- scription of the Habitations of Animals, abridged from ' Homes without Hands \ W ith 60 Illustrations. Crown 8vo, 35. ftj.

LONGMANS AND CO's STANDARD AND GENERAL WORKS, 3 1

Works of Reference.

ANNUAL REGISTER (THE). A

Review of Public Events at Home and Abroad, for the year 1902. 8vo, 18s.

Volumes of the Annual Register for the years 1863-1901 can still be had. 18s. each.

CHARITIES REGISTER. THE

ANNUAL, AND DIGEST: being a

Classified Register of Charities in or available in the Metropolis. 8vo, 5s. net.

CHISHOLM.— HANDBOOK OF

COMMERCIAL GEOGRAPHY, By George G. Chisholm, M.A., B.Sc, Fellow of the Royal Geographical and Statistical Societies. With 19 Folding-out Maps and numerous Maps in the Text. 8vo, 15s. net.

GWILT. - AN ENCYCLOPEDIA OF ARCHITECTURE. By Joseph Gwilt, F.S.A. With 1,700 Engravings. Revised (1888), with Alterations and Considerable Additions by Wyatt Papworth, 8vo, 21s. net.

LONGMANS' GAZETTEER OF THE WORLD. Edited by George G. Chisholm, M.A., B.Sc. Imperial 8vo, IBs. net cloth ; 21s. half-morocco.

MAUNDER (Samuel).

BIOGRAPHICAL TREASURY.

With Supplement brought down to 1889. By Rev. James Wood. Fcp. 8vo, 6s.

MAUNDER (Samuel) — continued.

THE TREASURY OF BIBLE KNOWLEDGE. By the Rev. J. Avrb, M.A. With 5 Maps, 15 Plates, and 300 Woodcuts. Fcp. 8vo, 6s.

TREASURY OF KNOWLEDGE AND LIBRARY OF REFERENCE. Fcp. 8vo, 6s.

THE TREASURY OF BOTANY. Edited by J. Lindley, F.R.S., and T. Moore, F.L.S. With 274 Woodcuts and 20 Steel Plates. 2 vols. Fcp. 8vo, 12s.

RICH. A Dictionary of Roman and Greek Antiquities. By A Rich, B.A, With 2,000 Woodcuts. Crown 8vo, 6s. net.

ROGET. —THESAURUS OF ENG- LISH WORDS AND PHRASES. Classified and Arranged so as to Facilitate the Ex- pression of Ideas and assist in Literary Composition. By Peter Mark Rooet, M.D., F.R.S. Recomposed throughout, enlarged and improved, partly from the Author's Notes, and with a full Index, by the Author's Son, John Lewis Rogbt. Crown 8vo, 9s. net.

WILLICH.— POPULAR TABLES

for giving information for ascertaining the value of Lifehold, Leasehold, and Church Property, the Public Funds, etc. By Charles M. Willich. Edited by H. Bench Jones. Crown 8vo, 10s. %d.

Children's Books.

ADELBORG. - CLLAN PETER

AND THE CH ILDREN OF GRUBBYLEA. By Ottilia Adeldorg. Translated from the Swedish by Mrs. Graham Wallas. With 23 Coloured Plates. Oblong 4to, boards, 3s. 6(/. net.

ALICK'S ADVENTURES.— By

G. R. With 8 Illustrations by John H assall. Crown 8vo, 3s. 6d.

BOLD TURPIN:

Sung by Sam Weller. in Colour by L. D. L. 6s.

a Romance, as

With 16 Illustrations

Oblong 4to, boards,

BROWN.— THE BOOK OF

SAINTS AND FRIENDLY BEASTS. By Arbie Farwell Brown. With 8 Illustra- tions by Fanny V. Cory. Crown 8vo, 4s. 6rf. net.

CRAK^ (Rev. A. D.).

EDWY THE FAIR: or, The First Chronicle of >Escendune. Crown 8vo, silver top, 2s. net.

ALFGAR THE DANE: or, The Second Chronicle of i4B8cendune. Crown 8vo, silver top, 2s. net.

THE RIVAL HEIRS: being the Third and Last Chronicle of iGscendune. Crown 8vo, silver top, 2s. net.

THE HOUSE OF WALDERNE.

A Tale of the Cloister and the Forest in the Days of the Barons' Wars. Crown 8vo, silver top, 2s. net.

BRIAN FITZ-COUNT. A Story of Wallingford Castle and Dorchester Abbey. Crown 8vo, silver top, 25. net.

32

LONGMANS AND CO. S STANDARD AND GENERAL WORKS.

Children's Books — continued.

DENT.— IN SEARCH OF HOME: a Story of Bast-End Waifs and Strays. By Phyllis O. Dent. With a Frontispiece in Colour by H ambl Listbr. Crown 8vo, 35. 6^. net.

HENTY (G. A.).— Edited by.

YULE LOGS: A Story-Book for Boys. By Various Authors. With 61 Illustrations. Crown 8vo, gilt edges, 3s. net.

YULE TIDE YARNS: a Story- Book for Boys. By Various Authors. With 45 Illustrations. Crown 8vo, gilt edges, 35. net.

LANG (Andrew). — Edited by. THE BLUE FAIUY BOOK. With

138 Illustrations. Crown 8vo, gilt edges, 6s.

THE Ri-D FAIRY BOOK. With

100 Illustrations. Crown 8vo, gilt edges, 65.

THE GREEs FAIRY BOOK.

