Using and understanding medical statistics 4ed

Edition Matthews D. download

https://ebookultra.com/download/using-and-understanding-medical-
statistics-4ed-edition-matthews-d/

Explore and download more ebooks or textbooks

at ebookultra.com
 We believe these products will be a great fit for you. Click
the link to download now, or visit ebookultra.com
to discover even more!

Applied Medical Statistics Using SAS 1st Edition Geoff Der

(Author)

https://ebookultra.com/download/applied-medical-statistics-using-
sas-1st-edition-geoff-der-author/

The Cambridge dictionary of statistics 4ed Edition Everitt

B.

https://ebookultra.com/download/the-cambridge-dictionary-of-
statistics-4ed-edition-everitt-b/

Discovering Statistics Using R 1st Edition Andy Field

https://ebookultra.com/download/discovering-statistics-using-r-1st-
edition-andy-field/

Business Statistics using Excel 2nd Edition Glyn Davis

https://ebookultra.com/download/business-statistics-using-excel-2nd-
edition-glyn-davis/
The A3 Workbook Unlock Your Problem Solving Mind Daniel D.
Matthews

https://ebookultra.com/download/the-a3-workbook-unlock-your-problem-
solving-mind-daniel-d-matthews/

Understanding and Using Trusts 3rd Edition Eric P. Hayes

(Editor)

https://ebookultra.com/download/understanding-and-using-trusts-3rd-
edition-eric-p-hayes-editor/

Using R For Introductory Statistics Second Edition John

Verzani

https://ebookultra.com/download/using-r-for-introductory-statistics-
second-edition-john-verzani/

Understanding and Applying Medical Anthropology 3rd

Edition Peter J. Brown

https://ebookultra.com/download/understanding-and-applying-medical-
anthropology-3rd-edition-peter-j-brown/

Medical Statistics at a Glance 2ed Edition Petrie A.

https://ebookultra.com/download/medical-statistics-at-a-glance-2ed-
edition-petrie-a/
Using and understanding medical statistics 4ed Edition
Matthews D. Digital Instant Download
Author(s): Matthews D., Farewell V.
ISBN(s): 9783805581899, 3805581890
Edition: 4ed
File Details: PDF, 2.40 MB
Year: 2007
Language: english
Using and
Understanding
Medical
Statistics
 David E. Matthews ⴢ Vernon T. Farewell

U U U U U U U U U U U U U U U U U U U U U U U U U U U

Using and
Understanding
Medical
Statistics
4th, completely revised and enlarged edition

42 figures and 113 tables, 2007

Basel ⴢ Freiburg ⴢ Paris ⴢ London ⴢ New York ⴢ

Bangalore ⴢ Bangkok ⴢ Singapore ⴢ Tokyo ⴢ Sydney
U U U U U U U U U U U U U U U U U U U U U U U U U U U

David Edward Matthews

BA, MA (Western Ontario), PhD (London);
Professor, Department of Statistics and Actuarial Science,
University of Waterloo, Waterloo, Ontario, Canada

Vernon Todd Farewell

B. Math, M. Math (Waterloo), PhD (London); Senior Scientist
MRC Biostatistics Unit, Cambridge, UK

Library of Congress Cataloging-in-Publication Data

Matthews, David E., 1948-

Using and understanding medical statistics / David E. Matthews Vernon T.
Farewell. – 4th, completely rev. and enl. ed.
p. ; cm.
Includes bibliographical references and index.
ISBN-13: 978–3–8055–8189–9 (hard cover : alk. paper)
ISBN-10: 3–8055–8189–0 (hard cover : alk. paper)
1. Medical statistics. I. Farewell, Vernon T. II. Title.
[DNLM: 1. Biometry–methods. 2. Statistics. WA 950 M438u 2007]
RA409.M39 2007
610.1’5195–dc22 2006100514

Bibliographic Indices. This publication is listed in bibliograpahic services, including Current Contents쏐 and
Index Medicus.
Disclaimer. The statements, options and data contained in this publication are solely those of the individ-
ual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the
book is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness,
quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property
resulting from any ideas, methods, instructions or products referred to in the content or advertisements.
Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accord with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of information
relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for
any change in indications and dosage and for added warnings and precautions. This is particularly important when
the recommended agent is a new and/or infrequently employed drug.
All rights reserved. No part of this publication may be translated into other languages, reproduced or
utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying,
or by any information storage and retrieval system, without permission in writing from the publisher.
© Copyright 2007 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland)
www.karger.com
Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel
ISBN-13: 978–3–8055–8189–9 (hard cover : alk. paper), ISBN-10: 3–8055–8189–0 (hard cover : alk. paper)
To Nancy and Jane
U U U U U U U U U U U U U U U U U U U U U U U U U U U

Contents

Preface to the Fourth Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XI

Preface to the Third Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XV
Preface to the Second Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . XVII
Preface to the First Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .XIX

1 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2. Random Variables, Probability Distributions and Some Standard Notation . . 2
1.3. Characteristics of a Distribution: Mean, Median and Variance . . . . . . . . . . . . . . 8

2 Tests of Significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2. An Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3. Common Features of Significance Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3 Fisher’s Test for 2 × 2 Contingency Tables . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2. Details of the Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3. Additional Examples of Fisher’s Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4 Approximate Significance Tests for Contingency Tables . . . . . . . . . . 28

4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.2. The ␹2 Test for 2 × 2 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.3. The ␹2 Test for Rectangular Contingency Tables . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.4. Using Statistical Tables of the ␹2 Probability Distribution . . . . . . . . . . . . . . . . . . 42

5 Some Warnings Concerning 2 × 2 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . 46

5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.2. Combining 2 × 2 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.3. Matched Pairs Binary Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6 Kaplan-Meier or ‘Actuarial’ Survival Curves . . . . . . . . . . . . . . . . . . . . . . . 54
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.2. General Features of the Kaplan-Meier Estimate . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
6.3. Computing the Kaplan-Meier Estimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6.4. A Novel Use of the Kaplan-Meier Estimator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

7 The Log-Rank or Mantel-Haenszel Test for the Comparison of

Survival Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
7.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
7.2. Details of the Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
7.3. Evaluating the Log-Rank Test – A Simple Example . . . . . . . . . . . . . . . . . . . . . . . . 71
7.4. More Realistic Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

8 An Introduction to the Normal Distribution . . . . . . . . . . . . . . . . . . . . . . . 76

8.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
8.2. Basic Features of the Normal Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
8.3. The Normal Distribution and Significance Testing . . . . . . . . . . . . . . . . . . . . . . . . 81
8.4. The Normal Distribution and Confidence Intervals . . . . . . . . . . . . . . . . . . . . . . . 84
8.5. Using Normal Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

9 Analyzing Normally Distributed Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

9.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
9.2. Some Preliminary Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
9.3. Analyzing a Single Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
9.4. Comparisons Based on the Normal Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 100
9.5. Testing the Equality of Variances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

10 Linear Regression Models for Medical Data . . . . . . . . . . . . . . . . . . . . . . . 111

10.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
10.2. A Historical Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
10.3. Multiple Linear Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
10.4. Correlation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
10.5. The Analysis of Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

11 Binary Logistic Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128

11.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
11.2. Logistic Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
11.3. Estimation in 2 × 2 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
11.4. Reanalysis of a Previous Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
11.5. The Analysis of Dose-Response Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136

12 Regression Models for Count Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

12.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
12.2. The Model for Poisson Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
12.3. An Experimental Study of Cellular Differentiation . . . . . . . . . . . . . . . . . . . . . . . . 145

Contents VII
13 Proportional Hazards Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
13.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
13.2. A Statistical Model for the Death Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
13.3. The Lymphoma Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
13.4. The Use of Time-Dependent Covariates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

14 The Analysis of Longitudinal Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

14.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
14.2. Liang-Zeger Regression Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
14.3. Random Effects Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
14.4. Multi-State Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

15 Analysis of Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174

15.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
15.2. Representing Categorical Information in Regression Models . . . . . . . . . . . . . . . 175
15.3. Understanding Two-Factor Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
15.4. Revisiting the INR Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

16 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191

16.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
16.2. Quality Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
16.3. Initial or Exploratory Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
16.4. Primary Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
16.5. Secondary Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
16.6. A Formal Discussion of Multiple Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
16.7. False Discovery Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

