Download the full version of the textbook now at textbookfull.

com

Learning SQL Master SQL Fundamentals Alan

Beaulieu

https://textbookfull.com/product/learning-sql-
master-sql-fundamentals-alan-beaulieu/

Explore and download more textbook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

T SQL Fundamentals Itzik Ben-Gan

https://textbookfull.com/product/t-sql-fundamentals-itzik-ben-gan/

textbookfull.com

SQL Easy SQL Programming Database Management for Beginners

Your Step By Step Guide to Learning the SQL Database Felix
Alvaro
https://textbookfull.com/product/sql-easy-sql-programming-database-
management-for-beginners-your-step-by-step-guide-to-learning-the-sql-
database-felix-alvaro/
textbookfull.com

SQL Primer: An Accelerated Introduction to SQL Basics

Rahul Batra

https://textbookfull.com/product/sql-primer-an-accelerated-
introduction-to-sql-basics-rahul-batra/

textbookfull.com

A Formula for Eradicating Racism: Debunking White

Supremacy 1st Edition Timothy Mcgettigan

https://textbookfull.com/product/a-formula-for-eradicating-racism-
debunking-white-supremacy-1st-edition-timothy-mcgettigan/

textbookfull.com
Pentecostal Politics in a Secular World: The Life and
Leadership of Lewi Pethrus Joel Halldorf

https://textbookfull.com/product/pentecostal-politics-in-a-secular-
world-the-life-and-leadership-of-lewi-pethrus-joel-halldorf/

textbookfull.com

An impossible dream racial integration in the United

States 1st Edition Sharon A. Stanley

https://textbookfull.com/product/an-impossible-dream-racial-
integration-in-the-united-states-1st-edition-sharon-a-stanley/

textbookfull.com

Beginning Data Science in R: Data Analysis, Visualization,

and Modelling for the Data Scientist 1st Edition Thomas
Mailund
https://textbookfull.com/product/beginning-data-science-in-r-data-
analysis-visualization-and-modelling-for-the-data-scientist-1st-
edition-thomas-mailund/
textbookfull.com

Female Urinary Tract Infections in Clinical Practice Bob

Yang

https://textbookfull.com/product/female-urinary-tract-infections-in-
clinical-practice-bob-yang/

textbookfull.com

Contexts for Music Learning and Participation: Developing

and Sustaining Musical Possible Selves Andrea Creech

https://textbookfull.com/product/contexts-for-music-learning-and-
participation-developing-and-sustaining-musical-possible-selves-
andrea-creech/
textbookfull.com
An Introduction to Clouds From the Microscale to Climate
1st Edition Ulrike Lohmann

https://textbookfull.com/product/an-introduction-to-clouds-from-the-
microscale-to-climate-1st-edition-ulrike-lohmann/

textbookfull.com
1. Preface
a. Why Learn SQL?
b. Why Use This Book to Do It?
c. Structure of This Book
d. Conventions Used in This Book
e. Using Code Examples
f. O’Reilly Online Learning
g. How to Contact Us
2. 1. A Little Background
a. Introduction to Databases

i. Nonrelational Database Systems

ii. The Relational Model
iii. Some Terminology
b. What Is SQL?

i. SQL Statement Classes

ii. SQL: A Nonprocedural Language
iii. SQL Examples

c. What Is MySQL?
d. SQL Unplugged
e. What’s in Store

3. 2. Creating and Populating a Database

 a. Creating a MySQL Database
b. Using the mysql Command-Line Tool
c. MySQL Data Types
i. Character Data
ii. Numeric Data
iii. Temporal Data
d. Table Creation

i. Step 1: Design
ii. Step 2: Refinement
iii. Step 3: Building SQL Schema
Statements
e. Populating and Modifying Tables

i. Inserting Data
ii. Updating Data
iii. Deleting Data
f. When Good Statements Go Bad

i. Nonunique Primary Key

ii. Nonexistent Foreign Key
iii. Column Value Violations
iv. Invalid Date Conversions

g. The Sakila Database

4. 3. Query Primer
 a. Query Mechanics
b. Query Clauses
c. The select Clause

i. Column Aliases
ii. Removing Duplicates

d. The from Clause

i. Tables
ii. Table Links
iii. Defining Table Aliases

e. The where Clause

f. The group by and having Clauses
g. The order by Clause

i. Ascending Versus Descending Sort

Order
ii. Sorting via Numeric Placeholders

h. Test Your Knowledge

i. Exercise 3-1
ii. Exercise 3-2
iii. Exercise 3-3
iv. Exercise 3-4
5. 4. Filtering
a. Condition Evaluation
 i. Using Parentheses
ii. Using the not Operator
b. Building a Condition
c. Condition Types

i. Equality Conditions
ii. Range Conditions
iii. Membership Conditions
iv. Matching Conditions
d. Null: That Four-Letter Word
e. Test Your Knowledge

i. Exercise 4-1
ii. Exercise 4-2
iii. Exercise 4-3
iv. Exercise 4-4
6. 5. Querying Multiple Tables

a. What Is a Join?
i. Cartesian Product
ii. Inner Joins
iii. The ANSI Join Syntax

b. Joining Three or More Tables

i. Using Subqueries As Tables
ii. Using the Same Table Twice
 c. Self-Joins
d. Test Your Knowledge
i. Exercise 5-1
ii. Exercise 5-2
iii. Exercise 5-3
7. 6. Working with Sets

a. Set Theory Primer

b. Set Theory in Practice
c. Set Operators

i. The union Operator

ii. The intersect Operator
iii. The except Operator
d. Set Operation Rules

i. Sorting Compound Query Results

ii. Set Operation Precedence
e. Test Your Knowledge

i. Exercise 6-1
ii. Exercise 6-2
iii. Exercise 6-3
8. 7. Data Generation, Manipulation, and Conversion

a. Working with String Data

i. String Generation
 ii. String Manipulation
b. Working with Numeric Data
i. Performing Arithmetic Functions
ii. Controlling Number Precision
iii. Handling Signed Data

c. Working with Temporal Data

i. Dealing with Time Zones
ii. Generating Temporal Data
iii. Manipulating Temporal Data

d. Conversion Functions
e. Test Your Knowledge
i. Exercise 7-1
ii. Exercise 7-2
iii. Exercise 7-3
9. 8. Grouping and Aggregates
a. Grouping Concepts
b. Aggregate Functions

i. Implicit Versus Explicit Groups

ii. Counting Distinct Values
iii. Using Expressions
iv. How Nulls Are Handled

c. Generating Groups
 i. Single-Column Grouping
ii. Multicolumn Grouping
iii. Grouping via Expressions
iv. Generating Rollups
d. Group Filter Conditions
e. Test Your Knowledge

i. Exercise 8-1
ii. Exercise 8-2
iii. Exercise 8-3
10. 9. Subqueries

a. What Is a Subquery?
b. Subquery Types
c. Noncorrelated Subqueries

i. Multiple-Row, Single-Column
Subqueries
ii. Multicolumn Subqueries

d. Correlated Subqueries

i. The exists Operator

ii. Data Manipulation Using Correlated
Subqueries

e. When to Use Subqueries

i. Subqueries As Data Sources
ii. Subqueries As Expression Generators
 f. Subquery Wrap-up
g. Test Your Knowledge

i. Exercise 9-1
ii. Exercise 9-2
iii. Exercise 9-3
11. 10. Joins Revisited

a. Outer Joins

i. Left Versus Right Outer Joins

ii. Three-Way Outer Joins
b. Cross Joins
c. Natural Joins
d. Test Your Knowledge
i. Exercise 10-1
ii. Exercise 10-2
iii. Exercise 10-3 (Extra Credit)
12. 11. Conditional Logic

a. What Is Conditional Logic?

