Software Testing and Analysis 1st Edition Mauro

Pezze download pdf

https://ebookultra.com/download/software-testing-and-analysis-1st-
edition-mauro-pezze/

Visit ebookultra.com today to download the complete set of

ebook or textbook!
We have selected some products that you may be interested in
Click the link to download now or visit ebookultra.com
for more options!.

Software Testing 2nd Edition Edition Patton

https://ebookultra.com/download/software-testing-2nd-edition-edition-
patton/

Exploratory Software Testing 1st Edition James A.

Whittaker

https://ebookultra.com/download/exploratory-software-testing-1st-
edition-james-a-whittaker/

Fundamentals of Software Testing 2nd Edition Revised and

Updated Bernard Homès

https://ebookultra.com/download/fundamentals-of-software-testing-2nd-
edition-revised-and-updated-bernard-homes/

Software testing a craftman s approach 2nd edition Edition

Jorgensen

https://ebookultra.com/download/software-testing-a-craftman-s-
approach-2nd-edition-edition-jorgensen/
Effective Methods for Software Testing 3rd Edition William
E. Perry

https://ebookultra.com/download/effective-methods-for-software-
testing-3rd-edition-william-e-perry/

Advanced Software Testing Vol 3 2nd Edition Jamie L.

Mitchell

https://ebookultra.com/download/advanced-software-testing-vol-3-2nd-
edition-jamie-l-mitchell/

Professional Software Testing with Visual Studio 2005 Team

System Tools for Software Developers and Test Engineers
Tom Arnold
https://ebookultra.com/download/professional-software-testing-with-
visual-studio-2005-team-system-tools-for-software-developers-and-test-
engineers-tom-arnold/

Multi criteria Decision Analysis Methods and Software 1st

Edition Alessio Ishizaka

https://ebookultra.com/download/multi-criteria-decision-analysis-
methods-and-software-1st-edition-alessio-ishizaka/

Software Abstractions Logic Language and Analysis 2nd

Edition Daniel Jackson

https://ebookultra.com/download/software-abstractions-logic-language-
and-analysis-2nd-edition-daniel-jackson/
Software Testing and Analysis 1st Edition Mauro Pezze
Digital Instant Download
Author(s): Mauro Pezze, Michal Young
ISBN(s): 9780471455936, 0471455938
Edition: 1
File Details: PDF, 3.90 MB
Year: 2008
Language: english
Software Testing and Analysis:
Process, Principles, and
Techniques
Software Testing and Analysis:
Process, Principles, and
Techniques

Mauro Pezzè
Università di Milano Bicocca
Michal Young
University of Oregon
 PUBLISHER Daniel Sayre
SENIOR PRODUCTION EDITOR Lisa Wojcik
EDITORIAL ASSISTANT Lindsay Murdock
COVER DESIGNER Madelyn Lesure
COVER PHOTO Rick Fischer/Masterfile
W ILEY 200 TH ANNIVERSARY LOGO DESIGN Richard J. Pacifico
This book was typeset by the authors using pdfLATEXand printed and bound
by Malloy Lithographing. The cover was printed by Phoenix Color Corp.
This book is printed on acid free paper. •

Copyright c 2008 John Wiley & Sons, Inc. All rights reserved. No part of this
publication may be reproduced, stored in a retrieval system or transmitted in any form or
by any means, electronic, mechanical, photocopying, recording, scanning or otherwise,
except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act,
without either the prior written permission of the Publisher, or authorization through
payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc. 222
Rosewood Drive, Danvers, MA 01923, website www.copyright.com. Requests to the
Publisher for permission should be addressed to the Permissions Department, John Wiley
& Sons, Inc., 111 River Street, Hoboken, NJ 07030-5774, (201) 748-6011, fax
(201) 748-6008, website http://www.wiley.com/go/permissions.
To order books or for customer service please, call 1-800-CALL WILEY (225-5945).

ISBN-13 978-0-471-45593-6
Printed in the United States of America
10 9 8 7 5 6 4 3 2 1
Contents

List of Figures xi

List of Tables xv

I Fundamentals of Test and Analysis 1

1 Software Test and Analysis in a Nutshell 3
1.1 Engineering Processes and Verification . . . . . . . . . . . . . . . . 3
1.2 Basic Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 When Do Verification and Validation Start and End? . . . . . . . . . 5
1.4 What Techniques Should Be Applied? . . . . . . . . . . . . . . . . . 7
1.5 How Can We Assess the Readiness of a Product? . . . . . . . . . . . 10
1.6 How Can We Ensure the Quality of Successive Releases? . . . . . . . 11
1.7 How Can the Development Process Be Improved? . . . . . . . . . . 11

2 A Framework for Test and Analysis 15

2.1 Validation and Verification . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 Degrees of Freedom . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3 Varieties of Software . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3 Basic Principles 29
3.1 Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2 Redundancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Restriction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4 Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.5 Visibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.6 Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4 Test and Analysis Activities Within a Software Process 39

4.1 The Quality Process . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2 Planning and Monitoring . . . . . . . . . . . . . . . . . . . . . . . . 41
4.3 Quality Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.4 Dependability Properties . . . . . . . . . . . . . . . . . . . . . . . . 43
4.5 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

v
vi CONTENTS

4.6 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.7 Improving the Process . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.8 Organizational Factors . . . . . . . . . . . . . . . . . . . . . . . . . 50

II Basic Techniques 53
5 Finite Models 55
5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.2 Finite Abstractions of Behavior . . . . . . . . . . . . . . . . . . . . 58
5.3 Control Flow Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.4 Call Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.5 Finite State Machines . . . . . . . . . . . . . . . . . . . . . . . . . . 65

6 Dependence and Data Flow Models 77

6.1 Definition-Use Pairs . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6.2 Data Flow Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.3 Classic Analyses: Live and Avail . . . . . . . . . . . . . . . . . . . 85
6.4 From Execution to Conservative Flow Analysis . . . . . . . . . . . . 91
6.5 Data Flow Analysis with Arrays and Pointers . . . . . . . . . . . . . 94
6.6 Interprocedural Analysis . . . . . . . . . . . . . . . . . . . . . . . . 96

7 Symbolic Execution and Proof of Properties 101

7.1 Symbolic State and Interpretation . . . . . . . . . . . . . . . . . . . 102
7.2 Summary Information . . . . . . . . . . . . . . . . . . . . . . . . . 104
7.3 Loops and Assertions . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.4 Compositional Reasoning . . . . . . . . . . . . . . . . . . . . . . . 108
7.5 Reasoning about Data Structures and Classes . . . . . . . . . . . . . 109

8 Finite State Verification 113

8.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.2 State Space Exploration . . . . . . . . . . . . . . . . . . . . . . . . 116
8.3 The State Space Explosion Problem . . . . . . . . . . . . . . . . . . 126
8.4 The Model Correspondence Problem . . . . . . . . . . . . . . . . . 129
8.5 Granularity of Modeling . . . . . . . . . . . . . . . . . . . . . . . . 131
8.6 Intensional Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
8.7 Model Refinement . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
8.8 Data Model Verification with Relational Algebra . . . . . . . . . . . 140

III Problems and Methods 149

9 Test Case Selection and Adequacy 151
9.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
9.2 Test Specifications and Cases . . . . . . . . . . . . . . . . . . . . . 152
9.3 Adequacy Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
9.4 Comparing Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
 CONTENTS vii

10 Functional Testing 161

10.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
10.2 Random versus Partition Testing Strategies . . . . . . . . . . . . . . 162
10.3 A Systematic Approach . . . . . . . . . . . . . . . . . . . . . . . . 167
10.4 Choosing a Suitable Approach . . . . . . . . . . . . . . . . . . . . . 174

11 Combinatorial Testing 179

11.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
11.2 Category-Partition Testing . . . . . . . . . . . . . . . . . . . . . . . 180
11.3 Pairwise Combination Testing . . . . . . . . . . . . . . . . . . . . . 188
11.4 Catalog-Based Testing . . . . . . . . . . . . . . . . . . . . . . . . . 194

12 Structural Testing 211

12.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
12.2 Statement Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
12.3 Branch Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
12.4 Condition Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
12.5 Path Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
12.6 Procedure Call Testing . . . . . . . . . . . . . . . . . . . . . . . . . 229
12.7 Comparing Structural Testing Criteria . . . . . . . . . . . . . . . . . 230
12.8 The Infeasibility Problem . . . . . . . . . . . . . . . . . . . . . . . 230

13 Data Flow Testing 235

13.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
13.2 Definition-Use Associations . . . . . . . . . . . . . . . . . . . . . . 236
13.3 Data Flow Testing Criteria . . . . . . . . . . . . . . . . . . . . . . . 239
13.4 Data Flow Coverage with Complex Structures . . . . . . . . . . . . . 241
13.5 The Infeasibility Problem . . . . . . . . . . . . . . . . . . . . . . . 243

14 Model-Based Testing 245

14.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
14.2 Deriving Test Cases from Finite State Machines . . . . . . . . . . . . 246
14.3 Testing Decision Structures . . . . . . . . . . . . . . . . . . . . . . 251
14.4 Deriving Test Cases from Control and Data Flow Graphs . . . . . . . 257
14.5 Deriving Test Cases from Grammars . . . . . . . . . . . . . . . . . . 257

15 Testing Object-Oriented Software 271

15.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
15.2 Issues in Testing Object-Oriented Software . . . . . . . . . . . . . . 272
15.3 An Orthogonal Approach to Test . . . . . . . . . . . . . . . . . . . . 280
15.4 Intraclass Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
15.5 Testing with State Machine Models . . . . . . . . . . . . . . . . . . 282
15.6 Interclass Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
15.7 Structural Testing of Classes . . . . . . . . . . . . . . . . . . . . . . 293
15.8 Oracles for Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
15.9 Polymorphism and Dynamic Binding . . . . . . . . . . . . . . . . . 301
viii CONTENTS

15.10 Inheritance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303

15.11 Genericity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
15.12 Exceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

16 Fault-Based Testing 313

16.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
16.2 Assumptions in Fault-Based Testing . . . . . . . . . . . . . . . . . . 314
16.3 Mutation Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
16.4 Fault-Based Adequacy Criteria . . . . . . . . . . . . . . . . . . . . . 319
16.5 Variations on Mutation Analysis . . . . . . . . . . . . . . . . . . . . 321

17 Test Execution 327

17.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
17.2 From Test Case Specifications to Test Cases . . . . . . . . . . . . . . 328
17.3 Scaffolding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
17.4 Generic versus Specific Scaffolding . . . . . . . . . . . . . . . . . . 330
17.5 Test Oracles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
17.6 Self-Checks as Oracles . . . . . . . . . . . . . . . . . . . . . . . . . 334
17.7 Capture and Replay . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

18 Inspection 341
18.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
18.2 The Inspection Team . . . . . . . . . . . . . . . . . . . . . . . . . . 343
18.3 The Inspection Process . . . . . . . . . . . . . . . . . . . . . . . . . 344
18.4 Checklists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
18.5 Pair Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . 351

19 Program Analysis 355

19.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
19.2 Symbolic Execution in Program Analysis . . . . . . . . . . . . . . . 356
19.3 Symbolic Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
19.4 Summarizing Execution Paths . . . . . . . . . . . . . . . . . . . . . 359
19.5 Memory Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
19.6 Lockset Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
19.7 Extracting Behavior Models from Execution . . . . . . . . . . . . . 365

IV Process 373
20 Planning and Monitoring the Process 375
20.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
20.2 Quality and Process . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
20.3 Test and Analysis Strategies . . . . . . . . . . . . . . . . . . . . . . 377
20.4 Test and Analysis Plans . . . . . . . . . . . . . . . . . . . . . . . . 382
20.5 Risk Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
20.6 Monitoring the Process . . . . . . . . . . . . . . . . . . . . . . . . . 389
 CONTENTS ix

20.7 Improving the Process . . . . . . . . . . . . . . . . . . . . . . . . . 394

20.8 The Quality Team . . . . . . . . . . . . . . . . . . . . . . . . . . . 399

21 Integration and Component-based Software Testing 405

21.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
21.2 Integration Testing Strategies . . . . . . . . . . . . . . . . . . . . . . 408
21.3 Testing Components and Assemblies . . . . . . . . . . . . . . . . . 413

22 System, Acceptance, and Regression Testing 417

22.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
22.2 System Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
22.3 Acceptance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
22.4 Usability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423
22.5 Regression Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
22.6 Regression Test Selection Techniques . . . . . . . . . . . . . . . . . 428
22.7 Test Case Prioritization and Selective Execution . . . . . . . . . . . . 434

