Full text of "A history of the war, the middle ages from the fourth to the fourteenth century"

A HISTORY OF
THE ART OF WAR

BY THE SAME AUTHOR

WARWICK THE KINGMAKER

A HISTORY OF EI-ROPE, 476-9,8

A HISTORY OF THE BYZANTINE EMPIRE

A HISTORY OF

THE ART OF WAR

THE MIDDLE AGES

FROM THE FOURTH TO THE

FOURTEENTH CENTURY

CHARLES OMAN, M.A., F.S.A.

FELLOW OF ALL SOULS COLLEGE, OXFORD

WITH MAPS, PLANS, AND ILLUSTRATIONS

NEW YORK: G. P. PUTNAM S SONS

LONDON: METHUEN & CO.

1898

RFF & REN:

PREFACE

THE present volume is intended to form the second of a series
of four, in which I hope to give a general sketch of the history of
the art of war from Greek and Roman times down to the begin
ning of the nineteenth century. The first volume will deal with
classical antiquity ; this, the second, covers the period between
the downfall of the Roman Empire and the fourteenth century.
In the third volume will be included the fifteenth, sixteenth, and
seventeenth centuries. The fourth will treat of the military
history of the eighteenth century and of the Revolutionary and
Napoleonic wars down to Waterloo.

These volumes are concerned with the history of the art
of war, and do not purport to give the complete military
annals of the civilised world. Each section deals with the
characteristic tactics, strategy, and military organisation of a
period, and illustrates them by detailed accounts of typical
campaigns and battles. There are also chapters dealing with
the siegecraft and fortification, the arms and armour of each
age.

The present volume should in strict logic have included two
more books, dealing the one with the military history of Central
and Eastern Europe in the fourteenth century (especially with the
first rise of the Swiss and the Ottoman Turks), and the other
with the invention of gunpowder and firearms. But the exi
gencies of space the volume is already more than six hundred
and sixty pages long have compelled me to relegate these
topics to the opening chapters of the third volume. It is
fortunate that the influence of the discovery of gunpowder on

PREFACE

the wars of Western Europe was so insignificant during the
fourteenth century that no serious harm comes from deferring
the discussion of the subject.

I have endeavoured to avoid overburdening the volume with
too voluminous foot-notes, but at the same time have given
references for all statements which might seem to require
justification or defence. In citing English chronicles my
references are, where possible, to the Rolls Series editions ;
French chronicles are mainly quoted from Bouquet s magnificent
Scriptores Rcrum Gallicannn ct Francicarum, German and
Italian from the collections of Pertz and Muratori respectively.

Much valuable aid given to the author requires grateful
acknowledgment. Most especially must I express my thanks
to two helpers : to the compiler of the index the fourth and
the largest which has been constructed for books of mine by the
same kindly hands and to my friend Mr. C. H. Turner, Fellow
of Magdalen College, who read the whole of the proofs, and
furnished me with a great number of corrections and improve
ments.

I have also to acknowledge my indebtedness to Mr. T. A.
Archer, who was good enough to go through with me the whole
of Book V. (the Crusades) and also chapter vii. of Book III.,
wherein certain topics much disputed of late years are dealt
with. I also owe some valuable hints to Professor York Powell
and to the Rev. H. B. George of New College. The former,
with his usual omniscience, indicated to me several lines of
inquiry, from which I obtained valuable results. The latter will
notice that in chapter ii. of Book VIII. I have adopted his
theory of the formation of the English army at Cregy. Mr. F.
Haverfield of Christ Church gave me some useful notes for the
opening pages of the first chapter of Book I.

All the maps and plans have been constructed by myself
from the best sources that I could procure. When possible, I
walked over important battlefields, e.g. Crecy, Bouvines,
Bannockburn, Evesham, in order to supplement the information

PREFACE vii

to be derived from maps by a personal acquaintance with the
ground. The English plans are derived from the Ordnance
Survey, the French from the maps of the Etat-Major, the
Syrian from the admirable publications of the Palestine
Exploration Society.

Of the seven plates illustrating armour, the first three are
sketches taken from the original manuscripts ; the last four I
owe to the kindness of Messrs. Parker of Oxford, who permitted
me to reduce them from the blocks of one of their most
valuable publications, Hewitt s Ancient Armour, a book from
which I derived much useful information when dealing with the
later Middle Ages.

OXFORD, March i, 1898.

TABLE OF CONTENTS

BOOK I

THE TRANSITION FROM ROMAN TO MEDIAEVAL
FORMS IN WAR, A.D. 235-552

CHAPTER I. THE LAST DAYS OF THE LEGION,
A.D. 235-450

PAGES

Collapse of the Frontier Defences of the Roman Empire Disasters of the
Third Century Reorganisation of the Army by Diocletian and Con-
stantine I. Final Success of the Barbarians Battle of Adrianople (378)
The Foederati Vegetius and the Decay of Infantry The Huns . 3-21

CHAPTER II. COMMENCEMENT OK THE SUPREMACY OF
CAVALRY, A.D. 450-552

The Army of the Eastern Empire The Isaurians Justinian and his Wars
The Horse-Archer Belisarius and his Tactics Battle of Daras
(530) Battle of Tricameron (535) Belisarius and the Goths Battle
of Taginae (552) Battle of Casilinum (554) . . . 22~37

BOOK II

THE EARLY MIDDLE AGES, A.D. 500-768
CHAPTER I. THE VISIGOTHS, LOMBARDS, AND FRANKS

Cavalry and Infantry among the Teutonic Peoples The Visigoths in Spain,
their Military Institutions and their Decay The Lombards, their Arms
and Tactics The Franks and their Early Methods of War They finally
adopt Armour and take to Horsemanship Their Weakness as an Offen
sive Power ........ 41-62

CHAPTER II. THE ANGLO-SAXONS

The Conquest of Britain Arms of the Old English They remain a Nation
of Foot-Soldiery Evidence of the5t u a//^Indecisive Nature of the Old
English Wars ... .... 63-72

x TABLE OF CONTENTS

BOOK II I

FROM CHARLES THE GREAT TO THE BATTLE
OF HASTINGS, A.D. 768-1066

CHAPTER I. CHARLES THE GREAT AND THE EARLY
CAROLINGIANS

PAGES

The Empire of Charles the Great His Military Legislation Growth of the
Importance of Cavalry Charles and his Burgs Carolingian Armour
The Beginnings of Feudalism , 75~^8

CHAPTER II. THE VIKINGS

The Coming of the Vikings Their Tactics Fate of their Invasions in the
Empire, England, and Ireland Disruption of the Carolingian Empire
The later Carolingians and their Efforts to restrain the Vikings
Charles the Bald and the Edict of Pitres Arnulf and the Battle of
Louvain ........ 89-100

CHAPTER III. THE VIKINGS TURNED BACK THE FEUDAL
HORSEMAN AN*D THE FEUDAL CASTLE THE THEGN
AND THE BURH

Importance of Cavalry in the Struggle with the Vikings Development
of Feudal Cavalry on the Continent Systematic Fortification Alfred
and the Danes Origin of the Thegnhood The flurhs of Edward the
Elder Origin of the English Fleet The Housecarles . . 101-115

CHAPTER IV. THE MAGYARS

Appearance of the Magyars on the Danube They ravage Germany, Italy,
and France Henry the Fowler and his Burgs Battle on the Unstrut
(933) The Last Raids Battle of the Lechfekl (955) The Magyars
turned back ...... 116-125

CHAPTER V. ARMS AND ARMOUR (800-1100)

Extension of the use of Armour The Byrnie and the Helm with Nasal

The Shield The Danish Axe . . . 126-130

CHAPTER VI. SIEGECRAFT AND FORTIFICATION
The Siegecraft of the Early Middle Ages The Ram and the Bore Mining
The Movable Tower Military Engines : the Balista and the Mangon
The Crossbow The Great Siege of Paris (885-886) and its Stages . 131-148

CHAPTER VII. THE LAST STRUGGLES OF INFANTRY
HASTINGS AND DYRRHACHIUM

Duke William invades England His Army and its Tactics The Senlac
Position Harold adopts the Defensive Battle of Hastings (1066)
Victory of the Horseman and the Archer over the Old English Infantry
Dyrrhachium (1081) : the Norman Horse and the Varangian Axemen 149-165

TABLE OF CONTENTS xi

BOOK IV

THE BYZANTINES, A.D. 579-1204

CHAPTER I. HISTORICAL DEVELOPMENT OF THE
BYZANTINE ARMY

PAGES

Strong Points of Byzantine Army Its Reorganisation by Maurice (circ.
579-580) The Strategicon Character and Composition of the East
Roman Forces in the Early Middle Ages Importance of Archery and
Heavy Cavalry The Struggle with the Saracens Creation of the
Themes ; Strength of the System . 169-183

CHAPTER II. ARMS AND ORGANISATION OF THE
BYZANTINE ARMY

Arms of the Heavy Cavalry Their Tactics Arms of the Infantry The
Auxiliary Services The Army in Camp and on the March The Line
of Battle Leo s Ideal Formation for Cavalry .... 184-197

CHAPTER III. STRATEGY AND TACTICS OF THE
BYZANTINE ARMY

Defensive Character of Byzantine Strategy Its Weak Points Methods of
dealing with the Franks, the Slavs, the Turks The long Saracen Wars
The Tactics by which the Moslems were turned back from Asia Minor
The Book of Nicephorus Phocas on Frontier Defence Military
Successes of the Tenth Century ..... 198-215

CHAPTER IV. DECLINE OF THE BYZANTINE ARMY
(1071-1204)

The Coming of the Seljouks Battle of Manzikert (1071) The Loss of Asia
Minor Disorganisation of the Army Alexius I. and the Battle of
Calavryta (1079) The Army under the Comneni . . 216-226

BOOK V

THE CRUSADES, A.D. 1097-1291
CHAPTER I. INTRODUCTORY

The Western Nations take the Offensive Conditions which rendered the

Crusades possible Faults of the Crusading Armies . . . 229-232

xii TABLE OF CONTENTS

CHAPTER II. THE GRAND STRATEGY OF THE CRUSADES

PACES

Lack of Geographical Knowledge among the Crusaders Their Mistaken
Choice of Itineraries The Land-Routes of Asia Minor The Lines
selected by the Leaders of the First, Second, and Third Crusades
Strategy of the Conquest of Syria The Crusading States and their
Boundaries The Causes of their Fall Zengi and Saladin The Attacks
of the Crusaders on Egypt King Amaury, John of Brienne, and Louis IX. 233-267

CHAPTER III. THE TACTICS OF THE CRUSADES
/. The Earlier Battles (1097-1 102)

The Turkish Horse- Archers and the Frankish Knights Battle of Dorylseum
(1097) Siege of Antioch and Combat of Harenc (1097-98) Battle of
Antioch (1098) Battle of Ascalon (1099) Battles of Ramleh (noi and
1102) Battle of Jaffa (1102) . . . . 268-293

CHAPTER IV. THE TACTICS OF THE CRUSADES continued
II. The Later Battles (1119-1192)

Regular Combination of Infantry and Cavalry Battle of Hab (1119)
Battle of Hazarth (1125) Battle of Marj-es-Safar (1126) King
Richard I. and his Tactics The March from Acre to Jaffa The
Triumph of Arsouf (1191) Combat of Jaffa (1192) . . . 294-317

CHAPTER V. THE GREAT DEFEATS OF THE CRUSADERS

The Causes of Defeat Battle of Carrhae (1104) Battle of Tiberias (1187)
Battle in front of Acre (1192) Battle of Mansourah (1250) The
Moral of such Disasters ....... 318-350

BOOK VI

WESTERN EUROPE FROM THE BATTLE OF HASTINGS
TO THE RISE OF THE LONGBOW

CHAPTER I. INTRODUCTORY

Complete Supremacy of Cavalry Neglect of the use of Infantry The

General Type of Battles in the Twelfth and Thirteenth Centuries . 353-356

CHAPTER II. THE ARMIES OF THE TWELFTH AND
THIRTEENTH CENTURIES

A. England. TheFyrd and the Feudal Host William the Conqueror and

Knight-service The I ettts Feoffdmentum The Cartae Saronum

Scutage Status of the Knight, and its Changes The Rise of Mercen-

aries The Braban$ons and Crossbowmen .... 357-360

B. Tin Continent. Different Fate of the term Miles in different Countries
The Clientes and Sergeants The Military Caste Importance of
Mercenaries The Flemish Pikemen The Crossbowmen of Italy 369- 77

TABLE OF CONTENTS xiii

CHAPTER III. ENGLISH BATTLES AND THEIR TACTICS
(1100-1200)

PACES

Battle of Tenchebrai (1106) Battle of Bremule (1119) Combat of Bourg
Theroulde (1124) Battle of Northallerton (1138) First Battle of
Lincoln (1141) The Wars of Henry II. : Alnwick and Fornham The
English in Ireland and their Enemies Battle on the Dinin (1169)
Battle of Dublin (1171 ) Surprise of Castle Knock (1171 ) Characteristics
of Irish War . . . . 378-406

CHAPTER IV. ENGLISH BATTLES AND THEIR TACTICS
(1200-1272)

Second Battle of Lincoln (1217) Combat of Taillebourg (1242) Battle of
Lewes (1264) Campaign of 1265 and Battle of Evesham Comparison
of the Merits ol Simon de Montfort and Edward I. . . . 407-435

CHAPTER V. CONTINENTAL BATTLES
(1100-1300)

The Types of Battle and their Variety Battle of Thielt (1128) Battle of
Legnano (1176) Battle of Steppes (1213) Battle of Muret (1213), the
Greatest Triumph of Unaided Cavalry in the Epoch Campaign of 1214 :
King John s Strategy: Causes of its Failure Battle of Bouvines (1214)
Battle of Benevento (1266) Battle of Tagliacozzo (1268) Tactics of
Charles of Anjou Battle of the Marchfeld (1278) . . . 436-509

CHAPTER VI. ARMS AND ARMOUR
(1100-1300)

The Hauberk and the Gambeson Development of the "Great Helm "-
Rise of Heraldry The Beginnings of Plate Armour Its Slow Develop
ment ......... 510-516

CHAPTER VII. FORTIFICATION AND SIEGECRAFT
(i 100-1300)

Scarcity of Stone Fortifications before the Eleventh Century The Early
Stockaded Mounds, Burhs and Mattes The Castles of William the
Conqueror The Rectangular Keep of Norman Times : the Tower of
London The Shell Keep Influence of the Crusaders on Fortification :
Byzantine Sources The Fortifications of Constantinople and Antioch
The Earlier Syrian Castles : Kerak-in-Moab Richard I. builds Chateau
Gaillard Its Siege by Philip Augustus ( 1 203-4) The Thirteenth Century
in East and West Concentric Castles : Krak-des-Chevaliers and Caer-
philly The Castles of Edward I. Siegecraft : Introduction of the
Trebuchet : its four Varieties The Balista, Mangon, and Springal
Greek Fire and its Use Its Employment by Byzantines and Saracens
Minir.g: the Siege of Carcassonne (1204) Use of the Mine in the
Levant General Ascendency of the Defensive over the Offensive in the
Thirteenth Century, and its Political Consequences . . . 517-553

xiv TABLE OF CONTENTS

BOOK VII

ENGLAND AND SCOTLAND, A.D. 1296-1333
DEVELOPMENT OF THE LONGBOW

CHAPTER I. ENGLAND AND SCOTLAND (1296-1328)

PACBS

The History of the Longbow Its probable Origin in South Wales The
Assize of Arms of 1252 Use of the Bow in the Welsh Wars of
Edward I. Difference of Welsh and Scottish Campaigns Wallace
and the Battle of Cambuskenneth (1297) The Longbow at Falkirk
(1298) Bruce at the Combat of Loudon Hill (1307) Bruce victori
ous at Bannockburn (1314) Mistaken Tactics of Edward n. "King
Robert s Testament " . . . . 557-580

CHAPTER II. ENGLAND ANL&gt; SCOTLAND (1328-1333) FIRST COM
BINATION OF ARCHERY AND DISMOUNTED CAVALRY

The English change their Tactics The Longbow at Dupplin Muir (1332)
Edward HI. victorious at Halidon (1333) Complete Ascendency of
the Bow over the Pike . ... 581-588

BOOK VIII

THE LONGBOW BEYOND THE SEAS

CHAPTER I. THE ARMIES OF EDWARD in.

Comparison between the Military Strength of England and France The
Methods by which English Armies were raised : Commissions of Array
and Indentures Character and Composition of the Armies of Edward in. 591-596

CHAPTER II. THE LONGBOW IN FRANCE: CRECY

Archers and Crossbowmen : Combat of Cadzand (1337) King Edward in
Flanders The Great Invasion of France : Edward marches from La
Hogue to Cre9y Battle of Crecy (1346) Its Tactical Meaning . 597-615

CHAPTER III. POICTIERS, COCKEREL, AND AURAY
The Effects of Cre$y : Combats of La Roche Darien and Ardres The Black
Prince invades France His March to Maupertuis Battle of Poictiers
(1356) : new Tactics of the French : their Complete Failure Cocherel
and Auray (1364; _ 616-636

CHAPTER IV. NAYARETTE AND ALJUEAROTTA
Arms and Tactics of the Spaniards : " the Genetours" The Black Prince
invades Castile and outgenerals Don Henry of Trastamara Triumph
of the Archers at Kavarette (1367) The Spaniards fail to profit by the
Lesson Battle of Aljubarotta( 1385) The Portuguese apply the English
System and win a great Victory over the Castilians . . 637-653

TABLE OF MAPS, PLANS, AND
ILLUSTRATIONS

PLATE I. Plans of Daras and Taginae . . To face page 28

,, II. Prankish Warriors . . . . ,, 36

III. The "Themes" of the Eastern Empire in 680

and 900 ..... ,, 180

IV. Byzantine Soldiery . . 186

,, V. Byzantine Soldiery . . 1 88

,, VI. Byzantine Cavalry Formation, and Neighbour
hood of Antioch . . ,, 196

VII. Siege and Battle of Antioch . ,, 282

,, VIII. Battles of Ascalon and Hab 288

IX. Battle of Arsouf 310

,, X. Battles of Tiberias and Acre . 326

,, XI. Map of Lower Egypt : Plan of Mansourah . ,, 342
XII. Battles of Tenchebrai, Bremule, Northallerton,

and Lincoln ... . 394

,, XIII. Battle of Lewes ; Campaign of Evesham ,, 420

,, XIV. Battles of Muret and Bouvines . .. 450

,, XV. Battles of Benevento and Tagliacozzo . ,, 484

XVI. The Marchfeld . . 504

XVII. Seals of William I. and William n. . ,, 510

,, XVIII. Seals of Richard I. and Henry in. . ,, 512

,, XIX. Thirteenth-century Armour . 514

,, XX. Thirteenth-century Armour . ., 516

XXI. Typical Castles of the period 1100-1300 ,, 530

XXII. Battle of Bannockburn . . . . ,, 572

,, XXIII. Battles of Crecy and Poictiers . 606

XXIV. Battle of Navarette . . . ,,644

BOOK I

THK TRANSITION FROM ROMAN TO MEDIAEVAL
FORMS IN WAR

THE ART OF WAR

THE MIDDLE AGES

CHAPTER I

THE LAST DAYS OF THE LEGION
A.D. 235-450

BETWEEN the middle of the third and the middle of the
fifth century lies a period of transition in military history,
an epoch of transformations as strange and as complete as those
contemporary changes which turned into a new channel the
course of political history and of civilisation in Europe. In war,
as in all else, the institutions of the ancient world are seen to
pass away, and a new order of things develops itself.

The most characteristic symptom of the tendencies of this
period is the gradual disappearance of the Roman legion, that
time-honoured organisation whose name is so intimately bound
up with the story of Roman greatness. In A.D. 250 it was still
the heavy-armed infantry of the empire which formed the core
of battle, and was the hope and stay of the general. By A.D.
450 the cavalry was all in all, the foot-soldiery had fallen into
disrepute, and the very name of legion was almost forgotten. It
represented a form of military efficiency which had now com
pletely vanished. That wonderful combination of strength and
flexibility, so solid and yet so agile and easy to handle, had
ceased to correspond to the needs of the time. The day of
the sword and pilum had given place to that of the lance and
bow. The typical Roman soldier was no longer the iron
legionary, who, with shield fitted close to his left shoulder and

4 THE ART OF WAR IN THE MIDDLE AGES [200

sword-hilt sunk low, cut his way through the thickest hedge of
pikes, turned back the onset of the mailed horsemen of the East,
and stood unmoved before the wildest rush of Celt or German
The old military organisation of Augustus and Trajan began
to fall to pieces in the third century ; in the fourth it was so
weakened and transformed as to be hardly recognisable ; by
the end of the fifth it had disappeared.

The change in the character of the Roman army which
ultimately substituted cavalry and light infantry for the solid
strength of the ancient legion was mainly caused by the
exigencies of border-warfare. From the time of Hadrian to that
of Severus, the system of frontier- defence \\hich the Roman
Government adopted was to fix the limit of the empire at a
great natural boundary, such as the Rhine, Danube, or Euphrates,
and to place behind the boundary at suitable points large
permanent camps, in which one or more legions were quartered.
These garrisons were placed many scores or even hundreds of
miles apart, and the long intervals between them were only
filled by minor posts occupied by small bodies of auxiliary
troops. Where natural obstacles, such as rivers or mountain-
chains, were wanting, the frontier was not unfrequently
marked out by long lines of entrenchments, like our o.vn
Northumbrian Wall, or the similar structure which stretches
across South Germany. The stations were connected with each
other by good military roads, and the alarm could be passed
from one to another at the shortest notice by a system of beacons
and mounted messengers. If the barbarous enemy across the
frontier, German, Sarmatian, or Parthian, essayed a raid on
Roman territory, he must first cross the obstacles and then cope
with the garrisons of the local posts. These would be able to
beat back any small plundering parties ; but if they found the
invaders too strong, they could at least endeavour to harass
them, and to restrict the area of their ravages, till the nearest
legion could march up from its great permanent camp.

This system worked well for more than a hundred years.
But it had its weak points ; there was a great want of a central
reserve, in case the legions of any frontier should be unable to
hold their ground against an attack of unusual strength For
the middle provinces of the empire were kept entirely & denuded
of troops, and new legions could not be improvised in a hurry
from the umvarlike subjects of the empire, as they had once

235] FAILURE OF ROMAN FRONTIER DEFENCES 5

been from the citizens of the early republic. Hence it came to
pass that a disaster on one point of the border had to be
repaired by drawing troops from another. This rather dangerous
device could only be employed so long as the enemies of Rome
were so obliging as to present themselves one by one, and to
refrain from simultaneous onslaughts on far distant tracts of
frontier. For more than two centuries the empire was fortunate
enough to escape this contingency ; its military system was
never tried by the crucial test of an attack all along the line ;
in the times of stress Germany could lend troops to Britain, or
Moesia reinforce the legions of Syria. Disasters were suffered
from time to time which threw a province for a moment into
hostile hands, but because they came singly they could always
be repaired. The rebellion of Civilis shook the Roman hole 1
on the Rhine frontier for a space ; the defeat of Domitian s
generals Sabinus and Fuscus let the Dacians into the interior
of the Danube provinces ; Marcus Aurelius once saw the Quadi
at the gates of Aquileia. But reinforcements were brought up
from frontiers where no war was in progress, and the incoming
flood of invasion was at length stemmed.

In the third century there was a complete change in the
face of affairs : the system of defence broke down, and the empire
well-nigh collapsed under the stress. From the da} of the
murder of Alexander Severus (235 A.D.) to the moment at which
Diocletian put down the last surviving rebel Caesar in the
remotest corner of the West (297) the empire was subjected with -
out a moment s respite to the double scourge of civil war and
foreign invasion. In the space of sixty years no less than sixteen
emperors and more than thirty would-be emperors fell by sword
or dagger. While the arms of the legions were turned against
each other, the opportunity of the enemies of the empire had
arrived. All its frontiers simultaneously were beset by the
outer barbarians, and the fabric reeled before the shock. F&lt; &gt;r
Rome s neighbours were growing more powerful just when Rome
herself was weak and divided. The new and vigorous Persian
kingdom had just replaced the decrepit Parthian power in the
East (A.D. 226). The Germans were already commencing to form
the confederacies which made their scattered tribes for the first
time really formidable. The names of the Franks, Alamanni
and Goths begin to appear along the Rhine and Danube.

So long as the frontier defence of the legions held firm, the

6 THE ART OF WAR IN THE MIDDLE AGES [259

empire presented to its foes a hard shell and a soft kernel. The
border was strongly held and difficult to pierce, but the rich
prsi inciae incrincs within were defenceless and ripe for plunder,
if only the shell could be pierced. When the legions were with
drawn from the frontier to take part in civil war, and marched
off time after time to enthrone some new usurper upon the
Palatine, it was impossible to keep back any longer the pressure
from without. The period 235-297 opens with a heavy and
long-continued onslaught of the Ouadi Carpi and Goths on the
Middle and Lower Danube (236;. It was beaten back by
Maximinus I. and Philip for a few years; but in 249, while a
vigorous civil war was distracting the Illyrian regions, the line
of resistance was at last broken through. The Goths crossed
Danube and Balkans, overran Moesia and Thrace, and scattered
the Imperial troops before than. The Emperor Decius, having
put down his rivals, hastened to meet them ; but he, his son,
and his whole army were cut to pieces in the disastrous battle
of Forum Trebonii in the summer of 251. No Roman emperor
had ever been slain before in battle with the barbarians ; no
Roman host of such strength had suffered defeat since the day
of Cannae. It seemed for a moment as if the empire was fated
to be cut in twain, or even as if some earlier Alaric were about
to present himself before the gates of Rome.

For the next twenty years the Goths ranged almost
unresisted over the middle provinces of the empire. The
troops that should have been called in to resist them were
occupied in civil wars in Italy, or were employed in defending
other menaced frontiers. For, while the Gothic war was at its
height, the Persian king Sapor overran Mesopotamia, defeated
and took captive the Emperor Valerian, stormed Antioch, and
ravaged Syria and Asia Minor ^258-259). Favoured by these
distractions, the Goths were able to carry all before them in the
central provinces of the empire. Not only did they harry the
whole Balkan peninsula as far as Athens and Dyrrachium, but
daring bands of plunderers crossed the Hellespont and sacked
Chalcedon, Alexandria Troas, Ephesus, and even the distant
Trebizond. With a little more guidance and a single leader at
their head, they might have made an end of the empire, for
usurpers were rising in every province. Civil war had become
endemic among the Romans ; the Germans of the Rhine frontier
were battering at the defences of Gaul and Rhaetia ; and the

297] DIOCLETIAN REORGANISES THE ARMY 7

indolent and frivolous Gallienus, who still maintained his
precarious seat on the Palatine, bade fair to be the Sardanapalus
of Rome, and to see city and empire go down together in one
universal conflagration of civil strife and foreign war. In the
years 260-268 all seemed lost. But deliverers arose the tough
Illyrians, Claudius, Aurelian, and Probus, reconquered the West
from rebel Caesars, cleared the Germans out of the Balkan
peninsula, and won back the East from the Persians and the
Palmyrenes. Soon after came Diocletian, the reorganiser and
restorer, and with the reconquest of Britain (A.D. 297) the empire
resumed its old external shape.

But the restoration was external only. In the sixty years
of battle, murder, and plague which had elapsed since the
extinction of the dynasty of Severus, the vital strength of the
empire had been fatally sapped. Half the provinces lay waste :
the other half had been drained dry of their resources. By
twenty years of incessant labour Diocletian restored a super
ficial semblance of strength and order ; his grinding taxation
enabled him to put an end to the chronic bankruptcy of the
Imperial exchequer, and to restore and regarrison the lon^
broken-down military frontier of the Roman world.

But the sixty years of anarchy and disaster had left-
indelible marks on the composition and organisation of the
Roman army. Though few of the old legions of Trajan and
Severus seem to have disappeared, most of their names arc
still found in the Notitia, a document a hundred years later
than Diocletian, yet they had apparently been much pulled
about and disorganised, by being cut up and sent apart in
detachments. Often the legionary eagle at headquarters must
have been surrounded by a mere fraction of the corps, while
detached cohorts were serving all about the world, drafted off
under the pressure of necessity. 1 All sorts of cohorts anrl
alae with new and often strange names had been raised
The old broad division of the army into legions and auxilia, the
former filled with Roman citizens, the latter with subjects of
the empire who did not possess the citizenship, could no longer
exist, for Caracalla in 212 had bestowed the franchise on all
provincials. Thus the ancient distinction between the legionary

1 So, at least, one would deduce from such facts as that the usurper Carausius in
Britain strikes coins to celebrate the fidelity to himself of legions whose proper head
quarters were in Germany or Moesia, e.g. IV. Flavia and XXX. Ulpia.

8 THE ART OF WAR IN THE MIDDLE AGES [300

who was a Roman and the auxiliary who was not had vanished :
the status of the one was now as good as that of the other.

Yet if auxiliary and legionary were now Romans alike, the
non-citizen element had not disappeared from the army. In
the days of anarchy the emperors had not been able to reject
any military resources that came to hand. They had enlisted
thousands of warriors from across the frontier, who were not
subjects of the empire at all, and only served for pay and
plunder. Broken German clans, Sarmatians, Arabs, Armenians,
Persian renegades, Moors from inner Africa, were all welcomed
in the time of stress and need. Corps formed of these foreigners
now stood to the Roman army in much the same relation that
the auxiliaries had once borne to the legions. Individuals
among the mercenaries rose to high rank in the army ; one of
them, said to be the son of a Gothic father and an Alan mother,
wore the purple for three short years under his adopted name
of Gaius Julius Verus Maximinus. But it is needful to note
that down to the beginning of the fourth century these foreign
elements in the Roman army, though growing perilously large,
were still entirely subsidiary to the native legions and cohorts.
In the words of a fourth-century writer, they were still praeliandi
magis adminiculum quam prindpale subsidium&gt;

But a tendency to increase the proportion of cavalry and light
infantry, arid to trust less and less to the legionary of the old
type, grows more and more apparent as the fourth century
commences. This is best shown by the fact that the name of
" legion " itself no longer commands its old prominence in the
empire. Instead of being considered superior to all other corps,
and taking precedence of them, the legionaries began to be treated
as what we should now call " troops of the line," and saw many
new bodies, which were in name, but not in fact, parts of the
Imperial guard, preferred to them. It was considered high pro
motion when Diocletian took two Moesian legions out of their
old numerical place in the army list, rechristened them the
Jovians and Herculians, and gave them under their new titles pre
cedence over all their former comrades. By the end of the fourth
century we learn from Vegetius that the legions had been so
neglected and thrust back that it was difficult to keep their ranks
" the large majority of recruits insist on enlisting among
the auxiliaries, where the discipline is less severe, where the work

1 Vegetius, i. 2.

3oo] INSTITUTION OF A CENTRAL RESERVE 9

is lighter, and where the rewards of good service come quicker
and are bestowed with a more bountiful hand." 1

In the Roman army as it was reorganised by Diocletian the
legionary infantry no longer formed, as of old, the wholly pre
ponderant part of the foot-soldiery of the empire, in spite of
the fact that he and his colleagues raised a very considerable
number of new legions. In the eastern half of the empire,
where Diocletian himself presided, he seems to have added eleven
new legions to the sixteen old ones which he found already
existing. But the non-legionary part of the army was developed
on an even larger scale. To the already existing auxiliary cohorts
and numeri other bodies were added in huge numbers. 2 But
they do not mainly belong to the frontier line of defence where
the legions lay. The institution of the Comitatenses or movable
Imperial army, as opposed to the limitanei or ripenses, the fixed
garrison troops of the frontier, belongs undoubtedly to Diocletian s
time. In this category were placed the flower of the new
regiments. They were mainly composed of provincials from the
Illyrian, Gallic, and Germanic provinces, though there was a con
siderable number of corps raised from the barbarians beyond the
Rhine and Danube. Quartered almost entirely in the interior
of the empire, they were to be used as a central reserve, free to be
transferred to any point of the border that chanced to be in peril.
To the Comitatenses raised by Diocletian numerous additions
were made by Constantine, who drafted off many cohorts and
fragments of legions from the frontier forces and added them to
the movable army. These were the corps which later genera
tions called the Pseudo-comitatenses, a curious name intended to
show that they ranked somewhat lower than the old comita-
tensian troops, though they had been raised to a higher standing
than the surviving limitary legions.

For some not fully known reason all the legions of the
Comitatenses were kept at a strength of only a thousand strong,
though those left on the border still retained their old comple
ment of six thousand men. Thus, though there were seventy such

1 Vegetius, ii. 3,

2 Of cohorts alone there were still fifteen existing when the Notitia was drawn up
which bear the names of Diocletian or his colleagues Maximian and Constantius (i.e.
Klavia, Valeria, fovia, Herculea) in the regimental name. See Mommsen, Hermes,
1889. How many new cohorts were made which did not bear the Imperial name one
cannot say. In the Notitia there were a hundred and five cohorts and forty-four
auxi/ia in the frontier garrisons, over and above the legions.

io THE ART OF WAR IN THE MIDDLE AGES [300

legions at the end of the fourth century, they did not represent
the enormous force which such a roll of names seems to imply.

But Diocletian not only raised the Comitatenses and gave
them precedence over the old legions. He was the first to raise
a huge Imperial guard, which stood as much above the Comita
tenses as the latter did above the limitary troops. These were
the Palatini, who practically superseded the old Praetorians, a
body which Diocletian rightly distrusted, as having for the last
century been far too much given to the making and unmaking
of emperors. He confined the Praetorians to Rome, a place
which neither he nor his colleagues often visited, and formed his
new Imperial guard out of picked men who did not inherit the
evil traditions of the old corps. How numerous the Palatini
were at their creation we cannot say ; but by the end of the
century they appear in the Notitia as a very considerable bod} ,
comprising twenty-four " vexillations " of horse (regiments of
five hundred eachj, and of foot twenty-five legions, each a
thousand strong, with a hundred and eight auxilia, each
probably five hundred strong. This was, no doubt, a very much
stronger force than the original Palatine regiments raised by
Diocletian. Each of his successors had added new units to it,
as the names " Honorian," " Theodosian," etc., show. Constantine
the Great is known to have raised the five scJwlae of horsemen
who formed the actual life-guard of the prince, and followed his
person whenever he went out to war. By the end of the century
the Imperial guard mustered about twelve thousand horse and
eighty thousand foot, all (or nearly all) cantoned round or within
the eastern and western capitals of the empire.