b. The Case Expression

i. Searched Case Expressions

ii. Simple Case Expressions
c. Case Expression Examples

i. Result Set Transformations

 ii. Checking for Existence
iii. Division-by-Zero Errors
iv. Conditional Updates
v. Handling Null Values
d. Test Your Knowledge

i. Exercise 11-1
ii. Exercise 11-2
13. 12. Transactions

a. Multiuser Databases
i. Locking
ii. Lock Granularities

b. What Is a Transaction?

i. Starting a Transaction
ii. Ending a Transaction
iii. Transaction Savepoints

c. Test Your Knowledge

i. Exercise 12-1

14. 13. Indexes and Constraints

a. Indexes

i. Index Creation
ii. Types of Indexes
iii. How Indexes Are Used
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
 iv. The Downside of Indexes
b. Constraints

i. Constraint Creation

c. Test Your Knowledge

i. Exercise 13-1
ii. Exercise 13-2
15. 14. Views
a. What Are Views?
b. Why Use Views?

i. Data Security
ii. Data Aggregation
iii. Hiding Complexity
iv. Joining Partitioned Data

c. Updatable Views

i. Updating Simple Views

ii. Updating Complex Views

d. Test Your Knowledge

i. Exercise 14-1
ii. Exercise 14-2

16. 15. Metadata

a. Data About Data

b. Information_Schema
 c. Working with Metadata

i. Schema Generation Scripts

ii. Deployment Verification
iii. Dynamic SQL Generation
d. Test Your Knowledge

i. Exercise 15-1
ii. Exercise 15-2
17. 16. Analytic Functions

a. Analytic Function Concepts

i. Data Windows
ii. Localized Sorting

b. Ranking
i. Ranking Functions
ii. Generating Multiple Rankings

c. Reporting Functions

i. Window Frames
ii. Lag and Lead

d. Test Your Knowledge

i. Exercise 16-1
ii. Exercise 16-2
iii. Exercise 16-3

18. 17. Working with Large Databases

 a. Partitioning

i. Partitioning Concepts
ii. Table Partitioning
iii. Index Partitioning
iv. Partitioning Methods
v. Partitioning Benefits

b. Sharding
c. Big Data

i. Hadoop
ii. NoSQL and Document Databases
iii. Cloud Computing
iv. Future of SQL
19. 18. SQL and Big Data

a. Apache Drill
b. Drill and MySQL
c. Drill and MongoDB
d. Drill with Multiple Data Sources
 Learning SQL
THIRD EDITION

Generate, Manipulate, and Retrieve Data

With Early Release ebooks, you get books in their earliest form—the author’s
raw and unedited content as they write—so you can take advantage of these
technologies long before the official release of these titles.

Alan Beaulieu
Learning SQL

by Alan Beaulieu

Copyright © 2020 Alan Beaulieu. All rights reserved.

Printed in the United States of America.

Published by O’Reilly Media, Inc., 1005 Gravenstein Highway

North, Sebastopol, CA 95472.

O’Reilly books may be purchased for educational, business, or

sales promotional use. Online editions are also available for
most titles (http://oreilly.com/safari). For more information,
contact our corporate/institutional sales department: 800-998-
9938 or corporate@oreilly.com.

Acquisitions Editor: Jessica Haberman

Development Editor: Jeff Bleiel

Production Editor: Deborah Baker

Interior Designer: David Futato

Cover Designer: Karen Montgomery

Illustrator: Rebecca Demarest

May 2020: Third Edition

Revision History for the Early Release
2019-12-11: First Release

See http://oreilly.com/catalog/errata.csp?isbn=9781492057611
for release details.

The O’Reilly logo is a registered trademark of O’Reilly Media,

Inc. Learning SQL, the cover image, and related trade dress
are trademarks of O’Reilly Media, Inc.

The views expressed in this work are those of the author, and
do not represent the publisher’s views. While the publisher and
the author have used good faith efforts to ensure that the
information and instructions contained in this work are
accurate, the publisher and the author disclaim all
responsibility for errors or omissions, including without
limitation responsibility for damages resulting from the use of
or reliance on this work. Use of the information and
instructions contained in this work is at your own risk. If any
code samples or other technology this work contains or
describes is subject to open source licenses or the intellectual
property rights of others, it is your responsibility to ensure that
your use thereof complies with such licenses and/or rights.

978-1-492-05754-3

[LSI]
Preface
Programming languages come and go constantly, and very few
languages in use today have roots going back more than a
decade or so. Some examples are Cobol, which is still used
quite heavily in mainframe environments, and C, which is still
quite popular for operating system and server development
and for embedded systems. In the database arena, we have
SQL, whose roots go all the way back to the 1970s.

SQL is the language for generating, manipulating, and

retrieving data from a relational database. One of the reasons
for the popularity of relational databases is that properly
designed relational databases can handle huge amounts of
data. When working with large data sets, SQL is akin to one of
those snazzy digital cameras with the high-power zoom lens in
that you can use SQL to look at large sets of data, or you can
zoom in on individual rows (or anywhere in between). Other
database management systems tend to break down under
heavy loads because their focus is too narrow (the zoom lens is
stuck on maximum), which is why attempts to dethrone
relational databases and SQL have largely failed. Therefore,
even though SQL is an old language, it is going to be around
for a lot longer and has a bright future in store.
Why Learn SQL?
If you are going to work with a relational database, whether
you are writing applications, performing administrative tasks,
or generating reports, you will need to know how to interact
with the data in your database. Even if you are using a tool
that generates SQL for you, such as a reporting tool, there may
be times when you need to bypass the automatic generation
feature and write your own SQL statements.

Learning SQL has the added benefit of forcing you to confront

and understand the data structures used to store information
about your organization. As you become comfortable with the
tables in your database, you may find yourself proposing
modifications or additions to your database schema.
Why Use This Book to Do It?
The SQL language is broken into several categories.
Statements used to create database objects (tables, indexes,
constraints, etc.) are collectively known as SQL schema
statements. The statements used to create, manipulate, and
retrieve the data stored in a database are known as the SQL
data statements. If you are an administrator, you will be using
both SQL schema and SQL data statements. If you are a
programmer or report writer, you may only need to use (or be
allowed to use) SQL data statements. While this book
demonstrates many of the SQL schema statements, the main
focus of this book is on programming features.

With only a handful of commands, the SQL data statements

look deceptively simple. In my opinion, many of the available
SQL books help to foster this notion by only skimming the
surface of what is possible with the language. However, if you
are going to work with SQL, it behooves you to understand
fully the capabilities of the language and how different features
can be combined to produce powerful results. I feel that this is
the only book that provides detailed coverage of the SQL
language without the added benefit of doubling as a “door
stop” (you know, those 1,250-page “complete references” that
tend to gather dust on people’s cubicle shelves).