With 99 Illustrations. Crown 8vo, gilt edges, 65.

THE GREY FAIRY BOOK. With

65 Illustrations. Crown 8vo, gilt edges, 65.

THE YELLOW FAIRY BOOK.

With 104 Illustrations. Crown 8vo, gilt edges, 6s.

THE PINK FAIRY BOOK. With

67 Illustrations. Crown 8vo, gilt edges, 65.

THE VIOLET FAIRY BOOK.

With 8 Coloured Plates and 54 other Illus- trations. Crown 8vo, gilt edges, 6s.

THE CRIMSON FAIRY BOOK.

With 8 Coloured Plates and 43 other Illus- trations. Crown 8vo, gilt edges, 6s.

THE BLUE POETRY BOOK.

With 100 Illustrations. Crown 8vo, gilt edges, 6s.

THE TRUE .^TO Y BOOK. With

66 Illustrations. Crown 8vo, gilt edges, 6s.

THE RED TRUE STORYBOOK.

With 100 Illustrations. Crown 8vo, gilt edges, 65.

THE ANIMAL STORY BOOK.

With 67 Illustrations. Crown 8vo, gilt edges, 65.

THE RED BOOK OF ANIMAL STORIES. With 65 Illustrations. Crown 8vo, gilt edges, 6$.

LANG (Andrew) Edited by — cont.

THE ARABIAN NIGHTS EN- TERTAINMENTS. With 66 Illustra- tions. Crown 8vo, gilt edges, 6s.

TH K BOOK OF ROMANCE. With

8 Coloured Plates and 44 other Illustrations. Crown 8vo, gilt edges, 65.

LYALL. — THE BURGES

LETTERS : a Record of Child Life in the Sixties. By Edna Lyall. With Coloured Frontispiece and 8 other Full-page Illustra- tions by Walter S. Stacey. Crown 8vo, 2s. 6rf.

MEADE (L. T.).

DADDY'S BOY. With 8 Illustra- tions. Crown 8vo, gilt edges, 3s. net.

DEB AND THE DUCHESS. With 7 Illustrations. Cr. 8vo, gilt edges, 3s. net.

THE BERESFORD PRIZE. With

7 Illustrations. Cr. 8vo, gilt edges, 3s. net.

THE HOUSE OF SURPRISES.

With 6 Illustrations. Crown 8vo, gilt edges, 3s. net.

PACKARD. -THE YOUNG ICE WHALERS : a Tale for Boys. By Win- throp Packard. With 16 Illustrations. Crown 8vo, 6s.

PENROSE.— CHUBBY: A NUI- SANCE. By Mrs. Penrose. With 8 Illus- trations by G. G. Manton. Crown 8vo, 3s. U.

PRAEGER (Rosamond).

THE ADVENTURES OF THE THREE BOLD BABES : HECTOR, HONORIA AND ALISANDER A Story in Pictures. With 24 Coloured Plates and 24 Outline Pictures. Oblong 4to, 3s. 6rf.

THE FURTHER DOINGS OF THE THREE BOLD BABES. With 24 Coloured Pictures and 24 Outline Pictures. Oblong 4to, 3s. 6d.

ROBERTS.— THE ADVENTURES

OF CAPTAIN JOHN SMITH: Captain of Two Hundred and Fifty Horse, and some- time President of Virginia. By E. P. Roberts. With 17 Illustrations and 3 Maps. Crown 8vo, 5s. net.

STEVENSO N.— A CHILD'S GARDEN OF VERSES. By Robert Louis Stevenson, Fcp. 8vo, gilt top, Ss.

LONGMANS AND CO. S STANDARD AND GENERAL WORKS,

33

Children's Books — continued.

UPTON (Florence K. and Bertha).

THE ADVENTURES OF TWO DUTCH DOLLS AND A 'GOLLI -i WOGG '. With 31 Coloured Plates and numerous Illustrations in the Text. Oblong I 4to, 6s.

THE GOLLIWOGG'S BICYCLE CLUB. With 31 Coloured Plates and i numerous Illustrations in the Text. Oblong 4to, 6s.

THE COLLI WOGG AT THE SEASIDE. With 31 Coloured Plates and ' numerous Illustrations in the Text. Ob- { long 4to, 6s. i

THE GOLLIWOGG IN WAR.' With 31 Coloured Plates. Oblong 4to, 6s. |

UPTON (Florence K. and Bertha)

— continued.

THE GOLLIWOGG'S POLAR ADVENTURES. With 31 Coloured Plates. Oblong 4to, 6s.

THE GOLLIWOGG'S AUTO-GO- CART. With 31 Coloured Plates and numerous Illustrations in the Text. Ob- long 4to, 6s.

THE GOLLIWOGG'S AIR-SHIP, With 30 Coloured Pictures and numerous Illustrations in the Text. Oblong 4to, 6s.

THE GOLLIWOGG'S CIRCUS.

With 31 Coloured Pictures. Oblong 4to, boards, 6s.

THE VEGE-MEN'S REVENGE.

With 31 Coloured Plates and numerous Illustrations in the Text. Oblong 4to, 6s.

THE SILVER LIBRARY.

Crown 8vo, 3s. 6^. each Volume.

Arnold's (Sir Bdwin) Seas and Lands. ! Buckle's (H. T.) History of Civilisation in

With 71 Illustrations. 3s. 6^. ' Enflfland. 3 vols. 10s. 6d.