17 The Question of Sample Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

17.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
17.2. General Aspects of Sample Size Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
17.3. Two Examples of Sample Size Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
17.4. Some Hazards of Small Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

18 The Design of Clinical Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

18.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
18.2. General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
18.3. Trial Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
18.4. Randomized versus Historical Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
18.5. Intention to Treat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
18.6. Factorial Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
18.7. Repeated Significance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
18.8. Sequential Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

19 Further Comments Regarding Clinical Trials . . . . . . . . . . . . . . . . . . . . . . 234

19.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
19.2. Surrogate Endpoints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234

Contents VIII
19.3. Active Control or Equivalence Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
19.4. Other Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
19.5. Multiple Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
19.6. Multiple Treatment Arms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
19.7. Sequential Designs for Efficacy-Toxicity Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
19.8. Stochastic Curtailment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250

20 Meta-Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
20.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
20.2. Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
20.3. Study Heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
20.4. An Illustrative Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
20.5. Graphical Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
20.6. Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

21 Epidemiological Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263

21.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
21.2. Epidemiological Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
21.3. Relative Risk Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
21.4. Odds Ratio Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
21.5. Confounding and Effect Modification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
21.6. Mantel-Haenszel Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
21.7. Poisson Regression Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
21.8. Clinical Epidemiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

22 Diagnostic Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

22.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
22.2. Some General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
22.3. Sensitivity, Specificity, and Post-Test Probabilities . . . . . . . . . . . . . . . . . . . . . . . . . 286
22.4. Likelihood Ratios and Related Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292

23 Agreement and Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298

23.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
23.2. Intraclass Correlation Coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
23.3. Assessing Agreement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
23.4. The Kappa Coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
23.5. Weighted Kappa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
23.6. Measures of Agreement for Discrete Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
23.7. The Dependence of κ on Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314

Contents IX
U U U U U U U U U U U U U U U U U U U U U U U U U U U

Preface to the Fourth Edition

Twenty-five years have elapsed since we first began drafting our thoughts
about using and understanding the statistical methods commonly employed
in medical research. In the meantime, AIDS, SARS and avian flu, to name only
three public health concerns, have burst onto the world stage. So, too, have
fears of terrorism, the consequences of global warming and the next pandem-
ic. Communication has been transformed, not least through the development
of the World Wide Web. Twenty-five years ago, it would have required remark-
able prescience to foretell, even in the vaguest of terms, how circumstances and
events would unfold.
Even to predict the changes in attitudes that have occurred concerning
smoking in public places, and the dangers of unprotected, excessive exposure
to sunlight might have seemed highly speculative when Using and Understand-
ing Medical Statistics first appeared in print.From the perspective of our goals
in conceiving the first edition, the fairly widespread adoption of the phrase
‘evidence-based medicine’ is particularly noteworthy. It implicitly reflects the
increased understanding, by physicians, that a basic grasp of statistical con-
cepts and a passing appreciation for what statistical analysis can and cannot
do is essential if one wants to be able to read and evaluate the medical litera-
ture.
When the first edition appeared, we hoped to make a small contribution
to such effective use of the medical literature. Never in our wildest dreams did
we imagine that four editions would be published in English, as well as foreign
language editions in Italian, Spanish and in Japanese. And when we first began
writing, we certainly did not anticipate the topics that would become com-
monplace in the medical research literature, and which we might therefore feel
were appropriate to try and explain to our readers.

XI
 How does this edition differ from its predecessor? In our opinion, the
fourth edition represents the most substantial revision of Using and Under-
standing Medical Statistics since the book was first published in 1984. As med-
ical research has evolved, the statistical methods that are used to sift study data
and adduce convincing evidence of improvement and innovation have become
increasingly sophisticated. As a result, we have added entirely new chapters on
Poisson regression, the analysis of variance, meta-analysis, diagnostic tests
and the subject of measurement agreement and reliability. In addition, there
are sections describing new topics in the chapters on longitudinal studies, data
analysis, and clinical trials. Because statistical software is now widely avail-
able, we have removed the nine pages of statistical tables pertaining to Fisher’s
exact test; there are now many computational tools that will evaluate the exact
significance level of this widely-used hypothesis test. Since there is now a chap-
ter describing Poisson regression, we have also been able to add a new section
to the chapter on epidemiological applications, one that describes the use of
this tool to analyze the classic cohort study of smoking and coronary mortal-
ity reported by Doll and Hill. The changes in the public attitude towards smok-
ing to which we previously referred are in large measure due to the pioneering
efforts of Drs. Doll and Hill, and their work provides an outstanding example
of fruitful collaboration between medical and statistical scientists. Finally, we
must admit that the goal we identified in the first edition to have most chapters
represent a single evening’s reading has proved increasingly difficult to
achieve.
First drafts of most of this new material were developed last year while
DEM was an antipodean sabbaticant in the Centre for Clinical Epidemiology
and Biostatistics at the University of Newcastle, and VTF made use of the ex-
cellent research facilities of his employer, the MRC Biostatistics Unit in Cam-
bridge, England. We want to thank the Director and staff of the Centre – and
particularly Professor Robert W. Gibberd – for generously providing DEM
with a quiet office, as well as access to library and computing facilities. We also
thank Professor Simon G. Thompson, the Director of the Biostatistics Unit,
and the Unit staff for their support of VTF’s efforts. Although the authors were
half a world apart, the tangible encouragement of colleagues made writing and
collaboration on a first draft of the fourth edition relatively easy.
Thanks are also due to our publisher, S. Karger AG in Basel, and especial-
ly to Rolf Steinebrunner, in production management, and Ms. Deborah Lau-
tenschlager with whom we worked on this edition. We are particularly grateful
to Rolf who has overseen the publication of all four editions of the book, and
plans to retire at the end of 2006. It has been our privilege to enjoy such a long
and fruitful relationship with Rolf, and everyone at Karger who has partici-
pated in publishing Using and Understanding Medical Statistics. We hope

Preface to the Fourth Edition XII

Rolf’s retirement years will be as fulfilling, and enjoyable, as his working ca-
reer at Karger seems to us to have been.
For both of us, this year, 2006, marks significant anniversaries of our wed-
dings. In recognition of those pivotal events in our personal lives, once again
we dedicate whatever our efforts may have achieved to the two special indi-
viduals whose love and support have sustained and enriched our lives for more
than half a century of unforgettable, shared experience.
D.E. Matthews
V.T. Farewell

Preface to the Fourth Edition XIII

U U U U U U U U U U U U U U U U U U U U U U U U U U U

Preface to the Third Edition

The world today is a very different place from what it was 12 years ago.
Setting aside all the political changes that have occurred since Using and Un-
derstanding Medical Statistics was first published, there have been extraordi-
nary developments in computing technology that have affected both the prac-
tice of medicine and the statistical methods that researchers use to evaluate
medical progress. We will leave it to you, the readers of this book, to reflect on
how medical practice has changed. From the statistical perspective, consider
that when the first edition was published, access to statistical packages that
would fit relative risk regression models to survival data was somewhat lim-
ited. More often than not, use of this software would have required the assis-
tance of a statistician familiar with the input-output quirks and other vagaries
of the routines. Now, the same models are a standard feature of many com-
mercial software packages. Data input is often accomplished using spreadsheet
editors, and most of the key aspects of model selection and fitting can be car-
ried out by simply clicking the mouse on the correct item in a pull-down menu.
Seconds later, the results of the analysis scroll across the screen. Perhaps the
most astonishing aspect of this revolution in statistical analysis is the fact that
one can carry out such analyses virtually anywhere – sitting in an airplane,
riding on a train, or logged in from an office on one continent to a remote ma-
chine halfway round the globe.
The ease with which statistical analyses can now be carried out makes the
focus of this book all the more important. Readers of the medical literature in
former times may well have thought that a statistically significant result was,
perhaps, the primary determinant in the editorial process of selection and
publication. However, it has not been, and is not, sufficient simply to carry out
some statistical analysis of medical data with this goal in mind.

