Full text of "Map Interpretation"

See other formats

DO =

u< OU 158843

7) ~X

^ DO

OUP 408 16-6-64 5,000.

OSMANIA UNIVERSITY LIBRARY

Call No. 5<5>'8/I> 96 Fl Accession No.

Author

Title MCXJD Vntetwetokon ,

This book should be returned on or before the date last marked below.

MAP INTERPRETATION

G. H. DURY, M.A., Ph.D., F.G.S.

Lecturer in Geography, Birkbeck College
(University of London)

WITH A FOREWORD BY
A. E. MOOD IE, B.A., Ph.D.

Reader in Geography, Birkbeck College
(University of London)

AND A CONTRIBUTION BY
H. C. BROOKFIELD, B.A., Ph.D.

LONDON

SIR ISAAC PITMAN & SONS, LTD.
PITMAN HOUSE, PARKER ST., KINGSWAY, W.C.a
BATH MELBOURNE JOHANNESBURG

First published 1952
Reprinted 1957

Second edition 1960
Reprinted 1962

G. H. Dury
1960

MADE IN GREAT BRITAIN AT THB PITMAN PRESS, BATH
F2 (E.5I9<5)

FOREWORD

THE statement that maps are the geographer's tools has been repeated
so often that it is in danger of becoming a platitude: yet there can be
no doubt as to the validity of its meaning. It is difficult, if not impos-
sible, to imagine a course of geographical study being followed
successfully without constant recourse to maps of a wide variety in
type and scale. Ability to use these maps is an essential requirement or
all geographers, but it is not an easy skill to acquire. After more than
twenty years of experience in teaching undergraduates, I am fully
aware of the difficulties which confront the uninitiated in map inter-
pretation. There is a world of difference between a Boy Scout finding
his way from A to B by map-reading and a geographer seeking to
illuminate his studies by map interpretation, although the map is a
"tool" in each case. The latter requires the ability not only "to see
solid" but also to grasp and explain the relationships between all the
conditions which the map represents in symbolical form.

Some people seem to have a flair in this direction, but competence
can be acquired only by application and frequent practice. Dr. Dury
does not claim to have covered all the problems of his subject, but the
student who masters the principles laid down here will be well on the
way to a mastery of map interpretation: he will have learned the use
of his tools in the sense that they are peculiar to geographical study.

Such an achievement will be commendable, but this book also
draws attention to a major problem in modern geography. Dr. Dury
lays emphasis on the importance of adequate knowledge of physical
geography in map interpretation, but he by no means neglects the
human aspects. Indeed, the close relationships between man and
nature are brought out clearly. This symbiosis, the core of geography,
is in danger of being overlooked, owing to the increasing tendency or
recent years towards specialized studies. Maps can no more show all
the conditions of an area than they can reveal all the relationships
which exist in it, but they represent the closest approach to an expres-
sion of its wholeness. It follows, therefore, that map interpretation
fosters the development of a balanced outlook and its practice is, in
consequence, a valuable discipline. When combined with field
investigations, as suggested by Dr. Dury, it becomes one of the most

rewarding of activities for geographers at all levels.

A. E. M.

ACKNOWLEDGMENTS

I WISH to acknowledge with gratitude the help I have received in
preparing this book. My colleagues Miss E. M. J. Campbell, M.A.,
and H. C. Brookfield, B.A., Ph.D., have read portions of the manu-
script, while A. E. Moodie, B.A., Ph.D., has offered constructive
suggestions on the whole book. Dr. Brookfield contributed a chapter
(XVI) to the first edition. Parts of his material have been incorporated
in the corresponding chapter in this edition, where considerable
revisions have been made in accordance with changes in map series.
The Controller, H.M.S.O., has permitted O.S. material to be repro-
duced or otherwise made use of in the illustrations; photographs have
been obtained from H.M. Geological Survey and from Hunting
Aerosurveys, Ltd.

G. H. D.

Birkbeck College,
February, 1960.

CONTENTS

CHAP. PAGB

Foreword v

Acknowledgments vi

I. INTRODUCTORY i

Relation of work with maps to work on the ground Map interpretation and
map analysis Background knowledge Bibliographical references Selection
of maps

PART I

PHYSICAL INTERPRETATION
II. PHYSICAL INTERPRETATION: GENERAL . . . .11

General characteristics of the landscape Interpretation in terms of the erosion-
cycle Physiographic subdivision Sketch-mapping

III. SCARPLAND TOPOGRAPHY 22

The scarp-formers The clay vale Structures Drainage Terraces of the
Avon The Stour and the Evenlode Underfit streams

IV. ERODED FOLDS 31

General aspect of Chalk country Limits of the Chalk outcrop Geological
structure The drainage pattern Evolution of the landscape

V. UNGLACIATED UPLAND 40

Erosion-platforms Former base-levels Denudational sequence Superim-
posed drainage

VI. GLACIATED HIGHLAND AND A DRUMLIN FIELD . . 46

Landforms of glaciated highlands Renewed normal erosion Glacial troughs
and through-valleys Diversion of drainage Relation of drainage to structure
Changes of base-level Lowland glaciation Drumlins

VII. CARBONIFEROUS LIMESTONE COUNTRY . . .59

Recognition of limestone country Sinks Other karstic features Thickness
of the limestone Underground cavities Rocks above and below the lime-
stone Evidence of glaciation River development

VIII. COASTS AND SHORELINES 67

The shoreline cycle The shoreline and subaerial cycles

Viii CONTENTS

CHAP. PAGE

IX. LANDSCAPE IN AN ARID CLIMATE 83

Distinctive character of desert landscapes Inselbergs Alluvial plain Pedi-
ments Intermittent drainage Desert basins

X. A COMPLEX LANDSCAPE BASED ON COMPLEX STRUCTURE 88

Hill country of the south-west The Levels Small cuestas and outliers amid
the levels The Mendips and other similar limestone hills Broken hill country
in the east The drainage system The shoreline

PART II

THE FEATURES OF OCCUPANCE
XI. FEATURES OF OCCUPANCE: GENERAL . . . .97

Interpretation of land use Routes The pattern of rural settlement Inter-
pretation of rural settlement: village forms Interpretation of rural settle-
ment: place-names

XII. RURAL SETTLEMENT STUDIES 115

Dominantly nucleated Dominantly dispersed Rural settlement in the
fenland Rural settlement in glaciated highland Recent primary dispersion
The transect chart

XIII. TOWNS: SITE, FORM, SITUATION 134

Effects of re-growth Form Function Situation Site The small market
town Ports Resorts Large towns An industrial town Geometrical
layout of a new town

XIV. PREHISTORIC OCCUPANCE 152

The question of dating Neolithic antiquities Bronze Age antiquities The
Early Iron Age Roman sites Earthworks of the Dark Ages Lynchets
Interpretation from an O.S. map

PART III

SPECIAL TOPICS

XV. MORPHOMETRIC ANALYSIS 167

Geometric analysis Arithmetic analysis Volumetric analysis Clinomctric
analysis

XVI. CARTOGRAPHICAL APPRECIATION (BY H. C. BROOKFIELD,

B.A., Ph.D, and G. H. DURY) 180

Scale The representation of reliefsThe representation of cultural features
Non-landscape features Margins and marginal information An example
of appreciation Comparison of map scries

Index 203

LIST OF ILLUSTRATIONS

FlG. PAGE

1. A scale of symbols suitable for use on physical sketch-maps . .19

2. Interpretation of physique: The Cotswolds and the Avon Valley . 27

3. Interpretation of physique: the western part of Salisbury Plain . 35

4. Diversion of drainage by the agency of ice 48

5. Interpretation of physique: Ingleborough and its surroundings . 61

6. Morphological subdivision : part of the Somerset Plain . . .91

7. Parish boundaries 109

8. A transect chart 130

9. Man-made features in an industrial district 149

10. Analysis of prehistoric evidence . . . . . . .159

11. Generalized contours for the area shown on Sheet 186 of the O.S.

Seventh Series 1/63,360 map 168

12. Some applications of profile-drawing 170

13. Long-profiles, drawn from the map, for some of the rivers shown on

O.S. Seventh Series 1/63,360, Sheet 186 172

14. Some graphic methods of summarizing relief distribution . .175

LIST OF PLATES

PLATE PACING PAGE

I. (A) Aerial view of the mouth of the Fowey River, showing low

plateau inland 100

(B) Ben Nevis rising above surrounding mountains . . .100

II. (A) General view of Ingleborough and Simon Fell . . . 101
(B) The shoreline near Sidmouth 101

III. (A) Aerial view of the Mendip Plateau, with Cheddar Gorge . 116
(B) Part of the fenland, near Boston, seen from the air . . .116

IV. (A) Aerial view of Stow on the Wold 117

(B) Part of Merthyr Tydfil: terraced housing, steelworks, and tip-
heaps 117

Straight mine eye hath caught new pleasures
While the landscape round it measures. MILTON

CHAPTER I
INTRODUCTORY

It is an hieroglyphical and shadowed lesson of the whole
world. SIR THOMAS BROWNE

IT has long been tacitly assumed that a map interpreter has no field
knowledge of the country with which he has to deal; but, now that
field-work has come to be generally accepted as an integral part of
geographical study, this antique fiction should be discarded. It is
clear, of course, that the most extensive and prolonged excursions can
sample only a little of even this country, and that only the most
fortunate will be able to observe, at first hand, contemporary glaciers,
deserts, or tropical landscapes. On the other hand, an impressive
variety of terrain is easily accessible from all the great centres of
population, so that school journeys, university field-classes, and less
formal explorations can exemplify, in the large, much of the material
which must inevitably be compressed and generalized in the text-book.
Hence it is reasonable to urge that students of geography should
become acquainted on die ground with many of the landforms
developed in the normal (humid) cycle of erosion, with the leading
features of limestone country, with glaciated landscapes, and with
shorelines, as well as with a wide range of land-use and settlement
forms. It is no exaggeration to claim that every locality has its own
opportunities for field study of the most rewarding kind. In due
course, then, the interpreter will be capable of bringing knowledge of
real country to bear on his problems not necessarily knowledge of
the actual country represented on a given map, but of country which
is essentially similar. The value of such an approach cannot be over-
stated. It implies the whole difference between performing an abstract
exercise and dealing in realities.

Relation of Work with Maps to Work on the Ground

The question of field-work as an aid to map interpretation devolves,
in part, upon the difference of scale. On the 1/63,360 maps of the
Ordnance Survey, which must be the customary tool for many geo-
graphical purposes, a large number of significant individual features

2 MAP INTERPRETATION

are necessarily suppressed. Even though they are represented, they
may not be distinctive. The only possible corrective is a comparison,
on the ground, between the object itself and its appearance on the map.
Practice in comparison of this kind will ensure a very high yield of
information. The map becomes astonishingly clarified and illuminated
by real experience. So true is this, and so partial and unsatisfactory is
work with the map alone, that one may insist on the ground itself as
the proper place to learn map reading. There are, admittedly, a few
excellent elementary texts, which set out to introduce their users to
the method of representing country on a greatly reduced scale by
means of symbols, and to the manner of transposing the symbols back
into words; but, however skilful one may become in this respect, the
exercise is valueless unless the symbols are understood in terms of
real objects. Thus texts on map reading should be regarded as com-
prehensive and formal additions to study of the home region with its
own maps.

Now the present book does not set out to teach map reading. It is
expected that the user will already possess an elementary vocabulary
of cartographical terms, and will be able, with some assistance from a
key, to perceive the meaning of the several notations employed. We
are concerned here, not with the translation of individual items, but
with the terrain as a whole. In map reading one learns a new language ;
in map interpretation one begins to speak it. The reduced scale of the
map is no longer merely a difficulty to be overcome as well as may be:
it is the means whereby a broad expanse of country can be surveyed at
one time. Herein lies the great value of work with maps. They
provide a *synoptic view and, to the skilled interpreter, reveal the
distributions and interrelations with which geography is peculiarly

concerned.

Map Interpretation and Map Analysis

Map interpretation, like a great deal of geography in general,
involves a synthesis in which complex ideas are deduced and combined
from simple ones. This point is worth emphasizing, for maps lend
themselves readily to a number of attractive exercises in analysis, and
it is important not to mistake the results of these for the results of
interpretation. For instance, one might transfer to a sheet of tracing
paper the roads, or the buildings, represented on one of the O.S.
1/63,360 sheets. The resulting map would indubitably reveal, with all
possible clarity, the distribution in question, but only because all other

INTRODUCTORY 3

distributions had been excluded. Interrelations would no longer
appear. If two distributions were mapped on a single tracing, for
example, those of relief and moorland, the result would be a partial
synthesis, the whole of which is provided by the original map. This
is not to say that selective mapping has no value. It can, on the con-
trary, be highly instructive, just because it throws into relief a single
distribution which on the more detailed map is obscured. A number
of statements in the following text are based on analysis of precisely
this kind, but the problem is ultimately one of cartographic representa-
tion and of skill in handling the map. In a rather different context,
analysis is the correct, indeed the only, method of operation. A
number of techniques have already been evolved for treating map
information statistically, and their number may well be usefully added
to in the course of time. Samples of these methods in their application
to the study of landform are briefly described in Chapter XV.

Background Knowledge

In addition to some knowledge of the elements of map reading, the
interpreter needs a grounding in physical geography and elementary
geology. Within the limits of Part I of this book, where physical
interpretation is treated, there is no room to discuss or summarize all
the fundamental concepts of this branch of study. It must, therefore,
be assumed that the reader is, or soon will be, acquainted, in outline,
with the scheme of the normal cycle of erosion. Moreover, the land-
forms of the normal cycle are likely to be accessible to a majority of
students. In the discussion of maps of shorelines, of glaciated, arid,
and karstic landscapes, however, rather more explanatory detail has
been included, partly because the features concerned offer special
problems to the cartographer and, therefore, to the user of the map.
Throughout the book, where technical terms necessary in description
are introduced for the first time, they have been printed in heavy
type, so that one may easily perceive the range of geographical
vocabulary required, and refer, if necessary, to the relevant pages of a
standard text.

As yet, physical geography is better served in the matter of text-
books than the features of human occupance, which are dealt with in
Part II. In the words of Professor S. W. Wooldridge, we as geo-
graphers are concerned with land genesis, land quality, and land use.
Thanks to the combined efforts of geologists and geomorphologists,
the manner of land genesis is sufficiently well understood. Although a

4 MAP INTERPRETATION

formidable body of material still awaits detailed attention, descriptions
in general terms and some noteworthy special studies are universally
available. Land quality, on the other hand, has long been regrettably
neglected, or disposed of in brief and inadequate phrases. This will
continue until the rudiments of soil science, the use of soil maps, the
detailed investigation of landscape facets, and the factors of micro-
climatology have been combined in an improved geographical dis-
cipline. It is evident, from the beginnings that have been made, that
land quality, like land genesis, will in time be accorded a systematic
treatment, with its elements clearly stated in an orderly form; but,
until then, map interpretation must perforce make do with a minimum
of factual information about land quality, and must move almost
directly from land genesis to land use. In dealing with land use one is
concerned with almost the whole field of human endeavour. On the
time-scale of man's history, even of his written history, the cultural
landscape is an impermanent and rapidly changing complex of features.
Consequently, the problem of interpreting land use, in its widest sense,
from maps, is one which involves a large number of complexly
related variables. This is why, in Part II, there can be no short list of
standard texts for background reading comparable to that suggested
for physical study, and why it has seemed desirable to state principles
more fully than in the earlier part.

In Part III two special topics are discussed: cartographical apprecia-
tion and the analysis, by measurement on maps, of landform. The
two chapters are intended for those more advanced students who
require to make a comparative study of different map series, or to
carry out an exploratory morphological survey, and who will norm-
ally have access to a large map collection.

Bibliographical References

Selected bibliographical references are provided at the end of each
chapter. They are of two kinds. The first includes a number of the
better-known text-books dealing with the subjects discussed, so that
the reader may revise, amplify, or extend his general geographical
knowledge at need. The second consists of important single papers
dealing with individual topics, including a number of specific points
which are treated in the main text. It is hoped that in this way state-
ments can be substantiated, if necessary, by reference to the original
authority, without recourse to cumbrous footnotes, and that the lists
will serve to locate useful material for further reading. It is not

INTRODUCTORY 5

suggested, or intended, that the student of map interpretation should
read them all. Indeed, the beginner is best advised to defer all such
work until he is well grounded in basic principles. Sooner or later,
however, he should attempt a modicum of detailed reading, in order to
appreciate the direction and scope of current geographical research,
and the nature of the ground which he encounters on the map. This
is particularly important since nothing has been included in the inter-
pretations but what can fairly be obtained from the map, except where
additional notes of the briefest sort have proved essential. Some of the
articles cited are, therefore, to be regarded as possible supplements to,
and extensions of, the present text.

It will be clear from the table of chapter headings that the arrange-
ment of Part I is broadly similar to that of several text-books of
physical geography. This consideration, as much as any other,
governed the selection of the maps treated herein ; but any selection
had necessarily to be a compromise. In addition to illustrating, within
the limits of the sheet lines, some leading classes of physical landscape,
the maps were required to exemplify a variety of settlement, both in
respect of density and of form. It was desired, furthermore, to keep
the list as brief as possible, on the grounds of expense, and from the
O.S. maps to select as far as possible those showing accessible pieces of
country. These various factors operated with different strength in
different cases. The inclusion of Ingleborough, for example a piece
of country more than once included in books on map reading is due
to its convenient occurrence in the centre of a 1/25,000 sheet. Other
things being equal, the precise area chosen is that best known to the
author.

Selection of Maps

In accordance with the view that work on maps and work in
the field are inseparable, the text which follows relates largely to
the maps of the Ordnance Survey on the scales of 1/63, 36O 1 and
1/25,000. The complete list of maps required is given in the table
on page 6.

The total cost of the maps required, with the exception of the last

1 When the first edition of this book was prepared, the then current New Popular
Edition of the 1/63,360 Map was used. This Edition is now superseded by the Seventh
Series, but the sheet lines remain unchanged, and the older series may still be employed.
Sheet-line changes, however, affect the map initially chosen for Scotland (Popular,
Sheet 47).

2 (.5196)

6 MAP INTERPRETATION

two, which arc used here only to illustrate cartographical appreciation,
is thus ^i I2s. 8d.

The prices given for O.S. maps are the current prices less one-third.
O.S. maps required for educational purposes may be obtained direct
from the Director General, Ordnance Survey, Chessington, Surbiton,
Surrey. Discount is allowed at the rate of 33 i per cent for 1/25,000

SERIES

SCALE

NUMBER AND TITLE

PRICE

O.S. (Seventh Series or
New Popular Edition)

1/63,360

114 (Boston and Skegness)
144 (Cheltenham and Evesham)
165 (Weston-super-Mare)
167 (Salisbury)
1 86 (Bodmin and Launceston)

2s. 4d.
per sheet

O.S. Tourist Map

1/63,360

Lorn and Lochaber

3s.

O.S.

1/25,000

NC/76
SY/I8
SO/oo
SD/77
NY/71

2S.

per sheet

Survey of Ireland

1/63,360

169 (Sheets 169, 170, 180, and
1 8 1, printed as a single
sheet)

2S.

U.S. Department of the
Interior: Geological
Survey

1/62,500

Arizona (Final County)
Casa Grande Quadrangle

6s.

Netherlands: Topogra-
phische Dienst (Topo-
graphic Service)

1/25,000

New Series No. 25A (Haarlem)

5*.

Germany: Landesver-
messungsamj; (Land
Survey Office) Nord-
rhein-Westfalen

1/25,000

4506 (Duisburg)

75. 6d.

and smaller-scale maps. Orders should be accompanied by O.S. form,
O.S. 318, countersigned by the appropriate educational authority.
Prices of foreign maps are liable to fluctuate with rates of exchange,
and those given should be taken as approximate. Agents dealing in
foreign maps include Sifton Praed & Co., Ltd., 67 St. James's Street,
London, S.W.I, and Edward Stanford, Ltd., 12-14 Long Acre,
London, W.C.2.

INTRODUCTORY 7

NOTES AND REFERENCES

The following may be recommended for background reading

SIR. C. CLOSE. The Map of England. Peter Davies, London, 1932.
F. DEBENHAM. Map Making. Blackie, London, 1940.
A. R. HINKS. Maps and Survey. University Press, Cambridge, 1942.
H. S. L. WINTERBOTHAM. A Key to Maps. Blackie, London, 1945.

The most abundantly illustrated works on map interpretation are
undoubtedly

R. D. SALISBURY and W. W. ATWOOD. The Interpretation of Topo-
graphical Maps. U.S. Geological Survey, Professional Paper 60,
Washington, 1908.

A. GARNETT. The Geographical Interpretation of Topographical Maps.
Harrap, London, 1935. (Accompanied by an atlas of selected maps.)

In the early years of this century H. R. Mill proposed that geo-
graphical memoirs should be produced, corresponding to the Sheet
Memoirs of the Geological Survey. A great deal of material additional
to that shown on the map would naturally be included, but Mill's
essay is nevertheless well worth study in connection with map inter-
pretation

H. R. MILL. "A Fragment of the Geography of England." Geogr.
Journ., xv, 1900, pp. 205, 353.

PART I

PHYSICAL INTERPRETATION

CHAPTER II
PHYSICAL INTERPRETATION: GENERAL

The form of the land-surface is perhaps the most funda-
mental of all geographical factors. WOOLDRIDGE

A STRICTLY analytical attack on the problem of interpreting the physical
landscape from maps will usually prove the most effective. The method
advocated here resembles qualitative analysis in chemistry, or the use
of mineralogical tests by the field geologist, and might indeed be aptly
described as the qualitative interpretation of landscape. Each operation
is carried out in three stages. First, the interpreter works systematically
over the map, identifying individual features, assemblages of features,
and types of terrain. Next, the kind of erosion-cycle in progress and
the most evident structures are determined. Finally, the map is care-
fully inspected for special features associated with the particular mode
of denudation, but not yet discovered, as well as for evidence of stage
in the current cycle and for signs of sculpture in previous cycles. With
practice, many of the essential facts are perceived at a glance, but
orderly treatment can never be dispensed with, especially if difficult
maps are likely to be encountered.

General Characteristics of the Landscape

W. M. Davis* s classic description of the physical landscape as "a
function of structure, process, and stage*' embodies the fundamental
principle to be observed, and serves as a basis for a scheme of operations.
It is, of course, quite clear that all three elements cannot be fully inter-
preted from every, or perhaps from any, topographical map. Nor is a
stereotyped procedure or treatment to be recommended. In some
areas the main interest will reside in a close correspondence between
structure and surface; elsewhere the chief concern will be landforms
developed in one of the special cycles; elsewhere again, the map will
provide striking evidence of erosional forms belonging to a certain
stage of the cycle, or of a combination of forms developed under the
control of more than one base-level. While, therefore, structure,
process, and stage must always be taken into account, they need not be
discussed in that order, and it is frequently desirable to lay emphasis

12 MAP INTERPRETATION

on one particular aspect. Examples of varying treatment and emphasis
will be found in the descriptive chapters which follow.

In order to provide a convenient scale of reference and a guide to
orderly interpretation of physique, several lines of reasoning are
brought together in the accompanying table. This is arranged, for
convenience, in summary form, not as a set scheme of headings under
which, in the order given, results should be presented. However,
there is no doubt that physical interpretation is most successful when
it is carried out systematically, or that the task is considerably simplified
when, by the use of some such scheme as this, many possibilities can
be at once eliminated. The table should assist the interpreter to identify
the kind of terrain represented, to locate critical features, to perceive
apparent anomalies, and to test his conclusions.

No attempt has been made to catalogue all the typical or critical
features of the various classes of physical landscape, for this would
have made the table unwieldy, in addition to usurping some of the
functions of text-books of geomorphology. On the other hand, the
table is so arranged that the relevant portions of standard texts can be
discovered easily from an index, and reading or revision carried out
as necessary. If a particular class of landscape is dealt with later, a
note is made of the relevant chapter in this book; if not, a reference is
given to a standard work which contains the description required.

It is, perhaps, as well to repeat that the table is designed for reference,
not as something to be assimilated before interpretation is begun. Its
frequent use should lead, in due course, to a considerable economy of
time and effort, and to a rapid perception of essential facts. It is meant
to assist in relating study of physical geography to the use of maps.
Like the rest of this chapter, it is intended to assist in the development
of a satisfactory technique.

I. PROCESS
Class of Cycle

What form of the erosion cycle is now in progress ?

(i) The normal cycle; or Systems of perennial streams observed,
(ii) One of the special cycles

(a) The glacial cycle Glaciers actually present.

(b) The arid cycle Intermittent or absent surface drainage,

associated with desert features. (See
Chapter IX and the texts there cited.)

(c) The karstic cycle; and, if Absence of surface drainage and
so, are full karstic features de- presence of sinks in a region of humid
veloped, or only the modified climate. (See Chapter VII and Chap-
forms typical of Chalk country? ter IV on Chalk country.)

PHYSICAL INTERPRETATION: GENERAL

If the normal cycle, does the land bear marks of

Corries, glacial troughs, etc. (See

Chapter VI.)

Drumlins or large terminal moraines.

(See Chapter VI.)

Raised abrasion-platforms. (See

Chapter V.)

(i) Former glacial erosion?
(ii) Former glacial deposition?
(iii) Former marine erosion ?

If any shoreline occurs, what is
(i) Its class and type?

(ii) The stage reached in the present
shoreline cycle?

Erosional and Depositional Features

These should be selected and specified, as required, to illustrate the general
argument, and to provide evidence of structure and stage.

I.e. whether submergent or emergent,
highland or lowland, etc. (See
Chapter VIII.)
(See Chapter VIII.)

II. STRUCTURE
Lithology

(i) Do the rocks as a whole seem
resistant or weak?

(ii) What evidence is there of differ-
ential rock resistance?

(iii) Assuming a humid climate, is
there any sign of permeable rock ?
If so, are there definite lines of
sinks?

Geological Structure

(i) Is there a definite alternation of
weak and strong outcrops? If
so, the rocks are likely to be
part of a sedimentary succession,
either

(a) Horizontal; or

(b) Uniclinal; or

(d) Openly folded; if so, note
arrangement of folds, for exam-
ple, if off-set or pitching.

(ii) Is the area one of complex struc-
ture, for example, strongly
folded or faulted?
If so, what structural grains can
be identified ?

and

AND LITHOLOGY

Strong relief corresponds generally to
resistant rocks, but feeble relief is not
confined to weak rocks.
In sedimentary rocks, the resistant
formations are likely to be limestones
and sandstones; the weak, clays.
Local absence of surface drainage.
Limestone with karstic features.
(See Chapters III, IV, and VII.)

Structural plateau.

Eroded into scarpland country. (See

Chapter III.)

When denuded, domes are enclosed

by in-facing scarps.

Eroded anticlines show in-facing

scarps; eroded synclines, out-facing

scarps. (See Chapter IV.)

For example, an oldland. See Chap-
ters V and VI.)

MAP INTERPRETATION

iii) What, if any, are the signs of
faulting?

(a) Full development of fault-
block topography;

(b) Offsetting of scarps;

and

(iv) Where the structure is generally
simple, for example, uniclinal,
what minor structures, if any,
can be identified ?

and

(v) What igneous masses occur, if
any?

and

(vi) If volcanic topography is devel-
oped, where are the vents, what
lava flows can be identified, and
how advanced is the dissection
of the cones?

(For a full discussion, refer to the text-
books cited below.)

(See Chapter VI.)

For example, transverse folds or faults
in a scarpland. (See Chapter III.)

For example, granitic bosses. (See
Chapter V.)

(For a discussion of volcanoes, which

are not dealt with in the present work,

see the texts cited below.

The fullest treatment is given in

Volcanoes as Landscape Forms by

COTTON.)

in. STAGE

Landscape

(i) If a special cycle is in progress,
what is the stage reached?

(ii) If the normal cycle, is the land-
scape

(a) Unicyclic? or

(b) Polycyclic?

and

(iii) What is the approximate stage
reached in

(a) The present cycle ?

(b) Any previous cycle whose
landforms can be identified?

Drainage

(i) Assuming the normal cycle, as
far as can be judged from the
map, is the drainage mature or
not?
and

(For descriptions of the cycle in karstic,
glacial, and arid conditions, see the
works cited below.)

I.e. still evolving in the first cycle
initiated by emergence.
I.e. comprising features developed in
more than one cycle ; for example, an
up-lifted and partly dissected pene-
plain, or an approximately mature
landscape with the lower parts of
valleys rejuvenated.

(See text immediately following.)

I.e. are the long-profiles of rivers
smooth throughout?

PHYSICAL INTERPRETATION: GENERAL 15

(ii) If breaks of slope occur in long-
profiles, do they appear

(a) Of cyclic origin ? or I.e. knickpoints. (See text immedi-

ately following, and Chapter XVI.)

(b) Non-cyclic? I.e. determined by structure alone,
and

(iii) What is the relation of drainage
to structure?

(a) Well adjusted;

(b) Partially adjusted;

Interpretation in Terms of the Erosion-cycle

A few supplementary comments are needed on the subject of stage.
There are good reasons for separating, under this head, landscape from
drainage. Rivers may easily attain maturity while the landscape is
still youthful, that is to say, there may be a phase-difference in the
respective cycles of landscape and drainage evolution, at a time when
relief is generally strong. Again, although many maps readily show
the stage reached by the evolving landscape, this fact is commonly
overlooked in map interpretation where undue emphasis may be
placed on drainage. Finally, it is impossible to tell, by mere inspection
of a map, whether the rivers shown are mature or not.

The landscape becomes mature in the normal cycle when all initial
forms are consumed, that is, when opposing valley walls meet along the
crests of divides. Although not all members of a divide system attain
maturity at the same time, it is usually possible to distinguish broadly
between those landscapes which do, and those which do not, retain in
the crest regions remnants of an initial surface or of an erosional
platform produced in an earlier cycle. When low divides are separated
by broad, flat-bottomed valleys, the landscape is indubitably post-
mature. In practice, various complications arise from differences in
rock strength. Furthermore, the problem of interpreting a normal
landscape from maps often resolves itself into one of understanding a
polycyclic landscape, which combines features produced in more than
one erosion-cycle, with the possible addition of the forms of earlier
shorelines; but, whatever the complexities, landscape is often more
easily interpreted than drainage, for contours show die qualities of
relief far better than those of river profiles. Where drainage is incised,
it is the landscape that has not yet become mature after rejuvenation.

16 MAP INTERPRETATION

A river is mature when it is graded, that is, when its long-profile is a
smooth curve decreasing in gradient from source to mouth. 1 Now a
graded state can be satisfactorily demonstrated only by accurate
levelling in the field. Very many rivers are in fact ungraded, either
because their profiles have never been smoothed into a continuous
curve, or more commonly because the profiles have been developed
in more than one cycle. When a river is rejuvenated, by a general fall
of base-level or by deformation of part of the earth's crust, the new
profile developing headwards from the new base-level intersects the
old profile in a kriickpoint. Unfortunately many irregularities of
profile, whether cyclic or not, do not appear on long-profiles
constructed from contoured maps. Hence the ungraded state of some
rivers can be proved only by levelling. All that the map interpreter
can do is to discover which rivers, according to his map, are certainly
ungraded in the present cycle, by locating marked irregularities of
profile.

Note that the form of the profile is the only criterion of a graded
state; the presence or absence of meanders has no significance here.

Braided streams can be graded. Again, rivers can begin to meander
well before maturity is reached in the first cycle, and, once meanders
are incised, their trace can obviously be preserved throughout a series
of successive rejuvenations.

Physiographic Subdivision

When the processes at work, the underlying geological structures,
and the stage reached in the current and earlier cycles of erosion have
been decided as fully as the map permits, the work of description may
begin. It is.usually necessary to subdivide the area represented into
parts which, each with its unifying characteristics, may be discussed in
turn, and to illustrate the subdivision by a sketch-map. The method
of subdividing must be carefully thought out. When, as in most map
interpretation, only small areas are in question, it must nearly always
depend on physical differences. Field studies in geomorphology
have revealed the basic principle, which is that a great part of the earth's
surface may be appropriately described in terms of flats and slopes. A
geomorphologist regards the physical landscape as composed of a
number of facets, each a slope or a flat, some depositional, others

1 Recent work indicates that a smooth long-profile may never be attained, and that
maturity may be indefinable in terms of slope. Obviously irregular reaches, however,
remain significant.

PHYSICAL INTERPRETATION: GENERAL 17

reflecting the structure of die solid rocks, and yet others entirely
erosional in origin. Once this principle is grasped, a contoured map
can be used in drawing boundaries between small contiguous parcels
of land, differing from one another in form, slope, aspect, geology,
soil quality, and frequently also in density and form of occupance and
in surface utilization.

The above statements merely express what has long been practised
in map interpretation, and record its sound foundation in morpho-
logical theory. As Linton 1 has so truly said, "a predilection for
morphological subdivisions may be considered as a characteristic of
British geographical method" ; but, rather surprisingly, no standard
nomenclature has arisen, comprising terms useful in identifying and
describing the units of country with which map interpretation deals.
The word "region" will not serve; "area" is too weak. In this book
the nomenclature adopted is that suggested by Linton (op. cit.), as
follows

Site. The unit of the smallest order; the geographical equivalent
of a single morphological feature, such as an individual hillside or a
patch of river terrace.

Stow. An assemblage of sites, possessing a geographical unity ;
cf. the examples suggested by Unstead, the valley stows and the
plateau stows of the North Downs.

Tract. A group of stows, again with a geographical unity, but
with a higher order of complexity ; for example the North Downs
as a whole.

Section. A unit of higher order, which derives its unity in part
from climate, vegetation, and land use; for example the highland
of N.W. France (Armorica).

Province. A group of sections with a fundamental geographical
resemblance; for example the Oceanic Uplands Province of
Linton, comprising S.W. Ireland, S.W. England and the coast of
S. Wales, and the N.W. Highlands of France.

Continental Subdivision. The major unit in the subdivision of
a continent; for example the Atlantic Highlands, comprising the
extreme west and north of Ireland, Scotland beyond the Lowlands,
and most of Scandinavia.

The three highest orders are illustrated in Linton (op. cit.).

1 D. L. LINTON. "The Delimitation of Morphological Regions." Published in London
Essays in Geography^ edited by L. Dudley Stamp and S. W. Wooldridge. Longmans,
Green, London, 1951, p. 199.

18 MAP INTERPRETATION

Needless to say, interpretation on medium scales will be concerned
chiefly with sites, stows, and tracts or parts of tracts.

These terms have a self-apparent value in providing much-needed
names for use in detailed physical description, and in bringing some
precision into descriptive work. When they are correctly employed,
there need be no doubt at any time as to the class of unit under dis-
cussion, or as to the order of geographical complexity which it is
likely to attain. One might discuss at considerable length the use of
physical criteria in' establishing major as well as minor geographical
boundaries, but Linton's scheme has the indisputable advantage of
coherence and of working upwards from the fundamental unit of
country, the individual feature. The physical basis of unity in a partic-
ular piece of country has been finely expressed by the same writer in
these words

"If in any area the physiographical conditions exist for the produc-
tion of a particular slope form they are usually rather widespread and
examples of that form are likely to be repeated fairly commonly over
the area." (Ibid. p. 209.)

Sketch-mapping

In drawing sketch-maps one should aim at representing the principal
steep slopes, such as scarp-faces, the walls of wide valleys, or the edges
of mountain blocks. When a sketch-map is prepared in this way for
one of the O.S. 1/63,360 sheets, the subdivision into stows will have
been at least partly effected, although further boundaries may be
necessary, for example at the edge of a fen which is bounded by low
but better-drained country. Steep slopes are physical features and
geographical sites ; the boundary between stows is usually to be drawn
at the foot of the slope.

Physical interpretation, however, demands more than boundaries.
Many significant features must be appropriately symbolized on a
sketch-map. If contoured maps were already adequate, other devices
for showing landform would be superfluous; but, since contours,
however close and accurate, are descriptive rather than interpretative,
there is need for a scale of symbols suitable for use in sketch-mapping,
such as that given here. (See Fig. i.) These symbols are already widely
used and widely understood. Those for steep slopes are all variants of
hachuring, which can be adapted at will for representing scarp-faces
developed on a variety of geological formations, the abrupt edges of

scarp
steep
gene

mean<
slip-o
incisec

incis(
mear
valle

Iwate
Jknick
direct
anticl
syncl
fault

1
e

selected flats,
e.g. terraces,
bevelled summits
glacier

glacial trough
corrie
drumlin
end- moraine
alluvial fan
delta

scree

exposed r
r rock surface

springs
sinks
cliffs

. . '. ....

>.s and

slopes
rally

der scar 1
ff slope [
J valley

**?

~o~

(T\

iLn

*%$

-U

idermg

rfall or
point
ion of dip
nal axis
nal axis

' outcrops

%>,

"SK

''-.*

dunes

' . * . '

shingle

.""

beach bar

*//*

sand , foreshore

<;?:;

lava flow

<u
* c

^=-=

^ T !

* 4- *
*'*'

FIG. i. A. SCALE OP SYMBOLS SUITABLE FOR USB ON PHYSICAL
SKBTCH-MAPS

2O MAP INTERPRETATION

uplands, the walls of glacial troughs, and the sides of incised valleys.
Where flats have to be emphasized selectively they may be blocked-in
or stippled. Certain kinds of terrain demand special attention to
special features, such as drumlins or dunes. When a symbol which
can be rapidly drawn is already widely used on published maps, it
should be preferred, but in other cases a new symbol may have to be
devised.

Several of the illustrations in this book are sketch-maps show-
ing physique, or. of physical analysis, using symbols from the
accompanying scale, while Fig. 6 shows a complete morphological
subdivision into stows. It should be noted that each sketch-map
depends for its effect on the selection and simplification of the material
contained in the original topographical map, and that the facts to be
shown can be adequately represented in black only. It may not be
out of place here to counsel those who are beginning the study of map
interpretation that in sketch-mapping of this kind the clear austerity
of monochrome is usually preferable to the confused gaiety of colour.

NOTES AND REFERENCES

Each of the following works deals generally with the evolution of
landscape, both in the normal and the special cycles. Since the mode of
treatment varies somewhat from one to another, it is desirable that
more than one should be closely studied.

C. Ar COTTON. Geomorphology. Third Edition. Whitcombe &
Tombs, Christchurch, New Zealand, 1942.

G. H. DURY. The Face of the Earth. Penguin Books, London, 1959.

O. D. VON ENGELN. Geomorphology. Macmillan, New York, 1949.

L. C. K*ING. South African Scenery. Oliver and Boyd, Edinburgh,
1951.

A. K. LOBECK. Geomorphology. McGraw-Hill, New York, 1939.

W. D. THORNBURY. Principles of Geomorphology. Wiley, New
York, 1954.

No student of physical geography can neglect the work of W. M.
DAVIS, in particular the papers reprinted in Part II of Geographical
Essays (Ginn, Boston (Mass.), 1909). Of these, the following should be
consulted at an early stage

"The Geographical Cycle." Reprinted from Geogr. journal, xiv,
1899, p. 481.

"Base-level, Grade, and Peneplain." Reprinted from Journal oj
Geology, x, 1902, p. 77.

PHYSICAL INTERPRETATION: GENERAL 21

Much of the Davisian thesis is now being strongly challenged. For
an important summary of opposing views, see

L. C. KING. "Canons of Landscape Evolution," Bull. geol. Soc.
Amcr., 64, 1953, p. 721.

The question of geographical subdivision is discussed in

D. L. LINTON. "The Delimitation of Morphological Regions."
Published in London Essays in Geography, edited by L. Dudley Stamp
and S. W. Wooldridge. Longmans, Green, London, 1951, p. 199.

J. F. UNSTEAD. "A System of Regional Geography." Published in
Geography, xviii, 1933, p. 175.

Specimens of morphological mapping may be found in

COMITE NATIONALS DE GEOGRAPHIE. Atlas de Prance. Planches 8,
8 A, 9, 9A: Morphologic. (Scale: 1/1,000,000.)

D. L. LINTON. Watershed Breaching by Ice in Scotland. Institute of
British Geographers, Publication No. 15. George Philip, London,
1951, p. 8. (Sec especially Figs. 4 and 5.)

S. W. WOOLDRIDGE and D. L. LINTON. Structure, Surface and
Drainage in South-east England. George Philip, London, 1955.

Techniques of morphological mapping in the field are discussed in

R. S. WATERS. "Morphological Mapping," Geography, xliii, 1958,
p. 10.

3-(E. 5 i 9 6)

CHAPTER m

SCARPLAND TOPOGRAPHY

A most patient and thorough examination of the structure is
made by the destructive forces. W. M. DAVIS

MAP: CXS. 1/63,360 (SEVENTH SERIES) SHEET 144
(CHELTENHAM AND EVESHAM)

ON this part of the English Plain scarpland topography is boldly
developed. There can be no mistaking the impressive scarp-face,
rising as much as 800 ft. above the sub-edge country to the north-
west. South-eastwards there is a general, steady descent, readily made
out although the country is considerably dissected. In broad outline,
then, the structure appears simple : the Avon valley has been excavated
in weak, impermeable rocks, while the Cotswolds are based on more
resistant strata, permeable at least in part and dipping gently towards
the south-east. It will be seen in due course that the foregoing state-
ment must be qualified in some important respects, but for the time
being it is well to accept it and to consider, in as much detail as the
map permits, the lithology and geological structure which so markedly

influence relief.
*

The Scarp-formers

Although the dry valleys of the back-slope indicate permeable rock,
and althoifgh the name "down" is given to many hills, 1 this is not
Chalk country. "Chalk Hill" (i326) 2 no more proves an outcrop of
Chalk than "Upper Slatepits" (1032) indicates the presence of meta-
morphosed shales. The texture of relief can be appealed to for
guidance. The short re-entrant valleys or combes in the Cotswold
scarp-face are more sharply and deeply cut than the corresponding
scalloped recesses in Chalk, while the contours of the back-slope are
less smooth than those, for example, of the comparable part of the
Chilterns. (See also the following chapter.) The Cotswold terrain,

1 The use of "down" in hill names is adduced, with surprising frequency, as evidence
of Chalk.
1 Throughout this book, grid references apply only to the kilometre squares.

SCARPLAND TOPOGRAPHY 23

with narrow, steep-sided valleys and tabular interfluves, suggests a
dissected plateau much more strongly than do the swelling undulations
of much Chalk country. Discontinuous drainage occurs much more
widely here than on the Chalk: a number of valleys are dry only in
certain reaches, as, for instance, that of the Dikler between 147303 and
167279, a characteristic which suggests that the rocks of the cuesta are
not permeable throughout, at least to a uniform degree. Some sand-
stones are highly permeable in mass, but rarely develop the localized
sinks which cause a stream to disappear; hence it may be concluded
that the rocks of the Cotswolds include limestones, which must be
mechanically stronger than Chalk in order to support the steep valley
walls.

There is some indication in the forms of the scarp-face that the
scarp-forming rocks are not homogeneous throughout. The steep
descent is broken at several places by benches, most clearly seen
perhaps on Oxenton Hill (963 1) and in Burhill east of Buckland (0836).
The main summit of the Oxenton Hill group, rising to 734 ft., is
presumably capped by resistant rock; but a different formation at a
lower level would seem to cap the three spurs at 500-600 ft., Dixton
Hill (9830), Crane Hill (9630), and the broad shoulder south of Ted-
dington (9632). Similarly at Burhill there is a sharp descent of some
250 feet from the brow of the main scarp to a distinctive flat, which
slopes gently to the south in rough accordance with the assumed
regional dip. It is, of course, natural to expect that lithological changes
should occur in a thickness of some 500 feet of sediments.

The Clay Vale

The strong formations are underlaid by weak rocks, which have
been deeply eroded into a strike vale. As stated previously, the weak
rocks may safely be regarded as of the clay family ; but the lowland
tract is not unvaried. The sub-edge country between the scarp and
the River Avon is divided into two stows, the Vales of Gloucester and
Evesham, by a group of detached hills. The flanks of these hills are
so obviously similar in general form and detail to the main scarp-face
that they must be geologically similar ; the resemblance is emphasized
by the distribution of woodland, which is very like that of the main
scarp. It follows that the hills are outliers, orographical as well as
geological, of the cuesta, and that they have survived for some reason
while the scarp generally has receded. At first sight there is nothing to
explain why they should not also have been destroyed as the strike

24 MAP INTERPRETATION

vale was opened out, but a useful clue may be obtained when a further
principle is applied.

Structures

The solution lies in minor structures, which may conveniently be
discussed as part of the question of cuesta structure as a whole. It is a
mistake to suppose that, in the English scarplands at least, the back-
slopes are usually structural surfaces: to this extent the customary
simple diagram illustrating elementary accounts of scarpland country
is misleading. As a rule, the dip of the strata is greater than the general
slope of the ground, so that progressively younger rocks come in
towards the foot of the back-slope. Again, in many cuestas minor
folds and faults occur, which may or may not be reflected in the form
of the ground. Once it is accepted that in actuality a cuesta is likely to
possess some internal structural variety, the interpreter is naturally on
the look-out for signs. The scarp-face is most likely to repay close
study, for its outline may be influenced by transverse folds or faults. In
an anticline, for example, the weak underlying rocks will be laid open
to attack and the scarp-face may be deeply indented; a syncline which
brings down the scarp-former may appear as a projection of the high
ground into a strike vale.

The structural pattern of the Cotswold cuesta is somewhat com-
plicated, as intermittent folding took place over a long period,
including the time when the sediments were being laid down. Little
of the total effect can be made out from Sheet 144, but it is at least
possible to see the deep re-entrant valley of the Isbourne above
Winchcomlj (0228) and that of the Badsey Brook south of Broadway
(0937). If these are anticlines, an ill-marked syncline appears to lie
between them, with a much deeper downfold west of Winchcomb
passing through Cleeve Hill (9826), Nottingham Hill (9828), and
Oxenton Hill. Bredon Hill is actually down-faulted, but this cannot
be seen from the map. One can perceive, however, that the cap-rock
here appears to stand lower than on the main crest of the south : the
profile of the Cotswold back-slope, if projected towards the north-
west, would pass above the summit of Bredon Hill.

With a little care, the interpreter may further make out signs of
earth-movement on a more considerable scale. The map evidence
available on this sheet is none too plentiful, but is nevertheless perfectly
sound. North of the Avon, a little west of the meridian of Evesham,
is a stow of low hill country in which stands Church Lench (0251).

SCARPLAND TOPOGRAPHY 25

The 2 soft, contour and the patches of woodland draw attention to the
steep outward-facing edges, which are appropriate to a denuded
syncline. The problem of linking this syncline to the one passing
through Cleeve Hill is less important, in the present connection, than
the remarkable similarity between these hills and the small cuesta
between Eatington (2649) and Loxley (2553). The resemblance
extends to the distribution of woodland as well as to height, but lies
fundamentally in the details of the scarp-slopes and in the amplitude
of curvature of the contours. It looks as if the two stows might be
based on a single formation. If so, a strong fold or fault must lie
between; for the strong rocks of the Lcnch Hills occur in the clays of
the strike vale and must dip under the scarp-formers of the Cotswolds,
but the small cuesta in the north-east is in line with the Cotswold
scarp-face. In other words, the north-eastern cuesta is considerably
offset down the dip. Such a distribution would be expected if a fault
or fold, near the line of the Stour valley, threw down the rocks on the
south-western side. In that event one would expect the main scarp
also to be offset. In reality certain complications occur, but this map
suffices to suggest that Chastleton Hill (2628) may correspond to
Ilmington Down (1842) and that displacement of the kind envisaged
has taken place. These suggestions are summarized in Fig. 2.

Drainage

The value of interpreting structure before drainage will now be
realized. Unless structures can be defined independently, there is no
way of telling whether or not the drainage is adjusted. It is an ele-
mentary error in logic to assume implicitly that adjustment has taken
place, and to use this assumption to define structures. On this view,
most of the trunk streams of a scarpland will seem to have become
adjusted to structure, whereas in fact many, in the English Plain, are
far from being so.

The Warwickshire Avon is often cited as a strike-stream. From the
atlas map there seems little doubt that it is; but it has already been
noted that transverse structures appear to run across the river, notably
in the Lench hills, and there is no alternative but to regard it as super-
imposed, at least in part. The topographical map provides no clue to
the truth, that it has cut into the "solid" rocks through a thick covering
of Pleistocene material. Nor is there any sign of the many fans of
sludge deposits, which in periglacial conditions spread over the sub-
edge plain from the numerous combes.