23 Automating Analysis and Test 439

23.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
23.2 Automation and Planning . . . . . . . . . . . . . . . . . . . . . . . 441
23.3 Process Management . . . . . . . . . . . . . . . . . . . . . . . . . . 441
23.4 Static Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
23.5 Test Case Generation and Execution . . . . . . . . . . . . . . . . . . 445
23.6 Static Analysis and Proof . . . . . . . . . . . . . . . . . . . . . . . . 445
23.7 Cognitive Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
23.8 Version Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
23.9 Debugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
23.10 Choosing and Integrating Tools . . . . . . . . . . . . . . . . . . . . 451

24 Documenting Analysis and Test 455

24.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
24.2 Organizing Documents . . . . . . . . . . . . . . . . . . . . . . . . . 456
24.3 Test Strategy Document . . . . . . . . . . . . . . . . . . . . . . . . 458
24.4 Analysis and Test Plan . . . . . . . . . . . . . . . . . . . . . . . . . 458
24.5 Test Design Specification Documents . . . . . . . . . . . . . . . . . 460
24.6 Test and Analysis Reports . . . . . . . . . . . . . . . . . . . . . . . 462

Bibliography 467

Index 479
x CONTENTS
List of Figures

1 Selective reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

1.1 Analysis and testing activities . . . . . . . . . . . . . . . . . . . . . 9

2.1 Validation and verification . . . . . . . . . . . . . . . . . . . . . . . 16

2.2 Verification trade-off dimensions . . . . . . . . . . . . . . . . . . . . 19

3.1 Unpredictable failure and predictable failure . . . . . . . . . . . . . 31

3.2 Initialize before use problem . . . . . . . . . . . . . . . . . . . . . . 34

4.1 Dependability properties . . . . . . . . . . . . . . . . . . . . . . . . 46

5.1 Abstraction coalesces execution states . . . . . . . . . . . . . . . . . 58

5.2 Constructing control flow graphs . . . . . . . . . . . . . . . . . . . . 59
5.3 Java method to collapse adjacent newline characters . . . . . . . . . 61
5.4 Statements broken across basic blocks . . . . . . . . . . . . . . . . . 62
5.5 Linear-code sequence and jump (LCSAJ) . . . . . . . . . . . . . . . 62
5.6 Over-approximation in a call graph . . . . . . . . . . . . . . . . . . 64
5.7 Context sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.8 Exponential explosion of calling contexts in a call graph . . . . . . . 67
5.9 Finite state machine specification of line-end conversion procedure . 69
5.10 Correctness relations for a finite state machine model . . . . . . . . . 70
5.11 Procedure to convert among Dos, Unix, and Macintosh line ends . . . 72
5.12 Completed FSM specification of line-end conversion procedure . . . 73

6.1 GCD calculation in Java . . . . . . . . . . . . . . . . . . . . . . . . 78

6.2 Control flow graph of GCD method . . . . . . . . . . . . . . . . . . 79
6.3 Data dependence graph of GCD method . . . . . . . . . . . . . . . . 80
6.4 Calculating control dependence . . . . . . . . . . . . . . . . . . . . 81
6.5 Control dependence tree of GCD method . . . . . . . . . . . . . . . 82
6.6 Reaching definitions algorithm . . . . . . . . . . . . . . . . . . . . . 84
6.7 Available expressions algorithm . . . . . . . . . . . . . . . . . . . . 86
6.8 Java method with potentially uninitialized variable . . . . . . . . . . 87
6.9 Control flow with definitions and uses . . . . . . . . . . . . . . . . . 88
6.10 Annotated CFG for detecting uses of uninitialized variables . . . . . 89

xi
xii LIST OF FIGURES

6.11 CGI program in Python with misspelled variable . . . . . . . . . . . 91

6.12 Powerset lattice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.13 Spurious execution paths in interprocedural analysis . . . . . . . . . 97

7.1 Binary search procedure . . . . . . . . . . . . . . . . . . . . . . . . 103

7.2 Concrete and symbolic tracing . . . . . . . . . . . . . . . . . . . . . 104

8.1 Finite state verification . . . . . . . . . . . . . . . . . . . . . . . . . 115

8.2 Misapplication of the double-check initialization pattern . . . . . . . 118
8.3 FSM models from Figure 8.2 . . . . . . . . . . . . . . . . . . . . . . 119
8.4 Promela finite state model . . . . . . . . . . . . . . . . . . . . . . . 120
8.5 Excerpts of Spin verification tool transcript . . . . . . . . . . . . . . 121
8.6 Spin guided simulation trace describing race condition . . . . . . . . 123
8.7 A graphical interpretation of Spin guided simulation trace . . . . . . 124
8.8 Dining philosophers in Promela . . . . . . . . . . . . . . . . . . . . 128
8.9 A simple data race in Java . . . . . . . . . . . . . . . . . . . . . . . 131
8.10 Coarse and fine-grain models of interleaving . . . . . . . . . . . . . 132
8.11 Lost update problem . . . . . . . . . . . . . . . . . . . . . . . . . . 133
8.12 OBDD encoding of a propositional formula . . . . . . . . . . . . . . 136
8.13 OBDD representation of transition relation . . . . . . . . . . . . . . 137
8.14 Data model of a simple Web site . . . . . . . . . . . . . . . . . . . . 141
8.15 Alloy model of a Web site. . . . . . . . . . . . . . . . . . . . . . . . 142
8.16 Alloy model of a Web site (continued) . . . . . . . . . . . . . . . . . 143
8.17 A Web site that violates the “browsability” property . . . . . . . . . 145

9.1 A Java method for collapsing sequences of blanks . . . . . . . . . . . 155

10.1 A Java class for finding roots of a quadratic equation . . . . . . . . . 165

10.2 A quasi-partition of a program’s input domain . . . . . . . . . . . . . 167
10.3 The functional testing process . . . . . . . . . . . . . . . . . . . . . 169

11.1 Specification of Check configuration . . . . . . . . . . . . . . . . . . 182

11.2 Specification of cgi decode . . . . . . . . . . . . . . . . . . . . . . . 195
11.3 Elementary items of specification cgi decode . . . . . . . . . . . . . 198
11.4 Test case specifications for cgi decode generated after step 2 . . . . . 201

12.1 The C function cgi decode . . . . . . . . . . . . . . . . . . . . . . . 213

12.2 Control flow graph of function cgi decode . . . . . . . . . . . . . . . 214
12.3 The control flow graph of C function cgi decode0 . . . . . . . . . . . 218
12.4 Deriving a tree from a control flow graph for boundary/interior testing 223
12.5 Buggy self-organizing list . . . . . . . . . . . . . . . . . . . . . . . 224
12.6 Control flow graph of C function search . . . . . . . . . . . . . . . . 225
12.7 Tree of boundary/interior sub-paths for C function search . . . . . . 226
12.8 Subsumption relations among structural test adequacy criteria . . . . 231

13.1 The C function cgi decode . . . . . . . . . . . . . . . . . . . . . . . 237

13.2 A C procedure with a large number of DU paths . . . . . . . . . . . 241
 LIST OF FIGURES xiii

13.3 Pointer arithmetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242

14.1 Functional specification of feature Maintenance . . . . . . . . . . . . 247

14.2 The finite state machine corresponding to Maintenance . . . . . . . . 248
14.3 Functional specification of feature Pricing . . . . . . . . . . . . . . . 252
14.4 Decision table for Pricing . . . . . . . . . . . . . . . . . . . . . . . 254
14.5 Set of test cases corresponding to the modified adequacy criterion . . 256
14.6 Functional specification of Process shipping order . . . . . . . . . . 258
14.7 Control flow model of Process shipping order . . . . . . . . . . . . 259
14.8 Node-adequate test suite . . . . . . . . . . . . . . . . . . . . . . . . 260
14.9 Branch-adequate test suite . . . . . . . . . . . . . . . . . . . . . . . 260
14.10 Functional specification of Advanced search . . . . . . . . . . . . . . 261
14.11 BNF description of Advanced search . . . . . . . . . . . . . . . . . 261
14.12 XML schema for Product configuration . . . . . . . . . . . . . . . . 263
14.13 BNF description of Product configuration . . . . . . . . . . . . . . . 264
14.14 Test case for feature Advanced Search . . . . . . . . . . . . . . . . . 264
14.15 The BNF description of Product Configuration . . . . . . . . . . . . 265
14.16 Sample seed probabilities for the BNF of Product Configuration . . . 266

15.1 Part of a Java implementation of class Model . . . . . . . . . . . . . 274

15.2 More of the Java implementation of class Model . . . . . . . . . . . . 275
15.3 Class diagram for the LineItem hierarchy. . . . . . . . . . . . . . . . 276
15.4 Part of a Java implementation of class Account. . . . . . . . . . . . . 278
15.5 Impact of object-oriented design on analysis and test. . . . . . . . . . 279
15.6 Statechart specification of class Model . . . . . . . . . . . . . . . . . 284
15.7 Finite state machine corresponding to the statechart in Figure 15.6 . . 285
15.8 Statechart specification of class Order . . . . . . . . . . . . . . . . . 287
15.9 Finite state machine corresponding to the statechart in Figure 15.8 . . 288
15.10 Class diagram of the Chipmunk Web presence . . . . . . . . . . . . 290
15.11 Use/include relation for the class diagram in Figure 15.10 . . . . . . 291
15.12 Sequence diagram for configuring an order . . . . . . . . . . . . . . 294
15.13 Partial intraclass control flow graph for class Model . . . . . . . . . . 296
15.14 Summary information for structural interclass testing . . . . . . . . . 299
15.15 Polymorphic method call . . . . . . . . . . . . . . . . . . . . . . . . 302
15.16 Part of a Java implementation of the abstract class LineItem . . . . . . 305
15.17 Part of a Java implementation of class CompositeItem . . . . . . . . . 307

16.1 Program transduce . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

16.2 Sample mutation operators for C . . . . . . . . . . . . . . . . . . . . 318
16.3 Sample mutants for program Transduce . . . . . . . . . . . . . . . . 320
16.4 Edit distance check . . . . . . . . . . . . . . . . . . . . . . . . . . . 325

17.1 JUnit tests in JFlex . . . . . . . . . . . . . . . . . . . . . . . . . . . 331

17.2 Test harness with comparison-based test oracle . . . . . . . . . . . . 333
17.3 Testing with self-checks . . . . . . . . . . . . . . . . . . . . . . . . 334
17.4 Structural invariant as run-time self-check . . . . . . . . . . . . . . . 336
xiv LIST OF FIGURES

18.1 Detailed description referenced by a checklist item. . . . . . . . . . . 349

19.1 A C program invoking cgi decode . . . . . . . . . . . . . . . . . . . 361

19.2 Purify verification tool transcript. . . . . . . . . . . . . . . . . . . . 362
19.3 Model of memory states . . . . . . . . . . . . . . . . . . . . . . . . 363
19.4 Concurrent threads with shared variables . . . . . . . . . . . . . . . 364
19.5 Lockset state transition diagram . . . . . . . . . . . . . . . . . . . . 365
19.6 A Java method for inserting a node in an AVL tree . . . . . . . . . . 367
19.7 Sample set of predicates for behavior program analysis . . . . . . . . 368
19.8 Test cases for an AVL tree . . . . . . . . . . . . . . . . . . . . . . . 369
19.9 Behavioral models for method insert . . . . . . . . . . . . . . . . . . 370

20.1 Alternative schedules . . . . . . . . . . . . . . . . . . . . . . . . . . 385

20.2 A sample A&T schedule . . . . . . . . . . . . . . . . . . . . . . . . 387
20.3 Typical fault distribution over time . . . . . . . . . . . . . . . . . . . 392

21.1 Chipmunk Web presence hierarchy . . . . . . . . . . . . . . . . . . 411

22.1 Version 1.0 of the C function cgi decode . . . . . . . . . . . . . . . 430

22.2 Version 2.0 of the C function cgi decode . . . . . . . . . . . . . . . 431
22.3 Coverage of structural test cases for cgi decode . . . . . . . . . . . . 432
22.4 Control flow graph of function cgi decode version 2.0 . . . . . . . . 433
22.5 New definitions and uses for cgi decode . . . . . . . . . . . . . . . . 434
22.6 Flow graph model of the extended shipping order specification . . . . 435

23.1 CodeCrawler code size visualization . . . . . . . . . . . . . . . . . . 450

24.1 Sample document naming conventions . . . . . . . . . . . . . . . . . 456

List of Tables

11.1 Example categories and value classes . . . . . . . . . . . . . . . . . 187

11.2 Test case specifications for Check configuration . . . . . . . . . . . . 189
11.3 Parameters and values for Display control . . . . . . . . . . . . . . . 190
11.4 Pairwise coverage of three parameters . . . . . . . . . . . . . . . . . 191
11.5 Pairwise coverage of five parameters . . . . . . . . . . . . . . . . . . 192
11.6 Constraints for Display control . . . . . . . . . . . . . . . . . . . . . 193
11.7 A test catalog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
11.8 Summary of catalog-based test cases for cgi decode . . . . . . . . . . 205

12.1 Test cases for cgi decode . . . . . . . . . . . . . . . . . . . . . . . . 215

13.1 Definitions and uses for C function cgi decode . . . . . . . . . . . . 238

13.2 DU pairs for C function cgi decode . . . . . . . . . . . . . . . . . . 239

14.1 A test suite derived from the FSM of Figure 14.2 . . . . . . . . . . . 249

15.1 Test cases to satisfy transition coverage criterion . . . . . . . . . . . . 285

15.2 Simple transition coverage . . . . . . . . . . . . . . . . . . . . . . . 289
15.3 Equivalent scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . 301
15.4 Pairwise combinatorial coverage of polymorphic binding . . . . . . . 302
15.5 Testing history for class LineItem . . . . . . . . . . . . . . . . . . . . 304
15.6 Testing history for class CompositeItem . . . . . . . . . . . . . . . . 306

20.1 Standard severity levels for root cause analysis . . . . . . . . . . . . . 397

21.1 Integration faults. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

xv
xvi LIST OF TABLES
Preface

This book addresses software test and analysis in the context of an overall effort to
achieve quality. It is designed for use as a primary textbook for a course in software
test and analysis or as a supplementary text in a software engineering course, and as a
resource for software developers.