Among the Palatini, as among the Comitatenses, there was a
very strong barbarian element, and this element was on the
increase all through the fourth century. As Mommsen remarks, 1
" each corps seems to have been valued more highly in proportion
as it differed the more in nationality, organisation, and spirit
from the old normal Roman legions."

Great as was the increase made by Diocletian and his col
leagues in the number of the non-legionary infantry, the additions
made to the cavalry were more striking still. An infinite number
of new bodies of horsemen, cunei, alae, vexillationes, etc., were
raised, alike for the limitary, the comitatensian, and the palatine
armies. Germans, Moors, Persians are more numerous amon"-

1 Hfrmes, 1889.

3zo] GROWING IMPORTANCE OF CAVALRY n

them than the born subjects of the empire. The old legionary
cavalry wholly disappears, 1 and the commands of horse and
foot are entirely separated. Yet under Constantine and his
immediate successors the infantry still remained the more impor
tant arm, though the cavalry was continually growing in relative
importance. When we read the pages of Ammianus Marcellinus,
we still feel that the Roman armies whose campaigns he relates
are the legitimate successors of the legions of Tiberius and Trajan,
though the names of the corps and the titles of the officers are
so greatly changed. In the last first-class victory which the
house of Constantine won over the barbarians Julian s great
triumph over the South German tribes near Strassburg it was
the infantry which bore off the honours of the day. The cavalry
were routed and driven off the field, but the foot-soldiery, though
their flank was uncovered, formed the testudo, beat off the
victorious German horse, and gained for their dispersed squadrons
the time to rally and retrieve the day. (357.)

Nevertheless, we find the cavalry continually growing in
relative numbers and importance. This is well marked by the
fact that when Constantine displaced the old Praefectus
Praetorio from his post as war- minister and commander -in -
chief under the emperor, he replaced him, not by a single
official, but by two a inagister peditum and a inagister equttuin.
By the time of the drawing up of the Notltia, the number of the
cavalry seems to have risen to about a third of that of the
infantry, whereas in the old Roman armies it had often been
but a tenth or a twelfth, and seldom rose to a sixth. The
figures of the Notitia show the results of the battle of Adrianople,
of whose military effects we have soon to speak. But long
before 379 the horse were high in numbers and importance.
The cause was twofold. The most obvious reason for the
change was that there was an increasing need for rapidly
moving troops. The Germans in the early fifth century
generally aimed at plunder, not at conquest. Comparatively
small bands of them slipped between the frontier posts, with
the object of eluding pursuit, gathering booty, and then making
their way homewards. It was as yet only occasionally that a
whole tribe, or confederation of tribes, cut itself loose from its
ancient seat, and marched with wife and child, flocks and herds
and waggons, to win new lands within the Roman border. To

1 Apparently under Constantine, as there are faint traces of it under Diocletian.

THE ART OF WAR IN THE MIDDLE AGES [35

hunt down and cut to pieces flitting bands of wary plunderers,
the fully-armed legion or cohort was not a very efficient tool.
The men marched with heavy loads, and were accompanied by
a considerable baggage train ; hence they could not, as a rule,
catch the invaders. Cavalry, or very lightly-equipped infantry,
alone were suitable for the task ; the mailed legionaries were as
ill-suited for it as were our own line-regiments to hunt down
the Pindaris of the Deccan in the present century.

But there was another reason for the increase in the numbers
of the cavalry arm. The ascendency of the Roman infantry
over its enemies was no longer so marked as in earlier ages, and
it therefore required to be more strongly supported by cavalry
than had been necessary in the first or second century. The
Germans of the days of the dynasty of Constantine were no
longer the half-armed savages of earlier times, who "without
helm or mail, with weak shields of wicker-work, and armed only
with the javelin," 1 tried to face the embattled front of the
cohort. Three hundred years of close contact with the empire
had taught them much. Thousands of their warriors had served
as Roman mercenaries, and brought home the fruits of ex
perience. They had begun to employ defensive armour ; among
the frontier tribes the chiefs and the chosen warriors of their
comitatus were now well equipped with mail-shirt and helmet.
The rank and file bore iron-bound bucklers, pikes, the short
stabbing sword (scramasax), as well as the long cutting sword
(spatho), and among some races the deadly francisca, or battle-
axe, which, whether thrown or wielded, would penetrate Roman
armour and split the Roman shield. As weapons for hand-to-
hand combat, these so far surpassed the old frainea that the
Imperial infantry found it no longer a light matter to defeat a
German tribe. At the same time, there is no doubt that the
morale of the Roman army was no longer what it had once been :
the corps were less homogeneous ; the recruits bought by the
composition - money of the landholding classes were often of
bad material ; the proportion of auxiliaries drawn from beyond
the frontier was too large. Nor can we doubt that the disasters
of the third century had left their mark on the soldiery ; the
ancient belief in the invincibility of the Roman Empire and the
majesty of the Roman name could no longer be held so firmly.
Though seldom wanting in courage, the troops of the fourth

1 See Tacitus, Annals, ii. 14.

378] THE BATTLE OF ADRIANOPLE 13

century had lost the self-reliance and cohesion of the old Roman
infantry, and required far more careful handling on the part of
their generals.

The end of this transitional period was sudden and dreadful.
The battle of Adrianople was the most crushing defeat suffered
by a Roman army since Cannae a. slaughter to which it is most
aptly compared by Ammianus Marcellinus. The Emperor
Valens, all his chief officers, 1 and forty thousand men were left
upon the field ; indeed the army of the East was almost
annihilated, and was never again its old self.

The military importance of Adrianople was unmistakable ;
it was a victory of cavalry over infantry. The Imperial army
had developed its attack on the great laager in which the Goths
lay encamped, arrayed in the time-honoured formation of
Roman hosts with the legions and cohorts in the centre, and
the squadrons on the wings. The fight was raging hotly all
along the barricade of waggons, when suddenly a great body of
horsemen charged in upon the Roman left. It was the main
strength of the Gothic cavalry, which had been foraging at a
distance ; receiving news of the fight, it had ridden straight for
the battlefield, and fell upon the exposed flank of the Imperial
host, " like a thunderbolt which strikes on a mountain top, and
dashes away all that stands in its path." s

There was a considerable number of squadrons guarding
the Roman flank ; but they were caught unawares : some were
ridden down and trampled under foot, the rest fled disgracefully.
Then the Gothic horsemen swept down on the infantry of the
left wing, rolled it up, and drove it in upon the centre and
reserve. So tremendous was their impact, that the legions and
cohorts were pushed together in helpless confusion. Every
attempt to stand firm failed, and in a few minutes left, centre,
and reserve were one undistinguishable mass. Imperial guards,
light troops, lancers, auxiliaries and legions of the line were
wedged together in a press that grew closer every moment, for
the Gothic infantry burst out from its line of waggons, and
attacked from the front, the moment that it saw the Romans
dashed into confusion by the attack from the flank. The
cavalry on Valens right wing saw that the day was lost, and

1 The grand masters of the infantry and cavalry, the count of the palace, and
thirty-five commanders of corps of horse or foot.
- Ammianus, xxi. 12.

14 THE ART OF WAR IN THE MIDDLE AGES [37

rode off without another effort, follo\ved in disorder by such of
the infantry corps on that side of the field as were not too
heavily engaged to be able to retire. Then the abandoned
foot-soldiery of the main body realised the horror of their
position : beset in flank and rear by the horsemen, and in front
by the mass which had sallied forth from the Gothic laager, they
were equally unable to deploy or to fly, and had to stand to be
cut down. It was a sight such as had been seen once before at
Cannae, and was to be seen once again, on a smaller scale, at
Roosbeke. Men could not raise their arms to strike a blow, so
closely were they packed ; spears snapped right and left, their
bearers being unable to lift them to a vertical position ; many
soldiers were stifled in the press. Into this quivering mass the
Goths rode, plying lance and sword against the helpless enemy.
It was not till two-thirds of the Roman army had fallen, that
the thinning of the ranks and the approach of night enabled a
few thousand men to break out, and follow the fugitives of the
right wing in their flight southward. (378.)

Such was the battle of Adrianople, the first great victory won
by that heavy cavalry which had now shown its ability to supplant
the heavy infantry of Rome as the ruling power of war. During
their sojourn on the steppes of South Russia, the Goths, first of
all Teutonic races, had come to place their main reliance on
their horsemen. Dwelling in the Ukraine, they had felt the
influence of that land, ever the nurse of cavalry from the day
of the Scythian to that of the Tartar and Cossack. They had
come to consider it more honourable to fight on horse than on
foot, and every chief was followed by his squadron of sworn
companions. Driven against their will into conflict with the
empire, whose protection they had originally sought as a
shelter against the oncoming Huns, they found themselves face
to face with the army that had so long held the barbarian world
in check. The first fighting about Marcianopolis and Ad Salices
in 377 was bloody, but inconclusive. Then, when Valens had
gathered all the forces of the East for a decisive battle, the day
of judgment arrived. The shock came, and, probably to his own
surprise, the Goth found that his stout lance and his good steed
would carry him through the serried ranks of the Imperial infantry.
He had become the arbiter of war, the lineal ancestor of all the
knights of the Middle Ages, the inaugurator of that ascendency
of the horsemen which was to endure for a thousand years.

380] THEODOSIUS SUBSIDISES THE GOTHS r 5

The battle of Adrianople had completely wrecked the army
of the Eastern Empire : Valens had stripped the Persian
frontier and the whole of Asia to draw together the great host
which perished with him. His successor Theodosius, on whom
devolved the task of reorganisation, had to restore the entire
military system of his realm. 1 He appears to have appreciated
to its full extent the meaning of the fight of Adrianople.
Abandoning entirely the old Roman methods of war, he saw
that cavalry must in future compose the more important half of
the Imperial army. To provide himself with a sufficient force
of horsemen, he was driven to a measure destined to sever all
continuity between the military system of the fourth and that of
the fifth century. After concluding a peace with the Goths c o
soon as he could bring them to reasonable terms, he began to
enlist wholesale every Teutonic chief whom he could bribe to
enter his service. The Gothic princes and their war-bands were
not incorporated with the Imperial troops or put under Roman
discipline: 2 they served as the personal retainers of the emperor,
whose " men " they became by making to him the oath of faith
ful service, such as they were wont to give to their own kings.
In return the princes received from the Caesar the atmonae
foederaticae, which they distributed among their horsemen.
Thus began the ruinous experiment of trusting the safety of the
empire to the Focderati, as the Gothic war-bands were non-
called : 3 for in their hands there lay the fate of the realm of
Theodosius, since they formed by far the most efficient division
of his army. From this moment the emperors had to rely for
their own safety and for the maintenance of order in the Roman
world, merely on the amount of loyalty which a constant stream
of titles and honours could win from the commanders of the
Foederati. No sufficient force of native troops was raised to
keep the Germans in check, and the remnants of the old national

1 I imagine that the enormous gaps in the numeration of the regiments of
the Eastern army in the Notitia largely proceed from the extermination of whole
corps at Adrianople. \Ve find, for example, of Sarmatian horse only Ala vn.
surviving, Ala \, Armeniorum is missing, and equites terlii Parthii, and nearly
all the regiments of the Zabdiceni and Cordueni. Of course other causes must
have extinguished many corps, but the slaughter at Adrianople was probably the
chief one.

2 See Jordanes, 28.

3 Hence they do not appear in the Notilia, though a few cohorts and alae of
Goths incorporated in the regular army are there to be found.

( 6 THE ART OF WAR IN THE MIDDLE AGES [393

army felt that they were relegated to a secondary place in the
scheme of military organisation.

Only six years after Adrianople there were already forty
thousand Gothic and other Teutonic horsemen serving under
their own chiefs in the army of the East It was on them that
Theodosius relied when a few years later he marched to
reconquer Gaul and Italy from the usurper Magnus Maximus.
In the two battles at Siscia and Aemona, which settled the
campaign of 387, he saw his confidence justified. On each
occasion the Roman army of the West, those Gallic legions
which had always been considered the best footmen in the world,
were finally ridden down and crushed by the Teutonic cavalry,
which followed the standard of the legitimate emperor. But
the West loved not to obey the East : there was a quasi-national
spirit of rage and resentment deep sunk in the breasts of the
Gallic legions : in 392 they rose again, murdered the young
Valentinian II., whom Theodosius had set over them, and tried
their luck once more against the Eastern emperor and his
hordes of Foederati. Under the nominal leadership of the
imbecile Eugenius, but really guided by a hardy soldier of
fortune named Arbogast, the Western armies faced Theodosius
at the battle of the Frigidus. They were beaten after a struggle
far more fierce than that of 387, 1 and again the chief part in their
defeat was taken by the twenty thousand Gothic horsemen who
formed the core of the host of Theodosius.

Henceforth the cavalry arm began to be as predominant in
the West as in the East If for a time the foot-soldiery of Gaul
and Britain maintained some of their ancient importance, it was
merely due to the fact that two Teutonic race.s which had not
yet taken to horsemanship the Franks and Saxons were at
once their most formidable adversaries and their favourite
recruiting ground. For in the Western no less than in the
Eastern realm the German mercenaries were for the future to
be the preponderant element in the Imperial army : the native
troops took a very secondary place. A glance down the lists of
military officers of high rank during the fifth century shows an
enormous numerical superiority of alien over Roman names. It
is true that since Constantine s day there had always been a large

1 So much more fierce, that the fortune of war ultimately leaned to Theodosius,
owing to the treachery of some of Eugenius officers rather than to the actual
fighting.

390] VEGETIUS AND THE ROMAN ARMY 17

sprinkling of half - Romanized barbarians among the corps
commanders the names of many of the generals in Ammianus
tell their own tale. 1 But it is only from the time of Theodosius
downwards that the alien names form the ever-increasing
majority. For some three generations after his death it is
hardly an exaggeration to say that the higher ranks in the army
were almost entirely in the hands of the Germans from the
day of Stilicho to that of Aspar and Ricimer. Aetius and
Marcellinus were the only first-class generals with Roman names
that we meet in the time : the rest are all aliens. It was but
natural, for the Foederati were the most important part of the
army, and they would not obey any leaders save their own
chosen chiefs and princes.

In the well-known treatise of Vegetius, De Re Militari, is
preserved a picture of the state of the Imperial army in the
Western provinces, painted probably in the time of Valentinian
II., and during his second reign in the West (388-392 ). 2 The
book would be of far greater value to us, if only Vegetius had
refrained from the attempt to describe things as they ought
to be instead of things as they were. He is far more con
cerned with the ancient history of the Roman legion, and with its
organisation, drill, and tactics in the days of its strength, than
with the degenerate corps that bore the name in his own day.
Instead of describing the army of A.D. 390, with its hordes of
Foederati, and its small legions and numeri, each only a
thousand strong, Vegetius persists in describing the army of the
early empire, when all the legions were five or six thousand
strong, and still formed the most important element in the
Imperial host. Apparently it was his wish to induce the young
Emperor Valentinian, for whose instruction he wrote, to restore
the ancient discipline and organisation. Accordingly we con
tinually find him describing the ideal and not the actual, as
is proved by his frequent confessions that " this custom has
long been extinct," or that " only part of these exercises are
now wont to be used."

1 e.g. Daglaif, Rhoemetalces, Hormisdas, Fullofaudes, Vadomar, Merobaudes
Nevitta, Immo, Agila, Malarich.

2 I am inclined to hold that the De Re Militari belongs to the time of
Valentinian n., and not, as many good authorities think, to that of Valentinian III.
In the days of the latter the whole military system had so far gone to pieces that
it is incredible that even an archaeologist like Vegetius should have described it in the
terms which he uses. But in 388-392 it was still holding together

i8 THE ART OF WAR IN THE MIDDLE AGES [39

Vegetius was a theoretical admirer of the old legion, and
wholly destitute of any insight into the meaning of the change
in military science which had taken place during the last
hundred years. His explanation of the decadence of the
Roman infantry is founded on a story that we can prove to be
untrue. " From the days of the Republic," he writes, " down to
the reign of the sainted Gratian, the Roman foot-soldiery bore
helm, cuirass, and shield ; but in Gratian s time regular drill and
exercise were gradually abandoned through negligence and
idleness. The soldier ceased to wear his armour habitually,
and grew to find it heavy when the time came to assume it.
Wherefore the men begged leave from the emperor first that
they might abandon the use of the cuirass, and then that of the
helm. So our soldiery went out with breast and head un
protected to meet the Goths, and perished beneath their missiles
on countless battlefields. And after so many disasters, and the
sack of so many great cities, no commander has yet been able
to persuade them to resume the salutary protection of helmet
and cuirass. So when our men, destitute of all defensive arms,
are drawn up for battle, they think of flight more than of victory.
For what can the footman armed with the bow, without helm or
breastplate, and even unable to manage shield and bow at once,
expect to do? ... Thus, since they will not endure the toil of
wearing the ancient armour, they must expose their naked
bodies to wounds or death, or what is worse surrender,
or betray the State by disgraceful flight. And the result is,
that, rather than bear a necessary toil, the} resign themselves
to the dishonourable alternative of being slaughtered like
sheep." l

Here Vegetius always more of a rhetorician than a soldier
has inverted cause and effect in the strangest fashion. It was
true that by his own day the Roman infantry had for the most
part become light troops and abandoned their armour. It was
true also that the change had begun about the time of Gratian,
for that emperor was reigning in the West when the disaster of
Adrianople destroyed the army of the East. But all else in the
story is obviously absurd and untrue. The Imperial foot-soldiery
were still wearing the full ancient panoply when it first met the
Goths. Ammianus, a strictly contemporary writer, twice speaks
of the defensive armour of the legions during his account of the

1 Vegetius, i. 20.

400] GROWING IMPORTANCE OF ARCHERY 19

battle of Adrianople. 1 More than ten years later the anonymous
writer on military equipment who dedicated his little work to
the three Augusti Theodosius, Arcadius, and Honorius takes
the breastplate for granted, when he gives some advice as to
thick underclothing to be worn beneath it for campaigning in
the winter or in cold and damp regions. 2 Ten years later,
the Roman soldiery on the column of Arcadius were still repre
sented in helm and cuirass.

It is of course ludicrous to suppose that, at a time when the
cavalry were clothing themselves in more complete armour, the
infantry were discarding it from mere sloth and feebleness. The
real fact was that the ancient army of mailed legionaries had
been tried in the battlefield and found wanting. In despair of
resisting the Gothic horsemen any longer by the solidity of a
line of heavy infantry, Roman military men had turned their
attention to the greater use of missile weapons for the foot-
soldiery, and to developing the numbers and efficiency of their
own cavalry. The scientific combination of bow and lance
against brave but disorderly swarms of horse was a fair device
enough as was to be shown a thousand years later on the fields
of Falkirk and Crecy.

If the new tactics failed first against the Goths of Alaric and
then against the Huns of Attila, their want of success must not
be attributed to their own intrinsic faultiness. The armies of
Honorius and Arcadius and their successors were generally
beaten because they were composed partly of untrustworthy
and greedy Teutonic Foederati, fighting for pay and plunder,
not for loyalty, and partly of native troops discouraged and
demoralised by being slighted and taught to consider them
selves inferior to their barbarian comrades. In the hands of a
Stilicho or an Aetius the Imperial army could still do some
good fighting. But it was more usually under the command
of self-seeking mercenaries or incapable court favourites, and
gradually sank from bad to worse all through the fifth century.
The deterioration was inevitable : as the Teutonic auxiliaries
grew more and more convinced of the weakness and impotence

1 (i) The heat of the day, "Romanes attenuates inedia silique confectos, ct
armorum gravantibus sarcints, exurebat. " (2) The lines of infantry close, and nostri
occursantes gladiis obtruncant : mutuis ictibus galeae perfringebatitur ct loricae." 1

2 Being dedicated to Theodosius and his two sons as joint Augu= .i, the work r.r, st
have been written in the years 394-395-

THE ART OF WAR IN THE MIDDLE AGES [406

of their masters, they became progressively greedier and more
treacherous. As the native troops saw the empire falling
deeper into the slough, they lost all self-respect and all hope
of victory, and as Vegctius complained came to battle with
their mind s fixed on discovering the safest and easiest line of

retreat.

In the reigns of Honorius and Arcadius the Roman army
finally ceased to be a regular and organised body. The Notitia
Digmtatum, a document drawn up during their joint reign,
somewhere about 406, still shows us the old arrangements
surviving. We find that many of the Flavian cohorts and numeri,
and many even of the legions of the early empire are still
surviving, though they are well-nigh swamped by the scores of
new barbarian corps, with extraordinary, magniloquent, and
sometimes grotesque 1 names, Honoriani and Theodosiani and
Valentiniani and Arcadiani, and so forth, not to speak of
regiments which more clearly betray their nationality cohorts
and alae of Chamavi or Juthungi, Franks, Alamanni, Taifalae,
Goths, and Alans (406-409). But chaos may be said to have set
in with the invasion of Alaric and the contemporary civil wars
caused by the subsequent rebellions of Constantine in Britain
(407-411), Maximus in Spain (41 1), and Jovinus and Sebastianus
on the Rhine frontier (411-412).

It was in these evil days, while the imbecile Honorius was
skulking behind the walls and marshes of Ravenna, that the
final disorganisation of the Imperial forces took place, and most
of the old native corps disappeared. It was not till the day of
Alaric that Italy came to know thoroughly the Gothic horsemen
whose efficiency Constantinople had already comprehended and
had contrived for the moment to subsidise. But now the Goth
became the terror of Rome, as he had previously been of the
East. His lance and steed once more asserted their supremacy :
the generalship of Stilicho, the trained infantry of the old
Western army, light and heavy, the native and Foederate cavalry
whose array flanked the legions, were insufficient to arrest the
Gothic charge. The last chance of salvation vanished when
Stilicho was murdered by his ungrateful master, and then the
conquerors rode at their will through Italy and sacked the
Imperial city herself. When they quitted the peninsula, it was

1 e.g. Leones Seniores, Ursi Valentiniani, promoti braccati seniores, Mauci
toiuntes, etc.

450] THE BATTLE OF CHALONS 21

by their own choice, for there were no troops left in the world
who could have expelled them by force (A.D. 409).

The day of infantry indeed was now gone by in Southern
Europe : they continued to exist, not as the core and strength
of the army, but as a subsidiary force used as light troops
in the day of battle, or to garrison fortresses, or to penetrate
woods or mountains where the horseman could not pierce his
way. Roman and barbarian alike threw their vigour into the
organisation of their cavalry.

This tendency was only emphasised by the appearance on the
Imperial frontier of the Huns, a new race of horsemen, formidable
by their numbers, their rapidity of movement, and the constant
rain of arrows which they would pour in without allowing their
enemy to close. In their tactics they were the prototypes of the
hordes of Alp Arslan, of Genghiz, and of Tamerlane. The in
fluence of the Huns on the Roman army was very marked :
profiting by their example, the Roman trooper added the bow to
his equipment ; and in the fifth century the native force of the
empire had come to resemble that of its old enemy the Parthian
state of the first century, the choicer corps being composed of
horsemen in mail armed with bow and lance. Mixed with these
horse-archers fought squadrons of the Teutonic Foederati, armed
with the lance alone. Such were the troops of Aetius and
Ricimer, the army which faced the Huns on the plain of Chalons.

That decisive battle was pre-eminently a cavalry engagement.
On each side horse-archer and lancer faced horse-archer and
lancer Aetius and his Romans leagued with Theodoric s Visi-
gothic chivalry Attila s hordes of Hunnish light horse backed
by the steadier troops of his German subjects, the Ostrogoths,
Gepidae, Heruli, Scyrri, and Rugians. The Frankish allies of
Aetius must have been the largest body of foot-soldiery on the
field, but we hear nothing of their exploits in the battle. 1 The
victory was won, not by superior tactics, but by sheer hard fight
ing, the decisive point having been the riding down of the native
Huns by Theodoric s heavier Visigothic horsemen (A.D. 450).
It was certainly not the troops of the empire who had the
main credit of the day.

1 Jordanes tells us, however, that the Franks had a bloody engagement with Attila s
Gepidae on the night before the battle, in which fifteen thousand men fell on the two
sides. There were no doubt many infantry in the host of Aetius. In Attila s harangue
before the battle Jordanes makes him bid the Huns despise the " testudines " of the
Romans, i.e. their infantry formed in solid masses.

CHAPTER II

COMMENCEMENT OF THE SUPREMACY OF CAVALRY.
EELISARIUS AND THE GOTHS

A.D. 450-552

7^O trace out in further detail the meaning of the wars of the
fifth century is unnecessary. But it must be observed
that, as the years of its middle course rolled on, a divergence
began to be seen between the tendencies of the Eastern
and the Western Empire. In the West the Foederati became
the sole military force of any importance. One of their chiefs,
the Suevian Ricimer, made and unmade emperors at his good
pleasure for some twenty years. A little later, another, the
Scyrrian adventurer Odoacer, broke through the old spell of
the Roman name, dethroned the last emperor of the West, and
ruled Italy as a Teutonic king, though he thought well to
legalise his usurpation by begging the title of Patrician from
Zeno, the emperor at Constantinople (476 A.D.).

In the East the decline of the native troops never reached
the depth that it attained in the West, and the Foederati never
became masters of the situation. That Byzantium did not fall a
crey to a Ricimer or an Odoacer seems mainly to be due to the
Emperor Leo I. (457-474), who took warning by contemporary
events in Italy, and determined that even at the cost of military
efficiency the native army must be kept up as a counterpoise
to the Teutonic auxiliaries. He unscrupulously slew Aspar, the
great German captain whose preponderance he dreaded, though
he himself owed his throne to Aspar s services. At the same time
he increased the proportion of Romans to Foederati in his hosts.
His successor Zeno (474-491) continued this work, and made
himself noteworthy as the first emperor who properly utilised
the military virtues of the Isaurians the rough and hardy pro-

489] ZENO AND THE ISAURIANS 23

vincials of the southern mountains of Asia Minor. 1 These wild
highlanders had hitherto been looked upon as intractable and
troublesome subjects. Zeno showed that their courage could be
employed to defend instead of to plunder their more quiet
neighbours. He dealt with them as William Pitt dealt with the
Celts of the Scottish hills thirteen hundred years later formed
them into numerous regiments and taught them to become
soldiers instead of mere cattle-lifters. Zeno also enlisted
Armenians and other inhabitants of the Roman frontier of the
East, and handed over to his successor an army in which the
barbarian element was adequately counterpoised by the native
troops. He had done another good service to the empire by
inducing the Ostrogoths, the most formidable of his Teutonic
auxiliaries, to migrate en masse to Italy. It would have been
an evil day for the East if Theodoric, after routing so many of
Zeno s generals and ravaging so many of his provinces, had
determined to stay behind in the Balkan peninsula. But, moved
by the emperor s suggestions and sent forth with his solemn
sanction, the Ostrogoth led off his people to win a new home,
and left Moesia and Macedonia ravaged and ruined indeed, but
free of barbarian settlers (489).

Under the comparatively peaceful reigns of Zeno s successors,
Anastasius and Justin (491-527), the Eastern Empire was able to
recover a considerable measure of strength, both military and
financial. A small pamphlet which has come down to us from
this time shows us how entirely the strength of its army now
lay in the cavalry arm. A certain Urbicius a tactician of the
closet, not a practical soldier dedicates to the Emperor
Anastasius "an original device to enable infantry to resist horse
men." Prefacing his remarks by a statement that a new theory
of the defensive is needed to meet the conditions of the day, he
proposes to resuscitate the ancient Macedonian phalanx. But
the projecting barrier of pikes, which formed the essential feature
of that body, is not to be composed of the weapons of the soldiery
themselves. The men are to retain their equipment with the
bow and javelin for apparently the whole Roman infantry were
by this time furnished with missile weapons. But each decury is
to take with it a pack-horse loaded with short beams set with
spear-blades. When the enemy comes in sight, the beams are
to be hastily placed in line before the front of the corps, so as to

1 Diocletian, however, had raised two Isaurian legions, which appear in the Notilia,

24 THE ART OF WAR IN THE MIDDLE AGES [s

form a continuous barrier of chevaux-de-frise. If the ground is
open, and attack may be expected from all sides, the infantry
are to range themselves in a hollow square, covered on all sides
by the spikes and beams. " The barbarians charging with their
usual headlong impetuosity, the dicraitx-de-frise will bring them
to a sudden stop, then the constant rain of missiles from our men
will strike down rank after rank before they can overturn the
machines, and they will infallibly be routed, more especially if the
corners of the square are strengthened with the balistae J which
each corps carries with it."

The weak points of this rather childish device are at once
obvious. It presupposes that the infantry will always have
time to form square, and that every pack-horse s burden will be
unloaded with equal celerity for obviously a single break in the
continuity of the line of obstacles would be fatal. Moreover, it
condemns the troops using it to complete immobility; their
square once formed, they cannot move, and must remain
rooted to the spot as long as the enemy has a single unbroken
squadron left. Moreover, if the barbarians under cover of a
charge send parties of dismounted men to pull away a few of the
chevaux-de-frise,\\. is practically certain that they must succeed at
some point or other. At the best the device only aspires to pre
serve the troops who use it from being cut to pieces it cannot
enable them to take the offensive, and an army condemned to
an eternal defensive can never deal a decisive blow.

As a matter of fact, the experiment was never tried, and the
army of the East continued to depend for victory on its horse
men, native and Foederate. By a fortunate chance, the wars of
the generation which followed that of Urbicius and his master
Anastasius are described to us in great detail by a capable and
observant eye-witness, Procopius. From him \ve learn all that
we can wish to know about the East-Roman army its disposi
tion, organisation, and tactics during the second and third
quarters of the sixth century.

The victorious hosts of Justinian, which reconquered for the
empire Italy, Africa, and Southern Spain, were composed in
about equal proportions of foreign auxiliaries serving under their
own chiefs and of regular native troops. The Foederati were

1 Large machines on the principle of the crossbow, each worked by several men and
throwing a heavy bolt to three times the distance that a javelin carries, as Urbicius is
careful to explain.

53] THE ROMAN HORSE-ARCHER 25

still mainly Teutonic Gepidae, Heruli, and Lombards; but there
was a not inconsiderable intermixture of Huns and a certain
number of Armenians among them. The native corps were
partly surviving numeri xaTa/.6-/oi is Procopius name for them
of the old standing army ; 1 but to these were added many new
bodies, raised for a particular service or emergency by officers to
whom the emperor gave a grant of permission to gather men.
This was something like the English mediaval system of com
missions of array still more like the seventeenth - century
arrangement by which a Wallenstein or a Mansfeld gathered
mercenaries under royal sanction, but by the attraction of his
own name.

Both among the Foederati and among the native corps the
cavalry were by far the more important arm. The mailed
cataphracti or cuirassiers of the Asiatic provinces win the special
admiration of Procopius. The paragraph in which he indicates
the superiority of the horse-archer of his own day over the
ancient infantry is so characteristic that it is worth reproducing.

" Men there are who call our modern soldiery mere bow
men, and can praise only the troops of old, the shielded
legionaries who fought hand to hand with the foe. They lament
that our ancient warlike courage has disappeared in these days,
and thereby show themselves to be mere ignorant civilians.
They say that bowman was from the earliest times a term of
contempt, not remembering that the archers of Homer s day
for of them they are thinking were light troops without horse,
lance, shield, or defensive armour, who came on foot to the battle
and skulked behind a comrade s shield or took cover behind a
stone. Such archers of course could neither defend themselves
adequately nor set upon the enemy with confidence : they were
mere furtive hoverers on the edge of battle. Moreover, they were
such weak and unskilled shooters that they only drew the bow
string to the breast, so that the arrow flew aimlessly and pro
bably did no harm.

" Now our horse-archers are very different men. They come
to the fight cuirassed and greaved to the knee. They bear bow
and sword, and for the most part a lance also, and a little shield
slung on the left shoulder, worked with a strap, not a handle.
They are splendid riders, can shoot while galloping at full speed,
and keep up the arrow-flight \vith equal ease whether they are

1 We hear of numeri still, but no longer of legions- all of them had disappeared.

26 THE ART OF WAR IN THE MIDDLE AGES [53

advancing or retreating. They draw the bow-cord not to the
breast, but to the face, or even to the right ear, so that the missile
flies so strongly as always to inflict a deadly wound, piercing
both shield and cuirass with ease. Yet there are men who in
antique prejudice despise our horse-archers, out of mere ignor
ance and folly. For it is clear and obvious that the grandest
military results in the wars of our own day have been attained
by the use of this very arm." 1

The professional soldiers of the sixth century were, in fact,
entirely satisfied with the system of cavalry tactics which they
had adopted, and looked with a certain air of superiority on the
infantry tactics of their Roman predecessors. They thought
that a cavalry force could be almost self-sufficient, if to the native
horse-archer were joined the heavier squadrons of the subsidised
Foederati, Lombards, Heruli, or Gepidae, led by their own princes
and armed with the lance. The one could act as light troops, the
other as supports, so that the infantry would hardly be needed
save for garrison duty or service in woods, mountains, or morasses
where the horseman could not penetrate. There was a certain
amount of justification for this belief; the hard-fought battle of
Daras in the first Persian war was mainly won by the cavalry.
The still more decisive victory of Tricameron, which made an
end of the Vandal power in Africa, was fought and won by the
horse alone ; the infantry were a march behind, and only arrived
in the evening when the battle was over.