26 MAP INTERPRETATION

Terraces of the Avon

Despite these serious limitations, the selected map provides a con-
siderable amount of useful information concerning the later evolution
of the Avon system. As always, the interpreter should be on the watch
for signs of re-grading in progress. Since the valley bottom is 100
feet or so below the surrounding countryside, it may be assumed that
rejuvenation has taken place. The rocks of the Vale are mostly weak,
but not too weak to preserve sufficient detail, at a number of sites,
to show the manner and the effect of the renewed downcutting.
Between Bidford on Avon (1051) and Offenham (0546) meanders
appear to have swept freely down-valley, trimming back meander-spurs
into straight lines of bluffs, clearing any valley fill, and developing a
continuous ribbon of flood-plain. In contrast, the run of the 5O-ft.
contour in the lobes of Pensham and Birlingham (9444, 9343) suggests
slip-off slopes of gentle gradient, which are opposed by undercut
slopes at least fifty feet high across the river. Here are certainly
ingrown meanders, which have enlarged themselves laterally during
incision without migrating downstream to any marked extent. The
lobes of ingrown meanders deserve very close inspection for signs of
terraces, which are often best preserved at such sites. Although a
standard topographical map will never reveal terraces with the clarity
of a geological sheet or a special morphological map, something may
perhaps^be read from the spot heights and the run of drainage channels.
Thus, here, spot heights of 81, 79, and 69 ft. are given on the lobe
immediately east of Pershore (9646), whereas the flood-plain alluvium
stands at ogly 50 ft. just below Pershore Bridge, i miles downstream.
Between the points at 79 and 69 ft. there is a fall of only 10 ft. in | mile;
at this rate the river would be reached at 60-65 ft- If these points are
on the flood-plain, there must be a pronounced break of slope a little
above the bridge. Further search, however, discloses that a patch of
terrace is present, for a height of only 61 ft. is marked on the crown of
a road, higher up the valley at 999457. Again, the tips of some lobes
are cut off by small channels, as near Birlingham, and below the spot
height 61 where Lench and Oxton Ditches are named. It is justifiable
to suppose that the crescent between these channels and the main
stream is part of the flood-plain, whereas on the landward side of the
channels the ground stands higher in other words, that the channels
define the edge of the alluvium, acting as drainage ditches for the
flood-plain and as catchwaters for ditches or streams flowing towards

SCARPLAND TOPOGRAPHY

the river. Confirmatory evidence is often forthcoming. The field
road through Broadway stops short at Oxton Ditch (cf. the similar

^
V

mom scorp
sub- scorp
minor scorp

'*$*

Avon flood plain and low terrace
incised meonderinq volleys
former head of Evenlode
dry volleys ( ~~

^Inferred anticlines
inferred synclines
.-^. inferred line of downward
displocement to west

FIG. 2. INTERPRETATION OF PHYSIQUE: THE COTSWOLDS AND

THE AVON VALLEY
(Based, by permission, on Sheet 144 of the O.S. New Popular 1/63,360 Map)

termination of a similar road at 948428). Evesham stands on the lobe of
a notably asymmetric and ingrown meander. On the southern side
of the town, building is roughly confined within the roo-ft. contour,
which curves smoothly round within the loop of the river. On the

28 MAP INTERPRETATION

south-western, downstream, side of the lobe the river appears to have
shifted a little down-valley, leaving a crescentic patch of alluvium
with poor natural drainage. The Evesham site seems to have acted
as a node, through which meanders have been unable to sweep freely
down the valley, for all the evidence goes to show that here and in
the downstream loops the alluvium is confined to the tips of the lobes.
Above the flood-plain, terrace patches indicate that downcutting has
been intermittent.

The Stour and the Evenlode

In the Stour system relief is more varied, and the valley less widely
opened than that of the Avon. It has already been inferred that the
Stour is probably cutting back along a structural line; but, although
downcutting has been vigorous, the meanders have not apparently
been able to sweep out an alluvial trough. Note the sharpened spurs
near Halford (2645) and below Shipston (2641), which are being
attacked on their upstream sides but are as yet little eroded.

The greatest contrast is that between the valleys of the Stour and
of the Evenlode, the one with numerous small hills 150-200 ft. higher
than their surroundings, the other with a wide, flat floor, which ends
abruptly in a steep edge defined by the 35O-ft. and 400-6. contours.
This is an erosional scarp. The Stour is pressing back the divide,
causing it to creep south-eastwards; it has also leapt, for the Knee
Brook head of the Stour has beheaded the Evenlode which once
flowed through the col at Campden Tunnel (Fig. 2). The topo-
graphical map can do no more than hint at the former course of
events, proof being obtained from work in the field.

Underfit Streams

The action of capture must have reduced the volume of the Even-
lode, but capture is not the explanation of underfit. The interpreter
notes that in the trace of the river below Adlestrop Station (2526)
meanders of small radius are superimposed on windings of much
greater size. The small meanders are described by the low-water
channel of the river, which appears too small for its valley and is
therefore described as a misfit, or underfit, stream. The implication
is that discharge has in some way been greatly diminished. Many
writers, following the early work of W. M. Davis, regard underfit as
the result of river capture, since beheaded streams are often found in an
underfit condition; but Sheet 144 shows that the capturing Stour, as

SCARPLAND TOPOGRAPHY 2p

well as the beheaded Evenlode, has meanders of the two orders of
size. Each river meanders within a meandering valley, and, in addition,
the Avon itself, although much regularized for navigation, displays
similar features, for example, in the Birlingham loop. The Coin,
which may well have been beheaded by the Isbourne, is no more a
misfit than the Avon, which should have been increased in volume by
any capture effected. Two useful conclusions may be drawn. Firstly,
capture is inadequate to explain underfit, which must be accounted
for in some more general fashion. Secondly, it is important to
distinguish, in description, between the meanders of the river and
those of the valley. It is the lobes of valley-meanders which provide
the interlocking spurs so frequently mentioned in introductory
studies.

Since underfit streams on the back-slope are not reliable indicators
of capture, the contest of streams along the crestal divide must be
interpreted with much care. In a tract such as this, where the sedi-
mentary scarp-formers are (as seen above) somewhat varied, and where
transverse structures have been identified, a notch in the crest may not
invariably be the result of capture. It is quite possible that two
streams, one on the scarp-face and the other on the back-slope, have
worked headwards along a single line of weakness, approaching the
same point on the watershed and forming a col. Consequently the
shallow dry gaps, as at Lyne's Barn (0627), may be noted without
explanation. There is no way of knowing, from the topographical
map alone, which of these are inherited from beheaded streams.
Capture appears certain only where, as between the Coin on the one
side and the Isbourne and Chelt on the other, the dry gaps are unusually
deep.

If in the foregoing paragraphs the limitations of the topographical
map seem to have been stressed, the effect is deliberate. It is useless to
approach the task of interpreting maps of real country in the hope that
structures, landscapes, and the results of human occupance will every-
where be simple. If they were, geographical study would fail to
stimulate and many cartographical problems would disappear. There
is no point in seeking complexity for its own sake, but the interpreter
who is prepared, on general grounds, to find the actuality varied will be
best fitted to discover most from the map. It seems particularly appro-
priate that this initial study of scarpland country, which at first glance
appears so uncomplicated and clear-cut, should provide the necessary
caution against an over-simple treatment.

30 MAP INTERPRETATION

NOTES AND REFERENCES

For a general account of the geology, including structure, see

G. A. KELLAWAY and F. B. A. WELCH. Bristol and Gloucester District.
Second Edition. British Regional Geology Scries, H.M.S.O., 1948.
(See especially Figs. 3 and 20 for maps of structure.)

F. H. EDMUNDS and K. P. OAKLEY. The Central England District.
Second Edition. British Regional Geology Series, H.M.S.O., 1947.

A summary of the orthodox views on the evolution of scarpland
drainage is

W. M. DAVIS. "The Drainage of Cuestas." Proc. Geol. Assoc., xvi,
1899-1900, p. 75. As stated in the text, this concept is not wholly
applicable to the Cotswolds and the Avon valley.

The capture of part of the Evenlode by the Stour is described in
W. J. ARKELL. The Geology of Oxford. Clarendon Press, Oxford,

1947.

W. J. ARKELL. "The Geology of the Evenlode Gorge, Oxfordshire."
Proc. Geol. Assoc., Iviii, 1947, p. 87.

M. E. TOMLINSON. "The Drifts of the Stour-Evenlode Watershed,
etc." Proceedings of the Birmingham Natural History and Philosophical
Society, xv, Part viii, 1929, p. 157.

Descriptions of the morphology and drift deposits of the sub-edge
plain occur in

W.J. ARKELL. The Geology of Oxford. (Above.)

W. W. BISHOP. "The Pleistocene Geology and Geomorphology
of Three Gaps, etc.," Phil. Trans. Roy. Soc., Ser. B, No. 682, cxli,
1958, p. 255.

G. H. DURY. "A 400-ft. Bench in South-eastern Warwickshire."
Proc. Geol. Assoc., Ixii, 1951, p. 167.

F. W. SHOTTON. "The Pleistocene Deposits of ... Coventry,
Rugby, and Leamington, etc.," Phil. Trans. Roy. Soc., Ser. B, No. 646,
cxxxvii, 1953, p- 209-

CHAPTER IV
ERODED FOLDS

. . . the shapely figured aspect of Chalk hills. GILBERT WHITE

MAP: O.S. 1/63,360 (SEVENTH SERIES) SHEET 167
(SALISBURY)

THE most cursory examination of this map sheet reveals that most of
the ground represented is based on Chalk. The easily recognized
assemblage of cartographic symbols and contour patterns stands for a
highly typified association of actual features, those of high open
downland a kind of landscape not found throughout the Chalk
outcrops but very widespread there. Somewhat paradoxically, a
map may allow no possible doubt that Chalk is present, while pro-
viding weak or unconvincing items of specific evidence. The best
guide to diagnosis is the texture of the country as a whole.

General Aspect of Chalk Country

The most striking attribute of much Chalk terrain is its lack of
surface drainage, which in itself proves no more than that die under-
lying rock is permeable. As in some other limestone tracts, valleys
occur although streams are lacking. On the Chalk these dry valleys
are found to be distinctively arranged in elaborate branching systems,
similar in plan to surface drainage and, with very few exceptions,
sloping continuously downwards to their mouths. Dry valleys
elsewhere, for example, on the Carboniferous Limestone, are much
less regular in form, particularly in long-profile. The smoothly
undulant appearance of many Chalk tracts, and the flowing contours
by which they are represented on the map are due to a characteristic
combination of the dry valleys the bottoms with the intervening
swelling hills. The dry valley systems are generally regarded as having
been cut by surface streams in conditions which no longer obtain.
The matter cannot be argued at length here, but one may note in
passing that the water table has evidently fallen so that percolation has
largely replaced run-off. As with other permeable outcrops, springs
are to be expected at the boundary of underlying impermeable strata,

32 MAP INTERPRETATION

or in the deeper valleys which have been cut down to the water-table.
Streams which head within the Chalk outcrop usually rise well below
the upper ends of valleys (cf. on Sheet 167 the River Till, which is
shown as rising at Orcheston St. George (0645), over five miles from
the watershed and more than three miles below the village of Tilshead).
In reality, the sources of streams of this kind are liable to shift up or
down the valley, in response to rises and falls of the water-table. An
unusual rise may cause temporary streams, called bournes or lavants,
to flow in valleys which are normally dry: in the wet spring of 1950,
which followed a wet autumn and winter, many such streams appeared
on Salisbury Plain. A "bourne" element may be looked for in place-
names, for example, the three Winterbournes (1634, 1635, and 1835)
which all stand on the River Bourne.

In some localities the Chalk outcrop is dimpled by sinks, which
rarely appear, however, on the standard map. Few streams are
swallowed, so that discontinuous drainage, like that of some other
limestone tracts, is not to be expected. The combination of many dry
valleys with few large localized sinks, which results from the wholesale
opening of joint-planes, is a means whereby Chalk can be distinguished
on the map from other permeable formations.

Some scarp-faces on the Chalk are remarkably straight, but others
are scalloped by short, rounded combes (cf. on Sheet 167 the form of
the scarp from Chirton Bottom (0655) eastwards). Combes of this
shape appear to have been eroded in the special conditions of a peri-
glacial environment and should not, therefore, be looked on as
essential features of denuded chalklands, but where they occur they
are useful pointers to the nature of the rock.

The evidence so far reviewed suffices, in the main, to prove perme-
able strata. The special characteristics the smooth outlines, the dry
valley systems, and the general permeability are those which suggest
Chalk, and which, moreover, are all to be discovered from the map
of contours and water only. Other evidence, for which one must
refer to the full topographic sheet, is almost entirely supplementary in
character and secondary in importance. Chalk quarries may be named,
as, for example, in "Chalkpit Hill" (2249), but a single example is not
always reliable. Similarly the names Broad Chalke (0325) and Bower
Chalke (0223) are helpful but not in themselves conclusive. Signs of
prehistoric occupance indicate no more than an easily cleared tract of
light, shallow soils. Similarly, the little that can be read of present
land use provides suggestions rather than facts, and is very far from

ERODED fOLDS 33

giving a definite indication of rock type. The interpreter, then,
should rely on the forms of the ground and on the texture of the relief
to show where Chalk occurs, treating any additional material as
tending to confirm inferences already drawn.

Limits of the Chalk Outcrop

It is nevertheless true that certain distributions, for example, that of
woodland, may be of use in interpreting the approximate boundaries
of the Chalk. One may expect the outer boundary to be associated
with a scarp-face, as generally throughout this tract, in which event it
is easily fixed: the belts of woodland on rocks beneath the Chalk, as
for instance in the upper Nadder valley in the south-west, serve to
emphasize (not to demonstrate) the grain of the relief; but, where the
Chalk is in part overlain by Tertiary rocks or by Clay-with-flints, the
soils are likely to be very different from those developed on the Chalk,
and the difference is usually, in fact, reflected to some extent in the
occurrence of woodland or heath. The possibility that such deposits
may occur deserves to be strongly emphasized, for the chalklands
cannot be understood if it is ignored.

On Sheet 167 the bounding scarps of the Chalk tract are seen to
flank the Vale of Pewsey in the north, the upper Nadder valley in the
south-west, and a small part of the Wylye valley above Upton Lovell
(9440). (See also Fig. 3.) One may assume that the Chalk tends to dip
southwards and eastwards away from the boundary thus defined.
Now within the line of scarp-faces, woodland is extensive only in the
east, and on the crest between the Wylye and the Nadder. East of the
Avon below Salisbury the woodland is based on impermeable rock,
as shown by the surface drainage: from what has been said, it may be
guessed that here the Chalk dips under younger, that is, Tertiary, rocks.
In the west, however, the woodland occurs on the highest ground, not
far from the outer limit of the Chalk tract. It is entirely possible, if not
indeed highly probable, that this woodland is underlain by Clay-with-
flints, the product of lengthy subaerial weathering of Chalk. Because
similar large woods are not found on the crests of other interfluves,
it should not be assumed that Clay-with-flints is confined to this stow
alone: former woodland may have been cleared for cultivation.

Geological Structure

Chalk having been identified and its limits broadly defined, an
interpretation of the geological structure may now be attempted.

34 MAP INTERPRETATION

Although much of the tract could be described in terms of scarp-face
and back-slope, as a whole it is not an area of uniclinal structure but
one of folding. This fact may be ascertained in either of two ways :
from a synoptic view of all the scarp-faces shown on the map, or from
a comparison of inferred dips.

The Chalk scarps which bound the Vale of Pewsey face one another
across the low ground and converge towards the east. The same is
true of the corresponding scarps of the upper Nadder valley. In each
case the forms are those of a denuded anticline which pitches, that is,
dies away, eastwards. The extremity of a third similar structure is
found in the Wylye valley, in the westermost part shown. If each
section of scarp-face is taken as running roughly along the strike, and
dip arrows inserted pointing down the back-slopes, it will be found
that, for example, the inferred dips are to the north on the northern
side of the Vale of Pewsey and to the south on the southern an
anticlinal axis must lie between them. An approximately south-
south-east dip near Bratton (9152) opposes one to the north-east at
Heytesbury (9242): here a synclinal axis must intervene. In these
ways a picture is obtained of a tract of open folding, wherein the axes
run approximately east-west and the anticlines pitch eastwards (Fig. 3).

This interpretation is confirmed and extended when the landforms
of the anticlinal vales are closely examined. Sub-scarps are found to
occur at the foot of the Chalk scarp-faces, that is to say, another resis-
tant formation appears from beneath the Chalk, standing out in a
narrow and lower cuesta. South of the Nadder a prominent bench
runs from Donhead St. Andrew (9225) to Barford St. Martin (0531),
more than a mile wide and clearly demarcated by the steep, tree-clad
edge on its northern side. The forms suggest a gentle southerly dip
similar to the inferred dip of the Chalk above. Corresponding features
occur north of the river, but the bench is replaced by a line of hog-
back ridges, here also marked by a line of woodland which runs
westward along Wick Ball (0032) and Ridge Hill (9532). The hog-
back form signifies a steep dip; or, in other words, the upfold is
asymmetrical, dipping steeply on the northern flank and gently on the
southern. No sub-scarp is visible in the nose of the Wylye anticline,
but it may be at once identified in the Vale of Pewsey. In the extreme
west the Chalk and the lower scarps are combined in a single edge, but
eastwards of Erdington (9253) the sub-scarp detaches itself and the line
of woodland begins near Erlestoke (9654). A line of scarp features is
easily traced across the mouth of the Vale, through Devizes, with the

ERODED FOLDS

help of the woodland symbol which shows where the crowded
contours should be looked for.

icorpi ruing above \r\ land above TOO ft

FIG. 3. INTERPRETATION OF PHYSIQUE: THE WESTERN PART OP

SALISBURY PLAIN
(Based, by permission, on Sheet 167 of the O.S. New Popular 1/63,360 Map)

The map indicates no more than a resistant formation, underlying
the Chalk and similar in respect of dip. One might well suggest that
in view of the general concordance of structure it also is likely to
belong to the Cretaceous. In the field the outcrop proclaims itself

36 MAP INTERPRETATION

from a distance. The woodland is largely coniferous (not distinguished
on the modern O.S. 1/63,360 sheet) and is very dark in appearance,
especially in the colder months. The observer is at no pains to
recognize the Greensand.

Further structural interpretation is not easy, except that the low
ground west of Devizes is seen to be based on weak, impermeable
rock presumably clay. Axes of folding additional to those described
are present in the area and affect the relief in varying degree, but, like
the faults which also occur, can scarcely be interpreted from the map.
It has already been suggested that the impermeable rocks in the
south-east may be of Tertiary age, in which event they would occupy
a syncline, but the suggestion cannot be confirmed without more
evidence. The geological basis of a number of hills in the east, which
rise above the Chalk plateau, must similarly remain obscure until the
geological map is consulted. Wexcombe Down (2757) attains 876 ft.,
unusually high for this tract, and well above the crest level of the
nearest bounding scarp. The line of hills from Windmill Down (245 1)
to Quarley Hill (2642) seems to present a scarp-face, marked by
woodland, towards the west, but Sidbury Hill (2150), Clarendon
Hill (2248), and Beacon Hill (2044) on the other side of the Bourne
valley are less clearly defined. The cuesta form is again found in the
high ground which runs north-eastwards from Pitton (2131) and
swings eastwards near Roche Court (2534), and also in Deal Hill
(2526). Since the rock composing these patches of high ground is not
likely to be less resistant than the generality of the Chalk, and, since
dry vall&ys are found, it might be that the scarped hills at least owe
their form to resistant beds high in the Chalk succession (cf. the South

Downs, where two scarp-formers occur within the Upper Chalk).

The Drainage Pattern

It is true to say that in this tract, as in the English Plain as a whole,
not every major structure is unequivocally expressed in the relief. All
the more caution is necessary, therefore, in an interpretation of the
drainage-pattern. If structures are in part unknown, a complete
description of the stream systems in terms of dip-, strike-, and scarp-
streams is obviously impossible, for it would imply not only that
structures are known but also that streams have attained some measure
of adjustment. Maladjustment of drainage, due to youth, glacial
interference, antecedence, or superimposition, is a text-book common-
place, but it is too seldom realized that much of the drainage of the

ERODED FOLDS 37

English scarplands is maladjusted. The effects of glacial interference
and of superimposition are widespread a fact never to be forgotten
in map interpretation.

Although the structures on the area represented on Sheet 167 have
been imperfectly discerned from the map, it can be stated that the
streams are not everywhere adjusted to them. The Nadder and Wylye
occupy anticlinal valleys, the Ebble valley may be synclinal, but the
Till, Avon, and Bourne flow across the lines of the east- west axes.
Now, if the adjusted streams are the older, it is highly improbable that
they would have thrown out tributaries across the structural grain:
the maladjusted Avon is likely to be the earlier, with the adjusted
streams developed as subsequents along the strike of weak rocks. In
other words, the drainage appears to have been superimposed on a
group of pitching folds, to which it is as yet only partially adjusted.

Wind-gaps in the scarp crest, on the southern side of the Vale of
Pcwsey at 9251 and 0151, testify to the capture which adjustment
involves. The Till and a feeder of the Wylye have been beheaded by
the Semington Brook, a tributary of the (Bristol) Avon, draining the
weak outcrop in the north-west. The beheaded streams once rose
north of the present crestline, as the (Wiltshire) Avon still does. Note
that the divide between the two Avons lies on the resistant formation
below the Chalk, somewhat east of the crest of the Devizes scarp.
The Bristol Avon is gaining ground here also.

Close inspection shows that the Chalk scarps, and the Great Ridge
between the Wylye and the Nadder, have even crestlines, higher than
650 ft. O.D. for considerable distances, but not often above 750 ft.
This accordance of level suggests that the flat crestal belt has survived
from a former more extensive erosion-platform. If so, the cycle in
which the platform was produced must necessarily have reached an
advanced stage. It follows, too, that the captures recorded in the wind-
gaps must have taken place in the present cycle, for gaps produced in
an earlier cycle would have been obliterated by planation.

The incised condition of the larger valleys of the Chalk tract proves
rapid downcutting in the present cycle, besides incidentally confirming
the resistant nature of the rock. Weak rocks could not support such
steep valley walls. Interpretation of drainage is somewhat hampered
by the numerous minor channels in the valley bottoms, but it is amply
shown that the rivers meander within meandering valleys. Since the
general problem of valley-meanders has been outlined in Chapter III,
further discussion can be dispensed with, but the present examples

4 -(E.5i9<>)

38 MAP INTERPRETATION

are well worth close attention if only by virtue of their bold, unmis-
takable forms. The inner valley of the Avon between West Chisen-
bury and Enford (1352) describes a great curve, within which the
river traces loops of much smaller amplitude. Similar relations are
found, for example, on the Bourne below Idmiston (1937) and on the
Nadder (9829). It is the valley-meanders which are ingrown, and the
spurs between them which have been eroded, that is, sharpened on
their upstream sides. Observe, for instance, the three successive spurs
in the Avon valley between Upper Woodford (1237) and Newton

.
The minor channels mentioned in the previous paragraph are much

too regular to be other than artificial: they are the "drawns" of water
meadows. Where they occur the valley must be flat-bottomed, and
the Avon both above and below Salisbury evidently flows over a
winding ribbon of alluvium. From Charlton (1723) downstream the
western side of the valley bottom is dry, however. Settlements occur
on the low ground and the main road runs along the valley floor.
These are signs of a terrace standing higher than the flood-plain, which
although not indicated by contours is identifiable by other distribu-
tions. One concludes that the downcutting in the present cycle has
been intermittent.

Evolution of the Landscape

The several inferences may now be combined in the following
minimiyn sequence of events required to account for the physical
landscape as mapped

1. Folding along east-west axes; anticlines pitching eastwards,
at least one upfold steeper on its northern side.

2. Prolonged denudation, producing a surface of low relief. This
surface is recorded in the accordant summits of the present Chalk
scarps.

3. Fall of base-level, initiating new cycle later followed by further
intermittent rejuvenation. It is probably the consequent drainage
of this cycle that has been superimposed.

Adjustment of drainage to structure is still incomplete in the
present cycle (cf. wind-gaps, beheaded consequents, streams flowing
across fold axes).

Dry valleys belong to this cycle, since they are related to the
present river system and lie well below the level of accordant crests.

ERODED FOLDS 39

As might be expected, on general grounds, the full sequence is
much more complicated, but the outline given here is valid as far as it
goes and of considerable value in the understanding of the terrain
represented.

NOTES AND REFERENCES

For a general account of the geology, see

C. A. CHATWIN. The Hampshire Basin and Adjoining Areas. Second
Edition. British Regional Geology Series, H.M.S.O., 1948.

The geomorphology is discussed, with more accuracy and in greater
detail than in the foregoing text, in

S. W. WOOLDRIDGE and D. L. LINTON. Structure, Surface and Drain-
age in South-cast England. George Philip, London, 1955. (See especially
Fig. 10 and the accompanying text.)

Certain aspects of Chalk topography are dealt with by

A. J. BULL. "Cold Conditions and Land Forms in the South
Downs." Proc. Geol. Assoc., li, 1940, p. 63.

C. C. FAGG. "The Recession of the Chalk Escarpment." Transactions
of the Croydon Natural History and Scientific Society, ix, 1923, p. 93.

F. K. HARE. "The Geomorphology of a Part of the Middle Thames/*
Proc. Gcol. Assoc. , Iviii, 1947, p. 294. (See especially pp. 326-8.)

J. F. KIRKALDY. "Solution of the Chalk in the Minims Valley,
Herts/' Proc. Geol. Assoc., Ixi, 1950, p. 219.

W. V. LEWIS. "The Pegsdon Dry Valleys/' Compass, i, No. 2, x,

1949, p- 53-

S. W. WOOLDRIDGE and J. F. KIRKALDY. "The Geology of the
Minims Valley/' Proc. Geol. Assoc., xlviii, 1937, p. 307.

The water meadows represented on Sheet 167 are among those
discussed in

H. P. MOON and F. H. W. GREEN. "Water Meadows in Southern
England." Appendix II, p. 373, to F. H. W. GREEN: The Land of
Britain, Part 89, Hampshire. Geographical Publications, London, 1940.

CHAPTER V
UNGLACIATED UPLAND

Besides the story of the rocks, we may try to trace that of the
surface itself. MACKINDER

MAP: O.S. 1/63,360 (SEVENTH SERIES) SHEET 186
(BODMIN AND LAUNCESTON)

WHEN this area is first inspected for evidence of "structure, process,
stage," one immediately notes strong contrasts with the areas of
tilted and folded sedimentary rocks treated in Chapters III and IV.
The area is a low plateau, undergoing dissection by numerous streams
which flow in valleys mostly deep, narrow, and steep-sided. There is
nothing corresponding to the broad strike vales or denuded anticlines
previously studied, where weak clays formed the basis of wide stows
of low ground. Here the rocks are generally resistant. Because of
this, and because little progress has been made in the present erosion-
cycle, geological structure must in part remain obscure. There are,
however, compensations for the interpreter, as will shortly be seen.
The processes now at work are those of normal erosion and the shore-
line cycje. Because the area bears none of the marks of regional
glaciation described below (see Chapter VI) it seems probable that
normal and shoreline processes between them have shaped the entire
landscape in its present form. The stage reached in the present sub-
aerial cycle is that of youth, shown, for example, by the deep youthful
valleys and the broad, flat crests which separate them. When the
landscape is mature, opposing valley walls will have retreated so far
that they meet along the crests of the divides, and the flat tops will
have been consumed. It can be seen at a glance that maturity is some
way off. Now the subdued relief of the crestal belts is the product of an
earlier cycle than the present. The general form of the landscape
produced in that earlier cycle would be realized if the existing valleys
were filled in. As the existing landscape combines features produced
in more than one cycle it is styled polycyclic. The map selected is
especially well suited to illustrate the broader features of a polycyclic
landscape, and the method of treatment employed in the following

UNGLACIATED UPLAND 41

paragraphs is intended to exemplify some of the possibilities of inter-
preting landscape history.

Erosion-platforms

The texture of relief is of two kinds. On Bodmin Moor, which lies
north-west of centre and shows up prominently as a blank in the
patterns of roads, settlement, and woodland, the valleys are wide,
shallow and marshy. Most of the Moor is higher than 800 ft. O.D.,
with single hills exceeding 1,000 ft. and two over 1,300 ft. This part
of the landscape is approaching a senile condition, for the higher hills
are mere residuals above a wide, gently sloping plateau surface. In
the tracts surrounding the Moor it is the interfluves that are broad and
flat. They are separated by the incised young valleys already mentioned.

The contrast of texture corresponds approximately to differences of
lithology. China clay pits, for example at 1970 and 1381 on Bodmin
Moor, and the extensive workings on Hensbarrow in the south-west
indicate that these uplands are based on granite. 1 Other signs of this
rock are the many tors, where residual piles of joint-blocks have
survived the attack of erosion. Tors are usually looked on as the result
of subacrial weathering, but it has also been suggested that some, at
least, may represent the stacks and islets of shorelines higher than the
present. Subaerial processes arc chiefly responsible for detaching the
boulders which clutter the steeper hillsides of granitic country, and
which are indicated on this map by the symbol for rock-strewn ground.

It is clear that the granite of Bodmin Moor stands above the sur-
rounding tracts in virtue of its greater strength, but even here one
finds evidence of severe denudation. The very fact that the originally
deep-seated granite is exposed at the surface proves that a great thick-
ness of cover has been stripped off in cycles of which no trace remains.
The earliest cycle to have left a recognizable mark on the present
landscape is that in which the granite has worn down to the surface of
low relief noted on first inspection. A closer examination shows
that, at c. 850 and 1,000 ft. O.D., the summit of the Moor is a
broad, open expanse with slight gradients and correspondingly wide-
spaced contours. Such a plateau surface, cut across the granite, cannot

1 One may note in passing that the origin of granite and china clay is wrongly stated
in many older geographical texts. The alteration of granitic felspar to kaolin appears to
have been a metamorphic process rather than the work of subaerial erosion. The forma-
tion of granite is a matter of geological controversy, but it may be said that many author-
ities regard this rock as incorporating much pre-existing material, rather than as being
simply a solidified magma.

42 MAP INTERPRETATION

be other than erosional. It requires, moreover, a long still-stand to
account for the extensive wearing-down of the resistant rock, whether
by subaerial or marine erosion. Since planation was nearly attained, it
may be inferred that the base-level by which the erosional processes
were controlled was not far from the bounding contour of the plateau,
say, c. 800-850 ft. O.D. The higher hills overtopping this erosional
platform may be residuals from an earlier cycle, for a number of
summits are fairly flat, but as a group the hills are too few and too
small to justify any attempted reconstruction of a higher platform.

On the flanks of the granite outcrop there is a sharp descent from
the c. 850-6. level, obvious enough on close inspection but much more
clearly revealed by generalized contours. 1 Outside Bodmin Moor and
Hensbarrow the ground is nearly all below 750 ft., with a general
slope coastwards to levels of 200-400 ft. at the cliff-tops. As already
noted, the surface is that of a low plateau, broadly preserved on the
flat interfluves. It is interrupted at 2757 and 3771 by steep-sided hills,
which, since they carry moorland vegetation, may again be granitic,
and is undergoing dissection by streams flowing in young valleys.
The summit surface, like the plateau of Bodmin Moor, is erosional,
not structural. It lies lower than the erosional platform cut across the
granite, and is developed across rocks in which the granite is cmplaced
and which are therefore likely to be considerably disturbed and in part
altered. Their structure cannot be discovered from this map, even in
outline, for the planation recorded in the summit surface seems to
have b^en complete, and differential erosion in the present cycle has
made little headway. There is a hint of structural graining in the
river pattern east of Bodmin Moor, where a number of east-west
reaches ar$ to be observed, but this fact is significant only if the streams
are adjusted to structure.

Former Base-levels

The non-granitic outcrops must have been denuded with reference
to a lower base-level than that of c. 800-850 ft. which is thought to
have controlled the levelling of Bodmin Moor. Presumably one or
more falls of base-level occurred, with still-stands long enough for all
the rocks except the granite to be worn down, whether by normal
erosion or by the sea. There is, in fact, fair indication on the map of at
least one intermediate still-stand. Near Tintagel (0588) a narrow
platform occurs between the cliff-top at c. 300 ft. and a marked
1 Sec below, Chapter XVI.

UNGLACIATED UPLAND 43

topographic riser on the landward side which extends from c. 400-
600 ft. O.D. or over. This steep slope might be the line of a former
cliff, cut when the strand-line stood some 400 feet above its present
level. In general, however, it cannot be claimed that any old shore-
lines are well indicated by the map, however accurately they may be
defined in the field. The closest scrutiny (on the map) of the immediate
coastland will often fail to produce evidence of base-level changes
which must have occurred to account for the forms of the valleys. It
is so here. The incised valleys are a sound indication that, after the
planation which is recorded on the flat intervening crests, base-level
fell considerably. As a result the streams were rejuvenated and cut
deeply into the erosional platform.

On the high moorland, however, the valleys are not deeply cut:
contours cross them at wide intervals, showing a gentle downstream
gradient. It is evident that the headward waves of rejuvenation,
propagated by the falls of base-level, have not yet reached the head-
waters. Somewhere the long-profiles must be broken by knickpoints,
where the older and newer elements intersect. The only reliable way
to locate a knickpoint is to survey a valley in detail on the ground, for
contours are apt to generalize the actual conditions and to that extent
to be misleading. However, where rejuvenation has so obviously
occurred as it has here, and where the stream profiles are generally
steep, major breaks of profile may be revealed to inspection alone. In
fact, contours crossing the valleys are significantly bunched, usually a
short distance above the boundary of the granite, as follows : on the
Inny, from 700 ft. downwards at 1786; on Penpont Water, below
c. 650 ft. at 2181; on the Lynher, below c. 700 ft. at 2379; on the
Fowey, below c. 600 ft. at 2268 where Golytha Falls are marked; on
the St. Neots River below c. 750 ft. at 1871 ; on the Warleggan below
c. 600 ft. at 1470; on the De Lank below c. 650 ft. at 1075; and on
another unnamed tributary of the Camel, below c. 650 ft. at 1078. The
rough similarity of height, taken in conjunction with the considerable
vertical range of each steep descent, suggests that most of these
plunges of long-profile are knickpoints related to a still-stand of base-
level. They do not appear to correspond to geological distributions.

Denudational Sequence

The polycyclic nature of the present landscape is by this time
sufficiently evident. There are, however, additional features to be
considered. It will be found when the shoreline is examined that the

44 MAP INTERPRETATION

deep inlets are drowned river valleys (Plate IA). In other words, base-
level was at one time lower than it now is, and has risen to its present
height.

Unaided inspection has now revealed evidence of the following
minimum sequence of denudational history

1. Unroofing of the granites (very lengthy).

2. Widespread but not complete planation of Bodmin Moor,
controlled by a base-level of 800-850 ft. O.D.

3. Fall of base-level to 600-6506. O.D.; shallow valleys cut in
the granite, non-granitic outcrops severely denuded.

4. Further fall of base-level, probably intermittent with one
still-stand detected at c. 400 ft. ; non-granitic outcrops further
planed off.

5. Relatively rapid fall of base-level to below O.D.; drainage
generally rejuvenated and valleys incised.

6. Rise of base-level to present position, valley mouths drowned.

While the refined techniques of morphometry are likely to elaborate
this sequence, and field-work is certain to do so, it can be claimed that
interpretation on sight, as it were, has made possible a useful genetic
description of the landscape represented. Individual valley features
can be placed in due perspective against a background of long-
continued but intermittent rejuvenation.

Superimposed Drainage

It must be stated at once that much less can be discovered about the
drainage system than about the major facets of the landscape. Certain
streams, or stream reaches, follow well-defined lines, but it cannot be
said how these lines have been determined. The Lynlier throughout
most of its length, and neighbouring streams to a greater or lesser extent,
pursue a south-easterly direction ; the Fowey runs westward for about
five miles below Doublebois (1964) ; the Tamar system in the north-
east combines east-west with north-south reaches in an approach to a
trellis or espalier pattern. Although on general grounds it is possible
that some reaches follow structures, there is no means of deciphering
the structural pattern. The Lynher to one side of Bodmin Moor and
the Camel on the other run in part near the boundary between the
granite and the surrounding rocks, but this relationship seems to be
accidental, since some of the streams parallel to the Lynlier flow across
the granite while others rise beyond its limits. In other words, there is

UNGLACIATED UPLAND 45

little connection between the direction of streams and the geological
distributions inferred from the map.

The most useful information comes from a study of the Tamar.
From the bridge between Crossgate and Pool (3488) to Greystone
Bridge (3680), a distance of about seven miles, the river flows on a
valley floor which for this area is relatively wide, and which appears to
have been swept over by meanders in their migration down-valley.
Below Greystone Bridge the river enters a defile, where the meandering
course is deeply incised between steep, close-set walls. In the course of
an incision of at least 150 ft. the two meanders which enclose Duntcrue
Wood and Wareham Wood (3878) have enlarged themselves laterally,
slipping oft the inner banks and undercutting the outer, but have not
shifted far downstream. The meander lobes are as yet little trimmed
on the upstream side. This assemblage of forms, which is paralleled
at Lamerhooe (3973), indicates a resistant outcrop where the meanders
have not succeeded in sweeping out an alluvial trough. The Tamar
passes from the weaker to the stronger formation at Greystone Bridge.

It is certain that the river is maladjusted to structure; but it has been
concluded from an examination of the landscape that this area was
severely denuded in an earlier cycle, during which one would expect
the drainage to have become adjusted. A likely possible explanation
is that the low plateau into which the Tamar is now incised was
covered by a thin veneer of terrestrial or marine deposits, from which
rivers were superimposed on to the outcrops now revealed.

NOTES AND REFERENCES

For a general account of the geology, see

H. DEWEY. South-west England. Second Edition. British Regional
Geology Scries, H.M.S.O., 1948,

Erosion-platforms in part of the area shown on the selected map
are described by

W. G. V. BALCHIN. "The Erosion Surfaces of North Cornwall/'
Geogr.Journ. xc, 1937, p. 52.

Other relevant material occurs in

J. A. STEERS. The Coastline of England and Wales. University Press,
Cambridge, 1946, pp. 254-260.

W. G. V. BALCHIN. "The Erosion Surfaces of Exmoor, etc.,"
Geogr. Jotirn., cxvii, 1952, p. 453. (See also part of Chapter XV
(below) including Figs, n and 13.)

CHAPTER VI
GLACIATED HIGHLAND AND A DRUMLIN FIELD

Hills peep o'er hills, and Alps on Alps arise. POPE

MAPS: O.S. TOURIST MAP, 1/63,360 (LORN AND LOCHABER);
O.S. 1/25,000, SHEET NY/71 (BROUGH)

IN the field, the landforms of highland glaciation are boldly defined,
not to say imposing. They are clearly distinguishable from one an-
other, and photograph well. Hence it is easy to become familiar, if
only at second hand, with the appearance of glaciated mountains, and
to identify on the spot features previously encountered only in the
text-book. In a general way the large erosional features also are
clearly represented on the map, by means of standard symbols, but
certain depositional features show up poorly, if at all.

Landforms of Glaciated Highlands

The fundamental landform of glaciated higliland is the corrie, the
hollow in which snow accumulated and consolidated before moving
down-hill. When freshly revealed by deglaciation the corrie is seen to
approacji, more or less closely, the form of a cylinder, with a shallow
concave floor and a precipitous partially enclosing wall. Although
there is some difference of opinion about the precise way in which
corries arejformed, there can be no doubt that the walls arc cut back
during glaciation, so that corries eat into the high ground. The sheer
slopes may be represented by rock-drawing, as on the Fort William
Sheet at the flank of Ben Nevis (1671) and along the high ground
eastwards to Stob Coire na Ceannain (2674), or by unbroken contours,
as, for example, in Coire Dubh (0882) and the nameless corries in
0471 and 1634, each of which is a fine example of the corrie form. The
slightest field knowledge of this type of country compels the interpreter
to remark also those less noticeable corries which are shown merely
by inflected contours, for example in grid squares 9570-9670, and the
unnamed corrie on the northern flank of Sgurr a'Mhaim (1666)
where the 3,ooo-ft. contour runs near the foot of the head wall.

In geography the term tarn is applied exclusively to corrie lakes.

GLACIATED HIGHLAND AND A DRUMLIN FIELD 47

They occupy shallow basins liable to infilling by peat, downwash, and
the coarse scree weathered off the corrie walls. Some have already
been drained as the effluent stream has cut down; and, since an enclosed
basin was not formed in eveiy corrie, in a given tract of glaciated
highland tarns may be uncommon, as they are here. Those which
occur are instructive. There are four between Binnein Mor (2166)
and Binnein Beg (2267), of which the largest draws attention to a well-
marked but shallow corrie. In Coire Leis, on the northern side of
Ben Nevis, one single tarn and a group of three indicate what can also
be read from other detail, i.e. that Coire Leis is composite. It is, in
fact, a short glacial trough, with corries running together along the
sides and at the head.

Where high ground is scalloped by encroaching corries, biscuit-
board topography results. In this tract very little remains of a pre-
glacial plateau surface, but a limited amount of subdued country at a
high level is to be observed on the divide running northwards through
Aonach Beg (1971) and Aonach Mor (1973). If glaciation goes on
long enough, neighbouring corries merge into one another. A pair of
hcadwalls on opposite sides of a divide, or a pair of juxtaposed side-
walls, intersect in an arete (knife-edge ridge). Three or four opposing
hcadwalls intersect in a sharp peak or horn, of which the Matterhorn
is the type. It is rather remarkable that in the glaciated highlands of
Britain the true horn is rare, while many ridges between adjacent
corries have not been completely sharpened into aretes. In the tract of
the Grampians represented here, the highest summits and the crests of
ridges are generally narrow, but nevertheless preserve rounded not
sharpened forms. From Stob Coire Easain (2372) to Stob Coire na
Ceannain (2674) the forms of the crests are entirely comparable to the
subdued relief noted on Aonach Beg and Aonach Mor a few miles to
the west, except that here less of the crestal belt has survived. On the
southern side of Ben Nevis itself, the relief is comparatively feeble
above c. 4,000 ft., with little beyond a field of riven blocks to
suggest frost-action. It would seem that, in some parts of the area
mapped on this sheet, glacial processes just failed to destroy completely
the pre-glacial topography.

During glaciation the corries were the headward catchment areas of
valley glaciers. Small tongues of ice from the corries coalesced into
great streams moving in U-shaped troughs. If corries are the funda-
mental landforms of a glaciated highland, glacial troughs are the
most distinctive, with their straight reaches and angular plan, steep

48 MAP INTERPRETATION

sides, truncated spurs, hanging tributary troughs, and ribbon lakes
(Plate IB). Their long-profiles, despite some alluvial fill, are character-
istically irregular in contrast to the smoother long-profiles of rivers.
Most of these characteristics are sufficiently apparent on the contoured
map, except perhaps for truncated spurs, which although numerous

Reconstructed by prc-glacial

relief.
(6) Corries, glaciers, and direction

of ice-movement,
(c) Present landforms.

(D) The Dubh Lighe.
(F) The Finnan.

i) Loch L-iJ.
(L.V) Loch Shicl.

FIG. 4. DIVERSION OF DRAINAGE BY THE AGENCY OF ICE
(Based, by permission, on part of Sheet 47 of O.S. Scotland (Popular) 1/63,360)

in the field and unmistakable in perspective along a trough reveal
themselves on the map only when very large and especially clear-cut,
for example, the eastern flank of Sgor Chalum (1369). The rock-steps
which typify many glacial troughs, and which are in part responsible
for the irregularities of long-profile, are not always revealed by
contours even when the contour interval is small. The major rock-
step or "trough's end," near the head, is the most likely to be identifi-
able, since as a rule it is high and stretches the whole way across the
trough (cf., in the valley of Allt Coire an Eoin, the small steps at 221733

GLACIATED HIGHLAND AND A DRUMLIN FIELD 49

and 218729 with the great step running through 210720, which rises
as far as 2,250 ft. O.D.). Like many small rock-steps, roches moutonnees
can scarcely be represented, unless by a special symbol on a large-scale
map. There are several noteworthy groups in the area of this map,
including a fine series in the Gallop valley between 910804 and 925793,
of which the map gives no hint. Terminal valley moraines, which are
numerous, are again unreflected by the contours. Their small, chaotic
mounds cannot be shown on the scale employed, even where the
range of height is sufficient. Lateral, medial, and ground moraines
are also unrepresented.

Renewed Normal Erosion

Since deglaciation there has not been time for streams to achieve
grade in the new normal cycle, and torrential reaches and falls are
numerous. The waste-mantle is similarly ungraded, for many slopes
remain oversteepened by glacial action and bare of the debris of
weathering. Nevertheless the details of relief have already been
considerably modified, especially by deposition.

On some upper slopes frost action, together with the rapid removal
of loose material, has tended to maintain a freshly denuded aspect,
but on the lower slopes and in valley bottoms the forms of glacial
erosion may be widely and often also thickly covered. Sheets of rock-
waste, in places bare scree and in others with a covering of soil and
plants, commonly lie banked against the lower parts of the trough
walls. These conditions are excellently displayed on the north-
eastern side of lower Glen Nevis, where debris-sheets conceal the true
steepness of the rocky sides. Elsewhere the masking waste is moraine
rather than scree, as in the Loch Eil valley to the west of Fort William.

The very many streams testify to impermeable rock and a heavy
rainfall. Run-off is rapid down the steep hillsides, even allowing for
the fact that hill peat acts as a reservoir, with the result that trunk
streams are very powerful at times of high water. The short lateral
tributaries bring down much rock-waste, which is deposited where
they are checked on the floor of the main valley or in a lake. Few of
the delta-fans on the land are perceptible on the map, partly because
the ground is in any case generally varied in fine detail, but lake deltas
on the other hand are distinctly outlined by the plan of the shore.
Lateral deltas may nearly approach the fan shape, as at Stronchreggon
House (0672), but lake-head deltas are frequently less regular (e.g. at
Kinlochctive, 1145). Damming of a lake may cause lateral deltas to

50 MAP INTERPRETATION

be drowned, and lead underwater contours to be omitted from the
map, as with the Blackwater reservoir on northing 60.

Most of the rock-waste in sheets of scree, deltas, alluvial flats, and
related forms is either re-sorted moraine or the product of weathering,
but corrasion by post-glacial streams has also provided some. On the
map, the effects of this corrasion are best detected where a rock-step
or the lip of a hanging valley has been deeply notched, as at 280734.
(Note the sharp re-entrant in the 1,750-6:. and adjacent contours.)

Glacial Troughs and Through-valleys

Soundings in the freshwater lochs of Scotland, as in other ribbon
lakes elsewhere, prove great depths. Loch Morar, a freshwater loch
adjacent to the west coast of Scotland, descends some 1,000 feet below
sea-level. Although detailed investigation is required to show how
much of the barrier at the lower end of a lake consists of moraine, and
how much rock in place, there can be little doubt that in many glacier
troughs erosion was deepest some way above the glacier snout.
Towards its lower end, a Pleistocene valley glacier must, like the
glaciers of to-day, have been much reduced by melting and ablation,
so that its cross-sectional area diminished down-valley. This is one
explanation of the "down-at-heel" erosion which excavated the great
lake basins. In their original form they must have been even more
impressive than they now are, for some have been completely silted
up, while in the remainder alluvial infilling has made progress to a
varyirlg degree.

Now these enormous hollows cannot have been formed unless the
basal ice moved locally up-hill. Such movement is made possible by
the very farge cross-sectional area of a glacier, by comparison with the
cross-sectional area of a river with an equivalent discharge. The
comparison is well illustrated at any glacier snout. Because ice-streams
are so wide and thick, they can override irregularities in their beds
behind which rivers would be impounded. It is necessary to emphasize
the phrase glacier bed, for the great troughs were the channels of glaciers.
This fact is usually advanced in explanation of hanging tributary
troughs, which are taken as the channels of tributary glaciers. The
tributary ice entered the main glacier accordantly, at the surface, but
required a much smaller channel hence the difference of height
between the channel floors now revealed by deglaciation.

The explanation thus briefly outlined is acceptable when the hanging
tributary leads back into the high ground, and terminates in corries,

GLACIATED HIGHLAND AND A DRUMLIN FIELD 51

but some lateral troughs merely pass through a divide into the next
trough (cf. that at 2257). As the l/-section of these cols testifies to the
passage of ice through them, it is evident that at some points ice has
spilled over the divide. Accessible examples of the resultant glacially
moulded col occur on either side of Dun Deardail (1270), where two
small gaps in the crest lead out of Glen Nevis and debouch above the
corrie at the head of the next valley on the south. It seems certain that
ice has passed southwards over the divide here.

Divergence of ice-streams in this manner, made possible by their
great thickness, may be associated with local uphill movement towards
the lateral outlets. Linton has convincingly demonstrated that, at a
number of sites, the sill of the over-ridden col has been much lowered,
whether by the lateral over-spill of distributary glaciers glacial
diffluence or by ice-movement across divides glacial trans-
fluence. At some points the rivers, recommencing the work of
normal erosion on deglaciation, have flowed through the lowered cols
across the lines of pre-glacial divides in the direction taken by the
diffluent ice. Elsewhere two rivers, flowing in opposite directions in a
glacial trough, are separated by a very low watershed. It can often be
shown, or at least inferred, that the pre-glacial divide has been des-
troyed and substituted by an imperceptible or temporary parting, for
example that formed by a delta-fan. Two distinct pre-glacial valleys
have been merged into a single trough or through-valley, so called
because it pierces a broad divide or runs through generally high
ground from side to side.