The main characteristics of this book are:

• It assumes that the reader’s goal is to achieve a suitable balance of cost, sched-
ule, and quality. It is not oriented toward critical systems for which ultra-high
reliability must be obtained regardless of cost, nor will it be helpful if one’s aim
is to cut cost or schedule regardless of consequence.

• It presents a selection of techniques suitable for near-term application, with suf-

ficient technical background to understand their domain of applicability and to
consider variations to suit technical and organizational constraints. Techniques
of only historical interest and techniques that are unlikely to be practical in the
near future are omitted.

• It promotes a vision of software testing and analysis as integral to modern soft-

ware engineering practice, equally as important and technically demanding as
other aspects of development. This vision is generally consistent with current
thinking on the subject, and is approached by some leading organizations, but is
not universal.

• It treats software testing and static analysis techniques together in a coherent

framework, as complementary approaches for achieving adequate quality at ac-
ceptable cost.

Why This Book?

One cannot “test quality into” a badly constructed software product, but neither can one
build quality into a product without test and analysis. The goal of acceptable quality
at acceptable cost is both a technical and a managerial challenge, and meeting the goal
requires a grasp of both the technical issues and their context in software development.

xvii
xviii Preface

It is widely acknowledged today that software quality assurance should not be a

phase between development and deployment, but rather a set of ongoing activities in-
terwoven with every task from initial requirements gathering through evolution of the
deployed product. Realization of this vision in practice is often only partial. It requires
careful choices and combinations of techniques fit to the organization, products, and
processes, but few people are familiar with the full range of techniques, from inspection
to testing to automated analyses. Those best positioned to shape the organization and
its processes are seldom familiar with the technical issues, and vice versa. Moreover,
there still persists in many organizations a perception that quality assurance requires
less skill or background than other aspects of development.
This book provides students with a coherent view of the state of the art and practice,
and provides developers and managers with technical and organizational approaches to
push the state of practice toward the state of the art.

Who Is This Book For?

Students who read portions of this book will gain a basic understanding of principles
and issues in software test and analysis, including an introduction to process and or-
ganizational issues. Developers, including quality assurance professionals, will find a
variety of techniques with sufficient discussion of technical and process issues to sup-
port adaptation to the particular demands of their organization and application domain.
Technical managers will find a coherent approach to weaving software quality assur-
ance into the overall software process. All readers should obtain a clearer view of the
interplay among technical and nontechnical issues in crafting an approach to software
quality.
Students, developers, and technical managers with a basic background in computer
science and software engineering will find the material in this book accessible without
additional preparation. Some of the material is technically demanding, but readers may
skim it on a first reading to get the big picture, and return to it at need.
A basic premise of this book is that effective quality assurance is best achieved
by selection and combination of techniques that are carefully woven into (not grafted
onto) a software development process for a particular organization. A software quality
engineer seeking technical advice will find here encouragement to consider a wider
context and participate in shaping the development process. A manager whose faith
lies entirely in process, to the exclusion of technical knowledge and judgment, will
find here many connections between technical and process issues, and a rationale for a
more comprehensive view.

How to Read This Book

This book is designed to permit selective reading. Most readers should begin with
Part I, which presents fundamental principles in a coherent framework and lays the
groundwork for understanding the strengths and weaknesses of individual techniques
and their application in an effective software process. Part II brings together basic tech-
 xix

nical background for many testing and analysis methods. Those interested in particular
methods may proceed directly to the relevant chapters in Part III of the book. Where
there are dependencies, the Required Background section at the beginning of a chap-
ter indicates what should be read in preparation. Part IV discusses how to design a
systematic testing and analysis process and incorporates it into an overall development
process, and may be read either before or after Part III.
Readers new to the field of software test and analysis can obtain an overview by reading
Chapters
1 Software Test and Analysis in a nutshell
2 A Framework for Test and Analysis
4 Test and Analysis Activities within a Software Process
10 Functional Testing
11 Combinatorial Testing
14 Model-Based Testing
15 Testing Object-Oriented Software
17 Test Execution
18 Inspection
19 Program Analysis
20 Planning and Monitoring the Process

Notes for Instructors

This book can be used in an introductory course in software test and analysis or as a
supplementary text in an undergraduate software engineering course.
An introductory graduate-level or an undergraduate level course in software test and
analysis can cover most of the book. In particular, it should include

• All of Part I (Fundamentals of Test and Analysis), which provides a complete

overview.

• Most of Part II (Basic Techniques), which provides fundamental background,

possibly omitting the latter parts of Chapters 6 (Dependence and Data Flow
Models) and 7 (Symbolic Execution and Proof of Properties). These chapters are
particularly suited for students who focus on theoretical foundations and those
who plan to study analysis and testing more deeply.

• A selection of materials from Parts III (Problems and Methods) and IV (Process).

For a course with more emphasis on techniques than process, we recommend

• Chapter 10 (Functional Testing), to understand how to approach black-box test-

ing.

• The overview section and at least one other section of Chapter 11 (Combinatorial
Testing) to grasp some combinatorial techniques.
xx Preface

• Chapter 12 (Structural Testing), through Section 12.3, to introduce the basic cov-
erage criteria.

• Chapter 13 (Data Flow Testing), through Section 13.3, to see an important appli-
cation of data flow analysis to software testing.

• The overview section and at least one other section of Chapter 14 (Model-based
Testing) to grasp the interplay between models and testing.

• Chapter 15 (Testing Object-Oriented Software) to appreciate implications of the

object-oriented paradigm on analysis and testing.

• Chapter 17 (Test Execution), to manage an easily overlooked set of problems

and costs.

• Chapter 18 (Inspection) to grasp the essential features of inspection and appreci-

ate the complementarity of analysis and test.

• Chapter 19 (Program Analysis) to understand the role of automated program

analyses and their relation to testing and inspection techniques.

• Chapters 20 (Planning and Monitoring the Process), 21 (Integration and Component-

based Software Testing), and 22 (System, Acceptance, and Regression Testing)
to widen the picture of the analysis and testing process.

For a stronger focus on software process and organizational issues, we recommend

• Chapter 10 (Functional Testing), a selection from Chapters 11 and 14 (Com-

binatorial Testing and Model-Based Testing), and Chapters 15 (Testing Object-
Oriented Software), 17 (Test Execution), 18 (Inspection), and 19 (Program Anal-
ysis) to provide a basic overview of techniques.

• Part IV, possibly omitting Chapter 23 (Automating Analysis and Test), for a
comprehensive view of the quality process.

When used as a supplementary text in an undergraduate software engineering course,

Chapters 1 (Software Test and Analysis in a Nutshell), and 2 (A Framework for Test
and Analysis) can provide a brief overview of the field. We recommend completing
these two essential chapters along with either Chapter 4, or a selection of chapters
from Part III, or both, depending on the course schedule. Chapter 4 (Test and Analysis
Activities within a Software Process) can be used to understand the essential aspects
of a quality process. The following chapters from Part III will help students grasp
essential techniques:

• Chapter 10 (Functional Testing) and a selection of techniques from Chapters 11

(Combinatorial Testing) and 14 (Model-Based Testing), to grasp basic black-box
testing techniques.

• Chapter 12 (Structural Testing), through Section 12.3, to introduce basic cover-

age criteria.
Random documents with unrelated
content Scribd suggests to you:
to be trusted, but by their overbearing manners they incurred the
dislike of the English, who were jealous of their superior industry
and efficiency. Even amongst British troops there was a tendency to
desert and join King James. Sir John Lanier, who was not
conciliatory, had much trouble with his men.[244]
Schomberg thought it might be possible to risk a
battle when his army had been reinforced by a Sufferings of the
army at Dundalk,
promised Danish contingent, who would be useless
if they arrived after a defeat. Owing to French
intrigues, and other diplomatic hindrances, these and at Belfast.
valuable allies did not reach England until November. They were
about 5000 foot and 1000 horse, ‘old disciplined soldiers, and very
civil, and the Duke of Wirtemberg their general.’ By that time
Schomberg had retired from Dundalk, where the mortality had been
frightful, and the state of things in the Belfast hospital was no better.
Story, who was an eye-witness, and who did what he could to help
the poor soldiers, has left notices of these scenes which, in their
simple brevity, vie with the descriptions of Thucydides and Manzoni.
They were quite demoralised, not caring much for the death of
comrades, but resenting their burial because they could then no
longer make shelters with the bodies. During the retreat, two men
died at Newry among a number who cowered in a ruined stable. The
survivors begged the chaplain to get them a fire, ‘which I did,
coming in about two hours after they had pulled in the two dead
men to make seats of.’ The cavalry suffered less than the infantry
because they had cloaks. Further on in the winter, when thousands
had already died, 18,000 great-coats were ordered in London.
Regimental surgeons at 4s. 6d. a day with assistants at 2s. 6d. were
not likely to be very efficient. Apothecaries received but 1s. The
supply of medical stores was altogether inadequate, very little for
wounds, and nothing at all for fever and dysentery, which were the
real destroyers. The mortality was not quite so great among the
officers, who were better clad and had more foresight than their
men, but many died. With some exceptions, the English officers,
many of whom drank hard, were not careful enough of those under
them, and compared unfavourably with the professional foreign
soldiers. Count Solms, in particular, was said to be a father to his
men. The fever, originating in the camp, spread all over the north.
Vessels lay off Belfast entirely filled with the dead, like the phantom
ships in Campbell’s poem, and the greatest mortality was among
people who lived near the hospital there. At first the Irish army
suffered less, for they escaped much of the rain which fell upon
Dundalk. The superstitious attributed this to a judgment, but Story
who was not superstitious, says ‘it was because we lay in a hollow at
the bottom of the mountains, and they upon a high sound ground ...
they were born in the country and were used to bad lying and
feeding.’ Between the camp, the transport, and the hospital about
half of Schomberg’s army perished. Before he retreated, the
mortality among James’s men was nearly as great, and they
confessed to a loss of 7000. Of Schomberg’s army, 1700 died at
Dundalk, 800 on ship-board, and 3800 in the Belfast hospital,
leaving only 7700 survivors.[245]
When military affairs go wrong, it is the common
practice to seek for a scapegoat. Schomberg was Mismanagement
and corruption.
disposed to blame the paymaster, William Harbord, Harbord and
whom he thought more attentive to his own Shales.
perquisites than to the business of his place. He
was superseded early in the following year, but was employed
elsewhere. A French apothecary was put under arrest for not
providing the necessary medicines. But both the general and the
public agreed in condemning Commissary-General Shales, whose
experience had been gained in victualling King James’s army during
the Monmouth affair, and afterwards in the camp at Hounslow. He
had lingered long in England after Schomberg took up his position at
Dundalk, and it was scarcely denied that he had been guilty, but he
said that if pressed he would put the saddle on the right horse. His
conduct was the subject of a very hot debate in the House of
Commons. ‘Whoever put this man into this trust,’ said Birch, ‘are
friends to King James and not to King William,’ and that was the
opinion of the majority who addressed the King for the removal of
Shales and for the name of the person who recommended him.
Somers suggested that His Majesty should be merely asked to
dismiss the person who had given the advice; but the more violent
course was adopted. William readily agreed to dismiss Shales and
impound his papers, but said it was impossible for him to name his
adviser. There is some reason to suppose that the statesman aimed
at was Halifax. Shales was sent over to England, but not until after
the dissolution, and nothing further seems to have been done.[246]
The Jacobites, though their challenge was not a
very determined one, claimed to have had the best Sligo taken and
retaken.
of the campaign. But between the two main armies
there was no serious fighting. Towards the end of September,
Colonel Lloyd left Sligo with a small force of Enniskilleners, passed
the Curlew hills, and defeated a much larger body under O’Kelly,
killing many and taking many prisoners, besides a great quantity of
cattle. So complete was the victory that he was able to occupy Boyle
and Jamestown on the Shannon. Schomberg, who was glad to have
any chance of encouraging his men, made much of this affair,
paraded all the Enniskillen men in camp, and rode along their ranks
with his head bare. Three rounds of musketry were fired, and also
salutes from some of the big guns. The noise excited wonder and
some alarm in the Irish camp, but James’s men professed to be
ignorant of any defeat. Sarsfield and Henry Luttrell were, however,
sent to the West a few weeks later with a considerable force.
Schomberg had not men enough to operate in that direction, and
Lloyd was soon driven from his new conquests back into Sligo, and
from thence to Ballyshannon. The fort commanding Sligo was
gallantly defended for three days by St. Sauveur, a French captain,
but provisions and water failed, and he was forced to surrender,
marching out with arms and baggage. Sarsfield kept his word
strictly, and as the garrison filed past he offered five guineas and a
horse and arms to any soldier who would serve King James. One
Huguenot accepted his offer, but carried the guineas, the horse, and
the arms to Ballyshannon. All the rest declared that they would
never fight for the ‘papishes.’ Sligo was the key of Connaught, and
the whole province remained in James’s hands until after the Boyne.
[247]