While the examples in this book run on MySQL, Oracle

Database, and SQL Server, I had to pick one of those products
to host my sample database and to format the result sets
returned by the example queries. Of the three, I chose MySQL
because it is freely obtainable, easy to install, and simple to
administer. For those readers using a different server, I ask
that you download and install MySQL and load the sample
database so that you can run the examples and experiment
with the data.

Structure of This Book

This book is divided into 15 chapters and 3 appendixes:

Chapter 1, explores the history of computerized

databases, including the rise of the relational model
and the SQL language.

Chapter 2, demonstrates how to create a MySQL

database, create the tables used for the examples in
this book, and populate the tables with data.

Chapter 3, introduces the selectstatement and

further demonstrates the most common clauses
(select, from, where).
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
Chapter 4, demonstrates the different types of
conditions that can be used in the whereclause of a
select, update, or deletestatement.

Chapter 5, shows how queries can utilize multiple tables

via table joins.

Chapter 6, is all about data sets and how they can

interact within queries.

Chapter 7, demonstrates several built-in functions used

for manipulating or converting data.

Chapter 8, shows how data can be aggregated.

Chapter 9, introduces the subquery (a personal

favorite) and shows how and where they can be
utilized.
Chapter 10, further explores the various types of table
joins.

Chapter 11, explores how conditional logic (i.e., if-then-

else) can be utilized in select, insert, update, and
deletestatements.

Chapter 12, introduces transactions and shows how to

use them.

Chapter 13, explores indexes and constraints.

Chapter 14, shows how to build an interface to shield

users from data complexities.

Chapter 15, demonstrates the utility of the data

dictionary.

Appendix A shows the database schema used for all

 examples in the book.

Appendix B demonstrates some of the interesting non-

ANSI features of MySQL’s SQL implementation.

Appendix C shows solutions to the chapter exercises.

Conventions Used in This Book

The following typographical conventions are used in this book:

Italic

Used for filenames, directory names, and URLs. Also used

for emphasis and to indicate the first use of a technical
term.

Constant width

Used for code examples and to indicate SQL keywords

within text.

Constant width italic

Used to indicate user-defined terms.

plainUPPERCASE
 Used to indicate SQL keywords within example code.

Constant width bold

Indicates user input in examples showing an interaction.

Also indicates emphasized code elements to which you
should pay particular attention.

NOTE
Indicates a tip, suggestion, or general note. For example, I use notes to
point you to useful new features in Oracle9i.

WARNING
Indicates a warning or caution. For example, I’ll tell you if a certain SQL
clause might have unintended consequences if not used carefully.

Using Code Examples

This book is here to help you get your job done. In general, if
example code is offered with this book, you may use it in your
programs and documentation. You do not need to contact us
for permission unless you’re reproducing a significant portion
of the code. For example, writing a program that uses several
chunks of code from this book does not require permission.
Selling or distributing examples from O’Reilly books does
require permission. Answering a question by citing this book
and quoting example code does not require permission.
Incorporating a significant amount of example code from this
book into your product’s documentation does require
permission.

We appreciate, but generally do not require, attribution. An

attribution usually includes the title, author, publisher, and
ISBN. For example: “Learning SQL, Third Edition, by Alan
Beaulieu. Copyright 2020 Alan Beaulieu, 978-1-492-05761-1.”

If you feel your use of code examples falls outside fair use or
the permission given above, feel free to contact us at
permissions@oreilly.com.

O’Reilly Online Learning

NOTE
For more than 40 years, O’Reilly has provided technology and business
training, knowledge, and insight to help companies succeed.