XV
 Now that complex statistical analyses are easy to execute, it is particularly
important that the focus for medical researchers shifts from computation to
interpretation and understanding. Readers of the medical journals, now more
than ever, need the ability to read and critically appraise the results of various
studies that deal with issues in the practice of medicine such as new treat-
ments for a specific disease, the natural history of conditions such as AIDS, or
the public health benefit of a new screening program for breast or prostate
cancer.
What is different about the third edition of Using and Understanding Med-
ical Statistics? First, there are two new chapters. One of these provides readers
with an introduction to the analysis of longitudinal data. We describe two
quite different approaches to the analysis of such studies; both methods are
beginning to find their way into the mainstream of medical literature. The
second new chapter augments material concerning the design of clinical trials
that appeared in the first and second editions. Readers are introduced to topics
such as the use of surrogate markers, multiple outcomes, equivalence trials,
and the design of efficacy-toxicity studies.
In addition to these new chapters, we have reorganized the last third of the
book so that the actual order in which topics are introduced precedes their
routine use. In this respect the second edition contained one glaring pedagog-
ical error that we are pleased to have the opportunity to rectify. As well, we
have taken great pains to carefully re-phrase sentences and paragraphs that did
not stand up to scrutiny. We were greatly assisted in this exercise by a Danish
medical statistician, Dr. Jørgen Hilden, who sent us several pages of construc-
tive remarks that revealed deficiencies in the second edition. We, and our read-
ers, owe Dr. Hilden a substantial debt of gratitude for this generous exercise of
his experience, insight, and labour.
We conclude with one final observation concerning the impact of techno-
logical change on normal patterns of work. When the first edition was being
prepared, Mrs. Joy Hoggarth typed the manuscript for us, and did a superb job.
The second edition was typeset by Ms Lynda Clarke, whose productivity at the
keyboard was astonishing. Now both of us write directly onto computers or X-
terminals that occupy a corner of the desk, so we have no one to thank for ca-
pably assisting us in the preparation of the manuscript for the third edition.
On the other hand, we are pleased to say that no technological change prevents
us from dedicating this third edition to the same very special people to whom
the first and second editions were also dedicated.
D.E. Matthews
V.T. Farewell

Preface to the Third Edition XVI

U U U U U U U U U U U U U U U U U U U U U U U U U U U

Preface to the Second Edition

Slightly less than four years have elapsed since the preface to the first edi-
tion was written. In the meantime, we have been surprised, and pleased, by the
response to the first edition. The letters and comments which we have received
from readers and reviewers on several continents have brought us much satis-
faction. Suggestions and criticisms have helped us to understand specific top-
ics where the first edition failed to meet the goals which we had established.
Despite our best intentions, there were inevitable errors in the first edition
which we were anxious to correct. Consequently, when the publisher inquired
about the possibility of a revised edition, we realized that it would be an op-
portunity to rectify both kinds of flaws simultaneously.
How do the two editions differ? Apart from minor corrections to Table 3.4
and the elimination of errors which appear to be randomly distributed through
the chapters, the principal differences may be found in the second half of the
book. The example in chapter 10 has been changed to one which we believe
suits better the purpose we intend to achieve. Sections have been added to
chapters 11, 12 and 14 which treat topics that were previously omitted. In some
ways, these additions reflect the changing face of medical statistics, and the
clinical investigations in which statistical methods play an important role.
However, the major difference between the editions is the addition of chapter
16, which concerns epidemiological studies. The topics treated in the final
chapter illustrate just how much the use of sophisticated statistical analysis has
permeated the recent practice of epidemiology. At the same time, this new
chapter knits together the fabric of the book, drawing on methods which we
have introduced in previous chapters to analyze data from various epidemio-
logical studies. In that respect, chapter 16 does what no chapter in the first edi-
tion was able to do. We hope its inclusion in the second edition will help all