Bagrehot's (W.) Biosraphicai Studies, j Churchill's (Winston S.) The Story of the 3^-6^ ; Malakand Field Force, 1897. With 6

Bairehot's (W.) Economic Studies. 3s. 6d.

Bagehot's (W.) Literary Studies. With Portrait. 3 vols., 3s. 6^. each.

Baker's (Sir S. W.) Elfifht Years in Ceylon.

With 6 Illustrations. 3s. 6d.

Baker's (Sir S. iV.) Rifle and Hound in Ceylon. With 6 Illustrations. 3s. 6^.

Baringr-Qould's (Rev. S.) Curious Myths of the Middle Agres. 3s. Sd.

Baringr-Gould's (Rev. S.) Oriein and De- velopment of Relisfious Belief. 2 vols. 3s. 6a. each.

Becker's (W. A.) Qallus : or, Roman Scenes in the Time of Augustus. With 26 Illustra- tions. 3s. Sd.

Becker's (W. A.)Charicles : or. Illustrations of the Private Life of the Ancient Greeks. With 26 Illustrations. 3s. 6d.

Bent's (J. T.) The Ruined Cities of Ma- shonaland. With 117 Illustrations. 3s. 6d.

Brassey's (Lady) A Voyage in the ' Sun- boam '. With 66 Illustrations. 3s. 6d.

Maps and Plans. 3s. 6^.

Clodd's (E). Story of Creation : a Plain Ac- count of Evolution. With 77 Illustrations. 3s. 6d.

Conybeare (Rev. W. J.) and Howson's {Very Rev. J. S.) Life and Epistles of SL Paul. With 46 Illustrations. 3s. 6^.

Dousrall's (L.) BeflTffars All : a Novel. 3s. Sd.

Doyle's (Sir A. Conan) Micah Clarke. A

Tale of Monmouth's Rebellion. With 10 Illustrations, 3s. 6d.

Doyle's (Sir A. Conan) The Captain of the Polestar, and other Tales. 3s. 6^.

Doyle's (Sir A. Conan) The Ref ugrees : A

Tale of the Huguenots. With 25 Illustra- tions. 3s. 6^.

Doyle's (Sir A. Conan) The Stark Munro Letters. 3s. 6^.

Fronde's (J. A.) The History of Enjrland,

from the Fall of Wolsey to the Defeat of the Spanish Armada. 12 vols. 3s. 6^. each.

[!iii;f-T

ffli<i<»l!|g»§Mt.»;.l T.-.w,«...

Kr.airjar.avjiK.atv.atr »

LONGMANS ANt> CO.'s STANDARD AND GENERAL WORKS. 35

THE SILVER LIBRARY— con^inw^^.

Knight's (E. F.) The 'Falcon' on the | Proctor's (R. A.) The Orbs Around Us.

Baltic : a Coasting Voyage from Hammer- 35. %d.

smith to Copenhagen in a Three-Ton Yacht. With Map and 11 Illustrations. 3s. %d.

Ko8tlfn*8 (J.) Life of Luther. With 62 Illustrations and 4 Facsimiles oF MSS. 3s. %d.

Lansr's (A.) Angrlinor Sketches. With 20 Illustrations. 3:. ^cT.

Langr^s (A.) Custom and Myth : Studies of Early Usage and Belief. 3s. 6i.

Langr's (A.) Cock Lane and Common- Sense. 3s. %d.

Land's (A.) The Book of Dreams and Ghosts. 3s. %i.

Langr's (A.) A Monk of Fife : a Story of the Days of Joan of Arc. With 13 Illustrations. 3s. 6d.

Langr's (A.) Myth, Ritual, and Religion. 2

vols. 7s.

Lees (J. A.) and Clutterbuck's (W. J.) B.C. 1887, A Ramble in British Columbia.

With Maps and 75 Illustrations. 35. %d.

Levctt- Yeats' iS.) D'Auriac. 3s. 6^.

The Chevalier

Macaulay's (Lord> Complete Works. ' Al- bany'Edition. With 12 Portraits. 12 vols. 35. %d. each.

Macaulay's (Lord) Essays and Lays of Ancient Rome. etc. With Portrait and 4 Illustrations to the ' Lays '. 3s. 6rf.

Macleod's (H. D.) Elements of Bankingr.

3s. %d.

Marshman's (J. C.) Memoirs of Sir tienry Havelock. 3s. 6d.

Mason (A. E. W.) and Langr's (A) Parson Kelly. 3s. 6d.

Merivale's (Dean) History of the Romans under the Empire. 8 vols. 3s. 6^. each.

Merriman's (H. S.) Flotsam : a Tale of the Indian Mutiny. 3s. 6<i.

Mill's (J. S.) Political Economy. 35. %d. Mill's (J. S.) System of Logrlc. 3s. W.

Milner's (Geo.) Country Pleasures: the

" Chronicle of a Year chiefly in a Garden. 3s. %d.

Nansen's (F.) The First Crossing of Green- land. With 142 Illustrations and a Map. 3s. %d.

Proctor's (R. A.) The Expanse of Heaven.

3s. 6rf. Proctor's (R. A.) Ligrht Science for Leisure

Hours. 3s. 6rf.

Proctor's (R. A.) The Moon. 3s. 6rf.

Proctor's (R. A.) Other Worlds than Ours.