Our unique network of experts and innovators share their

knowledge and expertise through books, articles, conferences,
and our online learning platform. O’Reilly’s online learning
platform gives you on-demand access to live training courses,
in-depth learning paths, interactive coding environments, and a
vast collection of text and video from O’Reilly and 200+ other
Random documents with unrelated
content Scribd suggests to you:
upon the enormity of the task which I have undertaken, with
inadequate powers myself to do anything, and without assistance of
any kind from the Spaniards[1].... I am apprehensive that all this may
turn out but ill for the Spanish cause. If, for any cause, I should be
overpowered, or should be obliged to retreat, what will the world
say? What will the people of England say?... That we made a great
effort, attended by some glorious circumstances; that from January
1st, 1812, we had gained more advantages for the cause, and had
acquired more extent of territory by our operations than any army
ever gained in such a period of time against so powerful an enemy;
but that unaided by the Spanish army and government, we were
finally overpowered, and compelled to withdraw within our old
frontier.’
It was with no light heart that Wellington faced the strategical
problem. In outline it stood as follows. Soult was now known to be
evacuating Andalusia, and it was practically certain that he would
retire on Valencia, where he would join King Joseph and Suchet.
Their three armies would produce a mass of veteran troops so great
that even if every division of the Anglo-Portuguese army were
concentrated, if Hill came up from Estremadura and Skerrett’s small
force from Cadiz, it would be eminently doubtful whether Madrid
could be held and the enemy thrust back. The French might advance
85,000 strong, and Wellington could only rely on 60,000 men of his
own to face them—though he might scrape together three or four
divisions of Spanish troops in addition[2]. It was true that Suchet
might probably refuse to evacuate his Valencian viceroyalty, and that
occupation for him might be found by utilizing Maitland’s
expeditionary force at Alicante, and Elio’s Murcian army. But even if
he did not join Soult and King Joseph in a march on Madrid, the
armies of the South and Centre might put 65,000 men into the field.
But this was only half the problem. There was Clausel’s Army of
Portugal, not to speak of Caffarelli’s Army of the North, to be taken
into consideration. Clausel had some 40,000 men behind the Douro
—troops recently beaten it is true, and known to be in bad order. But
they had not been pursued since the Allied army turned aside for the
march on Madrid, and had now been granted a month in which to
pull themselves together. Nothing had been left in front of them save
Clinton’s 6th Division at Cuellar, and a division of the Galicians at
Valladolid. And now the vexatious news had come that Clausel was
on the move, had chased the Galicians out of Valladolid, and was
sending flying columns into the plains of Leon. It was clear that he
must be dealt with at once, and there was no way to stop his
annoying activity, save by detaching a considerable force from
Madrid. If unopposed, he might overrun all the reconquered lands
along the Douro, and even imperil the British line of communication
with Salamanca and Portugal. If, as was possible, Caffarelli should
lend him a couple of divisions from the Army of the North, he might
become a real danger instead of a mere nuisance.
This was the reason why Wellington departed from Madrid on
August 31st, and marched with the 1st, 5th and 7th Divisions, Pack’s
and Bradford’s Portuguese, and Bock’s and Ponsonby’s dragoons—
21,000 sabres and bayonets—to join Clinton, and thrust back
Clausel to the North, before he should have leisure to do further
mischief. There was, as he conceived, just time enough to inflict a
sharp check on the Army of Portugal before the danger from the side
of Valencia would become pressing. It must be pushed out of the
way, disabled again if possible, and then he would return to Madrid
for the greater game, leaving as small a containing force as possible
in front of the Northern army. He summed up his plan in a
confidential letter in the following terms: ‘All the world [the French
world] seems to be intending to mass itself in Valencia; while I am
waiting for their plans to develop, for General Hill to march up from
Estremadura, and for the Spanish armies to get together, I shall hunt
away the elements of Marmont’s [i. e. Clausel’s] army from the
Douro. I shall push them as far off as I can, I shall try to establish
proper co-operation between the Anglo-Portuguese detachment
which I must leave on this side and the Galician Army, and so I shall
assure my left flank, when I shall be engaged on the Valencian
side[3].’
Here then, we have a time-problem set. Will it be possible to deal
handsomely with the Army of Portugal, to put it completely out of
power to do harm, before Soult shall have reached Valencia,
reorganized his army, and joined King Joseph in what Wellington
considered the inevitable scheme of a march on Madrid? The British
general judged that there would be sufficient time—and probably
there might have been, if everything had worked out in the best
possible way. But he was quite conscious that events might prove
perverse—and he shuddered at the thought—as he wrote to his
brother in Cadiz.
The last precautions taken before departing for the Douro were to
draw up three sets of instructions. One was for Charles Alten, left in
command of the four divisions which remained in and about
Madrid[4], foreseeing the chance of Soult’s marching on the capital
without turning aside to Valencia—‘not at all probable, but it is
necessary to provide for all events.’ The second was for Hill, who
was due to arrive at Toledo in about three weeks. The third was for
General Maitland at Alicante, whose position would obviously be
very unpleasant, now that Soult was known to be evacuating
Andalusia and marching on Valencia to join Suchet and King Joseph.
Soult’s march altered the whole situation on the East Coast: if
50,000 more French were concentrated in that direction, the Alicante
force must be in some danger, and would have to observe great
caution, and if necessary to shut itself up in the maritime fortresses.
‘As the allied forces in Valencia and Murcia’—wrote Wellington to
Maitland—‘will necessarily be thrown upon the defensive for a
moment, while the enemy will be in great strength in those parts, I
conclude that the greater part of those forces will be collected in
Alicante, and it would be desirable to strengthen our posts at
Cartagena during this crisis, which I hope will be only momentary[5].’
Another dispatch, dated four days later, adverts to the possibility that
Soult may fall upon Alicante on his arrival in the kingdom of Valencia.
Maitland is to defend the place, but to take care that all precautions
as to the embarking his troops in the event of ill-success are made.
He is expected to maintain it as long as possible; and with the sea
open to him, and a safe harbour, the defence should be long and
stubborn, however great the numbers of the besiegers[6].
With these instructions drawn out for Maitland, Wellington finally
marched for the North. His conception of the situation in Valencia
seems to have been that Soult, on his arrival, would not meddle with
Alicante: it was more probable that he and King Joseph would rather
take up the much more important task of endeavouring to reconquer
Madrid and New Castile. They would leave Suchet behind them to
contain Maitland, Elio and Ballasteros. It would be impossible for him
to lend them any troops for a march on Madrid, while such a large
expeditionary force was watching his flank. The net result would be
that ‘by keeping this detachment at Alicante, with Whittingham’s and
Roche’s Spaniards, I shall prevent too many of the gentlemen now
assembled in Valencia from troubling me in the Upper Country [New
Castile][7].’
Travelling with his usual celerity, Wellington left Madrid on August
31st, was at Villa Castin on the northern side of the Guadarrama
pass on September 2nd, and had reached Arevalo, where he joined
Clinton and the 6th Division on September 3rd. The divisions which
he was bringing up from Madrid had been started off some days
before he himself left the capital. He passed them between the
Escurial and Villa Castin, but they caught him up again at Arevalo
early on the 4th, so that he had his fighting force concentrated on
that day, and was prepared to deal with Clausel.
The situation of affairs on the Douro requires a word of
explanation. Clausel, it will be remembered, had retired from
Valladolid on July 30, unpursued. He was prepared to retreat for any
length—even as far as Burgos—if he were pressed. But no one
followed him save Julian Sanchez’s lancers and some patrols of
Anson’s Light Cavalry brigade. Wherefore he halted his main body
on the line of the Arlanza, with two divisions at Torquemada and two
at Lerma; some way to his left Foy, with two divisions more, was at
Aranda, on the Upper Douro, where he had maintained his forward
position, because not even a cavalry patrol from the side of the Allies
had come forward to disquiet him. In front of Clausel himself there
was soon nothing left but the lightest of cavalry screens, for Julian
Sanchez was called off by Wellington to New Castile: when he was
gone, there remained Marquinez’s guerrilleros at Palencia, and
outposts of the 16th Light Dragoons at Valtanas: the main body of G.
Anson’s squadrons lay at Villavanez, twenty miles behind[8]. The only
infantry force which the Allies had north of the Douro was one of the
divisions of the Army of Galicia, which (with Santocildes himself in
command) came up to Valladolid on August 6th, not much over
3,000 bayonets strong. The second of the Galician divisions which
had descended into the plain of Leon was now blockading Toro. The
third and most numerous was still engaged in the interminable siege
of Astorga, which showed at last some signs of drawing to its close