XVII
Random documents with unrelated
content Scribd suggests to you:
with great spirit. The Austrians amounted to about forty thousand,
including a fine body of cavalry, for which the ground was highly
favourable; the French were not more than thirty thousand, posted
strongly in and around the village of Marengo, in three divisions,
each stationed about a quarter of a mile behind the other. After two
or three attempts the Austrians drove the French out of the village
of Marengo, threw the second division, commanded by Lannes, into
confusion, and put to rout the left wing of Buonaparte's own
division, threw his centre into disorder, and compelled him to retreat
as far as St. Juliano. The whole tide of battle was running against
Buonaparte, and a short time must have completed his rout, when
the strength of the old general, Mélas—more than eighty years of
age—gave way, for he had been many hours on horseback. He
retired from the field quite secure of the victory, and left General
Zach to finish it. But, at this moment, General Desaix, who had lately
arrived from Egypt, and had been sent by Buonaparte to make a
diversion at Rivolta, came back with his detachment of twenty
thousand men. Kellermann, also, who was posted in the rear with a
body of reserve, marched up at the same time. A new and desperate
charge was made on the fatigued Austrians, and they were broken
and put to the rout. They retreated across the Bormida, towards
Alessandria, in a panic, the horse galloping over the infantry. Mélas,
dispirited by his defeat, but more by his age, gave up the struggle
and on the 16th of June concluded an armistice, resigning not only
Alessandria, where he might have stood a longer siege, but Genoa,
which had just surrendered to the Austrians, and all the Genoese
territory, agreeing to retire behind the line of Mantua and the Mincio,
and leaving to the French all Lombardy as far as the Oglio. The
French themselves could scarcely believe the reality of such a
surrender.
 GENOA.
[See larger version]
During this brilliant campaign in Italy, Moreau, in Germany, had
beaten General Kray in several engagements, advanced to Ulm, and
there, crossing the Danube, had overrun a great part of Bavaria, and
had made himself master of Munich and menaced Vienna. On
hearing of the armistice in Italy, the Emperor demanded one for
Austria, to continue till September; and Buonaparte, seeing that the
Czar Paul had ceased to support Austria, recommended the Emperor
to make peace with France. The Emperor required that Britain
should be included in it. But Napoleon demanded a separate
negotiation, which Austria was afraid to grant. No sooner was this
answer received in Paris than Buonaparte gave the word for
renewed and vigorous action, both in Italy and Germany. Moreau
advanced by Salzburg towards Vienna, whilst Brune drove the
Austrians from the Mincio, and over the Adige and the Brenta to the
very vicinity of Venice, whilst Macdonald occupied the passes of the
Tyrol, ready to march to the support of the army either in Italy or
Germany. The Archduke John met Moreau near Haag, and for a
moment worsted him; but on the 2nd of December the two armies
came to a general engagement at Hohenlinden, between the rivers
Iser and Inn, in which the Austrians were routed, with a loss of ten
thousand men. Moreau advanced and occupied Salzburg, and
trembling for the safety of Vienna itself, the Emperor hastened to
make peace. An armistice was signed on the 25th of December, and
the treaty was concluded at Lunéville on the 9th of February, 1801.
By this treaty all the conditions of the Treaty of Campo Formio were
renewed, and the frontier of the Rhine was again ceded to France.
In Britain there were terrible outcries in consequence of the scarcity
of bread. There were rioting and plundering of corn-factors' and
bakers' shops, and Government passed a number of Acts giving
premiums on the importation of grain, and forbidding the making of
any but mixed and coarse breads. Had not large subscriptions been
raised, and private benevolence been called forth to an immense
extent for the relief of the distress, the consequences would have
been more terrible. Pitt was in favour of remedial legislation, but
Grenville was against interfering with the laws of supply and
demand.
On the 22nd of January, 1801, the first Imperial Parliament met, and
Addington was re-elected Speaker. The king did not meet this
Parliament till the whole of its members had been sworn; his
opening of it for business took place on the 2nd of February, and his
speech had no cheering topics to give spirit to its first proceedings;
on the Continent there had been nothing but defeat on the part of
the Allies, of triumph on that of France. Our late ally, Paul, had not
only seized our merchant vessels in the ports of the Baltic, and the
property of our merchants in the Russian towns, but he had entered
into a league with Sweden and Denmark to close the Baltic
altogether to us, and to compel us to relinquish the right of search.
This confederacy, by stopping the supplies of corn from the North,
threatened us with great aggravation of the distresses at home; and
some members advocated the surrender of the right of search, or
the acceptance of the principles of an armed neutrality, such as
Catherine of Russia had endeavoured to establish. But Pitt plainly
showed that to allow neutral vessels to carry arms, ammunition, and
commodities of life into the ports of our enemies would render all
blockades of their forts useless, and enormously increase our
difficulties during war. Orders were immediately issued to send a
powerful fleet into the Baltic to chastise the insane Czar.
But there was another topic started in this first Imperial Parliament
which was as odious to George III. as the perfidious conduct of his
late Russian ally. As one means of bringing about the union with
Ireland, Pitt held out to the Irish Catholics the argument that by
having Irishmen in the united Parliament they would be most likely
to obtain a repeal of the Catholic disabilities. Both he and Lord
Cornwallis had sent circulars to this effect, anonymous, it is true, but
with a secret avowal of their authorship, amongst the leading
Catholics, which had a great effect in procuring their assent to the
Union. Lord Castlereagh, who as Secretary of State for Ireland had
helped to carry the Union, claimed the redemption of this pledge.
The matter was talked over in the Cabinet during the autumn of
1799, and again in September, 1800. Pitt introduced the subject
about the middle of January in the Privy Council. But in the interval
the Chancellor, Lord Loughborough, had betrayed the plan to the
king, and in conjunction with Lord Auckland had convinced his
Majesty that it would involve a violation of the Coronation Oath.
George was indignant, and almost furious. At the levee on the 28th
of January, when Lord Castlereagh was presented, he said to
Dundas, "What is this which this young lord [Castlereagh] has
brought over to fling at my head?" He alluded to a plan for Catholic
emancipation, and added, "I shall reckon every man my personal
enemy who proposes any such measure! This is the most jacobinical
thing I ever heard of." Dundas replied that his Majesty would find
amongst those friendly to the measure some whom he had never
supposed to be his enemies. On the 31st of January Pitt wrote to the
king, assuring him that the union with Ireland would render it
absolutely necessary that important questions regarding the
Catholics and Dissenters should be discussed; but, as he found how
extremely such topics were disliked by his Majesty, and yet how just
it was that Catholics should be admitted to Parliament as well as
Protestant Dissenters, who were already admitted, he begged to be
permitted to resign. At the same time, not to inconvenience his
Majesty, he was willing to hold office till his Majesty had
reconstructed a Cabinet wholly to his mind. George replied, the very
next day, that Mr. Pitt's letter had occasioned him the liveliest
concern; that, so far from exposing him to the agitation of this
question, he had flattered himself that the Union, by uniting the
Protestants of both kingdoms, would for ever have excluded the
question of Catholic emancipation. He expressed his ardent wish that
Pitt should continue to be his Minister as long as he lived; and he
only required, as a condition, that he should stave off this question.
Pitt replied, on the 3rd of February, that his Majesty's determined
tone on the subject of Catholic emancipation left him no alternative
but to resign, in compliance with his duty; and that, as his Majesty's
resolve was taken, it would certainly be best for the country that his
retirement should be as early as possible. On the 5th the king wrote,
accepting Pitt's resignation, though with expressions of deep regret.
Five days after this, February 10th, the matter was made public by
Lord Darnley rising in the Upper House, and moving for an inquiry
into the conduct of the Ministry. This roused up Lord Grenville, who
candidly avowed that, in consequence of their failure to introduce
the question of Catholic emancipation, the Ministers had resigned
and only held office till a new Cabinet was formed. On this, Lord
Darnley postponed his motion. On the same day, in the Commons, a
letter from Addington, the Speaker, was read, announcing his
resignation of the Speakership in consequence of the king's proposal
to nominate him to a situation incompatible with that post. Pitt then
rose and confirmed this, and proposed an adjournment till the next
day in order to prepare for the nomination of the new Speaker. The
House adjourned accordingly, and next day, the 11th of February,
elected Sir John Mitford, the Attorney-General, as Speaker. Before
the House could resume business, it was announced that the king
was ill—confined to the house by a severe cold; but it was soon
known that it was a return of his old malady, lunacy, in consequence
of his extreme agitation on the proposal of the Catholic question and
the resignation of Pitt. The report was soon augmented into the
startling rumour that the king was dangerously ill, and that a
regency must take place—if not superseded by his death. At this
news Fox, who had for some time absented himself from Parliament,
on the plea that all endeavours to carry sound and prudent
measures were hopeless with Pitt's great martial majority, hastened
up to town from St. Anne's Hill; and the Whig body was in a flutter
of expectation that he would soon be the Minister of the prince
regent, or of George IV. But all these hopes were speedily
overthrown by the news of the rapid improvement of the king, and
on the 12th of March the royal physicians announced him perfectly
recovered. He attributed his illness to Pitt's conduct, and the ex-
Premier thereupon wrote and promised never to re-open the
question again.
The new Ministry consisted of Addington, son of Chatham's old
physician, Dr. Addington, as First Lord of the Treasury and
Chancellor of the Exchequer: the Duke of Portland, President of the
Council; Lord Eldon, Chancellor; Earl St. Vincent, First Lord of the
Admiralty; the Earl of Chatham, Master-General of the Ordnance;
Lord Pelham, Secretary of the Home Department; Lord Hawkesbury,
the eldest son of the Earl of Liverpool, Secretary for Foreign Affairs;
Lord Hobart, Secretary for the Colonies. Several of Pitt's Ministers
remained, but the important members, Grenville, Dundas, Woodham
and Spencer retired with him. It was soon seen, however, that
though Pitt was out of office his principles dominated in it, and that
there was no chance of a change of system. The Cabinet was one of
mediocrities, and was probably regarded by Pitt as a convenient
makeshift until he could return to power.
 NELSON AT THE BATTLE OF COPENHAGEN. (See p. 481.)
[See larger version]
But long before this—as early, indeed, as the 15th of April—news
had reached London of the death of the erratic Emperor Paul, and of
the bombardment of Copenhagen by the British fleet. Paul had been
won over by Buonaparte to his views, and had been flattered by him
by being elected—though irregularly and illegally—Grand-Master of
the Knights of Malta. He had been persuaded that the conquest of
Malta by the British was an invasion of his rights, and by these and
other flatteries Buonaparte had influenced his weak mind to become
the agent of his plans in destroying the British ships in the Baltic,
and in closing that sea to British commerce. Paul pretended that we
had captured Danish convoys, these same convoys being engaged in
guarding vessels loaded with materials of war for France, and that
thus the independence of the North was menaced by us. On this
ground, and on that of the invasion of Malta, he immediately laid an
embargo on all British vessels in Russian ports, and as two vessels in
the harbour of Narva resisted the attempts to seize them, in
consequence of the embargo, he ordered all the British vessels in
that port to be burned. In consequence of this sudden and
unwarrantable order, contrary to all the laws of nations, about three
hundred British vessels were seized, and the officers and crews
dragged on shore, put into irons, and sent up the country under
menaces of Siberia. Paul next ordered all property of Englishmen in
Russia to be seized and sold. Denmark—with whom we had various
rencontres, on account of its men-of-war convoying vessels laden
with stores for French ports—soon joined Russia. We sent Lord
Whitworth to Copenhagen to endeavour to come to some
understanding on these matters in 1800, but though a convention
was signed, it was not satisfactory. Sweden followed the example of
Denmark, and the three Northern Powers entered into a treaty of
armed neutrality to resist our search of their vessels in any
circumstances. As the consequence of this policy would be to shut
us out of all trade with the ports of the Baltic, it was resolved to
send a fleet to chastise these Powers and break up their co-
operation with France. Mr. Vansittart was despatched to
Copenhagen, accompanied by a fleet of eighteen sail of the line,
with several frigates and smaller vessels, under command of Admiral
Sir Hyde Parker, with Vice-Admiral Nelson as second. The fleet left
the Yarmouth Roads on the 12th of March, 1801, and arriving at the
mouth of the Sound, Nelson recommended that they should sail
directly up to Copenhagen, and be prepared, on the refusal of our
proposals, to bombard the place, as this would not allow them time
to get ready their batteries, and thus do all the more damage to our
ships and men. But this was deemed too offensive before any
attempt at negotiation, and accordingly Mr. Vansittart was sent
forward in a frigate with a flag of truce, leaving the fleet at the
Scaw. He returned without effecting anything more than what
Nelson anticipated. Sir Hyde Parker wasted time in making the
needless inquiry by a flag of truce of the Governor of Elsinore,
whether the passage of the Sound would be disputed, who replied
that it would. It was then proposed to enter by the Belt. Nelson said:
—"Let it be by the Sound, or the Belt, or anyhow—only don't let us
lose an hour."