3s. 6d.

Proctor's (R. A.) Our Place amongr Infini- ties : a Series of Essays contrasting our Little Abode in Space and Time with the Infinities around us. 3s. %d.

Proctor'^ (R. A.) Other Suns than Ours. 3s. %d.

Proctor's (R. A.) Rougrh Ways made Smooth. 35. M.

Proctor's (R. A.) Pleasant Ways In Science. 35. 6^.

Proctor's (R. A.) Myths and Marvels of Astronomy. 3s. 6d.

Proctor's (R. A.) Nature Studies. 3s. 6 /.

Proctor's (R. A.) Leisure Readingrs. By

R. A. Proctor, Edward Clodd. Andrew Wilson, Thomas Foster, and A. C. Ran- yard. With Illustrations. 3s. 6i/.

Rossetti's (Maria F.) A Shadow of Dante.

3s. %d.

Smith's (R. Bosworth) Carthage and the Cartha^nians. With Maps, Plans, etc. 3s. %d.

Stanley's (Bishop) Familiar History of Birds. With 160 Illustrations. 3s. M.

Stephen's (Sir Leslie) The Playgrrpund of Europe (The Alps). With 4 Illustrations. fV3s. M.

Stevenson's (R. L.) The Strange Case of Dr. Jekyll and Mr. Hyde; with other Fables. 35. 6i.

Stevenson (R. L.) and Osbourne's (LI.) The Wrongr Box. 3s. 6d.

Stevenson (Robert Louis) and Stevenson's (Fanny van de Grift) More New Arabian Nights.— The Dynamiter. 3s. 6i.

Trevelyan's (Sir G. 0.)The Early History of Charles James Fox. 3s. %d.

Weyman's (Stanley J.) The House of the Wolf : a Romance. 3s. 6rf.

Wood's (Rev. J. G.) Pctland Revisited.

With 33 Illustrations. 3s. %d.

Wood's (Rev. J. G.) Stransre Dwelling;*.

I With 60 Illustrations. 3s. 6a.

Phillipps-WoUey's (C.) Snap : a Legend of 1 Oat of Doors. With U

the Lone Mountain, With 13 Illustrations, i Wood's (Rev. J. Q.) Out ol UOOrs. vv«n

3j g^ I Illustrations. 6s. ba.

>l

36

LONGMANS AND CO. S STANDARD AND GENERAL WORKS,

Cookery, Domestic

ACTON. — MODERN COOKERY.

By Eliza Acton. With 150 Woodcuts. , Pep. 8vo, 4s. %d.

ANGWIN.— SIMPLE HINTS ON 1

CHOICE OF FOOD, with Tested and Economical Recipes. For Schools, Homes, and Classes for Technical Instruction. By M. C. Anowin, Diplomate (First Cla'Cs) of , the National Union tor the Technical Train- ing of Women, etc. Crown 8vo, \s.

ASHBY.— HEALTH IN THE NUR- SERY. By Henry Ashhy, M.D., F.R.C.P., Physician to the Manchester Children's Hos- pital. With 25 Illustrations. Crown 8vo. 3s. ' net.

BULL (Thomas, M.D.). I

HINTS TO MOTHERS ON THe' MANAGEMENT OF THEIR HEALTH DURING THE PERIOD OF PREG- NANCY. Fcp. 8vo, sewed, Is. &/. ; cloth, gilt edges, 'Is. net.

THE MATERNAL MANAGE- MENT OF CHILDREN IN HEALTH AND DISEASE. Fcp. 8vo, sewed. Is. ft/. ; cloth, gilt edges, 2s. net.

DE SALIS (Mrs.).

X LA .MODE COOKERY : Up to- date Recipes. With 24 Plates (16 in Colour). Crown 8vo, 5s. net.

CAKES AND CONFECTIONS A LA MODE. Fcp. 8vo, Is. %d.

DOGS : A Manual for Amateurs. Fcp. 8vo, Is. M.

DRESSEDGAME AND POU LTRY X LA MODE. Fcp. 8vo, Is. (v/.

DRESSED VEGETABLES X LA MODE. Fcp. 8vo, Is. %d.

DRINKS X LA MODE Fcp. 8vo,

l5. &/.

Manasrement, etc.

DE SALIS (W.r9.)—continued.

ENTRIES X LA MODE. Fcp. 8vo. Is. %d.

FLORAL DECORATIONS. Fcp.

8vo, \s.%d.

GARDEMING X LA MODE. Fcp.

8vo, Part I., Vegetables, Is. 6rf. Part IL, Fruits, Is. %d.

NATIONAL VIANDS X LA MODE,

Fcp. 8vo, Is. 6rf.

NEW-LAID EGGS. Fcp. 8vo. Is. %d.

OYSTERS X LA MODE. Fcp. 8vo, Is. %d.

PUDDINGS AND PASTRY X LA

MODE. , Fcp. 8vo, Is. %d.

SAVOURIES X LA MODE. Fcp. 8vo, Is. W.

SOUPS AND DRESSED FISH X LA MODE. Fcp. 8vo, Is. %d.

SWEETS AND SUPPER DISHES X LA MODE. Fcp. 8vo, Is. 6rf.

TEMPTING DISHES FOR SMALL INCOMES. Fcp. 8vo, Is. &/.

WRINKLES AND NOTIONS FOR EVERY HOUSEHOLD. Crown 8vo,

Is. 6(/.