Justinian s army and its achievements were not unworthy
of the praise which Procopius lavishes upon it : its victories
were its own, while its defeats were generally due to the wretched
policy of the emperor, who persisted in dividing up the
command among many hands a system which secured
military obedience at the cost of military efficiency. Justinian
might, however, plead in his defence that the organisation of
the army had become such that it constituted a standing menace
to the central power. The system of the Teutonic comitatus, of
the " war-band " surrounding a leader to whom the soldiers are
bound by a personal tie, had become deeply ingrained in the
Imperial forces. Always predominant among the Foederati, it
had spread from them to the native army, owing to the system
by which distinguished officers were now allowed to raise corps
of their own for t he Imperial service, instead of being merely

1 De Bella Persia, I. i. 25-40.

530] BELISARIUS VICTORY AT DARAS 27

promoted to the command of old existing units. In the sixth
century the monarch had always to dread that the loyalty of the
troops towards their immediate commanders, in whose name they
had been levied, might prevail over their higher duties. For
generals of note came to be surrounded by bands of retainers
of a very dangerous size and temper, when they were allowed to
take into their own bodyguard any soldier of the line who
distinguished himself in action. Belisarius and even the eunuch
Narses were surrounded by large bodies of these devoted com
panions. 1 The personal followers of the former at the time of
his Gothic triumphs amounted to no less than seven thousand
veteran horsemen : it was no wonder that the Romans exclaimed
that " the household of a single man has overthrown the kingdom
of Theodoric."-

The existence of such corps of retainers rendered every
successful commander a possible Wallenstein to use a name
of more modern significance. Thus the emperor, in his desire
to avert the predominance of any single officer, would join
several men of discordant views in the command of an army
usually with disastrous consequences. This organisation of the
Imperial forces in " bands," B bodies attached by personal ties
to their leaders, is the characteristic military form of the sixth
century. Its normal prevalence is shown by the contemporary
custom of speaking of each corps by the name of its command
ing officer, and not by any official title. Nothing could be more
opposed than this usage to old Roman custom. 4

How entirely the efficiency of Justinian s army depended on
the combination of heavy cavalry with the bow, can best be
shown by a short description of the three chief victories which
it won in East and West over its most important foes;

Earliest in date is the battle of Daras (530), in which
Belisarius won his first decisive victory. Daras was an
important frontier fortress which was threatened by a Persian
army of forty thousand men. Belisarius had gathered about
twenty-five thousand to prevent the siege being formed. He

1 Procopius, De Bella Gotlhico, III. i.

* Procopius calls them SopvQopot and viraairitrrai. The usual Latin word for them
was Biifcellarii, from Rucccllum, the ration-biscuit, meaning retainers fed by their lord.

3 ptivdov is used by Procopius both for the standard of the regiment, and for the
regiment itself.

4 e.g., where Ammianus would still talk of the "cohors quarta Thracum," Pro
copius would call them "that catalogue of Thracians which Bryes led."

2 S THE ART OF WAR IX THE MIDDLE AGES [S3

put them in array close outside the city, so as to get easy pro
tection if he were beaten. The centre, composed mainly of foot,
was much drawn back and " refused " ; the wings, composed of
horse in equal strength, were thrown forward. To prevent^ a
breach of continuity between centre and wings, a reserve of six
hundred chosen Foederate cavalry (Huns) was placed at each
flank of the infantry, charged with the duty of supporting the
cavalry wing to which it was nearest. Behind the infantry was
the general and his personal bodyguard of cuirassiers. The
whole front of the line was protected by a ditch, broken by
many open passages left for the free exit or retreat of regiments
moving forward and back. That it was not a very formidable
obstacle is shown by the fact that both sides crossed it without
difficulty more than once in the day. One flank of the whole
line was covered by an isolated hill ; that the other had any
such protection we are not told.

The Persians came on in two lines apparently, like the
Romans, with horse on the flanks and foot in the centre ; but
this is not expressly stated, though we know that the hard
righting was all done by the former. The infantry were, as
Belisarius remarked, " half-trained rustics, only good for trench
work and long shooting." On the first day there was an
indecisive skirmish, on the second a pitched battle.

When the Persians advanced, they came into contact with
the Roman wings, but not with the " refused " centre, which was
so far drawn back that only arrow-fire was here exchanged when
the two cavalry divisions on the flanks were already heavily
engaged. On the Roman left the Persians made some impression
at first ; but when they had pushed forward beyond the trench,
they were charged in flank by the reserve of Hunnish cavalry
from the left of the line of infantry. At the same time a small
body of Herule Foederati, which had lain hid on the isolated hill,
charged them in the rear. They broke and retreated, but did
not disperse or leave the field. The Romans re-formed in their
first position.

On the right meanwhile the Persian attack had been far more
formidable ; their commander had placed there the famous corps
called the " Immortals " and the pick of his other horsemen. In
the first charge they drove the Roman cavalry right back to
the gates of Daras. But in so doing the victorious squadrons
became separated from their own centre, which was now engaged

PLATE I.

Henih

Rtuazes

BATTLE OF
PARAS

A.D. 530.

Roman

Persian:

Foot r^~i

Horse

I ;

Foot
Horse

-Bar* small es

A Heruli

^Xf_

Gothic Horse

Gothic Foot

Roman Horse I
Foot

i ic Horse
Foot

BATTLE OF TAG1 N ^g. .A.D.552.

535] THE BATTLE OF TRICAMEROX 29

in a duel of missiles with the Roman infantry behind the trench.
Into the gap between the centre and the victorious wing
Belisarius threw first the six hundred Huns who flanked his
infantry on the right, then the similar body from the left, which
he recalled the moment that the danger on that flank was
ended. He himself with his bodyguard followed. Charged in
flank and rear by these fresh troops, the Persian left wing fled
away diagonally, in a direction which completely separated
them from their own centre. Leaving the rallied right wing to
pursue the fugitives, Belisarius now threw his Huns and body
guard against the exposed flank of the Persian centre. The
infantry there stationed at once broke and fled, and suffered
horrible slaughter. For the rest of the war the Persians never
again would face the Roman host in the open for a pitched
battle.

The main tactical point to be noticed in this fight is the
deliberate purpose of Belisarius to keep his infantry out of the
stress of the fight, and to throw all the burden of the day upon
the horse. This was accomplished by " refusing " the centre
and protecting it with the ditch, while the wings were thrown
forward and so placed as to draw upon themselves the chief
impact of the enemy. As the Persian had also strengthened
his wings, all went as Belisarius desired, and the infantry in the
centre hardly came to blows at all. If the hostile commander
had adopted the opposite plan, that of reinforcing his centre
and making his chief assault on the corresponding part of the
Roman line, Belisarius would have been able to stop him by
charging from the flank with his cavalry on to the Persians,
when they had passed the level of his wings and had got into
the hollow space in front of the " refused " line of infantry.

Of the two fights which settled the Vandal war we need say
little ; that of Ad Decimum was a mere " chance medley,"
fought without premeditation in a series of isolated combats.
It is only noteworthy that the day was mainly won by the
charge of the Hunnish light cavalry. The second and decisive
battle, that of Tricameron, was a pure cavalry engagement.
The infantry was a march to the rear when Belisarius found
the Vandal host drawn out to oppose him. In spite of this, the
great general resolved to fight at once ; he placed his Foederate
horse on one wing, his regular native regiments on the other,
and his own bodyguard, the pick of the army and now several

30 THE ART OF WAR IN THE MIDDLE AGES [535

thousand strong, in the centre. The front was covered by a
small stream, which he hoped that the Vandals might be
induced to cross, purposing to charge them just at the moment
when they should be labouring through it. But King Geilamir
would not take the offensive, and remained unmoved beyond
the water. Belisarius sent several small detachments across the
brook, to harass the hostile centre and induce it to charge and
assault him. But the Vandals contented themselves with
throwing out slightly larger bodies of horse, which drove the
Romans back over the water, but refused to cross it in pursuit.
Seeing the enemy grown so cautious, Belisarius concluded that
they had lost their morale after their previous defeat at Ad
Decimum, and might be dealt with summarily. Accordingly he
bade his own centre cross the brook and advance for a serious
attack. The Vandals thronged around it and gave the general s
bodyguard very hard work for some minutes. But when all
their attention was engrossed in the attempt to surround and
destroy the Roman centre, Belisarius let loose his two wings and
bade them cross the brook and do their best. Unprepared for
a general assault all along the line, and apparently caught in
flank while endeavouring to encompass the Imperial centre, the
wings of the Vandal army broke at the first impact of the
enemy. Their flight uncovered their comrades of the middle
corps, who were nearly all cut to pieces, together with their
commander Tzazo, the king s brother. Geilamir himself played
a poor part, made no effort to rally his men, and escaped by the
swiftness of his horse ("535).

So ended the Vandal kingdom, wrecked in less than an hour
of cavalry fighting. The lesson of the fight was simply that in
a duel between two bodies of horse, the one which adopts a
passive defensive, and receives the enemy s charge at a halt, will
be scattered, in spite of a decided superiority in numbers.
Geilamir s obvious duty was to charge the Roman centre while
it was hampered in crossing the brook. He refused allowed
himself to be attacked, and lost the day. A similar example on
a small scale was seen in the English heavy cavalry charge
at Balaclava, thirteen hundred years later. There, too the
stronger _force of cavalry chose to stand still to receive an
attack : it bore up for some time against the frontal assault
of the Scots Greys and Inniskillings, but broke at once and
fled in disastrous confusion when its flanks were chared a

538] BELISARIUS AND THE GOTHS 31

few minutes later by the Royals and 4th and 5th Dragoon
Guards.

The Gothic war, the greatest of the three struggles waged by
Justinian, was essentially a war of sieges and not of battles. In
the first half of it, indeed, down to Belisarius capture of Ravenna,
there was no single general engagement between the Goths and
the Imperialists. The decisive event of this part of the struggle
was the long beleaguering of Rome, from which the Goths retired
foiled, partly because of their own unskilfulness in siegecraft,
partly because of the deadly fever of the Campagna, which had
thinned their ranks. But if the sieges were the chief events
in the struggle of A.D. 535-40, there were a good many skirmishes
and minor engagements which served to display the qualities and
tactics of the two armies. A glance cast round them shows that
on both sides the cavalry did almost all the fighting, and would
seem to have been the larger half of the host. 1 Infantry were, in
fact, so little used by Belisarius, that we read that during the
third year of the war - many of them procured themselves horses,
and learned to serve as light cavalry. On one occasion the com
manders of the Isaurian archers, who formed the choicest part of
the foot-soldiery, came to the general complaining bitterly of
being kept out of the best of the fighting. Belisarius therefore
gave them a prominent part in his next sortie, more (we are told)
to conciliate such gallant soldiers, than because he thought it
wise to put them in the forefront of the battle. The result was
not happy for the infantry : they were shaken by the headlong
flight of a party of their own horse, who rode through their
ranks and put them into confusion. Then the Goths fell on
them and routed them : the two officers, Frincipius and Tarmutus,
who had counselled the sortie, were both slain while trying to
rally their broken troops. 3 The event of the fight only served
to confirm Belisarius in his belief in the absolute superiority of
cavalry.

The great general s own verdict on the military meaning of
the war has fortunately been preserved to us. On one occasion
during the siege of Rome, 4 some of his officers asked him how
he had dared to attack the Gothic power with such a small
army, and wished to know the causes of the confidence in his

1 On one occasion we find a force composed of 4500 horse, and only 5000 foot.

2 Procopius, DC Bell. Gott. \. 28.

Ibid. i. 29. " Ibid. i. 27-

3 2 THE ART OF WAR IN THE MIDDLE AGES [552

final success which he had always shown. Belisarius answered
as reported by Procopius, who was himself present, in the follow
ing terms : " In the first small skirmishes with the Goths, I was
always on the look-out to discover what were the strong and
weak points in their tactics, in order to accommodate my own to
them, so as best to make up for my numerical inferiority. I
found that the chief difference between them and us was that
our own regular Roman horse and our Hunnish Foederati are all
capital horse-bowmen, while the enemy has hardlyany knowledge
whatever of archer}-. For the Gothic knights use lance and
sword alone, while their bowmen on foot are always drawn up
to the rear under cover of the heavy squadrons. So their horse
men are no good till the battle comes to close quarters, and can
easily be shot down while standing in battle array before the
moment of contact arrives. Their foot-archers, on the other
hand, will never dare to advance against cavalry, and so keep too
far back." Hence there was no coherence between the two
arms in the Gothic host ; the knights were always wanting to get
to close quarters, while the bowmen preferred long shooting, and
were nervously anxious not to be exposed to a cavalry charge.
Thus it generally came to pass that the former, teased by
the Roman arrows, were always making reckless and premature
charges, while the latter, when they saw the horsemen beaten,
absconded without thinking for a moment of retrieving the
battle.

The clear-sightedness of Belisarius, and his complete apprecia
tion of the weak point of the Gothic host, is best shown by a
short account of the one great pitched battle which distinguished
the war, though in that engagement the great general himself
was not present. The fight of Taginae (552), which finally
brought the struggle to an end, was won by the eunuch Xarses,
who, in spite of his training as a mere court chamberlain, showed
military talents not inferior to Belisarius own. His triumph
was all the more striking because the Goths were now headed,
not by the slow and incapable Witiges, with whom Belisarius
had to deal, but by King Baduila, a gallant and experienced
soldier, who had beaten the East-Romans in a score of minor
fights, and thoroughly knew the tactics and methods of his
adversaries.

Taginae lies just below the central watershed of the
Apennines, near the modern Gubbio. The Goth had wished

552] THE BATTLE OF TAGINAE 33

to defend the mountain-line, but while he guarded the main pass,
Narses slipped over by a side path, and appeared on the lower
spurs of the western side of the range, at the head of the narrow
valley down which runs the Chiascio, one of the affluents of the
Tiber. Baduila arrived in time to seize the outlets of the valley,
and to draw up his army so as to force Narses to fight, or else
to make a perilous retreat back over a difficult pass, and in the
face of a daring enemy. The scene of the battle was a small
upland plain pressed in between the hills, with a breadth of
perhaps two milts of ground suitable for the movement of
cavalry. The two armies seem to have stretched across the
level ground on an equal front, though the Imperialists had a
considerable superiority in numbers. In front of the extreme
left of Narses position there was a small steep isolated hill which
would have given good cover for an attack on that flank of his
army. This he occupied with a small body of infantry ; on the
night before the battle the Gothic king tried to seize it, but the
squadron of horse which he sent forward for that purpose could
not make its way up the steep path which led to the summit of
the mound, and was driven down with loss.

In accordance with Gothic custom, Baduila put all his con
fidence in his horsemen, who seem to have formed a good half
of his host. They included all the flower of his nation, and were
strengthened by many hundreds of German mercenaries who
had, at one time and another, deserted the Imperial standards in
order to serve under a leader in whom they recognised the last
of the hero-kings of old. Baduila ranged his horsemen in the
front line ; the whole of his infantry, mostly archers, formed a
second line in his rear. It was his purpose to carry all before
him by a single charge there was to be no skirmishing or slow
advance, but by a sudden unexpected onslaught he hoped to
break through the Roman centre, where, as he could see with
his own eyes, there appeared to be only infantry opposed to
him. It was his object to get at the enemy as quickly as possible,
in order to avoid the showers of arrows which were the strongest
defence of the Imperialist troops. Delaying his attack all the
morning, he suddenly hurled his whole army forward at the
time of the midday meal, when he hoped to find Narses off his
guard.

To meet the Gothic attack, the eunuch-general had adopted
an order of battle which seems to have been of his own invention ;

34 THE ART OF WAR IX THE MIDDLE AGES [55 2

at any rate it had not been hitherto employed by any general
in the wars of that age. He had composed his centre of the
pick of his Foederate troops, eight or ten thousand Lombards,
Gepidae, and Heruli, whom he had ordered to dismount from
their horses and use their lances on foot. This employment of
mailed horsemen as infantry recalls King Edward lll. s device
at Cregy ; still more so does the rest of N arses battle-array,
for on each flank of the dismounted Foederati he had ranged
his Roman foot-archers, four thousand on each wing ; they were
slightly advanced in a curved half-moon, so that an enemy
advancing against the centre would find himself in an empty
space, half encircled by the bowmen and exposed to a rain of
arrows from both sides. To protect the archers, the native
Roman horse-soldiery, not dismounted, were arrayed immedi
ately in their rear. Finally, on the left wing, where the isolated
hill already described projected in front of the line, two detached
bodies of cavalry were stationed, thrown out at an angle from
the main line. The object of these was to deliver a side attack-
on the Gothic infantry, if it should advance close in the rear
of its horse, and so expose itself to being rolled up from the
flank.

The peculiarity of this formation was the combination of
heavy masses of dismounted cavalry, armed with the lance and
arrayed in close phalanx, with flanking bodies of archers. In
fantry had so long given up any idea of resisting horse by a
level front of spears, that Baduila seems to have had no idea of
the strength of the tactics that were opposed to him. Even the
historian who wrote the tale of the campaign ascribes a political
and not a military purpose to Narses order of battle. Procopius
tells us that he distrusted the Lombards and Gepidae, thinking
that they might retire, or even join the enemy, because of their
sympathy and admiration for Baduila, and that he dismounted
them to prevent their moving. But this very inadequate reason
is evidently not the true one, for at Casilinum, the other great
victory of the eunuch-general, a similar order was employed
when there was no question of disloyalty among the Foederati.

At midday the Gothic king suddenly bade his horsemen
charge ; they made for the hostile centre, leaving the wings of
archers alone a terrible mistake, much like that which the
French knights committed at Cre9y. For when they reached
the centre of the semicircle formed by the Roman army, they

552] DEFEAT OF THE GOTHS 35

began to fall by hundreds beneath the converging fire from the
flanks. So disordered were the Gothic knights by their heavy
loss, and by the plunging and swerving of hundreds of wounded
or riderless horses in their ranks, that their charge slackened to
a very slow pace, and it was only after a long time, and with
great difficulty, 1 that they penetrated to the mass of dis
mounted Foederati in the Roman centre. Having lost all the
advantage of a sudden impact, they did not break the line of
spears, and the battle resolved itself into a hand-to-hand fight
along a contracted front. Here the horsemen surged up and
down for several hours, vainly trying to make a gap, and being
shot down all the time by the volleys of arrows from the flanks.
Their own foot, who should have helped them by keeping the
Roman archers engaged, did not advance far enough to the
front, being apparently afraid to expose themselves to the risk
of a side-stroke from Narses detached body of horse on the left
wing.

At last, at eventide, the Goths were thoroughly tired out,
and after one final effort the great mass of wearied and dis
heartened horsemen gave back and began to retire. Narses at
once charged them with his Roman cavalry, who had as yet done
no work and were quite fresh. Then the Goths broke and fled,
and in their disorderly flight rode over their own infantry, who
in the confusion did not open their ranks to let the fugitives
through, but stood helpless and amazed.

So ended in complete success the first experiment in the
combination of pike and bow which modern history shows. It
is an interesting point of speculation to decide what would have
happened if Baduila had either commenced the battle with the
advance of his foot-archers supported by part of his horse, or
launched some of his cavalry at the Roman bowmen before
charging the dismounted men in the hostile centre. The whole
conduct of the battle on his side is so unworthy of his previous
fame, that we are tempted to accept the story told by Procopius,
that he was mortally wounded at the beginning of the great
charge, and that his men fought all the afternoon without a
leader. But the alternative tale which tells how he escaped
unhurt from the field, fled through the night, and was slain in a
chance medley by a small body of pursuing horsemen, has

1 Tro\\uv re avr/KfyTuf KaKuv 61 ireipav tXffovrfs fyt re Kal /j.6\is es TU&gt;V
d^lxopTo TTJV irapdraijiv ^Proc. , De Bell. Gott. iv. 32).

36 THE ART OF WAR IN THE MIDDLE AGES [553

generally been accepted by historians perhaps merely because
it presents more picturesque details. 1

Narses had barely stamped out the last embers of the Gothic
war, and received the surrender of the few fortresses which held
out after the battles of Taginae and the Sarno, when he was
called upon to encounter a new and altogether different race of
antagonists. A great Prankish host, under the brothers Lothar
and Buccelin, the generals of Theudebert of Austrasia, came push
ing down into the peninsula, to prevent the Imperialists from
enjoying the fruits of their victories. Unlike the Goths, the
Franks were a nation of foot-soldiers armed with spear, sword,
and axe : we shall deal with their methods of warfare in the next
chapter. At Casilinum in Campania, not far from the battlefield
of the Sarno where the Goths had made their last stand, Narses
met and vanquished the eighty thousand men of Buccelin by a
varied application of the same tactics which he had used against
Baduila on the held of Taginae.

The Franks were wont to advance in a deep column or
wedge, which was too solid to be easily broken by a flank
attack : if assailed from the side during its advance, it halted,
fronted to the exposed point, and beat off the assailants.
Well acquainted with these tactics, Narses prepared a dread
ful snare for the Franks. He ranged his foot-archers and other
infantry in the centre, placed a chosen band of dismounted
Foederati behind them, and arrayed his native Roman cavalry,
all horse-archers, in two long wings. The Prankish column
came rushing down on the centre, and scattered the front line
of regular infantry and the second line of archers behind them
without any great difficulty. It then came into contact with the
Heruli and other Foederati who lay behind the light troops, and
began to push them back. But at this moment Narses wheeled
inwards both his wings of horse and threatened to charge the
flanks of the advancing mass. The Franks were at once forced
to halt, and made ready to receive the attack of the cavalry.
But instead of letting his horsemen close, Narses halted them
a hundred yards from the enemy, and bade them empty their
quivers into the easy target of the great weltering mass of
spearmen. The Franks could move neither to front nor flank, for
fear of breaking their array and letting the horsemen into the
gaps, hence they stood helpless, exposed to a shower of missiles

1 Proc., De Bell. Gott. iv. 35.

PLATE //.

FKANKISH WARRIORS

553] THE BATTLE OF CASILINUM 37

to which they could make no reply. Their stubborn bravery kept
them rooted to the spot for some hours, but at last they lost heart,
and began to tail off to the rear, the one side on which they were
not surrounded. Waiting till they were well shaken and lapsing
into disorder, Narses ordered a general charge. His horsemen
rode through and through the broken column, and made such a
slaughter that it is said that only five of Buccelin s army got
away from the field.

With this last victory of the Roman army of the East in Italy
we may close the transition period in the history of the art of
war. The old classical forms have long vanished, and with the
appearance of the Franks on the field we feel that we have
arrived at the beginning of the Middle Ages.

BOOK II

THE EARLY MIDDLE AGES
A.D. 500-768

CHAPTER I

THE VISIGOTHS, LOMBARDS, AND FRANKS

WHEN we leave the discussion of the military art of
the later Romans, and pass on to investigate that of
the Teutonic kingdoms which were built upon the ruins of the
Western Empire, we are stepping from a region of comparative
light into one of doubt and obscurity. If, in spite of our
possessing military manuals like that of Vegetius, official
statistics such as the Notitia Dignitatum, and histories written
by able soldiers like Ammianus and Procopius. we still find
difficult points in the Roman art of war, what can we expect
when our sole literary material in Western Europe consists of
garrulous or jejune chronicles written by Churchmen, a few
fragments of ancient poems, and a dozen codes of Teutonic
laws? To draw up from our fragmentary authorities an estimate
of the strategical importance of the Persian campaigns of
Heraclius is not easy ; but to discover what were the particular
military causes which settled the event of the day at Testry or
the Guadelete, at Deorham or the Heavenfield, is absolutely
impossible. We can for some centuries do little more than give
the history of military institutions, arms, and armour, with an
occasional side-light on tactics. Often the contemporary
chronicles will be of less use to us than stray notices in national
codes or songs, the quaint drawings of illuminated manuscripts,
or the mouldering fragments found in the warrior s barrow.

It is fortunate that the general characteristics of the period
render its military history very simple. By the sixth century
the last survivals of Roman military skill had disappeared in
the West. No traces remained of it but the clumsily-patched
walls of the great cities. Of strategy there could be little in an
age when men strove to win their ends by hard fighting rather
than by skilful operations or the utilising of extraneous

4- THE ART OF WAR IN THE MIDDLE AGES [500

advantages. Tactics were stereotyped by the national
organisations of the various peoples. The true interest of the
centuries of the early Middle Ages lies in the gradual evolution
of ne\v forms of military efficiency, which end in the establish
ment of a military caste as the chief power in war, and in the
decay among most races of the old system which made the
tribe arrayed in arms the normal fighting force. Intimately
connected with this change was an alteration in arms and
equipment, which transformed the outward appearance of war
in a manner not less complete. The period of transition may
be considered to end in the eleventh century, when the feudal
knight had established his superiority over all descriptions of
troops pitted against him, from the Magyar horse-bowmen of
the East to the Danish axemen of the North. The fight of
Hastings, the last notable attempt of unaided infantry to with
stand cavalry in Western Europe for two hundred years, serves
to mark the termination of the epoch.

The Teutonic kingdoms which were founded in the fifth
century within the limits of the Western Empire were some of
them established by races accustomed to fight on horseback,
some by races accustomed to fight on foot. All the tribes
which had their original habitat in the plains beyond the
Danube and north of the Euxine seem to have learned horse
manship: such were the Goths, both Eastern and Western,
the Lombards, Gepidae, and Heruli. The races, on the other
hand, which had started from the marshes of the Lower
Rhine or the moors of North Germany and Scandinavia were
essentially foot - soldiery ; the Franks, Saxons, Angles, and
Northmen were none of them accustomed to fight on horseback.
The sharp division between these two groups of peoples is all
the more curious because many tribes in each group had been
in close contact with the Romans for several centuries, and it
might have been expected that all would have learned a similar
lesson from the empire. Such, however, was not the case : the
Franks of the fifth century, though their ancestors the Chama\ i
and Chatti had been for four hundred years serving the Romans
as auxiliaries when they were not fighting them as enemies,
seem singularly uninfluenced by their mighty neighbours ; while
the Goths under similar conditions had profoundly modified
their armament and customs. In the days of the breaking-up
of the Western Empire the Franks seem no more advanced

SOD] THE VISIGOTHS IN SPAIN 43

than races like the Saxons and Angles, whose relations with
Rome had begun late and continued comparatively slight. To
a certain extent this must have come from the fact that the
emperors had been wont to encourage each band of auxiliaries
to keep to its own national arms and equipment. In the fourth
and fifth centuries, as Mommsen observes, each Teutonic corps
of mercenaries seems to have been valued more, in proportion
as it had assimilated itself less to the Roman model. In spite
of this, it is astonishing to find the Franks of Chlodovech still
destitute of all body-armour and wholly unaccustomed to fight
on horseback. Our surprise is only the greater when we find
that the Imperial host had actually included an ala or two of
Prankish cavalry 1 in the year 400. Evidently the Roman
teaching had taken no hold on the bulk of the race, and its
methods of fighting had remained unaltered.

(I.) The Visigoths, 500-711.

We have already spoken of the Goths, and their pre
ponderant use of cavalry in war. We have seen the Visigoths
of Theodoric charging the Huns on the Catalaunian plain, and
the Ostrogoths of Baduila fretting away their strength against
the horse-archers of Narses. The latter race disappear from
the stage of history in 553, but their Western kindred survived
and kept the same warlike customs down to the eighth century.
Considered as a military power, the Visigoths were not strong;
they generally failed in their contests with the foot-soldiery of
the Franks, and they were shattered with shameful ease by the
Saracens of Tarik and Musa. It would seem, however, that we
must ascribe their weakness to political rather than to purely
military causes. From the first they were too few to hold
firmly the enormous realm that they had conquered. The
Suevi could brave them for several generations in the
Galician hills: the weak chain of Imperial garrisons which
Justinian had established along the southern coast of the
Peninsula was able to hold out against them for seventy years.
The Visigoths of the sixth and still more of the seventh century
appear to have consisted of a not very numerous aristocracy of

* e.g. one cantoned in Egypt and another in Mesopotamia occur in the Notitia.
What is more curious still is that there occurs in the province of Phoenicia an ala
Saxonum " ; so that even the Saxons had been formed into cavalry. (Xot. Or.
Thebais, 31-53 ; Mesopotamia, 31-33 ; and Phoenicia, 32-37).

44 THE ART OF WAR IN THE MIDDLE AGES [S3*

nobles, surrounded by war-bands of their personal retainers,
buccellarii or clientes, without any solid national body below
them. The original army of Alaric and Ataulf had been small,
and the Gothic conquerors could not recruit their numbers by
amalgamating frankly with the Spanish provincials, owing to
the fatal bar of religion. Reccared s conversion to orthodoxy
(589) seems to have come too late to save the race from
perishing for want of numbers. From the military point of
view, the masses of provincials counted for little or nothing;
though they seem from the first to have been made liable to service
in the host, they were unwilling and inefficient auxiliaries. 1
Amalgamation between them and their masters began so late
that it was not quite complete even at the time of the Saracen
conquest in 711. The ruin of the kingdom was the want of a
solid middle class of free Goths. For lack of it the strength
and core of the Visigothic armies consisted of the counts and
the horsemen of their personal retinues, the oath-bound clientes
or buccellarii who had made themselves the "men" of the
nobles. This body showed all the faults of feudal armies of a
later age, for the spirit of loyalty was wanting. The old royal
house died out with the slaughter of Amalric in 531, and none
of the later kings succeeded in founding a permanent dynasty.
The throne passed rapidly from usurper to usurper, and each
great man might covet it, and hope some day to snatch at it by
the aid of his war-band. The provincials passed helplessly
from hand to hand without asserting any will of their own : the
later kings utterly failed in their effort to build up a strong
royal power based on the friendship of the Church and the
support of the masses. Towards the end of the seventh
century there seems to survive no free middle class at all ;
apparently a process like that which occurred in England after
the Danish invasions had driven the small freemen to "commend
themselves " to the local magnates and become their clients.

The Spanish nobles were at the first, like the English
thegnhood, an aristocracy of service, not of blood. The original
host of Ataulf which conquered Spain was Visigothic in name,
but in reality a mixed multitude of Teutons of all sorts. The
Visigothic nucleus which Alaric had originally commanded in
Epirus was quite small ; it only swelled to a great army by the

1 We hear of the Arvemi, all provincials without doubt, serving by themselves,
and under a native leader, in the Visigoth host that fought at Vougle as early as 507.

650] WEAKNESS OF THE VISIGOTHIC ARMIES 45

junction of adventurers of all sorts, especially that of the thirty
thousand Foederati in Italy who joined the invader after the
murder of Stilicho. Hence in this heterogeneous mass there \\ as
no generally recognised noble blood, such as was to be found
among more compact nationalities, like the Lombards, Bavarians,
or Saxons. The only original distinction came from being
promoted to official command by the king. But the men who
had once been given the appointment of " count " or " duke "
grew wealthy, acquired lands, and accumulated clients. Their
descendants in a few generations formed a true nobility based
on wealth and local influence. The majority of the provincial
governors were drawn from their ranks, and they resented in a
body the attempts of strong-handed kings to supersede their
class in office by the preferment of obscure but loyal members
of the royal comitatus. Chindaswinth (641-652; and Wamba
(672-680) tamed them for a short time, but the moment that
the sceptre passed to weaker hands, the aristocracy asserted
itself again. At the moment when the monarchy fell in 711, it
had become wholly feudalised : the nobles and bishops were
the real masters of the realm.

The stream of Spanish annals is such a scanty one that we
learn very little about the details of the interminable civil wars
of the sixth and seventh centuries. Towards the end of the
latter the chronicles fail altogether, and the Egicas and
Rodericks of the last days of the realm are mere names to us.
It is certain, however, that by the end of the seventh century
the Visigothic kings were at their wits end to keep up the
numbers of their army ; a notable law of Wamba gives the
best proof of it. He orders that " every man who is to go forth
in our host, duke or count or castellan, Goth or Roman, freeman
or freedman, or holder on a servile tenure of royal domain-land,
shall bring with him to the expedition a tenth part of his slaves
armed with weapons of war." * Nothing but the utter want of
a middle class of warlike small proprietors could account for
this desperate expedient being tried. A similar deduction may
be made from the fact that another law of Wamba orders even
clerical landholders to come to the host with their armed slaves.
Of the organisation of the army we know only that the counts
led the levies of their own districts, each of which corresponded
as a rule to an ancient Roman civitas. Under the counts were

1 Lex Visigothorum, UL. 2, 9.

46 THE ART OF \\~AR IN THE MIDDLE AGES [680

thiufads or thousand-men, and centenarii or hundred-men, whose
duty was to collect the host each in his own locality. In time
of peace the count and thousand-man were judges and governors,
like an English ealdorman ; in time of war they took the field
at the head of the whole levy en masse, Gothic or Roman, of
their district. Spanish armies, therefore, were often very
numerous, but they were disorderly, undisciplined, and generally
very half-hearted in their service. The masses of provincials
cared nothing for their ephemeral kings, and thought much
more of propitiating their local despots, the counts. Hence
rebellious nobles could generally rely on the service slack
and unwilling though it might be of the inhabitants of their
government. By the seventh century the majority of these
inhabitants had become the " men " of their rulers, who thus
reached such a pitch of greatness that we find them called,
even in state documents, tyranni? as if they were independent
princes.

The Gothic nobles and their war-bands fought on horseback,
" gaudent equorum praepeti cursu," as Isidore of Seville wrote in
6 1 5 ; 2 though, when necessary, they would dismount. Their great
weapon was the lance ; their bodies were covered with harness
of ring-mail or scale-armour, and their heads by crested helms,
probably of the same shape as those worn by their neighbours
the Franks. They bore round shields, swords, and daggers
(semispatha, scrama). The mace and axe were not unknown
to them ; the use of the latter they had learned from the Franks,
and they therefore called it francisca. That defensive armour
was fairly common may be deduced from the fact that King
Erwig (63o ; ordered that even of the slaves whom the bishops
and nobles led to the host, some should wear a mail-shirt,
though the majority were only expected to come with shield,
spear, sword, scrama, or bow and sling. 3 The word employed

1 e.g. in some of Wamba s rescripts.