COPENHAGEN.
[See larger version]
On the 30th of March the British cast anchor before Copenhagen,
between it and the island of Huen. On reconnoitring, the defences of
the place were found to be very formidable. Nelson was appointed
to make the attack with twelve line-of-battle ships, and some smaller
craft. He had asked for ten. The next morning—the 2nd of April—the
wind was favourable, and Nelson weighed and drew nearer to the
town—Sir Hyde Parker on the outside threatening the batteries and
vessels at the mouth of the harbour. At ten o'clock the firing
commenced, and at eleven it was general. Three of the British
vessels—the Agamemnon, the Bellona, and the Russell—stuck fast
on the shoal. For three hours the battle raged fiercely, for the Danes
fought with their well-known valour. It was necessary for Nelson to
silence or destroy the floating batteries and gunboats before he
could come at the ships of the line and the great land batteries. He
had ordered five hundred seamen, under the Hon. Colonel Stuart
and Captain Freemantle to storm the Kroner Battery as soon as it
was silenced; but at this moment Sir Hyde Parker, seeing the signals
of distress flying at the mast-heads of the three vessels aground,
and that three others, which he had sent forward as a
reinforcement, were making but slow way to the front, signalled for
the fleet to draw off, and cease the engagement. But Nelson took no
notice of the signal: he continued to walk the deck, and asked if his
signal for close action was still hoisted, and, being told it was, said:
—"Mind you keep it so." About half-past one o'clock the fire of the
Danes slackened, and by two it had nearly ceased. But the vessels
that had struck their flags recommenced firing on our boats sent to
take possession of them, and the fire of the batteries on land and on
Amager Island struck these surrendered vessels on one side, and
that of our ships on the other. To prevent the destruction of the
unhappy Danes placed in this fatal situation, Nelson sent on shore
Sir Frederick Thesiger with a flag of truce, and a letter to the Crown
Prince, entreating him to put an end to a contest that was uselessly
wasting the lives of the brave Danes. Within half an hour after
Thesiger's departure, the firing from the Kroner Battery ceased, and
Adjutant-General Lindholm came on board to learn the precise
object of Nelson's note. Nelson replied that his object was humanity.
He demanded that the action should cease, and that the wounded
Danes should be taken on shore; that then he would burn or carry
away the surrendered vessels, as he should think fit. It was agreed
that the combat should cease for twenty-four hours, during which
negotiations should be entered into. After five days' arduous
discussion, an armistice was concluded for fourteen weeks, during
which the treaty of armed neutrality with Russia was to be
suspended. Nelson was to have full liberty to purchase any
necessaries for his fleet, in Copenhagen or along the coast, and in
case of renewal of hostilities all the Danish prisoners were to be
again surrendered.
The ships being got afloat again, on the 12th of April Parker sailed
away with the main body of the fleet, leaving the St. George (with
Nelson) and a few other ships to repair their damages. Sir Hyde
Parker went in quest of the Swedish fleet, which consisted only of six
ships, and which had taken refuge behind the forts of Karlskrona.
Parker sent in a flag of truce, informing them of the armistice with
Denmark, and demanding an answer as to the intentions of Sweden.
Gustavus, the King of Sweden, hastened to Karlskrona, and on the
22nd informed the English admiral that he was ready to treat with
an envoy accredited to the Northern Powers. Admiral Parker then
proceeded towards the Gulf of Finland to attack the Russian fleet,
but was soon overtaken by a dispatch boat from the Russian
ambassador at Copenhagen, announcing that the Emperor Paul had
been murdered by his courtiers, and that his son, Alexander, had
accepted the proposals of Britain to treat. Parker considered the
news of Paul's death as tantamount to the conclusion of peace, and
proposed sailing down the Baltic again; but Nelson, who had joined
him at Karlskrona, thought very differently. He had blamed Parker's
slowness and easiness all through the affair of Copenhagen, and he
now wanted to push on to Revel, and destroy the Russian fleet
before the ice allowed it to retire into Cronstadt. Sir Hyde Parker
refused; and the fleet was on its way down the Baltic when an order
came recalling Parker and giving the command to Nelson. He
immediately put about and proceeded to Revel, but the thaw had
allowed the Russian fleet to get into Cronstadt. Nelson, however,
opened communications with the Emperor Alexander, and proposed
to land and terminate a convention with him at once. Alexander, not
liking to have Nelson's fleet too near, declined the proposal in terms
of courtesy, and Nelson took his leave in no complimentary mood.
The emperor thought it best to send after him Admiral Tchitchagoff,
to assure him that Alexander regretted that any misunderstanding
had ever taken place between Russia and Britain; that all the British
subjects seized by Paul should be immediately liberated, all their
property restored, and that the Czar would be glad to see Nelson at
St. Petersburg in any style which he liked to assume. But Nelson had
now resolved to return at once to Britain, his shattered health ill
bearing the severity of the northern climate; nor was his presence
necessary, for on the 17th of June, two days before Nelson went on
board the brig which took him to Britain, Lord St. Helens, who had
proceeded to St. Petersburg as ambassador, had signed a
convention, by which all subjects of dispute between the two
countries were ended. Denmark and Sweden came into the
convention as a matter of course.
General Kleber, whom Buonaparte had left in command of the
Egyptian army, was an excellent officer, and he had improved the
condition of the forces there. Instead of the French army in Egypt
being weaker than when Buonaparte left it, it was much stronger. In
1800 Kleber was attacked at the fort of El Arish, in the Desert, by a
strong Turkish force, supported by the British squadron under Sir
Sidney Smith. Being defeated, he agreed to a convention, by which
he promised to evacuate Egypt, on condition of his army being
allowed to return unmolested to Europe; but no sooner were these
terms communicated to the British Government than they disavowed
them, declaring that Sir Sidney had no authority to propose them.
Kleber, therefore, resumed hostilities and returned towards Cairo;
but being attacked by the Turks, he fought and routed them with
great slaughter, on the 20th of March, 1800, near the ruins of the
ancient city of Heliopolis. The Moslems of Cairo, encouraged by
Murad Bey, who still hovered about with his Mameluke cavalry, rose
on the French there, and massacred such as could not escape into
the citadel. Kleber hastened to Cairo, relieved the forces in the
citadel, and entered into a truce with Murad Bey, but whilst thus
busily engaged he was assassinated by an Arab, who declared he
was commissioned by Allah to free the country of the infidels. The
command was taken by Menou, whose administration of the army
and general affairs was far inferior to that of Kleber. At the time that
matters were changing thus for the worse, amongst the French,
Dundas, now Lord Melville, urged upon Ministers the good policy of
sending an army to Egypt and compelling the surrender of the
French. He contended that, whilst one army was sent from Britain,
another should be brought across the Persian Gulf from India, and
success made certain. The plan was much too bold, even for Pitt;
and the king opposed it energetically, as "a dangerous expedition
against a distant province." But the danger of having this French
army transferred to Europe at some critical moment—as it would
have been had the Convention of El Arish been carried out, by which
these twenty thousand seasoned men could have been landed in
Italy to act against Suvaroff—at length brought the British Ministry to
dare the attempt.
On the 8th of March, 1801, General Sir Ralph Abercromby landed in