Both sets of conditions appear to be represented on this map. At
271643 and 240695, for instance, the parting between opposed drainage
in a glacial trough is most ill-defined: ice must have passed along the
whole length of the trough, moving away from its gathering-ground
and eroding outlet channels. It is unnecessary at the present juncture
to comment on the significance of through-valleys in relation to
routeways, since the facts are evident enough in themselves and belong,
in any event, under another head; but it is well to emphasize that
through-valleys, formed in the manner described, are typical of
glaciated highlands and to be deliberately looked for in map
interpretation.

Diversion of Drainage

The effects of glacial diffluence in river diversion are probably
illustrated by the country between Glenfinnan (9080) and Drimsallie

52 MAP INTERPRETATION

(9578). The Dubh Lighe descends from a tributary trough into the
Loch Eil valley, where it turns sharply eastwards. The Gallop, des-
cending the opposite side, turns sharply westwards instead of joining
die Dubh Lighe, and enters a narrow defile through the high ground.
Near their elbow-bends the two streams are separated by a very low
divide, scarcely perceptible on the map but seen, in the field, to consist
in part of river-laid debris. Since the roches nwutonnees within the
defile (referred to above) present their stoss or onset sides to the east,
it would seem that the gap is due, at least partly, to ice which passed
westwards through it. Thus field evidence supports the hypothesis
that map interpretation suggests, namely that the Gallop flows west-
wards along a former line of ice-movement, and that its sharp turn is
the result of glacial diversion not of river capture in the normal cycle.
Although in a full study certain additional facts, for example, strand-
line movements, would have to be taken into account, the hypothesis
as stated is satisfactory as far as it goes. It does provide a possible
explanation of the drainage forms, while river capture docs not.

Relation of Drainage to Structure

It should now be seen that the problem of the drainage pattern in
glaciated highlands must be approached with caution. Even if the
Scottish Highlands had not been glaciated, many pitfalls would await
the unwary: the drainage net represented on this sheet cannot possibly
be interpreted in terms of the stream system of an ideal scarpland a
fact too often overlooked. A little explanation will make the matter
clearer. In areas of uniclinal strata, where drainage is to some extent
adjusted to structure, streams of any size will flow rouglily parallel to
the dip er else roughly parallel to the strike, either as a whole or in
their various reaches. The abrupt angle where a large stream changes
from flowing down-dip to flowing along the strike is most obviously
interpreted as an elbow of capture. But in the Scottish Highlands the
structural grain is very different from that of a scarpland. Although
there is an indescribable variety of detail, it is probably justifiable to
say that large areas are dominated by two sets of linear structures which
intersect at a high angle. Consequently the broad pattern of relief
of ridges, valleys, and the rivers in them itself tends to be angular.
In the part represented on this map a number of valleys run approxi-
mately east-west, while a second group trends from north-east to
south-west. 1 The scale of relief makes it clear that the rocks of the

1 The Ben Nevis mass, an old volcanic centre, does not display the same graining.

GLACIATED HIGHLAND AND A DRUMLIN FIELD 53

whole area are resistant. The lines of weakness which the valleys
follow are determined by structures not by outcrops. In such country,
especially after glaciation, the idea of dismembered consequent streams
ceases to be helpful in map interpretation, for angular bends in stream
courses are seen as the natural result of adjustment to structure.

Changes of Base-level

The two topics outstanding have in common a relation to former
base-levels of erosion. The Parallel Roads in the north-east, symbolized
on the map by double pecked lines, are so nearly horizontal as to seem
undoubtedly referable to some local series of base-levels, which can
scarcely have been provided by anything but bodies of water contained
in the valleys. Nothing more can be read from the map; but,
as is well known, the Roads are the littoral benches of temporary
lakes impounded by ice. Glaciers from the Ben Nevis group blocked
Glens Gloy, Roy, and Spean, which filled with meltwater up to the
level of the lowest open col. As lower cols were uncovered by the
melting ice the lake-levels intermittently fell and benches were cut-
and-built at the new shorelines. It may be added that the benches are
generally narrow, especially those intermediate ones which are not
shown on the map. The controlling spillway at about 850 ft. was a
lateral drainage channel, unspectacular but distinctly recognizable
on the ground at 288812, at the outlet at the lower end of Glen Spean,
where lake-water escaped past the wasting but still obstructive ice
which descended from the mountains to the low ground.

Earlier in this chapter it was stated that some of the highest ground,
above corrie level, displays subdued relief. Although frost-riving has
attacked these elevated sites they do not seem to have been overridden
by ice, and might therefore provide some slight indication of the form
of the pre-glacial landscape. Within the limits of this map, many
summits to the west of Glen More and in the extreme north-east are
found to lie within, or very little outside, the range 2,000-3,000 ft.
with a majority between 2,000 and 2,500. A wider area would have
to be studied before the possible significance of this distribution could
be fully appreciated, but one can at least make a note of the relatively
small range of height, and advance an extremely tentative suggestion
that the numerous summits in the range 2,000-2,500 ft. may be
remnants of an erosional surface of low relief. East of Glen More, in
the Ben Nevis group and in Mamore Forest, a number of sum-
mits rise above 3,000 ft. and some above 4,000. If the possible

3 (.5196)

54 MAP INTERPRETATION

subdued erosional surface at the lower level is authentic, the higher
mountains in the south-east might well have risen above it as
monadnocks.

Lowland Glaciation

It is a text-book commonplace to contrast highland glaciation,
dominated by erosion, with lowland glaciation, dominated by de-
position. The generalization should not be pressed too far, for some
lowland parts of the Laurentian and Baltic Shields were centres of ice
dispersion, and have in consequence been stripped of their waste-
mantle and deeply scoured besides; but the resulting terrain of multi-
tudinous rock knobs, lakes, and watercourses mammillated
topography is poorly represented in Britain, where ice did in fact
move outwards from the hills and leave thick deposits on the low
ground.

At the farthest limit of advance, where melting held the ice-front
stationary, debris accumulated in terminal moraines; temporary
halts in the decay of the ice-sheet are marked by recessional moraines.
In front of the moraines, the debris has in many places been redistri-
buted as outwash sands and gravels, in some areas pitted by the
kettles where detached masses of ice were buried and melted. Much
of the outwash material was transported by streams of meltwater
issuing from the ice-front, often from tunnels within or beneath the
glacier. Fans of debris laid down by streams of this kind now appear
as roughly conical hills of gravel, called kames, while the stream
courses themselves may be recorded in gravelly ridges eskers
which wind across country. On beaded eskers, kames occur from
place to place, marking points at which the ice-front rested for
a time.

Examples of all these forms have been located in Britain, but for
various reasons are not well shown by the standard map. It will be
realized that, comparatively speaking, none is of great size, so that they
easily escape representation with a contour interval of 50 ft. especially
as half the contours are interpolated. Again, the English Midlands,
where the effects of glacial deposition in a lowland tract could be well
displayed, have mostly remained ice-free since the older Pleistocene.
Their dritts have been deeply dissected and in part destroyed. Finally,
the features mentioned are characterized not only by their form, which
a sufficiently detailed topographical map would show, but also by
their geological composition. Just as the low sinuous ridge of an esker

GLACIATED HIGHLAND AND A DRUMLIN FIELD 55

is concealed on the topographical sheet by a general irregularity of
surface, so the material of which it is made is included, on the standard
geological map, in a wider spread of glacial drift. Work in the field,
and special maps, must provide the illustrations of these smaller
features of lowland glaciation.

Drumlins

The greatest sum effect on relief is produced by the sheets of boulder
clay, with their associated sands and gravels, which in places thickly
conceal the topography of the "solid" rocks. Some boulder-clay
spreads cannot be identified from the topographical map, but the
interpreter working on lowland parts of the Midlands, on East Anglia,
or on the eastern plains should be aware that he is likely to be dealing
with a drift cover, more or less complete. Valley trains of outwash,
like many river terraces, are also likely to escape notice. In some of
the glacially impounded lakes known to have existed on the English
Plain great amounts of sands, gravels, and laminated clays were laid
down, but neither the lake deposits nor the shoreline features are
usually distinguishable on the O.S. map. In contrast to all this lack of
information, drumlin swarms are unmistakable. As exemplified
on the selected 1/25,000 sheet, drumlins consist of low hills of oval or
elliptical plan, 50-100 ft. high and J-^ mile long. They occur in
close-set groups, as here, with their longer axes roughly aligned in a
single direction, giving rise to the "basket of eggs" topography.
Their exact mode of origin has been keenly argued, but there is little
doubt that they have been moulded by a vigorous ice-sheet, principally
from ground moraine, but in places from the solid rock. The ice near
the ground moved parallel to the long axes of the drumlins in this
tract, along N.W.-S.E. lines in the centre and north, and W.-E. in
the south, centre, and east. The actual direction of movement can be
found by close study, for drumlins usually taper more gradually at the
one end than at the other. The blunter is the stoss end, from which
the ice came, the sharper is the lee end, which points the direction of
movement. Unless contours have been carefully surveyed at close
intervals it is not always possible to detect on the map a systematic
tapering throughout the drumlin field, but the expected asymmetry is
suggested by die contours of Ketland (720184), Hemmel Hill (742133),
Bermer Hill (743149), and others. The point should be very carefully
checked, for the basal ice which moulded the drumlins did not
invariably move down-valley. Hollingworth has proved that in the

56 MAP INTERPRETATION

Eden valley, wherein the mapped area lies, basal ice-sheds did not all
coincide with the centres of dispersion on high ground. The drumlins
shown on this 1/25,000 sheet were shaped by basal ice which moved
south-eastwards from near Appleby and swung eastwards to pass over
the higher ground near Stainmore.

Immediately after deglaciation, drumlin fields typically contain
many small enclosed hollows, occupied at first by lakes. In suitable
climatic conditions these hollows tend to be filled by peat, often in the
form of raised bog- as in parts of the Irish drumlin country. Here in
Edenside the shallow basins have generally been drained, whether
naturally by tributaries of the vigorous Eden, or artificially. Note
that the pattern of the minor streams is reminiscent, on a small scale,
of the theoretical pattern of consequent drainage in an area of Jura-
type folding: streams run along the hollows, parallel to the long axes
of the hills, passing from one hollow to another by way of the low
saddles between adjacent drumlins.

In the south-west, south-east, and north-east, the drift is thin or
absent. In the south-west the land is shown to rise above 950 ft. O.D. ;
the drumlins appear to be confined below the yoo-ft. contour. Observe
also the deep narrow valley of Waterhouses Beck, which with the
quarries marked at several points indicates solid rock. The symbol
used for the quarries is that for rock-drawing, not for earth-slopes
(cf. 702111, 712128). At the second of these a short spur of railway
linking the quarry with the main line indicates that the stone is worth
transporting over some distance, which would not be true of boulder
clay or ill-assorted gravel. At 701128 a resistant outcrop forms
Swathburn Crag. Small quarries in the extreme south-east, served by
field roads, 'seem likely to have provided the building stone for the
houses of Winton. In the north-east the valley wall rises sharply for as
much as 1,000 feet: it seems to be based on almost horizontally bedded
Carboniferous Limestone, for the general assemblage of features is very
similar to that found on Ingleborough (see Chapter VII) : scars, springs
near the foot of the slope, rare and discontinuous streams above, caves
(718189), sinks (for example Carry Pot, 770190), abundant scree or
more scattered debris (for example Alme Bank, 768186), and possibly
also limestone pavement (775188).

The steep edge was obviously in being before the drumlins were
formed, for post-glacial re-grading in the normal cycle has not greatly
changed the general aspect of the drumlin field and cannot therefore
have had much effect on the more resistant "solid" rocks. The physique

GLACIATED HIGHLAND AND A DRUMLIN FIELD 57

of the whole area, as interpreted from the 1/2,5000 map, might there-
fore be summarized in terms such as the following

A drumlin field, moulded by basal ice moving in a broad arc from
north to east, occupies the floor of a wide valley which slopes towards
the north-west. The valley is bounded on the north-eastern side by
a steep edge, developed across (probably) Carboniferous Limestone,
which also underlies at least part of the drumlin field. Karstic features
are developed on the limestone to a somewhat limited extent, although
it does not seem to have been thickly covered by glacial drift. The
drumlin field is in an early stage of dissection by normal drainage ;
the trunk stream, the Eden, is cutting down rapidly and shows signs of
developing a meander-belt, but although any small lakes have been
filled in or drained, the tributaries do not seem to have effected much
erosion.

NOTES AND REFERENCES

The descriptions of glaciated landscapes given in the texts listed at the
end of Chapter II may be amplified from

R. F. FLINT. Glacial and Pleistocene Geology. Wiley, New York,
1957-

The map used to illustrate glaciated highland in the first edition of
this book was the old Popular Sheet 47. Because the Lorn and Lochabcr
Tourist sheet omits part of the area represented on Sheet 47 (Popular),
while giving a greater total coverage, little use is made here of the
abundant signs of glaciation including breaches which occur in
the south.

The extent to which field-work modifies and amplifies conclusions
drawn from maps alone can be judged by comparing the foregoing
interpretation ot the Callop breach with

G. H. DURY. "A Glacial Breach in the Northwestern Highlands/'
Scot. Geogr. Mag., Ixix, 1953, p. 106.

Other accounts of glacial breaching include

G. H. DURY. "A Contribution to the Geomorphology of Donegal,"
Proc. Geol. Assoc., Ixx, 1959, p. i.

D. L. LINTON. "Some Scottish River Captures Re-examined/'
Scot. Geogr. Mag., Ixv, 1949, p. 123.

D. L. LINTON. "Some Scottish River Captures Re-examined, II."
Scot. Geogr. Mag., Ixvii, 1951, p. 31.

58 MAP INTERPRETATION

D. L. LINTON. Watershed Breaching by Ice in Scotland. Institute of
British Geographers, Publication No. 15. George Philip, London,
1951, p. i.

Drift spreads of lowland areas, and certain marginal features, are
described in

S. E. HOLLINGWORTH. "The Glaciation of Western Edenside."
Quart. Journ. Geol. Soc. y Ixxxvii, 1931, p. 281. (Includes a map and
discussion of the drumlin field of which part is shown on the selected
map.)

C. P. CHATWIN. East Anglia and Adjoining Areas. Second Edition.
British Regional Geology Series, H.M.S.O., 1948. (See Chapter VI,
p. 57 f)

V. WILSON. East Yorkshire and Lincolnshire. British Regional
Geology Series, H.M.S.O., 1948. (See Chapter VIII, p. 71 ff.)

The geology of the country represented on the selected maps is
outlined in

T. EASTWOOD. Northern England. Second Edition. British Regional
Geology Series, H.M.S.O., 1946.

H. H. READ. The Grampian Highlands. Second Edition (Revised
by A. G. MacGregor). British Regional Geology Series, H.M.S.O.,

1948.

J. PHEMISTER. Scotland: The Northern Highlands. Second Edition.
British Regional Geology Series, H.M.S.O., 1948.

For an illustrated account of the Parallel Roads of Glen Roy, see
Scotland: The Northern Highlands (above), p. 74 ff., Fig. 22, Plate IX.

The survival of unglaciated parts amid a generally glaciated high-
land is discussed by

D. L. LINTON. "Unglaciated Areas of Scandinavia and Great
Britain/' Irish Geography, ii, 1949, p. 25.

CHAPTER VII
CARBONIFEROUS LIMESTONE COUNTRY

Behold a labyrinth of hydraulic pipes. BROWNING
MAP: O.S. 1/25,000, SHEET SD/yy (INGLEBOROUGH)

IN tracts underlain by limestone, solution tends to be the most effective
agent of erosion. When the limestone is thick, soluble, and well-
jointed, but also mechanically strong, systems of cavities are hollowed
out which absorb not only the local precipitation but also the largest
streams which pass on to the limestone from adjacent impermeable
outcrops. The texture of the landscape and the system of drainage
come to differ so markedly from those observed where creep, surface
wash, and surface streams are at work that limestone country is looked
on as evolving in a special cycle, the karstic cycle. The name is
from the type Karst on the eastern side of the Adriatic.

Recognition of Limestone Country

It is important that a map interpreter should recognize country
which is so highly specialized, both as a type of physical landscape and
as a form of environment for human life. The most reliable evidence
is provided by the sinks whereby drainage passes underground.
Surface drainage is discontinuous, intermittent, or absent; dry valleys
and enclosed hollows with no surface outlet can also be looked for,
together with exposed surfaces of limestone more or less severely
weathered. All these are true karstic features, by some or all of which
limestone outcrops should be identified on the map. Lime kilns,
where they occur and are named, provide useful supporting evidence,
especially when quarries adjoin them, but unspecified quarries alone
are of little help except in proving resistant rock. Certain additional
features are widely observed on the Carboniferous Limestone of this
country lead mines, numerous prehistoric sites, and heathland
vegetation but these are not confined to the limestone and cannot
therefore be taken as diagnostic. They should be regarded rather as
expected occurrences on limestone outcrops which have been recog-
nized by their physical features.

60 MAP INTERPRETATION

Sinks

The essential unit of a karst landscape is the doline, a conical hollow
which may be hundreds of feet across. Comparable and related
features in this country may be grouped as sinks, a term applicable
throughout the range from true dolines to enclosed marshy basins
and minute depressions. Many sinks 1 appear to be connected with
more or less vertical shafts (the pitches of speleologists), which
develop at the intersection of major joint-planes, and are best dis-
played where they are kept open by a disappearing stieam. In the
Chalk country some groups of localized sinks and shafts are known,
but the rock appears to be too soft to support and preserve large
systems of cavities. It is moreover permeable, and therefore soluble,
in mass, so that much rainfall is disposed of by wholesale percolation
and dry valleys are more typical than sinks. Karst country proper is
well represented in the selected area, which is not based on Chalk, and
where the several varieties of sink may be observed on Ingleborough.
Hundreds more occur than are shown on the 1/25,000 sheet. Some of
the larger hollows are symbolized by hachuring, as, for example, in
the kilometre square 7272. A number of deep conical sinks are found
here, although the surface drainage is concentrated in nothing more
than rills. On the eastern flank of the mountain perennial streams are
swallowed, as Fell Beck at Gaping Gill (751727) and Alum Pot Beck
at Alum Pot (775756). In the absence of the hachure symbol a well-
marked terminal sink may be recorded in a place-name, for example,
Hodge Hole (735756), or by a note, for example, "Enters Pot Holes"
(759768), "Mouth of Cave" (762730). Sinks of this kind are rarely
blocked. The disappearing stream tends to cut through the lip of the
shaft, lowering its profile upstream with reference to this local base-
level. Thus the valley is deepened as far down as the sink but not
below it, while the sink itself is enlarged as the sides are weathered
back. The blind valley resulting ends in a steep enclosing wall.
Other streams dwindle gradually through small openings in their
channels, ceasing to flow above ground at a point which varies with
the season and which is determined by the combined effect of discharge
and the condition of the sinks upstream. Many of the smaller streams
mapped here disappear in this manner, so that one cannot with com-
plete certainty interpolate sinks at their lower ends.

1 Sinks are known in various localities as swallows, swallow-holes, swallets, sink-holes
pots, and potholes.

CARBONIFEROUS LIMESTONE COUNTRY

Nevertheless, when marked sinks and probable sinks are plotted
(Fig. 5) it is found that a systematic arrangement can be made out.
The higher, steep-sided part of Ingleborough discharges surface

^ > -1

^JZ-s^a^

sink \_* scar

spring | /? scar with screej . roc ^Y 9L_ un ^L * wq ^ er f G lL_

Odrumlinl Q

miles

FIG. 5. INTERPRETATION OF PHYSIQUE: INCLEBOROUGH AND ITS

SURROUNDINGS
(Based, by permission, on Sheet SD/77 of the O.S. 1/25,000 Map)

streams, which nearly all vanish at a line of sinks encircling the moun-
tain at a height varying between 1,200 and 1,500 ft. O.D. The stow
thus enclosed is one of impermeable rock and die line of sinks approxi-
mately marks die upper boundary of the limestone, beyond which
the waters circulate underground to reappear at very well-marked
lines of springs low down in the valleys. As none remains at the

62 MAP INTERPRETATION

surface, the drainage of the limestone tract is mature in the karstic
cycle. In fact, the stage is that of late maturity, for normal erosion
is re-asserting itself on the impermeable rock which underlies the
limestone and which is being revealed in the valley bottoms. The
junction with the limestone is marked by the spring-lines referred
to (Fig. 5).

Other Karstic Features

On the limestone outcrop as thus delineated, other features of a
karst landscape may be sought. They may be dealt with under two
heads: those related to the circulation of underground water, and those
due to weathering at the surface. Although the cavern systems and
their associated surface features connect the sinks and springs already
mentioned, it is more convenient to deal first with the features of
weathered limestone, which are well expressed on the map and are
well understood.

Between the spring-line and the line of sinks one can make out, in a
very rough way, a succession of surface features (Plate HA). On the
steep valley walls above the springs, individual resistant beds of lime-
stone have been thrown into relief by differential erosion, and stand
out as precipitous rocky scars (Fig. 5). Debris weathered off the scar-
face tends to accumulate in a bank of scree lower down the slope.
Some of the many sheets of rock-waste are symbolized on the 1/25,000
sheet and on the sketch-map. An interesting modification of the scar-
scree association occurs on Keld Bank (7477), below Bent Hill Wood
(7777) and elsewhere; although the scar has been weathered back into
the hillside and covered with soil and vegetation, a bank of scree
survives, running across the slope, to indicate the near-by presence of
a strong limestone bed. Scars are naturally best formed and maintained
on steep slopes where the debris of weathering can be most quickly
removed. On the flatter hilltops above, the bare limestone beds are
exposed in limestone pavements, or clints, shown on the O.S. map
by a variant form of rock-drawing. Clints are fields of joint-blocks,
each block separated from its neighbours by grikes, the fissures
opened by solution along joint-planes. Away from the bounding
scarps clints pass into rock-strewn ground with a partial soil cover,
distinguished on the O.S. map by the symbol for scattered boulders,
and that in turn into expanses of peat moor. The "Moss" names (for
example on the south-west flank of Ingleborough) and the various
bogs are significant of climate as well as of vegetation type.

CARBONIFEROUS LIMESTONE COUNTRY 63

Thickness of the Limestone

Now the spring-line shown in Fig. 5 is not far from horizontal.
It lies close to 850 ft. in the Greta valley, where the springs are re-
markably aligned, ranges vertically through c. 100 ft. in the south-
west, and is included between the 9OO-ft. and i,ooo-ft. contours in
Crummack Dale in the south-east. One may infer that the base of the
limestone is nearly flat and level. The attitude of the scars, which run
sub-horizontally across the hillsides, confirms this inference by showing
that individual beds of limestone dip very gently. It follows that the
vertical distance between the spring-line and the line of sinks may be
taken as a rough guide to the total thickness of the limestone. A
figure of the order of 500 ft. seems the most reasonable.

Underground Cavities

A limestone formation of this thickness, late mature in the karstic
cycle, is bound to include extensive systems of cavities through which
water passes from the sinks to the springs ; but, on a map, information
is usually restricted to the names of a few well-known caves ; it is rare
for the outlines to be shown, as they are here at Ingleborough Cave
(7571). Without such guidance, the interpreter should refrain from
explaining certain features by the collapse of caverns, as he may at
first be tempted to do. The need for this caution arises from the
occurrence in a karst landscape of enclosed hollows, not only the
small sink and the larger doline, but in addition the uvala and the
polje. The polje, unknown in this country, 1 may be 100 square miles
or more in area and is of tectonic origin. The uvala (or valley sink) is
a basin of intermediate size, formed by the merging of two or more
dolines. The feature is not common in Britain, but is not unknown.
Now the underground cavity systems include three elements: the
joint-plane cave previously mentioned, the gallery (passage, or, in
gently-dipping strata, bedding-plane cave), and die large cavern or
hall which is long, wide, and deep. There must be a limiting size,
above which the roof of a cavity can no longer support its own weight.
Collapse of large bedding-plane caves would cause shallow depressions
at the surface, like the subsidence-hollows of coalfields; the unroofing
of halls would produce uvalas and gorges; more localized collapse at
a shaft might be reflected in a deep doline. Investigations show,
however, that the upper parts of cave systems tend to become choked
Unless possibly at Lane End ,Bucks.

64 MAP INTERPRETATION

by detritus and stalagmite, rather than to be enlarged still more by
solution, flaking, and corrasion. Hence in map interpretation, gorges
and enclosed hollows should be noted merely in descriptive terms.
Trow Gill (7571) on the present sheet (see also Fig. 5) is known to
have been cut, at least in part, by a surface stream which flowed during
the Pleistocene. Similarly, the common dry valleys of a karst may have
large sinks in the bottom, as has that between Gaping Gill and Trow
Gill, but it cannot be assumed that the former surface stream follows
its old line with the sole difference of flowing at a lower level and
underground. The dry limestone valleys of Britain should be regarded
generally as carved by surface streams, which have been captured to
the underground system, and which as a result may find an outlet in
entirely separate valleys at some distance away. The dry valley
symbols in Fig. 5 must therefore not be taken as suggesting under-
ground courses: they merely show features which on the 1/25,000
sheet are expressed by contours.

A number of the springs at the base of the limestone are likely to
fluctuate greatly in volume, in response to local rainfall, siphoning,
and flow under pressure, but unless the map gives descriptive notes
the interpreter can do little more than remark the issue of considerable
streams, for example, at Clapham Beck Head (754711).

Rocks Above and Below the Limestone

Limestone formations in Britain, of the order of thickness estimated
here and* with pronounced karstic features developed on them, may be
confidently regarded as Carboniferous Limestone. Of the underlying
impermeable rocks little can be said, except that they are older than
Carboniferous. The name "Ingleton Granite Quarries" (7175) is
misleading, including a quarryman's term, not a geologist's. Above
the Carboniferous Limestone in the higher part of Ingleborough are
younger strata, for the most part impermeable since there are many
surface streams, but including some permeable beds to account for the
occasional sinks. It is likely that the permeable rocks too are limestone,
and that they explain the few scars in addition to the well-defined
springs.

Evidence of Glaciation

In the north-east of the area mapped is part of a drumlin field. The
discussion of Chapter VI need not be amplified here : it is necessary
only to note that the arrangement of many drumlins indicates a

CARBONIFEROUS LIMESTONE COUNTRY 65

southward movement of ground ice down the Ribble valley. Sinks
in the grid squares 7974 and 7975 prove that the drift cover is in places
thin. Apart from the drumlins, the map affords no indisputable
evidence of glaciation. The straightness and trough-like form of
Kingsdale and the Greta valley might well be structurally determined
rather than glacially shaped. One cannot judge how widely the hill
peat is developed on boulder clay which, as it were, waterproofs the
limestone. The fact is that the whole area has been powerfully
glaciated, but does not lend itself readily to the development and
preservation of glacial landforms. In the field the most striking
evidence consists in the numerous erratic boulders of grit scattered
over the clints: they are unrepresented on the map.

River Development

On a map of this size and scale, the area represented (less than forty
square miles) is too small to afford the wide synoptic view required in
the interpretation of drainage and drainage history. Because of this
limitation, and because the landscape is complicated by karstic,
glacial, and structural features, certain important results of normal
erosion are obscured. The series of falls on the Greta, like the force
on Jenkin Beck (7173), are in reality located at knickpoints, the head-
ward limits of waves of rejuvenation. Another probable knick
belonging to an earlier cycle occurs on the Greta at Weathercote
(7377)- It has been shown by Sweeting that the great mountain
shoulders at c. 1,300 ft. O.D. are to be regarded as erosional rather
than structural: the close coincidence of an erosional platform and a
structural surface is mainly accidental. None of these facts can be
perceived on the map: they are mentioned here as a reminder that
interpretation, however careful and detailed, is subject to inevitable
limitations by comparison with work on the ground.

NOTES AND REFERENCES

For a general account of the geology, see

D. A. WRAY. The Pennine* and Adjacent Areas. Second Edition.
British Regional Geology Series, H.M.S.O., 1948.

The concept of the karstic cycle is briefly summarized in

E. M. SANDERS. "The Cycle of Erosion in a Karst Region (after
Cvijic)." Geogr. Review, xi, 1921, p. 593.

66 MAP INTERPRETATION

An illuminating morphological study of the whole district in which
Ingleborough occurs is

M. M. SWEETING. "Erosion Cycles and Limestone Caverns in the
Ingleborough District." Geogr. Joitrn., cxv, 1950, p. 63.

Other relevant material is to be found in

R. G. S. HUDSON et al "The Geology of the Yorkshire Dales."
Proc. GeoL Assoc., xliv, 1933, p. 227. (I: 'The Scenery and Geology
of North-west Yorkshire," by R. G. S. HUDSON; III: "Alum Pot,"
by H. W. HAYWOOD; IV: "The Glacial and Post-glacial periods in
West Yorkshire," by A. RAISTRICK.)

A very detailed study of Ingleborough was published under the
name of T. McKENNY HUGHES in Proc. Yorks. GeoL Soc. "Physical
Geography" appeared in 1901, "Geology" in 1905-8 inclusive.

Current research and exploration, illustrated by detailed maps, are
described in the periodicals Cave Science and Transactions of the Cave
Research Group.

Papers describing the modifications of karstic phenomena observed
in the Chalk country are listed at the end of Chapter IV.

CHAPTER VIII
COASTS AND SHORELINES

What's it now?

Changed like a rock-flat, rough with rusty weed,
At first wash-over o the returning wave. BROWNING

MAPS: O.S. 1/63,360 (SEVENTH SERIES) SHEETS 114 (BOSTON
AND SKEGNESS), 186 (BODMIN AND LAUNCESTON); O.S.
TOURIST MAP, 1/63,360 (LORN AND LOCHABER); O.S.
IRELAND, 1/63,360, SHEET 169, ETC. (WEXFORD); O.S.
1/25,000, SHEETS NC/y6 (SUTHERLAND), SY/i8 (SIDMOUTH)

THE concept of a cycle of erosion is no less useful in the study of
shorelines than in the study of landscapes. Youth, maturity, and old
age are marked by distinctive assemblages of forms. In some ways
maps of shorelines are easier to interpret than those of inland areas,
for many shoreline features are very simple to represent on a map, and
the principal changes in the course of the shoreline cycle are changes of
plan. Moreover, shorelines become more and more regular as maturity
approaches, whereas landscapes become progressively diversified.
On the other hand, while it is often unnecessary or impossible to
consider the effect on inland areas of changes of base-level, whatever
they may have been, evidence of such changes is critical in the inter-
pretation of shorelines. Away from the coast, the essential fact is that
the land lies above base-level and is subject to subaerial processes; at
the shore, neither the direction nor the amount of strand-line move-
ments can be ignored.

The Shoreline Cycle

Any considerable displacement of the strand-line initiates a new
cycle of shoreline erosion on the drowned landscape or the exposed
sea floor. Hence the customary distinction between shorelines of
submergence and shorelines of emergence. A third class, neutral
shorelines, includes the shorelines of deltas, volcanoes, and fault-
blocks, where neither submergence nor emergence is required to
account for the initial forms. In nature, compound shorelines are

MAP INTERPRETATION

widespread. They combine elements essential to any two, or all three,
of the foregoing classes. Shorelines with both emergent and submer-
gent features are especially common.

It must be understood that the terms "submergence" and "emer-
gence" mean nothing more than movements of the strandline,
upwards positive, and downwards negative, respectively. No-
thing is implied as to causes. Examples of local or regional uplift,
depression, or tilting, are well known from certain unstable parts of
the earth's crust, -but there is also powerful support for the view that
world- wide rises and falls of sea-level have taken place. These general
or eustatic movements of the strand-line appear to have been going
on at least since early Pliocene times. Some must be due to changes
in the capacity of the ocean basins, but others reflect the loss or gain
of water resulting from the growth or decay of the great Pleistocene
ice-sheets. It is easy to see that, quite apart from crustal deformation,
the record of strand-line movements on a given stretch of coast may
be lengthy and involved. Hence the wide distribution of compound
shore-lines, and the need to employ non-committal descriptive terms
in speaking of changes in base-level.

Besides the direction and extent of the strand-line movement with
which the new shoreline cycle begins, it is necessary also to consider

TYPE OF
STRAND-LINE
MOVEMENT

INITIAL RELIEF

HIGHLAND

LOWLAND

NEGATIVE
EMERGENCE

Emergent
highland
shoreline

Emergent
lowland
shoreline

POSITIVE
SUBMERGENCE

Submergent
highland
shoreline

Submergent
lowland
shoreline

the form of the emergent or submergent land. As far as the stage of
maturity, the shore-line cycle takes a different course on highland and
lowland coasts respectively. Some accounts over-simplify matters by
assuming or implying that emergent and lowland coasts on the one
side, and submergent and highland coasts on the other, are synony-
mous. This is frequently, but not universally, true. A eustatic rise of

COASTS AND SHORELINES 69

sea-level must obviously submerge lowland as well as highland, while in
places the bottom topography is far too rugged to constitute a lowland
if it emerged. Consequently the initial shorelines produced by strand-
line movements must be cross-classified in the manner illustrated in
the table on p. 68.

Compound shorelines combining emergent and submergent features
may be appropriately described in such terms as "compound highland
shoreline, dominantly submergent."

The Shoreline and Subaerial Cycles

Now a map interpreter will rarely confine his attention to the
shoreline the immediate zone of contact between land and water
but will need to treat the coast as a whole. In this connection the direc-
tion of strand-line movement is more than ever important, for a rise
of base-level influences the development of rivers and valleys differ-
ently from a fall. In tliis wider view, it is necessary to pay attention
not only to raised shoreline features, including perhaps extensive
wave-cut platforms, but also the breaks of river profile and to valley-
in- valley forms ; but, although a given strand-line movement may be
looked on as initiating a new cycle both on land and at the shore, it is
usually convenient to consider the landscape and shoreline individually,
for several reasons. Even if the two cycles begin simultaneously, it
does not follow that they will keep in phase after the initial stage.
Furthermore, the intermittent movements of emergence abundantly
proved in many areas have been relatively small and frequent, initiating
epicycles rather than completely new cycles of subacrial erosion.
Consequently a landscape and its limiting shoreline are commonly
found to be in very different stages of their respective cycles, a fact
wliich must be recognized in description. The physical characteristics
of a given piece of coast may be summarized under the following
heads

1. Class of shoreline submergent, emergent, neutral, or
compound.

2. Type of coast highland or lowland.

3. Stage reached in the shoreline cycle.

4. Form of subacrial cycle normal, karstic, glacial, or arid
and stage reached therein.

The areas discussed below can be no more than a selection of the
combinations possible. They are, however, intended to illustrate a

7O MAP INTERPRETATION

wide variety of coastal forms, and to exemplify the main features
developed in the shoreline cycle up to and including the stage of
maturity, with particular reference to highland coasts in different
stages of the subaerial cycle.

HIGHLAND COAST WITH FEATURES OF EMERGENCE. EARLY
YOUTH OF THE SHORELINE CYCLE. LANDSCAPE YOUTHFUL
IN THE NORMAL CYCLE AFTER DEGLACIATION
O.S. 1/25,000, SHEET NC/y6

This portion of the coast of Sutherland displays a remarkable combina-
tion of features. Although no submarine contours are given, the
amplitude of subaerial relief and the forms produced on the shore
indicate that the highland variety of the shoreline cycle is in progress.
Numerous lakes on the low plateau seem most likely to be the relics of
glaciation, which has also left its mark in the steep-sided flat-bottomed
trough glimpsed in the extreme south-west. It is not surprising,
therefore, to discover signs of re-grading in the new, post-glacial,
normal cycle, as at 738622, 783620, and 785629, where streams are
sawing through rock-steps or through the lips of hanging valleys.
Some additional factor may be needed to explain the well-marked
gorge in the east (798630), to which another in the south-east may
be tributary, but without more evidence this problem cannot be
explored.

One soon finds, however, that not all active valley-deepening can be
explained as the re-grading of glacial features. Several tiny valleys
hang nqjtably above the shore, discharging their streams in cascades
down the cliff. At 752658 and 753659 the "hang" is more than
100 ft. At some points the stream, in commencing to accommodate
itself to present base-level, has deeply notched the cliff-face, for
instance at 774655; but inspection of the contours proves that the
work is still in its early stage and that the long-profile still plunges
seaward. A number of these ungraded streams appear too long to
have developed headwards from the present shore: they are some-
thing more than rills. Consequently an explanation must be found
for the steep gradients near their mouths. The leading possibilities are
cliff recession and rejuvenation by emergence. To cause a stream to
hang or plunge in this way, cliff recession must be rapid and extensive.
But, as will shortly be noted, the forms of this shoreline indicate very
early youth and resistant rocks, whence it may be inferred that in the

COASTS AND SHORELINES ?I

present shoreline cycle the cliffs have receded very little. The alter-
native explanation, late emergence, is amply supported by the available
evidence.

At the head of Armadale Bay (7964) a low cliff is marked (in
hachuring) above HWMOST, backed by a gentle ascent up to 100 ft. or
thereabouts, after which there is a sharp rise. Near Kirktomy (74638)
a flat spur rises a little over 100 ft. O.D., and closed loo-ft. contours
occur at 735635 and 708625 in neighbouring valleys. The steep rising
slopes behind Farr Bay (7162) commence at about the same level of
100 ft. One may suggest subject to the test of work on the ground
that the features noticed are the remains of former cliffs and abrasion-
platforms, mostly referable to a base-level of c. 100 ft. above the
present one. If so, the low cliff in Armadale Bay appears to record a
still-stand at an intermediate level during emergence. Since heavy
glaciation would have been likely to destroy or to conceal evidence of
tliis kind, the inferred high base-levels are probably later than local
deglaciation.

These features of relatively late emergence of a highland coast are
associated with the shoreline forms of very early youth. The shoreline
as a whole is crenulate, i.e. minutely irregular. Besides the bold
irregularities of bay and headland there are very many small clefts
and points, etched out by wave-erosion working selectively along
lines of weakness. At this stage the details of shoreline form are to a
considerable extent structure-guided, as is well shown in the group
of parallel clefts just east of Port Mor (7363). As yet bay-head beaches
are little developed: powerful waves enter nearly every inlet and
maintain steep cliffs along most of the shoreline.

SUBMERGENT COAST OF SUBDUED RELIEF. HIGHLAND
SHORELINE, APPROACHING SUB-MATURITY. LANDSCAPE

POST-MATURE

O.S. IRELAND, 1/63,360, SHEET 169 (SHEETS 169, 170, 180, 181)
(WEXFORD)

Submergence is likely to produce embayed shorelines. Within
this group a very wide range of initial forms is possible. In schematic
outlines of the shoreline cycle on a submergent highland coast, it is as
a rule tacitly assumed that the landscape is approximately mature in
the subaerial cycle, so that the former valleys constitute deep bays
while the intervening divides stand out prominently in headlands.

72 MAP INTERPRETATION

In these circumstances, wave-attack on the promontories gives rise to
cliffs. The derived rock-waste combines with the load discharged by
rivers to provide material which is disposed as beaches, spits, and bars.
The chief process at work in cliff-cutting, in rolling and abrading the
beach material, and in moving it along the shore, is wave-action.
Despite the multitude of individual forms, both destructional and
constructional, one may observe a general tendency for beaches to
accumulate at the heads of bays and for headlands to be linked to one
another by coalescent spits. By the time the shoreline is sub-mature,
much of the headlands has been destroyed; the surviving bays are
almost cut off from the open sea by bars, which with the headland
beaches form a straight or gently curving, regular line. Maturity is
attained when the line of beach has been pushed back to the mainland
throughout its whole length, i.e. when the cliff line has receded as far
as the bayheads.

Although the selected area is not one of very strong relief, the shore-
line cycle is that appropriate to highland coasts. Submergence is
indicated by the drowned valleys : Bannow Bay and Wexford Harbour
are identical in kind with Tacumshin Lake and Lady's Island Lake.
Perhaps it is not out of place to insist that, although the outlines of
bays are smoothed out by shoreline processes, the actual inlets are
usually the mark of submergence. In other words, most of the excava-
tion has been done by subaerial not marine agents. The fact that a bay
coincides with an outcrop of weak rock does not prove that the inlet
has been carved out by the sea. Observation shows that shoreline
processes tend to regularize, not to diversify, the outline. Initially this
particular shoreline must have been deeply embayed, but a great deal
of regularization has already taken place and sub-maturity has been
reached in places. The different stretches may be taken one by one,
in order of advancement in the shoreline cycle.

Bannow Bay seems to have been a fairly deep, steep-sided valley,
cut by a river working to a lower base-level than the present. Sub-
mergence must have converted it into a deep-water inlet of the ria
type, with a bottom gradient determined by the form of the old
valley floor. Doubtless the bay originally deepened seaward, but the
old floor now lies deeply buried under rock-waste that has accumulated
since the submergence, filling the inlet up to mid-tide level and re-
placing deep water by tidal flats. It is often urged that tidal scour
militates against the silting-up of inlets of this kind, but tides can
sweep material in as well as out. Observation teaches that these

COASTS AND SHORELINES 73

inlets usually tend to become choked, and the limited investigations so
far made show that much of the fill consists of marine rather than
fluvial deposits. In some measure the silt is protected by the bar
typically developed across the mouth of the inlet. There is a difference
between the offshore bar, formed under water and driven landward,
and the spit which grows out from a headland. Here the channel on
the western side of Bannow Island is about two-thirds closed by a
large spit projecting north-eastwards from the opposite shore; a
mile to the southward, the bar at the head of Fethard Bay almost
completely encloses a small drowned valley.

Between Crossfarnoge and Carnsore Point the cycle is more
advanced. No powerful streams enter the sea ; bay mouth bars almost
seal off the two inlets. The forms are those of sub-maturity a smooth
line of beach passing from one cliffcd headland to another, with
portions of the initial inlets surviving. Note that spits appear to have
grown most vigorously from the central headland, which is now
winged, both to cast and west a fact which should prevent any
facile generalization on the dominant direction of longshore drift.
Although it may be possible to show that, on a long stretch of
shoreline, there is a net transport of beach material in a single
direction, local exceptions are to be expected if the shoreline is at
all irregular (cf. the opposed directions of the spits in Bannow and
Fethard Bays).

The shoreline of Ballyteige Bay is not far from mature. It may be
assumed that the headlands on either side, now \vell cliffed, once
projected farther, and that they have receded considerably under
wave-attack. The great dune-covered bar (Ballyteige Burrows) is
the counterpart of the baymouth bars previously noticed, but has
been driven inland as the anchoring headlands were destroyed until
it rests almost against the mainland. The sizeable bay of submergence
behind it is now represented by no more than a narrow valley bottom
with diminutive tidal flats. A further recession of less than a mile
would involve the burial of these flats, the destruction of the bar, and
the development of a smooth line of cliffs on the firm ground. This
stretch of shoreline would then be fully mature.

It seems as if the large bar has grown westwards from Crossfarnoge,
progressively displacing the mouth of the stream. Where records
exist, however, it is remarked that such growth and displacement have
usually been intermittent and subject to relapse, even without human
interference. The map can give no more than a statement of existing

74 MAP INTERPRETATION

conditions, showing effective growth and ignoring historical details,
unless a given spit or bar is obviously complex or compound. Com-
plex spits are those which send out branches during growth; com-
pound spits include elements of more than one earlier feature;
recurved (hooked) spits, for instance, not infrequently have "teeth"
on the landward side, at former positions of the terminal hook. On
the present map the record is somewhat obscured by dunes, which
incidentally prove that the spits are at least partly composed of sand,
and by sea walls at the edges of the tidal flats. The long spit terminating
in Rosslare Point is possibly compound, but the irregular inner edge
may also be due to tidal deltas (washovers), where beach material has
been swilled inwards at high tide. Small spits which would be more
easily understood with the help of bottom contours are St. Patrick's
Bridge near Crossfarnoge, St. Patrick's Bridge on Saltee Island Little,
and The Ring on Saltee Island Great.

The eastern shoreline remains to be considered in its entirety. The
low cliffs alternating with shallow bays between Carnsore and
Greenore Points call for little remark. About a mile beyond Rosslare
Harbour Pier a smoothly curved shoreline begins, based in the south
on the mainland but farther north on the spit already mentioned. It is
clear that the shoreline is being developed to some extent independ-
ently of the subaerial forms, a fact which emphasizes the utility of
dealing separately with the shoreline and with the landscape. The spit,
already protecting a great width of tidal flat from wave-attack, seems
likely o unite with The Raven to the north in a cuspate bar, separating
Wexford Harbour from the open sea. On this view, the shoreline
from Rosslare Harbour northwards is roughly sub-mature, but if it
were a Igwland shoreline the interpretation would be different (see
below, p. 77 ff ).

HIGHLAND COAST. SHORELINE MATURE IN THE PRESENT

CYCLE. LANDSCAPE YOUTHFUL
O.S. 1/25,000, SHEET SY/i8 (SIDMOUTH)

This part of the shoreline of South Devon is in reality compound.
The River Sid has a buried channel related to a lower base-level,
while raised abrasion-platforms in the vicinity show where the sea
formerly stood higher against the land; but maturity in the present
shoreline cycle is so nearly attained that for the purposes of this
discussion strand-line movements may be ignored, the shoreline being

COASTS AND SHORELINES 75

regarded as one of submergence, exemplifying a later stage in evolu-
tion than that just discussed.

There are no bays. Any headlands that may formerly have existed
have been cut back to a smooth line of cliffs, broken only at the valley
mouths and continuously fringed by beach (Plate HB). Such regularity,
and the abrupt truncation of the landscape are typical of shoreline
maturity. There is no doubt that neither geological structure nor the
texture of the subaerial relief has much effect on the plan of the shore,
nor do waves enter and enlarge the openings of river valleys.

The area seems to be one of uniclinal structure, with a more resistant
cap-rock overlying weaker beds. Note at the top of each valley wall
a steep descent through about a hundred feet, with a gentler slope
below. 1 Wave-attack is vigorous and the cliffs are receding rapidly.
Small streams hang above the shore, or plunge downward to their
mouths, having been unable to lower their profiles as fast as the cliffs
have been cut back. The discordant relationship of minor drainage to
present base-level is similar to that observed previously on the Suther-
land shoreline, but the cause is different. In Sutherland other evidence
supported the hypothesis of emergence; here, rapid and general
retreat of the cliffs is indicated not merely by the mature form of the
shore, but also by the partial destruction of hilltop camps at High
Peak Hill (104859) and Berry Cliff (188882). It seems probable that
marine erosion has made a great deal of progress in the last two
thousand years or so. The whole cliff-face, not the lower part alone,
is being brought down. Landslip symbols, particularly clear on Berry
Cliff, show how the weak rocks respond to undercutting.

HIGHLAND COAST, DOMINANTLY EMERGENT BUT ALSO
SHOWING FEATURES OF LATE SUBMERGENCE. COMPOUND
SHORELINE, YOUTHFUL IN THE PRESENT CYCLE. POLY-
CYCLIC LANDSCAPE, YOUTHFUL IN THE PRESENT CYCLE
O.S. 1/63,360 (SEVENTH SERIES) SHEET 186
(BODMIN AND LAUNCESTON^

The maps so far treated in this chapter represent many of die larger
features typical of submergent highland coasts, at various stages up to
and including maturity of the shoreline cycle. Stress has been laid on

1 This inference is correct as far as it goes, but a geological map would reveal additional
facts of great interest: the cap-rock rests unconfonnably on the weak formation; the flat
summit-surface, although apparently structural at first sight, truncates the resistant beds
at a very low angle, and is thought to be a plane of marine abrasion.

76 MAP INTERPRETATION

shoreline rather than on subaerial forms. It is now proposed to take
a more complicated example, where an upland in early youth of the
subaerial cycle has been affected by oscillations of the strand-line,
including a late positive movement, and where the present shoreline
is still immature.

The initial shoreline of submergence was very different from that
of a submerged mature landscape. Instead of the ample bays, corre-
sponding to wide and deep valleys, one observes long, narrow, branch-
ing inlets: this is. a ria coast. The partially submerged valleys of the
Fowey, Looe, and Lynher, comparable to Barrow Bay in County
Wexford, are indisputable criteria of submergence. Wave-erosion
cannot have shaped them. They inherit their form from the young
valleys of the subaerial cycle in progress before submergence. As
expected in a youthful landscape, the narrow valleys are separated by
broad divides, on whose flat tops part of the initial surface survives.
These divides end seaward in rather irregular lines of steep cliff', which
correspond to the cliffed noses of the headlands which result from the
drowning of a mature landscape.

Additional evidence of a rise inbase-levcl is given by the submerged
forest (2150) whose trees must have flourished above the sea-level of
the time. It is possible also that th e floor of St. Austell Bay is a drowned
bayhead beach; but the drowned valleys remain the most easily
interpreted sign of submergence, even in the field, as shoreline features
developed at lower levels than the present are necessarily under water.
Evidence of emergence, in contrast, is often unmistakable on the
ground: caves, arches, stacks, wave-cut notches, patches of abrasion-
platform and of raised beach may be located, all out of reach of the
highest waves.