James was inclined to cling to the position at Ardee

with its unfinished fortifications, which Avaux had Sufferings of
James’s army.
always said would be useless. But the ambassador
prevailed upon him to remove his headquarters to
Dublin, where at least the soldiers would not have State of Dublin.
to live in huts that did not keep out the rain. Scarcely any of them
had shirts, one-half were without shoes and stockings, and one-third
were bare-legged. The country was exhausted, and the magazines
recommended by Avaux and Rosen had never been built. When the
camp was evacuated, many dying men were left behind without food
or care. In the hospital established between Ardee and Drogheda,
there were 300 sick without provisions, wine, or beer. There was no
doctor, no baker or cook, not even an attendant to bring a glass of
water. At Drogheda there were over 200 more in a disused church.
One-third had palliasses, the rest lay on the ground, with scarcely
any food, and no drink but bad water. Dublin itself was given up to
riot and dissipation during the winter and early spring. The city, says
Stevens, ‘seemed to be a seminary of vice, an academy of luxury, or
rather a sink of corruption, and living emblem of Sodom.’ Other
Jacobite accounts are much to the same effect. Among the worst
drunkards was James’s son Henry, who enjoyed the empty title of
Grand Prior. He was Stevens’s colonel, but scarcely ever fit for duty.
Dining one day with some Irish officers he began to quarrel with
them, and Berwick tried to smooth matters by drinking confusion to
Melfort. The Grand Prior then declared that Melfort was his friend
and an honest man, and ended by breaking his full glass on Lord
Dungan’s nose. James was willing that his son should fight, but
Dungan very wisely passed the matter over as a childish ebullition.
[248]

King James did little to improve the state of affairs.

He seldom made up his mind until it was too late, Incompetence of
James.
and would scarcely listen to those who sought to
establish discipline. There were many French
officers who could be of little use, for they had no Lauzun is sent to
Ireland.
direct charge of the soldiers, and received
commissions as majors and colonels with no duties attached. Some
of them indeed, in Boisseleau’s language, were good neither to boil
nor to roast. There were also many French swashbucklers, who did
nothing but increase the ill-feeling between their countrymen and
the Irish. Tyrconnel, Avaux, and Nagle worked together to evolve
order out of chaos. They suppressed over a hundred loose
companies, and aimed at reducing the army to twenty battalions of
800 men, seven regiments of cavalry, and seven of dragoons. Louis
made up his mind to send over six or seven thousand men as soon
as he could spare them, receiving in exchange a like number of Irish
for his own army. About the beginning of 1690 it was known that
troops were going to Ireland and that Lauzun would command them.
The decision to send a force had been come to early in the previous
summer; the appointment of such a general was owing to Mary of
Modena, to whom Louis paid frequent visits. La Hoguette was made
second in command. He did not like the work or the general, but
prepared to obey. Bussy Rabutin said the exiled Queen was mad to
raise a man of so little merit as Lauzun, for himself he would always
have the meanest opinion of him, though he were given the Golden
Fleece in addition to the Garter and the Holy Ghost. Louis XIV.
realised that it was impossible to put Avaux and Lauzun together, as
they were on the worst of terms. The ambassador and Rosen were
to return with the Irish contingent in the ships which brought over
the French. Lauzun lingered in Paris as long as he decently could,
but at last followed his men to Brest, whence he sailed on St.
Patrick’s Day 1690.[249]
James was at first unwilling to have the assistance
of a French army, lest the control of the country Irish troops sent
to France.
should be taken out of his hands. But after less
than six months’ experience he despaired of doing anything without
this dreaded help, and as France could not spare men until the
continental campaign was finished, he thought of leaving Ireland.
Louis warned him that to do so would be to give up all hope of ever
regaining his crown. But jealousy of his great ally continued to
animate him. He did not like Sarsfield, whom he had promoted very
unwillingly; but when Avaux proposed to send him in charge of the
Irish going to France, he said the ambassador wished to steal all his
best officers. It was the same in Lord Kilmallock’s case, and in that
of every competent candidate. Louis refused to have any of the
Hamiltons, and the command was given to Mountcashel, who was
peculiarly fit for the work, and who, from the circumstances of his
escape from Enniskillen, could not serve again in Ireland. Very few
tolerable officers were to be had, and it was not easy to collect the
stipulated quota of privates, but in the end five strong regiments
were embarked, numbering 5300 men. Many of the officers were
shopkeepers and artisans, and they could not be refused for fear of
stopping the recruiting, but it was intended to change them in
France. As may be imagined under these circumstances, the health
and cleanliness of the rank and file were neglected, and many were
sick on arriving at Brest. Louvois gave orders to have them cared for
and to force their officers to cleanse them from the vermin by which
they were devoured. Yet these same men served gloriously in many
a continental battle.[250]
Even before his failure at Londonderry, many at the
French Court thought James’s presence in Ireland French Opinion.
did more harm than good. During the lull between
the arrival of Lauzun and the expedition of William, Advice of Louis
Madame de Sévigné reported the general opinion XIV.
that James had spoiled his own business there and earned all his
misfortunes. With a greatly superior force he was just able to check
Schomberg’s advance, and yet he talked of a descent on England or
Scotland. He hated Ireland, and lent a ready ear to secret emissaries
from both his lost kingdoms, who assured him that William was most
unpopular and that all were ready to welcome their rightful king. His
Queen received many messages to the same effect, but for some
months she did not think it would be safe for her husband to invade
England with less than 20,000 men. About the time that the two
armies were going into winter quarters, she thought it might be
attempted with any force he could command and without French
help, except at sea. He gave a list of the Irish troops which he
proposed to send. Avaux, Louvois, Vauban, and Louis XIV. all
impressed upon him that the business in hand was to make himself
master of Ireland, and the latter said he would never risk his ships in
St. George’s Channel until he had command of the sea. The
opportunity did not come till the battle of Beachy Head, but that was
on the eve of the Boyne. Quite late in the winter, when James feared
that Schomberg would be reinforced and that an English force might
land in Munster, he began to talk again about going to England.
Louis finally declared that this was not to be thought of until there
was a party under arms strong enough to resist William’s army, until
there was a fortified port ready, garrisoned and victualled, and until
all the conditions of his return were fully settled. Above all, he must
wait until the passage was made safe by a naval victory. James was
cautioned not to believe those who gave contrary advice and who
were very probably secret agents of the Prince of Orange, but he
went on talking of invading England even when William was making
his final preparations for attacking him in Ireland. We know from
Jacobite sources that the English Government was always well
informed about what happened at St. Germain.[251]
Pikes could be made in Ireland, where ash-trees
are plentiful, for 3s. 10d. apiece, but firearms and Brass money.
even swords had to be imported from France.
Textile fabrics of all sorts ran short. In the winter of 1689 some
armourers were at work in Dublin, but the supply of steel was
insufficient. Wool was abundant, and cloth for uniforms could be
produced. But the most important manufacture under James’s rule
was brass money. Needy governments have been tempted in all
ages to tamper with the currency. Gallienus, and other late Roman
emperors, carried the practice very far. Leather money was issued by
Indian princes at a very early date, with the usual result. The credit
of paper, which is the modern equivalent, depends upon the ability
of the government to make good its nominal value. In the American
Civil War, the notes of the seceding states became depreciated as
the end drew near, and when all was over were sold very cheap as
curiosities. Three months after James’s arrival in Dublin he issued a
proclamation setting forth that money was scarce, and that he
proposed to remedy this by coining sixpenny pieces out of brass or
copper. These were made legal tender, except for the payment of
duties on foreign commodities, of money held in trust, or of
judgment debts already due. Interest accruing thereafter on
mortgages, bills, bonds, or obligations might be satisfied with the
new currency, and also the principal of debts ‘where the debtor or
his goods are or shall be taken in execution of the same.’ Refusal to
accept the new coin was to be punished with the utmost rigour of
the law as contempt of the royal prerogative, but actual importers of
foreign goods were excepted for the first payment. The King
declared that the expedient was only temporary, and promised to
pay the full value in gold or silver when the base coin should be
cried down. A few days later an issue of shillings and half-crowns
was proclaimed on the same terms.[252]
The full effect of these measures in destroying
credit and paralysing trade was not felt at once. A depreciated
currency.
Within a month of the first proclamation Avaux
reported that the copper coins were everywhere taken as ready
money, and that this was a great relief to King James. The precious
metals soon disappeared from circulation. Even copper ran short,
and the ambassador applied to the French King for at least fifty tons.
Steel to make dies was also wanted, and men who could use it; for
the whole supply of money depended on a single Protestant
engraver who might go away at any moment. When French troops
were expected, it was decided, after much discussion, to pay them
in French money, and this made matters worse. Prices rose to an
undreamed-of height in anticipation of brass having to compete with
gold and silver. Cannon were converted into coin, and the total issue
ultimately reached a million or more. Every sort of rubbish was used
to make up for the want of good brass or copper. Half-crowns were
converted by re-stamping into crowns, and at last a guinea, which at
the beginning of 1689 was worth 24s. Irish, became exchangeable
for base metal to the nominal value of 5l. As in the old Greek tale,
gold made its way in spite of brass. The Protestants hoarded it or
smuggled it to England. Writing to the exiled Queen on December
12/22, Tyrconnel says, ‘Not a farthing of silver or gold is now to be
seen in this whole nation.’ All attempts to arrest the depreciation of
course failed. By one proclamation the Government undertook to
receive any quantity of currency by way of loan, to be repaid in
specie when it was decried, interest being fixed at six per cent., and
afterwards at ten; but the public liked not the security. All the
exceptions made in favour of creditors were abrogated, and brass or
pewter money was declared universal legal tender. To counterfeit it
was high treason, and to refuse it contempt of the prerogative. Only
a fortnight before the Boyne a guinea was officially rated at 38s.,
and no one was to give more on pain of death. When William gained
possession of Dublin he lost no time in crying down the base money.
The best of the crowns and half-crowns were made legal tender for
one penny, and the smaller pieces in proportion.’[253]
The military operations between Schomberg’s
retirement from Dundalk and the landing of King Fight at Newry.
William were not very important. Boisseleau made
an attempt to surprise the ruined town of Newry, Wolseley takes
but his party was beaten back by the small garrison Belturbet
consisting chiefly of sick men. Some who could not
stand managed to fire with their backs propped and Cavan.
against the walls of the roofless houses, and others shot from the
windows. Among the slain Irish was Magennis, who killed Tory
Hamilton at Down in 1686. After this Schomberg sent regular reliefs,
and Newry remained his outpost on the side of Belfast. The outpost
on the side of Enniskillen was Belturbet, which Wolseley surprised
early in December. More than two months later Berwick led an
expedition to recapture it, and concentrated a considerable force at
Cavan. Wolseley was well informed and determined on a night attack
before the whole of the enemy arrived. He had with him about a
thousand men, and the English accounts say that Berwick’s force
along with the garrison was four times as large. This is probably an
exaggeration, but the odds were certainly not less than two to one.
Wolseley was delayed on the march, and did not reach Cavan till
after daybreak. The surprise was not therefore complete, and the
assailants were met by a smart fire. The Irish retired through the
town to the fortified castle, and the Enniskilleners, who imagined
their victory complete, began to plunder in all directions. A sally
followed, and Wolseley had to set fire to the houses to get his men
out. They fell back on the reserve, and he then advanced in good
order. Berwick’s success was short-lived. His cavalry, as he tells us
himself, fled for a distance of twelve miles, and he owns to a loss of
500. Wolseley lost thirty men and two officers. Among the slain was
Brigadier Nugent, a brave soldier, much regretted by the Irish, and
many officers were taken prisoners. Berwick had a horse shot under
him. The victorious soldiers took 4000l. in brass money, but they
threw it about the streets as not worth carrying away. The Castle
was too strong to attack, and Wolseley marched back to Belturbet,
which was not molested afterwards. Soon after this Sir John Lanier
threatened Dundalk, but found it too strong to attack. A detachment
took Bellew Castle, and 1500 cattle were driven off. Schomberg had
garrisons at Clones, Monaghan, and Armagh, and his headquarters
were at Lisburn. There were frequent skirmishes along the line
between Lough Erne and Newry, but beyond it Charlemont was the
only place holding out for King James. The rest of Ireland was in his
power.[254]
Avaux and Rosen were both recalled to please James and because
neither of them could get on with Lauzun. Avaux did his best to hide
his contempt for the King, but did not quite succeed. Rosen was
scarcely civil to His Majesty and was moreover hated by every officer
in the Irish army.
Melfort was generally hated on both sides of St.
George’s Channel. In Scotland, Dundee was Intrigues in
France. Melfort.
thought to be his only friend, and was an
extremely candid one. But he had Mary of
Mission of Dover.
Modena’s ear, and he always worked against Avaux and against
Tyrconnel as head of the French party in Ireland. In the three
months preceding his journey to Fontainebleau at the beginning of
October, Louis paid the exiled Queen no less than fourteen visits at
St. Germain, and Melfort had influence in this way. Even when
recalling the unpopular favourite, the King of France rebuked Avaux
for being too hard on him. At Dublin Lord Dover tried to steer a
middle course, realising Melfort’s incompetence and working with
Tyrconnel, though he hated his French tendencies. In July, when
Londonderry was still unrelieved, James sent him on a mission to
France, and Avaux evidently feared his action there while defying
him to contradict anything he had said about the mismanagement of
affairs in Ireland. Dover was commissioned to ask for 6000 French
infantry, a considerable sum of money, a hundred thousand pounds
of powder, a train of artillery with the necessary officers, and a vast
quantity of small arms and other munitions of war. On reaching
Versailles he spoke slightingly of Tyrconnel and favourably of Melfort,
with whom he was supposed to have some understanding, but court
opinion was entirely with the former. Dover pressed Louis hard to
give all that James had asked for, but he was told that it was
impossible to do this with English and Dutch fleets at sea, but that
when December came the men should be sent and as much of the
other things wanted as could be spared. The visits to St. Germain
had done their work, and when they were resumed after the
excursion to Fontainebleau, the exiled Queen was informed to her
great joy that 6000 men were going, and that her favourite Lauzun
was to command. This had been known for some time in official
circles. That James and his wife should have been foolish enough to
wish for such a general is surprising, but that Louis should have
granted their prayer passes all understanding. Bussy Rabutin,
expressing the general opinion, says Lauzun was one of the smallest
of God’s creatures, both in body and mind. Dover was sent back to
Ireland with 2000 muskets and ammunition. He reached Kinsale
safely in December, but the vessel containing arms was captured by
the English off the Scillies. Avaux was afraid that Lauzun’s intrigues
would injure him at Versailles, but Louis reassured him on this point.
As neither he nor Rosen could serve with the new general, they
were ordered to return with the fleet that brought him to Ireland.
The King of France showed that he valued his ambassador’s services
by inviting him to all the much valued, but very uncomfortable
parties at Marly, and by sending him on a mission to Sweden. Rosen
obtained an important cavalry command.[255]
Exactly twelve months to a day after King James
Lauzun sailed from Brest and arrived in Cork Lauzun reaches
Ireland.
harbour with over 7000 French troops. One
regiment contained many Dutch Protestants, and had to be closely
watched. The general had not yet got the ducal coronet which he
had tried to stipulate for, but he wore the Garter and the Order of
the Holy Ghost. Tyrconnel warned Avaux that there would not be
horses for the officers nor carts to carry stores. It was no business of
the retiring ambassador’s and he could only warn Lord Dover, who
was responsible for embarkation and quarters. When Avaux and
Rosen were gone, Lauzun, who thought only of the King of France,
had difficulties with Dover, who thought of James as King of
England. Much merchandise had come with the fleet, and the
Frenchman sought favourable terms for the traders, while the
Englishman was chiefly anxious that his sovereign’s rights should not
be infringed. Lauzun thought, perhaps rightly, that under existing
circumstances not one guinea would reach King James out of the
duties thus insisted on, that Ireland should be regarded as a
besieged city, and that famine could only be averted by opening the
ports. This reasoning prevailed, and three weeks after his landing
Lauzun persuaded James to issue a proclamation remitting the
customs on all foreign goods except silk and tobacco. But the
difficulties about transport and storage continued. La Hoguette
thought King James improvident, but it was Lauzun’s cue to lay the
whole blame upon Dover. Cork, he said, was a tomb very hard to get
out of. As soon as Avaux and Rosen had embarked with the Irish
regiments, he and Dover went to Dublin, but the French troops could
not move for some time. Even the flour they brought with them had
to be stored in a ruined building, and half of it was washed away or
reduced to a condition in which the dough would not rise. Much of
what remained was lost in the carriage to Dublin on horses’ backs.
[256]