Egypt, where Nelson had fought the battle of Aboukir. Menou
brought down against the British twelve or fourteen thousand men,
including a fine body of cavalry. Sir Ralph Abercromby landed only
about ten thousand in effective order, but these were men full of
ardour and disciplined to perfection. On the 8th of March they
landed in face of the French, five thousand being put on shore at
once, these returning no single shot whilst in the boats, though
assailed by fifteen pieces of artillery from the opposite hill, and by
grape-shot from Aboukir Castle. They were led on by General
(afterwards Sir John) Moore; and running, or climbing on hands and
knees, up the steep sand-hills, they drove the French from their
cannon, and seized them. The French retreated, and posted
themselves on some heights between Aboukir and Alexandria. On
the 19th, having compelled Fort Aboukir to surrender, General
Abercromby advanced, and found Menou had concentrated all his
forces between them and Alexandria. On the 21st of March a general
engagement took place. It commenced as early as three o'clock in
the morning, whilst quite dark, by an attack on the British left, which
was meant to draw all attention to that quarter, then a desperate
charge was made on the right by the main body of the French
cavalry, which hoped to get into the rear of the British infantry; but
the attempted surprise failed: the French were driven back with
great loss. As the day dawned the battle became general, and the
French found themselves opposed not only by accustomed British
doggedness, but by a precision of fire and an adroitness of
manœuvre which astonished them. By ten o'clock the French were in
full flight for Alexandria, leaving seventeen hundred men on the
field. The loss of the British was stated at fourteen hundred killed
and wounded; and, unfortunately, the brave Abercromby was killed.
To complete the success, the Capitan Pacha's fleet in a few days
brought a Turkish army of between five and six thousand men, and
the Grand Vizier, posted at El Arish, began to march towards Cairo.
General Hutchinson, now chief in command of the British army,
hastened to join the Grand Vizier; but before he could accomplish
this, he had to drive four thousand French from a fortified camp at
Ramaneeh, and meanwhile five thousand French rushed out of Cairo
and attacked the Grand Vizier. On the 27th of June Cairo capitulated,
General Belliard obtaining the condition that his troops should be
conveyed to the ports of France on the Mediterranean with their
arms and baggage; yet they left behind them three hundred and
thirteen heavy cannon and one hundred thousand pounds of
gunpowder. On the 8th of June General Baird had landed at Cosseir
on the Red Sea with his Indian army, and was marching through the
burning desert for Cairo. Menou, cooped up at Alexandria, found it
useless to contend further and, before Baird could join the main
army, capitulated on the same terms as Belliard, and the Egyptian
campaign was at an end. The news of the French expulsion reached
France sooner than it did England, and created a strong sensation.
 SIR RALPH ABERCROMBY. (After the Portrait by J. Hoppner, R.A.)
[See larger version]
Britain was everywhere successful on the sea, and Lord Nelson, on
the 1st of August, made an attempt on the French flotilla lying at
Boulogne for the invasion of England. He was furnished with a flotilla
of gunboats for the purpose, and he was able to destroy two floating
batteries and a few gunboats, but found the fleet too strongly
posted under the batteries of the harbour to make further
impression. However, Napoleon saw that for the present an invasion
was out of the question, and the autumn of this year was employed
in endeavours to arrange a peace. Lord Cornwallis proceeded to
Paris for this object, and went to Amiens, which was appointed as
the place for the conference. The preliminaries were signed on the
1st of October, and General Lauriston, the schoolfellow and first
aide-de-camp of Buonaparte, brought them over to London. The
negotiations progressed slowly, being arrested now and then by the
conduct of the First Consul. Without waiting for the ratification of
peace, he sent off, on the 14th of December, 1801, only ten days
after the signing of the preliminaries, a strong fleet and army to the
West Indies to reduce the independent black Republic in St.
Domingo. Britain was obliged to send reinforcements to her own
West Indian fleet by Admiral Martin—so that it looked much more
like war than peace. Again, in January, 1802, came the news of the
election of Buonaparte to the Presidency of the Cisalpine Republic,
directly contrary to the Treaty of Lunéville, and betraying the
ambitious aims of Napoleon. Immediately followed the news that
Buonaparte had exacted from Spain a treaty by which Parma and
the island of Elba were made over to France on the death of the
present, already aged, duke; that Spain had been compelled to cede
part of the province of Louisiana in North America, by the same
treaty; and that Portugal, though the integrity of her dominions had
been carefully guaranteed by the preliminaries of peace, had by a
secret article given up to France her province of Guiana. A
Republican constitution was forced on Holland, and in Switzerland
instructions were given to the French Minister to thwart all efforts at
the formation of a stable constitution. These revelations startled the
British Ministers, but did not deter them from concluding the peace,
with the full approbation of Pitt. It was not that the First Consul,
who every day betrayed some fresh symptom of an insatiable
ambition, was disposed to offer them tempting terms; on the
contrary, though we were never more able to dictate measures at
sea, and he never less so, he was as haughty and dictatorial in his
demands as if Great Britain had been completely under his feet. Yet
the treaty went on, and was concluded and signed on the 27th of
March, 1802. It settled nothing, as Britain refused to acknowledge
the newly organised Republics, and declined to entertain Napoleon's
preposterous suggestion that Malta was to be occupied by
Neapolitan troops, under a neutrality guaranteed by all the chief
European Powers; since it was well known that Napoleon, when it
suited him, would cease to respect the conditions, and would readily
dispossess the troops of Naples. Though Pitt believed him to have
been sincere, Grenville, Windham, and Spencer saw that the
ambition of the "Little Corporal" was insatiable, and denounced the
treaty.
 CHAPTER XVIII.
REIGN OF GEORGE III. (continued).
Napoleon's Plans of Conquest—Sebastiani's Report—Napoleon's
Complaints against the British Press—Espionage and
Confiscation—He continues his Continental Aggressions—
Napoleon's Interview with Lord Whitworth—Imminence of War—
Negotiations for Pitt's Return to Office—War Declared—
Napoleon Arrests British subjects in France—Seizure of Hanover
—Emmett's Rebellion—Naval Attacks on the French Coast—The
Mahratta War—Battle of Assaye—Successes of General Lake—
Battle of Laswaree—Battle of Argaum—Conclusion of the War—
Renewed Illness of George III.—Increasing Opposition of Pitt—
He offers to undertake the Government—He forms a Tory
Ministry—Wilberforce's Abolition Motion—The Additional Force
Bill—Scheme for blowing up the French Fleet—War with Spain—
The Georges Conspiracy—Murder of the Duke D'Enghien—
Napoleon becomes Emperor—His Letter to the British King—The
Condition of Europe—Lord Mulgrave's Reply to the Letter—
Ministerial Changes—Weakness of the Ministry—Attack on Lord
Melville—Whitbread's Motion—Melville's Defence—His
Impeachment voted—Secession of Lord Sidmouth—The
European Coalition—Hastened by Napoleon's Aggressions—
Rashness of Austria—Invasion of Bavaria—Napoleon marches on
the Rhine—Capitulation of the Austrian Army at Ulm—
Occupation of Vienna—Battle of Austerlitz—Treaties of
Schönbrunn and Pressburg—The Baltic Expedition—Expedition
to Naples—Naval Affairs—Nelson's Pursuit of Villeneuve—
Calder's Engagement—Battle of Trafalgar—Death of Nelson—
Continuation of the Mahratta War—Lord Lake's Engagements
 with Holkar—Siege of Bhurtpore—Defeat of Meer Khan—The
Rajah of Bhurtpore makes Peace—Treaties with Scindiah and
Holkar—Death of Pitt—Payment of his Debts by the Nation.