LEAR.— MAIGRE COOKERY. By

H. L. Sidney Lear. 16mo, 2s.

POOLE. — COOKERY FOR THE DIABETIC. By W. H. and Mrs. Poole. With Preface by Dr. Pavy. Fcp. 8vo, 2s. 6rf.

ROTHERAM. — HOUSEHOLD

COOKERY RECIPES. By M. A. Rothe- KAM, First Class Diplom^e, National Training School of Cookery, London ; Instructress to the Bedfordshire County Council. ' Crown 8vo, 2s.

The Fine Arts and Music.

BURNE-JONES.— THE BEGIN- NING OF THE WORLD: Twenty five Pictures by Sir Edward Burne-Jones, Bart. Medium 4to, Boards, 7s. 6rf. net.

BURNS AND COLENSO.-

LIVING ANATOMY. By Cecil L. Burns, R.B.A., and Robert J. Colenso, M.A , M.D. 40 Plates, llj by 8^ ins., each Plate containing Two Figures— (a) A Natural Male or Female Figure; (b) The same Figure Anatcmatised. In a Portiolio, 7.s. %d. net.

HAMLIN.— A TEXT-BOOK OF THE HISTORY OF ARCHITECTURE. By A. D. F. Hamun, A.M. With 229 Illus- trations. Crown 8vo, 7s. 6rf.

HAWEIS (Rev. H. H.).

MUSIC

Portrait net.

AND MORALS. With

of the Author. Crown 8vo, 6s.

MY MUSICAL LIFE. With Por- trait of Richard Wagner and 3 Illustrations.

Crown 8vo, 6s. net. ^

LONGMANS AND CO, S STANDARD AND GENERAL WORKS. 37

The Fine Arts and Music — continued.

HUISH, HEAD AND LONG- MAN. SAMPLERS AND TAPESTRY EMBROIDERIES. Bv Marcus B. Huish, LL.B. : also ' The Stitchery of the same.' by Mrs. Hbad; and Foreign Samplers,' by Mrs. C. J. Longman. With 30 Reproduc- tions in Colour, and 40 Illustrations in Monochrome. 4to, £2 2s. net.

HULLAH.— THE HISTORY OF MODERN MUSIC. By John Hullah. 8vo, 8s. %d.

JAMESON (Mrs. Anna).

SACRED AND LEGENDARY

ART, containing Legends oF the Angels and Archangels, the Evangel ists.'^he Apostles, the Doctors of the Church, St. Mary Magdalene, the Patron Saints, the Martyrs, the Early Bishops, the Hermits, and the Warrior-Saints of Christendom, as repre- sented in the Fine Arts. With 19 Etchings and 187 Woodcuts. 2 vols. 8vo, 20s. net.

LEGENDS OF THE MONASTIC ORDERS, as represented in the Fine Arts, comprising the Benedictines and Augustines, and Orders derived from their Rules, the Mendicant Orders, the Jesuits, and the Order of the Visitation of St. .Mary. With 11 Etchings and K8 Wood- cuts. 1 vol. 8vo, 10s. net.

LEGENDS OF THE MADONNA, OR BLESSED VIRGIN MARY. De- votional with and without the Infant Jesus, Historical from the Annunciation to the Assumption, as represented in Sacred and Legendary Christian Art. With 27 Etchings and 165 Woodcuts. 1 vol. 8vo, 10s. net.

THE HISTORY OF OUR LORD,

as exemplified in Works of Art, with that of His Types, St. John the Baptist, and other persons of the Old and New Testa- ment. Commenced by the late Mrs. Jameson ; continued and completed by Lady Eastlake. With 31 Etchings and 281 Woodcuts. 2 vols. 8vo, 20s. net.

KRISTELLER.— ANDREA MAN-

TEGNA. By Paul Kristeller. English Edition by S. Arthur Strong, M.A., Librarian to the House of Lords, and at Chatsworth, With 26 Photogravure Plates and 162 Illustrations in the Text. 4to, gilt top, £3 10.9. net.

MACFARREN.— LECTURES ON

HARMONY. By Sir George A. Mac- parren. 8vo, 12s.

MATTHAY. — THE ACT OF

TOUCH IN ALL ITS DIVERSITY. An Analysis and Synthesis oF Pianoforte Tone Production. By Tobias Matthav, Fellow and Professor of the Royal Academy of Music, London, etc. With 22 Illustrations. 8vo, 7s. %d.

MORRIS (William).

ARCHITECTURE, INDUSTRY AND WEALTH. Collected Papers. Cr. 8vo, 6s. net.

HOPES AND FEARS FOR ART.

Five Lectures delivered in Birmingham, London, etc., in 1878-1881. Cr, 8vo, 4s. 6d.

AN ADDRESS DELIVERED AT THE DISTRIBUTION OF PRIZES TO STUDENTS OF THE BIRMINGHAM .MUNICIPAL SCHOOL OF ART ON 2l8T FEBRUARY, 1894. 8vo, 2s. %d. net. (Printed in ' Golden ' Type.)

SOME HINTS ON PATTERN- DESIGNING: a Lecture delivered at the Work'ng Men's College, London, on 10th December, 1881. 8vo, 2s. 6d. net. (Printed in * Golden ' Type.)

ARTS AND ITS PRODUCERS (1888) AND THE ARTS AND CRAFTS OF TO-DAY (1889). 8vo, 2s. Sd. net. (Printed in ' Golden ' Type.)