On February 25, 1689, Tyrconnel, having heard all

that Richard Hamilton could say, issued a The Protestants
disarmed.
proclamation for disarming Protestants. They had
to carry their weapons to their parish churches on
pain of being subjected to disorderly searches by Dr. William King.
the soldiery. Three thousand firearms besides bayonets, swords, and
pikes were seized, and horses were taken also. Both before and after
this, crowds went to England and others found their way to the
North. Many fled from their country homes to Dublin in hopes of
escaping thence or perhaps supposing that the law could protect
them there. The established clergy got away in large numbers,
Francis Marsh, Archbishop of Dublin, among them. He left Dean King
authority to act as his commissary, and the chapters of St. Patrick’s
and Christ Church submitted in spiritual matters to Dopping, Bishop
of Meath. Some other Dublin clergymen stood their ground, and with
the help of the fugitives from country districts King managed to
arrange for the duties of every parish in the diocese. When the Irish
Parliament had been prorogued just before the relief of Londonderry
and only a little before the landing of Schomberg, King was
imprisoned in the Castle. No evidence was ever produced against
him, and Sir Edward Herbert was for releasing him on bail, but
Nugent was hostile, and he remained in confinement for more than
four months. He was allowed to see his friends, and had many
visitors, Roman Catholics as well as Protestants, who kept him well
informed. The possible approach of Schomberg made his gaolers
stricter, but in November Nagle said the invading army was
mouldered to the Devil, and he saw no use in prisoners; in the
following month King was released. Even when the watch was pretty
close he mentions a venison pasty for supper, and Father Harold the
Franciscan, who helped to eat it. Lest his diary should fall into the
hands of the enemy he always entered James as King and William as
Prince of Orange. In June 1690 when the deliverer was at hand, a
state of siege was established in Dublin, and there were some 3000
Protestants in custody. Lists were made of all male Protestants from
16 to 80, any arms that still remained among them were ordered to
be given up on pain of death, none were to leave their houses from
ten at night till five in the morning, and it was a capital offence
cognisable by court-martial for more than five of them to assemble
anywhere or at any time.[257]
Vast numbers of Protestants had been leaving
Ireland ever since the death of Charles II., but Protestants in
Dublin. James
many remained because they could not get away Bonnell.
or because they had no means elsewhere. Many
placemen stood their ground, for patents could not be voided
without some process of law, and the depositaries of official
knowledge might reasonably hope to be found indispensable. Among
them was the accountant-general, James Bonnell, who took up the
active duties of his office in 1685. Clarendon, while acknowledging
him to be ‘ingenious,’ did not think him strong enough for the work,
but there were trained clerks, and he soon learned the business. He
had travelled, and saw that Versailles was sucking the life-blood of
France as clearly as Arthur Young did more than a century later. He
was a remarkably good and religious man, and his Anglican
orthodoxy is certified by many bishops, and by the fact that his
familiar friend was the Rev. John Strype. Bonnell was, nevertheless,
willing to meet the Presbyterians half-way on the question of orders.
He spent his salary and his spare time in relieving the wants of
others during the time of Tyrconnel and James II. The doctrine of
passive obedience weighed heavily with him, but he ‘could not but
secretly wish success to King William,’ and accepted the result gladly.
When Bishop Cartwright, of whom historians have little good to say,
died in Dublin in April 1689, Bonnell gives him credit for fidelity to
the Church of England, and a sort of disinterestedness—‘he was
buried decently from the Bishop of Meath’s house, and at his charge,
for he had no money.’ On July 3, 1690, Bonnell saw his fellow-
Protestants ‘congratulate and embrace one another as they met, like
persons alive from the dead.’ Later on, when Aughrim had been won
and Limerick taken, Bonnell wished to have a parliamentary union as
in Cromwell’s time and to make all English laws since Henry VII.
applicable to Ireland. By these means the English and Protestant
interests might be preserved.[258]
As the principal traders, the skilled artisans and the
officials were mostly Protestants, and as they were Refugees in
Dublin.
the chief sufferers the tradition of the Brass Money
has naturally been preserved among their
descendants. The crowd of fugitives from country Case of Trinity
College.
visitors added to the confusion. Men who had been
rich were reduced to penury, and the holders of power and influence
were either in exile or reduced to the condition of a conquered
population. As in 1641 the established clergy and laymen with
property guaranteed by the Act of Settlement were often surprised
at what happened. They found the conquered people friendly
enough in common life, and often failed to see that they were
perfectly certain to retake their own when they could, and in doing
so often to take what never belonged to them. Trinity College,
Dublin, though the fellows had escaped personal attainder, was not
spared. Under Tyrconnel no rents were paid and but one meal a day
was given in the hall, ‘and that a dinner, because the supper is the
more expensive by reason of coals, &c.’ But fourpence a day was
allowed to each fellow for kitchen and buttery. All arms and horses
were taken away. When James landed, Vice-provost Acton and his
three remaining colleagues waited on him and were promised
protection and encouragement. But six months later the college was
turned into a barrack and prison for Protestants. The government
grant to the scholars was stopped. The chapel plate—all that was
left of a rich store—was sent to the custom-house by Luttrell, but
preserved by a friendly commissioner of revenue. The chapel itself
was re-consecrated and Mass said there, but later it was made a
magazine. All the woodwork in the college was destroyed, first by
way of searching for arms, and then no doubt for fuel, of which
there was a famine in Dublin. Dr. Michael Moore, a distinguished
scholar and a man of high character, was made Provost by James,
but soon had to resign as a punishment for having preached against
the Jesuits. Another priest, Tiege MacCarthy, had charge of the
library, and is honourably distinguished for having preserved the
books and manuscripts. Provost Huntingdon and the fellows
returned immediately after the Boyne.[259]