The Peace of Amiens, instead of turning the attention of Buonaparte

to internal improvements, seemed to give it opportunity to range, in
imagination, over the whole world with schemes of conquest and of
the suppression of British dominion. There was no spot, however
remote, that he did not examine on the map with reference to plans
of conquest. Louisiana and Guiana, obtained from Spain and
Portugal, were viewed as ports whence conquest should advance to
Nova Scotia, Canada, the Brazils, Mexico, and Peru. Every station in
the West India Isles was calculated as a point for this purpose, and
for seizing some day all the British islands there. The Cape of Good
Hope, Madagascar, the isles of France and Bourbon, the Dutch spice
isles, and their settlements in Java, Sumatra, etc., were regarded as
a chain of ports which would enable Buonaparte to become master
of India. He sent out expeditions, under different officers, to
examine every island and region where the British had a settlement,
or where he might plant one, to oppose them. One of these
expeditions sailed in a couple of corvettes, commanded by Captain
Baudin, who was accompanied by a staff of thirty-three naturalists,
geologists, savants, etc., the ostensible object being science and
discovery—the real one the ascertaining of the exact possessions of
Britain, and of the best means of becoming master of them. The
head of the scientific staff was M. Péron. On their return their report
was published, and it is singular that in this report St. Helena,
destined to be the prison of Napoleon, is described in rapturous
terms as an earthly paradise.
Another expedition was that of Colonel Sebastiani, a Corsican, who
was despatched to Egypt, Syria, and other countries of the Levant.
Sebastiani reported to Buonaparte that the British were so detested
in Egypt that six thousand men would suffice to re-take it; that
Buonaparte's name was so venerated that it had procured him the
utmost honour everywhere, and especially with Djezzar Pacha,
Viceroy of Egypt. He asserted that General Stuart, the British envoy,
had endeavoured to excite the Turks to assassinate him. He
harangued the natives in the Ionian Isles, and assured them of the
protection of Buonaparte, and besides many calumnies against the
British officers, he told Napoleon that so hateful was the British rule
that both Greeks and Venetians in those islands were ready to rise
against them at the first word from France. On the appearance of
this base report, our ambassador at Paris made a strong
remonstrance; but Napoleon only replied by complaining of the late
account of the campaign in Egypt by Sir Robert Wilson, in which he
had detailed the butchery of the Turks and Arnauts at Jaffa, and
Napoleon's command to poison his own wounded on the retreat
from Acre. Through M. Otto, the French envoy in London, Napoleon
demanded that statements injurious to his character made by the
British press should be stopped by Government, that all French
emigrants should be expelled from England, that Georges Cadoudal
should be transported to Canada, and such princes of the House of
Bourbon as remained there should be advised to repair to Warsaw,
where the head of their house now resided. To these peremptory
demands the British Government, through Lord Hawkesbury, replied
that his Britannic Majesty did not possess the absolute power
necessary for these acts, and that whilst the statements charging
upon a British Ambassador instigations to murder were published in
the Moniteur, the official organ of the French Government, the
statements by the British press were protected by the freedom of
that press guaranteed in Great Britain, which the king was not
disposed to invade, but from which any man, British or foreign,
might claim redress by an action at law. To show the First Consul
how this might be done, the British Government commenced an
action against M. Peltier, a French emigrant, for a libel on Napoleon
in a newspaper published by him in London, called the Ambigu.
Peltier was found guilty; but this by no means answered
Buonaparte's object. He wanted the accounts of his darkest actions
suppressed by a power above the law, not thus made more public by
the action of the law. As Sir Walter Scott has observed, he wanted
darkness, and the British Government gave him light.
The audacity of Buonaparte still further excited the indignation of
the British Government. Under the name of consuls, he sent over to
England and Ireland a number of military officers, whose real
business was to act as privileged spies, to prepare plans of all the
chief ports, with soundings, and an exact account of the winds with
which vessels could go out or come in with most ease, and also at
what draught of water the harbours might be entered by large
vessels. These agents had been instructed to maintain the utmost
secrecy as to their real objects, but they became known, and
Ministers announced that any person coming in such a character to
this country should be ordered instantly to quit it. Neither was the
temper of the nation at all improved by the irritating proceedings of
the French authorities on the coasts of France. A law had been
passed by the Jacobins, in the most rabid time of the Revolution,
condemning any vessel under a hundred tons burden found within
four leagues of the French shores, having on board British
merchandise. It was taken for granted that this decree was virtually
annulled by the Peace of Amiens; but repeated seizures were now
made of British merchant vessels driven by stress of weather on the
French coasts, and the mere fact of having plates, knives, and forks
for the crew, of British make, was used as a plea for confiscation of
ships. It was in vain that remonstrances were made to the First
Consul: they passed without notice. Such a peace it was evident
could not last long. Napoleon was in a mood to brook no control
from any quarter; he at this time showed how completely he would
crush any creature who offended him when he had the power.
The Treaty of Amiens did not for a moment, even in appearance,
interrupt the unlimited plans of aggression which Buonaparte had
formed. Whether these plans tended to alarm Britain or not gave
him no concern whatever. The encroachments on Italy never
paused. Before the signing of the Peace of Amiens, Buonaparte had
made himself President of the Cisalpine Republic; and though he had
pledged himself to Alexander of Russia that he would not interfere
further with Piedmont, because Alexander would not entertain the
scheme of co-operating with France in the march to India, as his
father had done, Buonaparte seized on all Piedmont in September of
this year, annexed it to France, and divided it into six Departments.
Charles Emmanuel, the King of Piedmont, retired to his island of
Sardinia, and then abdicated in favour of his brother Victor
Emmanuel. But Victor Emmanuel would not have been left long king,
even of that small territory, had it not been for the protection of
Britain. In October he annexed Parma and Placentia. He next made
an agreement with the King of Naples for Elba, and took possession
of it. Every movement of this restless being showed his intention to
drive Britain out of the Mediterranean, and convert it into a French
lake. But on the mainland he was equally active. There was no
country on the Continent in which Buonaparte did not presume to
dictate, as if he already were universal monarch. In the Diet of
Germany his influence was prominently conspicuous, and he
prevailed to have towns and districts transferred as he pleased. To
have all the territory on the left bank of the Rhine secured to France,
Prussia received valuable compensation at the expense of the
German empire for the cession of the Duchy of Cleves and other
provinces transferred to France. Bavaria and other minor States were
benefited in the same way, because Napoleon already meant to use
these States against Austria and Russia, as he afterwards did. Every
endeavour was made, contrary to the articles of the Peace of
Amiens, to shut out the trade of Britain, not only with France—as he
had a right to do—but with Holland, Belgium, and Germany. It was
in vain that Britain remonstrated. Buonaparte, through his official
organ, the Moniteur, declared that "England should have the Treaty
of Amiens, the whole Treaty of Amiens, and nothing but the Treaty
of Amiens"; but he interpreted this treaty to give every advantage to
France to the exclusion of Britain. Half Europe was closed to British
trade. It was a condition of the Treaty of Lunéville that the
independence of Switzerland should be respected, and this was
guaranteed by the Batavian, Cisalpine, and Ligurian Republics, as
well as by France and Austria. But Buonaparte had already absorbed
all these republics into France, and Austria he set at defiance. He
had never withdrawn the French troops from Switzerland, but whilst
they remained French emissaries had continued to foment the feuds
between the people and the nobles, between one canton and
another. He now declared this state of things must end, and he
assumed the office of umpire, to settle the affairs of the Swiss for
them. He had no right to assume this office—if needed, it belonged
to the other Powers of Europe as well as France; but he knew that
he had the might—and he used it. At the end of September he sent
General Rapp to issue a manifesto announcing that Napoleon was
determined to put an end to all their differences. This manifesto was
immediately followed by the appearance of General Ney at the head
of forty thousand men, in addition to those already in the country.
Thus Switzerland was invaded, and its constitution trodden out by an
armed occupation. Buonaparte assumed the title of Mediator of the
Helvetic League, and dictated his own terms to the deputies of the
French party who were sent to Paris.
The king's speech at the opening of Parliament, and the martial tone
of the speeches by the members of both Houses, exceedingly
exasperated Napoleon; for though preparing for war he was scarcely
ready, and meant to have carried on the farce of peace a little
longer. Talleyrand demanded of Lord Whitworth the reason of this
ebullition of the British Parliament and of the Press. Lord Whitworth
replied, as he had done regarding the comments on the trial of