- The passages on weapons in Isidore of Seville s Etymologicon are so pedantic
and so stuffed with quotations from Virgil and Lucan, that we might be tempted at
first to dismiss them as wholly useless repetitions of Roman usage. But this would be
unjust to the author, who shows that he is not wholly neglectful of the thinjjs of his
own day by making notes on the scrama-semispatha, and adding a mention of the
secures quas et Hispani ab usu Francorum per derivationem franciscas v Ocan t " It is
o be noted also that he has no account of the old Roman breast and back harness of
p ate under lorica and only catalogues the mail-shirt of rings and the brica syuamea
of scales. See Etym. xviii. 11, 13, 18.

3 Lex Visigotho&gt;-nm, ix. 9.

8o] SIEGECRAFT OF THE VISIGOTHS 47

for the mail-shirt is zaba, the same which Maurice and Leo use
for the armour of the Byzantine cavalry-soldier, and not brunia
(byrnie], the common term of the FYanks, Saxons, and other
Teutonic tribes of the North.

The provincial levies, as opposed to the counts and their
clientes, were great masses of unarmoured infantry, like the old
English fyrd, armed with rude and miscellaneous weapons, and
serving much against their will. There was little or no infusion
of Gothic blood amongst them, and their service was perfunctory
unwilling, and inefficient.

The Visigoths seem to have had a greater skill in the poliorcetic
art than many of their Teutonic kinsmen. Probably it was
picked up from the East-Romans during the long sieges of
the haven-towns of South Spain during the reigns of Reccared,
Sisibut, and Swinthila, when for a whole generation (580-620)
the main political object of the kings was to recover the
ports of Andalusia and Algarve, which the folly of Athanagild
had betrayed to the generals of Justinian. We find that the
Visigoths were acquainted with the funda and balista, which
threw respectively stone balls and darts, that they used the ram
(aries), and aided its work with the pluteus ("shelter-hurdle) and
the musculus for digging into the foundations of walls. In the
one siege of which we have considerable details, that in which
Wamba took Nismes in 673, the ram, the stone-throwing
machine, and fire-arrows are described as in use. 1

The end of the Visigoths as a military power was sudden
and disgraceful. How far the immediate cause of the loss of
the battle of the Guadelete was disloyalty on the part of the
counts, or slackness on the part of their subjects in the provincial
levies, or a deficiency of properly - equipped fighting men, we
cannot tell. The details of the fatal day are lost ; nor have
we sufficient notices of any Spanish wars of the previous century
to enable us to construct a full account of the tactics of the
Visigothic army.

(II.) The Lombards, 568-774.

Concerning the Lombards, the last of the Teutonic races
whose strength lay in their horsemen, we have far more know
ledge. They were in much more direct touch with the Eastern

1 See Archbishop Julian s Life of Wamba, the last really detailed piece of
Visigothic history which survives.

48 THE ART OF \VAR IN THE MIDDLE AGES [650

Empire than any of their brethren during the sixth, seventh,
and eighth centuries, so that we have a certain amount of
information bearing on them from Byzantine sources. Their
early legends have been preserved by the excellent Paul the
Deacon, who also furnishes us with a sketch of their later
annals, abounding in those picturesque tales which, though they
may not be accurate history, are invaluable as giving the
manners and customs of the race. In addition we can draw on
the information contained in the code of laws dra\vn up by
Rothari (643; and the supplements appended by his successors.
Like all the races that have ever dwelt by the Middle Danube,
the} were essentially a race of horsemen. The primitive folk
tales recorded by Paul show it very clearly ; on their first actual
appearance on the stage of history it is equally manifest. Pro-
copius records how they sent to Narses two thousand five hundred
horsemen of noble birth, and three thousand of lesser race who
were the attendants and squires of the others. If they dismounted
at Taginae to stand the Gothic charge, it was by Xarses order ;
the old general had resolved to make his centre solid by placing
there his steadiest auxiliaries. 1 A little later, when they invade
Italy on their own account, we read of every king and duke and
hero fighting with lance and war-horse at the head of his men.
One interesting passage in Paul gives us the armament of the
Lombard knight helm and mail-shirt, and even greaves, which
la^t many Western races had not adopted even three centuries
later. 2 In another, we read of the great lance (contus}, so
strong that a Lombard champion, who had pierced a Byzantine
horseman through the body, actually lifted him from his saddle
and bore him aloft wriggling on the weapon s point. 3 The
other great Lombard weapon was the broadsword (spat/ia ),
which seems to have been worn at all times, 4 not merely when
the warrior was equipped for war. On one occasion only do
we hear of a hero fighting with a club, and then only because his
lance was not to hand. 5 Though acquainted with the bow, 6 they
do not seem to have used either it or the javelin to any extent

Not, we need hardly repeat, because he wished to prevent troops of doubtful
loyalty from leaving the field.

2 " Loricam suam, galeam, atque ocreas tradidit diacono, et caetera arma"
(Paul. v. 40).

3 Ibid. v. 10.

* In Paul. vi. 51 it is worn at the king s council board ; in vi. 38 at a feast.
5 Ibid. vi. 52. Ibid. v. 33.

So] THE ARMS OF THE LOMBARDS 49

in war. It was always on lance and war-horse that they placed
their reliance, like the Goths, who had held the plains of
Northern Italy before them. It was always on horseback that
their plundering bands crossed the Alps to ravage Provence and
Dauphine", faced the Bavarians on the Upper Adige, or pursued
the Slovenes of Carinthia when they dared to molest the borders
of Friuli. From a passage in the Tactica of Leo the Wise we know
that, when hard pressed and surrounded, the Lombard knights
would turn their horses loose, and fight back to back in a solid
mass, with spears levelled outwards. 1 It must have been only
in dire extremity that they would do so. Paul the Deacon
tells in one characteristic passage relating to a Lombard defeat,
how Argait the Schultheiss was slain with many of his men
because he must needs spur his horse up an almost inaccessible
slope to attack the plundering bands of the Carinthian Slavs.
His duke Ferdulf had taunted him with the words, " Arga
[slothful] is your name and your nature too." To vindicate his
courage, Argait and his horsemen charged up the steep slope and
were destroyed by the great stones which the Slavs rolled down
on them, whereas, if they had dismounted and turned the
position, they were " men many and brave enough to have
destroyed thousands of such foes." -

It is perhaps worth noting that the horse appears more fre
quently in the Lombard laws than in those of any other Teutonic
people. There are countless clauses relating to horse-stealing,
to horse-breeding, to the valuation of horses, to assaults such as
throwing a man off his horse (meerworphiri), to accidents caused
by the kick of a horse, to the buying and selling of horses. A
war-horse with its trapping was valued as high as one hundred
solidi, twice the value of the life of a household slave, and two-
thirds of that of a free Lombard of low degree. 3 The king s
breed of chargers was highly esteemed, and the gift of one of
them to a retainer or a high official was a great mark of favour.

The Lombards, unlike the Franks, Visigoths, and Saxons,
were not a collection of war-bands, nor a mixed multitude of
diverse races,* but a compact national body moving down en
masse with wives and children, flocks and herds, to occupy the

1 Leo Sapiens, Tactica, xviii. 81. " Paul. vi. 24.

3 See Laws of Rothari and Luitprand, passim.

4 Though there were many Saxons and broken men of small tribes with Alboin
(Paul. ii. 26, iii. 5), yet the great majority were Lombards.

50 THE ART OF WAR IN THE MIDDLE AGES [7 2

well-nigh depopulated plain of Northern Italy. But there was a
disintegrating force among them ; this was the want of a per
manent royal house. Even before the conquest of Italy by
Alboin, their dynasty, according to their own legends, had
changed several times. Alboin was only the second of his
race who had reigned over them. When he died heirless, and
his immediate successor, Cleph, was slain only a year later, the
nation could not agree on the choice of a king, and lived for ten
years without one. But they did not cease to advance and to
conquer, though they were only led by the " dukes " (heretogas
or ealdormen, as the Anglo-Saxons would have called them),
who were the heads of the various faras or families which
made up the nation. Under these princes the Lombards broke
up into tribal groups : some entered Gaul to ravage Burgundy ;
others pushed down the peninsula of Italy, and established the
duchies of Spoleto and Benevento. It was only the pressure of
a Prankish invasion, aided by the Byzantines, that drove them
into combination again, and forced them to crown Authari as
their king. The kingdom thus restored was never so strong as it
should have been ; the dukes of Spoleto were in practice, if not
in name, independent of it, and those of Benevento hardly
acknowledged its supremacy at all. It was only Luitprand
( 712-744), who reigned but shortly before the Frankish conquest
of Italy, that welded the Lombards north of Benevento into
a compact state.

The warlike organisation of the race, as was the case in all
the Teutonic kingdoms, was the same which served for civil
government. The Lombard realm was divided up into duchies ;
there are said to have been thirty-six, and the men of each
district rode to war under their duke. These chiefs were
generally of the old noble blood of the race, eorl-kin, as the
English would have called them. Chance has preserved the
names of some of these old noble families, the houses of Caupus
and Harodos, Beleo, Anawas, and Hildebohrt. As the realm
grew stronger, the king sometimes replaced a rebellious duke
by an officer more directly dependent on himself, a gastaldus ;
those who had borne this title at first seem to have been the
governors of cities in the royal domain, 1 and the guardians of the
royal domains within the duchies. There appears to have been a
large middle class of Lombard race, the thing that was so

1 Domus Xostrae Civitates, Codex Dip!. Lang. ii. 334.

750] THE LOMBARD HOST 51

much lacking among the Visigoths of Spain. All Lombards
small and great were exercitiales (or arimanni), bound to turn
out at the monarch s call to war, like the English fyrd. Many,
both noble and simple, had made themselves the king s " men,"
by the oath of personal devotion. They were called gaisindi?
a word corresponding of course to the Anglo-Saxon gesitli, and,
like fat gesith, rode in their lord s train, and had their place in
his hall. The chief of these military retainers were the marfahis
or constable, the scilpor or shield-bearer, and the banner-bearer
of the king. The dukes in a similar way kept smaller bands of
gaisindi, but they were never able to make henchmen of the
whole of the freemen resident in their duchies, as did the counts
of Visigothic Spain. The number of the Lombards of middle
fortune was too great to allow of such a usurpation taking place,
and the king s gastaldus and scliultlieiss ( reeve, as the English
would have called him) were present in each duchy, to keep its
ruler in check, and afford protection to any freemen whom he
might strive to oppress. 2

Having dealt with Goth and Lombard, we may now turn to
the Teutonic kingdoms of the North, where infantry and not
horsemen were the main power in war.

(III.) The Franks, 5OO-768.

The Prankish tribes whom Chlodovech had united by the
power of his strong arm, and who under his guidance overran
the valleys of the Seine and Loire, were among the least
civilised of the Teutonic races. In spite of their long contact
with the empire, the} were fas we have already had occasion to
mention; still mere wild and savage heathen when they began
the conquest of Northern Gaul. The Franks, as pictured to us
by Sidonius Apollinaris, Procopius, Agathias, and Gregory of
Tours, still bore a great resemblance to their Sigambrian or
Chamavian ancestors whom Tacitus described more than three
centuries earlier. The words in which Sidonius paints them in
460 are practically identical with those which Agathias used
more than a century later, so that even the conquest of Southern
Gaul seems to have made little difference in their military

1 Paul translates gaisimt by safeties, vi. 38.

- See the Law of Rothari, 23 : " Si dux exercitialem suum molestavit injusle,
gastaldus eum solatiet, quousque veritatem suam inveniat," etc.

52 THE ART OF WAR IN THE MIDDLE AGES [500

customs. The poetical bishop of Auvergne speaks of their
unarmoured bodies girt with a belt alone, their javelins, the
shields which they ply with such adroitness, and the axes which,
unlike other nations, they use as missiles, not as weapons for
close combat. He mentions their dense array and their rapid
rush, " for they close so swiftly with the foe, that they seem to
fly even faster than their own darts." Agathias is more detailed,
but he is evidently describing a race in exactly the same stage.
" The arms of the Franks," he says, " are very rude ; they wear
neither mail-shirt nor greaves, and their legs are only protected
by strips of linen or leather. They have hardly any horsemen,
but their foot-soldiery are bold and well practised in war. They
bear swords and shields, but never use the sling or bow. Their
missiles are axes and barbed javelins (u.y~/uv&lt;;}. These last are
not very long, they can be used either to cast or to stab. The
iron of the head runs so far down the stave that very little of
the wood remains unprotected. In battle they hurl these
javelins, and if they strike an enemy the barbs are so firmly
fixed in his body that it is impossible for him to draw the
weapon out. If it strikes a shield, it is impossible for the
enemy to get rid of it by cutting off its head, for the iron runs
too far down the shaft. At this moment the Frank rushes in,
places his foot on the butt as it trails on the ground, and so,
pulling the shield downwards, cleaves his uncovered adversary
through the head, or pierces his breast with a second spear." 1

The francisca or casting axe was even more typically a
Prankish weapon than the barbed ango&gt;i. Numerous specimens
have been found in Merovingian graves ; - it was a single-bladed
axe with a heavy head, composed of a long blade curved on its
outer face, and deeply hollowed in the interior. It was carefully
weighted, so that it could be used, like the American tomahawk,
for casting purposes, even better than for close combat. The
skill with which the Franks discharged the weapon just before
closing with the hostile line was extraordinary, and its effective
ness made it the favourite national weapon. A shield, sword,
and dagger completed the arms of the warrior: the first-named
was of a broad oval shape, and had a large iron boss and an
iron rim ; the sword was a two-edged cut-and-thrust weapon,

1 Agathias.

- One was in the first Krankish monument to which a definite date can be riven
Childeric s tomb at Tournay (481).

5oo] FRANKS AND GAULS 53

ranging from thirty to thirty-six inches in length ; the dagger
(scramasax, semispatka) was a broad thrusting blade of some
eighteen inches.

For some two centuries on from the time of Chlodovech,
these were the arms of the Prankish foot-soldiery ; they seem to
have borrowed nothing from their Roman predecessors. It is
true indeed that some of the Gaulish levies who served the
Merovings continued for a space to wear the ancient equipment
of the troops of the empire. Such, at least, is the statement of
Procopius, an author whose words are never to be lightly treated :
he says that many of the Gaulish cities, having surrendered
themselves on favourable terms to the Prankish conqueror, were
still in his own day sending their contingents to the host under
their ancient banners, and wearing the full Roman array, even 1
down to the heavy-nailed military sandals. There is nothing
incredible in this statement ; it is certain that from a very early
stage of the conquest of Gaul the Prankish kings strengthened
their armies from the ranks of the provincials, an experiment
which was far easier for them than for Lombard or Visigoth,
because they were not divided from their subjects by the fatal
bar of Arianism. 2 But it is quite clear that the conquerors
did not adopt the arms of the conquered, and that the survival
of the Roman garb and weapons among the Gauls disappeared
in the sixth century. Just as we find Gallo-Romans adopting
Prankish names by the end of that age, so we find them
assimilating Prankish military customs. The tendency among
the masses is towards the barbarising of the provincial, not to
wards the civilising of the Teuton. All through the Merovingian
times, and indeed down to the time of Charles the Great him
self, the Prankish armies were mainly great disorderly masses
of unarmoured infantry, fighting in dense column formation.

It is among the highest classes alone that the effect on the
invaders of their contact with the lingering civilisation of Gaul
is to be seen in things military as in all other things. The
epigram which the Gothic sage made concerning his own
tribesmen and the conquered provincials was true of the Franks

1 ;cal aijiifia ra mptrtpa. eirayo/j^voi ovrw is Atax 7 ? 1 " foMfrvm, i ay^^ua. TU&gt;K
(v rt roii dXXois dira&lt;n Kai v TOIS i&gt;To^/uurt oiaffufovtrtv (De Bell. Gott. i. 12).

- As Fustel de Coulanges points out, even Chlodovech himself seems to have had
Gauls in his army, especially a certain Aurelianus, whom he made ruler of Melun
(M. F. 495)-

54 THE ART OF WAR IN THE MIDDLE AGES [550

also: "The poor Roman tends to assimilate himself to the
German, and the wealthy German tends to assimilate himself
to the Roman." 1 While the masses in Gaul forgot the oJd
military habits of the empire, and degenerated into disorderly
tribal hordes, the kings and great nobles among the Franks
borrowed something from the externals of the vanishing
civilisation. Just as they appropriated relics of Roman state
and sho\v in things civil, so in certain military matters they did
not remain entirely uninfluenced by the Roman practice. In
the sixth and seventh centuries \ve find among them the feeble
beginnings both of the use of cavalry and of the employment
of armour, commencing around the person of the king, and
gradually spreading downwards.

Of the employment of horsemen among them the first mention
is in Procopius, 2 who says that King Theudebert, while invading
Italy in 539 with a hundred thousand men at his back, had a few
horsemen whom he kept about his person. They were armed
with the lance, but nothing is said of their wearing armour;
probably it was still very rare among them, and only used by
kings, dukes, and counts. It is remarkable that on the whole
there is very little mention of defensive arms in Gregory of
Tours, though he describes countless battle scenes. Even
chiefs engaging in single combat before their followers do not
always seem to have been provided with them. 3 But from the
middle of the sixth century onwards armour seems gradually
to grow usual among great men, and then among all the
wealthier classes. Bishop Sagittarius in 574 is blamed for
taking the field " armed not with the sign of the heavenly cross,
but with the secular cuirass and helm." 4 Count Leudastes
shocks the good Bishop of Tours by entering his house in
helm and breastplate, a quiver swinging at his waist, and a
lance in his hand. 5 The henchman of Duke Guntram wears a
breastplate, and is drowned by its weight in a ditch (AD 583)6
The usurper Gundovald Ballomer is saved by his body armour
from the stroke of a javelin (A.D. 585).* I n the Saxon war of
626 we read of both Clothar n. and his son Dagobert wearing

Romanus miser imitatur Gothum, et Gothus utilis imitatur Romanum -
- De Bell. Got/, n. 25.

GregoJ v^*" *"** ** ^^ ^ *"* ^"^ ^^ f Guntram and Dracolenu, in
Gregory , iv . 18. * //&lt;/. v. 48.

*. 26. -. Illid V

6oo] PRANKISH ARMOUR 55

helm and breastplate (A.D. 626). 1 The brunia, which com
posed the body armour, was no doubt usually the mail-shirt of
rings which we find among all Teutonic races in the Middle-
Ages. But scale armour sewn on to a leather foundation was
also known ; it was sometimes of the fish-scale shape, sometimes
square-scaled. In either case it was fixed so that each row of
scales overhung the one below it, and protected the upper ends
of it, where the thread fastened it to the leather. There seems
to have been no survival beyond the fifth century of the old
Roman lorica of plate ; perhaps Western armourers were not
capable of forging it ; but even at Byzantium, where the power
to make it was not wanting, this form of cuirass disappeared:
probably it was inconvenient for the horse-bowman, and was
dropped when he became the chief factor in war in the East,
that the more pliant mail-shirt might take its place.

The Prankish headpiece was of a peculiar form, very dis
similar both to the usual shapes of the Roman helmet and to the
pointed Byzantine casque with its little tuft. The typical form
among the Franks was a morion-shaped, round-topped head
piece, peaked and open in front, but rounded and falling low at
the back, so as to cover the nape of the neck. It was furnished
with a comb or crest, which may have been composed either
of thin metal or of leather. This very peculiar helm bears
more likeness to a sixteenth-century morion than to any shape
among the numerous headpieces of the Middle Ages. Its
prototype, however, was undoubtedly one of the less common
late Roman types, not the old classical helmet, which we see
on the head of Honorius or Justinian, but one more like that
worn by certain classes of gladiators, and occasionally represented
on coins of the fifth and sixth centuries. [See Plate No. n.]

Some German writers have doubted the existence of the
crested Prankish helm, such as appears in hundreds of
Carlovingian and pre-Carlovingian representations of military
figures. 2 They allege these drawings to be the mere slavish
copies of old Roman pictures, taken from fourth or fifth
century manuscripts. There was, no doubt, an immense
amount of such copying done, but that the crested helm never
existed is incredible. The Franks brought no headpiece of their
own into Gaul ; they had fought bareheaded when they dwelt

1 I fta Dagoberii, 13.

As, for example, those from the Utrecht Psalter on Plate II.

56 THE ART OF WAR IN THE MIDDLE AGES [550

on the moors of Toxandria. But they found the late Roman
helm in full use in their new re-aim, and there can be no doubt
that their kin-s and nobles borrowed it from their subjects.
From the first, as we have seen, the Franks used their
provincial vassals as auxiliaries in the field. The Roman con-
dottieri, like Count Aurelianus, who served under Chlodovech I.,
no doubt wore the crested headpiece ; so did the Gallic
contingents, whom Procopius describes as serving "with the
old Roman uniform and standards," in the army of Theudebert
in s^9- 1 ^ e cannot suppose that when the Gallo- Roman
Bishop Sagittarius equipped himself in a helm in 574, to fight
the Lombards, he put on some newly-invented Prankish head-
piece.* Undoubtedly the old crested helm of the late Roman
period was perpetuated among the leading men of the Gallic
provincials, and was taken directly from them by the Franks.
It only gave way to simpler forms of a more pointed shape in
the ninth" century. No doubt, however, this costly metal
helm was always rare ; when headpieces became more common,
cheaper productions, such as the leather caps of a plain round
shape, which the MSS. of the eighth and ninth centuries often
display, were more usual. But the helm which the eighth-
century Lex Ripuaria values at six solidi 3 half the price of a
mail-shirt must have been no leather makeshift, but an
elaborate piece of metal-work, to be worth such a great price. 4

The Prankish shield, it may be added, was usually round
and very convex. It was made of wood bound at the edges
with iron, and possessed a prominent boss, which was sometimes
spiked. It was only when the use of the horse in war became
common, that the round shield became kite - shaped. Before
the ninth century the circular shape was almost universal.

The use of the horse in battle seems to extend itself in
exactly the same proportion as that of body armour, spreading
downward through the sixth and seventh centuries, till, by the
close of the Merovingian age, it has become usual among the
upper classes ; the counts and dukes with their immediate

1 See p. 53. 2 See p. 54.

3 The hrunia is mentioned in the early Ripuarian law, and valued at twelve solidi.
the helm at six, the sword at seven (Lex Rip. xxxvi. 1 1 ). It is more surprising
to find bainiergae (greaves) mentioned, and valued at six solidi.

4 See illustration on Plate n. : the Utrecht Psalter is late, but its drawing? are
copied from Aferovingian originals

630] THE FRANKS LEARN HORSEMANSHIP 57

retinues were habitually fighting on horseback by the end of the
seventh century, though when pressed or surrounded they
would still dismount and fight on foot like their ancestors. The
first single combat on horseback related to us is that of
Guntram and Dracolenus in 578. Early instances of the
appearance of a considerable body of cavalry are found in the
army of Count Firminus in 567 ,* and that of Duke Leudigisl in
584 ; 2 but the first mention of a regular cavalry charge which
settled a battle is in the Saxon war of Chlothar II. in 626. The
king, irritated by the cries of the enemy, who from the other
side of the Weser kept pelting him with taunts and insults,
"put spurs to his horse and crossed the stream, all the Franks
following him and swimming through the water, though it was
full of fierce whirlpools." Chlothar engaged Bertwald, the Saxon
leader, before his men could come up with him ; " then all the
Prankish horsemen, who were still far behind their lord, shouted,
" Stand firm, O king, against thy adversary ! " Chlothar s
hands were wearied, " for he wore a breastplate, and the water
which had soaked all his garments rendered them very heavy,"
but he slew Bertwald before his men reached him, and then
together they made a vast slaughter of the Saxons. 3

That, as a rule, the proportion of horsemen in a Merovingian
army, even in the seventh century, was very small, can be
gathered from many pieces of evidence. The battle picture
which Fredegarius gives of the victory of Ziilpich in 612, when
Theuderich of Burgundy beat his brother of Austrasia, may
serve as a fair example, because the writer specifies it as the
most bloody and obstinate combat on a large scale which had
been seen in human memory. It appears that the fighting was
all on foot, for " so great was the press when the hostile masses
\pJialanges\ met and strove against each other, that the bodies
of the slain could not fall to the ground, but the dead stood
upright wedged among the living." 4 Obviously this could only

1 Gregory, iv. 30. In this case the horses are only mentioned as lost by their
riders after a defeat ; does this mean that they had dismounted to fight ? They are
described as swimming the Rhone on their backs.

- Ibid. vii. 35.

8 Vita Dagoberti, 13.

4 " Tanta strages ab utroque exercitu facta est, ubi phalangae ingressae certamine
contra se praeliabant, ut cadavera occisorum undique non habuerint ubi inclines
jacerint, sed stabant mortui inter ceterorum cadavera stricti, quasi viventes "
(Fredegarius, 38).

S 8 THE ART OF WAR IM THE MIDDLE AGES [73 2

have happened in an infantry fight. Still more interesting is
the account of the array of the Franks a hundred years later, at
the all-important battle of Poictiers, where Charles Martel
turned back the advancing flood of Saracen horsemen who had
swept so easily over the debris of the Visigothic monarchy.
" The men of the North," says the chronicler, " stood as
motionless as a wall ; 1 they were like a belt of ice frozen
together, and not to be dissolved, as they slew the Arabs with
the sword. The Austrasians, vast of limb, and iron of hand,
hewed on bravely in the thick of the fight ; it was they who
found and cut down the Saracen king." Obviously, therefore, at
Poictiers the Franks fought, as they had done two hundred years
before, at Casilinum, in one solid mass, 2 without breaking rank
or attempting to manoeuvre. Their victory was won by the
purely defensive tactics of the infantry square ; the fanatical
Arabs, dashing against them time after time, were shattered to
pieces, and at last fled under shelter of the night. But there
was no pursuit, for Charles had determined not to allow his
men to stir a step from the line to chase the broken foe.
Probably he was right, for an undisciplined army cannot
advance against cavalry without danger, and the Arabs, even
when repulsed, were too agile and brave to be allowed the
chance of penetrating into the mass. We must conclude,
therefore, that the Prankish chiefs and nobles had all dis
mounted and fought on foot in the " wall of ice " which they
opposed to the fiery onslaught of the Moslem horsemen. Such
tactics were, no doubt, exceptional by the eighth century, and
adopted only against an enemy all - powerful in horsemen.
Against armies of Saxons, or Frisians, or Bavarians, composed
wholly or almost wholly of foot - soldiery, the Franks would
employ their proportion of mounted men to advantage. We have
already seen King Chlothar, a hundred years before Poictiers,
lead a charge against a Saxon host at the head of his cavalry.
Perhaps a less able general than Charles Martel would have
tried the experiment against the Arabs, and courted disaster
thereby. For a few thousand Prankish knights could have

1 "Gentes septentrionales ut paries immobiles permanentes, et sicut zona rigoris
glacialiter adstricti gladio Arabes enecant. Gens Austriae mole membrorum
praevalida et ferrea manu per ardua pectorabiliter ferientes regem inventum cxanimant"
(Isidorus Pacensis).

" See p. 63.

700] MILITARY ORGANISATION OF THE FRANKS 59

done nothing against the swarms of invaders, while the infantry,
destitute of the backing of mailed men of high rank and
practised skill, might have been ridden down.

Nothing could have been more primitive than the military-
organisation of the Merovingian era. The count or duke who
was the civil governor of the civitas was also its military head.
When he received the king s command, he ordered a levy en
masse of the whole free population, Roman, it would appear, no
less than Prankish. From this summons, it seems that no one
had legal exemption save by the special favour of the king. In
practice, however, we gather that it cannot have been usual to take
more than one man from each free household. 1 That the " ban "
did not fall on full-blooded Franks alone, or on landholding
men alone, is obvious from the enormous numbers put in the field.
The levy of the county of Bourges alone was fifteen thousand
men, 2 and, as Fustel de Coulanges remarks, it is incredible that in
sucli a district, at a time when large estates were common, there
should have been fifteen thousand families holding their land
straight from the king. The fine for failing to obey the ban was
enormous : by the Ripuarian law it was sixty solidi for free
Franks, thirty for Romans, freedmen, or vassals of the Church. 3
At a time when a cow was worth only one, and a horse six
solidi, such a sum was absolutely crushing for the poor man, and
very serious even to the rich.

There is as yet no trace of anything feudal in the Merovingian
armies. The Franks in Gaul appear, as far as can be ascer
tained from our sources, to have had no ancient nobility of
blood, such as was to be found among the eorl-kin of England,
the Edilings of continental Saxony, and the Lombard ducal
families. The Franks, like the Visigoths, seem to have known
no other nobility than that of service. Chlodovech had made a
systematic slaughter of all the ruling families of the small
Prankish states which he annexed : apparently he succeeded in
exterminating them. Among all his subjects none seems to
have had any claim to stand above the rest except by the royal
favour. The court officials and provincial counts and dukes of
the early Merovings were drawn from all classes, even from the

1 Such would be the deduction from the document quoted by Fustel de Coulanges,
Monarchic Frartyue, p. 293, where a son is allowed to volunteer for a campaign in his
father s place.

- Gregory of Tours, vi. 31. 3 Lex Rip. Ixv. g 2.

60 THE ART OF WAR IN THE MIDDLE AGES [600

ranks of the Gaulish provincials. Great officers of state with
Roman names are found early in the sixth century ; by the end
of it, the highest places of all were open to them. One Gallo-
Roman, Eunius Mummolus, was King Guntram s commander-in-
chief ; a few years later, another, Protadius, was Mayor of Bur
gundy, and first subject of the crown. The Frankish king, like all
Teutonic sovereigns, had his own " men " bound to him by oath ;
they were called antrustions, and corresponded to the English
gcs ith, the Lombard gaisind, and the Gothic saio. But they do
not appear to have been a very numerous body, certainly not
one large enough to form the chief element of importance in the
host, though there were enough of them, no doubt, to furnish
the king with a bodyguard. The Frankish tariff of weregilds
shows that the antrustions were drawn from all classes. In each
rank of life their valuation was very much higher than that of
persons not included in the royal comitatus. Both the Salic and the
Ripuarian laws value a free Frank at two hundred solidi, but a
freeman " in the king s trust " at six hundred. That there were
also Gauls and letes (Treedmen; among the antrustions, is shown by
two clauses of the Salic law, which fine " anyone who, at the head
of an armed band, has broken into the house of a freeman in
the king s trust and slain him, eighteen hundred solidi ; and
anyone who has broken into the house of a Roman or a lete in
the kincr s trust and slain him, nine hundred solidi." * From the

ranks of the antrustions were drawn the counts and dukes who
headed the Frankish provincial levies in the field.

It seems clear that these officials had very imperfect control
over the men whom they led out to war. Being mere royal
nominees, without any necessary local connection with the
district which they ruled, their personal influence was often
small. When the counts, with their subordinates in the ad
ministrative government, the vicarii and cen tenant, took the field,
it was at the head of masses of untrained men. There was
neither pay nor even food provided for the army, the men being
supposed to bring their own rations with them even down to
the time of Charles the Great. Hence it was no marvel that
bad discipline, and a tendency to plunder everywhere and any-

1 Lex Sal. xlii. (ed. Hessels ) : "Si quis collecto contubernio hominem ingenuum in
domo sua Occident, si in irusle dominica fuit tile qui occisiis est, solidos MDCCC
culpabilis judicelur : solidos DCCCC si quis Romanum vel litum in truste dominica
Occident/

700] MILITARY ORGANISATION OF THE FRANKS 61

where, were the distinguishing features of a Merovingian army.
Having exhausted its own scanty food supply, the host would
turn to marauding even in friendly territory : the commanders
were quite unable to keep their men from molesting their
fellow-subjects, for hunger knows no laws. When in a hostile
country, they lived by open rapine, eating up the land as they
passed ; if therefore a long siege or a check in the field confined
them for some time to the same spot, they soon harried it bare,
and were then reduced to starvation. Gregory of Tours and
Paul the Deacon show one great host in Lombardy reduced to
such straits that the men sold their very clothes and arms to buy
bread. 1 Time after time large armies melted away, not because
they had been defeated, but merely because the men would not
stand to their colours when privations began. To this cause,
more than to any other, is to be ascribed the fact that after the
first rush of the Franks had carried them over Gaul, they failed
to extend their frontiers to any appreciable extent for more than
two hundred years.

The other great disease of Merovingian hosts was want of
discipline. Unless the king himself were in the field, there was
the gravest danger that the contingents of the various provinces
would fail to obey their commander- in -chief. One count
thought himself as good as another, and the local levies might
have some respect for their own magistrate, but cared nothing
for the man who ruled a neighbouring province. The Merovings
sometimes tried to secure obedience by creating dukes for the
frontier regions, and giving them authority over several counts
and their districts, so as to secure uniformity of action against
the enemy. But there was no proper hierarchy either of civil or
of military functionaries ever established, nor was subordination
of man to man really understood. The generals of King Gun-
tram answered to their master when he rebuked them for a
disgraceful defeat at the hands of the Visigoths : 2 " What were
we to do? no one fears his king, no one fears his duke, no one
respects his count ; and if perchance any of us tries to improve
this state of affairs, and to assert his authority, forthwith a
sedition breaks out in the army, and mutiny swells up." This
is almost the same language used by the Byzantine emperor,
Leo the Wise, when, three hundred years later, he describes the
Franks of his own day.

1 Gregory of Tours, x. 3. J Ibid. ix. 31.

62 THE ART OF WAR IN THE MIDDLE AGES [612

Even the kings themselves often found that the hereditary
respect of their people for the royal blood was insufficient to
secure obedience. Chlothar I. in 555 wished to make peace
with the Saxons, when they offered him tribute and submission.
But his army thought themselves sure of victory, and yearned
after the plunder that had been promised them. They forced
Chlothar to send away the Saxon envoys and to fight. 1 As
might have been expected, the disorderly host was well beaten.
An example of the opposite form of indiscipline was seen in
6 1 2, when the armies of Theuderich n. and Theudebert II. one
of the numerous pairs of unnatural brothers who disgrace the
annals of the Merovings were in presence. When Theuderich
bade his men advance, they broke their ranks, slew the Mayor
Protadius in the king s very tent, because he tried to urge them
on, and forced their unwilling master to make peace with the
Austrasians. It is marvellous that this phenomenon did not
take place more often ; so worthless were the Merovings, and so
futile their pretexts for war with each other, that one can only
wonder at the docility of the subjects who let themselves be
butchered in such a cause.