FOOTNOTES:
[239] Schomberg’s letters to William, calendared in State Papers,
Domestic, July 21 to August 3. Journal in Kazner’s Schomberg, ii.
282. Hamilton to Melville in Leven and Melville Papers, August 1,
1689.
[240] Journal in Kazner’s Schomberg. Story’s Impartial History.
Avaux to Louvois, September 10/20. Schomberg to King William,
August 16, State Papers, Domestic. Dean’s information is in
Clarke’s Life of James, Original Mem., ii. 374. Berwick’s Memoirs.
Contemporary letter in Benn’s History of Belfast, p. 171.
[241] Story’s Impartial History, pp. 7-10. Schomberg to King
William, August 27, State Papers, Domestic. Light to the Blind.
The articles of capitulation are in Story’s Continuation and in
McSkimmin’s History of Carrickfergus, part i. Letter printed in
Benn’s History of Belfast, p. 171. Nihill’s Journal in Macpherson’s
Original Papers, i. 222. Letter of September 2 in Le Fleming
Papers.
[242] Story’s Impartial History, pp. 10-16, 38—‘A regiment of
Dutch were so well hutted that not above eleven of them died the
whole campaign.’ Schomberg letters of September 20 and 21 and
January 9, 1689-90, in State Papers, Domestic, and Dalrymple.
He says the English were ‘si delicatement élevés,’ that in all
countries he had seen them die off at the beginning of a
campaign. Early in the journal, in Kazner appx. no. 85, it is said
that the English nation ‘veut assez être conduite à son sens et
n’aime que peu la subordination quoiqu’au reste très belliqueuse,’
and under September 9 the writer says the English soldiers liked
no law but ‘leurs fantaisies.’ Writing on October 8, 1689,
Schomberg says his levies were as raw as those of King James,
but the latter twice as numerous, Leven and Melville Papers.
General Douglas’s opinion of the English soldiers is in Evelyn’s
Diary, February 19, 1689-90: they were very brave and very badly
treated. Dumont de Bostaquet, whose Mémoires inédits were
published in 1864, was with Schomberg in September 1689, and
describes the Enniskilleners as very good troops, but ‘trop
picoreurs.’
[243] Story’s Impartial Hist., pp. 17-28. Light to the Blind. Avaux
to Louis XIV., August 20/30, August 28/September 7, September
10/20, September 17/27. In the last it is mentioned that Rosen
visited the outposts at midnight and found all sentries and
vedettes asleep, ‘sans en excepter pas un.’ Same to same,
October 21/31. Nihill’s journal in Macpherson, i. 222. Memoires du
Marquis de Sourches, September 19/29. Schomberg to William
III., September 15, 20, 27, October 3, 6, State Papers, Domestic;
and Dalrymple. Lord Lisburn to Shrewsbury, September 25, ib. A
Jacobite account is in A relation of what most remarkably
happened, 1689. On October 28 Dangeau notes that Avaux had
told the French King that James’s army was in a good state,
twenty-eight battalions of 600 men, sixteen squadrons of cavalry
and ten of dragoons. He offered battle in two lines, leaving a
reserve under Sutherland. Hamilton was at the centre of the first
line with the King, Tyrconnel on the right, Rosen and Galway on
the left. Berwick was at the centre of the second line with
Sarsfield on the right. A diagram sent by Avaux is in Dangeau’s
Diary, iii. 23.
[244] Schomberg to William III., September 20-27 and November
14, State Papers, Domestic. Caillemote to Shrewsbury, September
23, ib. Story’s Impartial History, p. 25. Letter in Le Fleming
Papers, October 24. Dumont, who had fled from Normandy to
Holland to escape the dragonnades, throws light on the sincerity
of official conversions in France: he had received absolution from
a conscientious priest at Rouen, who told him to take his time
and not to go to church till he had reflected, ‘ce que j’ai executé
fort religieusement, n’ayant jamais entendu de messe ni participé
à leurs mystères.’ Luttrell, i. 613.
[245] Story’s Impartial History. Schomberg’s letters from
September to December 1689, in State Papers, Domestic. W.
Harbord to William III., October 23, ib. Newsletter of November
28, ib. Luttrell’s Diary, October and November, particularly
November 15, where it is noted that letters from Ireland report a
mortality of at least 10,000 in the Jacobite army. Evelyn’s Diary,
February 19, 1689-90. Mr. Waller’s evidence in Grey’s Debates,
November 26. During the terrible days of December 1812, after
Napoleon deserted his army, Segur testifies to the extreme
demoralisation of the survivors: ‘Tels que les sauvages, les plus
forts dépouillaient les plus faibles: ils accouraient autour des
mourants, souvent ils n’attendaient pas leurs derniers soupirs.’
Dumont lay in the Dundalk hospital for four weeks with enteric
fever and actually recovered.
[246] Schomberg’s letters, ut sup. Story’s Impartial History, vol. i.
Commons Journal, November 26, December 2 and 16. Grey’s
Debates, November 26. A defence of Shales is attempted in
Walton’s Hist. of the British Army, p. 74. Foxcroft’s Halifax, ii. 82.
On February 19, 1689-90, Evelyn met General Douglas at dinner,
who mentioned ‘the exceeding neglect of the English soldiers,
suffering severely for want of clothes and necessaries this winter,
exceedingly magnifying their courage and bravery during all their
hardships.’
[247] Story’s Impartial History, pp. 25, 34. Avaux’s narrative sent
to Seignelay on November 24/December 6, 1689. Luttrell’s Diary,
October 3, November 15. State Papers, Domestic, November 28.
Clarke’s Life of James II., ii. 383.
[248] Stevens, p. 72. Light to the Blind, p. 90. Macariæ Excidium,
p. 38. Avaux to Louis XIV., November 14/24 1689, and February
1/11, 1689-90.
[249] Louis XIV. to Avaux, May 24 and November 16, 1689. Avaux
to Louis XIV., November 24. Dangeau, January 6, 1689-90. De
Sourches, November 19, February 20. Bussy Rabutin to Madame
de Sévigné, March 23, 1689.
[250] Avaux to Louvois, October 11/21, 1689, and April 2/12,
1690. Louvois to Buridal, May 11, 1690, in Rousset, iv. 383.
Schomberg considered that MacCarthy had broken his parole, but
he was acquitted by a Court Martial in France. A sergeant whom
he had bribed was executed. The regiments that sailed were
those of MacCarthy himself, Butler, O’Brien, Fielding, and Dillon.
[251] Avaux to Louis XIV., November 14/24 and January 15/25
1689-90. Louis XIV. to Avaux, December 25/January 4. De
Sourches, April 18, 1689. Madame de Sévigné, May 31, 1690.
Lauzun to Louvois, May 10/20, in Ranke’s appendix. Letter of
Rizzini in Haile’s Mary of Modena, p. 261. Louis privately
cautioned James against trusting Albeville, who was known to be
corrupt.
[252] Clarendon to Rochester, February 8, 1685-6. Proclamations
of June 18 and 27, 1689. On September 19 Dr. King notes in his
diary that ‘the great gun which lay in Castle yard was taken away
in order to be melted and coined.’ Avaux to Louis XIV., December
12/22. On December 26/January 5 Louvois wrote to Avaux:
‘Comme le roi a veu par vos lettres que le Roy d’Angleterre
craignait de manquer de cuivre pour faire de la monnoye; Sa
Majesté a donné ordre que l’on mist sur le bastiment qui portera
cette lettre une piece de canon du calibre de deux qui est
eventée, de laquelle ceux qui travaillent à la monnoye du Roy
d’Angleterre pourront se servir pour continuer à faire de la
monnoye, en attendant que les soixante et quinze milliers de
cuivre que le Roy envoye soient arrivez.’
[253] Proclamations of February 4 and 28, March 28, April 21,
June 9 and 15, 1690; and July 10 (William III.). Avaux to Louis
XIV., July 5/15, 1689; to Louvois, June 30/July 10; to Louis XIV.,
August 20/30 and September 10/20; to Louvois, November 1/11,
November 26/December 6, 1689, and January 22/February 1,
1689-90. Light to the Blind. King’s State of the Protestants, chap.
iii. section 11. Transactions of the Late King James in Ireland,
licensed July 7, 1690, p. 57. Character of the Protestants of
Ireland, licensed November 13, 1689. This last well-written tract
has been attributed to Halifax, but neither Miss Foxcroft nor Sir
W. Raleigh mention it. Story’s Impartial History, l. 93. Lauzun to
Louvois, June 16/26, in Ranke’s appendix. King makes the total
base coinage 965,375l. Story learned from treasury officials that
‘not much above’ 1,100,000l. had been coined. The True and
Perfect Journal, 1690, states the amount at about two millions.
Tyrconnel’s letter is in Haile’s Mary of Modena, p. 258.
[254] Captain Kennedy to the Scotch Council, December 12, in
Leven and Melville Papers. Story’s Impartial History, November to
February, 1689-90. The author of Light to the Blind says the
attack on Newry was a mere reconnaissance and that there was
no repulse. Schomberg says Boisseleau was there, State Papers,
Domestic, December 6. As to the action at Cavan, besides the
above and Berwick’s memoirs, there are accounts in State Papers,
Domestic, particularly Schomberg’s letter, February 19, and that
of Gustavus Hamilton, ambiguously calendared under March 21,
1689 (Addenda, p. 571).
[255] Melfort’s unpopularity is sufficiently shown by Dundee’s
letters to him, June 27 and 28, Napier, iii. 599. Notices in
Dangeau and De Sourches. Avaux’s letters, particularly that of
July 16/26, enclosing James’s requirements, Louvois to Avaux,
September 7/17. Madame de Sévigné marvelled greatly at
Lauzun’s ‘second volume.’ The reference to her letters and to
Bussy Rabutin’s concerning him are collected in the Grands
Ecrivains edition of La Bruyère, i. 335, 535, where he is
characterised under the name of Straton. Madame de Caylus in
her memoirs notes the good luck of Lauzun in being in England at
the critical time, gaining honour and glory for helping William by
assisting the flight of James.
[256] Lauzun to Seignelay, April 6/16, in appendix to Ranke’s
History and to Louvois, ib. June 16. Proclamation of March
25/April 4. It was known at the French Court that Lauzun was
‘extrêmement ulceré avec raison’ against Dover, De Sourches,
April 24/May 4. Compare the extracts in Miss Sandars’s Lauzun.
True and Perfect Journal, June 16.
[257] Simon Luttrell’s orders as Governor of Dublin, May 3 and
June 18, 1690, in appendix to King’s State of the Protestants,
nos. 30 and 31. Besides King’s principal book on this subject we
have his autobiography, the original Latin printed in English
Historical Review, vol. xiii., an English version in King’s A great
Archbishop of Dublin, and his diary edited by Dr. Lawlor in the
Irish Journal of Archæology, 1903.
[258] Archdeacon Hamilton’s Life of Bonnell, 3rd edition, 1707,
particularly pp. 60, 273. Bonnell to Strype, August 20, 1684,
January 21, and April 17, 1689, and August 5, 1690, in English
Historical Review, xix. 122, 299. Clarendon and Rochester Corr., i.
245, 266. Cartwright was buried in Christ Church with a full choral
service, all the principal people in Dublin attending, Athenæ
Oxonienses, p. 831. Bonnell to Harty, Portland Papers, November
3, 1691.
[259] College register for 1689-90 printed in Stubbs’s Hist. of the
University of Dublin, pp. 127-133. Harris’s Ware, ii. 288. King, iii.
15.
 CHAPTER LIV

WILLIAM III. IN IRELAND, 1690. THE BOYNE

Lauzun and Dover were in Dublin together early in

April, and continued to quarrel there. The The French
contingent. Dover
Englishman made light of the French contingent, and Lauzun.
saying that Louis was plainly deceiving King James,
who would be well advised to make terms with the Prince of Orange.
Uncle and nephew might then join their forces to those of the
Augsburg allies and attack the tyrant of Europe. The old courtier
proposed to go to William and make terms for himself, but James
could not countenance this, though willing to give him a pass for
Flanders, since he could not venture into France. In the end he was
allowed to live and die unmolested in England. As for Lauzun, he
had no hopes of successfully resisting the Prince of Orange, and
proposed to burn Dublin and destroy the country entirely while
retreating from point to point, but James thought this policy too
cruel. In the meantime the French general exerted himself in the
work of arming and organising the Irish, and in this he made
considerable progress. He could not speak or understand English,
and his attendance at the Council was waste of time, so he proposed
to do business with the King and Tyrconnel. The three accordingly
met daily, and Lauzun succeeded in making friends with the Lord
Lieutenant, who had been cautioned by Avaux not to trust him lest
he should usurp all power, seeing that he had already ruined his
career by vaingloriousness, and was not likely to be much changed
for the better. But he assured the French minister that he was a
chastened man and worked with a single eye to the interest and
wishes of his own King.[260]
While Lauzun and Tyrconnel tried to make up for
lost time amid the dissipations of Dublin, Siege of
Charlemont.
Schomberg was growing stronger every day by the
arrival of fresh troops from England and Scotland,
including 6000 Danish veterans under the Duke of Attempt
it.
to relieve