Peltier, that it was the direct result of the insulting articles in the
Moniteur, which was known to be the organ of the French
Government; whereas, in Britain, the Government had no direct
control, either over the speeches in Parliament or over the press.
Talleyrand and Whitworth again discussed all the vexed questions of
the retention of Malta, the conduct of Colonel Sebastiani in the East,
the aggressions of Napoleon in Italy, Germany, and Switzerland, in
violation of the Treaty of Amiens; and Lord Whitworth declared that
all Britain wanted was, that the Treaty should be faithfully carried
out on both sides; that we were ready to evacuate Malta, and recall
our complaints, on that being done. But this was what Napoleon was
resolved never to do, and he therefore resorted to the most
extraordinary insults to the British Ambassador. He requested Lord
Whitworth to call at the Tuileries at nine o'clock in the evening of the
day on which he had had his conference with Talleyrand. Napoleon
had, by an assumption of extreme hauteur and impetuosity,
frightened the Austrian Ambassador at Campo Formio, and he
probably thought of frightening the British one; but Britain had not
been beaten like Austria, and such a proceeding could only enrage
the British people. In this interview, Buonaparte ran over, in a rapid
and excited harangue of two hours' length, scarcely permitting Lord
Whitworth to interpose a word of reply, all the alleged causes of
dissatisfaction with England; at one moment threatening to invade it,
if it cost him his life; at another, proposing that France and England
should unite to rule the Continent, and offering to share with it all
the benefits of such an alliance. Lord Whitworth replied, as before,
that the British Government desired nothing but the bonâ fide
execution of the Treaty of Amiens, and could not for a moment
entertain such schemes of aggression and domination as the First
Consul proposed to her. He began to comment gravely on the
aggressions in Switzerland and Italy, but Buonaparte cut him short
angrily, saying those things were no business of his and that he had
no right to talk of them. There was a fresh interview with Talleyrand,
and fresh notes from him and Andreossi of the same character. A
similar though more violent scene occurred at a levee on the 13th of
March, in which Napoleon passionately accused Britain of driving
France into war. A shrewd observer, Madame de Rémusat, was of
opinion that his rage was simulated.
Everything in Parliament and in Ministerial movements now denoted
the near approach of the renewal of war. On the 8th of March a
message was received by both Houses of Parliament from his
Majesty, stating that great military preparations were going on in
Holland and France, and that his Majesty deemed it highly necessary
to take measures for the security of his dominions. It added that
negotiations were going on with France, the issue of which was
uncertain, but it neither stated what these negotiations were, nor
the measures called for. The message was taken for what it was—a
note of war, and both in the Lords and Commons strong expressions
of defiance were used to France. This seemed to have encouraged
Ministers to a plainer expression of their intentions, for only two
days later another message came down, calling for an increase of
the navy. The next day, the 11th, the Commons formed themselves
into a committee, and voted an addition of ten thousand seamen to
the fifty thousand already voted. The militia were embodied.
Sheridan was very zealous for war; Ministers, however, professed to
desire the continuance of peace if possible.
There were rumours of negotiations going on for a return of Pitt to
power; but as Mr. Addington showed no disposition to resign
altogether in favour of Pitt, these came to nothing. Already in the
previous November Canning and Lord Malmesbury, two of Pitt's most
zealous supporters, had set on foot an address to Addington begging
him to resign. But this plan was abruptly stopped by Pitt himself,
who felt that he was to a certain extent the creator of the Ministry.
Now Addington made overtures to Pitt through Dundas, become
Lord Melville. But his propositions were absurd. He stipulated that
Greville and Windham, who had opposed the Peace of Amiens,
should be excluded from the arrangements. He also wished that
Pitt's brother, Lord Chatham, should be the nominal Prime Minister,
while he and Pitt should be equal secretaries. At this, Pitt put a stop
to the envoy's conversation. "Upon my word," he said afterwards, "I
had not the curiosity to ask what I was to be." So England engaged
in war, with her greatest statesman excluded from office.
On the 6th of May Lord Pelham communicated to the Lords, and Mr.
Addington to the Commons, another message from his Majesty,
informing them that he had ordered Lord Whitworth, our
Ambassador, to quit Paris immediately, unless he saw a prospect of
closing the negotiations with the First Consul within a certain date;
and that M. Andreossi, the French Ambassador, had applied for his
passport, in order to quit London when Lord Whitworth should quit
Paris. In consequence of the uncertainty of the result there was an
adjournment, and then a second; but on the 16th of May all
suspense was terminated by the announcement of Ministers that
Lord Whitworth had quitted Paris, and M. Andreossi London. The
papers which had passed between this Government and France, in
the late negotiations, were ordered to be produced, and an Order in
Council was issued, directing reprisals to be granted against the
ships, goods, and subjects of the French Republic, and also for an
embargo not only on all French ships in British ports, but on all
Dutch vessels, and vessels of any Power under the military rule of
France. Britain was once more at war. On the 17th of June the king
announced, by message, that, in consequence of the Batavian
Republic refusing to order the French troops to quit Holland—which,
indeed, would have paid no attention to such orders—he had
recalled his Ambassador from the Hague and had issued letters of
marque and reprisals against that Republic. Thus, we were also at
war with Holland. At the same time a demand was made for a grant
of sixty thousand pounds, and a pension of sixteen thousand pounds
per annum to the Prince of Orange, the ex-Stadtholder, on the plea
that he was an exile and destitute; and the grant was voted.
Parliament was now daily occupied in passing fresh measures for the
defence of the country. It was voted, on the 20th of June, that a
reserve army of fifty thousand should be raised by ballot, like the
militia; and, indeed, it was no other than the extension of the militia:
for during the war this division was to serve only in Great Britain,
Ireland, and the Channel Islands. On the 18th of July it was
proposed to pass a Bill enabling his Majesty to raise a levy en masse
in case of invasion. Pitt strongly supported it, and proposed fresh
fortifications on the coasts.
 NAPOLEON AND HIS SUITE AT BOULOGNE. (See p. 490.)
[See larger version]
On the declaration of war, Buonaparte resorted to a proceeding that
had never been practised before, and which excited the most violent
indignation in England. He ordered the detention of British subjects
then in France, as prisoners of war. Talleyrand previously assured
some British travellers, who applied to him for information, that they
had nothing to fear; that their persons would be safe under the
protection of a Government which, unlike that of Britain, observed
the laws of nations, and Buonaparte caused his well-known agent,
Louis Goldsmith, the editor of a French paper, the Argus, published
in London, to insert the same assurance in that journal. Thus thrown
off their guard, all the British in France were seized by authority of a
proclamation of the 22nd of May. Numbers of these were families
and individuals not resident in France, but merely hurrying home
from Italy, Switzerland, etc. They numbered some 12,000, and were
kept confined till the close of the wars. The pretext was the capture
of two ships before war was declared, but they were not captured
until the Ambassadors had withdrawn, or until an embargo had been
laid by Napoleon on British shipping.
There was another point, besides the seizure of unsuspecting British
travellers, on which Buonaparte could deeply wound the honour of
the British monarch, and at the same time furnish himself with
considerable materials of war—the seizure of Hanover. George III.
held this hereditary territory distinct from his Crown of Britain, as a
State of the German federation. It was impossible to defend this
against France with the forces kept there, and Napoleon ordered
General Mortier to cross the Dutch frontier, and march into the
Electorate with twenty thousand men. The Duke of Cambridge, who
was Viceroy there, and General Walmoden, at first, put themselves
in an attitude of resistance; they called on the chief Powers of
Germany to protest against this invasion of the German Empire, and
to come to their aid, if this remonstrance was disregarded. The Duke
of Cambridge, seeing himself totally deserted by Germany, thought it
best to surrender Hanover to France, by agreement that the troops
should retire behind the Elbe, and not serve again till exchanged.
This was done at the end of May; the different towns made their
submission on the 3rd of June, and on the 5th Mortier entered
Hanover; the Duke of Cambridge had quitted the country; and the
British Cabinet refusing to ratify the Convention previously made
with him, he called on the Hanoverian army to surrender as
prisoners of war. Walmoden would have resisted with anything like
equal forces, but as that was impossible, he made the best terms he
could, which were that his army should give up their arms and
disband themselves.
Napoleon also exerted himself to excite a rebellion in Ireland. He
was the more bent on this, because he saw that it was hopeless to
make a direct descent on England itself. He had collected a great
fleet in the harbours of Boulogne, Dieppe, Havre, Dunkirk, Ostend,
and other smaller ports, many of them capable only of receiving the
gunboats in which he proposed to transport his soldiers. He had
assembled a very fine army on the heights above Boulogne, called
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookultra.com