1 Gregory, iv. 8.

CHAPTER II

THE ANGLO-SAXONS

IN their weapons and their manner of fighting, the bands of
Angles, Jutes, and Saxons who overran Britain were more
nearly similar to the Franks than to the German tribes who
wandered south. In blood and language, however, they were more
akin to the Lombards than to the Franks; but two or three
hundred years spent by the Danube had changed the Lombard
warriors and their military customs, till they had grown very
unlike their old neighbours on the Elbe from whom they had
parted in the third or fourth century. The Angles and Saxons,
even more than the Franks, were in the sixth century a nation
of foot-soldiery, rarely provided with any defensive armour save a
light shield. They had been in comparatively slight contact
with the empire, though they had made occasional piratical
descents on the east coast of Britain even before the year 300,
and though one "ala Saxonum" appears among the barbarian
auxiliaries of the Notitial

The arms and appearance of the war- bands which followed
Hengist or Cerdic across the North Sea can best be gathered
from the evidence of the countless Anglo-Saxon graves which
have been excavated of late years. We must trust the Fairford
or Ossengal cemeteries rather than the literary evidence of Bede
or the Beowulf, which are excellent for the seventh and eighth
centuries, but cannot be relied upon for the fifth and sixth.
Arms and armour had been profoundly modified in the interval.

It is doubtful whether even the chiefs of the first English
war-bands wore any defensive armour. Probably they, like their
gesit/is, used to go out to war in their tunics, with undefended
head and breast, and bearing the broad shield of linden tree

1 It is most curious to find these Saxons acting as cavalry, and stationed so far
eaat as Phoenicia. (See p. 43.)

64 THE ART OF WAR IN T THE MIDDLE AGES [500

alone. This was a round convex target like that of the Franks,
bound with iron at the rim, and furnished with a large projecting
iron boss. Often it seems to have been strengthened by a cover
ing of stout leather.

Of the offensive arms of the old English the spear was the
most prominent : they were in this respect still in the stage
which Tacitus had described four centuries back. The most
usual form of the weapon had a lozenge-shaped head, ranging
from ten up to eighteen or even twenty inches in length. Barbed,
leaf-shaped, and triangular spear-heads are occasionally found,
but all of them are far less common than the lozenge-headed
type. The shaft was usually ash, fastened to the head by rivets :
it seems to have averaged about six feet in length. The sword
appears to have been a less universally employed weapon than
the spear ; the usual form of it was broad, double-edged, and
acutely pointed. It had very short cross-pieces, which only
projected slightly beyond the blade, and a very small pommel.
In length it varied from two and a half to three feet. As an
alternative for the sword the old English often used in early
times the broad two-edged dagger eighteen inches long, re
sembling the scramasax of the Franks, which they called sea.r,
and associated with the Saxon name. The axe, the typical
weapon of the Frank, was rare in England, but the few specimens
that have been found are generally of the Prankish type, i.c. they
are light missile weapons with a curved blade, more of the type
of the tomahawk than of the heavy two-handed Danish axe of
a later day.

The organisation of the English conquerors of Britain differed
from that of the other Teutonic invaders of the empire in
several ways. They were not a single race following its
hereditary king like the Ostrogoths, nor were they, like the
Franks, a mass of small, closely-related tribes welded together
and dominated by the autocratic will of the chief who had united
them. They were not of such heterogeneous race as the so-
called Visigothic conquerors of Spain, nor, on the other hand, so
homogeneous as the Lombards of Italy. The Ostrogoths and
Lombards were nations on the march ; the Franks and Visigoths
were at least the subjects of one king. But the old English were
merely isolated war-bands who had cast themselves ashore at
different spots on the long coast-line of Britain, and fought each
for its own hand. They were but fragments of nations whose

5oo] THE CONQUEST OF BRITAIN 65

larger part still remained in their ancient seats. 1 Their chiefs
were not the old heads of the entire race, but mere keretogas,
leaders in time of war, whose authority had no ancient sanction.
No continental Teutonic State started under such beginnings :
the nearest parallel that we can point out is the time when the
Lombards, after the death of King Cleph, abode for ten years
without a king, and pushed their fortunes under thirty inde
pendent dukes. But this condition of things lasted but a few
years in Lombardy, and was soon ended by the outward pres
sure from Frank and East-Roman. In Britain it was more than
four hundred years before the Danish peril led to a similar
result.

The old English kingdoms, therefore, were the small districts
carved out by isolated chiefs and their war-bands. They were
won after desperate struggles with the Romano-Britons, who did
not submit and stave off slaughter like their equals in Gaul or
Spain, but fought valiantly against the scattered troops of the
invaders. If a mighty host commanded by one great king like
Alaric or Theodoric had thrown itself upon Britain in the fifth
century, the provincials would certainly have submitted : they
would have saved their lives, and probably have imposed their
tongue and their religion upon the conquerors within a few
generations. But instead of one Theodoric there came to
Britain a dozen Hengists and Idas, each with a small following.
The Romano-Britons were often able to hold the invaders back
for a space, sometimes to entirely beat them off. Even after the
Saxons had gained a firm footing on the southern coast, they
were unable to advance far inland for two generations. Hence
it came to pass that in its early stages the conquest was not a
matter of submission under terms, such as always happened on
the Continent, but a slow hunting of the Romano-Britons towards
the West and North. 2 In the first stage of the conquest, there
fore, the English kingdoms were almost wholly Teutonic, and the
survival of the Celtic element small ; yet it is certain that some
men of the old race still remained on the soil as laets and many
more as slaves. The realm of Kent or Sussex or Essex would
be composed of a heretoga who had become permanent and
adopted the title of king, of his personal oath-bound followers

1 At least this was the case with Jute and Saxon : the majority of the Angles did,
in all probability, cross the seas.

- This, one must certainly imply from Bede i. 1 5, and from Xennius.

66 [-Hi: ART OF WAR IN THE MIDDLE AGES [600

or gesiths, and of other freemen, some of noble blood (eorls), some
of simple blood (ceorls). Below them were the non-Teutonic
element a few laets and many more slaves. The kingdom of
Kent as it appears in the laws of King Aethelbert (A.D. 600)
still preserves the character of the days of the first conquest.
Having attained its full limits in a few years, and being cut off
from further expansion into Celtic Britain, its condition has
become stereotyped. In such a State the army consisted of the
whole free population, and was a homogeneous Teutonic body,
very unlike a contemporary Visigothic or Prankish host. The
simple freemen (ceorls) have a very important position in the
State : they possess slaves of their own (laws 16, 25) ; the fine for
violating their domicile is half that paid for violating an "cod s
tun" in the same way 1 Haws 13, 15); to put one of them in
bonds is a high crime and misdemeanour (law 24). Laets of
various standing exist, but evidently the free Teuton is the
backbone of the community. The king s dependants are but
slightly mentioned, nor does the word gesith occur in the code,
though it is found in the additions made to the Kentish law by
YYihtraed 2 ninety years later.

But the later and larger English kingdoms were of a some
what different cast. The picture of YVessex which we get in
Ini s Code, a production of about the year 700, gives us a less
simple and a less Teutonic realm than that of Aethelbert. 3 Even
before the coming of Augustine and the introduction of
Christianity, the English had begun to admit the Romano-Britons
to terms. 4 After a victorious campaign the cities were still
sacked and burned, but the Celtic country-folk were no longer
reduced to slavery or at the best to laethood, but were granted
an independent, though an inferior, status as freemen. The laws
of Ini speak of Welsh subjects of the king owning a half-hide
or even a whole hide of land. 5 They even serve in his retinue:
the horse-wealh who rides on his errands is specially mentioned, 6

1 So too for misdoings with a ceotl s slave the fine is half of that for meddling with
an eorl s (laws 14, 16).

- ^Yihtraed s laws, 5.

3 It has been lately suggested that Ini s Code is connected with the settlement of
newly- won British land rather than with the ordering of the whole of Wessex.

4 See, for example, Bede s account of the heathen Aethelfrith, "who conquered more
territory from the Britons, either making them payers of tribute, or driving them out,
than any other king or tribune of the English " (i. 34).

5 Law 32. 8 Law 33.

7oo] THE LAWS OF INI 67

and King Cynewulf had a Welshman among his gesiths. 1 We
are reminded at once of the Prankish king and his Gallo-Roman
antrustions on the other side of the Channel. 2 But something
more is to be noted in the Wessex of 700. Society seems to be
growing more feudal, and the nobility of service is already assert
ing itself over the old eorl-blood. We find not merely slaves
and Welshmen, but English ceorls under a Jdaford or lord,
to whom they owe suit and service. If they try to shirk their
duty to him, heavy fines are imposed on them. 3 We are
tempted to infer that a large proportion of ceorls were now
either the vassals of lords or the tribute-paying tenants OP royal
demesne land. 4 The king has geneats or landholding tenants,
who are so rich that they are twelve-liynde and own estates
even so large as sixty hides. 5 But the most important thing to
notice is that the king s comitatus seems to have superseded the
old eorl-kin as the aristocracy of the land. The " gesithcund man
owning land " is the most important person of whom the code
takes cognisance after king and ealdorman. Probably the
greater part of the old noble families had already commended
themselves to the sovereign, and entered the ranks of his sworn
companions. The actual name of the thegn only once appears
instead of that of gesith, but the thegnhood itself is evidently in
existence. There still exist, however, certain members of the
comitatus who have not yet become proprietors of the soil. The
" gesithcund men not owning land " inferior members of the
war-band who got but bed and board and weed and war-horse
from the king are valued at double a ceorl s price.

Military service is required from ceorl as well as gesith.
When the call to arms is heard, the landed gesith who neglects
it is to forfeit his estate and pay fyrdwite to the extent of
a hundred and twenty shillings. The landless gesith pays
eighty for such disobedience, the " ceorlish man " thirty shillings.

One clause (law 54) in the code is very important as giving
the first indication of the fact that armour is growing common.
A man weighed down by a great fine, it says, may pay part of it
by surrendering his byrnie [mail-shirt] and sword at a valuation.
Comparing this with the almost contemporary law of the

1 A.S. Chronicle, A.D. 755 ; but the event related occurred in 784.

- See p. 60.

3 Law 39. 4 Laws 59, 67, " paying gafol," rent or tribute, to him.

5 Law 19.

68 THE ART OF WAR IN THE MIDDLE. AGES [700

Ripuarian Franks, we note that Ini says nothing about the
helm and the bainbergce, whose price is settled under similar
circumstances by the continental code. 1 Apparently, there
fore, the byrnie was much more common than the helm in
A.D. 700.

From whence did the old English learn the use of their mail-
shirt? Possibly it was already known to them ere they left
Saxony and Jutland, though few but kings can have possessed it
at that early time. Conceivably it may have been borrowed
from the Welsh. If we can be sure that the Gododin poems are
fair reproductions of early originals, and were not wholly
rewritten, with new surroundings, five hundred years later, we
must hold that the use of armour no less than that of the war-
horse survived for some time in Britain as a legacy from the
Romans. A poem that claims a sixth-century origin speaks of
the " loricated legions " of the half-mythical Arthur : 2 another
praises at length the battle-steeds of Geraint, " whose hoofs were
red with the blood of those who fell in the thick of the battle."
Helm and corslet are mentioned almost as regularly as shield
and spear. 3 There is no antecedent improbability in believing that
such legacies from their old masters lingered on among the Celts
of Britain, as they certainly did among the Celts of Gaul. Perhaps
the Cymry taught the use of mail to the Englishmen, as the
Gallo-Roman taught it to the Frank. If so, the use of these
remnants of the old civilisation must have been mainly confined
to Eastern Britain. The wilder tribes of Wales, as we find them
in the later centuries, were neither wearers of armour nor com
batants on horseback. The loss of the plain-land of Loegria
and the gradual decay of all culture among the mountains of
the West, may account for the disappearance of the war-horse,
and even for that of the mail.

But, on the whole, it is more probable that the byrnie came to
England from the Franks rather than from the Celts. The
invaders seemed to have borrowed nothing save half a dozen
words of daily speech from the tribes whom they drove westward.

1 See p. 56. -Ancient Books of Wales, Taliessin, xv.

3 Take as examples Gododin, 14 (Battle of Cattraeth) : With his blade he would
in iron affliction pierce many a steel-clad commander." Or ibid, 38 : " From Edyrn
arrayed in golden armour, three loricated hosts, three kings wearing golden torques."
Ibid. 96 : " When Caranmael put on the corslet of Kyndylan and pushed forward his
ashen spear." Oi Taliessin, 14 : " Wrath and tribulation as the blades gleam on the
glittering helms."

THE SAXON HELM 69

It is noticeable, too, that mail begins to grow common in England
almost at the same moment when we saw it coming into ordinary
use on the other side of the Channel.

The Saxon helm, however, was certainly not borrowed from
the Franks. Though the crested helm of late-Roman type, such as
Merovingian warriors wore, is not unknown in English illustrated
MSS., yet the national headpiece was the boar-helm mentioned
so frequently in the Beowulf. A single specimen of it has been
preserved that dug up at Benty Grange in Derbyshire by Mr.
Bateman. This headpiece was composed of an iron framework
filled up with plates of horn secured by silver rivets. On its
summit was an iron boar with bronze eyes. 1 Another form of
helm was destitute of the boar ornament, and consisted merely of
a framework of bronze overlaid with leather and topped by a
circular knob and ring. Such was the specimen dug up on
Leckhampton Hill above Cheltenham in 1844. It is probable
that the composite headpiece of iron blended with horn or leather
is the early form of the Saxon helm, but that by the seventh or
eighth century the whole structure was solid metal. This at least
we should gather from the Beowulf, where " the white helm with
its decoration of silver forged by the metal-smith, surrounded by
costly chains," 2 the " defence wrought with the image of the boar,
furnished with cheek guards, decked with gold, bright and
hardened in the fire," 3 must surely refer to polished metal,
not to the less showy and less efficient helmet of composite
material. Unfortunately, in Christian times burial in full
armour ceased, so that the later helms are only preserved to us in
literary descriptions or in illuminated manuscripts. Many seem
to have been plain conical headpieces, quite unlike the classical
shapes ; others, again, resemble the crested Prankish helm of
which we have already spoken.

Both head armour and body armour appear so perpetually in
the Beowulf that we should be tempted to believe that they
must have been universal in eighth-century England. But in fact
the writer of the epic is using the poet s licence in making his
heroes so rich and splendid. Just as Homer paints Achilles
wearing arms of impossible beauty and artistic decoration, so
the author of the Beowulf lavishes on his warriors a wealth that
the real monarchs of the eighth century were far from owning.

1 Collectanea Antigua, vol. ii. - Bio. 1450.

3 Beo. 350.

70 THE ART OF WAR IN THE MIDDLE AGES [700

Helm and byrnie were still confined to princes and ealdormen
and great thegns.

Unmolested for several centuries in their new island home,
and waging war only on each other or on the constantly receding
Celt, the English retained the old Teutonic war customs long
after their continental neighbours had begun to modify them.
They never learned, like the Franks, to fight on horseback ; though
their chiefs rode as far as the battlefield, they dismounted for
the battle. Even in the eleventh century they still were so
unaccustomed to act as cavalry that they failed as lamentably
when they essayed it 1 as did Swiatoslafs Russians before
Dorostolon. One isolated passage in the Beowulf speaks of a
king s war-horse " which never failed in the front when the slain
were falling." 2 But we have no other indication of the use of
the charger in the actual battle ; perhaps the poet may have
been taking the same licence as Homer when he makes Greek
kings fight from the chariot, or perchance he is under some
continental influence. It is at any rate certain that in spite
of some pictures in English MS. copied from foreign originals,
the horse was normally used for locomotion, but not for the
charge.

Nor had the old English learned much of the art of fortifica
tion : they allowed even the mighty Roman walls of London
and Chester to moulder away. At best they stockaded strong
positions. The Anglo-Saxon Chronicle tells us that Bamborough,
the Bernician capital, was first strengthened with a hedge, 3 and
later by a regular wall ; but the evidence is late, and Bede tells
us that when in 65 1 Penda the Mercian beset it, he strove to
burn his way in by heaping combustibles against the defences
a fact which seems to suggest that they were still wooden. 4 The
plan, we read, must have succeeded but for the miraculous wind
raised by the prayers of St. Aidan, which turned back the flames
into the besiegers faces. If an actual stone wall was built across
the narrow isthmus of the rock of Bamborough, it was a very
unusually solid piece of work for old English engineers to take
in hand.

1 A.S. Chronicle, Year 1055.

"Then Hrothgar bade bring eight steeds within the enclosure with rich cheek-
trappings, on one of them was girt a saddle wrought with gold and bright treasures
the war-seat of Halfdan s son when he would enter on the sword-play : never did it
fail in the front when the slain were falling " (Beo. 1036-42).

3 A.S. Chronicle under A.D. 547. 4 Bede, iii. 16.

joo] THE OLD ENGLISH BATTLE 71

Hence it came that the wars of the English in the sixth,
seventh, and eighth centuries were so spasmodic and inconse
quent. Edwin or Penda or Offa took the field at the head of a
comparatively small force of well-armed gesiths, backed by the
rude and half-armed levies of the countryside. The strength
of their kingdoms could be mustered for a single battle or a
short campaign : but even if victory was won, there was no means
of holding down the conquered foe. The king of the vanquished
tribe might for the nonce own himself his conqueror s man and
contract to pay him tribute, but there was nothing to prevent
him from rebelling the moment that he felt strong enough. To
make the conquest permanent, one of two things was needed
colonisation of the district that had been subdued, or the
establishment of garrisons in fortified places within it. But the
English were never wont to colonise the lands of their own
kinsmen, though they would settle readily enough on Welsh
soil. Fortifications they were not wont to build, and garrisons
could not be found when there was no permanent military force.
No great warrior king arose to modify the primitive warlike
customs of the English till the days of Alfred and Edward the
Elder. Hence all the battles and conquests of a Penda or
an Offa were of little avail : when the conqueror died, his
empire died with him, and each subject State resumed its
autonomy.

The Anglo-Saxon battle was a simple thing enough. There
is no mention of sleight or cunning in tactics : the armies faced
each other on some convenient hillside, ranged in the " shield-
wall," 1 i.e. in cloje line, but not so closely packed that spears
could not be lightly hurled or swords swung. The king would
take the centre, with his banner 2 flying above his head, and his
well-armed gesiths around him. On each side the levies of the
shires would stand. After hurling their spears at each other
(the bow was little used in war), the hosts would close and
"hack and hew at each other over the war-linden," i.e.
over the lines of shields, till one side or the other gave way.
When victory was achieved, the conqueror thought rather of

1 1 he " Bord-weall " is of course merely a poetical expression for the wall-like line
of shielded men. It has nothing to do with locking shields after the manner of the
Roman ttstudo, with which it has been compared. Warriors in BcffiLitlJ 2980 hew
each other s helms to pieces " over the shield- wall."

2 The banner is mentioned both in Bede (King Edwin s) and in BeomulJ 2506.

72 THE ART OF WAR IN THE MIDDLE AGES [700

plundering the richest valleys in his adversary s realm than
of seizing the strategical points in it. Systematic conquest
as we have already observed never came within the
scope of the invader s thoughts : at the best he would make
the vanquished his tributaries.

BOOK III

FROM CHARLES THE GREAT TO THE BATTLE
OF HASTINGS

A.D. 768-1066

CHAPTER I

CHARLES THE GREAT AND THE EARLY CAROLINGIANS
(A.L). 768-850)

THE accession of Charles the Great serves to mark the
commencement of a new epoch in the art of war, as in
most other spheres of human activity in Western Europe. In
our second book we had to describe the military customs of
Frank and Goth, Lombard and Saxon, in separate sections. The
conquests of Charles combined all the kingdoms of the Teutonic
West into a single State, with the exception of England and the
obscure Visigothic survival in the Asturias. Races which had
hitherto been in but slight contact with each other are for the
future subjected to the same influences, placed under the same
masters, and guided towards the same political ends. The
rescripts of Charles were received with the same obedience at
Pavia and Paderborn, at Barcelona and Regensburg. For the
first time since the fall of the West-Roman Empire the same
organisation was imposed on all the peoples from the Ebro to
the Danube. The homogeneity which his long reign imposed
upon all the provinces of Western Europe was never entirely
lost, even when his dynasty had disappeared and his realm
had fallen asunder into half a dozen independent States. In the
history of the art of war this fact is as clear as in that of law,
literature, or art. In spite of all national divergences, there is
for the future a certain obvious similarity in the development
of all the Western peoples.

We have pointed out that under the later Merovings and the
great Mayors of the Palace the Franks were showing a decided
tendency towards the adoption of armour and the development
of cavalry service. It is under Charles the Great that this
tendency receives a definite sanction from the royal authority,

7 6 THE ART OF WAR IN THE MIDDLE AGES [768

and, ceasing to be voluntary, becomes a matter of law and com
pulsion. At the same time an endeavour is made to render the
old Prankish levy en masse more efficient, by making definite
provision for its sustenance and by enforcing discipline. Most
important of all is the introduction of a system under which the
universal liability to service remains, but the individuals on whom
the hereban falls are made to combine into small groups, each
bound to furnish one well-armed man to the host ; so that a
single efficient warrior is substituted for two, three, or six
ill-equipped peasants.

The reasons which led to the reforms of the great Charles
are not hard to seek. Under the later Merovings the Franks
were barely able to maintain their own borders : their usual foes
were the Saxon, Frisian, and Bavarian: expeditions against Spain
and Italy had almost ceased. This period of decay and unend
ing civil wars was brought to a sudden close by the onslaughts of
the Saracens in 725-732 : Charles Martel had fortunately come
to the front just in time to save the State. The next forty years
were a period of aggressive wars against the Saracen, the
Lombard, and the Saxon. Both Saracens and Lombards were
horse-soldiery, and we cannot doubt that in the wars with King
Aistulf and the Emirs of Spain the Franks were led to develop
their cavalry in order to cope with their enemies. They obtained
such marked success against each of their adversaries, that we
cannot doubt that their mounted men were growing more
numerous and more efficient than they had been in the seventh
century.

But Charles the Great undertook offensive wars on a much
larger scale than Pepin and Charles Martel. His armies went
so far afield, and the regions which he subdued were so broad,
that the old Prankish levy en masse would have been far too slow
and clumsy a weapon for him. An army of Neustrian and
Austrasian infantry could hardly have hunted the Avars on the
plains of the Theiss and the Middle Danube. The Prankish
realm had been so vastly enlarged that it extended, not as of old
from Utrecht to Toulouse, but from Hamburg to Barcelona.
To keep this mighty empire in obedience a more quickly-moving
force was required ; hence Charles did his best to increase the
number of his horse-soldiery. It was also incumbent on him to
raise the proportion of mailed men in his host : against the
well-armoured Lombard and Saracen, and later against the

786] CHARLES THE GREAT AND THE LOMBARDS 77

horse-bowmen of the Avars, troops serving without helm and
byrnie were at a great disadvantage.

The first ordinance bearing on military matters in the
Capitularies of Charles the Great is one showing his anxiety to
keep as much armour as possible within the realm. In 779 he
orders that no merchant shall dare to export byrnies from the
realm. This order was repeated again and again in later years,
in the Capitula Minora, cap. 7, 1 and again in the Aachen
Capitulary of 805 ; the trade in arms with the Wends and Avars
is especially denounced in the last-named document. 2 Any
merchant caught conveying a mail-shirt outside the realm is
sentenced to the forfeiture of all his property.

In the first half of his reign Charles issued a good deal of
military legislation for his newly-conquered Lombard subjects.
He imposed upon them the Prankish regulations on military
service, which made the fine for neglecting the king s " ban "
sixty solidi, the old Ripuarian valuation of the offence, and
the penalty for desertion in the field, " which the Franks call
keres/iscs" death, or at least to be placed at the king s mercy
both for life and property. 3 It is interesting to find in the
Lombardic Capitulary of 786 that the Lombards who are to
swear obedience to the royal mandates are defined as cavalry
one and all, being described as " those of the countryside, or men
of the counts, bishops, and abbots, or tenants on royal demesne,
or on Church property, all who hold fiefs, or serve as vassals
under a lord, all those who come to the host with horse and arms,
shield, lance, sword, and dagger." 4 The possession of this mass
of Lombard horsemen was of the greatest importance to Charles
in his wars with the Avars. Nearly all the fighting against
these wild horse-bowmen was done by the Lombards, under
Pepin, the king s son, whom he had made his vicegerent in
Italy. It was a Lombard host which in 790 pushed forward
into the heart of Pannonia, beat the Avars in the open field, and
stormed their camp. The slow-moving Austrasians meanwhile
had only wasted the Avaric borders as far as the Raab. A few
years later it was again the Lombard horsemen who practically
made an end of the Avaric power : under Pepin and Eric Duke
of Friuli they captured the great " Ring," or royal encampment
of the Chagan, hard by the Theiss, and sent its spoils, the

1 Cap. Min. 7 : " Ut bauga et bruniae non clentur negociatoribus."

- Cap. Aquisg. 7. 3 Cap. 7 ictttense, 3. 4 Cap. Langobardiae of 786, 7.

78 THE ART OF WAR IN THE MIDDLE AGES [803

accumulation of two centuries of plunder, to deck the halls of
Aachen, The Avars never raised their heads again, and fell
into decrepitude. If he had led only Prankish infantry levies,
Charles would never have been able to subdue this race of nomad
horsemen : the numerous Lombard knights, however, could both
pursue them and ride them down when caught, it is interesting
to note how the strong domineering spirit of the great king
inspired his new subjects to undertake and carry out an adven
ture which their own kings had never been able to achieve, for
the Avar raids had been a scourge to Friuli and Lombardic
u Austria " for two centuries, and no remedy had been found
against them.

The chief military ordinances of Charles the Great are
five rescripts dating from the later years of his reign the
Capitularc dc Exercitu Promovendo of 803, the Capitulare
Aquisgranense of 805, the later edicts issued from the same city
in 807 and 813, and the Capitulare Bononiense of 81 1. All these
deserve careful study.

The first of them, the edict of 803, is directed towards the
substitution of a smaller but better-armed force for the old
general levy. It ordains that the great vassals must take to the
field as many as possible of the retainers whom they have
enfeoffed on their land (Jiomines casati). A count may leave
behind only two of his men to guard his wife, and two more
to discharge his official functions. A bishop may leave only
two altogether. 1 Secondly, a new arrangement is made as to
the field service of all Franks holding land. Everyone who
owns four manst? or over, must march himself under his lord,
if his lord is serving on the expedition, under his local count
if the lord be busy elsewhere. To every man who owns
three mansi there shall be added another who has but one, and
these two shall settle between them for the service of one man
properly equipped: if the wealthier goes himself, the poorer
shall pay one-fourth of his equipment ; if the poorer goes, the
wealthier shall be responsible for three-fourths. Similarly, all
men owning two mansi are to be arranged in pairs : one is to
march, the other to provide half the equipment. And so, again,
holders of one mansus are to be arranged in groups of four : one
will go forth, the other three will each be responsible for one-

1 Cap. dc Exercitu Prommitndo, 4.

5 Cf. the English enactment about the man with five hides or over, on p. 109

805] THE CAPITULARIES OF CHARLES THE GREAT 79

fourth of his equipment. 1 The local counts are charged to see
that all men holding a mansus or more are placed in one of these
groups : those found unenrolled are to be heavily fined for
shirking the ban. 2 Thus we see that the service of the ill-
armed poor is lightened, and that of the well-armed rich strictly
enforced. The general result would be a decrease in numbers,
but a rise in average personal efficiency, in the host of the
realm.

The Capitulare Aquisgranense of 805 is intended to supple
ment the ordinance of 803. It orders that every man having
twelve mansi must come to the host in a mail-shirt : anyone
who has such armour and fails to bring it to the host is to forfeit
both the byrnie and any beneficium that he may hold from
the king. 3 The fine for neglecting the ban, or failing to be
enrolled in one of the contributary groups established in 803, is
to be half a man s substance ; three pounds of gold for anyone
holding land or chattels to the value of six pounds, thirty
solidi for a man owning three pounds, and so forth. 4 The
prohibition against selling arms outside the realm is re-enforced,
and it is enacted that no man shall carry weapons within his
own district in time of peace : " if a slave is found with a spear,
it shall be broken over his back." 5

The bulk of the army consisting of men owning less than
twelve niansi, it is obvious that the minority only were as yet
furnished with armour. All the men of the contributory groups
are evidently infantry armed with shield and spear alone.

Much more notable than the Capitulary of 805 is that of 807.
This carries the duty of providing warriors down to men holding
even less than the one mansus which was laid down as the base
of service in 803. For the future three owners of that limit, in
stead of four, are to furnish a man for the host, while six holders
of half a mansus, or possessors of ten solidi in chattels, are to
contribute to equip one of themselves. 6 Two separate clauses
deal with the service of the Saxons and Frisians. The former,
all apparently treated as belonging to the poorest class, i.e. being
all infantry, are to send one man in six for an expedition against
the Saracens or Avars, one man in three against the Slavs of
Bohemia ; but if the Wends and Sorbs, their immediate neigh
bours, are in arms, then the whole levy is to take the field.

1 Cap. de Exerc. Promov. I. : Ibid. 2. 3 Cap. Ayuisf. 805, 6.

4 Ibid. 19. 5 Ibid. 5. 6 Ibid. 807. S 2.

So THE ART OF WAR IN THE MIDDLE AGES [8n

The ordinance for the Frisians is quite different. The counts
all holders of a royal "beneficium," and all who serve on horse
back (caballarii omnes), are to march out whenever the ban is
proclaimed ; of the commons (pauperiores] every six men are to
join in equipping one warrior for the host. There is unfortunately
no statement of the limits of the class which served as mounted
men ; we should have been glad to learn its character. Not
improbably it may have consisted of the holders of twelve
inansi, and the personal retainers of the great vassals and
officials.

For the inner discipline of the host the Capitulare Bononiense
(811) is very important. We learn from it that those who
arrived late at the muster were punished by being compelled to
abstain from wine and flesh for just so many days as they had
fallen behind the appointed time. 1 Anyone found drunk in
camp was to be deprived of wine till the campaign was ended. 2
Every holder of a " beneficium " who deserted his comrades in
the hour of need, either from cowardice or from private feud,
was to forfeit his holding to the crown. 3 The provision of food
which each man was to bring to the host is defined as being
three months rations ; it consisted, as we learn from a later
document, of flour, bacon, and wine. 4 The three months were to
count from the border, with certain relaxations in favour of
those coming from afar. Thus anyone coming from beyond the
Rhine may count his three months commencing at the Loire,
and anyone coming from beyond the Loire may count his three
months from the Rhine. On the other hand, a dweller beyond
the Rhine going east may only count from the Elbe, and
a dweller beyond the Loire going south may only count from
the Pyrenees. 5 The Capitulare Bononiense is very clear on the
necessity for providing as many fully-armed men as possible : it
enacts that if any bishop or abbot finds that he has more byrnies
in store than he has to contribute men to the host, he must not
let them lie idle, but at once inform the king of their existence. 6
It also lays great stress on the necessity of all retainers follow
ing the host even when their lord is not present : if he neglects

1 Cap. Bon. 3. 2 Ibid. 6. 3 Ibid. $.

4 Cap. Aqziisg. 813, g 10. Cf. also the curious story about Charles and the
drunken guards in the Monk of St. Gall, book ii.

5 Cap. Bon. g 8.

6 Ibid. cap. 10. I presume that the king would either buy them at a valuation
or provide other men to wear them.

813] THE CAPITULARIES OF CHARLES THE GREAT Si

to forward them to the local count, he must pay the fine that
they have incurred by slighting the hereban. 1

The section on rations in the Capitulare Bononiense can be
supplemented by a clause of the edict De Villis Dominicis, which
lays down the rule that cars such as follow the host should each
be able to contain twelve bushels of corn, or twelve small barrels
of wine, and that each car should be furnished with a leather cover
pierced with eyelet holes, and capable of being turned into a
pontoon by being sewed together and stuffed (with hay?).
Each cart was to carry a lance, a shield, a bow and quiver pre
sumably to equip the driver in time of need. 2

Last of the military decrees of Charles the Great comes the
Capitulare Aquisgranense of 813, which contains several im
portant notices. It provides that the count, when his men are
mustered, must see that each has a lance, a shield, a bow, two
bowstrings and twelve arrows. No one is for the future to
appear carrying a club alone ; the most poorly-armed men must
at least have a bow. The stress laid on the bow in this document
and in the Capitulare de Villis Dominicis is important. The
weapon was practically new to the Franks, and the attempt to
make it universal was probably due to experience in war against
the Avars, 3 the only neighbours of the empire who made much
use of the weapon. Another clause provides that all the " men "
(obviously the household men) of counts, bishops, and abbots
must have both helm and mail-shirt. We get from section 10
of this document a glimpse at the existence of a military train :
on the royal cars are to be pickaxes, hatchets, iron-shod stakes,
pavises, rams, and slings (obviously machines, not merely hand-
slings). The king s marshals are to provide stones suitable for
casting from these fundibuli.

On all these documents the best commentary is the summons
which calls Fulrad, Abbot of Altaich, to the royal host in 806.
It is worth quoting at length. " You shall come to Stasfurt by
the Weser on May 20," writes the king, " with your men
prepared to go on warlike service to any part of our realm that
we may point out ; that is, you shall come with arms and gear
and all warlike equipment of clothing and victuals. Every
horseman shall have shield, lance, sword, dagger, a bow and a

1 Cap. Bon. caps. 7, 9. " Cap. de Villis, Dominicis, 64.

3 Rather the Avars than the Byzantines, I should imagine, as the contact with
the latter had been comparatively small, while the Avar wars were very long.

82 THE ART OF WAR IN THE MIDDLE AGES [806

quiver. On your carts you shall have ready spades, axes, picks,
and iron-pointed stakes, and all other things needed for the host.
The rations shall be for three months, the clothing must be able
to hold out for six. On your way you shall do no damage to
our subjects, and touch nothing but water, wood, and grass.
Your men shall march along with the carts and the horses, 1 and
not leave them till you reach the muster-place, so that they may
not scatter to do mischief. See that there be no neglect, as you
prize our good grace."