ARTS AND CRAFTS ESSAYS.

By Members of the Arts and Crafts Exhibition Society. With a Preface by William Morris. Crown 8vo, 2s. 6i. net.

**" For Mr. William Morris's other Works, see pp. 24, 27, 28 and 40.

ROBERTSON.— OLD ENGLISH

SONGS AND DANCES. Decorated in Colour by W. Graham Robertson. Royal 4to, 42s. net.

SCOTT.— PORTRAITURES OF

JULIUS C>ESAR: a Monograph. By Frank Jbsup Scott. With 3S Platt^s and 49 Figures in the Text. Imperial 8vo, 21i. net.

38 LONGMANS AND CO's STANDARD AND GENERAL WORKS,

The Fine Arts and Music — continued.

VANDERPOEL. -COLOUR PRO- BLEMS : a Practical Manual for the Lay Student of Colour. By Bhily Noybs Vandbrpobl. With 117 Plates in Colour. Sq. 8vo, 21 5. net.

VAN DYKE.— A TEXT-BOOK ON

THE HISTORY OF PAINTING. By John C. Van Dvkb. With 110 Illustrations. Crown Svo, 6s.

WOT TON.— THE ELEMENTS OP ARCHITECTURE. Collected by Hbnry Wotton, Kt.. from the best Authors and Examples. Royal 16mo. boards, 10.<t. 6^. net.

WELLINGTON. — A DESCRIP- TIVE AND HISTORICAL CATALOGUE OF THE COLLECTIONS OF PICTURES AND SCULPTURE AT APSLEV HOUSE. LONDON. By Evelyn. Duchess of Wel- lington. Illustrated by 52 Photo-Engravings, specially executed by Braun, Clement & Co., of Paris. 2 vols. Royal 4to, £6 65. net.

W I L L A R D. — HISTORY OF I MODERN ITALIAN ART. By Ashton

Rollins Willard. Part I. Sculpture, f Part II. Painting. Part HI. Architecture. t|With Photogravure Frontispiece and

numerous full-page Illustrations. Svo, 2l5.

net.

Miscellaneous and Critical Works.

AUTO DA f6 and OTHER EVANS.— THE ANCIENT STONE

Essasrs : some being Essays in Fiction By tlie Author of ' Essays in Paradox ' and ' Exploded Ideas '. Crown Svo, 55.

BAGEHOT. — LITERARY

STUDIES. By Waltbr Baorhot. With Portrait. 3 vols. Crown Svo. 3s. 67. each.

IMPLEMENTS, WEAPONS AND ORNAMENTS OF GREAT BRITAIN. By Sir John Evans, K.C.B. With 537 Illus- trations. Svo, 10s. 6(i. net.

FITZWYGRAM.— HORSES AND

STABLES. By Lieut.-General S?r F. Fitz- wygram, Bart. With 56 pages of Illustra- tions. Svo, 3s. net.

MYTHS OF THE MIDDLE AGES. By i c'DrkC'r A Mcrfcl CV X>ru\\Z o Rev. S. Baring-Gould. Crown Svo, 3s. %d. I FROST.— A MEDLEY BOOK. By

George Frost. Crown Svo, 3s. 6</. net

BARING -GOULD.— CURIOUS

MYTHS OF THE MIDDLE AGES. B

B A Y N E S. — SHAKESPEARE STUDIES, and other Essays. By the late Thomas Spencer Baynbs, LL.B., LL.D. With a Biographical Preface by Professor Lewis Campbell. Crown Svo, 7s. 6(/.

BONNELL. — CHARLOTTE

BRONTfi, GEORGE ELIOT, JANE AUSTEN: S*^udies in their Works. By Henry H. Bonnbll. Crown Svo, 7s. 6.t. net.

BOOTH.— THE DISCOVERY AND DECIPHERMENT OF THE TRILING- UAL CUNEIFORM INSCRIPTIONS. By Arthur John Booth, M.A. With a Plan of Persepolis. Svo, \Ai. net.

CHARITIES REGISTER, THE

ANNUAL, AND DIGEST: being a Clas- sified Register of Charities in or available in the Metropolis. Svo, 5s. net.

CHRISTIE. — SELECTED ESSAYS. By Richard Copley Christie, M.A., Oxon. Hon. LL.D., Vict. With 2 Por- traits and 3 other Illustrations. Svo, 12s. net.

DICKINSON. — KING ARTHUR IN CORNWALL. By W. Howship Dickin- son, M.D. With 5 Illustrations. Crown Svo, 4s. 6<i.

ESSAYS IN PARADOX. By the

Author of ' Exploded Ideas ' and ' Tinves and Days '. Crown Svo, 5s.

GEIKIE.— THE VICAR AND HIS FRIENDS. Reported by Cunningham Gbikib, D.D , LL.D. Crown Svo, 5s. net.

GILKES.— THE NEW REVOLU- TION. By A. H. GiLKBS, Master of Dulwich College. Fcp. Svo, Is. net.

HAGGARD (H. Rider).

A FARMER'S YEAR: being his Commonplace Book for 1898. With 96 Illustrations. Crown Svo, 75. 6£2. net.

RURAL ENGLAND. With 23 Agricultural Maps and 56 Illustrations from Photographs. 2 vols., Svo, 36s. net.