Würtemberg. Long before William left London the

old general saw that a stand would probably be
Fall of
made at the Boyne, and he was anxious to take Charlemont.
Charlemont, so that no enemy should be left in the Teague O’Regan.
rear. It was James’s last stronghold in Ulster, and
Mountjoy had chosen the position well. The castle, which stood on
the right or Armagh bank of the Blackwater, a few miles above
Lough Neagh, had been fortified in modern fashion, and was well
armed and manned. The town or village had been levelled, and the
fort was nearly surrounded by bogs and fields subject to flooding. It
was considered unassailable, except by placing batteries on the left
bank of the river, and Schomberg, who reconnoitred the place,
thought it too strong to attack with the means then at his command.
In March Colonel La Caillemote brought up a small force in boats to
stop the garrison from making incursions into Tyrone. He set fire to
the bridge, and drove the Irish out of two small outworks. Paul
Rapin, the historian, was wounded in this skirmish, and it is much to
be regretted that we have no account by him. As his force increased,
Schomberg massed troops all round Charlemont. Nevertheless, at
the beginning of May Colonel Macmahon, who held Castleblaney,
managed to elude the post at Armagh and brought 500 men, well
armed but badly clothed, with provisions and ammunition, to the
blockaded fort. Having got within the lines, they were quite unable
to break out again, and had to encamp miserably between the inner
and outer works, for the governor would not have them inside. This
relief only hastened the end, for men could not carry much food
through bogs and hills, and there were so many additional mouths.
At last starvation-point was reached, and Schomberg was glad to
have the place surrendered without a formal siege. The garrison
marched out with all the honours of war, and made their way to
Dundalk. As they passed, it was noticed that many were chewing
pieces of hide with the hair on. They left nineteen pieces of
ordnance behind them, but nothing eatable. Teague O’Regan himself
was a grotesque figure, with worn-out clothes and draggled wig. He
had been drinking brandy—and it naturally affected the head of a
half-starved man. His charger, a vicious old screw, would scarcely
allow him to salute Schomberg, who remarked that Teague’s horse
was very mad and himself very drunk. But William met no braver
enemy, and he afterwards defended Sligo with the same courage
and tenacity. The victorious general ordered bread to be distributed
among the vanquished. About 800 marched out, with 200 women
and children. When Schomberg was told that the Irish would not
stay in garrison without their wives and mistresses, he said there
was more love than policy in it. Story himself saw papers in the late
governor’s room which showed that he had information as to what
was going on outside. James very rightly knighted O’Regan as soon
as he reached Dublin.[261]
It was known at the beginning of 1690 that King
William had resolved to go to Ireland in person. King William and
Ireland.
There was strong opposition on the part of the
Whigs, who argued that there were too many active Jacobites in
England for the sovereign to leave it safely. Better to lose Ireland
than England, said some. Nor would he be safe himself, for his
courage led him into danger, in which he furnished a strong contrast
to the King of France. He was reminded of Richard II.’s fate and of
his own insecure position. ‘When any one at meat,’ said Delamere,
‘has unnecessarily risen from his chair to reach over to the other
side of the table, if by design or chance his stool has been removed,
who, suspecting no such thing, his breech has found the ground
instead of his chair—there has been more in the company who have
been pleased with it, than concerned for him.’ An address against
the Irish voyage was contemplated in both Houses, and might have
passed had not William prorogued and afterwards dissolved the
Convention Parliament. The general election was favourable to him,
and preparations began in earnest. The Commons did not give all
that the King wished, but they provided money enough for the
immediate purpose. Harbord was superseded for a time, and the
duties which Shales had neglected were committed to others. Both
Houses adjourned on May 23, and did not meet again for business
until after William’s return from Ireland. The Government was left in
Mary’s hands with a special council of four Whigs and five Tories.
[262]