This is a summons to a tenant-in-chief (the phrase is already
to be found in Carolingian documents) to come forth with his
retainers for general service. It is noteworthy that all Fulrad s
followers are expected to appear on horseback ; there is no
mention of any foot-soldiery, or directions as to their equipment.
It is not definitely stated that all the abbot s horsemen are to
appear in mail ; the summons being dated before the laws of
807 and 813, it naturally contains no such order. Any of
Fulrad s men who had twelve mntisi would have been bound to
serve in a byrnie by the edict of 805, but compulsion is not yet
put upon the rest. The command to bring the bow is to be
compared with the contemporary attempt to make the infantry
adopt the same weapon. In neither case did the experiment
succeed. The very large quantity of provisions and the heavy
entrenching tools must have made the waggon train very cumber
some. It was evidently contemplated that the camp might
have to be fortified, in order to protect the mass of bag^a^e

f i

it is for this purpose that the iron-shod stakes and the spades
are required. Charles is also, as the last clause of the summons
shows, very anxious to avoid the cardinal vice of the old Mero
vingian hosts the plundering of the districts through which the
troops had to march before reaching the frontier. Hence the
very heavy load of rations which Fulrad is directed to bring with
him. If the train made the army slow to assemble and slow to
move, it at any rate enabled it to carry on operations even in a
hostile or a devastated district for several months, long after the
date at which a Merovingian expedition would have commenced
to starve and then to disband.

When all the royal commands were carried out under the

1 Reading caballis instead of ca.balla.riis, which last does not make good sense.
The only way of giving it a rational meaning would be to suppose that Fulrad had
other followers beside his horsemen, which does not appear.

8oo] THE BURGS OF CHARLES THE GREAT 83

royal eye, and Charles was ubiquitous, it is obvious that the
host of the early ninth century must have been a very different
weapon from the tumultuary hordes of the Merovings. Its
efficiency is best shown by the great king s conquests, and the
fact that when made they were retained. Charles was untiring :
if one campaign did not bring him to the desired end, he recom
menced his work in the next spring. In a specially difficult
conquest, such as that of Saxony, he even wintered in the hostile
districts, to prevent the rebels from having any opportunity of
rallying in his absence. In 785-786, for example, he not only
built forts and cut roads, but conducted repeated raids against
the surviving insurgents even in the depth of mid-winter.

But perhaps the most important of all Charles innovations is
his systematic use of fortified posts. When a district had done
homage and given hostages and tribute, he did not evacuate it as
his predecessors would have done, and leave it free to revolt
again at the first opportunity. He selected a suitable position,
a hill by a riverside was his favourite choice, and there erected a
palisaded and ditched " burg," in which he left a garrison. Each
post was connected with the next, and with its base on the old
frontier, by a road. Charles and his officers at last acquired a
very considerable skill in the laying out of entrenchments ; it
was unfortunate for the empire that his successors neglected the
art, till a long series of Danish invasions compelled them to learn
it again. Probably the most ambitious work of entrenchment
which was undertaken in his reign was the great circumvallation
round Barcelona, which was constructed in 800 by the king s son
Lewis and the levies of Aquitaine and Septimania. The army
lay around the town for the whole winter of 800-801, hutted and
girt by a double trench and palisade, to guard against sorties
from within and diversions from without. The works were so
efficient that the Moorish garrison, after a gallant resistance, was
starved out and forced to surrender. The burgs of Charles were
indeed a very successful expedient : it was seldom that they
were taken ; that of Eresburg only fell by treachery in 776,
though that of Karlstadt seems to have been fairly stormed by
the desperate assault of the Saxons (778). The use of these
fortifications was a new lesson in the art of war for Western
Europe ; the Teutonic nations hitherto had never even fortified
their own camps, much less had they thought of employing the
spade and iron stake for the holding down of conquered lands.

84 THE ART OF WAR IN THE MIDDLE AGES [782

Hence it came to pass that Charles made permanent conquests
where his predecessors had merely executed raids and imposed
tribute. So well chosen were the sites of his posts that many of
them have remained the centres of political life in the districts
where they were established down to our own da}-. Such were
Magdeburg, Paderborn, Bremen.

There are many points in the Carolingian armies on which we
crave information that Einhard and his fellows do not vouchsafe
to afford us. Of the proportion of infantry to cavalry and of
unarmed to mailed men in the hosts of Charles we are unfortun
ately unable to give any statistics. That, owing to his continuous
legislation on the topic, the mailed riders must have been a much
more numerous part of the army in 814 than in 770, is all that
we can say. One interesting passage in a chronicle relating to
the Saxon war of 782 seems to show that at least in some ex
peditions a very considerable part of a Prankish host must have
been composed of horsemen. The Counts Geilo and Adalgis,
marching against the rebels, find that Count Theuderich with
another detachment is converging on the enemy from a different
base. Eager that they should have the sole credit of the
victory which they supposed to be in their hands, they bade their
men snatch up their arms, " and hastened on as if they were
about to pursue a beaten army, not to fight an intact one, eacJi as
fast as tiis horse would go^ so that they came all in disorder
against the Saxons, who stood ranged in front of their camp."
The reckless attack was beaten off, and four counts, two missi
doininid, and more than twenty other persons of account, fell
" with many of their men, who chose to follow them to the death
rather than to survive them." If these words do not imply that
the whole of Geilo s and Adalgis forces were cavalry, they must
at least mean that so large a proportion of them were horsed that
the counts hoped to win without the aid of their infantry, which
in such a mad onslaught must have been left miles behind. 2
The latter, in all probability, is the real meaning of the passage,
and the desperate courage of the Prankish horsemen is to be
accounted for by the fact that they were the henchmen and
cnfeoffed retainers (homines casati] of the counts, whom they

" Proul quemque velocitas equi sui tulerat, unus quisque eorum summa
foblinatione conlendit " (Ann. Einh. 782).

2 The army had been raised in Thuringia and among the Francunian districts,
where we should expect to find more foot than horse.

34] THE STRATEGY OF CHARLES THE GREAT 85

refused to desert even in the hour of certain death. Probably
the infantry were left so far behind that they never came into
the fight.

Of the order of Prankish hosts in battle, i.e. whether the horse
stood on the wings or in front of the foot-soldiery, we are equally
unable to speak with certainty. Whether there was any larger
unit in the assembled army than the count and his local follow
ing we are never informed. That the host marched in divisions
with a rearguard and vanguard may be deduced from the
account of the disaster of Roncesvalles, where the rear f ii qui
novissimi agminis incedentes, praecedentes subsidio tuebantur " l )
were so far from the main body that they were cut to pieces
before their comrades could return to help them. A march in
parallel columns over open country can probably be traced in one
of the Avaric campaigns of 791 and the Saxon campaign of 804.

Perhaps the most scientific disposition of forces recorded in
all the wars of Charles occurs in a campaign at which he was not
himself present the invasion of Catalonia in 800-801. On this
occasion his son Lewis, who held the command, while under
taking the siege of Barcelona with one-third of his forces, placed
another third, under William Count of Toulouse, some leagues
west of the town to act as a covering army, while he himself
with the remainder took post nearer his base of operations in
Roussillon, ready to aid either of the other fractions that might
require his help. The Caliph of Cordova advanced from
Saragossa, but found the covering army so strongly posted that
he turned aside, and invaded the Asturias instead of entering
Catalonia. When he had retired, the covering force joined the
besieging force in building the trenches and winter camp, which
we have already had occasion to describe.

The best description of the appearance of one of the hosts of
Charles is unfortunately not that of a contemporary, though the
writer is careful to state that he had been in communication with
old men who remembered the emperor and had served in his
campaigns. This author is the Monk of St. Gall, who wrote
some sixty years after Charles death, and dedicated his work to
Charles the Fat, the unworthy great-grandson of the conqueror.
He is describing the Prankish host as it approached Pavia in
the Italian campaign of 773. Borrowing his words, as has been
suggested, from some lost poem contemporary with Charles,

1 Einhard, 9.

86 THE ART OF WAR IN THE MIDDLE AGES [814

he describes King Desiderius and his henchman Ogier the Dane
u-atching the long column of the invading army draw near. As
each body comes into sight, the king asks whether his rival and
the main host have not now appeared. Ogier replies again and
again that Charles is not yet at hand the numerous warriors
that have passed by are but his vanguard. At last the plain
grows dark with a still mightier column than any that have yet
drawn near. " Then appeared the iron king, crowned with his
iron helm, 1 with sleeves of iron mail on his arms, his broad breast
protected by an iron byrnie, an iron lance in his left hand, his
right free to grasp his unconquered sword. His thighs were
guarded with iron mail, though other men are wont to leave
them unprotected that they may spring the more lightly on their
steeds. And his legs, like those of all his host, were protected
by iron greaves. His shield was plain iron, without device or
colour. And round him and before and behind him rode all his
men, armed as nearly like him as they could fashion themselves ;
so iron filled the fields and the ways, and the sun s rays were in
every quarter reflected from iron. Iron, iron everywhere, cried
in their dismay the terrified citizens of Pavia."

The interest in this description of ninth-century armour is
that we learn that the short byrnie, not reaching below the hips,
was usual not only in the day of the great emperor, but in that of
his great-grandson, Charles the Fat, to whom the Chronicle of St.
Gall was dedicated. Greaves (ocreae, bainbergae) were evidently
in full vise when the description was written, but the thighs were
generally unprotected. That the sleeve is spoken of apart from
the byrnie as if it was a separate piece of a rmour is notable. The
description is borne out by a passage in the will of Count
Eberhard of Frejus, %vho in 837 leaves a helm with a hauberk,
a byrnie, one sleeve, and two greaves. Probably the sleeve
(manica, was only needed for the right arm, the left being
guarded by the shield.

The reign of Lewis the Pious (814-40) is as poor in military
legislation as that of his father had been rich a fact that might
perhaps have been expected when the character of the two
emperors is taken into consideration. By far the larger part of
Lewis capitularies deal with matters ecclesiastical. That the

1 Does "ferrea cristatus galea " imply that the helmet was a crested one, like those
in contemporary Frankish drawings in MSS.
- Monachus Sangallensis, ii. 26.

832] GROWTH OF FEUDALISM 87

organisation introduced by Charles was to some extent kept up
may be deduced from an edict of Lewis and his son Lothar,
dated 828, which orders the counts to inquire accurately whether
all the smaller landholders are properly enrolled in contributary
groups for service in the host, such as had been instituted in Soj. 1
Another document issued by Lothar at Pavia in 832 for his sub-
kingdom in Italy, recapitulates the prohibition against selling
mail outside the kingdom, and restates the old regulation that the
holder of twelve mansi must come to the host wearing a byrnie.

The time of Lewis being one in which the central power was
rapidly growing weaker, and the independence of the local counts
growing more marked, we cannot doubt that the mailed and
horsed retainers of these notables must have been continually
growing in numbers and importance as compared with the
unarmoured infantry of the local levies. The perpetual civil wars
which occupied the later years of Lewis reign are so full of
sudden desertions and inexplicable changes from side to side on
the part of large bodies of troops, that we see that the self-
interest of the counts has become of more importance than the
general loyalty of their subjects. Docile obedience to the royal
ban has been replaced by the most open treason. Owing to the
emperor s foolish liberality to his sons, the realm had four rulers
at once, and ambitious nobles could cloak their private schemes
by pretending to adhere to one or other of the rebellious young
kings. When the will of the local ruler became of more import
ance than that of the nominal head of the empire, the day of
feudalism was beginning to draw nigh. Already in the time of
Charles the Great we find the counts accused of pressing hardly
upon the smaller freemen, exacting from them illegal impositions
and services misdemeanours against which the capitularies
declaim again and again. Under weak rulers like Lewis and
his sons the evil was perpetually growing worse. At the same
time, the other characteristic sign of feudalism, in its social as
opposed to its political aspect the commendation of an ever
growing proportion of the smaller landholding classes to th^ir
greater neighbours was steadily going forward. Probably the
heavy burden of military service on distant frontiers, which
Charles had imposed on his subjects, was not one of the least of
the causes of the decay of the free peasantry. The duty which
had been comparatively light in the lesser realm of the Mero-

1 See Cap. Papiense, 832, 15.

88 THE ART OF WAR IN THE MIDDLE AGES [840

vings was immeasurably increased by the vast extension towards
the Elbe and Danube.

But the tendencies towards feudalism in the State, with the
corresponding tendency towards the depreciation of the national
levies of foot-soldiery, would have been comparatively slow in its
progress if it had not been suddenly strengthened by new in
fluences from without. The transformation of Western Europe
from the military point of view was to a very large extent the
direct result of the incursions of the Northmen. The lesser
troubles caused by the Magyars on the eastern frontier and the
Saracens in Italy were co-operating causes, but not to be
compared in importance with the effect of the raids of the
Scandinavians.

CHAPTER II

THE VIKINGS (800-900)

HOSTILE relations between the peoples of the North and
the Frankish kingdom had begun three centuries
before, on the day when Theudebert of Ripuaria slew Hygelac
the Dane, the brother of the hero Beowulf, on the Frisian shore
(515). But it was seldom that Frank and Dane had met; the
barrier of independent Saxons interposed between the two
races had always kept them apart. Down to the time of
Charles the Great the Scandinavian peoples were mainly engaged
in obscure wars with each other. The} are seldom heard of in
the North Sea. But at last the Frankish power, with its wealth,
its commerce, and its Christian propaganda, swept over Saxony
and moved on its boundaries to the Eider. It was within a very
few years of Charles first conquest of Saxony that the Vikings
(Wickings, men of the shallow fiords that face the Cattegat and
Skager Rack) made their first appearance on the scene as
serious disturbers of the peace of Western Europe. Perhaps
the first seeds of trouble were sown when Witikind the Saxon
fled before the sword of the Franks and took refuge in Jutland ;
we need not doubt that he told his Danish hosts terrible tales
of the relentless might, the systematic and irresistible advance
of the iron king of the Franks. The danger was now at their
doors the fate of Saxony might soon be that of Denmark. The
kings of the southern Danes gave shelter to Witikind, but they
sent fair words to Charles and did their best to turn away his
wrath. Yet, when Witikind yielded and was baptized in 785,
they must have felt that their own turn to face the oncoming
storm had now arrived. But for the next few years the great
Avaric war, the repeated local risings in parts of Saxony, and
the troubles of Italy kept the Franks employed elsewhere.
The first offensive strokes in the long struggle of Frank and

90 THE ART OF WAR IN THE MIDDLE AGES [789

Norseman were struck by the latter. Strangely enough, the
earliest recorded Danish raids were not aimed against the realm
of Charles the Great, but at more distant lands. The isolated
piracy of the " three ships from Herethaland " which burned
Wareham in Dorsetshire in 789 l is the first note of the appear
ance of the Scandinavians on the offensive. Four and five years
later two small fleets burned the rich abbeys of Lindisfarne and
\Yearmouth on the Northumbrian coast. In 795 the Danes
appeared so far west as Ireland, and destroyed the monasteries
of Rechru on Dublin Bay. It was only in 799, ten years after
the descent on Wareham, that the first recorded raids of the
Vikings on Prankish territory are noted. In that summer they
are said to have landed and made havoc both in Frisia and in
Aquitaine : the ever-watchful Charles was soon on the spot, and
ordered a fleet to be built to guard the narrow seas and the
coast of Neustria. But the only serious trouble which the
empire suffered from the Danes was a daring invasion of Frisia
by the warlike king Godfred in Sio. With two hundred ships in
his train, Godfred overran the Frisian Isles and extorted from
their inhabitants a large tribute. He spoke in his hour of
triumph of visiting the emperor at Aachen, but one of his own
men murdered him not long after, and his nephew and successor
Hemming at once made peace with the Franks and sailed
home ; the Danes were not destined to see Aachen till seventy-
six years later. The peace which Hemming promised was ill
kept, and several small raids on the northern coast of the
empire are recorded between Sio and 814. But these were ail
trifling matters. It was not till the reign of Lewis the Pious that
the Viking raids began to grow serious. During the later years
of Charles, the favourite sphere of activity of the Vikings was
Ireland, where, from 807 onward, they were making sad havoc
of the whole coast-line, and harrying one by one the rich
monasteries which lay along its bays and islands.

During their first tentative raids the Scandinavians had not
yet learned their own strength, nor were they such practised
marauders as they afterwards became. It is strange enough,
however, to see how suddenly they asserted themselves as a new
military power. At first they were sailing in unknown seas,

1 If that is the exact date : perchance the event was a few years later, for, though
the A.S. Chronicle enters the fact under 789, it says merely that it was "in King
Beortric s days " that the Vikings came to Wareham.

Sio] THE SHIPS OF THE VIKINGS 91

and their ships were but long, light, undecked vessels, that
seemed unfitted to face the wild Atlantic. That such craft, less
than twenty years after their first appearance in the North Sea,
should be risking their slight frames in rounding the rocky
shores of Donegal and Kerry, is the most astounding proof of
the wonderful seamanship of the Vikings. The boats were
essentially rowing, not sailing vessels ; their masts could be and
often were unshipped; they were only used when the wind set fair.
For their propulsion the Viking ships relied on their oars, from
ten to sixteen a side, though a larger number was employed
when boat-building had become more scientific, in the tenth and
eleventh centuries : even a second tier of oars seems to have been
occasionally used in these later times. The prows and sterns were
both high and curved. The former were often fashioned into the
dragon-shaped figure-heads which are so famous in the sagas.
There was no helm, but the ship was steered by a long oar
lashed near the stern, as is a Shetland sixern of to-day. The
early Viking vessels probably carried from sixty to a hundred
men only the larger constructions of the tenth century could
contain as many as two hundred.

The Danes, Swedes, and Norsemen of the year 800 were in
a state of society very much resembling that in which their
Anglian and Saxon kinsmen had come to Britain three hundred
years before. The raiders were not compact tribal bodies, but
war-bands of adventurers enlisted under the banner of some
noted leader, who was, as often as not, a mere warrior of renown,
not a member of one of the old royal houses. There are few
examples in the early Viking age of hosts commanded by the
national king, though the first notable raid that which King
Godfred led to Frisia in 810 was an exception to this rule.
The so-called sea-king was a mere war-chief, who might
relapse into obscurity when the expedition was over

"Solo rex verbo, sociis tamen imperitabat, "

as Abbo wrote, describing the leader who beleaguered Paris in 886.
The first Viking adventurers must have been no better
armed than the English raiders of the fifth century. If their
chiefs had a few helms and byrnies, spoils of war or merchandise of
the south, 1 the main body must have been wholly unmailed.

1 Finds in Sweden of the pre-Viking period have included fragments of byrnies
and iron helms (Montelius).

92 THE ART OF WAR IN THE MIDDLE AGES [8ro

After gold and silver, helms and mail-shirts were the form of
plunder \vhich the raiders most yearned for. This did not
endure for long: in less than two generations the Northmen had
armed themselves from the spoils of their enemies, and their own
smiths too had begun to essay the armourer s art. So essential
was mail to the professional Viking, whose hand was against
every man, whose sole occupation was war, that by 850 or 900
it was the rule, and not the exception, in their hosts. Their body
armour seems to have been exactly of the Prankish model ; the
helm, however, was pointed and often furnished with a nasal, un
like the old semi-classical shape which had prevailed among the
Franks down to the ninth century. 1 The shield was at first
round, like those of most of the other Teutonic races ; it was only
in the tenth century that it took the kite-shape familiar to us
in the Bayeux Tapestry and other contemporary works of art.
Shields were often painted red or some other bright hue, and, hung
on the bulwarks of the war-ship when the warriors were at sea,
produced lines of brilliant colouring along the gunwale.

The Danes used for offensive weapons spear, sword, and axe.
Their swords seem at first to have been of the comparatively
short, leaf-shaped kind, without a cross-guard, and very small in
the grip, which are habitually found in Northern excavations.
Later, they took to the longer and broader spatha of the Franks.
The axe was the more characteristic national weapon ; it was
not the light missile tomahawk (francisca) which the Franks
had been wont to employ in the sixth century, but a very heavy
weapon, with a single broad blade welded on a handle five feet
long. For proper use it required both hands : wielded by
muscular and practised arms, it would cleave shield and helm in
the same blow, strike off heads and limbs, and fell a horse without
difficulty. Both sword and axe-head were occasionally marked
with runes, as the sagas tell ; and specimens so adorned are to
be found in most of the Northern museums. The javelins of
the Scandinavians do not seem to have differed in any essential
point from those of the Franks and Angles. The bow they were
accustomed to use more than any of the nations with whom
they fought, for the English had never taken to it kindly, and
the edicts of Charles the Great had not succeeded in making it
popular on the Continent. Even the most noted warriors of the

1 The helm with nasal, however, was probably known to the Franks in the ninth
century ; it was most likely the "helmum cum dincto" of the Ripuarian Code.

830] EARLY RAIDS OF THE VIKINGS 93

North were proud of their skill with the arrow ; it was held an
honourable weapon by them, while among their enemies it was
the mark of the poorest military classes. Readers of the sagas
will remember the marksmanship of Olaf Tryggeveson and his
henchman Einar, and the celebrated shot with which King
Magnus slew Earl Hugh the Proud on Alenai Strait.

It was only some time after their appearance in western
waters that the Vikings acquired a complete ascendency over
the peoples of the older Teutonic realms. They were at first
cautious, attempting no ravage deep in the land, but absconding
after the plunder of some one seaboard town or abbey. The
Franks, Irish, and English seem to have been more angered than
terrified by the first raids, and several times caught and destroyed
considerable bodies of the invaders. 1 But the fleets grew larger,
the raiders more daring and better armed, their knowledge of
the strong and weak spots of the line of defence more perfect.
About forty years after the first plunderings in England, and
thirty after the first assault on the Franks, Western Europe
began to awake to the fact that the Northmen were beginning
to be no mere pest and nuisance, but a serious danger to Christ
endom. The landmarks of this period are the first serious inva
sion of the interior of Ireland by a great host under Thorgils
(832), the plunder of the rich haven of Dorstadt and the famous
cathedral city of Utrecht among the Franks (834), and the
erection of the first fortified Viking camp in England on the isle
of Thanet in 851. The invaders were beginning to grow so
numerous and so daring that it was obvious that some new
measures must be taken if their progress was to be checked.

Among the faction-ridden tribes of Ireland it was hopeless
to look for union or skilfully-combined resistance. More might
have been hoped from the English and the Franks. But the
contemporary political situation of neither of those peoples was
favourable. In England there was no central authority: King
Egbert, to whom the other princes of the Heptarchy had done
homage, was really supreme in Wessex alone. He had no power
to protect Northumbria or even Mercia : if he kept the bounds
of his own realm, it was all that he could accomplish. His
victory at Hingston Down over the combined bands of the
Vikings and the Corn-Welsh was a considerable success (838),

1 e.g. the Northumbrians destroyed in 794 the band that had sacked Wearmouth.
In 81 1 the Irish defeated a host in Ulster, and in 812 another in Connaught.

THE ART OF WAR IN THE MIDDLE AGES [850

but it did not and could not save the north or the east from
plunder. When Egbert died and his weaker son Aethelwulf
succeeded him, the supremacy of Wessex became purely
nominal : only once in his reign did Aethelwulf lead an army
beyond his boundary to help one of the other English States
(853). He was, in fact, a worthy and a well-meaning king, but
there was no touch of genius in him. Though he fought con
scientiously enough against the Vikings whenever they appeared,
and was more than once victorious, yet the fortunes of England
were steadily failing all through his reign. London and Canter
bury were both sacked in 850, and though Aethelwulf destroyed
at Ockley in Surrey the band that had wrought these ravages,
yet three years later another host came down on Wessex, and,
most ominous step of all, fortified themselves so strongly in the
isle of Sheppey, behind the marshy channel of the Swale, that they
could not be dislodged. 1 This was the second wintering of the
Danes in Britain. Meanwhile, if Wessex was faring ill, Mercia
and Xorthumbria were in a far worse case : both realms were
ravaged from end to end, and there remained hardly a town or
a monastery unburnt within their borders. Yet this was but the
beginning of evils : the period of settlement had not yet succeeded
to the period of sporadic ravages.

The Prankish Empire should have borne the brunt of the
contest with the Northman. But its condition was in some ways
even more unpromising than that of England. In the latter
country the tendency was still towards union : Wessex had just
permanently absorbed Kent and Sussex ; Mercia had almost
succeeded in doing the same to East Anglia, and had quite
amalgamated with herself the former sub-kingdoms of the
Hwiccas and Lindiswaras. 2 But in the realm of Lewis the Pious
the spirit of the times was making for disintegration rather than
for union. The old separatist tendencies of Aquitaine and
Bavaria, and the dislike of the Lombards for the Prankish yoke,
had disguised themselves in new shapes, and taken the form of
rebellions in favour of the ungrateful sons to whom Lewis had
distributed the government of those provinces. However much

1 The first was the wintering in Thanet narrated in A.S. Chronicle sub anno 851.

- From Offa : s murder of King Ethelbert in 792, onward to 825, East Anglia seems
to have been subject to Mercia : Ihe defeat of the latter by the King of Wessex brought
about that rising of the East Anglians in which two kings of Mercia, first Beornwulf,
then Ludica, perished.

850] DECAY OF THE CAROLINGIAN EMPIRE 95

the foolish tenderness of the emperor and the unfilial ambition of
his children may have supplied the formal cause of disruption,
its essential cause was the desire for independence on the part of
the subject nationalities. In all the realm the Austrasians were
the only people who consistently stood up for the cause of union
and imperialism. The civil wars of the sons of Lewis had begun
in 830, and for some time the ever-thickening Viking raids
seemed to the statesmen of the empire tiresome diversions,
distracting them for the moment from the all - important
questions whether Lewis should subdue his children or lose his
throne, and whether his youngest son Charles should or should
not obtain the kingly crown along with his brothers. Lewis
died in 840, after having seen the Danes cut deep into Frisia
and push daring raids up the Meuse and the Loire. After his
disappearance from the scene the civil wars only became more
constant and more chaotic: the bloody battle of Fontenay 541
where the might of Austrasia was for ever broken, settled the
fate of the empire. It was to split up permanently into inde
pendent national kingdoms, and never again was one sovereign
will to sway all the military force of the West, from Hamburg to
Barcelona, for a common end. 1

Xow, from some points of view it might appear quite probable
that three or four compact national kingdoms would be better
able to cope with the Vikings than the vast but somewhat
unwieldy empire of Charles the Great. But the dynastic
interests of the Carolingian house were still too strong to
allow real national States to develop themselves. Each king
was snatching at his brother s or cousin s provinces, in a vague
hope of reconstituting the empire for his own benefit. It was
not till the male line of the eldest son of Lewis the Pious died
out in Italy T875), and that of his second son in Germany ^911),
that those intermittent projects of reunion died out. As long
as they lasted they were wholly evil : while Charles the Bald
was getting himself crowned at Metz or Rome, while Wido was
overrunning Burgundy, or Carloman and Arnulf devastating the
Lombard plain, the Dane and Saracen and Magyar were tearing
their realms to pieces behind their backs. Kings immersed in
Imperial politics could not find time to discharge the simple
duty of superintending the local defence of their own coast and

1 Charles the Fat, though king of Germany, West Francia, and Lombard)-, never
ruled in the Burgundies, so the above statement is literally correct.

96 THE ART OF WAR IN THE MIDDLE AGES [850

border. It was not unnatural, therefore, that the years from
840 to 900 were the veiy darkest that Christendom had known
since the first formation of the Teutonic kingdoms in the fifth
century. No sign of better days is to be seen till Alfred s ex
pulsion of the Danes from Wessex (878), Count Odo s successful
defence of Paris in 885-886, and King Arnulf s great victory at
Louvain (891).

\Vc must now investigate the tactics of the Northmen, and
the various expedients which their English and Prankish
adversaries employed against them. By the middle of the
ninth century the invaders had increased into a formidable
multitude : their expeditions had been so fortunate that the
whole manhood of Scandinavia had thrown itself into the
Viking career. The Northmen were now members of old war-
bands contending with farmers fresh from the plough veteran
soldiers pitted against raw militiamen. They were far better
provided with arms than their adversaries : the helm and byrnie
seem to have become universal among them, while the English
fyrd and the Prankish local levies were still mainly composed
of unarmoured men. Only the thegnhood on this side of the
Channel, and the counts and their retainers on the other, were
sufficiently well equipped to be able to face the invaders man
to man. With anything like equal numbers the Vikings were
always able to hold their own. But when the whole country
side had been raised, and the men of many shires or countships
came swarming up against the raiders, they had to beware lest
they might be crushed by numbers. It was only when a fleet of
very exceptional strength had come together that the Northmen
could dare to disregard their opponents, and offer them battle
in the open field. Fighting was, after all, not so much their
object as plunder, and, when the landsfolk mustered in over
whelming force, the invaders took to their ships again and sailed
off to renew their ravages in some yet intact province. They
soon learned, moreover, to secure for themselves the power of
rapid locomotion on land : when they came to shore they would
sweep together all the horses of the neighbourhood, and move
themselves and their plunder on horseback across the land. To
fight as cavalry they did not intend : it was only for purposes
of swift marching that they collected the horses. The first
mention of this practice in England comes in the year 866
when "a great heathen army came to the land of the East

866] THE FORTIFIED CAMPS OF THE VIKINGS 97

Angles, and there was the army a-horsed." l Curiously enough,
it is in the same year that we first hear of the Danes in the
Prankish realm 2 trying the same device. Their base of opera
tions, however, was of course their fleet, and such excursions
always ended in a swift return to the boats. It was only when
a waterway was not available that the raiders dared to cut them
selves adrift from their vessels. As a rule, their method was to
work up some great stream, sacking the towns and abbeys on
each shore of it ; when they got to the point where it was no
longer navigable, or where a fortified city stretching across both
banks made further progress impossible, they would moor their
ships or draw them ashore. They would then protect them with
a stockade, leave part of their force as a garrison to guard it, and
undertake circular raids with the rest. On the approach of a
superior force they were accustomed in their earlier days to
hurry back to their vessels, drop down stream, and escape to
sea. But as they grew more daring they began to fortify points
of vantage, and hold out in them till the hostile army disbanded
for lack of provisions, or was dispersed by the advent of winter.
These strongholds were generally islands. The bands who
afflicted Xeustria made their habitual refuge the isle of Giselle
[Oscellus] in the Seine, ten miles above Rouen. Here they
stood sieges at the hands of Charles the Bald in 858 and 86 1.
But on one occasion at least they dared to fortify themselves
farther up the stream, at the place called Fossa Givaldi, near
Bougival, which seems to have been a peninsula girt round with
marsh rather than an island. In England they used Thanet, and
also Sheppey, for the same purpose. On one famous occasion
(871) they chose the tongue of land at Reading between the
Thames and Kennet for their stronghold. At the Loire mouth
they used the isle of Noirmoutier ; at the Rhone mouth the isle
of La Camargue was their refuge. Walcheren was in a similar way
their base for attacks on Flanders and Austrasia. The great host
which pushed up the Rhine in 863 defied the combination of the
Austrasians of Lothar II. and the Saxons of Lewis the German by
holding an island in the river near Neuss, from which they only
retired at their own good time. Against an enemy not provided
with ships of war these island posts were almost impregnable.

1 A.S. Chronicle, 866.

2 Annales Berlincnses, p. 84: " Nortmanni circiter quadringenti de Ligeri cum
caballis egressi, commixti Britonibus Cenomannis civitatem [Le Mans] adeunt."

98 THE ART OF WAR IN THE MIDDLE AGES [8 9 r

Even when the Danish fortifications were not pitched in an
inaccessible island, it was but seldom that the landsfolk were able
to break through the stakes and foss, manned by the line of
well-armoured axemen. The failures of Charles the Bald at
Givald s Foss (852), of Charles the Fat at Ashloh (882), of
Kthelred of Wessex at Reading (871), are well-known examples
of the danger of besetting a Danish camp. All the more credit,
therefore, is due to the few Christian kings who succeeded in
storming one of those formidable strongholds. King Arnulfs
capture of the great camp of Louvain in ^91 was probably the
most brilliant achievement of this kind recorded in the ninth
century. The host of Northmen had harried all Austrasia and
routed the local levies at the battle of the Geule. At the news
of this defeat the German king came flying from the eastern
frontier, and found the enemy stockaded in a place where the
Dyle forms a loop, with a ditch scooped in the marsh from
bank to bank, and a high rampart behind it. Undeterred by
the formidable barrier, Arnulf dismounted, bade all his counts
and mounted warriors do the like, and with drawn sword waded
through the marsh and began to hew down the palisade. His
men pressed in so fiercely behind him that after a bitter
struggle the shield-wall of the Danes gave way, and the whole
mass of Vikings were driven pell-mell into the flooded Dyle,
where they perished by thousands. Such a blow was worth
many victories in the open field, for it made the Danes doubt
their own power of resisting behind entrenchments in the
inland. No really dangerous Viking host ever essayed to strike
deep into the German kingdom after this defeat. For this
reason the storming of the Louvain camp deserves perhaps
an even higher place in military history than our own
Alfred s victory at Ethandun thirteen years before. For
the great king of Wessex, though he had beaten the Danes
in the open, did not storm their camp at Chippenham.
The stronghold only yielded on terms, and terms that,
considering the relative positions of Alfred and Guthrum
at the moment, must be considered very favourable to the
Danes.

When the Danes were surprised at a distance from their
camp and forced to fight without protection, they would draw
themselves up in the best position they could find, on a steep
hillside, as at Ashdown (871) or Ethandun (878), or behind a

9i i] TACTICS OF THE VIKINGS 99

stream ; they formed their shield-wall, 1 and fought the matter nut
to the end. On many occasions, when broken in the open by the
charge of the Prankish horse, they would retire behind the
nearest cover, a village, as at Saucourt (88 1) ; a church, as at
Brisarthc (866); a large building, as in the fight in Frisia in 873,
and there hold out till they either beat off the enemy, were
themselves cut to pieces, or at nightfall were able to abscond.