—not because the battering of the place had been effective, but
simply because the garrison was growing famished. They had been
provisioned only as far as August 1, and after that date had been
forced upon half and then quarter rations. But the news of the
disaster of Salamanca had not, as many had hoped among the
Allies, scared their commander into capitulation.
Clausel on August 1 had not supposed it possible that he would
be tempted to take the offensive again within a fortnight. His army
was in a most dilapidated condition, and he was prepared to give
way whenever pressed. It was not that his numbers were so very
low, for on August 1st the Army of Portugal only counted 10,000 less
effectives than on July 15th. Though it had lost some 14,000 men in
the Salamanca campaign, it had picked up some 4,000 others from
the dépôt at Valladolid, from the many small garrisons which it had
drawn in during its retreat, and from drafts found at Burgos.[9] Its loss
in cavalry had been fully repaired by the arrival of Chauvel’s two
regiments from the Army of the North, and most of the fugitives and
marauders who had been scattered over the countryside after the
battle of July 22nd gradually drifted back to their colours. It was not
so much numbers as spirit that was wanting in the Army of Portugal.
On August 6 Clausel wrote to the Minister of War at Paris that he
had halted in a position where he could feed the troops, give them
some days of repose, and above all re-establish their morale[10]. ‘I
must punish some of the men, who are breaking out in the most
frightful outrages, and so frighten the others by an example of
severity; above all I must put an end to a desire, which they display
too manifestly, to recross the Ebro, and get back nearer to the
French frontier. It is usual to see an army disheartened after a check:
but it would be hard to find one whose discouragement is greater
than that of these troops: and I cannot, and ought not, to conceal
from you that there has been prevailing among them for some time a
very bad spirit. Disorders and revolting excesses have marked every
step of our retreat. I shall employ all the means in my power to
transform the dispositions of the soldiers, and to put an end to the
deplorable actions which daily take place under the very eyes of
officers of all grades—actions which the latter fail to repress.’
Clausel was as good as his word, and made many and severe
examples, shooting (so he says) as many as fifty soldiers found
guilty of murders, assaults on officers, and other excesses[11]. It is
probable, however, that it was not so much his strong punitive
measures which brought about an improved discipline in the
regiments, as the fortnight of absolutely undisturbed repose which
they enjoyed from the 1st to the 14th of August. The feeling of
demoralization caused by the headlong and disorderly retreat from
Salamanca died down, as it became more and more certain that the
pursuit was over, and that there was no serious hostile force left
within many miles of the line of the Arlanza. The obsession of being
hunted by superior forces had been the ruinous thing: when this
terror was withdrawn, and when the more shattered regiments had
been re-formed into a smaller number of battalions, and provided in
this fashion with their proper proportion of officers[12], the troops
began to realize that they still formed a considerable army, and that
they had been routed, but not absolutely put out of action, by the
Salamanca disaster.
Yet their morale had been seriously shaken; there was still a want
of officers; a number of the rejoining fugitives had come in without
arms or equipment; while others had been heard of, but had not yet
reported themselves at their regimental head-quarters. It therefore
required considerable hardihood on Clausel’s part to try an offensive
move, even against a skeleton enemy. His object was primarily to
bring pressure upon the allied rear, in order to relieve King Joseph
from Wellington’s attentions. He had heard that the whole of the
Anglo-Portuguese army, save a negligible remnant, had marched on
Madrid; but he was not sure that the Army of the Centre might not
make some endeavour to save the capital, especially if it had been
reinforced from Estremadura by Drouet. Clausel knew that the King
had repeatedly called for succours from the South; it was possible
that they might have been sent at the last moment. If they had come
up, and if Wellington could be induced to send back two or three
divisions to the Douro, Madrid might yet be saved. The experiment
was worth risking, but it must take the form of a demonstration rather
than a genuine attack upon Wellington’s rear. The Army of Portugal
was still too fragile an instrument to be applied to heavy work.
Indeed, when he moved, Clausel left many shattered regiments
behind, and only brought 25,000 men to the front.
There was a second object in Clausel’s advance: he hoped to
save the garrisons of Astorga, Toro, and Zamora, which amounted in
all to over 3,000 men. Each of the first two was being besieged by a
Spanish division, the third by Silveira’s Portuguese militia. The
French general judged that none of the investing forces was equal to
a fight in the open with a strong flying column of his best troops.
Even if all three could get together, he doubted if they dared face two
French divisions. His scheme was to march on Valladolid with his
main body, and drive out of it the small Spanish force in possession,
while Foy—his senior division-commander—should move rapidly
across country with some 8,000 men, and relieve by a circular
sweep first Toro, then Astorga, then Zamora. The garrisons were to
be brought off, the places blown up: it was useless to dream of
holding them, for Wellington would probably come to the Douro
again in force, the sieges would recommence, and no second relief
would be possible in face of the main British army.
On August 13th a strong French cavalry reconnaissance crossed
the Arlanza and drove in the guerrilleros from Zevico: on the
following day infantry was coming up from the rear, pushing forward
on the high-road from Torquemada to Valladolid. Thereupon Anson,
on that evening, sent back the main body of his light dragoons
beyond the Douro, leaving only two squadrons as a rearguard at
Villavanez. The Galician division in Valladolid also retired by the road
of Torrelobaton and Castronuevo on Benavente. Santocildes—to
Wellington’s disgust when it reached his ears—abandoned in
Valladolid not only 400 French convalescents in hospital, but many
hundred stand of small arms, which had been collected there from
the prisoners taken during the retreat of Clausel in the preceding
month. This was inexcusable carelessness, as there was ample time
to destroy them, if not to carry them off. French infantry entered
Valladolid on the 14th and 15th, apparently about 12,000 strong:
their cavalry, a day’s march ahead, had explored the line of the
Douro from Simancas to Tudela, and found it watched by G. Anson’s
pickets all along that front.
The orders left behind by Wellington on August 5th, had been
that if Clausel came forward—which he had not thought likely—G.
Anson was to fall back to the Douro, and if pressed again to join
Clinton’s division at Cuellar. The retreat of both of them was to be on
Segovia in the event of absolute necessity. On the other hand, if the
French should try to raise the sieges of Astorga or Zamora,
Santocildes and Silveira were to go behind the Esla[13]. Clausel tried
both these moves at once, for on the same day that he entered
Valladolid he had turned off Foy with two divisions—his own and
Taupin’s (late Ferey’s)—and a brigade of Curto’s chasseurs, to
march on Toro by the road through Torrelobaton. The troops in
Valladolid served to cover this movement: they showed a division of
infantry and 800 horse in front of Tudela on the 16th, and pushed
back Anson’s light dragoons to Montemayor, a few miles beyond the
Douro. But this was only a demonstration: their cavalry retired in the
evening, and Anson reoccupied Tudela on the 20th. It soon became
clear that Clausel was not about to cross the Douro in force, and was
only showing troops in front of Valladolid in order to keep Anson and
Clinton anxious. He remained there with some 12,000 or 15,000 men
from August 14th till September 7th, keeping very quiet, and making
no second attempt to reconnoitre beyond the Douro. Anson,
therefore, watched the line of the river, with his head-quarters at
Valdestillas, for all that time, while the guerrilleros of Saornil and
Principe went over to the northern bank and hung around Clausel’s
flank. Clinton, on his own initiative, moved his division from Cuellar
to Arevalo, which placed him on the line of the road to Madrid via the
Guadarrama, instead of that by Segovia. This Wellington afterwards
declared to be a mistake: he had intended to keep the 6th Division
more to the right, apparently in order that it might cross the Douro
above Valladolid if required[14]. It should not have moved farther
southward than Olmedo.
Foy meanwhile, thus covered by Clausel, made a march of
surprising celerity. On the 17th he arrived at Toro, and learned that
the Galicians blockading that place had cleared off as early as the
15th, and had taken the road for Benavente. He blew up the fort, and
took on with him the garrison of 800 men. At Toro he was much
nearer to Zamora than to Astorga, but he resolved to march first on
the remoter place—it was known to be hard pressed, and the French
force there blockaded was double that in Zamora. He therefore
moved on Benavente with all possible speed, and crossed the Esla
there, driving away a detachment of the Galician division (Cabrera’s)
which had come from Valladolid, when it made an ineffectual attempt
to hold the fords. On the 20th August he reached La Baneza, some
sixteen miles from Astorga, and there received the tiresome news
that the garrison had surrendered only thirty-six hours before to
Castaños. The three battalions there, worn down by famine to 1,200
men, had capitulated, because they had no suspicion that any help