HARVEY- BROOKS.-MAR-

RIAGE AND MARRIAGES: Before and After, for Young and Old. By E. C. Harvbv- Brooks. Crown Svo, 4s. net.

HODGSON.— OUTCAST ESSAYS

AND VERSE TRANSLATIONS. By Shadworth H. H0DO8ON. Crown Svo, 8s. &i.

HOENIG.— INQUIRIES CON- CERNING THE TACTICS OF THE FUTURE. By Fritz Hobnig. With I Sketch in the Text and 5 Maps. Translated by Captain H. M. BowBR. Svo, 15>. net.

HUTCHINSON.-DREAMS AND

THEIR MEANINGS. By Horacb G. HUTCHIN60N, Svo. gilt. top, 9s. 6^. net.

LOIfGMANS AND CO. S STANDARD AND GENERAL fVOR/CS.

39

Miscellaneous and Critical Works — continued.

JEFFERIES (Richard).

FIELD AND HEDGEROW: With Portrait. Crown 8vo, Ss. Sd.

THE STORY OF MY HEART;

my Autobiography. Crown 8vo, 3s. 6^.

RED DEER. With 17 Illustrations. Crown 8vo, Ss. %d.

THE TOILERS OF THE FIELD. Crown Svo, 3s. %d.

WOOD MAGIC : a Fable. Crown Svo, 3s. %d.

JEKYLL (Gertrude).

HOME AND GARDEN: Notes and Thoughts, Practical and Critical, of a Worker in both. With 53 Illustrations from Photographs. Svo, lOs. 6^. net.

WOOD AND GARDEN : Notes and Thoughts, Practical and Critical, of a Working Amateur. With 71 Photographs. Svo, 10s. 6^. net.

OLD WEST SURREY: Some Re- collections. With numerous Illustrations from Photographs by the Author. Svo.

JOHNSON (J. & J. H.).

THE PATENTEE'S MANUAL: a Treatise on the Law and Practice of Letters Patent. Svo, 10s. %d.

AN EPITOME OF THE LAW AND PRACTICE CONNECTED WITH PATENTS FOR INVENTIONS, with a reprint of the Patents Acts of 18S3, 1S85, 1SS6 and 1S8S. Crown Svo, 2s. %d.

JOYCE. —THE ORIGIN AND

HISTORY OF IRISH NAMES OF PLACES. By P. W. Joyce, LL.D. 2 vols. Crown Svo, 5s. each.

LANG (Andrew).

LETTERS TO DEAD AUTHORS. Fcp. Svo, 2s. %d. net.

BOOKS AND BOOKMEN. With

2 Coloured Plates and 17 Illustrations.

Fcp. Svo, 2s. M. net. OLD FRIENDS. Fcp. Svo, 25. 6rf.

net. LETTERS ON LITERATURE.

Fcp. Svo, 2s. %d. net.

ESSAYS IN LITTLE. With Por- trait of the Author. Crown Svo, 2s. 6i.

COCK LANE AND COMMON- SENSE. Crown Svo. 3s. 64.

THE BOOK OF DREAMS AND GHOSTS. Crown Svo, 3s. W.

MATTHEWS. —NOTES ON SPEECH-MAKING. Bv B asder Mat- thews Fcp. Svo, Is. M. net.

MAX MULLER (The Right Hon. F.)

COLLECTED WORKS. 20 vols. Vols. I.-XIX. Crown Svo, 5s. each. Vol. XX., 7s. 6d. net.

Vol. I. Natural Religion : the

Gifford Lectures, 1S88.

Vol. II. Physical Religion : the

Gifford Lectures, 1S90.

Vol. III. Anthropological Re- ligion : the Gifford Lectures, IS9I.

Vol. IV. Theosophy ; or, Psycho- logical Religion : the Gifford Lectures, 1S92.

CHIPS FROM A GERMAN

WORKSHOP.

Vol. V. Recent Essays and Ad- dresses.

Vol. VI. Biographical Essays.

Vol. VII. Essays on Language and ' Literature.

Vol. VIII. Essays on Mythology and Folk-lore.

Vol. IX. The Origin and Growth

OP Religion, as Illustrated by the Re- ligions of India : the Hibbert Lectures, I87S.

Vol. X. Biographies of Words,

AND THE Home op the Aryas. Vols. XI., XII. The Science of

Language : Founded on Lectures de- livered at the Royal Institution in 1S61 and 1S63. 2 vols. 10s.

Vol. XIII. India: What can it Teach Us?

Vol. XIV. Introduction to the Science op Religion. Four Lectures, 1870.

Vol. XV. RiMAK^ysHATA : his Life and Sayings.

Vol. XVI. Three Lectures on the

Vedanta Philosophy, IS94. Vol. XVII. Last Essays. First

Series. Essays on Language, Folk-lore, etc.

Vol. XVIII. Last Essays. Second

Series. Essays on the Science of Religion.

Vol. XIX. The Silesian Horse- herd (' Das Pferdebiirla ') : Questions of the Hour answered by F. Max MOllbr. Translated by Oscar A. Fbchtbr, Mayor of North Jakima, U.S.A. With a Preface by J. EsTLiN Carpenter.

*«* This is a translation of a work which was published some years back in Germany, but which is now for the first time translated into English. It consists of a controversy on re- ligion carried on between Professor Max MUller and an unknown correspondent in A merica.

Vol. XX. The Six Systems op Indian Philosophy. Crowa Svo, li. M. net.