Nothing shows better the stubbornness of the Danes than
the way in which the} often by a desperate rally repaired a lost
battle. At the great fight in front of York in 868 they were
thoroughly beaten by Osbert and Aella, and forced back on
the town, but, rallying among the houses, they drove out the
Northumbrians, and finally slew both kings and won the day.
So, too, at Wilton in 872 they had been seriously repulsed by
Alfred, and had gone back for some distance, when at last,
seeing the Wessex men losing their order in the excitement of
victory, they rallied and redeemed the day. 2 The same had
almost happened at Saucourt, where nothing but the praiseworthy
efforts of King Lewis in restoring order among his men
prevented a success being turned into a disaster by the last
desperate effort of the Vikings. At the battle by Chartres in
911 they had been thoroughly defeated, and had lost six
thousand men, yet, when their beaten but undaunted host was
assaulted by the newly-arrived horsemen of the Count of Poictiers,
they turned on him, drove him off, and actually stormed his
camp, ending a day of failure by a success at nightfall. It was
hard to say that a Viking host was really disposed of till its
last banner had been cast down and its last man slain.

The Northmen seldom appeared as the assailants in the open
field like the English in the Hundred" Years War, they preferred
to stand on the defensive. Indeed, foot-soldiery fighting an enemy
whose force grew year by year to be more entirely composed of
cavalry were almost compelled to adopt such tactics. If they
did attack, it was generally by a surprise, as at the battle on the
Geule (891). On this occasion the Austrasian levies, marching
in disorder to find the Northmen, whom they believed to be

1 The shield-wall (testudo, as Asser pedantically calls it) is of course not a wedged
mass like the Roman testudo, but only a line of shielded warriors.

- I cannot see in either of these battles, as related in Asser and the authorities who
copied him, any trace of the "feigned flight " which some have detected. The Danes
seem to have been honestly driven back, and then to have rallied.

ioo THE ART OF WAR IN THE MIDDLE AGES [900

flying, \vere suddenly set upon by the invaders, \vho had
advanced to meet them instead of waiting to be attacked. The
Franks, being entirely out of array, were easily scattered.

\Ve must now turn to a consideration of the methods by
which the Franks and English endeavoured to beat off the
Vikings, at first with poor success. The one patent fact which
the kings of the house of Charles the Great and the house of
Egbert had to face was that the half-armed local levies of the
fyrd or the ban were insufficient to cope with the invaders. The
Frankish counts and the English ealdormen made many a
gallant attempt to beat off the raiders : sometimes they were
successful, but much more frequently they suffered a disastrous
defeat. The Vikings were too well - armed, too wary, too
experienced in every shift of war, to be adequately faced by
the raw militia opposed to them. Some more efficient body
of troops had to be improvised to meet them, some system of
defence devised to keep them from overrunning the open
country. Down to the ninth century the Frankish towns,
unless they had old Roman walls, were not provided with any
systematic protection ; the English were even more exposed,
for such of them as had the Roman circumvallation had allowed
it to moulder away ever since the first conquest, 1 while those
which had arisen since Roman days had never been fortified
at all.

1 York, for example, the greatest centre of Northern Britain in Roman days, was
in 867, in the words of Asser (sub ann. 867) imperfectly protected, for " non enim
tune ilia civitas firmos et stabilitos muros eo tempore habebat " ; therefore the
Northumbrians were able " murum frangere " by a rush to hew down a palisade, I
suppose. Canterbury- seems to have had walls rather early, however.

CHAPTER III

THE VIKINGS TURNED BACK CA.D. 9OO-IOOO) THE FEUDAL
HORSEMAN AND THE FEUDAL CASTLE THE TIIEGN
AND THE BURH

THE military history, therefore, of the ninth century sho\vs
two all-important movements directly caused by the
need of repelling the Danes. The first is the substitution of a
professional class of fighting men for the general local levies ;
the second is the development of a system of regular and
systematic fortification of the most important points in the
realm. The combination of the two movements gives us the
feudalism of the later Middle Ages. Though both are felt
equally in the English and the Prankish kingdoms, they take
somewhat different shapes on the t\vo sides of the Channel.
. The English thegn of the tenth century is not quite the same
as the Prankish vassal ; the English burh is by no means
identical with the continental castle.

The primary need of the Christian realms of the West was
a large body of courageous and well-armed fighting men,
capable of meeting the Northman man to man. Fortifications
are good things in their way, but they need trustworthy
garrisons. The most elaborate entrenchments serve no end-
as King Lewis of West Frankland found in 88 1 if those set
to defend them have not their heart in the business. His great
castle at Etrun was quite useless because none of his nobles
would undertake to hold the post of danger. 1

Now for the purpose of repelling the Vikings, the national
levy with its great tardily-moving masses of foot-soldiery had
been tried and found wanting. It was too slow, too ill-armed,

1 Annales Berlinenses, S8i : "Quod magis ad munimentum paganorum quam ad
auxilium Christianorum factum fuit, quia ipse rex Hludovicus invenire non poluit cui
illud castellum ad custodiendum committere posset."

101

IO2

THE ART OF WAR IN THE MIDDLE AGES [831

too untrained. The Danes if in small numbers took to their
boats or their horses and slipped away ; if in strong force they
put the local levies to rout. The only other military body in
the realm was the magnates and their retainers. Y\ e have
already seen that by the year 800 both the Prankish and the
English realms possessed an aristocracy, originally dependent
on the kings, and wholly official in character a "nobility of
sen-ice," to use the phrase that we have already had so many
occasions to employ. On the Continent it now included not
only actual holders of countships or great offices about the
court, but large numbers of persons, both lay and clerical, who
held "beneficia," feudal grants of land, from the king. Each of
these counts and vassi of various sorts had his bands of personal
followers, landed or unlanded, homines casati, or sub-tenants
with holdings of various size. The vassal-class was steadily
growing: a family which had once held office and received
grants of " beneficia " did not drop back into the ranks of the
ordinary freemen. The class, too, was already tending to
encroach on its poorer neighbours ; the counts were using their
official position, the holders of " beneficia" their less legal but
equally efficient powers of bringing pressure to bear on the
smaller men. Above all, the Church was extending its
boundaries on every side so rapidly, that, as early as 831, Lothar,
the son of Lewis the Pious, began special legislation against the
handing over of land to the " dead hand." When the hideous
distress caused by the Danish invasions came to aid the already
existing tendency towards feudalisation, the result was easy to
foresee. By the end of the tenth century the vast majority of
the smaller freemen had passed under the control of their
greater neighbours, either by voluntary commendation, or as the
result of deliberate encroachment.

Nor were the Danish invasions less powerful in hastening
the development of the other side of feudalism, the establishment
of the counts and dukes as hereditary local potentates, who
practically could no longer be displaced by the crown. There
was an obvious convenience during the time of trouble in letting
the son succeed to the father s government ; none would know
so well as he the needs and capacities of the district in which
he had been brought up. Moreover, there was danger, in those
days of incessant dynastic war, in the attempt to remove a
powerful noble from his father s post ; he might at once transfer

860] CREATION OF THE GREAT FIEFS 103

his allegiance to some other member of the Carolingian house.
Charles the Bald and his short-lived successors habitually
bought respite from the peril of the moment by letting the son
succeed to his progenitor s office. In the next generation, the
counties of West Francia had become hereditary fiefs, in which
the right of succession was looked upon as fixed and absolute.
In every one of the great vassal States of the later middle age,
we find that the commencement of succession within the family
starts from the years between the fatal battle of Fontenay and
the deposition of Charles the Fat. The first ruler in the county
of Toulouse who passed on his lands to his son, dates from
852 ; in Flanders, the date is 862 ; in Poitou, 867 ; in Anjou,
870; in Gascony, 872; in Burgundy, 877; in Auvergne, 886.
In East Francia, the development was not so rapid ; among the
newly-conquered German tribes, the Saxons and Frisians, there
still survived great masses of small freemen. But the tribal
dukes, whom Charles the Great had such difficulty in clearing
away, begin to reappear again before the end of the ninth
century. They start with Liudolf (died 866), the first Dux
Saxonnin of the new kind, who passed on his government to his
son Bruno, a great fighting man, who fell by the hands of the
Danes in the disaster on the Luneburg Heath in 880. By
forty years after his time, Bavaria, Lotharingia, Thuringia,
Suabia, have once more got dukes, and there were hereditary-
counts in Hennegau, Rhaetia, and many other smaller districts.
In Lombardy the same phenomenon crops up at about the
same time, and Ivrea, Friuli, Modena, Spoleto, appear as
hereditary States.

Now, as we have already seen, the Frankish counts and
vassals were accustomed to serve on horseback, and were
expected to bring their retainers to the host mounted like
themselves, even before the death of Charles the Great. The
development of feudalism, therefore, meant the development of
cavalry ; we can place the dismissal of the infantry of the
local levies into obscurity and contempt, and the entire
supersession of them by the feudal horsemen, between the death
of Charles the Great and the end of the century. Two short
quotations from chroniclers, dating the one from 820, the other
from 891, show how complete was the change. In the former
year Bera Count of Barcelona was challenged to a judicial duel
by Sanila, another noble of the Catalonian March. They

io 4 THE ART OF WAR IN THE MIDDLE AGES [866

fought, as the chronicler remarks, " equestri praelio quia uterque
Gothus esset" 1 Coming from the old Visigothic stock of
Septimania, it was natural for them to fight on horseback ; but
obviously this did not yet seem the most natural thing to a
Frank. How different from this is the note of the Monk of
Fulda, who states that Arnulf, when attacking the camp of
Louvain in 891, doubted for a moment whether he should bid
his knights dismount, "quia Francis pedetemptim certare
inusitatum est," 2 because it is not usual for the Prankish nobles
to fight on foot.

We may therefore conclude that, during the last seventy
years of the ninth century, the infantry were always growing less
and the cavalry more, just as the freemen were disappearing and
the vassals growing ever more numerous. Already, by the
middle of the century, the cavalry were the most important arm ;
in Nithard s account of the manoeuvres of his patron Charles
the Bald before and after Fontenay, the language used leads
us to think that most of the young king s followers must have
been mounted. Thirty years later, when this same king invaded
Australia to snatch territory from his nephew Lewis, he is made
to exclaim that "his army was ^o great that their horses would
drink up the Rhine, so that he might go over dry-shod." 3

The definite date at which we may set the permanent
depression of the infantry force in West Francia, is in 866.
From this year dates the celebrated clause in the Edict of Pitres,
in which Charles orders that every Frank who has a horse, or is
rich enough to have one, must come mounted to the host. His
words are that, " pagenses Franci qui caballos habent aut habere
possunt cum suis comitibus in hostem pergant," 4 and no one in
future is to spoil a man liable to service of his horse under any
pretence. The phrase pagenses Franci is evidently intended
to cover the surviving freeholders due for service under the
count. The " men " of the seniores were already obliged to
come horsed, by much older edicts.

After the recognition of the all-importance of cavalry in the
Edict of Pitres, we are not surprised to find that, twenty-five
years later, Kin? Odo, calling out the forces of Aquitaine against
his rival, Charles the Simple, found himself at the head often
thousand horse and six thousand foot. The chronicler Richer,

1 I ita. H .udtn ici, 33. - Ann. Fuld. 891.

; .-Inn. Fuld. 876. &gt; Edict of Pitres, 2. 26.

THE ADVANTAGES OF FEUDALISM 105

who tells of this levy, calls the cavalry milites, as opposed to the
foot-soldiery, /W/fej. 1 This is the first indication of the use of
the word miles, the warrior par excellence, for the mounted soldier.
A few years before, it would have been applied to all fighting
men ; we now see it starting on its way to become the designa
tion of the knight of the later Middle Ages. By the time that
the tenth century has arrived, the infantry in West Francia
seem wholly ^to have disappeared ; in such battles as the bloody
field of Soissons, where King Robert was slain, both armies,
without exception, seem to have been composed of mounted
men.

It is easy to understand the military meaning of the change ;
it was not merely that the impetus of the mailed horseman alone
could break the Danish shield-wall. Almost more important
was the fact that the cavalry only could keep up with the
swiftly-moving Viking, when he had purveyed himself a horse,
and was ranging over the countryside at his wicked will. The
local count who could put a few hundred mailed horsemen of
approved valour in the field, men bound to him by every tie of
discipline and obedience, and practised in arms, was a far more
formidable foe to the invader than ten thousand men of the ban.
Even if he could not check the raiders in open fight, he could
hang about their path, cut off their stragglers, fall upon them
when they scattered to plunder village or manor, intercept them
at every defensible ford or defile, where the few can block the
passage of the many, or circumvent them by cross roads which
the native must know better than the stranger. The moment
that the Prankish cavalry had reached its full development, the
career of the Viking was terribly circumscribed. At last, his
only method of dealing with it was to learn to fight on horseback
himself; 2 the art was acquired too late to influence the general
course of history in Western Europe, but by the end of the
tenth century the Norman horse was equal to any in Christen
dom. In the eleventh it was the flower of the chivalry of
the first Crusade.

The other expedient which the Franks used against the

1 " Odo congregari praecepit milites peditesque : quibus collectis in decem millibus
equitum peditum vero sex millibus erat," etc. (Richer, Si).

2 The first mention of Danes fighting on horseback seems to be at the battle of
Montfaucon (888). Abbo distinctly mentions that their horse and foot were separated,
and fought Odo apart. At Soissons (923) the Norman contingent in the army of
Charles the Simple all fight on horseback.

io6 THE ART OF WAR IN THE MIDDLE AGES [885

Northmen was the systematic and elaborate fortification of
points of vantage. The deliberate adoption of this policy is
laid down in the same Edict of Pitres (866), which we have
already had to quote for its importance in the development of
cavalry. But the actual scheme had been begun as early as 862.
It had occurred to Charles the Bald that the Danish fleets might
be kept from running up the rivers by erecting at favourable spots
fortified bridges, through which they would be unable to force
their way up stream. Pitres, some miles higher up the Seine
than the Viking stronghold on the isle of Giselle, was the chief
point which he pitched upon. Here he began to build a great
bridge with tctes-du-pont at either end ; it took some years to
complete, and the Danes still dashed through its unfinished
centre when they chose. He therefore constructed another
less ambitious bridge higher up, at Trilbardou, and by means of
it blocked the return of the raiders. After trying to break
through in vain, Weland, the Northmen s chief, gave up his
prisoners and plunder, on condition of being allowed to drop
down stream under the bridge unmolested. 1 The great structure
at Pitres was finished in 866, and smaller ones at Auvers and
Charenton-le-Pont were erected to guard the Oise and Marne,
as additional precautions. Most important of all, Charles made
the island-city of Paris throw bridges across to the northern
and southern banks of the Seine. These structures were
destined to have more influence on the future of the Viking
invasions than any of the new buildings down stream. For the
weak point of the plan was that the new bridges required
garrisons, and that a permanent force to hold them was hard to
find. A city like Paris could find men to man its own defences,
but isolated fortifications, like those at Pitres, required special
bodies of troops, which \vere not always at hand. Apparently,
they were broken through during the civil wars at the end of
the reign of Charles. At any rate, we find the West Franks in
885 devoting all their attention to building, as a substitute for
them, a new fortification at Pontoise. When the Danes came
up the Seine for the great siege of Paris, they had first to destroy
this obstruction. It made a creditable resistance, but, getting
no succour from without, was compelled to surrender. 2 Then,
pushing up to Paris, the invaders began the eleven months
beleaguering of the place. Paris had been more than once in

1 Annalcs fieri iiicnscs, 862. - Annals of St. Vedast, 885.

9 oo] FORTIFIED BRIDGES 107

Viking hands before Charles the Bald fortified it, 1 but now its
new defences enabled it to make a very different resistance.
Its gallant defenders, Odo and Bishop Gozelin, held it against
every attack, though the Emperor Charles the Fat gave them
little or no help. It is true that the Danes ultimately succeeded
in getting up the river, by laboriously dragging their vessels
across the flat shore round the southern bridge-head. 2 But they
could not take the place, and were at last glad enough to receive
a bribe and depart, leaving Paris free [886]. This successful
defence was almost as great a landmark in the history of West
Francia as the victory of Ethandun in England, or that of
Louvain in Austrasia.

The Danish ravages in Germany are of little importance
after the year 900 ; in the Western realm they con
tinued much later, but were never so threatening again as
they had been in the years before 886. For the future, the
Prankish victories are almost as numerous as those of the
Northmen. The fights of Montfaucon 888,,, Montpensier (892),
and Chartres (911), are all worthy of notice. They show that
the Franks were now no longer wont to shirk the ordeal of
battle, as they had been thirty years before, but fought
whenever they had the chance. As often as not the}- beat
back the invader, and kept the land free for a space from his
ravages. But it was the new fortifications, even more than the
battles, that saved France from utter ruin. When every town
had surrounded itself with a ring-wall, and endeavoured to
block its river with a fortified bridge-head, the Danes found
their sphere of operations much limited. They wanted plunder,
not year-long sieges with doubtful success at the end ; a gallant
resistance like that of Paris in 886, or Sens in 887, not only
saved the particular town that was holding out, but was of
indirect benefit to every other place that might have to stand a
siege hereafter, since it lessened the self-confidence of the Danes,
and forced them to contemplate the possibilities of similar
failures in the future. There was little gain in harrying the
open country ; not only had it been plundered already by fifty
previous raids, but now the peasantry flocked into fortified
places with all that was worth carrying away. The refuges and
strongholds were now numerous enough to afford shelter to the

1 It had been plundered in 845 and 856.

J/t/: Annuls, 8SS, and Abbo. See pp. 141-6 for a detailed narrative of the siege.

io8 THE ART OF WAR IN THE MIDDLE AGES [911

whole countryside ; for during several generations, bishops,
counts, abbots, and great vassals were hard at work, fortifying
every point of vantage. Not only great towns but small were
soon wall-girt, and private castles supplemented them as points
of resistance. A good deal of this work was only woodwork or
palisading, 1 not solid stone ; but if properly held, it yet served
its purpose.

It was the increasing difficulty and barren results of their
raids in France which led the Danes of Rolf in 91 1 to come to
the same bargain with Charles the Simple which the Danes of
Guthrum had made with Alfred of Wessex in 878. When the
king offered them a great Danelagh (as the English would
have called it), reaching from the river of Epte to the Western
Sea, Rolf and his followers accepted the bargain, and agreed to
draw together, settle down, and make a peace with the Franks.
Contrary to what might have been expected, the settlement was
on the whole a success from the point of view of Charles the
Simple. Gradually all the other Danish bands, leaving the
Loire and the Garonne mouths, gathered in to settle along with
Rolfs men. Like Guthrum in England, Rolf in Normandy was
a more faithful vassal than might have been expected, and even
sent his bands on several occasions to help the king against
native rebels. It was only when Charles had fallen into the
deadly snare of Count Herebert of Vermandois that the
Normans were turned loose again on the land (928). The
Franks proved now well able to defend themselves, and King
Rodolf cut to pieces at the battle of Limoges (929) the host
that tried to open once again the old route of the raiders into
Aquitaine. From the time of William Longsword onward, the
Normans appear no longer as heathen invaders from without,
but as unruly vassals within. By the year rooo they may for
most purposes be regarded as assimilated to their neighbours,
and Normandy is but the most important fief of the French
crown.

We must now turn back to the Danish invaders of England
and see how Alfred and his descendants faced the problem
which Charles the Bald endeavoured to solve by the aid of
cavalry, walled towns, and fortified bridge-heads. England had

1 For some account of the palisaded mounds of the continental nobles see Book
vi. chapter vii. The famous tower at the bridge-head round which so much fighting
raged during the great siege of Paris was only woodwork (see Abbo, i ).

9 oo] THE ENGLISH THEGNHOOD 109

no force of horsemen when the Viking raids began ; Ecgbert s
army was in this respect wholly unlike that of Charles the
Great. There was no question of reinforcing the cavalry arm
in England, for no such force existed. But in other respects
we find the Prankish methods reflected, with some variations,
on this side of the Channel. If VVessex had no mailed horse
men to serve as models for the reorganisation of the whole host,
she had heavily-armed foot-soldiery. The " gesithcund man
holding land," as Ini would have called him, the "thegn," as
the laws of Alfred name him, was practically equivalent to the
vassus or holder of a beneficiuin of the Continent. As among
the Franks the tendency of the ninth century was to drive all
men into the feudal hierarchy, the more important freeholders
becoming vassals, the less important serfs, so in England the
middle classes tend to be divided in a similar way. The richer
ceorls are absorbed into the thegnhood, the poorer sink into
subjection to their greater neighbours. In the laws of Alfred
it is easy to detect the fact that the free middle class is far less
prominent than it had been even in the time of the laws of Ini. 1
There were already " hlafords " and dependants in the day of
the elder code ; by the day of the later they must have been
the most important part of the population. How the change
came about may be gathered from the two important but
anonymous documents of the early tenth century, the one
dealing with Weregelds, the other with " The People s Ranks
and Laws," printed on pp. 79-8 1 of Thorpe s Early Englisli Laws.
In the Weregeld document the first draft states that " if a ceorl
thrive so that he have a helm and a coat-of-mail and a sword
ornamented with gold, but have not five hides of land to the
king s utu- are, he is nevertheless a ceorl. But if his son and
son s son so thrive that they have so much land afterwards,
the offspring shall be of gesithcund race, and the weregeld
2000 thrymsas." 2 The second draft, however, alters this into
"if the ceorl acquire so much that he have a coat-of-mail and a
helm and an overgilded sword, though he have not that land
[five hides] lie is sitlicund, etc. etc." 3 These two passages are
to be compared with the third in the " Ranks and Laws "
document, which states that " the ceorl who throve so that he

1 See Alfred s Laws, I and 37, particularly the latter.

-Weregeld Document, 9, 10, n.

Weregeld Document, 2nd version, 9, 10, n.

no THE ART OF WAR IN THE MIDDLE AGES [900

had fully five hides of his own land, church and kitchen, bell-
house and burhgeat, place and duty in the king s hall, was
henceforth of thegn-right worthy." l So was, it will be re
membered, " the merchant who fared thrice over sea at his own
expense." 2

The obvious meaning of these passages is that all holders of
five hides and upwards who were not already in the thegnhood
were now absorbed into it, and became charged with its duties
a? well as its privileges. Nay, even more, the ceorl who is fully
armed, though he have not the full five hides, is apparently
allowed to come into the gesithcund class, if the second version
of the Weregeld document is to be trusted. This is obviously
an endeavour to increase the thegnhood by encouraging all
ceorls to arm themselves as well as possible, and so obtain the
right to enter it. A similar object is served by allowing the
merchant to qualify for the same promotion.

The chief charge of the thegnhood was, of course, the duty
of following the host in full mail whenever the king took the
field. At all costs it was intended to raise the proportion of
well-armed men in the army to a maximum. It is worth noting
that we find, in the "Ranks and Laws " document, sub-tenants
holding under a " hlaford " who have reached the assessment of
wealth necessary to qualify for gesithcund rank : though not
directly sworn to the king, they are yet reckoned part of the
thegnhood, being called " medial thegns." 3

This new military force, therefore, which was produced by
incorporating all men of wealth and energy among the ceorls in
the enlarged thegnhood, was the main weapon with which Alfred
and his descendants faced the Danes. The great national levy
of the fyrd, though it still retained its miscellaneous armament
and its comparative inefficiency, was made somewhat more
useful by being divided into two halves, each of which was to
take the field in turn while the other tilled the countryside. 4
It served but as the shaft of the weapon of which the thegnhood
formed the iron barb.

Alfred did not neglect to follow the example of Charles the
Bald in the matter of building strongholds. Though the English
fortifications were as a rule mere palisades, the art of building

1 Ranks and Laws, 2. = Ranks and Laws, 6.

3 A phrase to be found in Canute s Herint-law, Leges C. 12.

4 A.S. Chronicle, 894.

907] THE VIKINGS TURNED BACK in

in England being far behind that of the Continent, they seem
to have been very effective in checking ravages. In a few cases
solid masonry seems to have been used for example, in patching
up the Roman wall of London, which Alfred "honorifice restaur-
avit 1 in 887. Alfred s warlike daughter Ethelflaed followed his
example in this respect at Chester in 907, where her rude repairs
can still be discerned among the Roman masonry. Canterbury,
too, had walls very early. But it was mainly by stake and foss
in concentric rings, enclosing water-girt mounds, that Alfred and
his children protected their frontier. Edward the Elder worked
against the Danelagh with such strongholds in a most systematic
way. His first line of burhs was to guard his own border, but
gradually he and his sister Ethelflaed pushed forward a second
line of forts of offensive purpose. These ixiTir/jeii.u.rct, as a Greek
would have called them, were built opposite every Danish town,
and furnished with garrisons to contain the sallies of the inhabit
ants and hold down the neighbourhood. Hardly one fell in
twenty years of war, so ineffectual were the siege operations of
the Danes.

It would seem that the system by which the burhs were
maintained was somewhat like that which Henry the Fowler "
established in Germany a few years after Edward had begun his
system of fortification. To each burh was allotted a certain
number of hides of the surrounding region, and all the thegns
resident in that district were responsible for the defence of the
stronghold. Each of them was bound to keep within the palisade
of the burh a house, which he must either inhabit himself, or fill
with a trustworthy representative able to bear arms in his stead.
Thus the original inhabitants of the burhs were a race of warriors,
though in later years, when the land settled down into quiet, and
town houses grew to be valuable property, the thegn might let
his tenement to a merchant or craftsman whose primary occupa
tions were not warlike. But in the early ninth century the burh-
men were essentially military in their pursuits. It would seem
that the cnihten-gilds, as we find them at Cambridge, London,
and elsewhere, were the original association of the settlers, who,
coming in from all sides to hold reconquered land, had no
common local tradition, and had to start new bonds of unity
among themselves. 3

1 Asser, 887. 2 See p. 120.

3 All these suggestions I get from Professor Maitland s invaluable Domesday Book

ii2 THE ART OF WAR IN THE MIDDLE AGES [896

One of Alfred s devices of fortification deserves a special
note, as being exactly copied from a feat of Charles the Bald.
In 896 a great Viking hosthad ascended the river Lea with all their
vessels. The king, choosing a place near the point where the Lea
runs into the Thames, rapidly erected two burns on each side of the
river, and then joined them so effectually whether by floating
booms or bridgework, we are not told that the Danes were
sealed up in the river, and, being unable to return to the Thames,
had ultimately to abandon their fleet, and retire overland, leaving
the Londoners to bring the ships in triumph back to their city. 1
This is perfect reproduction of the doings of the Prankish king
on the Marne in 862, 2 and it cannot be doubted that Alfred had
remembered the device, and deliberately copied it when the
opportunity came to him.

Far better, however, than any mere fortification of the
inland was the third great plan which Alfred adopted for bring
ing his Danish wars to a successful conclusion. He began to
build a strong fleet, able to contend at sea with the Vikings. In
the very first years of his reign he had seen that this was the
one really effective way of keeping the coast secure. As early
as 8/6, long before the peace of Wedmore, he gathered a few
ships and chased off a small raiding squadron. 3 After he had
gained some leisure by the peace with Guthrum, he kept con
tinually enlarging this force ; by 885 he had apparently some
dozens of ships afloat, though not enough to cope with the main
Viking fleets. 4 Later, as the Chronicle tells us, he built " long
ships that were full nigh twice as long as others ; some had sixty
oars, some more ; they were both swifter and steadier, and also
higher than others, and they were shaped neither as the Frisian
nor as the Danish vessels, but as it seemed to himself that they
might be most useful." The first successful doings of the new
squadron are recorded under the year 897. The nucleus of a
well-built fleet was perhaps the most precious legacy of all that
Alfred left to England ; his son steadily increased it. In 911

and Beyond. The " Burgal Hidage " which he gives in full, seems to belong to a
period early in Edward s reign, when the reconquest of Mercia and Essex was just
commencing. It has very full details of the division of all the shires south of Thames
into districts depending upon burhs, but becomes incomplete as we advance into the
regions which were beginning to be reconquered from the old enemy. There the system
%vas but just being built up.

1 A. S. Chronicle, 896. - See p. 106.

3 A.S. Chronicle, 876. * A.S. Chronicle, 885.

905] CAMPAIGNS OF EDWARD THE ELDER 113

Edward was able to send out some hundred ships to guard
the coast of Kent ; twenty years later the navy was so large and
so well practised, that /Ethelstan, Alfred s grandson, was able to
coast up the whole eastern shore of Britain unresisted, to invade
the domains of Constantine, King of the Scots. 1 The Danes of
Northumbria were in rebellion at the time, but they were
evidently unable to launch any squadron large enough to molest
his armament.

Among the Franks, then, mailed cavalry and systematic
fortification, among the English, mailed infantry, well-built
burhs, and a fleet, ultimately succeeded in curbing the raids of
the Northmen. It must not be forgotten, however, that to a
certain extent this triumph of the defensive over the offensive
was due to a change of conditions among the invaders themselves.
The success of the first Vikings was very largely due to the
fact that they were a mere army, with no homes or treasures
of their own to defend ; their wives and children and stored
property were all over seas in inaccessible Scandinavia, and they
had no base to defend save their fleet. Their sons, however,
who had rooted themselves down to a greater or less extent on
the Seine or the Humber, were in a very different case. The
moment that they began to make permanent encampments on
this side of the North Sea, they commenced to lose some of their
advantages. When they brought over their families, and began
to till the land in an English or a Prankish Danelagh, they
completely forfeited their strategical superiority. A Dane of
Normandy or the " Five Boroughs " had to protect his own
homestead as well as to endeavour to harry Neustria or Wessex.
An enemy who has towns to be burned, and cattle to be lifted,
is much more easily to be dealt with than a mere marauder who
has nothing to lose, and whose base of operations is the sea. In
the tenth century the tables were completely turned between
Englishman and Dane. Contrast with the dismal records of
the years 840-880 the following extract from the Anglo-Saxon
Chronicle, covering the fifth year of Edward the Elder :

" A.D. DCCCCV. In this year the " army" in East Anglia
[i.e. the Danes of Eoric, Guthrum s son] harried Mercia till they
came to Cricklade, and then went over Thames, and took about
Braden forest all that they could carry off, and then went home.
Then went after them King Edward, as speedily as he could

1 A.S. Chronicle, 933.

ii4 THE ART OF WAR IN THE MIDDLE AGES [924

gather his men, and harried all their lands between the Dikes
and Ouse, as far north as the Fens."

The retaliatory raid now followed an invasion as surely as
effect follows cause, and Eoric and hundreds of his warriors were
slain in the mere attempt to cut off Edward s last retreating
column, when the English wheeled round to return to Wessex,
after burning out every Danish farm in the East Midlands.
It is easy to understand the kind of reasoning that nineteen years
later caused all the English Northmen to take King Edward " to
father and lord," after he had gradually subdued East Anglia
and the " Five Boroughs " [924].

The later Danish wars in the time of Ethelred the Redeless and
Sweyn Forkbeard are no true continuation of the struggles of
Alfred and Edward a hundred years before. The later invaders
came for political conquest, not for plunder or land ; they were
in their ends more akin to William the Bastard than to Ingwar
and Guthrum. If Cnut conquered England, it was not the
individual superiority of his warriors that made him king. Dane
and Englishman were now armed alike, and fought with the
same weapons and in the same array. Ethelred fell because his
realm was in an advanced stage of feudal decomposition, due to
the mistaken policy of Edgar in cutting up England into great
Ealdormanries, whose rulers had grown too independent, and
failed to help each other in the hour of need. Instead of the
king heading the united thegnhood of England, backed by the
fyrd, we find great provincial satraps each at the head of his
local levy, maintaining a spasmodic resistance without mutual
aid. The fall of the Saxon house was due to the repudiation of
Ethelred by his own subjects, who disowned him and took Sweyn
and Cnut as their masters.

The rule of Cnut was notable in England not merely for his
temporary suppression of the danger of feudal disintegration, by
the rough method of summarily slaying the turbulent earls
Uhtred and Eadric, but for the introduction of a new military
element into the kingdom. He retained with him, when he
dismissed the rest of his host to their Danish homes, a small
standing army of picked mercenaries, his "huscarles," or military
household. To the number of several thousands, they constantly
followed the king, and formed the nucleus of any force that he
had to raise. They had a considerable advantage over the
thegnhood, as they had not to be called in from distant estates,

1054] THE HUSCARLES 115

but were always ready under the king s hand for any sudden
need. The institution survived the extinction of Cnut s house ;
Edward the Confessor and Harold Godwineson maintained
under arms this body of picked men. They were the core of the
hosts which smote Griffith the Welshman and Macbeth the Scot. 1
Their glorious end was to fall to the last man fighting round the
Dragon banner of Wessex, on the fatal field of Senlac.

The influence of the Danes had marked itself in English
warfare not only by causing the reorganisation of the military
force of the realm, and by precipitating the growth of feudalism,
but by certain novelties of equipment. It seems to have been from
the Vikings that the English got the kite-shaped shield which
superseded the round buckler in the tenth century. Still more
notable was the adoption of the Danish axe, a heavy two-handed
weapon utterly different from the light casting-axes of the early
English. By the time of Edward the Confessor it seems to have
been as common as the sword among the English thegnhood.
At Hastings it was the characteristic weapon of Harold s host.
In the far East it was so peculiar to the English and Danes of
the Byzantine Caesar s Varangian Guard, that they are habitually
described by their employers as the Tii/.r/.vq&gt;6pot.

1 In the battle against Macbeth there were slain " Osbem and Siward the Younger,
and some of Earl Siward s huscarles, and also many of the king s, on the day of the
Seven Sleepers" (A. S. Chronicle, 1054).

CHAPTER IV

THE MAGYARS (A.D. 896-973)

THOUGH the most formidable, the Vikings were by no
means the only dangerous enemies of Christendom in the
evil days of the ninth and tenth centuries. While the raids of
the Scandinavians were still terrifying the Franks and the
English, other enemies were thundering at the gates of the
southern and the eastern realms. With the Saracens who so
afflicted Italy in the days of Lewis II. and Berengar we need
not much concern ourselves. They are the same Cretan and
African Moslems with whom the Byzantine fought, and their

J O

methods of war are described in the chapters in which we deal
with the wars of the Eastern Empire. The more formidable
invaders of Germany require a longer notice.