was near. The Spanish general had succeeded in concealing from
them all knowledge of Foy’s march. They laid down their arms on the
18th, and were at once marched off to Galicia: Castaños himself
accompanied them with all his force, being fully determined not to
fight Foy, even though his numbers were superior.
The French cavalry pushed on to Astorga on the 21st, and found
the gates open and the place empty, save for seventy sick of the late
garrison, who had been left behind under the charge of a surgeon. It
was useless to think of pursuing Castaños, who had now two days’
start, wherefore Foy turned his attention to Zamora. Here Silveira,
though warned to make off when the enemy reached Toro, had held
on to the last moment, thinking that he was safe when Foy swerved
away toward Astorga. He only drew off when he got news on the
22nd, from Sir Howard Douglas, the British Commissioner with the
Galician army, to the effect that Foy, having failed to save Astorga,
was marching against him. He retired to Carvajales behind the Esla,
but was not safe there, for the French general, turning west from
Benavente, had executed a forced march for Tabara, and was
hastening westward to cut in between Carvajales and the road to
Miranda de Douro on the Portuguese frontier. Warned only just in
time of this move, Silveira hurried off towards his own country, and
was within one mile of its border when Foy’s advanced cavalry came
up with his rearguard near Constantin, the last village in Spain. They
captured his baggage and some stragglers, but made no serious
endeavour to charge his infantry, which escaped unharmed to
Miranda. Foy attributed this failure to Curto, the commander of the
light horse: ‘le défaut de décision et l’inertie coupable du général
commandant la cavalerie font perdre les fruits d’une opération bien
combinée’ (August 23rd)[15]. This pursuit had drawn Foy very far
westward; his column turning back, only reached Zamora on August
26th: here he drew off the garrison and destroyed the works. He
states that his next move would have been a raid on Salamanca,
where lay not only the British base hospital, but a vast accumulation
of stores, unprotected by any troops whatever. But he received at
Zamora, on the 27th, urgent orders from Clausel to return to
Valladolid, as Wellington was coming up against him from Madrid
with his whole army. Accordingly Foy abandoned his plan, and
reached Tordesillas, with his troops in a very exhausted condition
from hard marching, on August 28th. Clausel had miscalculated
dates—warned of the first start of the 1st and 7th Divisions from
Madrid, he had supposed that they would be at Arevalo some days
before they actually reached it on September 4th. Foy was never in
any real danger, and there had been time to spare. His excursion
undoubtedly raised the spirit of the Army of Portugal: it was
comforting to find that the whole of the Galicians would not face
8,000 French troops, and that the plains of Leon could be overrun
without opposition by such a small force. On his return to join the
main body Foy noted in his diary that Clausel had been very inert in
face of Anson and Clinton, and wrote that he himself would have
tried a more dashing policy—and might possibly have failed in it,
owing to the discouragement and apathy still prevailing among many
of the senior officers of the Army of Portugal[16].
Wellington had received the news of Clausel’s advance on
Valladolid as early as August 18th, and was little moved by it. Indeed
he expressed some pleasure at the fact. ‘I think,’ he wrote to Lord
Bathurst, ‘that the French mean to carry off the garrisons from
Zamora and Toro, which I hope they will effect, as otherwise I must
go and take them. If I do not, nobody else will, as is evident from
what has been passing for the last two months at Astorga[17].’ He
expressed his pleasure on hearing that Santocildes had retired
behind the Esla without fighting, for he had feared that he might try
to stop Foy and get beaten. It was only, as we have already seen,
after he obtained practical certainty that Soult had evacuated
Andalusia, so that no expedition to the South would be necessary,
that he turned his mind to Clausel‘s doings. And his march to
Arevalo and Valladolid was intended to be a mere excursion for a
few weeks, preparatory to a return to New Castile to face Soult and
King Joseph, when they should become dangerous. From Arevalo
he wrote to Castaños to say that ‘it was necessary to drive off
Marmont’s (i.e. Clausel’s) army without loss of time, so as to make it
possible to turn the whole of his forces eventually against Soult.’ He
should press the movement so far forward as he could; perhaps he
might even lay siege to Burgos. But the Army of Galicia must come
eastward again without delay, and link up its operations with the
Anglo-Portuguese. He hoped to have retaken Valladolid by
September 6th, and wished to see Castaños there, with the largest
possible force that he could gather, on that date[18]. The Galician
army had returned to Astorga on August 27th, and so far as
distances went, there was nothing to prevent it from being at
Valladolid eleven days after, if it took the obvious route by Benavente
and Villalpando.
The troops from Madrid having joined Clinton and the 6th Division
at Arevalo, Wellington had there some 28,000 men collected for the
discomfiture of Clausel, a force not much more numerous than that
which the French general had concentrated behind the Douro, for
Foy was now back at Tordesillas, and the whole Army of Portugal
was in hand, save certain depleted and disorganized regiments
which had been left behind in the province of Burgos. But having the
advantage of confidence, and knowing that his enemy must still be
suffering from the moral effects of Salamanca, the British general
pushed on at once, not waiting for the arrival on the scene of the
Galicians. On the 4th the army marched to Olmedo, on the 5th to
Valdestillas, on the 6th to Boecillo, from whence it advanced to the
Douro and crossed it by various fords between Tudela and Puente
de Duero—the main body taking that of Herrera. The French made
no attempt to defend the line of the river, but—rather to Wellington’s
surprise—were found drawn up as if for battle a few miles beyond it,
their right wing holding the city of Valladolid, whose outskirts had
been put in a state of defence, their left extending to the ground
about the village of La Cisterniga. No attempt was made to dislodge
them with the first divisions that came up: Wellington preferred to
wait for his artillery and his reserves; the process of filing across the
fords had been tedious, and occupied the whole afternoon.
On this Clausel had calculated: he was only showing a front in
order to give his train time to move to the rear, and to allow his right
wing (Foy) at Simancas and Tordesillas to get away. A prompt
evacuation of Valladolid would have exposed it to be cut off. On the
following morning the French had disappeared from La Cisterniga,
but Valladolid was discovered to be still held by an infantry
rearguard. This, when pushed, retired and blew up the bridge over
the Pisuerga on the opposite side of the city, before the British
cavalry could seize it. The critics thought that Clausel might have
been hustled with advantage, both on the 6th and on the morning of
the 7th, and that Foy might have been cut off from the main body by
rapid action on the first day[19]. Wellington’s cautious movements
may probably be explained by the fact that an attack on Clausel on
the 6th would have involved fighting among the suburbs and houses
of Valladolid, which would have been costly, and he had no wish to
lose men at a moment when the battalion-strengths were very low all
through the army. Moreover, the capture of Valladolid would not have
intercepted the retreat of the French at Simancas, but only have
forced them to retire by parallel roads northward. At the same time it
must be owned that any loss of life involved in giving Clausel a
thorough beating on this day would have been justified later on. He
had only some 15,000 men in line, not having his right-wing troops
with him that day. If the Army of Portugal had been once more
attacked and scattered, it would not have been able to interfere in
the siege of Burgos, where Wellington was, during the next few
weeks, to lose as many men as a general action would have cost.
But this no prophet could have foreseen on September 6th.
It cannot be said that Wellington’s pursuit of Clausel was pressed
with any earnestness. On the 8th his advanced cavalry were no
farther forward than Cabezon, seven miles in front of Valladolid, and
it was not till the 9th that the 6th Division, leading the infantry,
passed that same point. Wellington himself, with the main body of
his infantry, remained at Valladolid till the 10th. His dispatches give
no further explanation for this delay than that the troops which had
come from Madrid were fatigued, and sickly, from long travel in the
hot weather, and that he wished to have assurance of the near
approach of the Army of Galicia, which was unaccountably slow in
moving forward from Astorga[20]. Meanwhile he took the opportunity
of his stay in Valladolid to command, and be present at, a solemn
proclamation of the March Constitution. This was a prudent act; for
though the ceremony provoked no enthusiasm whatever in the
city[21], where there were few Liberals in existence, it was a useful
demonstration against calumnies current in Cadiz, to the effect that
he so much disliked the Constitution that he was conspiring with the
serviles, and especially with Castaños, to ignore or even to
overthrow it.
While Wellington halted at Valladolid, Clausel had established
himself at Dueñas, fifteen miles up the Pisuerga. He retired from
thence, however, on the 10th, when the 6th Division and Anson’s
cavalry pressed in his advanced posts, and Wellington’s head-
quarters were at Dueñas next day. From thence reconnaissances
were sent out both on the Burgos and the Palencia roads. The latter
city was found unoccupied and—what was more surprising—not in