40 LOl^GMANS AND CO.'s STANDARD AND C.ENKRAL n*ORKS,

Miscellaneous and Critical Works — continued.

M I LNER.— COUNTRY PLEAS- 1

UKBS : the Chronicle of a Year chiefly in a Garden. By George Milnbh. Crown 8vo,

MORRIS.— SIGNS OF CHANGE.

Seven Lectures delivered on various Occa- sions. By William Morhis. Post 8vo, 45. U.

PARKER AND UNWIN.— THE

ART OF BUILDING A HOME: a Collec- tion of Lectures and Illustrations. By Barry Parker and Raymond Unwin. With 68 Full-page Plates. 8vo, 10^. M. net.

POLLOCK.— JANE AUSTEN : her Contemporaries and Herself. By Walter Herri Bs Pollock. Crown 8vo, 3s. 6<f. net.

POORE (George Vivian, M.D.). ESSAYS ON RURAL HYGIENE.

With 13 Illustrations. Crown 8vo, 6s. 6<^.

THE DWELLING HOUSE. With

36 Illustrations. Crown 8vo, 3s. 6rf.

THE EARTH IN RELATION TO THE PRESERVATION AND DES- TRUCTION OF CONTAGIA: being the Milroy Lectures delivered at the Royal College of Physician** in 1899, touether with other Papers on Sanitation. With 13 Illustrations. Crown 8vo, 5s.

COLONIAL AND CAMP SANI- TATION. With 11 Illustrations. Crown 8vo, 2s. net.

ROSSETTL— A SHADOW OF

DANTE : being an Essay towards studying Himself, his World and his Pilgrimage. By Maria Francesca Rossetti. Crown 8vo, 3s. 6 J.

SERIA LUDO. By a Dilettante. 1

Post 4to, 5s. net.

*^'^ Sketches and Verses, mainly reprinted from the St. James's Gazette.

SHADWELL. DRINK: TEM- PERANCE AND LEGISLATION. By Arthur Shadwell, M.A., M.D. Crown 8vo, 55. net.

SOULSBY (Lucy H. M.).

STRAY THOUGHTS ON READ- ING. Fcp. 8vo, cloth, 2s. 6d. net ; limp leather, gift edges, 3s. 6^. net.

STRAY THOUGHTS FOR GIRLS.

Fcp. 8vo, cloth, 2s. 6d. net. ; limp leather, gilt edges, 3s. Gd. net.

*** Copies of the Original Edition can still be had. 16mo, Is. 6d. net.

STRAY THOUGHTS FOR MOTHERS AND TEACHERS. Fcp. 8vo, cloth, 2s. Sd. net. ; limp leather, gilt edges, 3s. Sd. net.

25,000—2/04.

SOULSBY (Lucy H. }A.)— continued.

STRAY THOUGHTS ON CHAR- ACTER. Fcp. 8vo, cloth, 2s. 6^/. net.: limp leather, gilt edges, 3s. %d. net.

STRAY THOUGHTS FOR VALIDS. 16mo, 2s. net.

IN

SOUTHEY.— THE CORRESPON- DENCE OF ROBERT SOUTHEY WITH CAROLINE BOWLES. Edited by Ed- ward Dowden. 8vo, 14s.

STEVENS.— ON THE STOWAGE OF SHIPS AND THEIR CARGOES. With Information regarding Freights, Char- ter-Parties etc. By Robert White Ste- vens. 8vo, 21s.

THUILLIER. — THE PRIN- CIPLES OF LAND DEFENCE, AND THEIR APPLICATION TO THE CON- DITIONS OF TO-DAY. By Captain H. F. Thuillier, R.E. With Maps and Plans 8vo, 12s. Gd. net.

TURNER AND SUTHERLAND.

—THE DEVELOPMENT OF AUS- TRALIAN LITERATURE. Bv Henry Gyles Turner and Alexander Suther- land. With Portraits and Illustrations. Crown 8vo, 5s.

WARD. PROBLEMS AND PER- SONS. By Wilfrid Ward, Author of ' The Life and Times of Cardinal Wiseman/ &c. 8vo, 14s. net.

Contents.— The Time-Spirit of the Nine- teenth Century— The Rigidity of Rome— Un- changing Dogma and Chanceful Man — Balfour's • The Foundations of Belief — Candour in Bio- graphy— ^Tennyson — Thomas Henry Huxley — Two Mottoes of Cardinal Newman — Newman and Rn^an — Some Aspects of the Life-work of Cardinal Wiseman — The Life of Mrs. Augustus Craven.

WEATHER S.— A PRACTICAL GUIDE TO GARDEN PLANTS. By John Weathers, F.R.H.S. With 159 Diagrams. 8vo, 21s. net.

WINSTON.— MEMOIRS OF A

CHILD. By Annie Steobr Winston.

Fcp. 8vo, 2s. Gd. net.

Contents— I. The Child and the Child's Earth. — II. People. — III. The Garden and a few Re- lated Things.— IV. Divers Delights.— V. The Child and 'The Creatures '.—VI. Playthings. — . VII. Portable Property.— VIII. Pomps and Vanities. — IX. Social Divertisements. — X. Con- duct and Kindred Matters. — XI. Dreams and Reveries. — XII. Bughears. — XIII. Handicraft. —XIV. School, Slightly Considered. — XV. Books. — XVI. Language.- XVII. Random Reflections. — Conclusion.

.d