On the map, however, such comparatively small features do not
show well, if at all, and arc rarely labelled. Hence emergence has to be
proved from inland areas. It is unnecessary to recapitulate die argu-
ments of Chapter V, where it was seen that in this area strongly
rejuvenated drainage is associated with a scries of probable erosional
platforms, the whole indicating intermittent falls of base-level. The
interpreter would do well to note that except for the presumptive
evidence of a former low base-level afforded by the rias, shoreline
forms as represented on maps are best able to indicate submergence.
Away from the shore, however, emergence may be clearly recorded
in the erosional forms of valleys while submergence is concealed by an
alluvial fill which may not be apparent from the topographical sheet.

COASTS AND SHORELINES 77

The shorelines mapped here are by no means mature in the present
cycle, but the work of regularization has begun. Bayhead beach is
slightly developed in Mevagissey and St. Austell Bays, while the rias
are being filled in; the mouth of the Seaton River (3054) is indeed
already choked by alluvium behind a small enclosing spit. It seems
likely that the shoreline will attain a smooth plan long before the cliff
line has been cut back to the heads of the rias; the inlets will disappear
mainly by being silted up, until the tidal flats are replaced by flood-
plain, i.e. until shoreline forms are locally replaced by subaerial forms.
The shoreline cycle is most advanced in the south-east. East of
Downend Point (2251) cliff recession from a shelving shore has left an
abrasion-platform, which passes eastwards in Whitcsands Bay into a
gently arcuate beach. The line of cliff on this stretch is approaching
the regular plan of maturity.

LOWLAND COAST. COMPOUND SHORELINE, DOMINANTLY

SUBMERGENT, IMMATURE

O.S. 1/63,360 (SEVENTH SERIES) SHEET 114

(BOSTON AND SKEGNESS)

The initial outline of a lowland coast is likely to be highly irregular or
generally smooth, according to whether the cycle begins with the
submergence of a land-surface or with the emergence of a flat sea-
floor. As strand-line movements at various times in the late Pliocene,
the Pleistocene, and the post-Pleistocene have been widely demon-
strated in Britain, it is probable that most immature lowland shorelines
will be compound. It is best to remain content with the fact that the
lowland shoreline cycle is in progress, and to ascertain die stage
reached with present base-level.

Whether emergent or submergent, the lowland shore has a bottom
gradient so gentle that large waves break offshore. In doing so they
throw up a bar, submarine at first, but appearing above water at the
end of the initial stage. The bar grows by accretion as more debris is
thrown up, but at the same time time tends to move inshore as the
seaward face is combed down and fresh material is washed over the
top towards the mainland. During the stage of youth, the lagoon
enclosed between the bar and the mainland becomes filled with rock
waste and organic debris silt and peat over which the bar is driven
inwards. Maturity is reached when the bar has been pushed back to
the mainland, which is itself attacked. Conditions at this stage are

78 MAP INTERPRETATION

generally similar to those at maturity of the highland shoreline cycle,
except that on the mature lowland shore the cliffs can be of no great
height.

It is of course possible for the gentle bottom gradient required for
evolution of the lowland-shore type to exist locally off a highland
coast, for example, where slight emergence brought the floor of a bay
within reach of the breaking waves. For this reason care is necessary
in interpreting such features as those of Wexford Harbour, where the
shoreline is knowji to be compound. On the present map, however,
the very wide tidal flats in the Wash leave no doubt that the lowland
shoreline cycle is being pursued.

In Britain generally, lowland shores have been greatly modified by
reclamation and drainage. The flat fenlands here, and the comparable
Somerset Levels, represent former lagoon and marsh. The remarkably
wide bar, easily distinguished from the fen by its contrasted pattern of
settlement and roads, is the compound product of several strand-line
movements, which cannot however be inferred from the map. This
shoreline will not be mature until the bar has receded across the fen,
to rest against the firm ground which commences near the 5O-ft.
contour. One may doubt whether recession is taking place at the
present time, for south of Gibraltar Point (5557) a strip of salt marsh has
been reclaimed on the seaward flank of the bar (see also Chapter XII),
and the work is still going on. North of the Point conditions are
different. A much narrower foreshore corresponds to a steeper bottom
gradient. No salt marsh was available for reclamation here, for the
shoreline runs along the edge of a sandy bar which is, in places at least,
under wave-attack. There is nothing to say whether the "-ness" of
Skegness was an earlier tip of the bar, but one may certainly note the
apparent southward diversion of the Steeping River (5460-5567) and
of the drainage now carried in the artificial Cow Bank Drain (5560).
The groynes near Chapel St. Leonards indicate a longshore movement
of beach material as well as the likelihood of marine erosion, while the
outlines of Chapel Point and Ingoldmells Point suggest that the move-
ment is towards the south. It seems justifiable to infer that the seaward
face of the bar tends to be combed down, and that the material trans-
ported along the beach accumulates at the tip of Gibraltar Point,
extending it southwards.

In spite of the redistribution of material on the outer side, great bars
of this kind appear to be, as features, fairly stable. Inshore movement
is slow, often perhaps because of artificial defensive works. Some

COASTS AND SHORELINES 79

offshore bars of shingle, or banks of sand and mud, are relatively fast-
moving, changing noticeably in their low-tide outlines between
successive surveys. It is very difficult to generalize about their develop-
ment. In map interpretation it is advisable merely to note them as
indications of shallow water and of the form of shoreline cycle to be
expected.

FIORD COAST
O.S. TOURIST MAP, 1/63,360 (LORN AND LOCHABER)

Fiords are glacier troughs invaded by the sea. They are frequently
compared to the rias of some unglaciated areas, presumably because
they indent the shoreline and allow the sea to penetrate deeply inland,
but the comparison needs to be qualified. Rias necessarily imply
submergence, fiords do not. Rias are river valleys, excavated when
base-level was lower and drowned by a positive strand-line movement :
fiords are glacier channels, occupied by the sea when the ice melted.
Because of their great size glaciers can corrade well below sea-level.
Now as many ria coasts are thought to have remained stable un-
affected by earth-movement since late Tertiary times, the sub-
mergence which has affected them would seem to have been eustatic,
the manifestation of a general change of base-level. Some heavily
glaciated areas, on the other hand, responded to the massive load of
Pleistocene ice-caps by regional subsidence from which they are still
intermittently recovering. Movements of this kind, due to loading or
unloading, are termed isostatic. It will readily be appreciated that,
although submergence is not an essential factor in the origin of fiords,
many fiord coasts bear the impress of positive movements as well as
displaying the more noticeable signs of emergence. Hence in addition
to noting those features which derive from glaciation, or from the
beginnings of re-grading in the normal cycle, one must pay close
attention to the shoreline in the attempt to discover evidence of recent
changes in base-level.

In many glacier troughs the "solid" rock is deeply buried under
moraine and alluvium, which conceal irregularities in the floor; but
in ribbon lakes and fiords deposition is concentrated at the mouths of
rivers, with the result that enclosed hollows survive to be recorded
in the underwater contours. On the Scottish O.S. sheets bottom
contours, given in feet for all freshwater lochs except where the level
has been raised by damming, clearly indicate the forms present. In

8O MAP INTERPRETATION

that part of Loch Shiel mapped on Sheet 47 there are three closed
contours at depths of 400, 400, and 300 ft. Similar subsidiary basins
occur in Lochs Arkaig and Lochy in the north, and in Loch Eildc Mor
in the south-east. Some of these hollows reflect the presence of rock-
steps, but each basin as a whole owes much of its depth to the "down-
at-heel" erosion characteristic of valley glaciers. In the fiords submarine
contours, drawn at 5 and 10 fathoms only, can often record no more
than the shallowing towards the lower end: in Loch Eil both are
closed ; in upper Loch Linnhe both converge at, but pass through, the
Corran Narrows (0263) ; the floor of Loch Leven is better represented,
with the lo-fathom line closed at least four times and with two sets of
narrows less than 5 fathoms deep.

Just as the sides of a glaciated trough plunge below an alluvial fill
or beneath the surface of a ribbon lake, so they descend sharply into
the waters of a fiord. Hanging tributary valleys and deltas at the fiord
heads or at the mouths of lateral tributaries bear the same relationship
to the main trough as do comparable features inland, which have
already been reviewed. It is not to be expected that in the sheltered
waters of fiords marine erosion will have advanced far since dcglacia-
tion, so that wave-cut cliffs are unlikely to be extensive. The influence
of communication with the sea is found in the features of occupance
(discussed elsewhere) and in the beaches.

All these sea-lochs are fringed between tide-marks by a narrow
strip of foreshore, very narrow where the bottom plunges but wider
where the larger deltas provide gentle gradients. In the freshwater
basins, beach development, if any, is small. Above the present sea
beaches, close examination of the fiord shorelines discloses a narrow
bench at a slightly higher level (Plate IB). Where it is best developed
this feature might well be interpreted at first sight as a delta flat, an
accumulation of moraine, or as the enclosing rock at the lower end of a
glacier trough. The small flat at North Ballachulish (0560), with a
spot height of 46 ft., could be based on the rocky or morainic barrier of
lower Loch Lcven; the low ground projecting eastward at the Corran
Narrows might be similarly related to the basin of upper Loch Linnhe;
the narrows at the lower end of Loch Eil seem to combine the terminal
shallowing of a glacier trough with the delta of a heavily laden
tributary. At other points, however, explanations of this kind cannot
apply. Opposite Fort William at Trislaig (0974) a small flat appears
between the edge of the water and the line of buildings. The fiord
appears to be deep hereabouts, and receives no delta-building stream.

COASTS AND SHORELINES 8l

Much of the ground up to 50 ft. is obscured by the road symbol,
but hints of comparable sites appear north of the Klachnish delta
(0669) and along the southern shore of Loch Eil. Indeed, the course of
the coast roads, and of the railway on the north of Loch Eil, itself
suggests that low ground is available. On the south side of Loch Leven
the 50-ft. contour runs along the water's edge, indicating a continuous
slope, and the main road climbs the valley wall. Elsewhere careful
inspection discovers that the so-ft. and zero contours are more widely
separated than those above. In fact, the coast roads skirting the water's
edge run along raised beach, except when crossing streams and deltas,
or passing over rocky knobs.

The interpretation is not easy on this map scale, but a careful inter-
preter might expect to locate the flat at Trislaig and to consider it in
relation to possible isostatic movement. Much of the rest follows.
Account may also be taken of spot-heights on the coast roads, but
caution is necessary here. Those on bridges and deltas should be
rejected as likely to be misleading. The whole scries, in order along
the coast road from 9359 to the head of L. Eil is: 15, 52, 41, 24, 38,
45, 31, 12, 27, 15, 45, 21, 40, 27, 15, n, 12, 30, 19, 10 ft. Some of
these would certainly be eliminated in the field, as marking crests
where the road crosses a protuberance, but in map interpretation
they must be taken as they stand. It would be straining the evidence
to claim that they are helpful : in twenty-five readings there is a range
of 35 ft. and little sign of grouping. The sample is too small to suggest
what is actually the case, that more than one raised beach exists as a
result of intermittent isostatic recovery. Again, the map scale is too
small and the contour interval too large to show what is at once noted
in the field, that rejuvenated streams have partly destroyed their raised
deltas, building anew at the lower level. The resulting "cone-in-
cone" form is well displayed on the ground at the northern side of
the Loch Eil narrows.

It is thus seen that die topographical map is not wholly successful
in representing on tin's fiord coast certain features which, although
small in extent, throw much light on recent shoreline evolution and
considerably modify the setting of human activity.

NOTES AND REFERENCES
For general accounts of geology, see

J. PHEMISTER. The Northern Highlands. Second Edition. British
Regional Geology Series, H.M.S.O., 1948.

82 MAP INTERPRETATION

V. WILSON. East Yorkshire and Lincolnshire. British Regional
Geology Series, H.M.S.O., 1948.

C. P. CHATWIN. East Anglia and Adjoining Areas. Second Edition.
British Regional Geology Series, H.M.S.O., 1948.

G. A. KELLAWAY and F. B. A. WELCH. The Bristol and Gloucester
District. Second Edition. British Regional Geology Series, H.M.S.O.,

1948.

H. DEWEY. South-ivest England. Second Edition. British Regional
Geology Series, H.M.S.O., 1948.

A full treatment of shorelines is

D. W. JOHNSON. Shore Processes and Shoreline Development. Wiley,
New York, 1919.

Much helpful material may be obtained from

J. A. STEERS. The Coastline of England and Wales. University Press,
Cambridge, 1946. (In particular, wave-action, longshore drift, and
spits are discussed on pp. 44-63, and strand-line movements in
Chapter XII, pp. 475-501. Some of the stretches of coast treated in
the foregoing text are considered on pp. 198-204, 211-12, 254-60,
420-438.)

The account of constructional forms given in this chapter might
usefully be supplemented by STEERS, op. tit., pp. 345-64 and by

J. L. DAVIES. "Wave Refraction and the Evolution of Shoreline
Curves," Geographical Studies, v, 2, 1958, p. i.

C. A. M. KING. "The Relationship between Wave Incidence, etc.,"
Trans. Inst. Brit. Geog., 1953, P- J 3-

Papers relating to the shorelines described in the foregoing chapter
include-?-

J, F. N. GREEN. "The High Platforms of East Devon." Proc. Geol.
Assoc., xlii, 1941, p. 36.

O. D. KENDALL. "The Coast of Somerset." Proceedings of the Bristol
Naturalists' Society, viii, 1936, p. 186.

H. GODWIN. "Studies in the Post-glacial History of British Vegeta-
tion, III and IV." Phil. Trans. Roy. Soc., B, 230, 1940, p. 239.

H. GODWIN. "Correlations in the Somerset Levels. New Phyto-
logist, xl, 1941, p. 108. (See also STEERS, op. tit., pp. 490-1, Figs. 105,
106.)

CHAPTER IX
LANDSCAPE IN AN ARID CLIMATE

The wind
Shoulders the pillared dust. BROWNING

MAP: DEPARTMENT OF THE INTERIOR, U.S. GEOLOGICAL

SURVEY, 1/62,500 (FINAL COUNTY (ARIZONA) CASA GRANDE

QUADRANGLE)

AN arid climate is one in which potential evaporation exceeds pre-
cipitation. It follows that deserts are regions of interior drainage,
except where very large rivers succeed in crossing them, without
drying up on the way; but, paradoxically enough, it is impossible to
understand much of the desert landscape without allowing a major
role to running and percolating water. Few areas are entirely without
rain : rain in the desert, when it does come, takes the form of violent
convectional downpours which, running off as heavily loaded torrents
or sheets of water, profoundly modify die relief.

Distinctive Character of Desert Landscapes

Landscapes shaped by desert weathering and erosion differ among
themselves as signally as do the glaciated landscapes of Britain. Just
as in glaciated country, however, the landforms produced are highly
characteristic and easily recognizable, whether in the field or on the
map. The nature of the geographical environment is never in doubt,
and a descriptive account presents few problems. The difficulties are of
another kind. As yet, the precise ways in which erosion operates in
arid conditions are not fully understood, nor has the scheme of the
desert cycle been satisfactorily worked out beyond the stage of
maturity. It seems best, therefore, to take the selected map as exem-
plifying certain leading features of a single class of desert landscape,
rather than a particular stage in the desert cycle, and to direct the
reader to the works listed below for a discussion of the morphological
problems involved, and for descriptions of contrasted arid tracts.

This map represents an assemblage of inselberg, pediment, and
alluvial plain. These are major landforms, sufficient proof in

84 MAP INTERPRETATION

themselves that the processes of arid erosion have been at work. On
the ground one would observe an abundant variety of detail, for, as
on shorelines, bare rock is exposed to differential attack ; but in both
cases limitations of scale compel the map interpreter to concentrate
chiefly on the texture of relief as a whole.

Inselbergs

The inselbergs are the Sacaton Mountains in the north-east, and a
similar mass in the south-west lying mostly outside the Quadrangle
and represented only by Double Peaks and near-by smaller hills.
Inselbergs are residual forms, rising abruptly here, to as much as
700 ft. above the surrounding gentler slopes. Some are based on
outcrops of particularly resistant igneous rock, and may be compared
to the tors found in humid regions; others are outliers of a partly
denuded formation, analogous to the outliers in front of escarpments
in the normal landscape. Whatever their composition, inselbergs
undergo severe weathering. The powerful sun by day, and unimpeded
radiation by night cause great changes of temperature in the exposed
rock-faces, which tend to disintegrate into angular slabs or coarse
grains. When water is available, heat promotes chemical change. By
these means, and with the aid of the natural sand-blast, rock can be
speedily worn away. Since the vegetation cover is necessarily sparse
and discontinuous, desert landscapes are made up of clear-cut features

with associated large accumulations of debris.

Alluvial Plain

It might be thought that the scree would remain banked up against
the inselbergs until it was sufficiently comminuted to be removed by
the wind. This might be so in completely rainless climates, but is
demonstrably not the fact in the Casa Grande neighbourhood. Loose
rock-waste has been carried away from the inselbergs, leaving them
continually exposed to further attack, and has been transported towards
the lowest ground, where it has accumulated in a constructional plain
with very low surface-gradients. On this map sheet the constructional
surface, nicely defined by widely spaced contours, is approximately
bounded by the railway on the north-east and by the upper part of
the Santa Rosa Wash on the south-west. The deposits beneath the
plain are certain to be generally permeable, and in suitable climatic
conditions may contain accessible reserves of ground-water (cf. the
numerous wells, especially towards the east).

LANDSCAPE IN AN ARID CLIMATE 85

It is clear that the supply of fine-grained waste is limited. Patches
of bare sand, as marked on the map, are mostly small, and dunes are
restricted in extent. Moreover, drainage channels occur, even though
the streams which flow in them are intermittent, and part of the plain
has been settled impossible conditions in a sand sea. One might
reasonably infer, therefore, that the rock-waste on which the plain is
based is partly coarse in calibre. It can scarcely have been transported
to its present position by wind.

Pediments

Between the eroded inselbergs and die depositional plain lie the
pediments, gently sloping surfaces cut across the solid rock and usually
covered with the thinnest veneer of rock-waste. The pediment
surrounding the Sacaton Mountains lies between the 1,500-6:. and
i,275-ft. contours on the western side, and between 1,600 and 1,400 on
the cast, while that below Double Peaks is observed to fall as low as
1,325 ft. towards the north. Inselbergs are separated from pediments
by a remarkably clear break of slope, below which the pediment sur-
faces descend smoothly towards the plain at gradients decreasing
slightly downwards. Their lower boundaries, where the flat plain
begins, arc scarcely less well defined although the change of gradient
is much less.

The mode of origin of pediments is far too controversial to be
debated at length here. Some authorities regard them as planed oft by
the lateral corrasion of intermittent streams, which work over the
pediments much as the streams of humid regions work over deltas,
repeatedly dividing and swinging from side to side. In support of
tins view it is urged that many pediments (including, as is at once
noted, those mapped here) are very shallowly dissected by numerous
tiny channels. It has been suggested that occasional streams, fed by
the rare convectional rainstorms, are capable of transporting the
heavy loads of rock-waste derived from the hills, but not of cutting
deeply down. On this view, the streams can corrade in the hill belt,
just succeed in transporting their load across the pediments, and aggrade
the alluvial plain where they disappear by evaporation and percolation.
Other writers regard sheet-flood as the chief process at work the
movement of occasional floodwater, heavily charged with the tools of
corrasion, over the whole pediment at once. Others again maintain
that a pediment is produced essentially by the bodily retreat, under the
attack of desert erosion, of the rock-wall above, and would compare

S6 MAP INTERPRETATION

it to the strike vales in front of scarps in humid regions. Whatever the
truth of the matter, it is generally agreed that pediments are progres-
sively extended as the higher ground is denuded. The useful term
pediplanation connotes the sum of the processes at work and their
total effect. The selected map is seen, then, to depict a tract where
pediplanation is going on. The inselbergs now observed are certain to
be destroyed. Final stages in their elimination are represented by
knobs of rock, isolated or in groups, such as the hill at 1,556 ft. about
two miles north-west of Casa Grande, or the detached remnants south
of the Sacaton Mountains.

Intermittent Drainage

All the streams of this tract are intermittent. Fed by surface run-off,
they cut gullies in the high ground and occupy shifting, braided
courses on the alluvial plain, where they lose most of their volume
before passing the bounds of the Quadrangle. It is evident that the
floor of this particular basin of centripetal drainage rises well above
the i,ooo-ft. contour.

Desert Basins

This consideration enables the interpreter to regard the landforms
identified in due perspective against a wider background of desert
erosion in general; for deserts are commonly enclosed by mountain
walls, constituting orographic as well as drainage basins. Hence the
considerable elevations of floor and walls alike. The encircling
mountains serve to accentuate an aridity, which might be expected
because of distance from the sea or the prevalence of high atmospheric
pressure. Many of the desert basins in the south-western United
States are, in fact, the tectonic depressions of highly faulted country,
and are bounded by fault-blocks. As the basin walls retreat before the
attack of pediplanation, detached portions survive for a time as
inselbergs, while the products of weathering accumulate except for
a fraction exported as wind-blown sand in the hollows. Plain,
pediments, rock knobs, inselbergs, basin walls, and the channels of
intermittent drainage thus make up a unitary assemblage of forms.
The pattern of drainage, the form of slopes, depositional features,
and features of erosion are not only characteristic in themselves, but
are combined in a distinctive manner. The resulting assemblage is
not least instructive in virtue of proving the qualities of the climate
in which it has evolved.

LANDSCAPE IN AN ARID CLIMATE 87

NOTES AND REFERENCES

Accounts of landscape development in arid and semi-arid climates may
be found in the texts listed at the end of Chapter II. The following
should be consulted for a fuller treatment of the problems involved

R. A. BAGNOLD. The Physics of Blown Sand and Desert Dunes.
Methuen, London, 1941.

C. A. COTTON. Climatic Accidents in Landscape Making. Whitcombe
& Tombs, London, 1947.

L. C. KING. South African Scenery. Oliver and Boyd, London and
Edinburgh, 1951.

Selected papers dealing with the nature and formation of pediments
are

W. G. V. BALCHIN and N. PYE, "Piedmont Profiles in the Arid
Cycle," Proc. Geol. Assoc., Ixvi, 1956, p. 167.

W. M. DAVIS. "Rock Floors in Arid and in Humid Climates."
Journ. Geology, xxxviii, 1930, pp. i, 136.

D. W.JOHNSON. "Rock Planes of Arid Regions." Geogr. Review,
1932, p. 656.

J. L. RICH. "Origin and Evolution of Rock-fans and Pediments."
Bull. Geol. Soc., Amer., xlci, 1935, p. 999.

CHAPTER X

A COMPLEX LANDSCAPE BASED ON
COMPLEX STRUCTURE

When we come to subdivide areas in detail, to examine the actual
differentiation of the earth's surface, the first of our considerations is
generally its, morphological diversity. S. W. WOOLDRIDGE

MAP: O.S. 1/63,360 (SEVENTH SERIES) SHEET 165
(WESTON-SUPER-MARE)

THE constituent stows of this piece of country contrast very markedly
with one another, in respect of geological structure, lithology, soil
property, form of relief, system of drainage, settlement pattern, and
surface utilization ; but, although a remarkable variety of country is
thus to be observed within a small compass, the task of physical
interpretation is not unduly difficult, especially in the light of the
foregoing studies. All the kinds of landform present here have already
been encountered. This map may therefore be treated in a somewhat
summary fashion, stow by stow, in accordance with the morphological
analysis sketched in Fig. 6. The subdivision made is based on the
principles recommended by Linton 1 and summarized in Chapter II.
Each boundary, as may be seen by comparing Fig. 6 and the O.S. map,
is drawn along a marked break of slope between two very different
types of country. The nature of the differences, that is to say the
rational basis of the subdivision, will be made clear in the following
descriptions.

I and 2: Hill Country of the South-west

These stows are undergoing dissection by many short streams,
which in the extreme south-west are deeply incised in narrow valleys.
Here the land rises to more than 900 ft. O.D. and the rocks appear to
be especially resistant. Note also the considerable extent of woodland
and moor. The lower ground, on the other hand, has been largely

1 D. L. LINTON. "The Delimitation of Morphological Regions." Published in London
Essays In Geography, edited by L. DUDLEY STAMP and S. W. WOOLDRIDGB. Longmans,
Green, London, 1951.

A COMPLEX LANDSCAPE 89

cleared: wide shallow valleys separate belts of low hills and the under-
lying rocks are evidently of no more than moderate strength.

3. The Levels

This is fen country, closely similar in many ways to the fenland
bordering the Wash. Much of the drainage is artificial, and the trace
of many natural streams has been lost. Note the extremely low
altitudes above mean sea-level that are recorded far inland. The tidal
range in the Bristol Channel is very great, as suggested by the wide
tidal flats and the long tidal reach of the River Parrett: hence an
increased need of land-drainage works, of embankments along the
rivers, and of strengthened coast defences. The strongest defence of
the fen is however a natural one. This is a lowland coast, on which
bars have been anchored by headlands and islands, with fen deposits
accumulating in the former lagoon behind.

The "islands" rising from the fen are of two kinds. The sizable
but very low-lying dry patches, only slightly higher than the rest of
the Levels, are based on permeable deposits, as suggested by the
absence of streams and ditches and the presence of nucleated settlement
and roads (cf. the sites of Chedzoy, Middlezoy, and Wcston Zoyland
(3437 to 3733) ). "Islands" of the second kind belong under the next
heading.

4. Small Cuestas and Outliers amid the Levels

The Polden Hills constitute a small cuesta, based on rock which is
not wholly impermeable. The scarp, which faces S.S.W., is in part
marked by strips of woodland. In the neighbourhood of Wedmore
(4347) a generally similar block of hills slopes gently towards the
south-west and has a steep scarp-face on the north-eastern side. If the
two stows are based on a single geological formation, they are to be
interpreted as the flanks of a denuded syncline with its axis roughly
along the line of the River Brue. Contrast the denuded anticlines
studied in Chapter IV, where the scarps were observed to face inwards.
Now as Brent Knoll, Glastonbury Tor, and Pennard Hill lie somewhere
near the supposed synclinal axis, it may be supposed that they are
outliers of a formation younger than that in the Polden Hills and at
Wedmore. A synclinal structure is compatible with the fact that they
are scarped on all sides. Note also that Glastonbury Tor and Brent
Knoll both display a tabular flat c. 300 ft. below the summit. It
seems likely that a resistant cap-rock is underlain by weaker beds, and

90 MAP INTERPRETATION

those again by strong rocks which crop out in structural benches.
Similar features were noted, and a similar inference drawn, in the
study of outliers from the Cotswolds (p. 23).

The hills around High Ham (423 1) in the south resemble the south-
eastern part of the Polden Hills very closely. If the central syncline
were followed on the south-west by a complementary anticline, one
might expect a cuesta similar to the Poldens but facing in the opposite
direction; but, since the highest and best-formed scarp in these
southernmost hills faces north-westwards, at right angles to the front
of the Poldens, it is clear that some structural complication has super-
vened, i.e. that the fold system is not composed of simple, parallel
elements.

5. The Mendips and Other Similar Limestone Hills

The Mendips rise abruptly above the valley flats of the Axe on the
south and the gentle slopes of the Yeo valley bottom on the north, to
a wide, flat plateau surface lying mostly between 850 and 900 ft. O.D.
(Plate HIA). As surface drainage is almost entirely lacking the under-
lying rock must be highly permeable, and also mechanically strong,
for dry valleys become gorges at the edge of the upland with precip-
itous, rocky walls. These features, together with the caves and caverns
noted on the map a mere selection of the total known are those of
karst country, which one may suppose to be developed on Carbon-
iferous Limestone (cf. Chapter VII).

Although sinks are not marked by a special symbol, the streams
flowing northwards off Blackdown (Fig. 6) evidently vanish into the
limestone. In actuality, sinks occur at many points but are not shown
on the map : one might expect them to be excluded at this scale. The
limited development of steep rocky slopes and the apparent lack of
limestone pavement are more difficult to account for, but like the
paucity of dry valleys are to be related to the remarkably perfect
planation of the plateau surface.

One soon realizes that the Mendip plateau is an erosional not a
structural surface; for the valley of the Lox Yeo, hemmed in by the
western hills, cannot be other than a denuded dome or basin. In other
words, the structures in the limestone block are far from uniclinal,
not to mention horizontal. The plateau surface is developed across,
and regardless of, high and inconstant dips. Brief consideration shows
that the Lox Yeo valley has been carved from a short pitching anticline
or elongated dome, with the longer axis running approximately

A COMPLEX LANDSCAPE

east-west. If the structure were synclinal, the impermeable rocks
which now floor the centre should be traceable some way to the east
by means of a visible stream system. There is, in fact, only a small

FIG. 6. MORPHOLOGICAL SUBDIVISION: PART OP THE

SOMERSET PLAIN

Sec text for explanation of numbers

Based, by permission, on Sheet 165 of the O.S. New Popular 1/63,360 Map)

impermeable outcrop on the high ground of Black Down, which
supports the short streams already noted as disappearing underground
on the north, into the overlying permeable beds. 1 The small enclosed

1 The dome was first excavated at least as early as Triassic times, as the 1/63,360 geo-
logical sheet proves at a glance, but the topographical map records only the net effect of
all erosion and infilling.

92 MAP INTERPRETATION

lowland is therefore taken as an anticline with its weak core eroded
away.

North of the Yeo valley, the western extremity of a group of hills
reproduces the association of physical and cultural features which is so
well displayed by the Mendips. Dry hilltops between dry valleys with
steep and rocky sides (Goblin Combe, 4765, Brockley Combe, 4766,
Healls Scars, 4966), at least one cave (in Brockley Combe), woodland
on the outer slopes, cleared woodland or heath on the flat summits,
and a long barrow permit of no doubt that this too is limestone
country of the same kind, differing from the Mendips in standing
lower and in being less regularly planed off by erosion. Carboniferous
Limestone also underlies the belt of hills stretching westwards from
Bristol and turning sharply north-eastwards at Clevedon to follow
the coast, but the evidence of rock type is less plentiful here, consisting
chiefly in the absence of surface drainage. Detached outliers of the
Limestone are seen in Middle Hope, Worlcbury Hill, and Brean
Down, as also in Steep Holme and Flat Holme; but, apart from the
"Celtic Fields" on Brean Down, which might correspond to limestone
soil, the only positive indications are of resistant rock of indeterminate
type.

6. Broken Hill Country in the East

Heavily dissected hill country comes in on the east, on both sides
of the Mendips. The texture of the landscape, the pattern of occupancc,
and surface utilization are all very different from those observed on
the outcrops of Carboniferous Limestone. The underlying rocks
appear to be of only moderate strength, for dissection is considerably
more advanced than in the limestone areas just reviewed. Broken
lines of scarp, and many valleys containing surface streams, indicate
sedimentary formations, in part weak and impermeable but in part
resistant. One may conjecture that, on the northern side at least, these
formations are younger than the Carboniferous Limestone; for the
latter, in the outcrop north of the Yeo, forms scarps which face south
and west, and should therefore possess an easterly dip which would
carry it under the hill belt to the south of Bristol.

The Drainage System

Although rivers of the levels (or the large artificial channels by
which they are now represented) are aligned roughly parallel to the
grain of relief, they cannot be described as adjusted to structure, for

A COMPLEX LANDSCAPE 93

they are not in contact with the solid formations. They flow, instead,
over fen deposits which insulate them from the deformed rocks
beneath. These rivers should, therefore, be regarded as occupying
the infilled valleys of a former landscape wherein the form of the ground
was related to structure, and where some measure of drainage adjust-
ment had been attained.

Since little can be discovered of the geological structure of the
south-western hills, no comment can be made on the relation of
drainage to it, but one may note as a fact that streams here flow in
roughly parallel courses towards the north-east. In the Mendips and
other outcrops of Carboniferous Limestone the question of adjustment
does not arise in the same way. It is more relevant to observe that the
drainage may be taken as mature, having gone completely
underground.

The map represents an area just large enough to bring out the
discordant relationship of drainage to structure in the north-east, where
the Avon turns north-westward from a broad valley to pierce a belt
of limestone hills in a gorge c. 250 ft. deep. There would seem to
be an easier route on weaker rocks, already picked out by the diminu-
tive Land Yeo, through the gap at Flax Bourton (5169). Since the
Avon must have become established in its course across the limestone
before this gap was eroded, the gorge is seen to be the result of super-
imposition, which has not yet been followed by readjustment of
drainage to the structures now revealed.

The Shoreline

Submergence is proved by the Submarine Forest (2246), while
emergence is strongly suggested by the wide expanse of fen behind
beach bars, and strikingly demonstrated in the field, for example, by
raised beaches north-west of Clevedon: the shoreline is therefore com-
pound. It is dominantly, however, a lowland shoreline of emergence,
far from mature in the present cycle, and diversified by headlands
where tongues or outliers of resistant rock run down to the sea.

A low offshore gradient and a great tidal range combine to produce
a wide foreshore, sandy, at least in part. Dunes have been formed
north of die Brue estuary. Wave-built features occur both at and off
the shore (cf. the banks such as Langford Grounds off the mouths of
the Yeo and Kcnn, and the shingle ridge which runs from Stolford
(2346) to Steart Point (2847), where it recurves).

At a few places, retreating cliffs cut in resistant outcrops have left a

94 MAP INTERPRETATION

visible abrasion-platform, nowhere widely developed in the north-
east but, finely shown in the extreme south-west, where it is mile
wide.

NOTES AND REFERENCES

For a general account of the geology, see

G. A. KELLAWAY and F. B. A. WELCH. The Bristol and Gloucester
District. Second Edition. British Regional Geology Series, H.M.S.O.,
1948-

Special aspects of landform are discussed in

H. E. BALCH. The Caves ofMendip. Fold Press, London, 1926.

A. E. TRUEMAN. "Erosion Levels in the Bristol District, etc." Proc.
Bristol Naturalists' Society, viii, 1938, p. 402.

H. GODWIN. "Correlations in the Somerset Levels." New Phyto-
logist, xl, 1941, p. 108.

The form and distribution of settlement, which are not dealt with
in the preceding text, are analysed and described by

B. M. SWAINSON. "Rural Settlement in Somerset." Geography, xx,
1935, p. 113.

B. M. SWAINSON. "Dispersion and Agglomeration of Rural Settle-
ment in Somerset." Geography, xxix, 1944, p- i-

PART II

THE FEATURES OF OCCUPANCE

CHAPTER XI
FEATURES OF OCCUPANCE: GENERAL

And beware how in making the portraiture thou breakest the
pattern. FRANCIS BACON

MAN'S activities have not greatly altered the form of the ground,
except locally, where quarrying or tipping has been extensive; but
reclamation, drainage, clearing, cultivation, and building have signally
altered the observed landscape. To the extent that it fails to portray
the visible scene, the map fails to show how little the present country-
side resembles the "natural* ' countryside of prehistory. Forests,
woodlands, heaths and marshes of the British lowlands, and of some
parts of the highlands, have been replaced by a patchwork of fields,
interlaced by man-made roads and thickly studded by settlements.
Even the high moorlands, which still remain largely open, have been
greatly changed: their treelessness is due in part to the grazing of
domestic animals, or to climatic changes which have occurred since
the arrival of man.

The study of cultural features lends itself well to map analysis, in the
conventional sense of that term the abstraction from a given map of a
certain class of data, such as roads, settlements, or mines. Selective
maps produced in this way have an undoubted value in representing,
with all possible clarity, distributions which may be somewhat
obscured on the full topographic sheet. The interpreter might well
make a full analysis of a chosen map at an early stage in his work,
taking off the several distributions in turn, in order to demonstrate to
himself the wealth of data available and the way in which patterns
emerge when various elements are separately studied. The ultimate
aim, however, must be to take a synthetic, not an analytic, view to
interpret the map as it stands, and to perceive relationships which are
necessarily suppressed or minimized when distributions are taken
singly.

Interpretation of Land Use

On the O.S. topographical maps, evidence of agriculture is mostly
of a negative kind. With few exceptions, the most highly productive

98 MAP INTERPRETATION

agricultural land is left blank. The existence of a special series of Land
Utilization maps is, in itself, a measure of the deficiency. A little
thought will show that most of the features of contemporary occupance
recorded on the standard maps fall into the category of Waste, i.e.
agriculturally unproductive land, in the classification of the Land
Utilization Survey. Six other categories are distinguished in that
scheme. Two, Inland Water and Forest (which includes all wooded
land), are marked by special symbols on the topographical map, so
that the interpreter is left in no doubt of their extent and distribution.
On the Land Utilization maps, Gardens include nurseries and orchards
in addition to domestic kitchen or flower gardens; topographical
maps record only orchards by a special symbol, except where the name
"nursery" occurs or where, on the 1/25,000 and larger scales, tiny
enclosures attached to dwellings are presumably for the most part
cultivated. The distribution of heath and moorland, as shown by the
familiar tuft symbol on the topographical sheet, is broadly similar to
the extent of Heath (land suitable for rough grazing) recorded by the
Land Utilization Survey for upland areas, but in the lowlands the
topographical maps are liable to omit much, while Arable and Meadow
(permanent pasture), accounting for the greater part of the best agri-
cultural land, are not symbolized at all. All the interpreter can do,
therefore, in dealing with the map of an area which has little wood,
moor, heath, or marsh, and which carries a rural population, is to
assume that most of the land is devoted to agriculture.

Needless to say, any specific points of relevant information should
be discovered and made use of. The nature of the boundary between
enclosed and improved land on the one side, and open, unimproved
land on, the other, is particularly interesting. Since field boundaries
are given on the 1/25,000 series, the extent of enclosure can be found;
but, as there is no positive evidence of the quality of the worked land,
one must rely on such data as the symbols for limestone pavement or
for rocky ground to show which fields are likely to be in pasture
rather than in tillage. In areas of crofting settlement, even this evidence
is likely to be misleading. Where woodland or heath is widespread in
a low-lying tract, the map may also provide inferential evidence of an
outcrop of sand or clay, but the possibly more favourable soils of adja-
cent cleared areas can only be guessed at not, as a rule, very accurately
by the absence of heath or wood.

Sometimes the beginner, encouraged by having identified the
underlying structures and types of rock, is tempted to suggest an

FEATURES OF OCCUPANCE! GENERAL 99

agricultural response. Such a practice would be determinism of the
crudest sort. It leads directly into a number of pitfalls. Soil type is not
determined solely, or necessarily at all, by solid geology. The initial
work of the new Soil Survey reveals, as any rational geographer must
have foreseen, a most complex and minutely diversified distribution
of soil series. Again, land use is not influenced by soil alone. Quite
apart from the influence of superficial deposits, of vegetation, of
climate, and of cultivation on the qualities of farmland, one must
always remember that agricultural distributions are in part a function
of historical and economic factors. The Land Utilization maps, drawn
from a field-by-field survey in the nineteen-thirties, have already
been relegated to the status of historical documents by the agricultural
changes which accompanied and followed the war of 1939-45.
Bearing in mind that there have been other periods in which no less
marked alteration in the structure of farming occurred, the map inter-
preter must content himself with concluding that a cleared, settled
rural area is presumably devoted to agriculture, and with observing
the probable extent and limits of cultivation. Within die bounds of
worked land, positive evidence is likely to exist in the form of named
farms, symbolized orchards and glasshouses, the channels of water
meadows, and, in places, such establishments as sugarbeet factories,
which by their nature indicate one of the crops raised; but all these,
even in the most favourable areas, provide no more than a small
fraction of the total possible information.

Routes

Railways and canals, which are features of industrialism, present
few problems of interpretation, except that on some map series a more
detailed classification might be conveniently employed. Their modes
of development, their functions, and their varying success are good
general guides. Natural waterways and road systems, however, are
more difficult to deal with, partly because their use dates from very
early times.

Canals. Canals are usually closely adapted to the form of the
ground, running along valley bottoms or parallel to the contours on a
slope. The existence of a canal implies that at the time of construction
a demand for transport facilities existed, or was expected to arise, but
in this country canal traffic has so greatly declined that parts of the canal
system now lie derelict. Where canals appear on the map they should
be examined for notation (disused), which is by no means uncommon.

100 MAP INTERPRETATION

Apparent signs of present use may be unreliable. Some factories
alongside canals are themselves derelict; others prefer road or rail
transport, even for the bulky goods which canals are designed to
carry. Canal names are worth particular attention: the original
purpose of the "Coal Canals" is self-evident, while "Navigation
Canals" were often intended as a substitute waterway for rivers on
which, for some reason, navigation was difficult. Many canals take
off from rivers at points determined by physical obstacles of some
kind, but the map evidence is not always clear. One may assume that
below the canal junction a river is navigable for canal-boats, as weirs
and locks may show. Nothing can be said of larger craft, however,
unless the map represents port facilities, for example, coal staithes.
It is significant of the small use made of inland waterways in this
country that the scale of map symbols makes no distinction on the
basis of capacity. Coastal ports are better treated: large harbour works
can be shown in plan, while the upstream limit of tides is recorded, on
the 1/63,360 (Seventh Series) Series, by a change from the black line
of HWM to the blue of inland water.

Railways. The railway network in Britain, laid down piecemeal
by competing companies, shows unnecessary duplication of routes in
some parts and deficiencies in others. A number of sizeable towns,
for example, lie off the main lines. These anomalies apart, railways
connect the great towns and industrial districts witli one another and
with the ports, as well as providing transport services of varying
quality in rural areas. Map interpretation is usually concerned with a
small part of the system at one time. Consequently, through-connec-
tions must be taken for granted, and attention confined to the class of
installation shown, the relation of routes to the form of the ground,
and the apparent effect of rail services on the location and growth of
settlement.

The scale of symbols used for railways varies with the map series,
the editions of a single scries, and even with the style of a single edition.
Thus on the Provisional Edition of the O.S. 1/63,360 (New Popular)
maps, railways with two or more tracks are marked by a solid black
line, whereas on the later New Popular Edition proper a chequered
line is used, similar to that on the 1/25,000 (Later Style). Single-
track lines are recorded on both series by a lighter chequer. The
distinction between the two classes corresponds to a useful division in
railway practice, between lines where two-way working is or is not
possible without by-passing. Other track symbols demand careful

Photograph: Aerofilms, Ltd.

(A) AERIAL VIEW OF THE MOUTH OF THE FOWEY RIVER,
SHOWING Low PLATEAU INLAND

Photograph: Aerofilms, Ltd*

(B) BEN NEVIS (CENTRE DISTANCE) RISING ABOVE SURROUNDING
MOUNTAINS

Note cultivation on raised beach, left foreground

PLATE I

Geological Survey photograph, reproduced by permission of the Controller, H.M.S.O.
[A) GENERAL VIEW or INGLEBOROUGH AND SIMON FELL
Glints in the foreground

Geological Survey Photograph, reproduced by permission of the Controller, H.M.S.O.

(B) THE SHORELINE NEAR SIDMOUTH

The band of beach can be seen to cross the mouth of the River Sid at the extreme
right. Abrasion-platform in the foreground

PLATE II

FEATURES OF OCCUPANCEI GENERAL IOI

reference to the key of the map: a hatched line on the 1/63,360 O.S.
sheets means a siding or tramway, but on the 1/25,000 scale stands for a
narrow gauge railway; sidings on the 1/25,000 appear as solid black
lines. For narrow gauge tracks the Seventh Series 1/63,360 uses
an adaptation of the chequer symbol. The 1/25,000 indicates clearly
which factories have rail connections; the 1/63,360, on close examina-
tion, will reveal the location of marshalling yards, goods yards, and
railway depots.

Railways are less closely adapted to relief than are canals, partly
because sharp bends in the track are impracticable and because a certain
gradient is permissible. Nevertheless, where relief is at all varied one
usually finds that the actual course of a railway line is a compromise
between the straight line from point to point and the line of least
gradient. Conversely, a long straight track is a guarantee of very low
natural gradients, but with this exception each stretch must be dis-
cussed on its merits in respect of guidance by relief features.

In examining the relation between railways and die growth of
settlements, one must recall that in long-settled countries most villages
and many towns already existed as villages and towns, before the
railways were laid. Consequently, railways as a factor in locating
settlement need little attention, but they can be shown, on the other
hand, to have influenced the manner and rate of expansion of pre-
existing settlements. The remarkable growth of towns in occidental
countries, wliich began in the nineteenth century and has continued
into the twentieth, cannot be looked on as caused by the construction
of railways : urbanization and railways alike are the outcome of that
complex of developments, the Industrial Revolution; but, since, for
the purposes of industry, large quantities of goods and large numbers
of people must be transported from place to place, it is natural to look
for some connection between the spread of towns and the existence
of a rail link.

The connection is most obviously revealed where a town has grown
towards, and across, a railway. When the lines were first laid they
usually skirted the towns then in being, so that where they now run
through a town it is often found that they have been included by recent
growth. Sometimes the early stage of the process is illustrated, as
where a town has not yet spread beyond the tracks, or where the
original nucleus of settlement lies at some distance and a newer group
of dwellings has been established near the station. Many of the clearest
examples of the stimulating effect of a rail connection on the rate of

8 (.5196)

102 MAP INTERPRETATION

growth, and on the direction of growth, are to be found in dormitory
areas, where small towns or villages are within easy reach, by rail, of a
large industrial town. Swollen dormitory villages along the railway
correspond to ribbons of housing along main roads, where motor
transport is the stimulating factor.

Roads. The pattern of roads is essentially composite. It has had a
far lengthier evolution than have the networks of canal and rail, and
combines antique elements with those of very recent origin. In the
broadest sense and as a whole, it is the most intimately associated with
the pattern of settlement. Furthermore, although canal and rail have
exercised their effects on the distribution of factories and on the direc-
tion of urban growth, roads are now recovering their dominance.
The most recent industrialization, generally post- 1920 and by no
means at an end, demands vast facilities for road transport of goods.
Travel by road has increased no less remarkably. It is not surprising
that in Britain the road system, alone of the three networks, still tends
to grow. The additions include new arterial roads and by-passes as
well as new streets in expanding towns.

Completely new roads, however, are rare, for changing needs have
been met by widening and re-surfacing rather than by entirely new
construction. The great road-makers of the later eighteenth and early
nineteenth centuries also achieved a compromise between improve-
ment and substitution. Only under the Romans was a single planned
system imposed on the whole country. Many elements of that system
can be distinguished in the present road network, but even where the
Roman line is preserved it may be followed in part by bridle tracks
and footpaths instead of by metalled roads. The dismemberment of
the Roman roads clearly illustrates the manner in which the present
system, t symbolized on the map according to present condition and
function, results from long-continued growth, selection, adaptation,
superimposition, and abandonment.

In map interpretation, therefore, the classification of roads should
be looked on as a rough guide to present use. A better general view
of the whole system is obtained when all roads, of whatever grade,
are traced off the map and indicated by a uniform symbol, when
unsuspected through-routes are likely to appear. Frequently the rela-
tion of the road system to the forms of the ground will also be more
clearly seen, for the various elements of the modern system are, as it
were, re-distributed among the different nets from which they have
been derived. In particular, this exercise is well suited to re-constitute

FEATURES OF OCCUPANCE: GENERAL 103

the Roman lines, and to reveal a great many ridgeways which tend to
be suppressed on the topographical sheet.

It is generally known that prehistoric men circulated for the most
part along the relatively open hilltops. Where the crests are very
narrow it is possible that definite trackways were beaten out, but the
early ways on the wider uplands should not be thought of as "roads'*
in the present sense. Very many of the ridge-routes, which in some
form or other are incorporated in the present road system, owe their
survival, if not their very existence, to medieval traffic. This point
deserves to be strongly emphasized, especially in connection with the
interpretation of town sites: many a so-called "gap town" was origin-
ally located, not at a focus of valley- ways, but where a ridge-route was
interrupted by a crossing (cf. Chapter XIII). A not dissimilar case is
that of the Anglo-Saxon herepaths, harepaths, or hard ways, occasionally
named on the map. These, literally "army paths" but in fact surfaced
or at least firm roads, run generally along the crests and show where
overland movement was easiest in early historical times. Valley-ways
are a later development : although they account for so much of the
existing main-road system, they must not be allowed to confuse
interpretation of origins.

The rectilinear Roman system and the irregular net of medieval
and modern times are both guided by relief, but in different ways.
Despite their long straight stretches and their small concession to
details of landform, the Roman roads are remarkably well adapted to
major physical features, as may be clearly seen on the J-in./mile maps.
The Fosse Way, for example, runs near the foot of the Cotswold dip-
slope, avoiding the more deeply incised valleys of the plateau to the
north and the ill-drained claylands to the south. Passing through the
Moreton Gap, the best defined break in the Jurassic hills for many
miles in either direction, it continues below the main scarp-face but
above the bottom of the Avon valley, which is floored with marly
boulder clay. Post-Roman roads, tortuous as a whole, are at many
points guided by minor topographical features but are also instructively
related to the broader lie of the land, for example in clay country,
where they are frequently seen to run along the very low divides.
The manner of approach to obstacles, such as rivers, is especially
informative, and well repays study in connection with die siting of
towns: it is proper to inquire why the roads should make for a
particular point in order to cross, rather than another in the same part
of the valley.