the least damaged by the retreating enemy. The whole of Clausel’s
army had marched on the Burgos road, and its rearguard was
discovered in front of Torquemada that evening. The British army
followed, leaving Palencia on its left, and head-quarters were at
Magaz on the 12th. Anson’s light dragoons had the interesting
spectacle that afternoon of watching the whole French army defile
across the bridge of the Pisuerga at Torquemada, under cover of a
brigade of chasseurs drawn up on the near side. Critics thought that
the covering force might have been driven in, and jammed against
the narrow roadway over the bridge[22]. But the British brigadier
waited for artillery to come up, and before it arrived the enemy had
hastily decamped. He was pursued as far as Quintana, where there
was a trifling cavalry skirmish at nightfall.
At Magaz, on the night of the 12th, Wellington got the tiresome
news that Santocildes and Castaños, with the main body of the Army
of Galicia, had passed his flank that day, going southward, and had
continued their way towards Valladolid, instead of falling in on the
British line of march. Their junction was thus deferred for several
days. Wellington wrote in anger, ‘Santocildes has been six days
marching: he was yesterday within three leagues of us, and knew it
that night; but he has this morning moved on to Valladolid, eight
leagues from us, and unless I halt two days for him he will not join us
for four or five days more[23].’ As a matter of fact the Galicians did not
come up till the 16th, while if Santocildes had used a little common
sense they would have been in line on the 12th September.
The pursuit of Clausel continued to be a very slow and
uninteresting business: from the 11th to the 15th the army did not
advance more than two leagues a day. On the 13th the British
vanguard was at Villajera, while the French main body was at
Pampliega: on the 14th Anson’s cavalry was at Villadrigo, on the
15th at Villapequeña, on the 16th near Celada, where Clausel was
seen in position. On this day the Galicians at last came up—three
weak divisions under Cabrera, Losada, and Barcena, with something
over 11,000 infantry, but only one field-battery and 350 horse. The
men looked fatigued with much marching and very ragged:
Wellington had hoped for 16,000 men, and considering that the total
force of the Galician army was supposed to be over 30,000 men, it
seems that more might have been up, even allowing for sick,
recruits, and the large garrisons of Ferrol and Corunna.
The 16th September was the only day on which Clausel showed
any signs of making a stand: in the afternoon he was found in
position, a league beyond Celada on favourable ground. Wellington
arranged to turn his left flank next morning, with the 6th Division and
Pack’s brigade; but at dawn he went off in haste, and did not stop till
he reached Villa Buniel, near Burgos. Here, late in the afternoon,
Wellington again outflanked him with the 6th Division, and he
retreated quite close to the city. On the 18th he evacuated it, after
throwing a garrison into the Castle, and went back several leagues
on the high-road toward the Ebro. The Allies entered Burgos, and
pushed their cavalry beyond it without meeting opposition. On the
19th the Castle was invested by the 1st Division and Pack’s brigade,
while the rest of Wellington’s army took position across the road by
which the French had retreated. A cavalry reconnaissance showed
that Clausel had gone back many miles: the last outposts of his rear
were at Quintanavides, beyond the watershed which separates the
basin of the Douro from that of the Ebro. His head-quarters were
now at Briviesca, and it appeared that, if once more pushed, he was
prepared to retreat ad infinitum. Wellington, however, pressed him
no farther: throwing forward three of his own divisions and the
Galicians to Monasterio and other villages to the east of Burgos,
where a good covering position was found, he proceeded to turn his
attention to the Castle.
This short series of operations between the 10th and the 19th of
September 1812 has in it much that perplexes the critical historian. It
is not Clausel’s policy that is interesting—he simply retired day after
day, whenever the enemy’s pursuing infantry came within ten miles
of him. Sometimes he gave way before the mere cavalry of
Wellington’s advanced guard. There was nothing in his conduct to
remind the observer of Ney’s skilful retreat from Pombal to Ponte
Murcella in 1811, when a rearguard action was fought nearly every
afternoon. Clausel was determined not to allow himself to be caught,
and would not hold on, even in tempting positions of considerable
strength. As he wrote to Clarke before the retreat had begun, ‘Si
l’ennemi revient avec toute son armée vers moi, je me tiendrai en
position, quoique toujours à peu de distance de lui, afin de n’avoir
aucun échec à éprouver.’ He suffered no check because he always
made off before he was in the slightest danger of being brought to
action. It is difficult to understand Napier’s enthusiasm for what he
calls ‘beautiful movements’[24]; to abscond on the first approach of
the enemy’s infantry may be a safe and sound policy, but it can
hardly be called brilliant or artistic. It is true that there would have
been worse alternatives to take—Clausel might have retreated to the
Ebro without stopping, or he might have offered battle at the first
chance; but the avoidance of such errors does not in itself constitute
a very high claim to praise. An examination of the details of the
march of the British army during this ten days plainly fails to
corroborate Napier’s statement that the French general ‘offered
battle every day’ and ‘baffled his great adversary.’ His halt for a few
hours at Celada on September 16th was the only one during which
he allowed the allied main body to get into touch with him late in the
day; and he absconded before Wellington’s first manœuvre to
outflank him[25].
The thing that is truly astonishing in this ten days is the
extraordinary torpidity of Wellington’s pursuit. He started by waiting
three days at Valladolid after expelling the French; and he continued,
when once he had put his head-quarters in motion, by making a
series of easy marches of six to ten miles a day, never showing the
least wish to hustle his adversary or to bring him to action[26]. Now to
manœuvre the French army to beyond Burgos, or even to the Ebro,
was not the desired end—it was necessary to put Clausel out of
action, if (as Wellington kept repeating in all his letters) the main
allied army was to return to Madrid within a few weeks, to watch for
Soult’s offensive. To escort the Army of Portugal with ceremonious
politeness to Briviesca, without the loss to pursuers or pursued of
fifty men, was clearly not sufficient. Clausel’s troops, being not
harassed in the least, but allowed a comfortable retreat, remained an
‘army in being’: though moved back eighty miles on the map, they
were not disposed of, and could obviously come forward again the
moment that Wellington left the north with the main body of his
troops. Nothing, therefore, was gained by the whole manœuvre,
save that Clausel was six marches farther from the Douro than he
had been on September 1st. To keep him in his new position
Wellington must have left an adequate containing army: it does not
seem that he could have provided one from the 28,000 Anglo-
Portuguese and 11,000 Galicians who were at his disposition, and
yet have had any appreciable force to take back to Madrid. Clausel
had conducted his raid on Valladolid with something under 25,000
men; but this did not represent the whole strength of the Army of
Portugal: after deducting the sick and the garrisons there were some
39,000 men left—of these some were disarmed stragglers who had
only just come back to their colours, others belonged to shattered
corps which were only just reorganizing themselves in their new
cadres. But in a few weeks Clausel would have at least 35,000 men
under arms of his own troops, without counting anything that the
Army of the North might possibly lend him. To contain him Wellington
would have to leave, in addition to the 11,000 Galicians, at least
three British divisions and the corresponding cavalry—say 16,000 or
18,000 men. He could only bring back some 10,000 bayonets to join
Hill near Madrid. This would not enable him to face Soult.
But if Clausel had been dealt with in a more drastic style, if he
had been hunted and harassed, it is clear that he might have been
disabled for a long time from taking the offensive. His army was still
in a doubtful condition as regards morale, and there can be little
reason to doubt, that if hard pressed, it would have sunk again into
the despondency and disorder which had prevailed at the beginning
of the month of August. The process of driving him back with a firm
hand might, no doubt, have been more costly than the slow tactics
actually adopted; but undoubtedly such a policy would have paid in
the end.
There is no explanation to be got out of Wellington’s rather
numerous letters written between the 10th and the 21st. In one of
them he actually observes that ‘we have been pushing them, but not
very vigorously, till the 16th[27],’ without saying why the pressure was
not applied more vigorously. From others it might perhaps be
deduced that Wellington was awaiting the arrival of the Galician
army, because when it came up he would have a very large instead
of a small superiority of force over the French. But this does not fully
explain his slowness in pursuing an enemy who was evidently on the
run, and determined not to fight. We may, as has been already
mentioned, speak of his wish to avoid loss of life (not much practised
during the Burgos operations a few days later on!) and of the
difficulty of providing for supplies in a countryside unvisited before by
the British army, and very distant from its base-magazines. But when
all has been said, no adequate explanation for his policy has been
provided. It remains inexplicable, and its results were unhappy.
 SECTION XXXIV: CHAPTER II
THE SIEGE OF BURGOS.
SEPTEMBER 19th-OCTOBER 20th, 1812