The King, accompanied by Portland, set out from

London on June 4, and slept that night at William reaches
Ireland.
Northampton. On Sunday, the 8th, he attended
service in Chester Cathedral, and heard a sermon
from Dr. Stratford, who had succeeded Cartwright He marches
towards Dublin.
in that see. On the 12th he took ship at Hoylake
and arrived with 300 sail at Carrickfergus on the He Maintains
14th. An eye-witness says that the total number of discipline.
vessels assembled was 700, and that Belfast Lough
looked like a wood. William mounted his horse as soon as possible
and rode amid cheering crowds through the town on his way to
Belfast. At Whitehouse, Schomberg met him with his coach, and
they drove together; a second carriage was sent by the General to
bring up some of the grandees who had landed. At the north gate of
the town the illustrious visitor was met by the Corporation in their
robes, accompanied by Dr. Walker and a dozen other clergymen. All
the way to the castle there were shouts of ‘God save King William,’
‘God save our Protestant King.’ At night the streets and all the
country round blazed with bonfires. They were seen, and the signal
guns heard by one of Lauzun’s spies, who brought him the news two
days later. Next day being Sunday, William heard Dr. Royse preach in
the Cathedral on ‘Who through faith subdued kingdoms,’ and on the
Monday received an address from the clergy, with Walker at their
head. Good order was kept, and necessaries were cheap, for the
ships brought vast quantities of provisions, and even of hay and
straw. ‘We fear no more Dundalk wants,’ says one letter, and the
army was thoroughly well provided; but of money there was no
great plenty. William spent four days at Belfast, reviewing the troops
and making arrangements. Sick of inaction and not fully paid,
officers and soldiers longed for active service, and were not
disappointed. On the 19th William dined with Schomberg at Lisburn,
having previously issued a proclamation against plundering or taking
goods without payment, and on the next day he was at
Hillsborough, spurring those in authority under him to fresh efforts.
He had not, he said, come to let the grass grow under his feet. Lest
there should be any doubt about the meaning of his proclamation,
he here issued a special order against pressing horses belonging to
the country people without permission under the sign manual, which
was afterwards refused even for ambulance purposes. A soldier
transgressing this order was to run the gauntlet thrice through the
whole regiment. A few months before Schomberg had rather made
light of seizing the little country horses. On the 22nd William was at
Loughbrickland, and by the 27th the whole army, mustering about
36,000 men, encamped a little to the south of Dundalk. During the
whole campaign the King and Prince George of Denmark lived each
in a wooden hut designed by Sir Christopher Wren, and capable of
being carried on two wagons. When William inspected his troops he
was not satisfied with seeing them march past from a comfortable
eminence, but went in among the ranks, regardless of heat, wind,
and clouds of dust. When a fuss was made about the wine for his
table, he said he would drink water rather than that the men should
suffer. He was deficient in courtly graces, but he was the kind of king
whom soldiers will follow cheerfully against any odds.[263]
Before making a general advance, William took
care to have the line of march thoroughly Skirmish near
Newry, June 22.
reconnoitred. At a boggy spot about half-way
between Newry and Dundalk, where there was a broken bridge, a
party of 200 foot and dragoons fell into an ambuscade on the day
that the King reached Loughbrickland. Lauzun takes credit for having
laid the snare, and he had reason to know the place, for his horse
had fallen under him there only two days before. The morning was
foggy and the surprise complete. Captain Farlow, who led the
infantry detachment, was taken prisoner with several others, and
Colonel Dempsy, who commanded the Irish, was mortally wounded.
There was a sharp skirmish, and the English were decidedly worsted,
but not pursued. From Farlow James had the first certain news of
William’s landing.[264]
Two days after King William’s landing, King James
left Dublin to join his army near Dundalk. They James leaves
Dublin, June 16.
were encamped about Roche Castle, and the
prisoners taken with Captain Farlow reported that
He falls back
William was on the road to Newry with 50,000 without fighting.
men, which was an exaggeration. On the day after
the skirmish there was a general retreat to the old position at Ardee,
where entrenchments had been left unfinished the year before.
James’s main object in advancing had been to exhaust the country
through which his rival would have to march, but William, with the
sea open, was in no want of supplies. The guns of the English fleet
could be heard by both armies. The difficult ground about Moyry and
Ravensdale had been the scene of much fighting in Elizabethan
times, and had been slightly fortified by James, who was blamed for
not trying to stop the invader there; but Berwick says that, with the
force at his disposal, William could easily have turned the position
from the Armagh side. Dundalk itself, though well fortified, was
judged to be untenable, and Lauzun evacuated it five days before
the final struggle. He abstained from burning the soldiers’ huts
because some of last year’s infection still hung about them and
might do the enemy more harm than want of shelter at midsummer.
But both Dundalk and Ardee were thoroughly sacked by the Irish.
On June 28, twelve days after leaving Dublin, James recrossed the
Boyne, half of his army marching through Drogheda and the other
half over the ford at Oldbridge, where entrenchments were begun
but not finished, owing to the want of labour. Lord Iveagh was
Governor of Drogheda, with 1300 men, and had he been an
enterprising man he might have done much to cover the Jacobite
right. The left wing, extending up the river, was evidently open to a
flank attack, but James rightly says that the country afforded no
better position. Sarsfield’s division, which had been detached to
guard against a possible attack on Athlone, joined the main body on
June 26, their leader having satisfied himself that all the troops
about Cavan and Belturbet had drawn towards Armagh, so as to fall
in with William’s line of march.[265]
On June 27 William’s army was encamped a little to
the south of Dundalk. He intended to attack the William’s march to
the Boyne.
enemy at Ardee, but a party of cavalry found that
position already abandoned. On the 30th the whole
army marched towards the Boyne, the King himself He is wounded,
June 30.
diverging a little to the left so as to view Drogheda
and the course of the river from the hill at Tullyesker. Schomberg
was with him, and also Prince George, the Duke of Ormonde, Sidney,
Solms, and Scravenmore. The latter, who had seen many armies,
remarked that James’s was a small one, but William said there might
be more in the town and behind the hills. A deserter said they were
about 25,000, and the Williamite chaplain admits that his King had
some 36,000. The line of march was through a deep depression,
where a modern road runs to the east of Townley Hall, which is still
known, and will always be known, as King William’s Glen. Thomas
Bellingham, an officer who had connections in the country, took the
opportunity of paying Mr. Townley a visit. About noon the head of
the column came out into the open, and took up ground facing
Oldbridge on the other side of the Boyne. William sat down to eat
and rest a little higher up. A party of five officers, of whom Berwick,
Tyrconnel, and Sarsfield were three, were observed riding slowly
along the opposite bank, and shortly afterwards two field guns were
quietly brought up and fired as soon as William was in the saddle
again. A six-pound shot ricochetted and struck him on the right
shoulder, tearing his coat and breaking the skin. He merely remarked
in Dutch that it was near enough. Thomas Coningsby, afterwards an
earl, applied a handkerchief to the bleeding wound, and William
made light of it, retiring to a tent to have it dressed and then
remounting. He remained on horseback for three hours without
changing his coat, and laughed at one Dr. Sangrado who proposed
to bleed him. The enemy, says Captain Parker, ‘concluded he was
killed, and this news soon flew to Dublin and from thence to Paris,
where they had public rejoicings for it.’ About three o’clock his
artillery came up, and both shot and shells from small mortars were
sent across the river, doing some mischief, but without altering the
situation. At sunset there was a council of war, and Schomberg
advised that a strong force should be sent up the river at midnight,
so that James’s army might be taken in flank and rear and cut off
from Dublin. William, however, who was supported by Solms and
other Dutch officers, decided upon a frontal attack, somewhat to the
veteran’s disgust. Many of the Enniskillen officers knew all the fords,
and with their help the order for next day’s battle was arranged. At
twelve o’clock William rode by torchlight through the whole army. He
was a man who kept his own counsel, but his unwillingness to take
Schomberg’s advice and perhaps gain a victory as complete as Ulm
or Sedan may fairly be ascribed to his dread of catching James. As
at Rochester, a means of escape was provided, and experience had
shown that it would not be neglected. The necessity of sparing
Mary’s feelings and the political danger of a captive king might well
prevail against purely military considerations.[266]
Whatever William may have said or thought at the
evening council, he did not entirely reject the idea Battle of the
Boyne, July 1.
of a flank movement. Very early in the morning of
July 1 he despatched Meinhart Schomberg with a strong body of
horse and foot and five guns to cross at the bridge of Slane. They
marched by the straight road, leaving the bend of the river far to
their left. Sir Neill O’Neill with his dragoons were sent to guard this
pass, and the bridge itself had been broken down, but there had
been several very hot days, and the river, not being affected by the
tide above Oldbridge, was fordable in many places. Schomberg’s
men crossed with ease, partly near Slane and partly at Rossnaree
lower down, the dragoons were beaten back, and O’Neill himself
mortally wounded. This was at about half-past nine. Warned by the
trumpets and drums of Schomberg’s force, Lauzun had already
begun to extend to his left, and when he saw what had happened
developed this movement gradually in order to secure the means of
retreat. Seeing that Schomberg’s party was in danger, William sent
Lieut.-General Douglas with a much larger force of infantry to his
aid. A bog prevented the hostile wings from coming to close
quarters, but Lauzun gained Duleek, which commanded the Dublin
road. In the meantime the passage at Oldbridge, where Richard
Hamilton commanded with eight battalions, had been forced, and
soon after noon most of the Irish infantry were in full flight, nor is
this to be wondered at, for less than a week before, many of them
had not learned how to fire their pieces. The baggage had been sent
off at daybreak; the tents and knapsacks became the prey of the
victors. Stevens saw the hills covered with fugitives running past like
sheep before a wolf. ‘The shame of our regiment,’ he says, ‘only
afflicted me before, but now all the horror of a routed army, just
before so vigorous and desirous of battle, and broke without scarce
a stroke of the enemy, so perplexed my soul that I envied the few
dead.’[267]
Schomberg was over the Boyne before the left and
centre of William’s army began to move, but at a Victory of William.
quarter past ten the Blue Dutch Guards, eight or
ten deep, entered the water opposite the unfinished works at
Oldbridge, their drums beating until they reached the bank. They
were up to their waists, and crossed under a heavy but ineffectual
fire, reserving their own until they reached dry land. The first to
climb the bank was a lieutenant who formed up the leading files,
and then crouched down for them to shoot over his head. The Irish
foot abandoned the first ditch, but their cavalry, under Berwick’s
command, charged the Dutch furiously before they were fully in
order. They stood firm against this and several other attacks,
gradually pressing the Irish infantry backwards, and in the meantime
the French and Enniskillen foot passed the river a little farther down,
several English and Danish regiments still lower. The tide was rising,
so that some of the men were up to their arm-pits, and on the
extreme left, horses had to swim. Some of the Danish infantry
carried their guns over their heads, but others fired steadily as they
waded over. William was looking on, and said he had never seen
anything better done. They were at once attacked by the Irish
cavalry, and there was hard fighting for half an hour. A regiment of
French Huguenots was broken by a charge, and Colonel La
Cailemotte was carried off the field mortally wounded, but still
encouraging his men, ‘A la gloire, mes enfants, à la gloire!’ Seeing
his friends in difficulties, Schomberg crossed himself, reminding
them that their persecutors were before them. He fell, shot through
the neck, and with sabre wounds on the head. Dr. Walker, the still
unconsecrated Bishop of Derry, was killed soon afterwards, and his
brother clergyman Story, offers as an excuse for his presence, that
he was going to look at the wounded general. Walker, says the
chaplain, was stripped at once, ‘for the Scots-Irish that followed our
camp were got through already, and took off most of the plunder.’
When the news of Schomberg’s death was brought to William, he
laid his finger on his lips, and lost no time in passing the river
himself with the left wing of his cavalry, Dutch and Dane chiefly, with
Wolseley’s Enniskilleners and Cutts’s English regiment. His right arm
was stiff from yesterday’s wound, and he carried a stick only. He was
unable to bear his cuirass, and when he drew his sword later, had to
hold it in his left hand. He crossed where the little Drybridge stream
enters the Boyne, but his horse stuck fast in the boggy ground
beyond the river, and he had to dismount before it could be
extricated. He was at once engaged in the thickest of the fight, and
a bullet which struck the heel of his boot killed a horse close by. He
put himself at the head of the Enniskillen cavalry, saying, ‘What will
you do for me?’ Owing possibly to a mistake, the Enniskilleners were
driven back for a short distance, and then William led on his steady
Dutch. The Enniskilleners soon recovered themselves, and the Irish
foot were pressed backwards, but the cavalry for the most part
fought bravely, making repeated and often successful charges, but
being gradually overborne by the disciplined troops opposed to
them. Lord Dungan was killed early in the fight, and his dragoons
would do nothing afterwards. Lord Clare’s yellow dragoons also ran
away, and some of them never stopped until they got far beyond the
Shannon. The broken troops rode right through the retreating foot
as if they had been enemies. But Tyrconnel’s and Parker’s regiments
of horse performed prodigies of valour. The latter was wounded, and
Sheldon, who commanded the former corps, had two horses killed
under him. Berwick’s was shot, and rolled over his rider. Hamilton,
who headed the last charge, was wounded and taken prisoner near
Plattin House, which stands two miles back from the river. William
said he was very glad to see him, and asked if the cavalry would
make any more fight. ‘Upon my honour,’ said the prisoner, ‘I believe
they will.’ ‘Your honour!’ said the King; and that was his only
revenge. Hamilton was sent to the Tower as a prisoner of war, and
was exchanged for Lord Mountjoy in the spring of 1692. Neither of
them saw Ireland again, and Mountjoy, whom William made Master
of the Ordnance, was killed at Steenkirk soon after his release from
the Bastille. He had had enough of passive obedience. There was no
more fighting, but the Irish cavalry rallied to protect the retreat with
the unbroken French contingent. The flying infantry threw away their
arms, and even their boots, and not many were overtaken, though
little quarter was given. The loss of the victors was about one-third
as great. The pursuit continued as far as Naul, when the light began
to fail. Drogheda surrendered the next day, the garrison marching
out without arms, rather than undergo the horrors of an assault. The
terms offered were pretty much the same as Cromwell’s, forty years
before, and the memories attaching to his siege were not favourable
to resistance.[268]
On the fatal morning King James posted himself
near the church at Donore, whence he could see Flight of James.
both armies. He took no part in the battle, and as
appears from his own account was chiefly concerned lest his retreat
should be cut off. As soon as the danger seemed imminent he drew
off to the left and joined Lauzun, who strongly advised him to take
care of himself. He needed but little pressing, and with four troops of
horse and four of dragoons he passed Duleek first and led the way
back to Dublin. The French kept their ranks and prevented the
victors from pressing too hard upon the routed army. Berwick
reached Duleek about the same time as William himself, and had to
gallop hard to avoid being intercepted. Lauzun and Tyrconnel kept
together. The loss in James’s army was perhaps 1500, that in
William’s about 500. To compare the conduct of the two Kings, it
need only be said that one led the advance and the other the
retreat.[269]
From the military point of view, the battle of the
Boyne is not interesting, and French writers dismiss Importance of the
battle.
it as a skirmish, in which Marshal Schomberg
happened to be killed. With a much superior force, both in numbers
and quality, William forced the passage of a small river which was
fordable in many places. The importance of the action lies in its
international character, and its political effect was enormous in
checking the overweening ambition of France. There have been
other occasions on which very small battles have decided very great
causes. At Valmy the forces engaged were greater than at the
Boyne, but the number of casualties was less than one-half, and yet
the effect is felt to this day. At Calatafimi the killed and wounded
altogether were only about 400, but that fight went far to change
the map of Europe. The great French victory at Fleurus and the
great English disaster off Beachy Head were both neutralised on the
banks of the Boyne. Lauzun’s despatch is dated sixteen days after
the battle, and it was a fortnight later that the full news reached
Louis XIV. But King James had arrived at Brest, with the tidings of
his own defeat, laying all the blame on the Irish, and giving faint
praise even to the cavalry who had fought so well. Soon after this it
was known that the Prince of Orange had been hit, and confidently
reported that he was dead. Without any encouragement from the
authorities, the Parisians abandoned themselves to rejoicing. How
much the French feared William, said Bolingbroke, ‘appeared in the
extravagant and indecent joy they expressed on a false report of his
death.’ The citizens dined in the streets, casks of wine were
broached, there were bonfires and fireworks everywhere. Effigies of
William were cast on dunghills, thrown into the Seine, or broken on
the wheel. First President Harlay and Advocate-General Talon had to
drink the King’s health, and Bossuet, though he protested that he
was on his way to say mass, was forced to do the same. Police
officers sent to suppress the unauthorised rejoicings had to drink
with the rest. Even in the inner court at Versailles the guards could
hardly prevent the people from lighting a fire. The excitement spread
to remote villages, and was not allayed for weeks. Even after the
middle of August the Abbé de Choisy made a bet with La Fontaine
that the Prince was dead, staking the price of the poet’s works
against the books themselves. The report reached Modena, but with
the puzzling addition that James was at St. Germain. Then the truth
came in English and Dutch papers. At Rome, too, the event was long
uncertain. Melfort at first heard that the Prince of Orange was killed,
and he enlarged on King James’s opportunity. This was the time to
take the power of the purse from Parliament, to repeal the Habeas
Corpus Act, and to abolish trial by jury in cases of treason. If an
amnesty was found necessary, the list of exceptions should be as
long as possible, and not one of those excepted should ever be
pardoned on any consideration. Alexander VIII., who thought more
of enriching his family than of rescuing England, was horrified that
Te Deums should be sung in Austrian cathedrals for William’s victory;
but he had no money to spare, and could not venture to go against
the general sense of European sovereigns. Even the French, though
they would have welcomed the death of their great antagonist, had
very little sympathy with his dethroned rival.[270]
In the morning of July 1 Dublin was full of rumours
that a battle was imminent. The gates were closely State of Dublin.
guarded, and Protestants kept their houses. Every
hour brought a fresh report. The French were in Dublin Bay. An
express from Waterford had announced that the Isle of Wight was in
French hands, and the victors going to Dover. The English right wing
on the Boyne had been completely routed. But at five in the
afternoon stragglers arrived on tired horses, who said the Irish had
the worst, and an hour later others declared that the rout was
complete. ‘Till one that night all the entries of the town were filled
with dusty, wounded, and tired soldiers and carriages perpetually
coming in.’ A little before ten King James arrived with 200 horse in
disorder, and was received by Lady Tyrconnel at the Castle gate. He
was followed two hours later by the bulk of the Irish cavalry in good
order, ‘with kettledrums, hautboys, and trumpets.’ Next morning
came the French with all their guns and many of the Irish foot. But
the King was already gone. He saw some of his Council—Herbert,
Fitton, the Duke of Powis, Price, Nagle, and Albeville being among
them—and asked whether the news of the battle of Fleurus was not
a reason for going to France. He seems to have thought that Louis
would seize the opportunity of invading England while William was
away. His advisers urged him to run no risk of capture, since the
victorious enemy might appear in the morning. At midnight a
message came from Berwick to say that he had rallied some of the
fugitives and asking for cavalry. His father sent a few troops that had
not been in the battle, but the gathering soon dispersed. Tyrconnel
sent his chaplain to advise His Majesty to lose no time, and to send
all the troops to meet him and Lauzun at Leixlip. La Hoguette and
other superior officers appeared in Dublin without their men, which
was explained by a mistaken order having been given to meet
Lauzun at Dunboyne. At five in the morning, after a few hours’ rest,
James sent for the Mayor and made a speech to him and others
present. Everything, he said, was against him. In England he had an
army that would have fought if they had not proved false; in Ireland
his soldiers were loyal enough, but would not stand by him. He had
now to seek safety for himself, and advised his hearers to do the
same. They were not to wreak present vengeance on the
Protestants, nor to injure a city in which he still had an interest. He
then took horse for Bray, ordering Simon Luttrell to evacuate the
town and to do no mischief. La Hoguette and the other French
officers asked for horses, but he had none to give them, and they
were left to follow as best they could. Brigadier Wauchop was posted
near the north end of Dublin to turn the stream of fugitives towards
Limerick. Luttrell was the last man to leave his post, and by sunset
the Castle was in sole charge of Captain Farlow, who had been a
close prisoner since the skirmish near Ravensdale.[271]
Louvois had strictly charged Lauzun not to attempt any dazzling
exploit, but to devote himself entirely to gain time and to prolong
the war. From the slavish way in which he addressed the great
minister, belittling himself and claiming no merit but in strict
obedience, it is evident that Lauzun distrusted his own powers. He
had no belief in the cause for which he was fighting, and his main
objects were to get King James safely back to his wife and to restore
to King Louis his money, his guns, and as many of his soldiers as
possible. Above all things he longed to get out of Ireland himself.
The glory of defending Limerick was left to Boisseleau, the credit of
keeping the French troops together after the retreat from the Boyne
chiefly belongs to the Swiss Colonel Zurlauben and to a captain
named La Pujade, of whom little else seems to be known. John and
Anthony Hamilton as well as Tyrconnel accompanied Lauzun in the
retreat to Limerick. La Hoguette and several other field officers
seemed only anxious to get to the sea. During the battle the only
French officer of rank killed was the Marquis d’Hoquincourt, who
commanded an Irish battalion. Finding that his men would not
stand, he charged alone and fell covered with wounds. Lauzun
certainly gained no glory, and was quite unfit for the task in hand,
but he maintained order during the retreat on the day of the Boyne,
and the rear was then the post of honour. Long imprisonment may
have shaken his nerves, but it seems hard to call him a coward, as
Rousset has done, and he is more fairly to be judged by what he
wrote to Louvois from Galway shortly before sailing for France:—‘The
bad state of affairs and my small capacity will cause me to make
many mistakes, but I beg you to excuse me to His Majesty; and at
least I can assure you that death would be sweeter than what I
suffer here.’[272]
King James had been most careful to provide for his own escape.
More than a week before the battle he sent Sir Patrick Trant to
prepare a ship at Waterford, and on the day after it he rode hard in
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