104 MAP INTERPRETATION

A concluding general remark may be offered on the subject of
nodality as expressed in the pattern of routes. Because roads are the
most numerous of routes, they are the best able to show convergence
on a particular focus; but it is a mistake to imply that a town has
arisen or expanded because roads, as such, converge on it: usually the
roads have become better defined and more frequented as the town
has grown larger. It is inevitable that roads should reveal the nodality
which is essential to all towns, but the nodality itself is a matter of
situation rather than of connections.

The Pattern of Rural Settlement

Interpretation of settlement can be highly rewarding, if the inter-
preter will recognize the fact that, with few exceptions, the map
attempts to show present distributions only. Whatever can be
inferred, nothing is actually stated about how the observed patterns
came into being. There is, for example, little direct evidence that the
different parts of a town differ in age or in function, or that different
elements in the pattern of rural settlement were introduced at different
times. Here is at once the interpreter's task and his limitation.

The remainder of this chapter will be given over to a discussion of
those various aspects of rural settlement which most closely affect map
interpretation. Prehistoric occupance and the analytical study of
towns are made the subjects of separate chapters, where they can be
conveniently illustrated by reference to maps selected for the purpose.

Units of Rural Settlement. The development, stage by stage, of
a given complex of settlement and routes is to be traced by the
historian, the historical geographer, and the archaeologist; but the
pattern cannot be perceived until the roads, buildings, historic and
prehistoric sites have been mapped. To suppose that patterns, especially
patterns of settlement, can be made out implies that habitations are
grouped and distributed in characteristic and recognizable ways.
With these the interpreter should be broadly acquainted. The units
of rural settlement are the isolated farm, the hamlet, and the village.
Small market towns are closely linked with their surroundings, both
economically and socially, but basic functional differences compel
them to be treated under a different head. Although there is continuous
gradation through the farm-hamlet-village series, with the result that
a strict classification is not always possible, it is nevertheless often found
that a given stow or tract is dominated by one of these elements, or
by more than one in a definite combination. In parts of the English

FEATURES OF OCCUPANCE: GENERAL 105

Midlands, for example, villages lie near the centres of their parishes
with outlying farms around them; elsewhere the hamlet is dominant;
elsewhere again, most rural habitations are disseminated single farms,
as is finely illustrated on the O.S. Ireland 1/63,360, Sheet 169.

The fundamental distinction is that between nucleated settlement
and dispersed settlement. The observed pattern need not, however,
be the original pattern. It expresses merely the cumulative effect of a
number of factors in the complex relationship of man to ground.
The problem of interpretation is far from simple, and can scarcely be
attacked unless certain principles are clearly understood.

Some environmental factors appear to have exerted a powerful
but by no means exclusive effect on the form of settlement. We may
remark, with Demangeon, that even relief and limited water-supply
favour nucleation by limiting the effective choice of sites, whereas
broken terrain and abundant water encourage dispersion. In this
connection one may note the distinction commonly made between
wet-point sites, where water is available in a relatively dry tract,
and dry-point sites, for example, "islands" of relatively good
drainage in a marsh. A common relationship between the siting of
settlement and the natural water-supply is, however, not an invariable
rule. Social and ethnic factors, although not yet well understood and
rather uncertain in their effects, have undoubtedly influenced the
form of some rural settlement. Demangeon considers the chief of
them to have been the form of agricultural economy, which although
itself open to the influence of the environment, appears to have been
capable on occasion of determining the form assumed by pioneer
settlement. Thus although "the unit of settlement on the English
Plain was the village community," consisting in its purest form of a
cluster of houses standing in the midst of its territory, 1 the Jutish
system of land tenure and agriculture was associated with hamlet
settlement in the lower Thames valley and in the Continental
Angle.

It is no part of the present task to discuss the origin of the early field
systems, the way in which they were linked with the feudal system of
government and administration, or the effects on agricultural practice
of improvements in tools and systems of cropping or stock manage-
ment. The mere reference must suffice to show that the historical
problem of rural settlement is exceedingly difficult and complex.

1 H. C. DARBY, in An Historical Geography of England Before A.D 1800. Edited by
H. C. DARBY. University Press, Cambridge, 1936, p. 189.

106 MAP INTERPRETATION

To suppose that a crude and direct relationship exists between
settlement and land is to be guilty of gross over-simplification.

The observed pattern of rural settlement in the English Plain has
evolved, in large measure, from the pattern established by Saxon,
Scandinavian, and contemporary pioneers. The evidence of place-
names and Domesday Book proves the antiquity of many settlements,
while Darby has shown that much of the forested low ground had
been cleared before the Norman invasion. In other words, the
conquest from nature of what is to-day the best agricultural land was,
to an important degree, the work of the Dark Ages. There have of
course been later changes. Some village communities were destroyed
by a catastrophe such as war or plague; others were removed to make
way for the sheep farming of the fourteenth and fifteenth centuries.
Some new settlements have been established, elaborating the ancestral
pattern. In the eighteenth and nineteenth centuries openfield working
was finally superseded by wholesale enclosure, and new farms were
built on the newly consolidated holdings.

Forms of Rural Settlement. These most recent enclosures are
principally responsible for the combination, in many lowland tracts,
of nucleated villages and outlying farms. It is important to recognize
that the two elements date from widely separate periods. Where the
village is the original settlement, the form is that of primary nuclea-
tion, while the farms represent secondary dispersion, i.e. a later
re-distribution of population in single scattered dwellings or hamlets.
Since the village survives, secondary dispersion has been incomplete.
Dispersion is also possible as an original form, but is rather difficult to
prove. Again following Demangeon, we may identify old-estab-
lished primary dispersion, intercalated dispersion, and recent
primary dispersion, in addition to the secondary dispersion already
mentioned. Old-established primary dispersion is most appropriate
to moorland, woodland, and mountain country, which is unstated
to openfield working by a village community and where potential
grazing is more abundant than potential tillage. Intercalated dispersion
results from piecemeal colonization, for example, during the Middle
Ages, of the forests and wastes remaining between the cleared land of
the villages. Recent primary dispersion is the form of relatively
modern agricultural settlement, where single farmsteads have been
established on the newly won land.

The problem of old-established primary dispersion arises in discus-
sions of Highland Britain, where the systems of land tenure descend

FEATURES OF OCCUPANCE: GENERAL 107

from Celtic times, never having been superseded by Romano-
British or later practices. In these generally less hospitable and wetter
tracts, the environmental factors thought to favour dispersion exert
their strongest influence, especially where individual parcels of land
have been reclaimed at the edge of the waste. Even here, however,
investigation shows that primary dispersion cannot be taken for
granted, no matter how widely disseminated dwellings may be at
the present day. The custom of gavelkind in North Wales, and
the breakdown of the rundale system of agriculture in Ireland, are
now known to have assisted in dispersion from primary nuclei.
Map evidence can be adduced for the complete dispersion of two
Donegal villages between 1834 and I9O3. 1 In view of these
findings, one is well advised in dealing with rural settlement in
Highland Britain to attempt nothing beyond a statement of existing
distributions.

Parish Boundaries. It has long been known that a map of parish
boundaries may bring out significant and useful facts. Despite
numerous small adjustments, and some greater changes effected for
administrative convenience, many boundaries remain substantially
the same as they were when first determined. Hence it is possible to
tell, in a general manner, what kinds of land were held by many of
the original village settlements. For, although the same caution must
be observed as in discussing the qualities of present agricultural land,
one may justly recall that the medieval openfield systems tended to
produce stereotyped methods of cropping and rotation, practised on
the best arable land, and demanded in addition meadow, woodland,
and common pasture. Where villages lay along a scarp-foot, there
were obvious advantages in the division of land into long, narrow
parishes running across the grain of the country, with meadow and
woodland on the lowest ground, common grazing and probably
woodland again on the hills, and the open arable around the village
itself in the scarp-foot zone, which is typically one of favourable soils.
Thus parish boundaries came to cross the boundaries of adjoining
stows. A similar fact may be noted on parts of the fenland margin,
where each group of villagers and each manorial lord wished to
secure rights of exploiting part of the marsh. Where the differences
in land quality were smaller, or were less apparent to the early settlers,
and again where villages were founded in forest clearings, parishes
tend to be more nearly hexagonal, although the actual boundaries
Sec the work of E. ESTYN EVANS, cited on p. 113.

108 MAP INTERPRETATION

are usually irregular in detail. Examples of both forms in juxtaposition
are given in the annexed Fig. 7 (/).

Rectilinear boundaries are uncommon. They occur in those parts
of the fens which were drained late in historical times, where no
landmarks had existed before reclamation, and in some mountainous
tracts, where they are of limited interest or influence. Some, however,
which run along Roman roads, neatly demonstrate that the road
existed before the parish was delimited. Parts of Watling Street, for
example, are used to mark county as well as parish boundaries.

Interpretation of Rural Settlement: Village Forms

The forms of individual villages, no less than their actual existence,
may be influenced both by the environment and by the agricultural
organization of the original settlers. As yet, the study of village forms
is far less advanced in Britain than in some European countries, where
the material is on the whole far more promising: large compact
villages unaffected by secondary dispersion are easier to classify
according to form than the small, shrunken agglomerations of rural
England. In the interpretation of British maps, therefore, it is usually
sufficient to distinguish elongated from massed villages. Where the
forms are clear and detail is abundant, as on the 1/25,000 scries, a
number of subclasses can usefully be made out. Thus, the elongated
or linear villages include the street-village, where houses line the
street on either side, and the street-green village, of cigar-shaped plan
and with an elongated central green. Massed villages are less easily
subdivided, but one may separate those which are close-knit (with or
without a central green) from those which are loose-knit or of indeter-
minate shape. Selected forms are illustrated, by village plans drawn
from the O.S. maps used with this book, in Fig. 7 (a-e).

The interpreter should not attempt to classify all the villages on a
given map according to this scheme, which is by no means compre-
hensive, but should seek to identify the dominant form(s) and to
select a few marked exceptions for detailed examination.

Interpretation of Rural Settlement: Place-names

Place-name evidence is conveniently taken in conjunction with the
pattern and lorm of rural settlement. Like the settlements themselves,
the names have been modified in the course of time. Since the map
records only the modern name (except at a few Roman sites), and,
since place-name study is a highly specialized task, a map interpreter

FIG. 7. PARISH BOUNDARIES

(Based, by permission, on O.S. maps)

Crown Copyright reserved

(a) Willerscy, a compact village (O.S. New Popular 1/63,360, Sheet 144, 1039).

(b) Hagworthingham, a loose-knit village (O.S. New Popular 1/63,360, Sheet

114, 3469).

log).

(/) Pattern of parish boundaries at the side of the Vale of Pewsey. Parish
churches and the scarps of Chalk and Greensand arc shown.

MAP INTERPRETATION

should attempt inferences only where certain common elements are
widespread. In this way one may discover something of early settle-
ment, avoiding at the same time most of the manifold errors which
arise from guesses at the derivation of particular names. The following
list contains a selection of the commonest elements in English place-
names, together with meanings and original forms, derived mostly
from Saxon, Norse, and Danish, but with a few Celtic additions.
The Ordnance Survey publishes a very useful Glossary of the Most
Common Welsh Words Used on the Ordnance Survey Maps (1949, 4d.),
and formerly issued a Glossary of the Most Common Gaelic Terms Used
on Scottish Maps, now unfortunately out of print.

A. ELEMENTS OF OLD ENGLISH ORIGIN

Element Meaning

berry, borough, (primarily)

borrow, brough, fortified place

burg, burrough,

bury

cester, Chester

chipping

cot(c)

dean, den, dene

don, dun
ea, ey

field
ford
ham *

hampstead,
hempstead *
harepath,
herepath,
hardway
hirst, hurst

holt
ing

ingaham, ingham
ley

port

stead, sted
stow
strat, street,

large city or town

market
cottage
valley

down, hill

island; also "land in midst of
marshes"

generally "farm** or "estate"

site of chief house of a farm or

manor

literally "army path" but used herepaej?

generally of through-roads

(i) hillock, knoll, bank (ii) copse,
wood possibly originally "wood-
ed height"
wood

often originally ingas, "X's
people"

the ham of X's people
originally woodland; later, clear- teah
ing in woodland

town, especially with market rights
and rights of minting

place, position, site stede, styde

place, site __

primarily a road of Roman con- strxt

Derivation

burh (dative singular byrig); but

liable to confusion with Old

English beorg, and Old Norse

berg, hill

ceaster, caester; loan-word from

Latin castra, camp, settlement

cleping

cote

denu; but may be confused with

Old English denn, (i) lair of wild

beast (ii) pasture for swine

dun; but liable to confusion

with denu and tun

ea, eu, eig; also from Old Norse

ham; but liable to confusion

with hamm, enclosed possession,

fold

ham + stede

FEATURES OF OCCUPANCE: GENERAL

III

Element

tret, etc.
thorp (c), throp

ton

wood

Element

booth

ergh

force, foss

garth

gill

ings

thorp (e)
thwaite

toft
ton

Meaning

MI action, later any made-up road
village or hamlet

wide range of meaning; primarily
enclosed piece of ground, then en-
closed land with dwellings on it,
i.e. estate, manor, vill, village
wood

Derivation

]?orp, ]?rop; but many thorpes
were so named by the Danes
-tun; but possibly also from
Scandinavian

wudu

B. ELEMENTS OF SCANDINAVIAN ORIGIN

Meaning

originally shieling (centre for sum-
mer pasture)
village, town (Danish)
homestead (Norse)
see booth
island, etc.
waterfall
enclosure
ravine, cleft

meadow-land, especially in marshy
places (North-country dialect
word)

a hamlet or daughter-settlement
dependent on, i.e. colonized from,
an older village

clearing; but many thwaites date
from the thirteenth and fourteenth
centuries, the word having passed
into local speech

piece of land, homestead; often
clearing

see under "Elements of Old English
Origin"; sometimes from Scandin-
avian tun

Derivation

both (Danish)
buth( Old Norse)
byr

fors
gardr
geil
cng

thorp

]?veit

topt

C. ELEMENTS OF CELTIC (CORNISH) ORIGIN

Element Meaning

pen head, top, summit, promontory;

but also possibly from Old English
perm, pen or enclosure

pol stream (cf. Welsh poll); also pos-

sibly from Old English pol and
Welsh pwll

tre, trev homestead, village, town; often

hamlet

Derivation

pen

a pool, deep place in a river

Wooldridge has shown that three phases of Anglo-Saxon settlement
can be distinguished, each associated with specific place-name elements:
the entrance phase of skeletal infiltration to the areas of early settlement,
the expansion phase of secondary colonization around and within the
early settled tracts, and of territorial expansion by conquest, and the
terminal phase, when the settlement plan was filled out and assumed
for the time being a relatively static condition. Place-names are most

112 MAP INTERPRETATION

informative about the entrance phase, which is recorded in many
names of early form, including -ing names (where these are derived
from names in -ingas), a large number of names in -ham and in the
combinations -ingaham and -ingham. Naturally enough, these early
elements are commonest on the eastern side of the country. Scandin-
avian forms occur widely on the north-eastern side of a line joining
the Cheshire Dee and the Lea River, i.e. north-east of Watling Street.
Norse elements are frequent on the western side from Cumberland
to the Wirral, and also extend farther southwards along the coast;
the rest of the last formed part of the Danelaw and Danish forms are
expectedly frequent. In the western highlands, Celtic (including
Gaelic) names are abundant along the old frontiers of conquest.

NOTES AND REFERENCES

Much information relevant to the subject of this chapter is to be found
in

H. C. DARBY. An Historical Geography of England Before A.D. 1800.
University Press, Cambridge, 1936. (Sec especially Chapter V, The
Economic Geography of England ', A.D. 1000-1250, by H. C. DARBY,
p. 165.)

Settlement patterns and forms are discussed in

A. DEMANGEON. "L'Habitat Rural." Comptes Rendus, Congres
Internal, de Geographic, Cairo, 1925. Tome IV. Reprinted in A.
DEMANGEON, Problemes de Geographic Humaine. Colin, Paris, 1947,
p. 156-

A. DEMANGEON. "La Geographic de 1'Habitat Rural." Reprinted
from Ann. de Geog., 199 and 200, xxxvi, 1927, in Problemes de Geographic
Humaine^ (above), p. 159.

A. DEMANGEON. "Types de Pcuplement Rural en France." Re-
printed (torn Ann. de Gt ( og., 271, xlviii, 1939, in Problemes de Geographie
Humaine (above), p. 291.

W. PAGE. "Notes on the Types of English Villages and their
Distribution." Antiquity, i, 1927, p. 447.

F. SEEBOHM. The English Village Community. Longmans, Green,
London, 1883.

Seebohm's work contains much helpful material on the relationship
between agricultural systems and the structure of the community. On
this point, reference may also be made to

C. S. and C. S. ORWIN. The Open Fields, Oxford. Clarendon Press,
Oxford, 1938.

FEATURES OF OCCUPANCE: GENERAL 113

C. S. ORWIN. A History of English Farming. Nelson, London, 1949.
H. PEAKE. The English Village. Benn, London, 1922.

Accounts of certain processes which have greatly modified the
cultural landscape occur in

H. C. DARBY. "The Clearing of the English Woodlands." Geo-
graphy, xxxvi, 1951, p. 71.

H. C. DARBY. "The Changing English Landscape." Geogr. Journ.,
cxvii, 1951, p. 377.

See also

M. W. BERESFORD. "The Lost Villages of Medieval England."
Geogr. Journ. , cxvii, 1951, p. 129.

Individual studies include

E. ESTYN EVANS. Irish Heritage. Tempest, Dundalk, 1945. (See
p. 47 ff., Figs. 13 and 14, for evidence of rapid dispersion.)

E. Jones. "Some Aspects of the Study of Settlement in Britain."
Advancement of Science, viii, 1951, p. 59.

B. M. SWAINSON. "Dispersion and Agglomeration of Rural
Settlement in Somerset." Geography, xxix, 1944, p. I.

In the study and interpretation of place-names, one should make
use whenever possible of the relevant county handbook of the Place-
name Society. The whole series has not yet appeared, however. A
great deal of help is obtainable from

A. MAWER and F. M. STENTON. Introduction to the Survey of English
Place-names. (English Place-name Society, Vol. I, Part i.) University
Press, Cambridge, 1925.

A. MAWER. The Chief Elements Used in English Place-names. (English
Place-name Society, Vol. I, Part 2.) University Press, Cambridge,
1924.

E. EKWALL. The Concise Oxford Dictionary of English Place-names.
Clarendon Press, Oxford, 1951. (Contains a useful introductory
study.)

ORDNANCE SURVEY. Glossary of the Most Common Gaelic Words used
on Ordnance Survey Maps. (Out of print.)

ORDNANCE SURVEY. Glossary of the Most Common Welsh Words
used on Ordnance Survey Maps. 1949.

These works have been freely drawn on in compiling the Table
given above.

The scheme of land-use classification adopted by the Land Utilization
Survey is set out in
L. DUDLEY STAMP and E. C. WILLATTS. The Land Utilization Survey

114 MAP INTERPRETATION

of Britain; an Outline Description of the First Twelve One-inch Maps.
Land Utilization Survey, London, 1935, pp. 5-7.

The extremely great geographical value of the Soil Surveys will be
conceded by all with experience of working over ground for which
soil maps are published. The pattern of soil-distributions shown on the
maps is complex, but complex reality is no excuse for neglect or scorn.
Among the soil maps already available on 1/63,360 are Sheet 296
(Glastonbury) of the Soil Survey of England and Wales, which relates
to part of the area represented on Sheet 165 of the O.S. series, and
Sheet 22 (Kilmarnock) of the Soil Survey of Scotland, which refers
mainly to drumlin country. Both of these maps are strongly
recommended.

CHAPTER XII
RURAL SETTLEMENT STUDIES

It sited was in fruitful soil of old SPENSER

MAPS: O.S. 1/63,360 (SEVENTH SERIES) SHEETS 114, 144, 167,
186; O.S. TOURIST MAP, 1/63,360 (LORN AND LOCHABER);
O.S. 1/25,000, SHEET NC/y6. U.S. GEOLOGICAL SURVEY,
1/62,500 (FINAL COUNTY (ARIZONA) CASA GRANDE QUAD-
RANGLE)

I. Dominantly Nucleated

O.S. 1/63,360 (SEVENTH SERIES) SHEET 144
(CHELTENHAM AND EVESHAM)

THE area shown on Sheet 144 is dominantly one of nucleated rural
settlement. There are towns, Cheltenham sizable and the rest small,
as well as numerous scattered farms, but the village is the typical unit.
This fact does not preclude differences between stows in the size and
form of village or in the relation of settlement to ground.

To the east of the Cotswolds, that is to say roughly east of the Grid
Line 16, numerous villages of no great size lie in the valley stows
drained by the Windrush, Evenlode, and Avon. Many sites near the
foot of hills obviously had the original advantage of water from
springs, for example the three Rissingtons (1921, 1919, 1917) and the
corresponding villages of Icomb, Westcotc, Idbury, and Fifield on the
eastern side of the same hill. Streams can be seen to head near, and to
flow away from, the base of the slope, while the name Springhill Farm
(2516) is significant. Villages of this kind are conveniently styled wet-
point settlements. If the rocks forming the high ground were perme-
able throughout, a definite line of springs would be looked for at the
geological boundary. Villages grouped around springs would then
constitute spring-line villages or scarp-foot villages, according to
whether emphasis were placed on relation to water-supply or relation
to terrain as a whole. In this locality, however, the resistant rocks are
variously permeable in themselves as well as being separated by clays,
so that springs occur at various elevations on the sides of hills and in

Il6 MAP INTERPRETATION

belts rather than in lines. The wells 1 of a single village may tap a
number of water-bearing horizons.

Also in this eastern part occur villages on or very close to streams,
particularly to the Stour, in valley-bottom sites. Halford stands where
the Fosse Way crosses the river, but the remaining valley-bottom
villages on the Stour appear to have little relation to the control of
crossings: in the first place, not every village has its road bridge, while
secondly, there are seventeen crossings of some kind in the 12 miles of
valley between Tidmington (2638) and the confluence with the Avon.
Evidently the difficulty of crossing was insufficient to concentrate
traffic at a few points only. Mills and weirs show that these riparian
sites had the advantage of potential water power, which was in due
course developed and must have assisted the survival and growth of
the mill-owning villages; but potential water power cannot alone
account for the selection of the sites in question. Place-name evidence
shows that the settlers were Anglo-Saxons. It follows that the stow
was found generally suited to cultivation. Field investigation is
required, since published geological maps are inadequate to the pur-
pose, to prove that these valley-bottom villages are, in fact, sited on
patches of low terrace. Faint pointers towards this conclusion arc
given by the sites of Halford (2645) and Tredington (2543), each in
the inner side of a meander-loop : such siting confirms the suggestion
of the contours that the Stour is ingrown, but there is nothing to show
whether the ingrowth has been continuous or spasmodic.

The shortcomings of the topographical map are again evident a
little? to the south, where Stretton-on-Fosse (2238), Todenham (2436),
the Wolfords (2434, 2635), and Barton on the Heath are sited on hills
capped by sandy and gravelly drift, which provides a perched water-
table and drier ground than the surrounding claylands. Only the
"Heath" element in the last-mentioned place-name provides a dubious
clue.

Along the north-western side of the Cotswolds wet-point settle-
ments are strung out near the scarp-foot, from Mickleton (1643) to
Great Witcombe (9114). They are linked by the Stratford-on- Avon-
Cheltenham-Bath road, but particularly north-eastwards of Winch-
comb (0228) tend to lie on one side of it or at a little distance.
In this north-eastern part the typical village site occurs at the mouth
of a combe, the short re-entrant valley of a scarp stream. In the field,

1 Water-supply schemes are progressively reducing the number of small wells in use,
but the historical argument remains unaffected.

Photograph: films, Aero Ltd
(A) AERIAL VIEW or THE MENDIP PLATEAU, WITH CHEDDAR GORGE (CENTRE)

Photograph: Aerofilmi, Ltd.

(B) PART OF THE FENLAND, NEAR BOSTON, SEEN FROM THE AIR

Note the regularized Witham, the rectangular field pattern, and dispersed
settlement.

PLATE III

(A) AERIAL VIEW or STOW ON THE WOLD

Photograph: Aerofilms, Ltd.

B) PART OF MERTHYR TYDFIL: TERRACED HOUSING,
STEELWORKS, AND TIP-HEAPS

PLATE IV

117

RURAL SETTLEMENT STUDIES 117

Buckland (0837) an< l Stanton (0634) in particular seem to be almost
enclosed by hills. Broadway, by far the largest member of the group,
lies where the London-Evesham road descends the scarp-face; the
influence exerted by traffic on this road on the growth of Broadway is
well shown by the remarkable elongation of the village in the direction
of movement.

In the dissected plateau stow of the Cotswolds, nucleated settlement
avoids the tops of interfluves, and the small villages are located chiefly
at valley-bottom sites as wet-point settlements, where roads cross the
larger valleys, or near the heads of tributaries. Examples of the former
class are Temple Guiting (0928) and Naunton (1123), lying within the
incised valley of the Windrush, and of the second Notgrove (1020)
and Aston Blank 1 (1219) near the heads of feeder streams. It is notice-
able that some valley-bottom villages, narrowly restricted by the
valley walls, have grown along the valley. Withiiigton (0315) is
especially interesting : the village extends in an arc across the Coin, in
conformity with the shape of the meandering valley.

In the claylands flanking the Avon different factors operate. The
scarp-foot settlements encircling Bredon Hill, and similar settlements
associated with similar but smaller hills to the south, may be regarded
as outliers of the main belt of scarp-foot villages, just as the hills them-
selves are outliers of the Cotswolds. For the rest, there is a noticeable
tendency for villages to be sited on the higher rather than on the lower
ground, as dry-point settlements. Among the clearest examples are
Pebworth (1347) and Long Marston (1548), each on the crest of a low
mterfluve, and Aston Somerville (0438) on one side of a low spur. In
reality this sub-edge country received a great deal of soliflucted material
during the Pleistocene. The deposits are by no means without influence
on the agricultural quality of the ground and on its suitability for
settlement, but the topographical map fails to indicate either these or the
true glacial deposits which are also present.

Close to the Avon the record is clearer. Physical interpretation has
indicated that terraces are present. It is now seen that, on the flood-
plain proper, buildings and roads are alike rare. Below Evesham,
where the Avon describes great loops, nucleated settlements are
systematically disposed on the intervening spurs: Charlton (0145),
Fladbury (9946), Wick (9645), the town Pershore (9546), the hamlet
Pensham (9444), Birlingham (9343), and Eckingtoii (9241). The
immediate neighbourhood of Fladbury illustrates particularly well the
1 Locally called Cold Aston.

$-(.5196)

Il8 MAP INTERPRETATION

manner in which settlement and trackways tend to cease at the boun-
dary of the flood-plain, as here defined by Oxton Ditch and Lench
Ditch. A contrasted factor of siting is the navigability of the river,
demonstrated by weirs and locks. A number of villages are placed
where meanders impinge on the higher, firmer ground, so that deep
water can be reached without crossing the flood-plain. Such sites are
those of Wyre Piddle (9647) and Cropthorne (0045) in the reach just
discussed, and of Bidford on Avon (1051) farther upstream.

Crossing-places below Evesham are differently placed from those
above. In the lower reach Eckington Bridge (9242), Pershore Bridge
(9545), and Jubilee Bridge (0045) he near points of inflexion of the
great river loops, that is to say at about where the direction of curvature
is changing. Since the flood-plain alluvium is disposed in crescents on
the inside of the loops, crossings at such points are made between the
tips of adjacent crescents and involve the minimum traverse of ill-
drained ground. Upstream of Evesham the valley bottom has been
more extensively cleared by the sweeping meanders, with the result
that a wider and more regular belt of flood-plain must be traversed.
The ford at 065470 and the two road bridges (1052, 1453) carry roads
which descend meander-spurs where the ground is at least a little
dryer than on most of the flood-plain. In addition, place-names
along this reach show that a number of settlements grew in association
with fording-places Salford Priors, Bidford, and Welford.

Although many of the villages in the area of this map sheet are small,
the most casual inspection in the field leaves no doubt that, as stated at
the outset, rural settlement is dominantly nucleated. Because of their
smallness, little useful comment can be made on the form of villages.
Most are rather formless agglomerations, 1 a well-defined shape being
imposed only in exceptional cases by close adaptation to the qualities
of the site. An influence not to be read from the map was the
decline of the Cotswold production of wool and woollen cloth, a
decline which has been reflected in the declining population of the
plateau.

The result of secondary dispersion, on the other hand, is easily
perceived. If it is assumed on the basis of place-name suffixes that many
of the existing settlements were established by Anglo-Saxons, and on
general grounds by way of corollary that openfield agriculture was
practised for many centuries, it appears probable that the single farms
are in large part the result of movement outwards from pre-existing
1 Haufendorfer of continental writers.

RURAL SETTLEMENT STUDIES 119

agglomerations. 1 The assumptions and the inference are in fact
justified by historical evidence. As in many other respects the map
record is incomplete: not every farm which exists in the field has the
word "farm" attached on the map, besides which it is common to
find some farmhouses within the villages themselves. Nevertheless,
the available information is sufficient to show the presence of outlying
farms, for example, four named in the parish of Stretton on Fosse
(2238) and four also in Aston Blank parish (1219).

Quite apart from the naming of farms, it goes without saying that a
widely cleared, well-peopled countryside of this kind is very largely
in tillage or pasture. The precise mode of land use, which is itself
liable to considerable change in response to economic pressure, can
be read only when it is indicated by a symbol. The hangers of wood
on steep slopes may well represent the remnants of a former extensive
woodland cover in the Cotswolds, although many small patches with
rectilinear bounds and lying on flatter ground are likely to be planta-
tions (cf. Stanway Ash (0832) ). Again, in formerly wooded country
it is often found that a park is partly ringed by narrow woods, as at
Northwick (1636), Trafalgar Farm (2911), and Bcmbro (1027). The
last two have evidently been put to agriculture while all three may well
represent early clearance of woodland granted by the Crown to an
individual. Recent clearance appears in the combination of the
symbols for trees and for rough grazing, for example, in the kilometre
square 1226; but Brockeridge Common (8838) is a more doubtful
case because of the name and is likely to be covered by heath or
scrub.

It is well known that the Vale of Evesham is noted for specialized
cash crops, market-garden produce and orchard fruit. Such specializa-
tion is due to a whole complex of factors which lie in part in the proper
field of the economist. The geographer engaged in map interpretation
should confine himself to remarking the facts shown on the map, in
this instance the considerable areas under orchards, and to the tacit
assumption that the local climate, soil, and aspect are favourable to
tree-fruit growing. The chief concentration of orchards is around
Evesham, except on the lowest ground, which has the double dis-
advantage of poor drainage and of liability to frost 2 ; extensions from
this central area run up and down the Avon. Where the orchards are

1 This dispersion was of course the concomitant of enclosure.

1 As might be expected, some orchards do lie within frost-pockets (see the work of
RAYMOND BUSH cited at the end of this chapter).

I2O MAP INTERPRETATION

more scattered it is seen that (for obvious reasons) they tend to abut on
metalled roads. Away from the river, or more properly from the
terrace spreads, the scatter of orchards is less dense, but continues up to
and includes the line of scarp-foot villages, with incipient penetration
of the upper Stour valley. The abrupt cessation of orchards at the
scarp-foot line is a fair sign that here marked agricultural and physical
boundaries coincide.

2. Dominantly Dispersed

O.S. 1/63,360 (SEVENTH SERIES) SHEET 186
(BODMIN AND LAUNCESTON)

The very frequent place-name element "Tre-," with the less common
"Pen-" and "Pol-" show that the pattern of rural settlement here has
developed not from Anglian, Danish, or Scandinavian origins, but
from Celtic. Much of Highland Britain, including this part, resisted
the militant settlers of the Dark Ages, so that the array of villages
representing original nucleation, with intervening farms representing
secondary dispersion, so familiar in parts of the English Plain, is not to
be sought on this map. At the same time it may well be emphasized
that the wide dissemination observed is not necessarily the effect of
primary dispersion. It is true that the generally impermeable rocks
and the heavy rainfall of Highland Britain made water-supply no
great problem, and that therefore wet-point settlements grouped
round springs are unlikely to occur in numbers, but the matter is far
mor$ complicated than this. Until the agrarian economy of Celtic
times is better understood, it is best to accept the present fact of
dissemination as it stands, noting at the same time that several recent
studies point to the law of gavelkind as a powerful factor in causing
dispersion in parts of the Celtic west.

Disseminated rural settlement may be dominated either by the
hamlet or by the single farm. In the area of Sheet 186 the hamlet
appears typical to the north-east of Bodmin Moor, while separate
farms are most numerous in the south, say, south of Grid Line 65. The
hamlets are characteristically sited high up on the sides of tributary
valleys, close to the subdued tops of the interfluves but at the same time
in sheltered positions. True valley-bottom sites have rarely been
selected. The many single farmsteads of the south, being presumably
placed somewhere near the centres of their respective land, occupy
less obviously selected sites, but in the field are nevertheless seen to take

RURAL SETTLEMENT STUDIES 121

advantage of folds of the ground too small to be represented by the
available contours.

The major part of Bodmin Moor may be taken as unimproved
rough grazing on heath or bog. Single farms occur on the flanks and
avoid the highest ground. From the manner in which rectilinear
patches free of the moorland symbol are interspersed with land which
is still heath, one may infer that piecemeal enclosure and improvement
has gone on (cf. "New Closes" (2071) ). If this is so, the single farms
established in the cleared areas would represent primary dispersion,
in the familiar guise of squatter settlement on the margins of poor land.
It is noticeable that in prehistoric times the higher parts were more
hospitable than they now are: besides many tumuli and various arrays
of standing stones, there are groups of hut circles and one "Ancient
Village" (2376) to prove former extensive habitation.

Any attempt to explain the location of villages in this area by local
nodality is bound to fail. In the first place, as nearly all are so diminutive
that they lie very close to the ill-defined division between hamlet and
village, any factors tending to increase their size seem to have operated
but feebly. Secondly, there appears little to choose between the road
connections of most villages and those of a large number of hamlets.
The fenced roads, which cover the whole area, except for Bodmin
Moor, in a close network, are mostly classified as "Under 14 ft. of
metalling, bad" or "Minor roads," and some villages arc served by no
roads of higher grade, for example, Treneglos (2088), St. Clether (2084),
and Cardinham (1268). On the map, the two obvious differences
between villages and hamlets are that the former are named in larger
and different type wliich, indeed, is the easiest means of identifying
them and that each has a church. It would appear, therefore, that
in this area the settlements to be recognized as villages arc ecclesiastical
centres, and presumably also centres of local administration, of parishes
where most dwellings lie away from the centre. It is for the historical
geographer and others to discover whether, when the parishes were
delimited, settlement was more strongly nucleated.

Dispersed agricultural settlement extends right to the coast, with
diminutive villages at intervals among the farms and hamlets. As there
is very little low ground immediately behind the shore, small groups
of buildings find themsleves somewhat precariously placed along the
base of the cliff (Portwrinkle (3553), Downderry (3154)) or uncom-
fortably confined in small youthful valleys (Polperro (2151), Polruan
(1251)). Pentewan (0247) also clings to the foot of the slope behind,

122 MAP INTERPRETATION

probably because the flat ground is too sandy for building. The
improved harbour here, like those at Charlestown (0451) and Par
(0752), which has rail connections, evidently provide outlets for the
china clay fields north of St. Austell.

In conclusion one may turn to the Fowey to remark the very small
agglomerations characteristically located at the head of the lesser
inlets: Lerryn (1457), Penpoll (1454), and Pont (1451). Because of
the smaller volumes of the tributary streams these sites are relatively
little affected by silting, and at high tide can be reached by fishing
craft and small coasting steamers.

3. Rural Settlement in the Fenland

O.S. 1/63,360 (SEVENTH SERIES) SHEET 114
(BOSTON AND SKEGNESS)

Each of the several kinds of stow represented on this map is clearly
distinguished from the others by differences of physique, of drainage,
of form and distribution of settlement, and of the road nexus. However
complex the relation between man and land, the sharp differentiation
observed here leaves no possible doubt that the terrain has exerted a
powerful influence on the history and nature of occupance: the
effect of occupance on the ground is no less clear.

The only land above 50 ft. O.D. lies in the north, outside the fenland
boundary. This stow has already been noted as one of fairly weak
rocks, generally impermeable and heavily dissected into country of
many wide valleys and small divides. The irregular parish boundaries
and the winding roads show early occupation and unplanned develop-
ment. Nevertheless a definite connection is to be observed between
the form gf the ground on the one hand, and the location of settlement
and the course of roads on the other. The roads avoid valley bottoms,
except to cross a stream, running instead near the foot of the valley
walls or along the crests of divides; villages also show a general pre-
ference for the high ground. It is of the greatest importance to note
that the finer details of location would amply repay intensive study
in the field, for the place-names Gravel Pit Hill (3669), Sandhill Farm
(4364), and the Brick and Tile Works at Hundleby (3866) mean that
marked local differences are certain to occur in the qualities of the
waste-mantle and of the soils developed thereon. Although the soil
survey of Britain is not yet far advanced, enough is known to justify
the statement that, unless variations of soil can be taken into account,

RURAL SETTLEMENT STUDIES 123

the pattern of rural settlement cannot be fully understood. Since a
vast quantity of fundamental data must thus be awaited, and since in
any case soils are not recorded on the topographical map, the inter-
preter needs to discuss rural settlement with these reservations in mind,
and should not fail to support the general argument by drawing
attention to such facts as those indicated.

Four groups of tumuli (4072, 4171, 4371, 4471) testify to early
penetration of possibly Bronze Age date. The road with long straight
stretches from 4075 to 448697 invites inquiry, especially as it is con-
tinued by a mile of field road near Burgh le Marsh (494657). The
straightness suggests a Roman origin; but the nearest site actually
described as Roman lies some miles to the south near Wainfleet All
Saints, where the ancient Salt Works locate the shoreline of some
2,000 years ago. Place-names ending in -by, which are very common,
show that Scandinavian invaders settled here in force, apparently
displacing, subjecting, or destroying earlier Anglo-Saxon inhabitants,
whose traces remain, for instance, in the place-name terminations
ham (for example, Greetham, 3070) and -ingham (for example, Hag-
worthingham, 3469). Names in -ton are less helpful, since, where
Scandinavian and Anglo-Saxon forms occur in association, this
termination may be of mixed derivation.

As in the Cotswolds and the Avon valley, studied on Sheet 144,
primary nucleation may be taken as highly probable, with Dark Age
villages sited on patches of drier soils above the wooded valleys. The
outlying farms would then once more represent secondary dispersion.
Another mode of expansion may be interred from the thorpes (for
example, Mid Thorpe and Far Thorpe, 2673 an d 2674), which are
likely to have been founded as daughter-settlements some way from
the original village.

On the seaward side, this northern stow is bounded by the fen,
wliich is thrown into prominence, on the map, by its sub-rectangular
pattern of roads, parish boundaries, and drainage channels. (In the
field the physiographic boundary is sharply and narrowly defined.)
This regularity in the features of occupance (Plate Ills), the scale of
the drainage works effected, and the obliteration of most of the natural
water-courses are sufficient to indicate that reclamation came relatively
late in historical times, even were the general facts less well known.
It is permissible to take the ubiquitous name "fen," together with the
flatness and very low altitudes, as presumptive evidence of the nature
of the soil, which must be developed largely on fen peat. One may

124 MAP INTERPRETATION

pass rapidly over the form of settlement, which is characteristic of
tracts reclaimed by major drainage operations: individual farms occur
at intervals on the roadside, where the risk of flooding or waterlogging
is least, or, more rarely, at a short distance from it. The pattern is one
of recent primary dispersion. Nucleated villages occur only where
firmer ground rises through the fen deposits, as at Stickford (3560),
Stickney (3457), and Sibsey (3551). The -ey element shows that at one
time two of these sites appeared as "islands" in the fen. Differences of
level, as shown by the spot heights, are very slight but significant:
the main road linking the three villages takes an irregular line, while
the houses are distinctly grouped into street-villages.

East and north of Wainfleet and Burgh le Marsh the ill-drained belt
is narrower than to the south-west, and less obviously modified by
the work of man, but here, too, a contrasted stow intervenes between
the low hill country and the coastal belt.

It was observed in the discussion of physical distributions that the
coastal belt itself comprises two tracts, one of salt marsh already
reclaimed or in process of reclamation, and one of firmer ground where
settlement is old-established. In the stows of reclaimed marsh, between
Gibraltar Point (5558) and Freiston Shore (3943), between Freiston
Shore and the Welland, and between the Welland and the Ncne, the
pattern of occupance recalls that of the fenland: regular drainage
channels, straight roads, and the single farms of recent primary dis-
persion are again observed. Progressive reclamation is attested by the
"Old Sea Bank" (4955) and the "Roman Bank" marked in several
places (but possibly later than Roman, despite the name), as well as by
the "Old Marsh" and "New Marsh" (5158, 5257). The Roman Salt
Works have been mentioned before. Note also the strip of marsh
represented above high-water mark.

The next tract inland, between the salt marsh and the drained fen,
stands little above the sea. It is nevertheless one of firmer ground,
which is well settled with old-established villages and a complex net
of local roads, and which retains some elements of the natural drainage
system. Individual stows in this tract vary somewhat from the general
pattern. In the southernmost, immediately south of the outer belt of
reclaimed marsh between the Nene and Welland outfalls, the road
network is more open and rather more regular than in the others,
while the largest settlement is a hamlet. Between Wainfleet and the
extreme south-west of the area mapped, the belt of firm ground is as
much as five miles wide in places. It is traversed by reclaimed valley

RURAL SETTLEMENT STUDIES 125

bottoms and inlets which formerly connected the marsh of the shore-
line with the fen behind: Bicker Haven (2533), The Haven (3540),
Friskney Low Ground (4454) and the neighbouring Wrangle Low
Ground, and the mouth of the Steeping River. Scandinavian place-
name elements such as toft, beck, wick, and thorpe indicate one period of
conquest and settlement. The -ey of Friskney bears witness to the
working distinction made in early times between the better and the
less well-drained land. The several "Low Grounds," "Commons,"
and "Fens" must have been held communally by the villages after
which they are named, under a medieval system of agronomy.
Many minor place-names imply a close relation between man and
land, and a keen appreciation of the varied qualities of the setting,
for example, the common "dyke," Fishmere End (2837), Freiston
Shore where the firm ground reaches the sea, the ings which are, in
this tract, patches of meadow in the marshland, and the name "hill"
given to very low rises. After what has previously been said, it is
unnecessary to stress the facts that the settlement pattern results from
primary nucleation, with secondary (or perhaps intercalated) disper-
sion, and that villages tend to straggle along almost imperceptible
crests.

Between Wainfleet and Burgh a stow of firm ground provides a
link with the hilly tract in the north-west. Once more, it is marked
by irregular roads and some nucleated villages. The pattern recurs
north of Gibraltar Point, where the dryer ground abuts directly on
the sea. Salt marsh is replaced by a sandy beach, with which is asso-
ciated the only new element that requires attention, i.e. the signs of
recent growth in the villages of Ingoldmells and Chapel St. Leonards.
Planned streets of modern layout at 570692 and 559731 indicate that
these two small centres have shared, to a limited extent, in the modern
growth of seaside resorts which has been responsible for the present
size and form of Skegness.

4. Rural Settlement in Glaciated Highland

O.S. TOURIST MAP 1/63,360 (LORN AND LOCHABER);
O.S. 1/25,000, SHEET NC/y6

Both the distribution and the form of settlement in glaciated mountains
are powerfully, although not exclusively, influenced by the setting.
It is in the nature of such tracts to include large negative areas, devoid
of soil, very steeply sloping, inaccessible, or climatically formidable.

126 MAP INTERPRETATION

Within the occupied parts, however, there is room for considerable
variety, according to the qualities of the ground, the system of agri-
culture practised, and the relations of the tract with other tracts outside.
A general concentration of settlement in the valleys is inevitable.
One may also safely postulate for earlier times an almost complete
dependence on subsistence farming, in which the relative importance
of livestock and tillage varied with the abundance of good land and the
qualities of the climate. Since good tillable land is generally scarce,
one may look for a selective concentration of rural settlement on
favourable sites.

A marked contrast between adjacent sites is characteristic of glaciated
highlands, and ensures that some influential facts of the environment
can be clearly read. The problem of aspect, however, demands careful
analysis for a complete solution. While the broad difference between
the northern and southern sides of a glacial trough is obvious enough,
Garnett has shown that significant differences of aspect between the
sites on one side only can be discovered from the map when certain
techniques are employed. As the required treatment is somewhat
lengthy, the reader is referred to the works cited at the end of this
chapter for a description of method and results ; but it must be borne
in mind that, unless analysis of this kind is attempted, questions of
aspect can be discussed only in general terms.

The various environmental factors of location are likely to operate
most powerfully near the upper limit of settlement, where conditions
are most difficult. Local differences of aspect serve to modify the
general limitation imposed by low summer temperatures, severe
winters, and long duration of snow cover, which are themselves
partly determined by height and latitude. Some Alpine villages are
situated far up the valleys, at heights which in Scotland would be on
or above the peaks. In the difficult tract represented on Sheet 47, the
sparse settlement is not confined merely to the valleys but nearly
everywhere to the valley bottoms, and the problem of distribution is
to that extent simplified.

Rural settlement occurs both in dispersed and in nucleated form,
but where grouping occurs it is dissimilar to that of the English
Lowlands. Trislaig (0874), Blarmachfoldach (0969), South Garvan
(9977), and Inverroy (2581) may be taken as examples. In each the
houses straggle along the glen. Trislaig has already been identified as
occupying a patch of raised beach. The houses lie along the old shore-
line, with cleared land in front and open moor rising behind. Assuming

RURAL SETTLEMENT STUDIES 127

this to be an agricultural settlement, one may suppose that the limited
area of flat ground, with its soil developed from raised beach deposits,
is likely to carry tillage crops and hay, while livestock is pastured on
the lower hillside. This is so in actuality (1950) : the organization is
that of crofting, the form of subsistence agriculture proper to the
remote Highlands. However, even though the environment imposes
strict limits on the possible organization of farming, marked differences
are possible within those limits, and it must not be assumed that the
crofting system holds good everywhere. One may perhaps repeat the
caution that the topographical map shows only the distribution of
buildings and the extent of cleared land, with, at some sites, physical
features likely to provide tillable soil. Furthermore, the map alone
cannot show how greatly the system of crofting may have been
modified: in its fullest development, crofting involves an economic
isolation and a low standard of living which are both rare to-day. A
further characteristic, again not to be read from the map, is the black
house, a single-storeyed chimneyless building containing a family of
crofters at one end and the livestock at the other. The type form may
be modified by the addition of chimneys, and by the complete
partitioning of the two ends, but the general structure is distinctive
enough, as at Trislaig.

At South Garvan there is apparently a less abrupt change from flat
to slope, and enclosures have been cleared and improved as high as the
loo-ft. contour. It would appear from the irregular moorland
boundary (better shown at Blaich (0377) ) that the land has been taken
in piecemeal. At Blarmachfoldach the clearance has, in total at least,
been more regular, extending upwards from the stream bank well past
the line of houses to a smooth fence at the edge of the moor. Inverroy
is different again. There are two lines of houses, one at a sharp break of
slope between the valley wall and the valley floor, which also coincides
approximately with the limit of moorland, and one below the main
road, at the edge of an alluvial valley flat or haugh.

Alluvium provides other sites capable of cultivation in deltas and
delta-fans, which must be numerous where so many ungraded streams
are well supplied with rock-waste. Deltaic sites at the heads of lakes
or at the mouths of large tributaries are identified without difficulty.
At many of them, buildings lie near the apex where natural drainage is
good, with easy access to, and command of, the whole cone, and safe
from flood (cf. Fassfern (0279) ). Much of the dispersed settlement is
also located on sites of this kind, but the cones are frequently too

128 MAP INTERPRETATION

small for the map to show (cf. Gucsachan (8879), obviously on deltas
built into Loch Shiel, with Tighnocomairc (9469), where there is
actually a small corrom or delta-fan).

A scale of 1/63,360 is able to reveal that some houses are considerably
larger than those of the crofting settlements. The contrast in size, and
also in plan, is well shown by Sallachan (9863) and the Crofts of
Sallachan to the east of it. The map does no more than record the
numerous large "Houses": the social order which produced them is a
subject for historical study, except that one may regard them as a
likely outcome of an aristocratic or squirearchical system.