The Castle of Burgos lies on an isolated hill which rises straight

out of the streets of the north-western corner of that ancient city, and
overtops them by 200 feet or rather more. Ere ever there were kings
in Castile, it had been the residence of Fernan Gonzalez, and the
early counts who recovered the land from the Moors. Rebuilt a dozen
times in the Middle Ages, and long a favourite palace of the Castilian
kings, it had been ruined by a great fire in 1736, and since then had
not been inhabited. There only remained an empty shell, of which the
most important part was the great Donjon which had defied the
flames. The summit of the hill is only 250 yards long: the eastern
section of it was occupied by the Donjon, the western by a large
church, Santa Maria la Blanca: between them were more or less
ruined buildings, which had suffered from the conflagration. Passing
by Burgos in 1808, after the battle of Gamonal, Napoleon had noted
the commanding situation of the hill, and had determined to make it
one of the fortified bases upon which the French domination in
northern Spain was to be founded. He had caused a plan to be drawn
up for the conversion of the ruined mediaeval stronghold into a
modern citadel, which should overawe the city below, and serve as a
half-way house, an arsenal, and a dépôt for French troops moving
between Bayonne and Madrid. Considered as a fortress it had one
prominent defect: while its eastern, southern, and western sides look
down into the low ground around the Arlanzon river, there lies on its
northern side, only 300 yards away, a flat-topped plateau, called the
hill of San Miguel, which rises to within a few feet of the same height
as the Donjon, and overlooks all the lower slopes of the Castle
mount. As this rising ground—now occupied by the city reservoir of
Burgos—commanded so much of the defences, Napoleon held that it
must be occupied, and a fort upon it formed part of his original plan.
But the Emperor passed on; the tide of war swept far south of Madrid;
and the full scheme for the fortification of Burgos was never carried
out; money—the essential thing when building is in hand—was never
forthcoming in sufficient quantities, and the actual state of the place in
1812 was very different from what it would have been if Napoleon’s
orders had been carried out in detail. Enough was done to make the
Castle impregnable against guerrillero bands—the only enemies who
ever came near it between 1809 and 1812—but it could not be
described as a complete or satisfactory piece of military engineering.
Against a besieger unprovided with sufficient artillery it was
formidable enough: round two-thirds of its circuit it had a complete
double enceinte, enclosing the Donjon and the church on the summit
which formed its nucleus. On the western side, for about one-third of
its circumference, it had an outer or third line of defence, to take in
the lowest slopes of the hill on which it lies. For here the ground
descended gradually, while to the east it shelved very steeply down to
the town, and an external defence was unnecessary and indeed
impossible.
The outer line all round (i.e. the third line on the west, the second
line on the rest of the circumference) had as its base the old walls of
the external enclosure of the mediaeval Castle, modernized by shot-
proof parapets and with tambours and palisades added at the angles
to give flank fire. It had a ditch 30 feet wide, and a counterscarp in
masonry, while the inner enceintes were only strong earthworks, like
good field entrenchments; they were, however, both furnished with
palisades in front and were also ‘fraised’ above. The Donjon, which
had been strengthened and built up, contained the powder magazine
in its lower story. On its platform, which was most solid, was
established a battery for eight heavy guns (Batterie Napoléon), which
from its lofty position commanded all the surrounding ground,
including the top of the hill of San Miguel. The magazine of
provisions, which was copiously supplied, was in the church of Santa
Maria la Blanca. Food never failed—but water was a more serious
problem; there was only one well, and the garrison had to be put on
an allowance for drinking from the commencement of the siege. The
hornwork of San Miguel, which covered the important plateau to the
north, had never been properly finished. It was very large; its front
was composed of earthwork 25 feet high, covered by a counterscarp
of 10 feet deep. Here the work was formidable, the scarp being steep
and slippery; but the flanks were not so strong, and the rear or gorge
was only closed by a row of palisades, erected within the last two
days. The only outer defences consisted of three light flèches, or
redans, lying some 60 yards out in front of the hornwork, at projecting
points of the plateau, which commanded the lower slopes. The
artillery in San Miguel consisted of seven field-pieces, 4- and 6-
pounders: there were no heavy guns in it.
The garrison of Burgos belonged to the Army of the North, not to
that of Portugal. Caffarelli himself paid a hasty visit to the place just
before the siege began, and threw in some picked troops—two
battalions of the 34th[28], one of the 130th; making 1,600 infantry.
There were also a company of artillery, another of pioneers, and
detachments which brought up the whole to exactly 2,000 men—a
very sufficient number for a place of such small size. There were nine
heavy guns (16- and 12-pounders), of which eight were placed in the
Napoleon battery, eleven field-pieces (seven of them in San Miguel),
and six mortars or howitzers. This was none too great a provision,
and would have been inadequate against a besieger provided with a
proper battering-train: Wellington—as we shall see—was not so
provided. The governor was a General of Brigade named Dubreton,
one of the most resourceful and enterprising officers whom the British
army ever encountered. He earned at Burgos a reputation even more
brilliant than that which Phillipon acquired at Badajoz.
The weak points of the fortress were firstly the unfinished
condition of the San Miguel hornwork, which Dubreton had to
maintain as long as he could, in order that the British might not use
the hill on which it stood as vantage ground for battering the Castle;
secondly, the lack of cover within the works. The Donjon and the
church of Santa Maria could not house a tithe of the garrison; the rest
had to bivouac in the open, a trying experience in the rain, which fell
copiously on many days of the siege. If the besiegers had possessed
a provision of mortars, to keep up a regular bombardment of the