The relation of certain Highland settlements to the land on which
they lie is illustrated in more detail on the O.S. 1/25,000, Sheet 29/76.
Nearly all the habitations shown arc comprised in five groups:
Armadale (7864), Kirktomy (7463), Swordly (7363), Farr (7263), and
Betty hill (7062). On this map, where field boundaries are shown, it
is at once seen that the houses in Armadale are strung out along the
lower hillside, with small enclosures carved out of the moor above or
running down to the stream and the shoreline below. The moorland
symbol within some enclosures indicates either that the part of the
land proved intractable, or that it has been allowed to revert, for
example because of a decline in population. At Kirktomy and at
Farr, buildings lie roughly at the break in slope between the steep
hillside and the flatter low ground a relationship already observed
on the 1/63,360 sheet. Bettyhill is the only group which has expanded
sufficiently to require an attack on the hilltops. Here the subdued
summits at c. 300 ft. O.D. have been enclosed and to some extent
cleared in the significantly named Newlands: the spread of clearing
has evidently been associated with the establishment of dwellings at
this higher level. One might suggest that the vigour of Bettyhill, by
comparison with the other groups, has something to do with its
situation on the main road (cf. the presence of an hotel). Because of
the expansion, it is not clear whether Bettyhill may not include more
than one original group. Farr, Armadale, and Kirktomy are more
truly representative of a settlement form indigenous to the Scottish
Highlands, the clachan a loose agglomeration of precisely the kind
seen here, but often no more than a small, loose-knit hamlet. On this
sheet the name "clachan" appears once only, applied to a small group
of buildings at the eastern end of the modern Bettyhill.

To explain the origin of the clachan, it would be necessary to take
account not only of a former dependence on subsistence farming, but

RURAL SETTLEMENT STUDIES 129

also of the former social structure of clans, and of family groups within
the clans. Thus it is seen that even here, in country where physical
circumstances impose strict limits or severe difficulties on land use, the
pattern of rural settlement cannot be understood without appeal to
additional factors.

5. Recent Primary Dispersion

U.S. GEOLOGICAL SURVEY, 1/62,500 (FINAL COUNTY
(ARIZONA) CAS A GRANDE QUADRANGLE)

The pattern of rural settlement here is geometrical. Systems of
numbered squares, in which the rural habitations lie, relate to the
arbitrary division and allocation of land at the time of settlement by
white men: each square is a range, i mile by I mile in size, with
a block of thirty-six squares making up a township. The town of
Casa Grande and the small, rather loosely grouped smaller settlement
of Chiu-Chiuschu in the south show a tendency to develop converging
roads, while other short lengths of road lead to wells, but in general
the rectangular pattern of land holdings is repeated in, and emphasized
by, a rectangular net of minor roads. Most houses lie by the roadside.
With such an arrangement nucleation can scarcely begin, for, if a
square block of four ranges is owned or held by four farmers, the
largest possible grouping at the central crossroads would include
only four farms plus the dwellings, if any, of farm-workers. The
factors militating against nucleation here are comparable to those
responsible for the primary dispersion noted in drained fen.

In this climate, proved to be arid by the nature of the landscape,
and without a supply of water for irrigation, farming is certain to
be pastoral. The wells are needed for domestic water-supply and for
watering stock. They tap the water-table in the deposits of the alluvial
plain, which is fed by percolation from intermittent streams and occa-
sional rainstorms. Note the absence of wells on the pediment in the
south-west, where settlement may have been deterred by the lack of
ground-water in a very thin cover of rock-waste. Below the Sacaton
Mountains, on the other hand, superficial deposits seem to extend
across the foot of the pediment, for wells have been sunk as high as the
i,5OO-ft. contour. It should be remarked that, although water-supply
is possibly the most pressing of all problems for settlers in this tract,
the special qualities of the water-table in the alluvium and the use of
modern techniques of well-sinking result in the almost complete

RURAL SETTLEMENT STUDIES 131

absence of anything comparable to the wet-point settlements studied
elsewhere.

6. The Transect Chart

O.S. 1/63,360 (SEVENTH SERIES) SHEET 167
(SALISBURY)

Structure, surface, and human occupance are seldom wholly uncon-
nected, however complex the relationship may prove to be. The
transect chart provides a useful and graphic means of sampling and
summarizing the various distributions, and of revealing their simil-
arities and contrasts. Because the similarities tend to be prominent in
the finished chart, it is necessary to repeat die caution against crude
determinism. The first aim of the map interpreter is to perceive and
define the association, for example, between settlement and landform,
and to discover, within the limits of his data, how closely the one is
adapted to the other. Man is everywhere the active partner, limited,
aided, or excluded by land, but never compelled.

The transect chart was developed for use in presenting the results of
intensive local survey, in which, among other tilings, rainfall, vegeta-
tion, and agriculture can be fully studied. A transect chart constructed
from a map is more limited in scope. Suitable headings are : profile of
relief, with inferred geology where possible; physical features;
nature drainage; surface utilization ; settlements and communications ;
prehistoric occupance.

The chart is constructed in the manner of a graph, with the profile
drawn across the bottom and other information arranged above in
columns, against the appropriate headings (Fig. 8). The profile refers,
of course, to a single line on the ground: additional matter may be
drawn from a belt of country, for example that defined by arbitrary
lines parallel to the line of profile. In the example given, the profile is
drawn from north to south along the Grid Line 94. Further informa-
tion is taken from the area included between the lines of easting 90
and 98. A north-south transect is chosen as crossing the grain of relief
(and of structure) approximately at right angles. Where a large feature
is crossed obliquely, as, for exaipple, the Wylye valley, it and its
associated features of settlement, etc., are "projected" on to the line
of profile and noted in the relevant column.

Since the chart is in itself a tabular descriptive summary, the facts
presented need not be elaborated. The reader is strongly advised,

132 MAP INTERPRETATION

however, to check the chart against the map in order to understand
precisely how the data have been obtained.

NOTES AND REFERENCES
On the nature and analysis of aspect and related matters, see

R. BUSH. "Frost and the Fruitgrower." Geography, xxx, 1945,
p. 80.

A. GARNETT. "Insolation, Topography, and Settlement in the Alps."
Geogr. Review^, xxv, 1935, p. 601.

A. GARNETT. Insolation and Relief. Institute of British Geographers,
Publication No. 5. George Philip, London, 1937.

The significance of scarp-foot and dip-foot belts, and of river
terraces, is demonstrated in

S. W. WOOLDRIDGE and D. L. LINTON. "The Loam-terrains of
South-eastern England in their relation to its Early History." Antiquity,
vii, 1933, p. 297.

S. W. WOOLDRIDGE and D. L. LINTON. "Some Aspects of the Saxon
Settlement in South-east England Considered in Relation to the
Geographical Background." Geography, xx, 1935, p. 161.

This question is also referred to in

S. W. WOOLDRIDGE. "The Anglo-Saxon Settlement." Published in
Historical Geography of England Before 1800. Edited by H. C. DARBY.
University Press, Cambridge, 1936, p. 88. (This essay discusses the
relation between place-name elements and the phases of penetration.)

Drainage of the Fenlands is reviewed in

H. C. DARBY. "The Draining of the Fens, A.D. 1600-1800,"
Chapter XII, p. 444 ff., in Historical Geography of England (above).

Specimens of the transect chart, together with the method of con-
struction from field survey, may be found in

C. C. FAGG and G. E. HUTCHINGS. An Introduction to Regional Sur-
veying. University Press, Cambridge, 1930. (See especially pp. 112,

5-)
C. A. SIMPSON. "A Venture in Field Geography." Geography, xxx,

1945, p. 35.

Dispersion of settlement in the Celtic west is described in the works
of EVANS and JONES listed at the end of the preceding chapter. The

RURAL SETTLEMENT STUDIES 133

studies of settlement form there specified could be usefully supple-
mented by

H. THORPE. "Some Aspects of Settlement in County Durham."
Geography, xxxv, 1950, p. 244.

H. THORPE. The Green Villages of County Durham. Institute of
British Geographers, Publication No. 15. George Philip, London,
I95i, p. 153.

io-(E. 3 i 9 6;

CHAPTER XIII
TOWNS: SITE, FORM, AND SITUATION

What is the meaning of this city? T. S. ELIOT
We have come to regard the town almost as an organic unit.
Its origin, where we can trace it, is almost completely controlled
by very definite and very local circumstances. An early growth
beyond that of its immediate neighbours is usually to be associated
with a quite definitely superior site. The modern road, the canal,
and the railway focused on such sites, and each such artificial
addition became a factor in the towns growth often more potent
than any local physical circumstance. RODWELL JONES

MAPS: O.S. 1/63,360 (SEVENTH SERIES) SHEETS 114, 144, 165,
167; O.S. 1/25,000, SHEETS SY/i8, SO/oo; O.S. TOURIST
MAP 1/63,360 (LORN AND LOCHABER); O.S. IRELAND
1/63,360, SHEET 169, ETC.; U.S. GEOLOGICAL SURVEY,
1/62,500 (FINAL COUNTY (ARIZONA) CASA GRANDE
QUADRANGLE)

As Darby has pointed out, 1 there seems to have been little continuity
between the towns of Roman Britain and those of later times. The
manner in which urban life renewed itself in the Dark Ages is complex
and obscure, but it may be said that "long before the Norman Con-
quest a force had begun to operate which was ultimately to give the
English borough its most permanent characteristic, i.e. that of a
trading centre." 2 Whatever other functions towns discharged,
marketing remained a principal concern during several hundred years,
when most people lived in the country and urban growth was slow.

Effects of Re-growth

The extent and form of the pre-industrial towns are still recognizable
in the towns of to-day. Although the fabric has been renewed, the
plan survives in the compact massing of buildings and in the narrow,

1 H. C. DARBY. An Historical Geography of England Before A.D. 1800. Edited by H. C.
DARBY. University Press, Cambridge, 1936, p. 214.
Ibid., p. 215.

134

TOWNS: SITE, FORM, AND SITUATION 135

irregular streets. The close-packed inner part, especially well marked
where the old town was confined by a wall, may be styled the core.
The remarkable increase in the country's population during the last 150
years has been effectively an increase in urban population, for some
80 per cent of the present total live in towns. This almost explosive
growth is expressed on the map in the integument, that part of a town
which surrounds the core. Streets and buildings are more widely
spaced than in the centre, forming a separate and distinctive pattern.
Not in every town is a core of medieval growth surrounded by a
later integument. A number of minor centres have neither been
industrialized nor converted into dormitories: they remain, as it
were, all core. Towns which are almost entirely the product of
industrialism seem to be all integument, but it is usually practicable
to separate the regular, crowded inner part with its nineteenth-century
layout from the more open and varied later portion.

Form

Identification on the map of the distributional patterns witliin a
town is part of the study of town form, with which interpretation is
most simply commenced. The problem involves much more, how-
ever, than the separation into integument and core: the integument
may be subdivisible, to some extent, according to function. Now the
main present functions of a town cannot be considered without
reference to past function, site, and situation. All of these items are
closely and complexly interrelated, but should not be confused: the
interpreter should be quite clear at all times which of them is being
discussed. So much regrettable confusion has in fact arisen in the past
that a few explanatory comments may be given.

Function

The function of a town implies its whole life and work, social and
economic. Various classifications of towns according to function
have been proposed, but if past functions are taken into account
function becomes confused with site, while if attention is confined to
present conditions the resulting classification is likely to be unsatis-
factory. The division into towns of extractive industry, towns of
manufacturing industry, and towns of service industry 1 is too general;
that into manufacturing, retail, diversified, wholesale, transport,

1 See R. E. DICKINSON. City Region and Regionalism. Kegan Paul London, 1947,
p. 45 ff., for a reference and comments.

136 MAP INTERPRETATION

university, resort, and retirement centres 1 tends to conceal the funda-
mental variety of urban life. Furthermore, it can hardly be disputed
that the functional classification of towns must have a statistical basis
which the map cannot provide. Thus it appears that the interpreter
is best occupied, not in trying to affix a single functional label, but in
discovering something of the structure and internal diversity of a
given town. It is in this latter task that a classification of functions is
required, in order that the different parts of a town may be properly
described. The following list, adapted from Aurousseau, may serve
the purpose

Production.

Communication and transport.

Marketing.

Residence.

Recreation.

Administration.

Culture.

Defence.

When different functions have come to be localized in different parts
of a town one may expect the map to show which are the chief
manufacturing, residential, and recreational areas, as well as the main
facilities of transport and communication. All these tend to lie within
the integument. Administration, culture, and marketing are more
typical of the core, but as a rule the only positive map evidence
consfsts in the names of individual establishments or the distinctive
representation of public buildings. A defensive role belongs to the
past and should be considered under the head of site. In any case,
medieval towns were commonly walled for defence, whether or not
they were founded as strongholds or had castles.

The question of function arises in another way. Just as there is no
universally clear distinction between hamlets and villages, so villages
grade into towns. Recent studies have revealed that some agglo-
merations fail to discharge certain functions proper to towns in general,
although they have urban administrative status. For purposes of work
with O.S. maps, it is necessary to rely on the form of lettering employed
in place-names to suggest which of the smaller centres can be classed
as towns. On the current New Popular Edition of the 1/63,360 map
the styles of lettering are not yet entirely standardized, but, since the
1 Ibid., p. 23, Footnote 2, for reference to authorities and brief comments.

TOWNS: SITE, FORM, AND SITUATION 137

smallest administrative area for which capital letters are used is the
Urban District, all places with their names in capitals have been taken
as towns for the purposes of this discussion.

Situation

All towns are nodal. The convergence of roads seems merely to
respond to the guidance of physical features, but the interpretation of
nodality nevertheless is not always easy. Under the heading of
situation the map interpreter can attempt to define and describe
subject always to the limitations of the map the geographical setting.
The economic and social relations of a town, which may be regarded
as external functions, are (like internal functions) a matter for statistical
treatment. In map interpretation they must be taken largely for
granted. The task of the interpreter is to state how and why the town
is physically accessible, how the converging routes wliich express
nodality are related to the form of the ground, and how the town is
placed in relation to the boundaries of stows and tracts. For obvious
reasons, small market towns are often centrally placed in the areas they
serve, but larger collecting and distributing centres are more typically
peripheral. This is obviously true of a great port or a major collecting-
centre like Winnipeg, but a number of lesser towns are also significantly
placed near the boundary of contrasted tracts, where dissimilar products
can be marketed.

Many nodal points lie at the focus of a sheaf of valleys. The crude
and obvious interpretation is that the town site is a node of valley- ways,
but it is often at least as probable that the early routes followed the
crests of divides, avoiding damp and wooded valley-bottoms. Ridge-
routes would, of course, converge on die same point as the later
valley-roads. 1 Similarly with many "gap towns": admittedly such a
town commands the gap, on which modern roads converge, but
formerly movement overland may have been principally across the
gap, along the line of the high ground on either side.

Site

When the site of a town is examined, two things should be looked
for: the nucleus around which growth has taken place, and the
manner in which the growing town has adapted its structure to the
form of the ground. On a small-scale map the nucleus alone may be

1 The author is indebted to Dr. H. C. Brookfield for suggesting Hertford as a leading
example.

138 MAP INTERPRETATION

visible. Consider the frequent case of a small country town lying at
one end of a river bridge : the bridgehead, commonly with the addi-
tion of defensive works, constitutes the nucleus; but, when the term
"bridgehead town" has been applied, interpretation is not much
further forward. Careful examination will often reveal, or at least
suggest, why, if the river had to be crossed in the locality, the bridge
should stand where it does instead of half-a-mile or so upstream or
downstream; that is, in what way the point selected is superior to
others available near by. Much can usually be done towards inter-
preting the advantages of the particular site, but none of these can in
themselves explain the growth of a town. The question that should be
asked is: assuming that a town is to grow somewhere in this neigh-
bourhood, which is the likeliest site, and why?

Similar treatment should be applied to a port. The physical setting,
however advantageous, can in no sense account for the existence of a
port, which is due to powerful economic and historical factors. On
the other hand most (but not all) large ports are clearly located in
close relation to geographical features ; but, since few ports have been
deliberately founded as towns, it is often difficult to find the nucleus
around which they have grown, modern harbour works having
destroyed or obscured the original advantage of position. If a port is
of any size, its site to-day will probably combine favourable with
unfavourable factors, for example, tidewater far inland with little
room for expansion, which the interpreter should try, as far as possible,
to perceive.

On the 1/63,360 and comparable or smaller scales it is not always
possible to make out in detail any correspondence between the form
of a town and that of the ground, especially where parts stand on
river terraces which do not show well on the topographical map. The
attempt should always be made, however ; it is rare in unplanned towns
to find no evidence whatever of what may be called "preferential
growth*' in certain directions. Where a town is functionally differ-
entiated, a rough correspondence is often found between the functional
subdivisions and the type of ground on which they are located. Thus
the study of site returns to that of functions and structure, with which
this discussion opened.

The maps already used show forty-seven towns, with part of a forty-
eighth, Bristol. There is more than enough material to illustrate

TOWNS: SITE, FORM, AND SITUATION 139

specifically the principles set out in the earlier part of this chapter, so
that it will not be necessary to study all the towns in detail. Since,
however, no dominantly industrial town is wholly represented,
Sheet SO/ooof the O.S. 1/25,000 Series has been selected in addition,
for an interpretation of Merthyr Tydfil.

The Small Market Town

Small market towns raise few problems of interpretation, except that,
paradoxically, maps on the smaller scales cannot show whether or
not a market survives; but, even if cattle and produce markets have
ceased to operate, the market square is part of the nucleus around which
growth has taken place, and the towns continue to discharge the
functions of marketing in the wider sense of retail distribution to the
surrounding countryside. By definition, towns of this kind are not
greatly industrialized, and have not been greatly affected by the urban
spread of modern times. Nevertheless, the interpreter is well advised
to look for factories, which not infrequently occur. Some are related
to an essentially local industry, for example, flax mills or processing
plants for dairy produce; others are outposts of the major industry
carried on in a neighbouring large centre, for instance, the leather works
in several market towns of the Midlands ; others again are the result of
decentralization of industry, a complex economic process greatly
stimulated by the recent war.

In or near the easternmost part of the Cotswolds (O.S. 1/63,360
(Seventh Series), Sheet 144) there are six small country towns:
Chipping Campden, Shipston on Stour, Moreton in Marsh, Stow on
the Wold, Northleach, and Burford. Although the plan differs from
one to another, they have in common their compactness and lack of
integument. Their sites bear a family resemblance to the sites of
neighbouring villages, but their situations are more highly nodal.

Chipping Campden (1539), whose market function is recorded in
its name, is centrally placed amid a group of villages which, like the
town, occupy original wet-point sites. The large number of villages
within a short radius of Chipping Campden is associated with the
convergence of the two scarps, facing respectively east and north-west,
and with the outlying scarp of Ilmington Hill, for most of the villages
in question are located at the scarp-foot.

Moreton in Marsh and Stow on the Wold lie on the Fosse Way at
nodes of cross-routes. If the official classification of roads is disregarded,
the convergence of routes is found to be far more pronounced than

I4O MAP INTERPRETATION

the map at first suggests. The situation of Stow is especially instructive,
for the town is sited on a hilltop approached from all sides by ridge-
ways (Plate IVA). The road from Evesham and Broadway runs along
the plateau top east of the headwater valleys of the Dikler; that from
Tewkesbury via Stanway on the western side crosses only two valleys.
There seems to have been an alternative to the second route through
Winchcombe, across the Windrush near Guiting Power (0924) and
thence eastward to Stow, but as a through-road this has been super-
seded by the main road from Cheltenham, A.436, which is in part
only a ridgeway and which makes more difficult crossings of the
Windrush and Slaughter Brook valleys. East of the Fosse Way two
ridgeways on either side of the Hazelford Brook valley run northwards
to unite near Wyck Beacon (2020), whence the route continues across
the col between the valleys of the Evenlode and Dikler. The main
road from Oxford comes in from the east along the flank of Chastleton
Hill, traverses a narrow part of the valley bottom, and mounts to
Stow along the side of St. Martin's Hill. Although the roads from
Blcdington (2422) and Evenlode (2220) are not ridgeways they do
not at least keep to the lowest ground, and are in any event less direct
than those previously mentioned. Moreton in Marsh is very differently
sited from Stow, lying on low ground where the Oxford-Evesham
road crosses the Fosse Way. The immediate approaches cannot
naturally take the form of ridgeways, in the ordinary sense, but are
nevertheless distinctly related to the low divides of clay country.

Northleach (1114) offers an interesting contrast with Moreton and
Stow, both of which stand on the Fosse Way whereas Northleach
stands to one side. The site is in a valley bottom with a natural water-
supply, between dry tabular interfluves. The town has grown east-
wards and westwards along the valley, apparently from a nucleus near
the church; the central cross-roads lie at the intersection of the
Burford-Cheltenham-Tewkesbury road, A.4O, with what is now a
minor road from Bourton on the Water (1620) through Farmington
(1315), branching at Northleach to Chedworth (0512) and Coin St.
Denis (08 n). It seems likely that at one time the less direct road was
preferred to the Fosse, possibly because of easier gradients, for example,
in the stretch between Northleach and Bourton.

Besides standing to the side of the Fosse, Northleach is also off the
ridgeways. The significantly named Salt Way, running southwards
from Evesham, passes about a mile away on the south-west; the road
from Burford, instead of descending into the valley at Northleach,

TOWNS: SITE, FORM, AND SITUATION 141

could continue to the north, rejoining the modern Cheltenham road
at Puesdown Inn (0717). Thus Northleach is less directly nodal than
might at first appear; its nodality arises, in part, from the fact that
several ways pass close by, so that Northleach provides an approxima-
tion to the several cross-roads.

Shipston on Stour (2540) further illustrates the effect of movement
along other roads than the Fosse Way. No nucleated settlement occurs
on the Fosse itself, either at the crossing of the Chipping Campden-
Banbury road, or at that of the road from Stratford on Avon up the
Stour valley towards Oxford. Instead, a small market centre has arisen
where these last two intersect, at one end of a bridge over the Stour.
Note at Shipston, as at Chipping Campden, isolated buildings which
appear to stand in the main street. These are probably covered market
halls or exchanges. Burford (2512) resembles Shipston in controlling
a river crossing, but also possesses an ancient priory, which may
well have encouraged early growth and have assisted Burford to
outpace the alternative crossing-settlement at Barrington, 3 miles
upstream. The road system south of the Windrush well repays close
study.

Enough has now been said to illustrate the fact that, within the class
of small towns, great variety of setting is possible within the limits of a
single tract of country. That is why the interpreter should deal with
each such town on its merits, refraining from hasty classification and
using to the full the information obtainable from the map.

The six towns discussed are very small and show few or no signs of
recent growth. Melksham and Devizes, in the north-west of the area
of the O.S. 1/63,360 (Seventh Series) Sheet 167, although by no
means big towns, are larger than the previous examples and appear to
be expanding. The characteristic openwork pattern of recent suburban
housing appears on the eastern side of Devizes, in the kilometre Grid
Square 1061, while Melksham is spreading to the south-east and
north-east. Note also at Melksham the rubber factory. It is difficult
to imagine that the town offers specific advantages for rubber manu-
facture it is far more probable that the general factors of rail transport
and a potential labour-supply were relied on when the industry was
established.

Some country towns, without being industrialized, have yet grown
fairly vigorously in modern times. They often serve as the economic
centres of whole tracts or of large stows and their wider influence is
not infrequently recognized in place-names (cf. Salisbury and Salisbury

142 MAP INTERPRETATION

Plain, Evesham and the Vale of Evesham). They are more markedly
nodal than the smallest towns, and in addition usually occupy dis-
tinctive sites, for their eminence dates from medieval times when they
were strongholds of defence, noted ecclesiastical centres, or both at
once; but a town which stood very high in the medieval urban
hierarchy, for whatever reason, may have experienced very little re-
growth, for example, the cathedral town of Wells and the neighbouring
Glastonbury with its renowned Abbey (O.S. 1/63,360 (Seventh Series)
Sheet 165). We are concerned at present not with these, but with
towns of comparable antiquity and former size which have developed
vigorous modern functions of industry and trading.

Salisbury is a case in point (O.S. 1/63,360 (Seventh Series) Sheet 167).
The nuclear area includes the cathedral; the core is approximately
defined by the very densely built-up southern part of the present
town, with an outlying portion in the angle of confluence between the
Avon and Nadder. Outside the core there has been considerable recent
growth towards the north-west, in the Pembroke Park area, and
towards the north in the direction of Paul's Dene. On the eastern side
also a more open pattern of streets and buildings can be seen, between
the core and the railway. As no marked industrial development
appears to have taken place it seems that the modern growth of
Salisbury is related to its situation, for the town is well placed to serve
a large number of rural settlements in a tract where no other town
exists. Both site and situation illustrate the contrast between the con-
ditions of Roman and post-Roman times respectively. The Roman
roads, now in large part disused or reduced to the condition of bridle
ways, radiate not from Salisbury but from the earlier centre of
Sorbiodunum (Old Sarum). The modern Salisbury lies at the focus
of two sets of roads, those running along the crests of interfluves and
those following the valleys. Since nucleated settlements are con-
centrated in the valleys it is the valley-ways that best express the
nodality of Salisbury as a market centre, even though the easiest
approaches in the early days must have been along the ridges. The
frequent elements -ton, -ford, and -bury in the place-names of the valley
settlements are associated with Saxon penetration. Thus, although the
observed pattern of settlement and communications belongs to post-
Roman times, it has evolved from an original pattern of considerable
antiquity a fact which serves to emphasize the utility, in a discussion
of Salisbury, of separating the core of the town from the integument.
The recent growth of the latter corresponds not to an increase in rural

TOWNS: SITE, FORM, AND SITUATION 143

population but to closer and more extensive commercial links between
the town and its environment.

Ports

Except for Bristol, which appears only in part on one of the selected
maps and will not be discussed, the ports available for study are small.

Bridgwater (O.S. 1/63,360 (Seventh Series) Sheet 167) stands at the
lowest crossing of the tidal River Parrett. Roads converge on the
bridge from the west, running across the broken hill country, and
also from the east where they are more directly guided by the belts of
dry land amid the fen. Since there is a relatively short traverse of
fenland between the extremity of the Polden ridgeway and the river
crossing at Bridgwater, it is readily understood that the town has con-
siderable nodality in respect of movement by land, quite apart from any
traffic on the Parrett.

The site of Bridgwater may usefully be compared with that of
Combwich (2642). Both settlements are based on firm ground which
approaches the river on the western side, and both originally possessed
small natural inlets to serve as harbours. Combwich is several miles
nearer to the open sea, and presumably has the longer period of high
water, but can scarcely have provided a crossing in the very early days
when fording not bridging was the rule. Furthermore, the natural
drainage of Pawlett Hams, across the river from Combwich, appears
likely to have been considerably worse than that of the ground
opposite Bridgwater.

The small core of Bridgwater is almost surrounded by integument,
wherein ribbon building may be identified along all the radiating
roads. Modern expansion has been industrial, for in addition to the
railway-carriage works one observes a number of factories alongside
the Parrett, from 305384 in the north to 320353 in the south-east. It
seems that the navigable river has been more influential in locating
factories than either the railway or the canal. Since no factories are
named or otherwise described, apart from the railway-carriage works
and the brick and tile works at Chilton Trinity (3039) it is impossible
to suggest what the dominant industry might be. The many patches
of inland water on the eastern side of the town might be flooded
clay-pits, as they are at Chilton Trinity, but might equally well be
gravel workings.

Few of the facilities of a small port can be adequately represented
on a map of this scale : the 1/25,000 Series enjoys a better scope. Since,

144 MA] P INTERPRETATION

however, factories are located on the Parrett it may be inferred that
wharves have been constructed along the river banks, providing
accommodation additional to that of the small dock. The canal
leading out of the dock would not appear to have been a great success,
since factories avoid it. The more useful railway link is secured by
spurs to the dockside and to the quays on the east of the river in
Castle Field.

In many respects Boston (O.S. 1/63,360 (Seventh Series) Sheet 114)
resembles Bridgwater. It also is an estuary-head port, with obvious
local nodality. Narrowly confined between the old fen on one side
and the former estuarine marsh on the other, the Boston crossing of the
Witham carries roads which converge along the broad lanes of firm
ground, as well as that running southwards along the "islands" of
Stickney and Sibsey. Although the shoreline has been pushed seaward
for a considerable distance by reclamation within historical times, the
Witham is still tidal up to Boston bridge. Since a canal takes off at
Dogdyke (2155) the regularized Witham above Boston must be
navigable for canal-boats, but unlike the Parrett is not flanked by
factories. The modern growth expressed in the integument of Boston
is to be associated with industrial development on the southern side
of the town, near the small rail-served dock and also on the west of
the river. Here no factories are named or described, but the two
electricity transmission lines suggest that a power station may be
located at 335431.

Lil^e Bridgwater, Boston serves a rural hinterland: hence their
relatively small size; but, just as Bridgwater surpasses Combwich,
so Boston is larger and better connected than Wainfleet All Saints
(5059). JThe general setting of Wainfleet is broadly comparable to
that of Boston, but the former now lies more than three miles from
tidewater on a smaller river than the Witham. Whatever historical
factors may have operated, it is clear enough that as a port Wainfleet
suffers grave physical disadvantages. Indeed, it has been effectively
reduced to the status of a small inland market centre, located at a river
crossing.

Wexford (O.S. Ireland, 1/63,360, Third Edition, Sheet 169) is
another small port in a rural tract. Since the map was last revised in
1898, it can hardly be expected to record a large integument. Very
little can be read from the map of port facilities, which appear to be
restricted. They may possibly be represented by the jetty immediately
downstream of the road bridge, together with the large buildings

TOWNS: SITE, FORM, AND SITUATION 145

on the seaward side of the single-track railway. Undoubtedly the
port of Wexford is unsatisfactory in some ways probably because of
too shallow water at low tide for an outport has been established at
Rosslare Harbour on the open coast. As Wexford seems very little
industrialized, and as the cargo traffic of small ports is not usually too
urgent to await a suitable height of tide, one may infer that the rail-
served pier at Rosslare Harbour is designed for passenger traffic.

Other specialized ports on the selected maps include Kinlochleven
and Fort William (O.S. Tourist Map, 1/63,360 (Lorn and Lochaber)),
and Par and Charlestown (O.S. 1/63,360 (Seventh Series) Sheet 186).
Each of the first pair has a deep-water pier, connected by rail with an
aluminium factory, whence it would appear that bauxite is probably
brought in by sea. The second two are Cornish ports which serve as
outlets for the china clay field of Hensbarrow, north of St. Austell.
They are too small to rank as towns.

Also in the area covered by Sheet 186, Lostwithiel (1050) and
Fowey (1251) well exemplify a relationship of settlement to ria which
is so common as to be properly regarded as typical. Lostwithiel is
sited where roads converge on the crossing of the deep valley (cf. the
place-name "Bridgend" at the eastern side). The main road from
Liskeard is a ridgeway, which secures gentle gradients by skirting the
heads of southward-flowing streams, and passes over the Fowey river
at the (present) tidal limit, i.e. at the lowest point which could be
crossed with relatively little difficulty. An east-west road nearer to
the coast would involve many bridges and steep hills, and would
approach the tidal Fowey where it is much wider. The town of Fowey,
although provided with two ferry services, is not primarily a crossing-
place, but a coastal settlement with a sheltered, deep-water harbour,
directly connected on one side of the ria only to the town at the ria-
head. An interesting modification of this association of towns with
rias occurs on the Looe, which is again represented on Sheet 186.
The river mouth is so narrow that bridging was not unduly difficult:
in consequence East and West Looe combine the functions of bridge-
town and port, and there is no settlement of any size at the ria-head.
The principal road crossing of the Looe valley, however, is still located
well inland, near Liskeard.

Resorts

Most resorts of this country lie at the seaside, for which reason they
may conveniently be discussed next. Whereas the nucleus of a port

I4<> MAP INTERPRETATION

may be regarded as an inlet, or as a conveniently sheltered portion of
shoreline, the nucleus of a seaside resort is the central part of the front.
The actual beach did not, as a rule, constitute a factor in the growth
of a resort before an efficient system of sewage disposal had been
acquired; but whatever its individual peculiarities of location and
history, a seaside resort is usually distinguished by being sited on the
open shore, and by possessing far more integument than core since it
is largely a product of the last century, or even of the last fifty years.

Weston-super-Mare, Clevedon, and Burnham on Sea (O.S.
1/63,360 (Seventh Series) Sheet 165) may be contrasted in respect
of site with the port of Bridgwater discussed above. Burnham,
indeed, might be more aptly considered in relation to the small
neighbouring town of Highbridge, a crossing-place and minor port
on the River Brue. All three resorts are characterized by a relatively
open pattern of streets and buildings, with very little sign of industri-
alization. All have golf links; Clevedon has one pier, Weston two.
None is served by an inland waterway, nor does the railway approach
the waterfront.

The considerable spread of building along the southern flank of
Worlebury Hill, on the northern side of Weston-super-Mare, illus-
trates a typical development of resorts where suitable ground is
available. The hillside is a "desirable residential area." The same
phenomenon is illustrated in greater detail, although on a smaller
scale, by Sidmouth (Sheet SY/i8 of the O.S. 1/25,000 Series), where
detached houses are scattered over the eastern side of the valley.

Clevedon and Weston each have a small core where the buildings
are compactly massed; Burnham, lacking a comparable centre,
appears ,to be of more recent growth, although the fact is not very
clearly illustrated on the 1/63,360 map (cf. however the representation
of Skegness on the same scale (Sheet 114): there can be no doubt that
this resort has grown recently as well as vigorously, for no core of
dense, older building is to be found).

When non-industrial inland centres display small cores and large
integuments resembling those of seaside resorts, the leading prob-
abilities are that the towns in question are also resorts, for example,
spas, or that they function as dormitories. The fact offering itself for
interpretation is once again the great extent of residential building.
Although spas, as such, are no longer fashionable, they may have
survived as residential or resort towns of a specialized character.
Cheltenham (O.S. 1/63,360 (Seventh Series) Sheet 144) is one such.

TOWNS: SITE, FORM, AND SITUATION 147

Although its history can only be guessed at, the map indubitably
represents a town with a large, unindustrial integument very similar
in cartographic appearance to the outer parts of flourishing seaside
resorts.

Large Towns

As was made clear in the early part of this chapter, the large modern
town performs many functions. The towns discussed so far, simply
because all are highly specialized and none is of more than moderate
size, have been conveniently treated under summary descriptive
headings, but it has already been observed that a useful distinction can
be made between the inner (commercial) and the outer (mostly
residential) portions. The core of Clevedon, for example, is identical
with the shopping centre. The resorts of very recent growth, and the
very small market towns, provide apparent exceptions, but maps on a
larger scale (for example, 1/10,560) would show that in them also the
central part is more densely packed than the outer, while a survey of
urban land use would reveal the functional differentiation. 1 The next
example provides more complex material. While it does not illustrate
the large, fully diversified, "general-purpose" modern town, it
combines extensive industrial and commercial areas with residential
districts of two distinct kinds.

An Industrial Town

O.S. 1/25,000, SHEET SO/oo (MERTHYR TYDFIL)
On this scale the form of the town is very clearly shown. The core
consists of the compact mass of buildings and narrow streets lying
within the bend of the River TafF, in the north-east of die kilometre
square 0405. The nucleus seems to have been located near the church
at 050048. Early growth was probably responsible for the tongue of
close building which fringes the main road to the north-east, but most
of the expansion has undoubtedly been associated with the working of
coal and iron ore, and with the manufacture of iron, which are so
abundantly attested by the pits, levels, works, and spoil-heaps (Plate
IVfi). The development has been remarkably asymmetrical, and
considerably guided in direction by the form of the ground: Dowlais
(0607) is based on a broad spur between two small but deeply incised
valleys, while Georgetown (0406) occupies part of the flat valley floor

1 Functional maps of towns may be consulted at the offices of many planning
authorities.

14$ MAP INTERPRETATION

of the Taff. Despite the irregular outline of the whole, certain general
conclusions are easily reached. Some quarters, for instance, that on
the valley side south-east of Pen-y-bryn Water Works (0507), include
an older and particularly compact element, but consist chiefly of the
regular streets of terraced houses which are a typical product of rapid
urban growth in the nineteenth and early twentieth centuries. With
slight modifications, this pattern is repeated in much of Dowlais and
in the Taff Valley both to north and to south of the core. That part
of Merthyr immediately east of the main railway station, spreading
north-eastwards up the hillside to Thomas Town, appears certainly to
have been built later than the railway. A concentration of public
buildings (shown by solid black) immediately north of the station
suggests that some important cultural and administrative functions
are discharged here, but the more easterly portions are less easy to
interpret. On higher slopes where the land is free of industrial waste,
or in valleys beyond the older residential quarters, one may note the
wider, curving streets of suburbs constructed in the twentieth-century
manner, as, for example, in the kilometre square 0508. The plan of
these streets still has a certain geometrical uniformity, in contrast with
the irregular scatter of large, detached houses on the hill flank above
Cefh-coed-y-cymmer (0308).

The industrial functions of Merthyr Tydfil can be interpreted with
unusual clarity. It is evident from the numerous old coal levels and old
ironstone levels that, at least in the first days of industrial development,
gently dipping seams of coal and beds of iron ore were worked in
adits not by shafts. The Taff, and the Cynon in the next valley to the
west have been incised into productive Coal Measures not violently
disturbed by earth-movements and containing economic bands of
ironstone. The presence of ironworks, as opposed to steelworks,
indicates either that manufacture was established early, or that the
local iron ore proved unsuitable for the acid Bessemer process which
came to dominate steel manufacture from about 1860 onwards. The
very numerous abandoned coal workings would also suggest early
exploitation and relatively small individual concerns. The fewer but
large pits now in work may be located by means of the notation on
the map, by a characteristic pattern of railway sidings and pithead
buildings (usually small), and by their association with active spoil-
banks; for derelict tips are symbolized by hachures, while those still
in use are shown by stipple and usually carry at least one line of rail.
The features in question can be seen at Cwm Bargoed Pits (086060).

TOWNS: SITE, FORM, AND SITUATION

149

Ironworks require much larger buildings and more elaborate systems
of railway lines (cf. those on the southern side of Dowlais at 065074).
The huge quarries in the north, on the flank of Morlais Hill (0509)
and elsewhere, appear to provide limestone, whose obvious destination

FIG. 9. MAN-MADE FEATURES IN AN INDUSTRIAL DISTRICT
Roads, railways, canals, buildings, quarries, and tip-heaps shown in solid black
Based, by permission, on Sheet SO/oo of the O.S. 1/25,000 Map)
Crown Copyright reserved

is the smelting works, for the highest ground hereabouts is marked,
on the map, with the symbols for rocky edges and rock-strewn ground
which have been encountered on the maps of Carboniferous Lime-
stone country. One might, perhaps, advance the very tentative
suggestion justified, in point of fact that the rock quarried is indeed
die Carboniferous Limestone, which would, therefore, seem to dip
southwards under the Coal Measures.

xi~(E. 5 i96)

150 MAP INTERPRETATION

The exploitation of minerals, the tipping of spoil, the construction
of works, and the expansion of towns have vastly transformed Merthyr
Tydfil and its environment. Simple inspection of the map gives a
powerful impression of the extent of this transformation, but its
extent can be fully appreciated only by means of selective mapping,
as shown in Fig. 9.

Geometrical Layout of a New Town

U.S. GEOLOGICAL SURVEY, 1/62,500 (FINAL COUNTY
(ARIZONA) CASA GRANDE QUADRANGLE)

Many of the younger towns in the central and western provinces of
North America have come into being very differently from the old
towns of Western Europe. Instead of a long period of slow urban
growth in which the street plan evolved, as it were, by natural selec-
tion, and a later recrudescence in the railway age, these new towns of
the New World have often arisen after the railway had been laid.
The order of succession has been reversed : the first arrival, the railway,
has been followed by streets, and the buildings have come last. The
streets in Casa Grande were not, at the time of the survey, fully built
up; they form a rectangular pattern, aligned in part on the railway
and in part on the meridian. Such a town, laid out as a whole or
expanding according to plan, differs from European towns as
greatly in form as in history. Core and integument are no longer
differentiated: they lie within the province of sociological not of
cartographic study.

NOTES AND REFERENCES

Much useful material on the early establishment and growth of towns
occurs in

H. C. DARBY (Editor). Historical Geography of England before A.D.
1800. University Press, Cambridge, 1936. (See especially Chapter V,
"The Economic Geography of England," A.D. 100-1250, by H. C.
DARBY, p. 214 ff.)

The classification and functions of towns are discussed by

M. AUROUSSEAU. "The Distribution of Population; a Constructive
Problem." Georg. Review, xi, 1921, p. 567.

R. E. DICKINSON. City Region and Regionalism. Kegan Paul,
London, 1947.

TOWNS: SITE, FORM, AND SITUATION 151

R. E. DICKINSON. The West European City: A Geographical Inter-
pretation. Routledge and Kegan Paul, London, 1951.

A. E. SMAILES. "The Urban Hierarchy of England and Wales."
Geography, xxix, 1944, p. 41.

The economic relations of towns with their surroundings are
attracting much notice at die present time. Among the papers which
have so far appeared, the following may be cited

R. E. DICKINSON. "The Distribution and Functions of the Smaller
Urban Settlements of East Anglia." Geography, xvii, 1932, p. 19.

A. E. SMAILES. The Urban Mesh of England and Wales. Institute of
British Geographers, Publication No. n. George Philip, London,
1946, p. 85.

A. E. SMAILES. "The Analysis and Delimitation of Urban Fields."
Geography, xxxii, 1947, p. 151.

Many individual studies have been made from time to time. The
following is a very brief selection

H. C. BROOKFIELD. "Worthing: A Study of a Modern Coastal
Town." Town Planning Review, xxiii, No. 2, July, 1952, pp. 145-62.

E. JONES. "Tregaron, A Welsh Market Town." Geography, xxxv,
1950, p. 20.

S.J.JoNES. The Growth of Bristol Institute of British Geographers,
Publication No. n. George Philip, London, 1946, p. 55.

M. J. WISE. "Some Factors Influencing the Growth of Birming-
ham." Geography, xxxiii, 1948, p. 176.

CHAPTER XIV
PREHISTORIC OCCUPANCE

For monuments as for men, position is everything. BALZAC
The laws of probability apply only to large numbers. (Statistical
axiom)

MAP* O.S. 1/63,360 (SEVENTH SERIES) SHEET 165
(WESTON-SUPER-MARE)

ON the bounds of archaeology the map interpreter must tread with
care. Excavation and dating are the task of the field archaeologist,
who makes use of a great body of material which the topographical
map cannot record, and who is able to subdivide prehistory minutely.
Map interpretation of prehistoric features depends on a fraction of the
evidence that actually exists. It can take no account of artefacts but
must rely on earthworks, and only on those earthworks which have
survived, have been identified, and are represented on the map. These
are the facts employed in an attempt to discover which parts of the
land were occupied by man in each of the great cultural stages of
prehistory.

The Question of Dating

The special maps of the Ordnance Survey, such as the Map of
Neolithic Wessex, naturally raise no problems of dating, but the
standard topographical map shows remains of very different age on
the same sheet. The help afforded by distinctive notation is very
limited. Roman sites and remains, where they are represented, are
named in a special type

UPRIGHT SANS SERIF CAPITALS on the New Popular One- inch
map, although policy is now to employ EGYPTIAN TYPE for
this purpose. Pre-Roman antiquities are shown in Olb (tt0li$l)
type, those of post-Roman date in erman text.

On the New Popular Edition the difference between Old English
type and German text is not always clear, but the 1/25,000 Series is
wholly successful in this respect.

Fortunately for the purposes of interpretation, the nature of the

152

PREHISTORIC OCCUPANCE 153

remains is usually clear from the name or from the symbol, or both.
Archaeology gives exactly the lead required in showing that each
great class of earthworks is likely to belong to a certain one of the
broad divisions of prehistoric and early historic times. The classes of
earthwork, the cultural stages, and approximate dates applicable in
Lowland Britain are summarized in the accompanying table. It must
be stated as emphatically as possible, that the table has very little value
in the classification of a single feature. There is, for example, no
justification for referring a particular round barrow to the Bronze Age,
but where a number of round barrows occur in a given tract of country
it is highly probable that they are of Bronze Age date. The more
numerous and closely grouped the barrows, the greater the probability.
In dealing with prehistoric evidence, the interpreter should concern
himself first and chiefly with groups of features not with isolated
examples, that is to say, with the signs of undoubted and effective
occupance.

The classes of earthwork listed differ widely in purpose. A few
brief remarks are called for to relate each to the life of its period and to
point its precise significance. The following paragraphs are by no
means intended as an adequate review of prehistoric time in Lowland
Britain, for which the reader should consult the useful elementary
texts listed at the end of the chapter, but are meant to show how the
broad relation of man to ground altered from period to period. A
study of prehistoric distributions is in fact, to a large extent, a study
of changing geographical values.

Neolithic Antiquities

In Neolithic times, the earliest from which earthworks are known
in this country, agriculture was already being practised by people
living in hill-villages. The organization was probably one of semi-
nomadism, with tillage subordinate to pasture. It seems certain that
the climate was more oceanic than it is to-day, and the water-table in
uplands based on permeable rock the Chalk in particular is thought
to have stood higher than it now does. Although the downlands
carried timber they were much more easily penetrated than the
lowlands, which, except for sandy outcrops and river terraces were, as
a whole, damp, ill-drained, and densely forested. The soils of the
open hilltops were deep enough and fertile enough for primitive
cultivation, and movement was easiest along the ridges. There was
little pressure of population on the best areas, and in any event

154 M AP INTERPRETATION

Neolithic men were ill equipped to clear or to cultivate the low-lying
claylands, which, indeed, remained forested for many centuries. The
commonest evidence of Neolithic occupance consists in long barrows,
the characteristic mound of ceremonial burial. Long barrows are not
numerous, partly because only a limited number were built by a small
population, but partly also because some have been exploited for
building-stone and in that manner destroyed. The hilltop sites,
encircled by ditch and bank in Neolithic times, were generally re-
occupied and their defences elaborated in the Iron Age. As yet no
Neolithic corn-plots are definitely known, possibly because the same
ground continued in use subsequently. It was during the Neolithic
Age that Britain received the religion associated with megalithic
(big stone) monuments, some of which, again, are known to have
been destroyed in the last few centuries. Long-chambered tombs, or
dolmens, were probably originally covered with earth. Erection of
megaliths continued during the Bronze Age, to which belong many
standing stones in single lines or avenues, solitary (menhirs:
these are, however, difficult to date) or in circles (cromlechs). A
highly organized society is implied. The rarity or absence of fortified
sites of this period is taken by archaeologists to signify a long period of
peace. The greatest megalithic monuments of all, such as Avebury
and Stonehenge, are similar to the great cathedrals of to-day, in that
they were altered and added to during a lengthy use and cannot be
ascribed to a single period.

Bronze Age Antiquities

The Bronze Age in Britain was introduced by numerous immigrants,
whose period of dominance roughly coincided with the sub-Boreal
climatic phase, when the climate was drier than it now is, and the
summers were warmer. A number of authorities hold that a lower
water-table encouraged pasture rather than tillage on the downlands,
and suggest that man tended to live near the springlines, that is, in the
valleys. It is important to bear this point in mind in considering the
round barrows of certain uplands. Excavation has shown that where
round barrows are numerous they usually belong, as a group, to the
Bronze Age, although secondary burials of later date but in the same
barrows are not uncommon. The geographer taking a synoptic
view may expect to find round barrows concentrated on the higher
ground, often in precisely those parts where the long Neolithic
barrows occur, and may justifiably infer that in the Bronze Age, as in

PREHISTORIC OCCUPANCE 155

earlier times, the permeable uplands were the most favourable tracts;
but he should think in terms of hilltop pasture, remembering that
although the barrows prove exploitation of the land, they are burial-
places and not dwelling-sites. It is on some impermeable uplands,
Dartmoor, for example, that Bronze Age houses and villages are
coming to be known on the plateau top, together with walled cattle
enclosures such as Grimspound, near Moreton Hampstead. The
settlement of Dartmoor is itself a measure of the climatic difference
between Bronze Age and present times.

The Early Iron Age

In about 750 B.C., at approximately the time that a more oceanic
climate was re-established in the sub-Atlantic Phase, the Celtic inva-
sions began. The earlier arrivals were of Bronze Age culture, but iron
implements became dominant from about 500 B.C. onwards. The
heavier rainfall and higher water-table allowed tillage to spread
widely over the downlands, but promoted peat growth on the im-
permeable rocks of many highland tracts. Doubtless much evidence
of Bronze Age occupance is buried under sub-Atlantic peat, while
many Neolithic plots must have been obliterated by Celtic ploughing.
For about seven centuries after the first invasions, the plough used
was the light Mediterranean form, which did not turn the sod and
was well adapted to the shallow upland soils. Cross-ploughing of
small, roughly rectangular patches gave rise to the characteristic
pattern of Celtic Fields. Adjacent plots are separated by low banks or
lynchets, 1 originally faced with stone, where earth accumulated at
the downhill side of individual plots. "Scratch agriculture'* of this
kind put a great deal of the Chalk uplands under tillage, and is also
known, for instance, from the outcrops of Carboniferous Limestone,
where natural drainage was also good and the soil light and shallow.
It was associated with a shift of habitation back to the higher ground,
as occasionally shown on the O.S. map by the notation "British
Village" (now being discontinued). This upland cultivation seems to
have continued into, and possibly throughout, the Roman occupation,
while according to Childe the "Celtic Fields" of Highland Britain
were still cultivated in the Middle Ages or later.

The Celtic influx ended the lengthy peace of the Bronze Age,
replacing it by militarism, raids, and war. The unsettled conditions
are reflected by the many fortified hilltop sites usually called "camps"

1 See also p. 157.

156 MAP INTERPRETATION

on O.S. maps which, as already stated, include some formerly
occupied in Neolithic times. Authorities differ on whether these
strong-points were permanently inhabited or were merely refuges for
men and livestock in time of danger, but, whatever the fact, their
existence points to a time of much disorder. Where they occur, the
map interpreter should note their siting, wherein marked tactical as
well as strategic advantages are often apparent. In Scotland these
strongholds include vitrified forts, usually so named on the map. The
original walls included substantial timbers, which when fired caused a
partial fusion of the stones.

By about 100 B.C. a widening political grouping had caused many
hill forts to be abandoned. Defended cities were becoming established
on the lower ground. The downhill shift, although still slight, was
emphasized from about 75 B.C. onwards, when Bclgic invaders
introduced the heavy plough, an implement fitted with wheel and
coulter and capable of working the claylands. It is precisely because
the two forms of plough are suited to contrasted soils that the Celtic
field patterns remain visible to-day, for the heavy Belgic model,
which was used to turn the land in long strips, would have obliterated
the outlines of the squarish plots.

Roman Sites

Clearance of the damp, low-lying areas was at first slow, and is
known to have been effected mostly in post-Roman times. In this
connection the sites of Roman towns are less significant than the sites
of villas, which were largely self-contained and self-sufficient agri-
cultural establishments. They lie almost exclusively outside those
belts which are, in the natural state, badly drained; but the Roman
period is recorded most distinctively in features of civil and military
occupation, superimposed on the terrain rather than adapted to it.
However strikingly the Roman roads avoid or circumvent major
obstacles, and however strategically placed the camps, forts, and towns,
the patterns remain geometrical, corresponding with the texture of
the country only in the broadest manner.

Earthworks of the Dark Ages

The withdrawal of the legions in about A.D. 450 re-opened the
country to active immigration. Anglo-Saxon and Scandinavian
penetration and settlement have been discussed in another chapter.
Here it remains to mention only linear earthworks, often named

PREHISTORIC OCCUPANCE 157

"dyke" or "ditch," of which many belong to the Dark Ages. Al-
though some may have been constructed in haste for immediate
tactical purposes, others had strategic or political value, delimiting a
frontier or the boundary of a domain. Geographically a number are
of great interest, in that they stop short at the edge of low clayland,
whence it is inferred that, when the works were constructed, the clay
was still wooded and not easily penetrable.

Towards the west of the country many earthworks, linear earth-
works in particular, bear names reminiscent of the Arthurian legends;
towards the east they are more commonly credited to the Danes or to
Grim, the Devil. In most cases the implied dating or origin is wildly
inaccurate and should on no account be accepted.

Lynchets

The one class of earthwork not yet reviewed in its entirety is the
lynchet. The term is applied to at least three distinct forms of different
date. The narrow, striplike, near-horizontal terraces of steep slopes
(for example, on the scarped edge of Salisbury Plain near Mere) are
usually taken as cultivation-terraces of unknown date. On O.S. maps
at 1/63,360 and 1/25,000 there is room only for single rows of hachures,
which have to stand for whole systems of lynchets. The low banks
which bound Celtic fields are not always indicated on the i-in. map,
if the name is inserted; elsewhere hachures are again used. In the
Yorkshire Dales a third type of lynchet occurs, this time on the lower
slopes and on parts of the valley bottoms. Like the first type, it
consists of a narrow flat strip terminating in a steep descent on the
downhill side, but although some groups run across the slope others
run almost directly down. It has been urged that these lynchets are of
Anglian date.

It is hoped that this brief review, summarized in the Table on
page 158, will be of use in the interpretation of prehistoric occupance.
Deficient though it may be, the map record is capable of revealing
something of the former relationship of man to ground, and of showing
that in earlier times, even more markedly than to-day, the qualities
of the setting were not without influence on man's activity.

Interpretation from an O.S. Map

The principles stated above will now be applied in a specific inter-
pretation. The map selected is O.S. 1/63,360 (Seventh Series) Sheet 165
(Weston-super-Mare), on which physical distributions have already

8 ^

Ml!

l_i C t*
tO T3
n g
i||

^^ s

ZQ
<o

Roman
Antiquities

D < a

" &

.||

c S

H ^

fx, :

a "5

2 Q

*u>

< p

S r)

13
IE

Sb ti

*"t> ctf

cq *>

<s>

|-s

^|B

'C ^

"<1 "

CULTURE
PERIOD

CLIMATIC
PHASE

PREHISTORIC OCCUPANCE

159

been studied in Chapter X. Certain morphological boundaries
relevant to the present aim have been added, in Fig. 10, to a map of

C limits of Cofboftifeious limestone ptoteoux |

FIG. 10. ANALYSIS OP PREHISTORIC EVIDENCE

Early antiquities as shown on p.S. Seventh Series 1/63,360, Sheet 165, with the
addition of two lake villages

Crown Copyright reserved

earthworks taken from the O.S. sheet. The close general relation
between the form of the ground and the distribution of remains is
immediately obvious: it is emphasized by a detailed inspection.
The eight long barrows are assumed to be probably Neolithic.

160 MAP INTERPRETATION

Five occur on the Mendip plateau, three on the dissected scarpland to
the north. All are prominently sited, either on summits or on false
crests, where they appear on the skyline as seen from lower ground.
In themselves these burial-places prove no more than Neolithic
penetration of the upland, but it is entirely reasonable to suppose, in
view of the distribution of later features, that in Neolithic times the
low ground was untouched forest and marsh.

The four aligned circles, possibly dating from Neolithic or Bronze
Age times, suggest the ritual practices of a relatively stable and well-
organized society. The abundant round barrows show that in (pre-
sumably) the Bronze Age, the Mendip plateau was broadly and easily
penetrable, i.e. not more than lightly wooded, if not indeed widely
grazed and tilled. Of a total of 145 round barrows, 137 lie on the
Mendips, nearly all on the flat summit-plane where, however, the
gentle north-easterly slopes are avoided. Two barrows lie near the
summit of Bleadon Hill, a detached portion of plateau, and three
more in the dissected northern scarpland where long barrows have
been previously noted, giving a total of 142 on high ground. Of the
remaining three, that at 2045 (marked "tumulus" on the map) is
exceptionally placed very close to the fen. Walborough (3157) and
Castle Batch (3663), each on low ground a little above the 5O-ft.
contour, may not be tumuli at all, although they are marked by the
hachure symbol, for the names suggest a post-Roman origin.

Whatever reservations are necessary on the grounds that not all
round barrows have survived, or have been located and mapped,
there is very good evidence of selective occupance of the uplands.
Although, as remarked above, barrows are not dwelling-sites, there is
little doubt that the damper valley bottoms as well as the fen were
avoided by die people who made the round barrows. In this connec-
tion it should be noted that there is little dry ground between the
steep edges of the limestone uplands and the bottom lands, so that
grazing at least must have been practised chiefly on the plateau-tops.

Where summits are so wide and flat it is pointless to look for an
alignment of barrows along a ridge-top, such as may be found in
other parts of the country. The Mendip plateau offers negligible
obstacles to circulation, providing a belt rather than a line of hilltop
between the forests and marshes on either side.

The probable Iron Age antiquities recorded on Sheet 165 and in
Fig. 10 include twenty-four specified "camps" or large circular
earthworks. As a group the "camps" are remarkably well sited,

PREHISTORIC OCCUPANCE I6l

occupying steep-sided single hills or prominent spurs, invariably
commanding a wide range of country. As far as may be judged from
the mapped distributions, this area was more deeply penetrated in the
Iron Age than in earlier times, for defensive works appear on the coastal
hills where no barrows are shown, and "Celtic Fields," in so far as they
belong to the Iron Age, prove cultivation in that period of hilltops
beyond the mapped extent of round barrows. As noted previously,
the "Celtic Fields" should be regarded more as relics of an agricultural
system than as the trace of a particular people, but when they and the
"camps" are considered together it seems justifiable to conclude that
during the Iron Age the upland stows were still the most attractive.

The O.S. map omits two Iron Age village sites, at Meare and
Glastonbury respectively, which have been marked in Fig. 10. The
omission is a serious one, for the sites are those of two very well-known
marsh villages, which exemplify the beginnings of lowland occupance
in Iron Age times. Their absence from the map, corresponding to an
extremely faint trace on the ground, is a fact that illustrates the need for
caution in treating negative evidence.

The three Roman villas marked signify civil occupation. They lie
severally near the crest of the Poldens (4824), on the southern side of
Banwell Hill (3958) and alongside the embanked and tidal Yeo (4065).
This last site is unusual for a villa, but its natural drainage may have
been better in Roman times than it is now. A lowland coast of this
type is capable of changing considerably in two thousand years. No
recognizable trace remains of any roads which may have linked the
villas with the national road system, which is here represented by
part of the Fosse Way in the south-east and by portions and traces of
Roman roads on the Mendips. The paved ways on the plateau served
a Settlement and a Camp, with an associated Amphitheatre near
Charterhouse on Mendip (4955) evidently a civil as well as a military
establishment existed here. This map fails to suggest that the Roman
settlement was connected with lead mining, although it is well known
that this was so. The modern mine buildings, disused but still extant,
have also been omitted, the only sign of mining being the three pools
impounded by spoil in the spinney at the head of Velvet Bottom.

Among the linear earthworks, some of which are likely to belong
to the Dark Ages, New Ditch (5033) is significantly named and
strikingly placed. It lies precisely where the crest of the Polden
cuesta narrows sharply westwards. It may be wondered if the Old
Ditch recorded in a settlement-name near Wells (5049) is an earlier

162 MAP INTERPRETATION

defensive work of the same people. "Intrenchments" above Clevedon
on Castle Hill are rather indistinctly shown, and should possibly be
associated with the systems of hill-forts on the two ridges enclosing
Walton Moor. In any event they mark a site of great tactical, if not
strategic, value. Finally, the "earthworks" inland of Weston-super-
Mare include some kind of mound as well as a ditch, and occupy a
small rise which commands the shortest way across the fen between
the Mendips and Worlebury Hill.

It has now been briefly demonstrated that an analytical treatment of
the distribution of early antiquities can reveal something of the con-
quest of the land in prehistoric times, and of the profound changes
which have occurred in man's relationship to the geographical setting.
Thus, despite the inherent defects of work with partial evidence, a
certain limited success may be claimed.

NOTES AND REFERENCES

Eminently readable, but at the same time authoritative, general accounts
of prehistoric times in Britain are

GRAHAME CLARK. Prehistoric England. Batsford, London, 1945.

V. GORDON CHILDE. Prehistoric Communities of the British Isles.
Chambers, Edinburgh, 1940.

SIR CYRIL Fox. The Personality of Britain. Fourth Edition. National
Museum of Wales, Cardiff, 1943.

JACQUETTA and CHRISTOPHER HAWKES. Prehistoric Britain. Chatto
and Windus, London, 1949.

STUART PIGGOTT. British Prehistory. University Press, Oxford, 1949.

S. E. WINBOLT. Britain B.C. Penguin Books, Harmondsworth
(Middx.), 1943.

Life in Roman Britain is surveyed by

R. G. COLLINGWOOD. Roman Britain. University Press, Oxford,
1942.

The summary Table given in the text is based chiefly on the works,
cited above, of SIR C. Fox and J. and C. HAWKES, together with the
following very useful booklet

ORDNANCE SURVEY. Field Archaeology, Some Notes for Beginners
Issued by the Ordnance Survey. O.S., Professional Papers, New Series,
No. 13, H.M.S.O., 1951.

The question of natural vegetation in early times is broadly reviewed
in Fox, op. cit., p. 53 ff.

PREHISTORIC OCCUPANCB 163

Post-Roman clearing of woodland is discussed by

H. C. DARBY. "The Clearing of the English Woodlands." Geo-
graphy, xxxvi, 1951, p. 71.

An accessible account of the Gaulish method of constructing forts
with stone and timber, relevant to the problem of vitrified forts, may
be obtained from

CAESAR. The Conquest of Gaul Translated by S. A. HANDFORD.
Penguin Books, Harmondsworth, (Middx.), 1951.

Short descriptions of the two marsh villages in Somerset may be
found in the books by CLARK and CHILDE referred to above; but, for
an instructive comparison between the results obtainable from map
interpretation and those yielded by field archaeology, see

D. P. DOBSON. Somerset (County Archaeologies Series). Methuen,
London, 1931. (This account bears out the interpretation given here,
but supplements it considerably in some respects, for example, by
recording finds of implements, and more Roman villas than are shown
by the O.S. 1/63,360 sheet (cf. Fig. 10, p. 159, and Dobson's map,
pp. 132-3). Dobson also analyses the Saxon finds.)

The study of archaeological distributions generally is greatly
assisted by reference to the maps in Fox, op. cit. (p. 162), and to the
special maps of the Ordnance Survey; for example, Map of Roman
Britain, Map of Neolithic Wessex, Map of Monastic Britain, etc.,
which can be obtained in annotated form.

PART III

SPECIAL TOPICS

13 (.5196)

CHAPTER XV
MORPHOMETRIC ANALYSIS

We require the permeating accuracy of scientific methods as well
as knowledge of basal scientific facts. FAIRGRIEVE

MORPHOMETRY is the measurement of shape. Morphometric analysis
of maps is intended to reveal and define, more clearly and precisely
than can be done by unaided inspection, the general form of the
ground as represented on the map. The techniques employed
vary in difficulty. Some involve laborious measurement and a
certain amount of calculation, but others are rapid, graphic, and
simple. All attain a degree of exactitude which is impossible in map
reading.

There can, of course, be no substitute for detailed field-work in the
study of landform. Very many features of the highest significance
appear only on those maps drawn by the field geomorphologist ;
but the techniques of morphometry are of great use in the rapid
exploratory treatment of large areas, in helping the map interpreter
to reduce the forms of diversified country to some kind of order, and
in the treatment of certain classes of data which are not well derived
from field study.

For convenience' sake, the methods reviewed in the following
paragraphs are grouped under the four somewhat arbitrary heads of
geometric, arithmetic, volumetric, and clinometric analysis, according
to whether the landscape element measured is the general form of the
ground, the relation of area to height, the volume of specific features,
or the degree of slope. The first group comprises most of the rapid
methods of analysis that the interpreter might expect to find useful
in a first approach to a given problem: some of these should undoubt-
edly be standard practice. The remainder, requiring more time and
lengthier measurement, are certain to be less generally applied. They
are included here, not so much in order to provide additional tools
for frequent use, as to demonstrate the considerable possibilities of
this kind of work, and to explain the construction of various
kinds of diagram which are coming to be widely used in descriptive
accounts.

167

168

MAP INTERPRETATION

I. Geometric Analysis

Work of this kind is designed to simplify the relief distributions of
the contour map, and to reveal the major landform patterns of the

FIG. ii. GENERALIZED CONTOURS, FOR THE AREA SHOWN ON SHEET 186

OF THE O.S. SEVENTH SERIES 1/63,360 MAP

Inset, method of constructing generalized contours

(Based on the O.S. map, by permission)

area treated. The simplest exercise, and the most widely useful, is the
construction of generalized contours, which touch the actual contours
at the tips of spurs but are carried across existing valleys (see Fig. n).
The surface defined by generalized contours is that which would be

MORPHOMETRIC ANALYSIS 169

observed if the valleys were filled in to the general level of the inter-
fluves. If the crests of divides are remnants of a platform, whether
erosional or structural, its form is approximately reconstructed in the
generalized contours. When platforms demonstrated in this manner
truncate structures they must be erosional.

It is often found that in a given highland tract the major forms of
the ground, however greatly diversified by present dissection, consist
in broad flats or very gentle slopes, separated by much steeper slopes
or topographic risers. The flats may be miles wide, die risers tens or
hundreds of feet high : the whole constitutes an impressive physio-
graphic stairway. Generalized contours are eminently suited to reveal
such major forms: they lie far apart on the flats and bunch together at
the risers, as shown in Fig. n, where they are drawn for the area
of the O.S. 1/63,360 (Seventh Series) Sheet 186. It has already been
seen in Chapter V that this landscape can be interpreted in terms
of denudation controlled by a sequence of higher base-levels: con-
clusions drawn from inspection of the map are amplified as well as
confirmed by generalized contours. The residual hills above the wide
plateau of Bodmin Moor are encircled by the generalized i,ioo-ft.
contours ; the plateau itself extends to the poo-ft. line on the eastern
side and to the 8oo-ft. on the west; the steep edge is well brought out
in close spacing, which also marks the possible old cliff line inland of
Tintagel. In this general view there is a very marked topographic
break between the bounding slope of the granite boss and the broad
flat of the lower ground. Both on the east and on the south this break
occurs between 600 and 700 ft. O.D. Note also the similar break of
general slope, at a similar height, on the north-eastern side of Hens-
barrow. If more detailed work for example, on the 1/25,000 map,
but primarily in the field proved that the change from slope to flat
was nearly horizontal, one would suspect the presence of an old shore-
line, related to a base-level higher than, but parallel to, the present one.
This would imply that the fall of base-level was not, in this area,
accompanied by tilting, in which case any old shorelines at lower
altitudes should also remain horizontal : now the run of the generalized
contour at 400 ft. O.D. in the basins of the Looe and Fowey is not
inconsistent with a former shoreline at about this level, such as has
already been suggested by a study of the north coast.

Serial profiles are drawn along equally spaced parallel lines, which
run in the direction of general slope, if any. When closely spaced
enough, such profiles may indicate or suggest the presence of erosional

170 MAP INTERPRETATION

or structural platforms. If geological detail be added, they illustrate
the relation between structure and relief.

As serial profiles are not easily compared with one another at a
glance, especially if the series is large, superimposed and projected
profiles have been devised for the purpose of rapid comparison. In a
superimposed profile the entire series of profiles is plotted in a single
diagram. If any platforms exist, of whatever nature, they are likely to
be reflected in the grouping of lines (Fig. 120), whereas topographic
risers are distributed over die whole figure; but the superimposition
of many lines may result in a confused diagram. Hence the practice

FIG. 12. SOME APPLICATIONS OF PROFILE-DRAWING

(a) Superimposed profiles.

(b) The same profiles projected.

Signs explanation appear at 400 and at 900 ft.

of simplifying, in the projected profile. Here the first of the series of
profiles, as it were the nearest to the observer, is drawn complete;
the second, or next nearest, only where it rises above the second ; and
so on throughout the series. Thus a kind of panoramic sketch (without
perspective) is obtained, which reveals accordance of level at specific
altitudes (Fig. 126). Care is needed in laying out the lines of profile,
for if a great deal of high ground occurs in the earlier profiles the
later may be heavily obscured. Some detail must in any event be lost,
for simplicity is achieved only by means of sacrificing part of the data.
Profiles drawn along arbitrary lines are likely to miss significant
breaks of slope. Hence the use of spur-top profiles, drawn along the
axes of interfluves. It is doubtful, however, if these have any advantage
over generalized contours, for the breaks of slope which they are de-
signed to reveal are precisely those by which the spacing of generalized
contours is governed. Needless to say, the matter is different with
spur-top profiles surveyed on the ground, whereby certain features of
Ate terrain may be fittingly sampled.

MORPHOMETRIC ANALYSIS lyi

Serial cross-profiles of valleys can bring to light the valley-in-
valley forms produced by rejuvenation, if these forms should be
present and sufficiently well marked to be recorded on the map. It
has been suggested that, where former cross-profiles can be recon-
structed, they may provide a basis for reconstructing former long-
profiles also, but in the writer's opinion their value in this connection
is not great.

One of the chief aims in constructing the long-profiles of rivers
is to discover any irregularities that may exist. If a river has been
rejuvenated, as for example by a rapid fall of base-level, it is often
found that the surviving portion of the earlier profile meets the
developing newer profile in a knickpoint of characteristic form. On
general grounds one should expect very many rivers to display compo-
site profiles, which like poly cyclic landscapes are the product of more
than one erosion-cycle ; but not every break of profile is a knickpoint
of cyclic origin: it is one thing to determine the profile form, another
to interpret it. If base-level falls, the head ward wave of rejuvenation
works its way up all the stream systems affected, so that a knickpoint
on one trunk stream should be associated with knickpoints on other
trunk streams and also on tributaries. A synoptic plot of the profiles
of a single-stream system, or of a group of trunk streams, should bring
to light any general correspondence of form. In practice, however,
complications are introduced, for example, by differences of rock
resistance, gaps in the record, the peculiarities of individual streams and
the diversity of actual sequences of rejuvenation. Thus, even when
long-profiles have been constructed by the only reliable and satisfactory
method, detailed levelling of the ground, the task of interpretation is
difficult. A map interpreter must rely on contours, not all of which
may have been instrumentally surveyed, and must reconcile himself
to the fact that long-profiles constructed from maps are certain to
omit many significant details of form. The interval between instru-
mental contours on the O.S. 1/63,360 series is too wide for any but the
coarsest work. Consequently any long-profiles drawn therefrom
are to be read with extreme caution, except perhaps where the river
gradient is generally steep and breaks of profile are very well marked.
Extrapolation of parts of a composite profile, so as to link them with
former base-levels, should not be lightly undertaken. Research
proves the futility of supposing that surviving profile curves can be
accurately extended in this manner. The interpreter should therefore
confine himself to suggesting, where the evidence is particularly clear,

172 MAP INTERPRETATION

that a certain profile element may be related to a former base-level
which has been identified by other means, for example, by generalized
contours.

For what they are worth, a number of long-profiles plotted from
Sheet 186 of the O.S. 1/63,360 (Seventh Series) are given in Fig. 13,
where marked breaks of gradient appear.

A further method which relies on the cartographic record of stream
gradients, and which is therefore to be employed with reservation, is
that of mapping the migrational tendencies of divides. It is assumed
that, other things being equal, a stream with a steeper gradient will
extend its catchment headward more rapidly than one with a gentler.

FIG. 13. LONG-PROFILES, DRAWN FROM THE MAP, FOR SOME OF THE
RIVERS SHOWN ON O.S. SEVENTH SERIES 1/63,360, SHEET 186
(Based, by permission, on the O.S. map)

(L) TheLynher.

(S) The Seaton River.

(F) The Fowey.

(/) One of the tributaries of the Fowey.

By comparing the gradients of opposing streams on either side of a
divide, where fortunately for the purpose in hand the gradients are
likely to be steep and streams much of a size, one may discover where
the advantage lies, and map the apparent direction of divide movement
as toward the streams with gentler gradients. Since relative gradient
is not the only factor which ought to be considered, the result obtained
is subject to some inaccuracy; but it has been found to work well in
practice, and to be capable of showing which streams tend to enlarge
their catchments at the expense of which neighbours.

2. Arithmetic Analysis

Under this head will be noted analytical techniques which employ
numerical values, whether of height alone or of height-plus-area.
Mathematically speaking, the simplest item is the summit-plane,
which is commonly indicated or suggested by a close similarity of
summit heights in a given area. If the accordant summits are numerous,
broad, and flat-topped, they may be looked on as the remnants of a
once extensive platform, which, if it truncates reconstructed folds,

MORPHOMETRIC ANALYSIS 173

faults, and rock formations, must have been the product of erosion;
but, where the summits are peaks, as in a great deal of glaciated
highland, the problem is more difficult. One usually finds that,
although the summits are roughly accordant in height, any recon-
structed summit-* 'plane" must be irregular in form and of considerable
relief. Here is the real difficulty: by no means all former cycles of
erosion approached completion, so that some former landscapes, if
accurately reconstructed, would appear hilly or even mountainous.
Unless the evidence is abundant and satisfactory, the interpreter can
do no more than point out a general similarity of summit heights.

Where a gieat many points have been levelled, their frequency
may be plotted against altitude. They are grouped within selected
ranges of height, for example, 0-99 ft., 100-199 ft., and plotted on a
graph, the altimetric-frequency curve. A number of such curves
have been constructed for parts of England by Hollingworth, who em-
ployed a wealth of unpublished O.S. data, arid for parts of France by
Baulig, who showed thereby the possibilities of morphometric work
with hachured maps. The method of treatment is purely statistical,
being designed to show at which altitudes there are relatively large
frequencies of levelled points. When large areas are treated, and when
more than one area gives high frequencies at corresponding heights,
it becomes possible that the high frequencies relate to erosional plat-
forms referable to former high base-levels.

The assumption made is that remnants of platforms, in dissected
country, lie on hilltops whose heights are determined in the course of
survey. A similar assumption is made in plotting the frequency of
closed contours. In the area selected for treatment, all contours closed
round outlying hills or spur-tops are identified, and the totals plotted
against height as in the altimetric-frequency curve. As before, the
diagram may be expected to reveal something of the general form of the
ground, and of the presence of any platforms.

Planimetric data may be similarly treated. Areas between successive
contours are measured, most rapidly and satisfactorily by the plani-
meter itself. 1 When the total area between each pair of contours is
plotted against height, in the same way as for the altimetric-frequency
curve, die resulting figure is the height-frequency curve (to give it
a distinctive name) whereby the general form of the ground is well
summarized. If the several areas between pairs of contours are

1 None of the other means of measurement seems in any way preferable to the use of
this effective instrument.

174 MAP INTERPRETATION

expressed as percentages of the total area measured, the familiar hypso-
metric (or hypsographic) curve is produced. Major topographic
breaks can appear on this curve although less clearly than on the
height-frequency curve which is far more sensitive to small differences
of area. Moreover, the hypsometric curve gives a misleading im-
pression of the form of the ground at the highest levels, for which
reason the clinographic (or hypsographoid) curve has been pro-
posed as an alternative. Here the total area above a given contour
is regarded as a circle* and the distance between the y-axis of the
graph and the point plotted against that contour is proportional
to the radius of the circle. A clinographic curve drawn with true
vertical scale would show the true mean slope between successive
contours.

These several forms of altimetric- and height-frequency curve are
illustrated in Fig. 14.

Relative relief, the vertical distance between the tops of divides
and the bottoms of valleys, may be readily determined from maps.
The map is gridded, the difference of height between the highest and
lowest points in each square is plotted, and the values obtained are
used to construct an isopleth map of relative relief. Such a map is
obviously capable of expressing, in numerical terms, a significant
element of landscape texture.

3. Volumetric Analysis

The expression of relative relief by the method just outlined un-
fortunately takes little account of the form of divides, apart from their
height. It is easy to imagine two contrasted lansdcapes with identical
relative relief, one with narrow valleys deeply cut between broad,
flat-top'ped interfluves, the other with wide, flat-bottomed valleys
separated by narrow, but still high, residual divides ; but the greater
the bulk of rock contained in the divides, the heavier the task of
denudation still to be performed. The volume of divides may be
allowed for by calculating the mean available relief. Available
relief is zero on a surface passed through the bottoms of the large
valleys the streamline surface, which is defined by generalized
contours linking points where the larger rivers cross actual contours.
Now if this surface is planimetered, one may easily calculate the volume
of rock between it and base-level. Similarly, the actual surface is
planimetered and a similar calculation made. The difference between
the two volumes gives the total volume of the divides, which if divided

400

FIG. 14. SOME GRAPHIC METHODS OF SUMMARIZING RELIEF
DISTRIBUTION

(a) Hypsometric curve.

(b) Clinographic curve.

All curves arc drawn from the figures given at the bottom left

Ij6 MAP INTERPRETATION

by the area treated gives the mean height of divide over the whole
area, i.e. the mean available relief.

In practice a certain economy of working is possible. Let the cumu-
lative areas above specific contours be / , / lf l n . Then the mean height
of a planimetered area is given by

where VI is the vertical interval between contours; and the mean avail-
able relief is given by

where M a , M, = / -

for the actual and streamline surfaces respectively.

The first equation is also suitable for use in determining, for
example, the volumes of reservoirs.

4. Clinometric Analysis

None of the techniques so far described gives a quantitative indica-
tion of the roughness or smoothness of the terrain. A rapid and simple
method of calculating the mean slope, regardless of direction, is as
follows: the map is gridded and the number of points at which
contours cross the grid lines is counted; the tangent of the mean slope
is given by

- Vl
G X 3661

where n is the number of grid/contour crossings, G is the aggregate

length of the grid lines in miles, and VI is the contour interval in feet.
The techniques outlined may be varied or extended in a number of
ways, which will not however be described here. Enough has been
said to illustrate the chief ways of handling the cartographic data
embodied in a map. Of the methods listed, the drawing of generalized
contours and of projected profiles, and the calculation of mean slope
can be recommended as rapid in practice and highly informative in
result. They are, in addition, suitable for use with most contoured
maps. As they do not require laborious measurement or computation,
they might well be chosen as the first methods of analysing physical

MORPHOMETRIC ANALYSIS 177

distributions, with the reservation that generalized contours may be
difficult to construct for areas of glaciated highland.

NOTES AND REFERENCES

The general problem of strand-line movements is discussed by

H. BAULIG. The Changing Sea Level. Institute of British Geo-
graphers, Publication No. 3. George Philip, London, 1935.

Some of the types of diagram referred to in this chapter are already
sufficiently familiar from standard texts. For full details of the various
methods, and for examples of their application, see

S. W. WOOLDRIDGE and D. L. LINTON. Structure, Surface, and Drain-
age in South-East England. George Philip, London, 1955. (Generalized
contours: Figs. 15, 27.)

J. BARRELL. "The Piedmont Terraces of the Appalachians." Atner.
Journ. Science, xlix, 1920, p. 227. (Projected profiles.)

A. E. TRUEMAN. "Erosion Levels in the Bristol District," etc. Proc.
Bristol Naturalists' Society, viii, 1938, p. 402. (Superimposed profiles.)

R. F. PEEL. "The North Tyne Valley." Geogr. Journ. xcviii, 1941,
p. 5. (Serial cross-profiles of valleys.)

G. H. DURY. "Remarks on the Migration of Divides in the Neigh-
bourhood of Northampton." Journ. Northamptonshire Natural History
Society, xxxi, 1949, p. 115. (Migrational tendencies of divides.)

D. L. LINTON. "Problems of Scottish Scenery." Scot. Geog. Mag.,
Ixvii, 1951, p. 65. (Summit-plane.)

H. BAULIG. "Les Hauts Niveaux d'rosion Eustatique dans le
Bassin de Paris." Ann. de Geog., xxxvii, 1928, pp. 288, 385. (Alti-
metric-frequency curve.)

S. E. HOLLINGWORTH. "The Recognition and Correlation of High-
level Erosion-surfaces in Britain." Quart. Journ. Geol. Soc., xciv, 1938,
p. 55. (Altimetric-frcquency curve.)

E. ROMER. "Une Nouvelle Representation Graphique de T Hypso-
metric." Comptes Rendus, Congres Internal, de Geographic, Paris, 1931,
Tome I, p. 328. (Clinographic curve.)

J. HANSON-LOWE. "The Clinographic Curve." Geol. Mag., Ixxii,
1935, p. 1 80. (Hypsometric and clinographic curves.)

W. G. V. BALCHIN. Contributions to the Geomorphology of South-
west England. Ph.D. Thesis, University of London Library, 1950.
(Closed contours.)

G. H. SMITH. "The Relative Relief of Ohio." Geogr. Review,
xxxv, 1935, p. 272. (Relative relief.)

178 MAP INTERPRETATION

G. H. DURY. "Quantitative Measurement of Available Relief and
of Depth of Dissection/' GeoL Mag., Ixxxviii, 1951, p. 339. (Available
relief.)

C. K. WENTWORTH. "A Simplified Method of Determining the
Average Slope of Land Surfaces. ' Amer. Journ. Science, Fifth Series,
xx, 1930, p. 184. (Slope analysis.)

The following must be included in a list of references to work on
long-profiles

H. BAULIG. The Changing Sea Level. (Above.)

H. BAULIG. r "Le Profil d'quilibre: Histoire et Critique." Comptes
Rendus, Congres Internat. de Geographic, Cairo, 1925, Tome III.

H. BAULIG. "The Reconstruction of Stream Profiles." Journ.
Geomorph., iii, 1940, p. 3.

V. S. JovANOVi6. Les Profils Fluviatils en Long, etc., Colin, Paris,
1940.

A. AUSTIN MILLER. "Attainable Standards of Accuracy in the
Determination of Pre-glacial Sea Levels, etc." Journ. Geomorph., ii,
1939, P. 95-

As an example of the reconstruction of past profiles, one may refer
to R. F. PEEL: The North Tyne Valley (p. 177). Work with a particular
bearing on the area represented on O.S. 1/63,360 (New Popular) Sheet
1 86, from which Fig. n has been drawn, includes

W. G. V. BALCHIN. "The Erosion Surfaces of North Cornwall."
Geogr. Journ., xc, 1937, p. 52.

J. F. N. GREEN. "The Terraces of Southernmost England." Quart.
Journ. GeoL Soc., xcii, 1936, p. Iviii.

Other references are given at the end of Chapter V.

Papers which, in their several ways, are relevant to the questions
raised frere are

G. H. DURY. "Methods of Cartographical Analysis in Geomorpho-
logical Research." Journ. Indian Geogr. Soc., Jubilee Volume, 1951.

A. AUSTIN MILLER. The Dissection and Analysis of Maps. Institute of
British Geographers, Publication No. 14. George Philip, London,
1948, p. i.

The analysis of aspect, not discussed in the foregoing text, may be
studied in

A. GARNETT. "Isolation, Topography, and Settlement in the Alps."
Geogr. Review, xxv, 1935, P 60 1.

A. GARNETT. Insolation and Relief. Institute of British Geographers,
Publication No. 5. George Philip, London, 1937.

MORPHOMETRIC ANALYSIS 179

Panorama construction is described by

D. SYLVESTER. "A Method of Panorama Construction from
Contoured Maps." Geography, xxviii, 1943, p. 12.

Block-diagrams are considered at some length in

A. K. LOBECK. Block Diagrams. Wiley, New York, 1924.

A briefer treatment is given in

C. B. BROWN and F. DEBENHAM. Structure and Surface, etc. Edward
Arnold, London, 1929. (Chapter XI.)

While the references cited above remain of historical interest, and
while they bear directly on the cartographical exercises suggested in
the foregoing chapter, neither they nor the above text indicate the
nature, aims, and scope of recent developments in morphometry.
Among the papers describing modern applications of statistical method
are

A. N. STRAHLER. "Hypsometric (Area-height) Analysis of Erosional
Topography," Bull. Geol. Soc. Amer., Ixiii, 1952, p. 1117.

A. N. STRAHLER. "Quantitative Slope Analysis," ibid., Ixvii, 1958,
P- 57-

CHAPTER XVI
CARTOGRAPHICAL APPRECIATION

By H. C. BROOKFIELD, B.A., PH.D., AND
G. H. DURY, M.A., PH.D., F.G.S.

Lecturers in Geography, Birkbeck College, University of London

In every work regard the writer's end,

For none can compass more than they intend. POPE

MAPS: O.S. 1/63,360 (SEVENTH SERIES) SHEET 114 (BOSTON
AND SKEGNESS), 1954; O.S. TOURIST MAP, 1/63,360 (LORN
AND LOCHABER), 1959; U.S. GEOLOGICAL SURVEY 1/62,500
(PINAL COUNTY (ARIZONA) CAS A GRANDE QUADRANGLE),
1925; O.S. 1/25,000, SHEET SO/oo, 1948; GERMANY
1/25,000, SHEET 4506 (DUISBURG), 1958; NETHERLANDS
1/25,000, NEW SERIES, SHEET 25A (HAARLEM), 1952

CARTOGRAPHICAL appreciation is the critical assessment of maps. It
involves study of cartographical methods and techniques, and estima-
tion of their success or failure in the representation of land. Like
regional geography, it can be executed on more than one level, but, -
also like regional geography, it depends both upon the use of facts and
upon subjective judgments. In recent years the subjective element in
regional geography has become highly suspect, and strenuous efforts
have been made to eliminate it, but the subjective element in carto-
graphical appreciation has suffered merely by neglect.

To identify the forms of notation used on a particular map is simple
enough, even if the key to symbols is scanty or absent. To identify
the several notations of a number of map-series takes longer, but the
drawing of comparisons still requires nothing but moderate pains.
When the suitability of notations comes in question, however, judg-
ments have to be passed. Although full agreement on all points of
failure, success, and suitability cannot be looked for, critics are likely
to pass many verdicts in common, even though the final task of
appreciation the criticism of the map as a whole will inevitably be

1 80

CARTOGRAPHICAL APPRECIATION l8l

affected by the taste, experience, and aesthetic susceptibilities of
individuals.

The final step is a form of art-criticism. Just as a work of art cannot
be described in terms of an analytical list of its parts, so it is not enough
merely to analyse the range of information presented on a map, and to
identify the scales of notation employed. Art critics can draw on an
extensive vocabulary, which has no parallel in the geographical world
fortunately, perhaps, since the vocabulary of art-criticism can suffer
from rapid changes of fashion and from injections of pretentiousness.
But these very circumstances are little more than responses to a funda-
mental difficulty, namely, that there are very few common words
designed to express what the art critic has to say. A geographer,
possibly untrained in aesthetics and long encouraged to view subjec-
tivity with deep suspicion, can be very badly equipped to criticize a
whole map.

Cartographical appreciation cannot be learned by rule of thumb.
Long practice and considerable patience are both needed. The largely
mechanical business of studying representation is so straightforward
that it might seem pointless, did it not invariably induce respect for
map-makers, reveal the wide range of choice in the designing of maps,
and make clear the wide range of actual practice. We strongly urge
that this analytical work should be performed with scrupulous care,
and that free comparison should be made between different scales of
single surveys, different series of single surveys, and maps of different
surveys on a single scale. We think it quite justifiable to react favourably
or adversely to the first sight of an unfamiliar map, but insist that the
initial reaction should be followed both by a study of notation and by
an explicit statement of a general assessment.

Maps, like words, have uses rather than meanings. Though the
topographical map be the geographer's principal tool, it is made
neither by him nor specifically for him. A modern topographical map
is a compilation of results obtained by field surveyors, office draughts-
men, archaeologists, local-government surveyors, transport officials,
and recording sections of bodies, public and private. In many
countries, the original object of a national survey was military, and
potential military use still strongly affects the design of maps today.
After the military, the next most important class of map-users is
perhaps formed by travellers especially travellers by road. However,
it is now common for motorists to use special road maps, one variety
of the huge group of special maps which includes Admiralty charts,

13 (.5196)

182 MAP INTERPRETATION

aeronautical maps, geological maps, land-use maps, soil maps, large-
scale plans, and cadastral maps. We are concerned in the present
context with none of these. This discussion is limited to the general
map, the ordinary topographical map used in most geographical
interpretation. Since such a map is a common tool, it is quite unreason-
able to expect that it should be designed primarily for geographical
purposes. Unless this principle be kept in mind, criticism may be
unfair.

The published map must omit a great deal of potential material, and
ought to achieve some kind of balance in the material actually repre-
sented. Some of the resulting compromises are, by general consent,
successful: others are distinct failures. Many map series achieve
praiseworthy success with certain classes of information for instance
the representation of height while failing signally in dealing with
others for instance built-up areas. Again, it is quite possible for
individual classes of notation to look clear and handsome in the key,
but to fuse together in the map, to combine in a faint array of pale
tints, or to constitute an over-vigorous and repellent display of solid,
clashing polychrome.

At the outset the critic is liable to take his own national maps as the
norm, simply because he is used to them. Adverse criticism is then
directed at foreign maps which use unfamiliar systems of notation, or
which omit reference to certain kinds of information. As foreign
maps become familiar, and as the critic achieves a certain sophistication
of cartographical taste, tolerance will grow; but full tolerance is
uhlikely to be attained without experience of using foreign maps on
their own ground. To take an example, the 1/50,000 map of Sweden
relies on black hachures and black form-lines for much of its representa-
tion *of relief, and seems to offer little to the map-reader trained on
contours. In fact, however, the techniques employed are admirably
suited to the terrain, and it is easy to follow a cross-country route
across the terraced drift and glaciated rocky bosses of south-west
Sweden with the aid of a standard map.

This is merely one of innumerable examples of the variation of
cartographic need from country to country. On the negative side, a
given survey will never need to consider the mapping and representa-
tion of some classes of feature: symbols for ricefields or vineyards are
not required by the Ordnance Survey of Great Britain. On the positive
side, features which could be represented on the finished map may vary
in apparent significance from country to country, so that there is variety

CARTOGRAPHICAL APPRECIATION 183

in the care, frequency, and prominence with which particular classes
of item are shown. It is to be expected that watercourses of all kinds
will be indicated on maps of Holland and Belgium, where they are sub-
classified far more elaborately than on British maps of comparable
scales. In this general connexion, the critic can usefully note which
classes of feature seem to be given special attention, although he cannot
go so far as to reconstruct the decisions on policy taken by the map-
designers. Furthermore, it is often possible to comment constructively
on the treatment of features which are always included buildings,
relief, and lines of communication among them and to identify
accomplishment or deficiency.

Numbers of surveys have traditions of their own. They have
developed highly characteristic styles, which enable their products to
be identified at once. Several factors contribute to the development,
and to the change, of style. Style is affected by the inspiration or
competence of heads of surveys, by technical developments in methods
of surveying and in methods of printing, and by the financial strength
of the survey organization. At any time, the current style usually
represents an evolved form of styles formerly used, as can readily be
seen from the changing practice of the Ordnance Survey. Wholesale
changes are rare; it is quite common to find ancient and modern
techniques used together on current maps. Thus, on the O.S. Tourist
Map (1/63,360) (Lorn and Lochaber), relief is shown by contours,
rock-drawing, partial layer tinting, and a modernized kind of oblique
illumination, but the woods, orchards, rough pasture, quarries, and
sandhills are marked by symbols which would not have been out-of-
place on the maps of private surveyors in the late eighteenth century.

Although many modern surveys use numerous common techniques,
there is no reason to suppose that map styles will so converge as to
become identical. So long as the perfect map remains unmade, so long
as there is no agreement on cartographical perfection, and so long as
improvements in printing-processes continue, styles of cartography
will persist. The way in which technical improvements occasion
changes, and the way in which problems of representation encourage
experiment, are well exemplified by the changes introduced by the
Ordnance Survey in the quarter-inch series. Sheet 10 (North Wales
and Lancashire) of the new 1/250,000 Series, published in 1957, owes
much to the true quarter-inch map which it superseded (1/253,440
Map, Sheet 4, North Wales and Manchester, published in 1935). But
the style of the Wales and Manchester Special Sheet of the 1/250,000

184 MAP INTERPRETATION

Map, published in 1959, strongly resembles the style of the current
Tourist sheets on 1/63,360. All four maps, however, are recognizable
as products of the Ordnance Survey.

Map style can vary sharply from country to country. In view of
the mixed emotions aroused by references to nationality, nationalism,
and national character, we wish to state firmly that national styles of
cartography are just as likely to develop as are styles of book-publica-
tion, cartoon-drawing, or poster-painting. A possible influence here
is that cartographic style within a country will be affected by the style
of the government's maps, so that people come to expect the kind of
map provided by the national survey. If so, nationals of a country
may well be far qasier with their national maps than are geographers
who come to the maps as foreigners. Growing familiarity with style
will remove difficulties of map-reading, but can do nothing to help
the critic in assessing the impalpable quality now in question. To our
hopes that the examples of appreciation given later in this chapter will
be of assistance, we link a plea for common sense; some maps are
certainly too skimpy, others are just as certainly overloaded, but is it
not rational to complain that a map is already congested and at the
same time to deplore numerous omissions.

Scale

The scale of the map is the most important factor in determining
the amount of detail that may be shown and the order of accuracy
that can be attained. Topographical scales have been variously
defined, but between 1/20,000 and 1/80,000 may be taken as limits for
most purposes, while about 1/50,000 is the optimum. Within these
limits of scale a map can show all the larger, essential features of the
landscape such as may be seen by a walker, including all buildings,
and every road and track. The larger the scale, the greater the amount
of detail that may be represented, at the expense however of a larger
coverage, and also of that wider and more synoptic view of the major
features of the countryside which can be gained from smaller-scale
maps. For example, in embarking on the morphological division of
an area the size of an English county (see Chapter II), it would be
unwise to attempt the preliminary division into sections with a map
on a scale greater than 1/125,000. As subdivision proceeds, however,
larger- and larger-scale maps need to be employed, until for the final
subdivision into stows a map on about 1/50,000 and preferably on
1/25,000 is required.

CARTOGRAPHICAL APPRECIATION 185

Within general range of topographical scales, there are four chief
scales in common use

1. 1/25,000, long used by Surveys on die Continent, notably the
German, and recently adopted by the Ordnance Survey. This
scale was found to be of particular use during the recent war, when
it was widely employed.

2. 1/50,000, used by several Continental Surveys as the standard
topographical scale.

3. 1/62,500, an approximate i-in. mile scale employed by the
U.S. Geological Survey. This scale has the advantage of being a
factor of 1,000,000, the series running 1/62,500, 1/125, ooo, 1/250,000,
1/500,000, 1/1,000,000.

4. 1/63,360, the exact i-in. mile scale employed by the
Ordnance Survey and most Commonwealth Surveys.

The latter is, perhaps, a rather small scale for topographical maps in
so closely settled a country as Great Britain, and the somewhat larger
1/50,000 scale would make for less crowding on maps of densely
settled areas.

Maps on these scales deserve close study and comparison. A useful
exercise is to examine the same piece of country as represented on
maps of different scales. By this means, the oft-repeated question on
the "limitations of scale" can best be approached. The problem of the
limitations of scale has its mechanical aspects, which are beyond the
scope of this chapter, but is also directly relevant to the subject of
cartographical appreciation. Detail such as field boundaries and full
land-use information is quite appropriate on a scale of 1/25,000 but
can heavily overcrowd a topographical map on a smaller scale. Full
land-use information is nowhere attempted on O.S. topographical
maps, but appears on many continental maps, for the widespread
prevalence of traditional peasant farming in Europe prevents the
land-use pattern there from changing rapidly. Even then, this informa-
tion is barely comprehensible on some of the sheets of the French
1/50,000 map, despite a generous use of colour, and is definitely
excessive on the all-black German 1/100,000 map, although the
draughtsmanship and printing are remarkably fine. On the German
1/25,000 map, however, land-use information is beautifully clear. The
two German series should be compared in this respect.

To take another example, the bends and twists of the "rolling
English road" cannot adequately be represented on topographical

186 MAP INTERPRETATION

scales, and are generalized even on the O.S. 1/25,000 map. Road
widths are greatly enlarged on the smaller-scale maps, the degree of
distortion increasing as scale diminishes, for the smaller scales are of
particular use to travellers, and other detail is subordinated. In 1945,
at the rime of the launching of the new 1/25,000 map, there was a
certain amount of controversy over the question of exaggerating road
widths on this scale. The Ordnance Survey held that exaggeration
was necessary in order to make the map of use to the traveller, but
critics maintained that it would destroy the advantages of pin-point
accuracy on a topographical scale which was one of the major features
of the new map.

The Representation of Relief

This is, to the geographer, the most important single function of the
topographical map. There are many methods, but most are variants
of the three basic techniques of hachuring, contouring, and oblique
illumination. Oblique illumination is in part a survival from ancient
maps, where hills were drawn in profile, one side being shaded to
give an impression of bulk. The method assumes a light placed
close to the surface of the ground in the north-west. All east- and
south-facing slopes are therefore in shadow. To-day, this method
is seldom employed except in conjunction with more accurate
devices.

Hachuring also is a semi-pictorial method. Basically, hachuring
consists of a series of parallel lines, drawn at right-angles to the direc-
tion of slope and usually of a thickness and intensity roughly propor-
tional Jto the steepness of the slope. The eighteenth-century
cartographer, Lehmann, developed a precise scale of hachuring
exactly proportioned to gradient, and a scale of this kind is employed
on the topographic sheets of The Netherlands. Some remarkably
fine hachuring was employed on the early maps of the Ordnance
Survey, though its effectiveness was much greater in lowland than in
upland country. The very finest use ever made of hachures was in
the Swiss "Dufour" Series. Bad hachuring can be very heavy and ugly.
The hachuring of the Austrian General Staff maps the only available
maps for large parts of central and eastern Europe for many years is
bad. Clumsy or inadequate hachures fail to bring out the distinction
between steep and gentle slopes, and can never be made to give any
precise effect of elevation. The hachuring of the Third Edition of the
Ordnance Survey of Ireland is a good example of feeble hachuring

CARTOGRAPHICAL APPRECIATION 187

which is largely ineffective. 1 The very worst form ever taken
by hachuring is the so-called "hairy caterpillar" a ridge shown
by a heavy, smudgy, overcrowded mass of hachures, inaccurate
and meaningless in itself and obscuring all other features in the
neighbourhood.

Hachures can be combined very effectively with other methods,
as will be shown below.

Hill shading is merely a tone shading of intensity proportional to
the degree of slope. The principle is the same as in hachuring, and
hill shading is often used in place of hachuring on smaller-scale maps.
Very frequently it is combined with oblique illumination to give a
good plastic impression of relief.

For potential accuracy, no method of relief representation can equal
the contour, a line passing through points of equal elevation. Its
apparent scientific accuracy has dangers, however, for many contours
convey an impression of precision which they do not in fact possess.
On O.S. maps, for example, only the so-ft., ioo-ft., 2oo-ft., to i,ooo-ft.,
i,25O-ft., i,5OO-ft. contours are actually surveyed on the ground. The
others are interpolated from spot heights and sketches, and are of a
much lower order of accuracy than the instrumental contours. The
same criticism is true of the intermediate contours drawn on the new
provisional 1/25,000 series, which ate seriously inaccurate in places.
In the new resurvey of Britain, now taking place, all contours will be
instrumentally surveyed.

The contour interval the vertical distance between contours is of
the greatest significance in the value of a map. A map having contours
100 ft. apart may miss many significant landscape features altogether.
On O.S. 1/63,360 maps of the Fourth Edition, and later, the interval
is 50 ft. ; on the new 1/25,000 map it is 25 ft. The French and Swiss
1/50,000 maps both employ contours at a lo-metre interval, while the
American Geological Survey normally employs an interval of 20 ft.,
decreasing to 5 ft. in some low-lying areas such as the Mississippi
delta, and to I ft. in part of the Texas coastal plain. A remarkably
detailed and vivid impression of relief can be gained from a closely
contoured map, and the very clear picture of physical features
obtained makes this series one of the best in the world for the study
of geomorphology.

Some of the most artistic results are obtained from combinations of
two or more of these methods. The O.S. 1/63,360 Fifth (Relief)
1 See Sheet 169, used elsewhere in this book.

188 MAP INTERPRETATION

Edition, first published in 1929, employed contours at 5O-ft. intervals,
together with buff hachuring and a grey overprint to the hachures on
the south- and east-facing slopes to give the effect of oblique illumina-
tion. There was also layer tinting in buff, the tint changing at
each 500 ft. The result was a very effective map, but it did not appeal
to the public and was never completed for more than a small part of the
country. Some of its finest sheets were in south-east England, particu-
larly in the North Downs area, and one or two very beautiful special
sheets were issued for Scotland, such as the Oban and Cairngorms
Sheets. This latter map employed brown hachures and brown tinting,
and provided a remarkably beautiful and effective map of wild and
spectacular country. Unfortunately the whole series is now out of
print.

The Ordnance Survey also issued a number of 1/63,360 Tourist
Sheets during the inter-war period, of special areas, and usually
employing contours in brown at 50-ft. intervals, layer colouring in
green and brown, and hachures in brown. These, though effective,
were less attractive than maps of the Fifth (Relief) Edition. One of
the present Tourist maps is discussed below.

The Swiss 1/50,000 is perhaps the most effective of all maps in the
representation of relief. Contours at lo-metre intervals are in brown,
except for blue on ice and permanent snow, and black on scree. In
addition, this series is characterized by remarkably good rock-drawing
on broken rock slopes which are too abrupt to be represented by
means of contours. Rock-drawing here reaches its finest development
that on British maps is very poor by comparison. In addition,
the Swiss map employs a faint brown hill shading, with a grey
overprint on the southern and eastern slopes, except where ice occurs,
when blue is substituted. The draughtsmanship is impeccable. This
map should be studied most carefully as a very fine example of
cartography.

The relief of the land may also be depicted by means of point
elevations. On O.S. maps these fall into the three groups of trigono-
metrical points, bench marks, and spot heights. Trigonometrical
points are stations in the triangulation of the country. Normally they
are found on hilltops commanding a wide view. They appear on the
ground as concrete pillars having a metal tripod-rest on top and a
bronze plate carrying a bench mark let into die side. The elevations
of the bench marks are known exactly. Bench marks at other places
are elevations of the second order usually marked in a wall or

CARTOGRAPHICAL APPRECIATION 189

gate-post. They are usually one or two feet above the surface of the
ground. Spot heights are mere points on the map, usually strung
along roads, and they provide a useful guide to the form of the ground
between contours, of particular use in interpreting the existence of
river terraces and/or erosional flats. Those actually shown on the
map are a mere fraction of the total number available on the original
survey.

The Representation of Cultural Features

In their broadest sense, the features of occupance cover the whole
landscape of settled lands. The very form of the woods, and in some
areas the details of the stream courses, are the product of man's work.
Although the "cultural overlay" is as complete as the surface of the
land which underlies it, the object of cartography is not to represent
the whole, but to select those elements of the overlay which are of
greatest significance to the user of a topographical map, and to re-
present them in such a way that their pattern is clear, without obscuring
the details of the underlying relief. In great urban areas the cultural
overlay is the whole map, and here in many cases the representation of
underlying relief is greatly subdued. On the O.S. 1/25,000 map
contours are shown by broken lines in built-up areas, while, on the
Tourist and Fifth (Relief) Edition maps of London, hachures and
tinting were largely omitted, and relief was shown by contours only.
Thus the whole of the Crystal Palace ridge, to quote an example,
disappears from all but the closest scrutiny.

Upon the success or failure of the cartographer in the task of properly
representing the cultural overlay in balance with the physique depends
most of the value of the map for the purposes of geographical inter-
pretation, for the cultural patterns must be both clear in themselves
and clear in their relationship to the underlying terrain if correlations
are to be understood and the basic purpose of map interpretation
achieved.

The topographical map was, and is, intended primarily for military
purposes, and must represent all those features which have military
significance. Thus buildings, roads, railways, and field boundaries of
all kinds are included wherever possible, and in as much detail as the
limitations of scale will allow. Different policies have been adopted
on this point by the different national surveys: the Ordnance Survey
has been slow to represent field boundaries on its topographical
scales, and did so for the first time on the new 1/25,000 map in 1945.

I9O MAP INTERPRETATION

All other features of the map, however, including the emphasis placed
on the classification of roads and trackways according to the traffic
they will carry, and the extra significance given to prominent buildings,
spring from a military origin, even though the maps have been adapted
to the needs of die walker and tourist.

Of first importance in the representation of cultural features is the
nature of the symbols employed. Ideally, cartographical symbols
should be self-explanatory, requiring no reference to the key. Many
symbols show in plan the features which they represent, for example
the symbols for buildings of all kinds, and for bridges, roads, and
railways. Others, such as the symbols used for pylons, windmills,
lighthouses, and (on many maps) for trees, are drawn in profile.
This practice marks the persistence of a venerable tradition: such
symbols are usually retained on modern maps only for particularly
prominent objects.

A very large range of symbols is employed on the modern map.
The increasing use of maps by travellers has led to a great development
in the representation of routeways. Railways are usually classified
according to number of tracks, and roads are now very fully classified
in respect of width and condition, or according to an official classifica-
tion adopted by the transport authorities. Guide- and mile-posts also
are frequently marked. Many foreign surveys attempt much more
detail than does the Ordnance Survey.

Appreciation of cultural representation on maps demands careful
study of the characteristic sheets of the major national surveys and of
representative maps. In this way only can the student acquire that full
and wide knowledge of the different methods employed on which
enlightened appreciation and comparison may be based.

Non-landscape Features

Names, boundaries, and the classification of land by administra-
tion or ownership are the principal non-landscape features commonly
represented on topographical maps. In general, the relative importance
of these features increases with the scale of the map, for on the larger
scales there is both more room and greater need for that class of
information which is of use particularly to the indoor map reader.
The characteristic sheet of the O.S. 1/2,500 map includes fifteen
different administrative boundary symbols with fourteen corresponding
styles of lettering for use in naming places and districts, while eight
further styles of lettering are employed in naming topographical

CARTOGRAPHICAL APPRECIATION 191

features. The 1/10,560 (6in./i mile) map uses a similar range of
styles, but the 1/63,360 map employs only three styles of boundary
and correspondingly fewer styles of lettering. The new 1/25,000 map,
which is primarily topographical, attempts a compromise between
the two adjacent scales of symbols for non-landscape features, with
but indifferent success. Only three styles of boundary are shown, as
on the 1/63,360, but the naming of districts and places corresponds
more closely to the system on the 1/10,560 than to that of the 1/63,360,
with the result that the names of most districts appear twice, once in
upright capitals for the place itself, and once in sloping capitals for
the administrative area. If there is also a small part of the larger
district known by the same name as the whole but requiring separate
identification, the name may well appear a third time, in smaller
print with lower-case lettering.

In the assessment of non-landscape features one must first consider
how far the information provided suits the scale of the map and the
type of terrain. Thus the representation of the least significant non-
landscape features would be justified in a map of a part of central
Australia, but the same detail would be quite out of place on a map of
south Lancashire on a similar scale. Just as it is easy to overload a map
with such information, it is equally easy to be too sparing. The
American Geological Survey maps are grave offenders in respect of
omission.

It is important that the information given should have some degree
of permanence. The Ordnance Survey has decided against the
practice adopted by some continental surveys of giving the population
of individual communes by means of a figure on the map. Without
frequent revision, this information rapidly becomes out-dated, and
must be used with great care.

The method and style of representation vary widely. Some reference
has already been made to the common practice of employing different
styles of lettering to convey different information. On the O.S.
1/2,500 map, different styles of lettering are used to denote counties,
county boroughs, parliamentary county divisions, poor law unions,
parliamentary boroughs, municipal boroughs, wards, urban districts,
rural districts, civil parishes, towns, and districts. Some national
surveys vary the style of lettering employed according to the popula-
tion of the place named, but this is open to the same objection as
stating the population itself. A style based on administrative status is
better. A wide variety of styles of lettering is possible. The interested

192 MAP INTERPRETATION

reader should consult Captain Withycombe's paper of 1929 (see
p. 202), in which the evolution of the present forms of lettering is
traced from early times. The modern tendency is toward lighter and
more open lettering, and it is a great pity that the international
1/1,000,000 map, with its very wide and varied use of lettering, was
designed in the period of heavier styles before the movement of
reform was fairly started.

Even now, the lettering employed by the Ordnance Survey on its
modern maps (Fifth, Sixth, and Seventh Editions, and 1/25,000) is very
slowly and carefully built up with fine steel pens. A practised draughts-
man is unable to draw more than twenty names per day. The Royal
Geographical Society has introduced a quicker style adapted to quill
pens, but this has not yet found favour with official bodies. Style of
lettering varies greatly from country to country. In analysing and
appreciating its value account should be taken of differing national
styles, and also of the age of the map, for the reform movement is not
yet twenty-five years old.

There is one further important aspect of names on maps spelling.
The modern map user, accustomed to the standardized forms of
spelling, might easily fail to realize that these have in many cases been
arbitrarily imposed, and are often at variance with a long tradition in
which a different spelling, or even a different name, was employed.
The modern "Dorking," for example, was for centuries "Darking,"
while the names of many rivers were taken from the eighteenth-
century topographers and not from local custom. In Gloucestershire,
the village of "Aston Blank" (O.S. 1/63,360 (New Popular Edition)
Sheet 144, Grid Reference 1320) is still known locally as Cold Aston,
this being also its name for postal purposes. Nineteenth-century
bowdlerizations of local names are almost always enshrined on O.S.
maps, perhaps to the detriment of interest. There is also the factor of
human error, as where High Barnet Station appears as High Barton
Station on Sheet 51/29 of the Provisional Edition of the 1/25,000 map.
Such errors as this are usually speedily corrected.

The arbitrary rendering of place-names has far-reaching results.
Consider the rendering of Gaelic names in different parts of western
Britain. In Ireland, very few Gaelic names are given in their original
forms: almost all have been done into a kind of phonetic English
which gives rise to frequent absurdities. Welsh names suffered less,
while in many parts of north and west Scotland Gaelic place-names
are correctly spelt out, even though the Gaelic tongue may have long

CARTOGRAPHICAL APPRECIATION 193

since vanished. The possibilities and dangers of place-name inter-
pretation from maps have been dealt with elsewhere, but it may be
well to repeat that the map interpreter should beware of taking the
present form given on the map as etymologically correct.

The classification of land according to administration or ownership
also falls into the category of non-landscape features. The modern
O.S. topographical maps represent, by a clear symbol, all land in the
possession of the National Trust, and on the larger scales all public
buildings are distinguished in heavy black. Similar conventions are
also employed on foreign surveys and are of considerable advantage
to the map user.

Margins and Marginal Information

In former times map margins were most elaborately decorated, but
in these more austere days they have lost almost all artistic detail.
Even to-day, however, there are few maps which are bounded merely
by straight lines, without any form of embellishment whatsoever.
Commonly, margins are divided according to scale so as to provide
an easy means of measurement. The margins also contain reference
systems of latitude and longitude, grid or otherwise. Sometimes
decoration may still be found.

Margins have the dual function of limiting the sheet and linking it
to the next. For this reason some information is carried from the map
across its margins, such as road names on large-scale plans, and road
and railway destinations on smaller-scale sheets. On some maps,
margins are opened where they cut inconveniently across some major
feature, such as a city, a coastline, a river, or a main highway. This
has been a practice particularly of the French 1/50,000 series. The
Ordnance Survey have employed it widely only once, in die special
sheets of the Fifth Edition published during the later 1930'$.

Again, the margin includes much information of great importance
in the proper understanding of the map. The title of the map is
normally stated in the upper margin, together with details of the
series to which it belongs and its number within the series. Some
larger-scale maps carry no tide, merely their number. At the foot
of the map it is customary to give the scale, usually in all three
forms verbal statement, representative fraction, and scale bar, the
latter usually divided in metric units as well as the national units if
these are not metric. A key should be given, including the commonest
symbols used on a map and also the most obscure. As it is clearly

194 MAP INTERPRETATION

impossible, within the ordinary limited margin, to provide a full list
of all signs and symbols employed, a characteristic sheet is usually
published separately; but unless a brief key is given the map loses
some of its usefulness.

Other information that must be provided includes a north point,
with grid north, if any, also shown. Magnetic north is commonly
given as at the date of publication, and a statement appended of its
approximate variation from year to year. Secondly, the date of
survey must be stated. Much that has been written above emphasizes
the fact that a map is already out of date when published, and becomes
progressively more out of date as time goes on. A note of the date is
particularly important in an area of rapidly expanding urban settle-
ment. The dates of any revisions are commonly stated, but frequently
the extent of the revisions are inadequately described. Revision may
be exhaustive, or it may apply only to certain features, such as roads
and railways: a fact of this kind should be stated. If a map has been
compiled from several surveys the date of each must be given. With-
out this the map loses much of its value. Omissions under this head
are a common weakness of much private cartography, and of the
Provisional O.S. 1/25,000 map which is a reduction from previous
1/10,560 and 1/2,500 surveys, so that parts of any one sheet may have
been revised later than others. This is not at present stated on the
maps, to their detriment in detailed work.

Thirdly, some description of the grid-reference system, if any,
should be provided, and the datum of all elevations should be stated
clearly and precisely. The Sixth Edition of the O.S. 1/63,360 map
merely* states that heights are given in feet above mean sea-level a
datum which has no existence as a single plane for the whole country.
The present Ordnance Datum is mean sea-level at Newlyn, Cornwall,
that is, alL heights are referred to the mean between high and low
water observed at Newlyn for a number of years. But the fenland is
low-lying with respect to adjacent sea-levels. In fact the base-level for
drainage at any point is the level of low-water spring tides at the out-
fall, while "sea-level" from the point of view of the reclamation of
land and its protection from the sea is the level of high-water spring
tides at each point of the coast. Along considerable stretches of the
shoreline of Britain these two levels are more than twenty feet apart,
and reach forty feet apart in the upper Bristol Channel. Students
unaware of this have interpreted the levels east of Newport, Mon., at
20-22 ft. above O.D., as a raised beach, and are perplexed to find the

CARTOGRAPHICAL APPRECIATION 195

25-ft. contour of the 1/25,000 map crossing the tideway of the Severn
below Gloucester. These facts are of some significance to the map
interpreter, and it is important to note that the topographical map
provides no information about them whatsoever. Admiralty charts
must be consulted in each case where such a problem arises.

A similar consideration applies to the interpretation of foreign maps.
For example, in maps of the Netherlands all heights are referred to
the summer water-level at Amsterdam; those in Belgium originate
from low water of spring tides in Ostend Harbour; the German
datum is referred to a point below Berlin observatory. To correlate
the elevations given on the maps of different countries is, in effect, a
problem for geodesy.

Final points of marginal information should include the name of the
projection used, the origin of any grid system, the origin of the
longitude by no means always Greenwich and also the price.

Examples of Appreciation

O.S. 1/63,360 (SEVENTH SERIES) SHEET 114, 1954
(BOSTON AND SKEGNESS)

Relief is shown by brown contours at intervals of 50 feet, strengthened
at every 250 feet. Point elevations in black include two kinds of spot
height; freely distributed on the low ground, they are helpful when
it is recalled that they occur inland mainly on embanked roads. Con-
tours for the hilly ground in the north are elaborate enough to suggest
accurate survey; they are finely drawn and adequately numbered.

Bright blue is used for water-names, natural rivers, creeks on the
foreshore, and the marsh symbol. Canals, drains, and artificial rivers
are in pale blue, between double bright blue lines if they are wide.
Sea and tidal estuarine water is in dull pale blue. Symbols are provided
for quarries, pits, sandhills, cliffs, rocky foreshore, sand- and mud-
banks, and accumulations of shingle, the last being omitted from
the key.

Cultural information is abundant. Roads are subdivided into seven
groups, partly by Ministry of Transport classification and partly by
width. Main roads appear in red, secondary roads in brown, and minor
country roads in yellow. For farm roads, access roads, and minor
roads in towns no colouring is used. Railways are classed in four
groups. The bold red cross formerly printed for level-crossings has
been abandoned; on this map, the road and rail symbols thin to points.

Ip6 MAP INTERPRETATION

Churches are indicated in solid black symbols; other town buildings,
and houses in the country, are in grey, but outbuildings and isolated
barns in country areas are in black. Bus and coach stations carry a red
symbol. Information on land use is scanty; apart from items already
mentioned, it is confined to stipple for parks, a simplified tussock
pattern for rough pasture, a blue cross-hatching in a black frame for
glasshouses, a grey pattern of symbolical trees for orchards, and
symbols for coniferous and deciduous trees under a green wash for
woodland.

Remaining landscape features relate chiefly to service installations
and to antiquities-; The style of lettering varies among pre-Roman,
Roman, and post-Roman antiquities. Service installations listed in
the key include three kinds of telephone kiosk, two of which belong to
motoring organizations; Youth Hostels are also marked. Masts,
windpumps, windmills, lighthouses and lightships are all shown by
pictorial symbols.

Three types of boundary appear in black. Finely-drawn and
unobtrusive grid lines are in grey. Latitude and longitude are given by
linear scales in the margin, and by intersections at 5-minute inter-
vals on the face of the map. Marginal information includes a revision
diagram, information about the grid and about the true, grid, and
magnetic norths, scale bars, county names, distances to nearby towns
off the map, and a beautifully-lettered key.

Lettering on the map itself varies in an obviously systematic but
unexplained way. Minor names generally are in sans serif in which all
strokes are of uniform thickness; this lettering is clear but not elegant
contrast the small lettering in the key. Use is made of sloping and
upright characters to sub-classify features, but the variation, like
variation in kind of lettering, is not keyed.

In several ways, this map is far more attractive and far easier to use
than the New Popular sheet which preceded it. Its general effect is by
far the less harsh, largely because of the free use of quite small lettering
and because of the liberal employment of grey. It would be unjust to
call the sheet pallid without comparison with other sheets of the same
series where great amounts of complex detail have to be shown. Never-
theless, it can fairly be said that the tangle of minor roads in the silt
fen is not easily perceptible at any distance in artificial light, while in
close-up the map seems overloaded with a not particularly attractive pale
yellow. A slightly more solid tint, and narrower lines, might have
been better. In rural areas, too, some individual buildings seem unduly

CARTOGRAPHICAL APPRECIATION 197

large. On the actual sheet used in the writing of this text, the grey
plate had failed to register correctly, especially in the north-east.

O.S. 1/63,360 TOURIST MAP (LORN AND LOCHABER), 1959

Since much of what has just been said of the Seventh Series Sheet 114
applies also to this Sheet, attention will be concentrated on the tech-
niques of representing relief, which are special to the Tourist Map.

Brown contours at intervals of 50 feet, strengthened at every
250 feet, are carried right up to the summits. The intermediate contours
are so finely drawn that they do not congest the map, yet they are
strong and clear enough to be easily read. The strengthening does not
produce that oyster-shell effect to which it has led on some past maps.
Since the green wash used on the Seventh Series for woodland is here
applied to layer-colouring for areas under 100 feet, the woodland tint
has been darkened and the tree symbols strengthened, but contours can
almost everywhere be read through the darker green with complete ease.

Apart from the partial layer colouring, this map is distinguished by a
highly effective use of oblique illumination and by the generous
application of rock-drawing. While the rock-drawing is inferior to
the best products of Swiss cartographers, it seems to be more varied
and more elaborate than on past productions of the Ordnance Survey.
In oblique illumination, on the other hand, the Survey has here suc-
ceeded beyond praise. The plastic shading ranges, by gradation, from
very pale buff to light yellow-brown on the lit slopes, and is purple-
brown of varying intensity on the shaded slopes. Summits are in part
untinted. Seen from a distance, the map achieves a most convincing
three-dimensional effect. In close-up it is no less effective, for the
shading has been applied with great care and delicacy, quite small
changes in slope or aspect being accurately reflected in changes of
tinting.

Although buildings are marked in grey, their firm but finely-drawn
black frames make them identifiable. It is, of course, unlikely that the
Tourist style would be applied to maps of heavily-peopled areas, so
that one possible cause of congestion does not arise. In any event,
allowance has been made for the free use of shading, for example in
the strengthening of the footpath symbol on the map, but not,
however, in the key.

It continues to be a matter of regret that submarine contours are not
shown below the lo-fathom line, that submarine contours are in
fathoms while subaqueous contours for freshwater lochs (not keyed)
14 (.5196)

198 MAP INTERPRETATION

are in feet, and that defects of spelling should recur Inversanda (9359)
should be Inversands. Such deficiencies, however, do little to detract
from the fine achievement of the whole map.

U.S. GEOLOGICAL SURVEY, 1/62,500, EDITION OF 1924
(FINAL COUNTY (ARIZONA) CASA GRANDE QUADRANGLE)

The representation of relief is by means of contours in brown at
25-ft. intervals, every fourth contour being strengthened, and by
point elevations of three types triangulation stations (trigonometrical
points) and bench marks, both in black, and spot heights, mainly at
the intersections of the land-survey lines, given in brown. In addition,
sand dunes are indicated by means of a brown-stipple symbol. The
impression gained is one of great accuracy, notably on the lowlands.
Water is shown in blue, intermittent streams being shown by blue
pecked-and-dotted lines and wells by blue circles.

Cultural information is very sparse, partly because of the very low
density of settlement in this area. Apart from the classification of
roads, black alone is employed. All buildings are shown in solid
black, their forms normally being generalized as small squares. Roads
appear in fine hair lines, broken in the case of unmetalled roads.
Classified through-routes and secondary routes are overprinted in red,
and, if in bad condition at the date of survey, are lined with red dots.
Railways are shown by hatched lines, with double and single tracks
separately distinguished. On the whole, cultural information is
definitely subordinated to the relief, and the occupance seems sparser
than it in fact is. Detailed examination reveals, however, a clearly
and very precisely drawn map of cultural features, of which a very full
interpretation may be made.

Non-landscape features also are sparingly represented. Eight forms
of boundary are shown, but naming is restricted to the larger adminis-
trative units, such as the Indian Reservations, and to more important
settlements, such as Casa Grande and Chiu-Chuischu. Individual
farms and ranches are not named, nor does there appear on this map
the profusion of rural schools that is so marked a feature of maps of
the eastern U.S.A. Styles of lettering employed are clear and effective*

A graticule is printed on the face of the map, and it is difficult to
separate this at a glance from the section lines which are very clearly
marked in black. The margins are plain, and are used for graticule
information, scale bars in miles and kilometres, a grid which is not

CARTOGRAPHICAL APPRECIATION 199

carried across the map, and some information regarding the destination
of roads (printed in red). North point, date of survey, projection, and
datum are given, the latter rather inadequately. A full characteristic
sheet, together with a description of the topographical maps of the
U.S.A., is printed in grey on the reverse.

In conclusion, one may say that this is primarily a relief map, second
importance being given to land-survey lines and the cultural features
being under-emphasized. Close examination, however, reveals a
detailed and accurate map of the cultural landscape, on a map which,
for relief representation, has few equals among the topographical maps
of the world.

Comparison of Map Series

It is customary in cartographical appreciation to refer freely to
comparable series of other surveys. The interpreter needs to be closely
familiar with such series and their characteristic sheets. In addition to
O.S. maps, the major productions of all the national surveys of Western
and Central Europe and North America should be studied. In particu-
lar, the maps of France, Germany, Switzerland, the Austrian General
Staff, and the U.S.A. deserve close examination. The object of this
concluding illustration is to suggest lines on which comparisons may
be made between maps of different countries on comparable scales,
in order to obtain a better appreciation of any one.

At the outset, a brief summary of the characteristics of each map
will be given.

GREAT BRITAIN. O.S. 1/25,000, SHEET SO/oo, 1948
(MERTHYR TYDFIL)

Relief is by point elevations in black and by contours in brown at
25-ft. intervals, each fourth strengthened. Rough rock, tip-heaps,
etc., are given in black, water in blue. Woods are represented by a
tree symbol in grey. All major roads, tracks, and footpaths are shown
in black, and the classified roads filled in brown. Buildings are out-
lined in black and filled in solid grey, except for public buildings which
are in solid black. Many names are recorded, some providing much
industrial information. Almost all farms, factories, inns, woods and
physical features, and many antiquities are named. The result is a
crowded, but clear, map, the relief being subordinated to the cultural
features in closely built areas, but dominating in rural areas.

200 MAP INTERPRETATION

GERMANY 1/25,000, SHEET 4506 (DUISBURG) 1958
(PART OF NORDRHEIN-WESTFALEN)

Brown contours are thickened at every 20 metres, broken at odd
5 metres, and supplemented by form lines in places at intervals of
i -25 metres. Some waterside slopes are hachured in brown; water is
in blue, and woodland is shown by black tree symbols overprinted in
green. All other information is given in black. Suburban building is
solid black, with gardens in light black stipple and an unobtrusive
orchard symbol; congested building is hatched in black, and factories
are in solid black.

Both the range of notation and the key are elaborate. Four types of
administrative boundary, four types of railway line, and eight types of
road (excluding footpaths) are distinguished. Fourteen classes of land-
use information are indexed, three of them further subdivided and two
others capable of variation: these are all additional to gardens. Nine
types of linear feature, not already specified, include seven types of
enclosure-boundary. Further symbols, mainly for landscape features,
are classed under twenty-nine heads and number forty-three in all,
and thirty-four literal abbreviations are listed. The key to water-
ways and associated features gives another thirty-odd explanations.
Remaining marginal information is equally carefully presented.
Free use is made on the face of the map of variations in style of
lettering.

NETHERLANDS 1/25,000, NEW SERIES, SHEET No. 25A
, (HAARLEM), 1952

Relief is scarcely present in the area represented, except in the belt of
sand-dunes in the west. Sand is marked by dull yellow which vanishes
under thfc green wash for woodland; plastic shading for dunes, in a
purplish brown, seems to have been applied by air-brush and thus to
be rather coarse, although it is likely that the form of the ground is
accurately indicated. Black woodland symbols resemble those on the
German map, but are even further simplified. Part of the polderland
is marked in green, part left blank, without explanation. All water is
in a uniform blue.

Most roads are in a rather heavy red, and are prominent for suburban
areas where buildings are black. In town centres, roads are left blank
and most building cross-hatched in pink. Factories and similar
installations appear in black. A notable feature of the key is the

CARTOGRAPHICAL APPRECIATION 2OI

provision made for showing dykes, lateral ditches, lock-gates, sluices,
and similar structures. Bridges over the minor open drains of the
polderland are recorded in great number.

Marginal information includes the graticule, numbered clearly in
black, and grid numbers effectively shown in light red.

In a comparison of these three maps, the most striking point is the
much greater dependence on fine engraving on the Continental maps
than on the British. Fine engraving is a characteristic of German
cartography, and by means of it a great deal of information is conveyed
with much greater clarity than on the British map, which both
attempts less and is less successful. Whereas the grey tone used on
British maps is adequate for buildings, it is far less satisfactory when
used for woods and field boundaries, which are by no means clear and
tend to be suppressed. On the British map, principal emphasis is
given to relief and to buildings, the map being at its most effective in
areas devoid of settlement or in closely built towns. It is much less
effective in settled rural areas and in areas of mixed urban and rural
settlement. The almost complete absence of information on agri-
cultural land use on the British map is noteworthy, although the more
rapid change of land-use patterns in Britain than on the continent
would diminish the utility of such representation.

The Netherlands map, while obviously constructed from a careful
and detailed survey, suffers from the traditional emphasis of road-
widths in suburbs, where colouring is also heavy. Town centres seem
to be suppressed. The map is perhaps at its most effective in open
polderland and in factory areas. Variation of lettering is used to good
effect, but the range of colour, while superior to that on the O.S.
1/25,000 map, results in a not altogether pleasing combination.

Despite its limited use of colour, the German map is the most
satisfactory of the three. It provides a most detailed and exact repre-
sentation of a wide range of features, and, as comparison of the keys
and of the actual maps makes obvious, has a far greater scope than have
the other two maps. Slight exception may perhaps be taken to the
use of hachures of uniform width (i.e., hachures which do not taper)
for embankments and earthen slopes generally, and to the hatched
railway-line symbols which cause groups of sidings, and marshalling-
yards, to be recorded in congested patches. On the other hand, the
type of symbol used is designed precisely to admit the sub-classification
of railways which is one of the admirable characteristics of this

202 MAP INTERPRETATION

production. Contours are less prominent than on the British 1/25,000
sheet, but urban features are more elaborately treated.

This comparison makes it possible to indicate some of the means
which the various designers have chosen in their efforts to present great
quantities of information. The German map, drawn in great detail
and with great care, demands equal care in reading. Although there
is a great deal of cultural information, represented throughout in black,
fine engraving and the limited use of solid black undoubtedly prevent
the map from looking overloaded. The Netherlands sheet relies
heavily on contrasts in colour. The British map, effectively employing
tapered black hachures and black stippling, and using solid black lines
for railway sidings and spurs, is effective with much of the industrial
landscape. The grey tint for most buildings enables the whole map
to be lightened, but grey is too faint for woodland where light green
wash would greatly improve the map.

NOTES AND REFERENCES

The style of maps is dealt with, either generally or in particular aspects,
in the following works

G. CHEETHAM. "New Medium and Small Scale Maps of the
Ordnance Survey." Geogr.Journ. cvii, 1946, p. 211. (An informative
discussion is printed with this paper.)

A. R. HINKS. Maps and Survey. University Press, Cambridge,
1942. (See Chapters II- VI inclusive.)

ORDNANCE SURVEY. A Description of Ordnance Survey Large-scale
Plans. ,1954; A Description of Ordnance Survey Medium-scale Maps.
1955; A Description of Ordnance Survey Small-scale Maps. 1957.

E. RAISZ. General Cartography. McGraw-Hill, New York, 1948.
(See especially Parts Two, Three, Four, Six, Seven, and Eight.)

UNITED 'NATIONS, DEPARTMENT OF SOCIAL AFFAIRS. Modern Carto-
graphy. Lake Success, 1949. (Provides an up-to-date map of the
status of topographical mapping in different countries.)

J. G. WITHYCOMBE. "Lettering on Maps." Geogr. Journ., Ixxiii,
1929, p. 428. (Followed by a useful discussion.)

INDEX

ABRASION-PLATFORM, 71, 74, 93-4

Accordant summits, 1723

Adjustment of drainage, 15, 25, 36-7,
52-3, 92-3

Administration, 190, 193

Agriculture, 97-9. See also Land use

Alluvial plain, 83-5

Alluvium, 118

Altimetric-frequency curve, 173

Analysis, morphometric, 167-79

Anticlinal valley, 34

Anticline, 24-5, 34, 90-2

Antiquities, 152-63

Appreciation, exemplified, 195 ff.

Archaeology, 153

Arete, 47

Arid-
climate, 83
cycle, 12

Arithmetric analysis, 172-4

Aspect, 126

Asymmetrical fold, 34

Available relief, 174

BACKGROUND knowledge, 3-4
Back-slope, 22, 24
Bar, 72-4

offshore, 77-9
Barrow, 153-5, 158-60
Basal ice, 55
Base-level, 16, 42-4, 169

change, 67. See also Emergence

and Submergence
Bay, 72
Beach, 71-2

material, 72-3
Belgic people, 156
Ben Nevis, 46-7
Bench mark, 188
Biscuit-board topography, 47
Blind valley, 60
Bodmin Moor, 41-2, 168-9

settlement, 121
Boston, 144
Bottom

contours, 79-80

gradient, 77-8
Bottoms, 31

Boulder clay, 55
Boundaries, 190-1
Bourne, 32
Break

of profile, 15, 43. See also Knick-
point

of slope, 85, 169
Bredon Hill, 24
Bridgehead, 138
Bridg water, 143-4
Bristol Avon, 93
Broadway, 117
Bronze Age, 160

antiquities, 154-5
Burford, 141

Burial mounds. See Barrows
Burnham on Sea, 146

GALLOP, 51-2

Camps. See Hilltop sites and Roman

sites

Canals, 99-100
Capture, 28-9, 37
Carboniferous limestone, 56, 59 ff. t

90,92
Cartographical appreciation, 180-96

exemplified, 192 ff.
Casa Grande, 150
Cave, 63-4
Celtic fields, 155, 161
Celts, 155

Chalk country, 3 1-9
Characteristics of coast, 69
Charlestown, 145
Cheltenham, 146-7
China clay, 145
Chipping Campden, 139, 141
Clachan, 128-9
Classification of shorelines, 68
Clay, 23-4
Clay-with-flints, 33
Clevedon, 146-7
Cliff, 71, 75

recession, 70-1
Clinographic curve, 174
Clinometric analysis, 176-7
Clint, 62
Closed contours, 173

203

204

MAP INTERPRETATION

Coal mines, 148

Coast, characteristics of, 69

Coasts, 67-82

Col, 51

Combe, 22, 32

Combwich, 143

Comparison of maps, 199-202

Continental subdivision, 17

Contour interval, 53, 171, 184-5

Contours, 187-8

closed, 173

generalized, 168-9

instrumental, 187

interpolated, 187
Core, 135-6

Cornwall, rural settlement, 120-2
Corrasion, 50, 85
Corrie, 46-7
Corrielake. See Tarn
Corrom, 128
Cots wolds, 22-30

settlement, 117
Crenulate shoreline, 71
Crest, 37, 40
Crestal belt, 47
Criticism of maps, 192 ff.
Croft, crofting, 127-8
Cromlech, 154
Cross-profile, 171
Crossings of Avon, 118
Crustal deformation, 68
Cuesta, 23, 34

in Somerset, 88-90
Cultural features, 189-90

landscape, 4

overlay, 189
Cycle of erosion, n, 14

DARK Ages, 106, 156-7
Datum, 191
Deglaciation, 49
Delta, 49

as settlement site, 127-8
Denudational history, 44
Denuded folds, 34, 89-92
Desert

basin, 86

cycle, 83

landscape, 83-4, 86

weathering, 83-4
Deserts, 83
Devizes, 141

Dip, 22, 33-4, 63
Disappearing streams, 60-1
Discontinuous drainage, 23
Dispersed settlement, 105
Dispersion, 106-7
Dissected plateau, 23
Dissection, 40, 92
Diversion of drainage, 51-2
Divide migration, 28, 172
Doline, 60
Dolmen, 164
Dome, 90-1

"Down-at-heel" erosion, 50, 80
Downcutting, 37-8
Drainage, 14-15

diversion, 51-2

in the Highlands, 52-3

intermittent, 85-6

maladjusted, 36-7

superimposed, 37, 44~5> 93
Drift, 56

Drowned valley, 76
Drumlin, 55-7, 64-5

field, 55-6

swarm, 55
Dry-
gap, 29

valley, 22-3, 31-2, 64, 92
Dry-point

settlement, 117

site, 105
Dune, 73-4, 85, 93

EARLY Iron Age, 155-6, 160-1

Earthwork, 153 ff., 161-2

Eden, 56-7

Edenside, 56-7

Elements of place-names, 108-12

Elongated village, 108

Embayed shoreline, 71

Emergence, 68, 71, 76

Erosion-platform, 37, 41-2, 65, 169,

173

Erosional scarp, 28
Esker, 54

Eustatic movement, 68
Evenlode, 28
Evesham, 27
Evidence

of emergence, 76

of submergence, 76
Evolution of landscape, 38-9

INDEX

205

FACET, 16-17

Factors in rural settlement, 105

Farm, 104-6, 118-19, 120-1

Fault, 14

Features of occupance, 189-91

Fen, 89

Fenland, 78

rural settlement, 122-5

Field-
boundaries, 98
systems, 105-6
work, 1-2

Fifth (Relief) Edition, 187-8

Fiord, 79-81

Flat, 16, 169

Flood plain, 26, 117-18

Fold, 13, 34

Ford, 118

Form

of towns, 135
of villages, 108

Fort William, 145

Fowey, 145

Frost action, 47, 49

Functions of towns, 135

GAVELKIND, 107, 120

Generalized contours, 168-9

Geological boundaries, 33

Geometric analysis, 168-72

Glacial-
cycle, 12
diffluence, 51
transfluence, 51
trough, 47-8, 50-1

Glaciated highland, 46 ff.
rural settlement, 125-9

Glaciation, 12-13, 64-5
of lowland, 54-7

Glacier, 50-1, 53

Graded river, 16

Grampians, 46-58

Granite, 41

Grid, 191

Grike, 62

Groyne, 78

HACHURB, 60, 186-8

Hamlet, 104-6, 120-1

Hanging-
stream, 70, 75
valley, 50

Harepath, herepath, 103

Haugh, 127

Height-frequency curve, 173-4
Hensbarrow, 41-2, 169
Highbridge, 146
Highland coast, 68, 70-7
Hill shading, 186-7, 197
Hilltop sites, 155-6, 160-1
Hinterland, 144
Hogback, 34
Horn, 47
Hypsometric curve, 174

IMPERMEABLE rock, 33
Improved land, 127-8
Incised meander, 16

valley, 37, 42-3
Industrial interpretation, 148
Ingleborough, 60-4
Inlet, 71-3
Inselberg, 83-4
Instrumental contour, 187
Integument* 135
Intercalated dispersion, 106
Intermittent drainage, 85-6
Interpolated contour, 187
Iron Age. See Early Iron Age
Iron ore, 148
Isostatic movement, 79, 81

JOINT-PLANE, 60, 62

KAMB, 64
Karst, 59

country, 90
Karstic

cycle, 12, 59

features, 59, 62
Kettle, 54
Kinlochleven, 145
Knickpoint, 15-16, 43, 65, 171

LAGOON, 78-9
Lake, 56, 79-80

Lake village. See Marsh village
Land-
genesis, 3-4
quality, 3-4
use, 3-4, 119
categories, 98
information, 182
interpretation, 97-9
Landforms of glaciation, 46-58

206

MAP INTERPRETATION

Landscape

characteristics, 11-12

cultural, 4, 189-90

evolution, 38-9

facets, 4

polycyclic, 14-15, 40
Landslip, 75
Large towns, 147
Lateral moraine, 49
Lavant, 32
Lead mine, 161
Lee, 55

Lettering, 136-7, 152, 191-2
Limestone, 23, 59 ff., 90, 92, 149

pavement, 62
Limitations

of maps, 29, 54-5, 65, 76, 81, 116, 171

of scale, 184-6
Linear

earthwork, 156-7, 161-2

village, 108
Lithology, 13
Local base-level, 60
Location of settlement. See Settle-
ment sites
Loch, 50, 79-80
Long barrow, 154, 159-60
Long-profile, 14-16, 171

of glacial trough, 48
Longshore drift, 73
Looe, 145
Lostwithiel, 145
Lowland

glaciation, 54-7

shoreline, 77-9
Lynchet* 155, 157

MALADJUSTED drainage, 36-7, 45
Mammillated topography, 54
Map *

analysis, 2-3, 97

reading, 2

series compared, 199-202
Margins, 189-91
Marine erosion, 13
Market town, 139-43
Marsh village, 161
Massed village, 108
Maturity, 15-16, 40

of shoreline, 72-4, 77-8
Mean slope, 176-7
Meander, 16, 26

bluff, 26

Meander (contd.)

incised, 16

ingrown, 26, 38

spur, 26, 28, 38
Meandering valley, 29, 37-8
Meanders and settlement, 117-18
Medial moraine, 49
Medieval agriculture, 107
Megalith, 154
Melksham, 141
Mendips, 90-2

antiquities, 152-63
Menhir, 154
Merthyr Tydfil, 147-50
Migration of divides, 172
Minor structure, 24-5
Misfit stream. See Underfit stream
Moor, 127-8
Moraine, 49, 54
Moreton in Marsh, 139-40
Morphometric analysis, 167-79
Morphometry, 167

NADDER, 33

Names, 191-3

Negative movement, 68

Neolithic antiquities, 153-4, 159-60

Nodality, 137

and roads, 104

Non-landscape features, 190-3
Normal

cycle, 3, 12

erosion, after deglaciation, 49-50
North point, 194
Northleach, 140-1
Nucleated settlement, 105
Nucleation, 106
Nucleus, 137-8

OBLIQUE illumination, 186-7, 197
Occupance

features of, 189-91

prehistoric, 152-63
Offshore bar, 73, 77-9
Old age. See Senility
Onset, 52
Orchards, 119-20
Outlier, 23, 92
Outport, 145
Outwash, 54
Ownership, 187, 189

PAR, 145
Parallel roads, 53

INDEX

207

Parish boundaries, 107-8
Parrett, 143-4
Pediment, 83, 85-6
Pediplanation, 85-6
Percolation, 31, 60
Periglacial conditions, 25, 32
Permeable rock, 22-3, 60
Physiographic

stairway, 169

subdivision, 16-18

exemplified, 88-94
Pitch, 60

Pitching fold, 34, 90
Place-names, 108-12

elements, no-n

spelling, 192-3
Planation, 37, 42
Planimetry, 173-6
Planned town, 150
Platform

of abrasion, 71, 74, 93-4

of erosion, 37, 41-2, 65, 169, 173
Ploughs, 155-0"
Point elevation, 188-9
Polden Hills, 89
Polje, 63

Polycyclic landscape, 14-15, 40
Ports, 138, 143-5
Positive movement, 68
Post-glacial erosion, 56-7, 70
Preferential growth, 138
Pre-glacial topography, 47
Prehistoric

evidence, Somerset, 1 57 ff.

occupance, I52ff.

remains, table, 158

settlement, 121
Primary

dispersion, 106-7

nucleation, 106
Process, 12-13
Profile, forms of, 170-1
Projected profile, 170
Province, 17
Purpose of maps, 181-2

QUARRY, 56

RAILWAY, 100-2
symbols, 100-1
Railways and settlement, 101-2

Raised beach, 81
Recession
of cliffs, 70-1, 75
of scarps, 23

Recessional moraine, 54
Reclamation, 123-5
Re-grading, 70, 79
Re-growth of towns, 134-5
Rejuvenation, 15-16, 26, 43, 171
Relative relief, 174
Relief
available, 174
relative, 174

representation on maps, 186-9
Representation

of cultural features, 1 89-90

of relief, 186-9
Residual, 41-2
Resistant rock, 40
Resorts, 145-7
Revision of survey, 194
Ria, 72, 76, 145
Ribbon lake, 48
Ridgeway, 103
Riser, 169

River capture, 28-9, 37
Rivers, 14-16
Roads, 102-4
Roche moutonnte, 49
Rock-drawing, 46, 188
Rock-step, 48

Rock-waste, 49-50, 62, 84-5, 127
Roman

antiquities, on Mendips, 161

roads, 102-3, 142

sites, 156

villas, 161

Rosslare Harbour, 145
Round barrows, 154, 160
Routes, 99-104
Run-off, 49
Rural settlement

dominantly dispersed, 120-2

dominandy nucleated, 115-20

forms, 106-7

in Arizona, 129-30

in fenland, 122-5

in glaciated highland, 125-9

pattern, 104-8

place-names, 108-12

sites, 105

studies, 115-33

units, 104-5

208

MAP INTERPRETATION

SALISBURY, 142-3
Salt marsh, 78
Sandstone, 23
Scale, 184-6

of hachuring, 186
Scar, 62
Scarp, 13, 34
Scarp-face, 33

Scarp-foot village, 115, 117
Scarp-former, 22-3
Scarp recession, 23
Scarpland, 22-30
Scree, 49, 62
Sea-level, 191
Sea-loch, 80

Secondary dispersion, 106
Section, 18
Senility, 41
Serial profiles, 169-70
Settlement

and railways, 101-2

prehistoric, 152-63

rural. See Rural settlement

sites, 115-33

studies, 115-51

urban. See Towns
Shipston on Stour, 141
Shoreline, 13, 67-82

compound, 69

cycle, 13, 67-82

initial, 68

neutral, 67

of emergence, 67

of Somerset, 93-4

of submergence, 67
Sidmouth, 146
Sink, 31, 59-61,90
Site, 17-18
Sites *

of rural settlement, 105

of towns, 137-8
Situations of towns, 137
Sketch maps, 18-20
Slip-off slope, 26
Slope, 1 6, 1 8

mean, 176-7

oversteepened, 49
Soil, 99
Soil map, 4
Spa, 146
Special cycles. See Arid, Glacial,

Karstic, and Shoreline
Special maps, 180

Spelling of place-names, 192-3
Spit, 71, 73-4
Spot height, 26, 185
Spring, 31-2, 64
Spring-line, 61-2, 115-16

village, 115
Spur-top profile, 170
Stage, 14-16
Standing stone, 154
Still-stand, 42
Stoss, 52, 55
Stour, 28

crossings, 116
Stow, 17-18

Stow on the Wold, 139-40
Strandline movement, 68, 77
Stream gradient, 172
Strike vale, 23
Structure, 13-14

minor, 24-5

of Cotswolds, 22, 24-5

of Highlands, 52-3

of Salisbury Plain, 3 3-6

of Somerset Plain, 88-94
Style of lettering, 188
Submarine contour, 80
Submature shoreline, 72
Submergence, 68, 71-2, 76
Sub-scarp, 34
Summit

heights, 53-4. i?2-3

plane, 172-3
Superimposed

drainage, 37, 44-5

profiles, 170
Superimposition, 93
Sutherland, 70-1
Symbols

for railways, 100-1

for roads, 102

of cultural features, 189-90

of land use, 98

of non-landscape features, 190-2

on sketch maps, 18-20
Syncline, 24-5, 34

TAFF valley, 147-8
Tamar valley, 45
Tarn, 46-7

Terminal moraine, 48, 54
Terrace, 26-8, 116
Texture of relief, 22
Through-valley, 51

INDEX

209

Tidal-
flat, 89

scour, 72

Topographical scales, 184-6
Tor, 41

Tourist Edition, 188, 197-8
Towns, 134-51

form, 135

function, 135-7

site, 137-8

situation, 137

studies, 138-50
Tract, 1 8

Transect chart, 130-2
Trigonometrical point, 188
Trough's end, 48-9
Truncated spur, 48
Tumuli. See Barrows

UNDERFIT stream, 28-9, 37-8
Underground cavity, 63-4
Ungraded stream, 70
Uniclinal structure, 13
Urban, Urbanization. See Towns
Uvala, 63

VALE of Evesham, 119-20
Vale of Pewsey, 33
Valley

anticlinal, 34

cross-profile, 171

incised, 37, 42-3

long-profile, 171-2

meander, 29, 37-8

sink, see Uvala

young, 41-2

Valley-bottom site, 116-17
Villa, 161
Village, 104-6

form, 1 08

sites, H5-33
Vitrified fort, 156
Volume, calculation from map,

174-3
Volumetric analysis, 174-6

WAINFLEET All Saints, 144

Warwickshire Avon, 22, 26-8

Water-
meadows, 38
supply, 129
table, 31

Waterfall, 65

Wave action, 72
of rejuvenation, 43

Weak rock, 36

Well, 116, 129

Weston-super-Mare, 146

Wet-point

settlement, 115-17
site, 105

Wexford, 144-5

Wiltshire Avon, 33

Wind-gap, 37

Winged headland, 73

Witham, 144

Wylye, 33

YOUNG valley, 41-2
Youth, 40

of shoreline, 71, 77

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Map Interpretation"

See other formats