The Magyars first came upon the horizon of the Western
Empire in 862, when the first of their bands which pushed across
Hungary made a transient irruption into the Bavarian Ostmark.
But they did not make a permanent appearance on the Imperial
frontier till 896, just when the worst of the Danish inroads were
ended in East Francia. King Arnulf had asked their aid in
892 against his enemies, the Slavs of Moravia, 1 and apparently
the easy success which they won over these tribes tempted the
Magyars to move westward. They had just been defeated by
their neighbours the Patzinaks, and, being driven out of their
previous homes on the Bug and Dnieper, came flooding through
the passes, of the Carpathians into the valleys of the Theiss
and Danube. The Avars had long sunk into nothingness, and
the Slavs who had succeeded them on the Middle Danube seem
to have been perfectly helpless before the invaders. So the
kingdom of " Hungary " came into existence in a single year,
with little fighting or opposition.

Ann. Fuld. 892.
116

896] THE COMING OF THE MAGYARS 117

The new neighbours of the East Franks were a people of
horse-bowmen, ever in the saddle, and entirely given up to war
and plunder. They were formidable on account of their swift
movements, their proneness to stratagems and surprises, their
wariness on the march, and their horrible greed and cruelty. .\&gt;
the chronicler Regino observed, " no man could stand against
them if their .strength and their perseverance were as great as
their audacity." 1 But they were incapable of besieging a walled
town, or of standing firm in the shock of hand-to-hand fi r htin rr .

i-&gt; O O

Their tactics in the West, as in the East, were to hover round the
enemy in successive swarms and overwhelm him with flights of
missiles. When charged by the heavy Prankish horse, they fled,
still pouring their arrows behind them.

The Magyars had been established for no more than three
years in their new abode, when they turned to plunder their
Christian neighbours. The poor spoil to be won from the Slavs
did not content them, and they were well acquainted with the
comparative wealth of the Franks and Lombards. The
ambassadors whom they sent to King Arnulf are said, indeed,
to have been mere spies, whose real object was to learn the
routes into the empire. 2 But their great irruption into Venetia
in 899, followed by an almost equally destructive raid into
Bavaria in 900, was a complete surprise to the Christians, who
had never suffered a serious invasion from the East since Charles
the Great had crushed the Avars ninety years back.

The moment which the Magyars chose for their invasion
was an unhappy one for Italy and Germany. In the former
country King Berengar was but lately freed from his first rival,
Lambert of Spoleto, and was just about to start on his contest
with a second pretender, Lewis of Provence (900-901). He was
also much distracted by Saracen raids on Latium and Tuscany.
In the German kingdom Lewis the Child wore the crown he was
a boy of no more than seven years old, the first minor who had
worn the Carolingian crown. No strong regent governed for
him, and the great vassals who had of late established themselves
in the new duchies were about to plunge into a series of bloody
and useless civil wars.

The extraordinary successes which the Magyars obtained

1 Regino, 889, i. 600.

2 " Missos illorum sub dolo ad Baioarias pacem optando, regionem illam ad explo-
randum transmiserunt " (Ann. Fiild. 900).

Ti8 THE ART OF WAR IN THE MIDDLE AGES [910

during the first thirty years of the tenth century were far more
the result of their enemies divisions and ill-governance than
of their own strength. The marvellous swiftness of their
incursions made it hard to catch them ; but if the eastern
frontier of Germany and the passes of the Venetian Alps had
been properly guarded by the systematic fortification of the
chief strategical points, and if the mounted levies of all the
frontier districts had been taught to act in unison, they could
have been held back. Neither in Italy nor in Germany were
these measures taken : the perpetual civil wars of the period
900-918 prevented any common action against the enemy.
The fortification of Ennsburg (901) to protect the eastern
frontier of Bavaria was an isolated and a wholly insufficient
precaution, but the only one which the reign of Lewis the Child
can show. Only once was a general levy of all Germany called
out against the Magyars (910), and then it fought in three
separate divisions many miles apart. The main body, with
which was the young king himself, was routed near Augsburg
by one of the usual " Turkish stratagems " so well known to
the Byzantines. While half the Magyars offered battle, and
turned to fly after a trifling resistance, the rest of their horde lay
hid in ambush till the German horse swept by them in the
disorder of victory. Then, pouring out on the flank and rear of
King Lewis s men, while their comrades wheeled and charged
the front, they won a great victory. 1

Pitched battles, however, were rare in the Hungarian wars,
for the raiders were more set on plunder than fighting. Xor
had they any bases (like the Danish ship-camps) to which they
were accustomed to return with their booty, and in which they
could be brought to bay. Carrying off only what could be borne
on pack-horses, they swept across the open country like a whirl
wind, and were often gone before the ban had time to assemble.
Ekkehard, describing the devastation of the lands by the Lake
of Constanz in 926, gives us a good picture of a Magyar raid.
" They went," he writes, " not in one mass, but in small bands,
because there was no Christian army in the field, spoiling the
farms and villages and setting fire to them when they had spoiled
them : they always caught the inhabitants unprepared by the
swiftness of their appearance. Often a hundred of them or less

1 A fair description of this fight is in Luitprand, Antapodosis, ii. 3, 4, much loaded
unfortunately with Virgilian quotations.

JS4] THE GREAT RAIDS OF THE MAGYARS 119

would come suddenly galloping out of a wood on to the prey :
only the smoke and the nightly sky red with flames showed
where each of their troops had been." 1

It was their rapid movement, far swifter even than that of
the Danes, which alone made the Magyars formidable. The
wide sweeps which some of their expeditions made far exceed
in length any Viking raid. The most formiddble of all were
those of 924, 926, and 954. In the former they swept through
Bavaria and Swabia, crossed the Rhine, ravaged Elsass and
Lorraine, penetrated into Champagne, turned eastward again
from the Ardennes, and returned across Franconiato the Danube.
In the second raid a still more astonishing feat of horseman
ship they passed the Venetian Alps, swept over Lombardy
(taking Pavia on their way), and then endeavoured to cross the
Pennine Alps into Burgundy. Checked in the passes by Rodolf of
Little Burgundy and Hugh Count of Vienne, they turned south,
and, taking a more unguarded route, burst into Provence and
Septimania. On their return journey Rodolf and Hugh cut off
many of them, but the bulk seem to have got safely back to
the Danube. 2 But the expedition of 954 was the most dreadful,
as it was the last, of all the great Magyar raids. In that year
the invaders wasted first Bavaria, then Franconia : they crossed
the Rhine near Worms. Then the rebel Duke Conrad wickedly
made a pact with them, and sent them guides to lead them to
the lands of his private enemy, Reginald-Duke of Lower Lorraine.
After harrying that duchy as far as Maestricht, they turned south,
and suddenly descended the Meuse into France, where no one
was expecting them. After burning every open village in the
territories of Laon, Rheims, and Chalons, they swooped down on
Burgundy. Here they met considerable resistance, but, forcing
their way through the Burgundians, they dropped down into Italy,
apparently by the Great St. Bernard, and finally hurried across
Lombardy and over the Carnic Alps back to their own land.
It was fortunate for Christian Europe that the Lechfeld victory
was to fall into the next year, and that the wings of the Magyar
vultures were to be for ever clipped by Otto the Great (95 5). 3

The remedies against the Hungarian raids were obviously the
same that were required against the Danish, swift cavalry to
chase the raider, and fortified places to afford shelter for the

1 Ekkehavd, c. 52. J Flodoard Ann. 924.

3 For this raid see Witikind, iii. 30, and Con!. Reguu, 954.

120 THE ART OF WAR IN THE MIDDLE AGES [925

population of the countryside, and place their wealth out of the
raiders reach. Unfortunately for Germany, its eastern frontier
was almost destitute of strong towns, and the Saxons and
Thuringians (as also the Bavarians to a lesser degree) were, of
all the Teutonic races, the least educated in cavalry tactics.
The Saxons, indeed, Were still for the most part foot-soldiery.

It was not till the advent of Henry the Fowler (or Henry the
Builder, as contemporaries more wisely called him) that any
check was set to the Magyars by either of the necessary
expedients. Henry from his first accession showed himself a
far more powerful prince than his unfortunate predecessors,
Conrad of Franconia and Lewis the Child ; but it was not till
he had been five years on the throne that he found leisure to
devise a system of defence against the invaders. Having in
924 concluded a truce with them, on the ignominious terms of
paying a large " Magyargeld " (if we may coin the word by
analogy from " Danegeld " j, he set to work to garnish the
frontier with new fortresses. In Saxony and Thuringia he
made every ninth man of the agrarii milites i.e. all men in
the countryside liable to the ban in time of need remove into
a walled place. He set the whole population to work day
and night to build these strongholds, and to construct houses
inside them : these being finished, he settled that each ninth
man should dwell therein, and take care of the eight neighbour
ing houses which his companions were to occupy in time of war,
while the eight were to pay the indweller in return one-third of
the net products of their lands. 1 All the legal and festal meet
ings of the district were to take place inside these new fortified
places, so as to induce the population to haunt them as much as
possible. Among these foundations were Merseburg, Ouedlin-
burg, Goslar, Nordhausen, Grona, and Pohlde. Henry also
compelled the abbeys to wall themselves in, and repaired the
fortifications of the older centres of population which dated
back to the burgs of Charles the Great. At first the new
strongholds were little more than thinly-inhabited places of
refuge, but ere long most of them became real towns. The
founding of Merseburg, the easternmost and the most exposed
bulwark of Saxony, deserves a special notice. Henry peopled it
by sparing the life of every " strong thief" that he caught, on
condition that he should go to dwell at Merseburg and receive a

1 All this is told very elaborately in Witikind, i. 35.

933] THE BATTLE ON THE UNSTRUT .21

grant of land in its environs. Strangely enough, this " legio
collecta a latronibus," as the chronicler calls them, 1 did very well
in their new settlement, and, like Romulus robber band, made
their city the centre of a strong community in a very few years.

Henry also devoted his years of peace to inducing the
Saxons and Thuringians to learn the art of fighting on horse
back. We are unfortunately without information as to the
means he employed whether he compelled the royal vassals
alone to serve mounted, or whether he also put pressure on the
freeholders who still abounded between the Elbe and YVeser.
We only know that when the next Magyar raid came, in 933, it
found North Germany for the first time possessed of " milites
equestri praelio probates," - as well as of a formidable range of
new fortresses.

The result was most satisfactory. When the invaders threw
themselves on Thuringia, their smaller bands were cut to pieces
by the local forces, who were now able to follow them at equal
speed. Their main army was attacked by Henry himself, who
had called up the cavalry of the neighbouring Franconian and
Bavarian lands to join the Saxons and Thuringians. By show
ing only a small force, the levy of Thuringia alone, "cum raro
milite armato," i.e. with few mail-clad men, he enticed them to
attack him. But when the whole German host suddenly
displayed itself and charged, the Magyars broke and fled with
out staying to fight. A few were caught and slain, a good many
were drowned in the Unstrut (which lay behind them 1 , but the
majority got off in safety and returned to Hungary. Such was
the battle at Riade, which modern historians have generally
called the battle of Merseburg, though it seems really to have
been fought nearer to Erfurt than to the other city.

Three years later Henry the Builder died, and was succeeded
by his still more famous son, Otto the Great. It may seem
strange that under such an able ruler the Magyar raids should
still have continued for more than twenty years after the day on
which his father had shown the true way of salvation. A closer
consideration of the facts shows that they are not so surprising
as they appear. The inroads after 933 are, with two exceptions,
by no means so formidable as those of the earlier years of the
century. These two really important invasions were carried
out, the one before Otto was firmly seated upon his throne, the

1 Witikind, ii. 3. - Ibid. ii. 38.

122 THE ART OF WAR IN THE MIDDLE AGES [950

other in the midst of a great civil war, and with the traitorous
co-operation of the rebels. For the greater part of the early
years of his reign (936-95 5 ) the realm was fairly free from raids, if
\ve except a continual bickering along the Bavarian frontier, in
which the Germans were more often victorious than unsuccessful.
The change in the spirit of the times since the battle of Riade is
sufficiently shown by the fact that the Bavarians are found
entering Hungary and wasting it as far as the Theiss in 950,
instead of waiting helplessly to see their own lands plundered, as
they had been wont to do thirty years before. 1 Saxony, safe
behind its new line of fortresses, seems to have held its own
without difficulty.-

The great Magyar invasions of 954 and 955 were a last
rally of the plundering hordes, conscious that their prey was
escaping them, and determined to try one more bold stroke before
it was too late. The chroniclers record the fact that they had
put every available horseman into the field, and that no such
host had ever been seen before. 3 We may compare the
Hungarian army that marched on Augsburg in 955 to the
Turkish army that marched on Vienna in 1683 it was the last
desperate effort of a power conscious that its superiority was
slipping from it.

Nevertheless, King Otto had every right to be proud of his
victory on the Lechfeld on St. Lawrence s Day. His realm was
still disturbed with the last throes of the great rebellion which he
had put down in the previous year, and, as there were dangerous
movements still working among the Slavs of the Lower Elbe
and on the Lotharingian frontier, he had not been able to
call out the full levy of his kingdom. There were hardly any
Saxons, Thuringians.or Lotharingians, and very few Franconians
with him. His army was composed of the cavalry of Bavaria
and S wabia, with a thousand Franconians, and the same number of
his Slavonic vassals the Bohemians, under their prince Boleslav.
Hearing that Augsburg was besieged, and that its garrison was
in great danger, Otto marched rapidly to its rescue, without
waiting for further reinforcements. He divided his army into
eight corps, legioncs as Witikind calls them, each entirely com-

1 Witikind, ii. 36.

- The Magyars raid into Saxony in 938 was most disastrous to themselves
(Witikind, ii. 14).

Gerh. v. Oudalr. 12.

THE BATTLE OF THE LECHFELD 123

josed of cavalry, and each mustering about one thousand men.
Three " legions " were Bavarian, two Svvabian, one Franconian,
3ne Bohemian ; the eighth was composed of the king s personal
following and of picked men from the other divisions ; it was
somewhat larger than the rest. The army was small compared
with that which had accompanied Otto on his invasion of France
in 946, when (as he boasted) " thirty-two legions had followed
him, every man wearing a straw hat," for in the summer heat
the Germans had marched unopposed through Champagne with
their helms at their saddlebows, and the peaceful headgear of
straw shading their brows. 1

On hearing of the king s approach, the Hungarians hastily
raised the siege of Augsburg, and drew themselves up on the
broad and level Lechfeld, a region very well adapted for the
practice of their usual Parthian tactics. Otto, however, moved
to meet them through broken ground which was unsuitable for
their manoeuvres, and then camped by the side of the Lech.
He drew up his army in a single line of corps, his own chosen
band in the centre, on its right the three Bavarian " legions " and
that of the Franconians, on his left the two Swabian divisions.
The Bohemians, whether because their loyalty was doubted or
because they were considered less solid troops, were placed behind,
in charge of the baggage. They were a camp-guard, not a reserve.

The Magyars soon came in sight a confused weltering mass
of hundreds of small troops ; the German chronicles say that they
were a hundred thousand strong, and, however exaggerated the
figure may be, they no doubt many times outnumbered Otto s
host. They had crossed the Lech far sooner than had been
expected. Their first manoeuvre was characteristic : while some
of them threatened the German front, a great body slipped off
to the left, apparently unseen, and suddenly fell upon Otto s
camp. The Bohemians left there on guard were routed after a
short struggle. The Magyars then suddenly changed their
direction, and charged in upon the rear of the two Swabian
corps of the king s left wing. Taken by surprise by this attack
from an unexpected quarter, the Swabians were defeated, and
driven towards the German centre : Otto then sent the Franconian
corps from his right wing to aid them. Led by Duke Conrad, a

1 Witikind, iii. 2. The straw hat was a specially Saxon head-dress for summer
wear. See the passage from Rather of Verona, quoted in Pertz s edition of Witikind,
p. 451.

124 THE ART OF WAR IN THE MIDDLE AGES [955

lately pardoned rebel who had to win back his reputation for
loyalty, the Franconian horse charged with such a fierce shock
that the Magyars were completely routed, and fled in disorder
to join their main body. Otto meanwhile, with his own division
and the Bavarians, had been watching and containing the rest of
the Magyars. When he saw the horde which had turned his flank-
crushed by Conrad, he hastily rearranged the disordered left
wing, and ordered a general charge of his whole line. 1

The Magyars, dismayed by the disaster which had befallen
their detached corps, made a poor resistance. They were
indeed wholly incapable of standing up to the Germans man to
man : their horses were smaller, and very few of them wore any
defensive armour.- After letting fly a few volleys of arrows, they
wheeled off and fled. Many were overtaken and slain, for their
horses were fatigued by the first fight ; more were drowned in
the Lech, for its farther bank was steep, and they could not
readily climb the slippery slope ; they had easily descended it as
they attacked, but found it almost impossible to mount on
their retreat.

Otto s host had suffered severely in the first fight, but lost
few men in the second ; Duke Conrad, however, who had
loosened his hauberk to take the air, received a Parthian shaft in
his throat at the very moment of victory, and was left dead on
the field. On the same evening the Magyar camp was taken
and plundered. For the next two days the army pursued the
flying foe, many of whom were cut off as they fled by the
Bavarian peasantry. Three great chiefs who fell into Otto s
hands were incontinently hung.

So ended, as Witikind remarks, the greatest victory which
Christendom had won over the heathen for two hundred years ;
he was thinking, no doubt, of Poictiers [723] as the last fight that
could fairly be compared with the Lechfeld. 3 It is only fair,
however, to remember that Henry the Builder s success at
Riade, though less showy and less complete, was far more truly
the turning-point of the history of the Magyar invasions than
the battle of the Lechfeld. Since 933 Germany had found the
raiders much less formidable than before, and the invasion of 955

r Thietmar is apparently wrong in making the battle last two days ; in Witikind the
whole of the fighting takes place on St. Lawrence s Day, August 10.

" Maxima enim ex parte nudos illos armis omnibus cognovimus," says Otto in the
speech which Witikind puts into his mouth (iii. 46).

: Wit. iii. 49.

955] THE MAGYARS TURNED BACK 125

was a desperate final rally. Just as in the history of the Otto
man assaults on Christian Europe we place the real moment of
greatest danger during the siege of Vienna in 1529, not during
that in 1683, so the most threatening time of the Magyar attack
was undoubtedly in 933, when they had never yet received a
check of importance, and not in 955, when they had already been
met and turned back many times by Otto and Otto s generals.

The danger, at any rate, was now wholly past. That it ever
had grown great was owing to the anarchy of the reigns of
Lewis the Child and Conrad the Franconian. In less than a
generation after the Lechfeld the roles of German and Magyar
were wholly changed : the Christian is always advancing and the
pagan recoiling. Otto, too, was able to cut a new " march " out
of the Pannonian lands which the Magyars had devastated and
occupied in his grandfather s time. This was the new Bavarian
Ostmark (973), destined to be famous under the name of Austria
for many a future generation.

CHAPTER V

ARMS AND ARMOUR (SdO-IIOO)

WE have seen that down to the time of Charles the Great
there had been comparatively little alteration in the
character of arms and armour since the days of the first founda
tion of the Teutonic kingdoms in the fifth century. In the ninth
century, however, we find a gradual change coming over the outer
appearance of the warriors of Christendom. Not only do a
much greater proportion of them wear defensive arms, but the
arms themselves begin to change in appearance. All the altera
tions are in the direction of securing greater protection for the
wearer. The short byrnic reaching to the hips and the open
Prankish helm seem to have been regarded as insufficient against
the Danish axe and the Magyar arrow.

One of the first changes consists in the adoption of the
hauberk (" hals-berge," or neck-protection) for the defence of the
throat, neck, and sides of the face. The earliest form of it was
simply a thick leather covering hiding the ears and neck, and
probably was fastened to the rim of the helm, like the camail of
modern Sikh or Persian headpieces. In this primitive shape it
is merely an appendage of the helm ; and when Count Eberhard
of Frejus records in his will (837) a helmnm cum halsberga,
we must think of it as meaning no more than this. Representa
tions of such hauberks may be seen in the St. Denis chessmen
figured by Viollet-le-Duc in his Mobilier Franqais? or the
warriors in the Stuttgart Psalter. The next form was more
complete : the material of the hauberk was changed to fine chain-
mail, and it was fitted more tightly to the head and brought
forward to cover the chin and neck. In this shape it was
probably formed into a coif or hood, the part covered by the
helmet being now leather, and the mail beginning where the

1 Vol. v. p. 67. But their date is much later than Yiollet supposed.

126

THE HAUBERK 127

headpiece no longer protected the skull. The lower edge of
the hauberk was sometimes tucked under the upper edge of the
byrnie and sometimes hung above it, for the two had not yet
become one garment.

This was the universal wear of well-armed warriors in the
tenth and eleventh centuries. The poorer men had only the
short mail-shirt, the richer supplemented it by the hauberk. We
find clear traces of its use in incidents such as that at the battle
of Soissons in 923, where King Robert, to make himself known,
" pulled out his long beard from under its covering," l that the
enemy might see it. So, too, Duke Conrad at the Lechfeld
received a mortal arrow-wound in the throat, because, overcome
by the heat, he had loosened his hauberk to take the air in
the moment of victory. 2

The next step in the development of this piece of armour
was that it was joined to the mail-shirt so as to form a single
garment, like an Esquimaux skin-coat. But this did not occur
till the end of the eleventh or beginning of the twelfth century.
Most of the warriors of the Bayeux Tapestry wear mail-shirts not
joined to their hauberks, for in several representations of byrnies
not in actual use we see that they have no hoods. When in the
twelfth century hauberk and byrnie became one, the name of the
former was often used to cover the whole suit a fact which has
caused much confusion to those who, knowing the term in this
late use, have not seen that it was at first a mere cheek-guard
hanging from the helm.

The helm itself changed entirely in shape in the ninth century.
The open crested Prankish helmet with its peak disappears, and
is superseded by a crestless conical headpiece. The latter shape
is better for turning off sword or axe blows, but it is probable
that it came in not merely for that reason, but because it could
be worn more easily with the hauberk. The older crested helm
stood out too far from the face and was too open to go well with
the new appendage. Probably, too, it did not fit so tightly to the
head, so that if worn above a hauberk of the later shape it would
be more likely to be knocked off than the new conical helm.
After the ninth century we never find the old crested Prankish

1 " Barbam obvelatam detegit, seseque esse monstrat " (Richer, i. 46). The other,
reading " barbam lorica extraxit," presupposes a lorica covering the chin, i.e. furnished
with a complete mail coif, which does not seem to have yet existed in 923.

2 Witikind, iii. 47.

i28 THE ART OF WAR IN THE MIDDLE AGES [950

shape in real use, though it still occurs occasionally in illustrated
manuscripts, copied from originals of an earlier time with too great
fidelity.

In the tenth century the conical helm receives a new addition
in the shape of the nasal, a projecting iron bar to guard the nose
from down-cuts which had been turned by the headpiece. The
device had been known earlier, 1 but only became really common
after 950. It prevailed from that date till the second half of the
twelfth century, when it was superseded by the " pot-helm " cover
ing the whole face, such as that seen on the great seal of Richard I.

Not only headgear and throatgear began to change in the
ninth century, but also the mail-shirt itself. It had hitherto
reached to the hips alone, but now began to lengthen itself
towards the knees. Horsemen fighting foot-soldiery armed with
heavy striking weapons Hike the Vikings), are specially liable to
receive cuts at and just above the knee. It was no doubt to
guard against this danger that the byrnie grew longer and longer
till it touched the calves. To make riding possible, it had to be
split at back and front, for a space of some thirty inches or two
feet from its lower edge. This divided shirt when drawn by an
incapable artist gives the impression of a pair of mail breeches,
but such garments were not common till much later.

The sleeves of the byrnie were still wide and short in the
tenth century, and far into the eleventh, so that the lower arm
had no protection. How wide they were about 923 may be
gathered from the fact that King Robert was killed at Soissons
by a lance which went up his sleeve, and then bore downwards
into his side and through his liver. 2

From this short sketch it can easily be seen that the warrior
of 1050, with conical helm and nasal, hauberk covering his ears
and throat, and long mail-shirt reaching below the knee, was
entirely different in appearance from the Carolingian fighting
man, who still preserved a certain resemblance to the late- Roman
soldier. He was also, it must be. owned, more effectively armed,
if less sightly to look upon. The covering of ring-mail was not
yet growing so heavy as to incommode or fatigue the wearer.

To complete the contrast, we must add that by 1050 the
kite-shaped shield had wholly superseded the round shield for
cavalry, though the latter was still often used by the despised
foot-soldiery. A large round shield is a great encumbrance to a

1 Helnium cunt directo occurs in the Ripuarian Laws. " Richer i 46

looo] AXE AND SWORD 129

rider, 1 who can only wield it with his upper arm, since his hand
is busy with the reins : while a small round shield gives poor
protection against arrows and javelins, though when used by a
skilled warrior it is effective enough against sword or lance.
The kite-shaped shield, on the other hand, has the advantage of
covering the greater part of the body without swelling to the
unnecessary breadth of the round shield, or hindering the outlook
on the left side to the same extent. Thus its advantages were
just those which led the Romans, twelve hundred years earlier,
to substitute the oblong scittum for the round Argolic shield.
The last people to preserve the circular targe were those of the
Scandinavians who did not settle in the South. As late as 1171
the Danes who fought Strongbow s Normans at Dublin had the
round red shield which their ancestors had carried three hundred
years before. 2

Offensive arms did not alter their shape nearly so much as
defensive during the years 800-1 100. The double-handed axe,
as we have already seen, was introduced by the Danes, and
adopted by the English and in a lesser degree by other races.
The missile taper-axe did not, however, entirely disappear: it is
mentioned in a charter of Cnut s, and appears again in William
of Poictiers description of the battle of Hastings, as hurled by
the English at the oncoming Normans. 3 The sword grew
decidedly longer, and had by 1050 received a rounded point
instead of a sharp one, so that it was wholly a cutting weapon.
The horseman s lance was not yet of any great length ; at
Hastings the Norman knights used.it to cast as well as to thrust.
In some countries the bow was in fairly common use, though it
was always the short-bow, not the formidable six-foot weapon of
the fourteenth century. The Scandinavian peoples, the South-
Welsh, and the races in touch with Byzantium seem to have used
it most. The Danish blood of the Normans accounts for the
large proportion of archers whom they employed at Hastings.
Neither the Germans, the English, nor the French seem to have
taken to it kindly. Abbot Ebolus, the defender of Paris in 886,
is the only notable archer among these peoples who occurs to my

1 Unless it is made of very light stuff, wicker or cane, for example, such as those
of the Turks of the eleventh and twelfth centuries. But the Western shield was a
heavy solid affair of wood and leather.

Giraldus Camb., Exf. Hib. i. 21 : " Clipeis quoque rotundis et rubris, ferro
circulariter munitis."

3 W. P. 201.

i3 THE ART OF WAR IN THE MIDDLE AGES [noo

memory. 1 At the end of the period which we are now discussing,
the crossbow had already been added to the longbow as an
infantry arm. But by uoo it was only just beginning to assert
the ascendency which it was to enjoy in the twelfth and still more
in the thirteenth century.

1 See Abbo, Bell. Put: ii. 405, for his lucky shot at a Danish pilot. He was also a
good marksman with a balista (ibid. i. no).

CHAPTER VI

SIEGECRAFT A.D. SdO-I IOO

THERE is on the whole a greater continuity in the history
of siegecraft and siege-machines through the whole
Middle Ages down to the invention of gunpowder, than in the
history of any other province of the military art. When we read
the account of Witiges siege of Rome in 537, of the beleaguering
of Gundovald Ballomer in Comminges in 585, of Wamba s
capture of Nismes in 673, of the Northman Siegfried s siege of
Paris in 885-886, of the operations of the Crusaders against
Jerusalem in 1099, we are struck with the astonishing similarity
of the proceedings of men so far apart in age and in nationality.
To take, for example, the first and the last of these five sieges
we find Witiges and Godfrey of Bouillon relying on exactly the
same methods. W T hen the rude expedients of striving to fill the
town-ditch and swarm up the wall on ladders do not avail, the
besieger in each case falls back on two main resources. The
one is that of breaching the fortress with rams, the other that of
clearing the ramparts of their defenders not only by the missiles
discharged by engines placed close at the foot of the wall, but
by the concentrated volleys of men posted in high movable
towers brought up close to the fortifications, so as to overtop them
and to allow them to be swept by arrows from above. If
Witiges failed and Godfrey succeeded, it was mainly because the
Goth never succeeded in getting his towers right up to the walls,
while the Crusader gradually filled the ditch with debris, and
finally pushed his engines into such close contact with the town
that he could throw his bridges down on the rampart, and cross
them at the head of his knights.

All through the Dark Ages there were two great weapons of
offence in siegecraft, the ram and the bore. The former worked
by gradually battering to pieces the point of the wall on which

131

132 THE ART OF WAR IN THE MIDDLE AGES [500

it was set to play : it shook the whole structure till the mortar
gave way and the ramparts crumbled into a breach. The bore
(terebrus), on the other hand, consisted of a massive pole furnished
with a sharp iron point: it was intended to work piecemeal,
picking out or breaking up the individual stones till it produced
a round hole in the tower or the front of curtain which it assailed.

The ram was often a vast bulk, the largest tree of the
countryside, fitted with an enormous head, and requiring forty or
sixty men to swing it. It was slung by ropes or chains from two
solid perpendicular beams, drawn back by the workers as far as
the chains allowed, and then released to dash itself against the
wall. As the besiegers could not hope to live close under the
ramparts, beneath the deadly hail of stones and shafts which the
defenders poured upon them, it was necessary to cover the ram
with a shelter. Accordingly it was provided with a large pent
house which usually ran upon wheels or rollers, though sometimes
it seems to have been carried forward by main force, and set
down again and again as the ram moved on. The sides of the
penthouse were usually made of hides, or of hurdles covered with
hides, to make the structure as light and portable as could be
managed. The roof, however, had to be more solid, as the
defenders were wont to pour on it liquid combustibles, such as
pitch or boiling oil. If the assailant made it very strong, with
solid beams covered by raw hides, tiles, or earth to keep off the
burning liquid, the only resource of the defenders was to drop
heavy stones upon it or to destroy it by a sortie.

But even if the penthouse could not be harmed, the ram
itself might be disabled : a favourite device descending, like the
engine to which it was opposed, from Roman times was to let
fall on its head, while it struck the wall, heavy forked beams,
which caught it, held it firm, and prevented it from being drawn
back. We shall see this plan tried in the Viking siege of Paris.
A less effective palliative was to hang from the wall, over the
point on which the ram was playing, thick mattress-like sheets of
sacking filled with straw, or broad and thick beams. The ram
spent its strength on these without progressing in its attempt to
make a breach. Both beams and sacking are heard of in the
great siege of Jerusalem in 1099, and both ultimately proved
more harmful to the besieged than to the assailant.

It is confusing to find the ram and its penthouse spoken of
in chronicles under names which hide the true nature of their

x&gt;] THE BORE AND THE MINE 133

work. Such are cancer and testudo, both employed as synonyms
for this machine, but both referring properly not to the ram but
to the penthouse, whose rounded upper surface suggested the
comparison to the two creatures.

The bore (teretnis, terebrus, terebro] worked less ostentatiously
and less effectively than the ram ; it required an immense
amount of labour before it could make its hole, and was exposed
no less than the ram to the dangers from above. It had, how
ever, the not inconsiderable advantage of being much lighter and
easier to transport. Moreover, it did not require the enormous
number of men to work it which the ram demanded. It was, of
course, always covered with a penthouse on a smaller scale than
that required for the battering engine, but constructed on the
same lines.

The bore and its shelter appear under many names in the
chronicles. It is sometimes called uiusculus, the mouse, 1 because
its object was to gnaw a round hole in the lower courses of the
lampart. At other times it is called a "cat," because it clawed
its way into walls. A third and very usual name was the
"hog" or "sow" (scrofa, sus*), 2 applied either because of the
resemblance of the round-topped penthouse to a hog s back, or
because it worked with its tusks like a boar. The word vitlpes
is less commonly used for it : "&gt; in this case, as in that of inuscnlus,
the allusion is to the capacity of the engine for making neat round
holes in the surface that it attacked.

Like the later Romans, the men of the Dark Ages sometimes
supplemented the ram and the bore by the device of mines.
Before the invention of gunpowder these were invariably worked
on a single plan. The besieger removed as much earth as he
could carry away from beneath some exposed corner of the
fortifications, and shored up the hole with beams. He then filled
the space between the beams with straw and brushwood, and set
fire to it. When the supports were consumed, the wall crumbled
downwards into the hole, and a breach was produced. Early
writers often call the mine a " furnace," the general effect of the
lighted mine breathing out smoke and sparks from its orifice

1 As in Abbo, i. 99.

- Many readers will remember the joke of Black Agnes of Dunbar when she had
smashed the penthouse and saw its occupants scampering away from beneath: " Behold,
the English sow has farrowed."

3 Albert of Aix uses it in his account of the siege of Niciea, 1097.

134 THE ART OF WAR IN THE MIDDLE AGES [800

reminding them of the oven of everyday life. Mines \vere of
course very effective against places built on soft soil : the diggers
could work undisturbed by the storm of stones and darts from
above, which made the use of the ram and bore so dangerous.
On the other hand, they were entirely useless against fortresses
built on high ground or upon a foundation of solid rock. The
best device which the besieged could employ against mining
was to countermine, and then attack the diggers below ground,
drive them back, and fill up the hole they had excavated. If,
however, the besieger had commenced his mine at a consider
able distance from the wall, and carefully hidden the mouth of
it, so that its exact locality and direction could not be easily
discerned, he had a very fair chance of success. For an early
example of the mine in use on this side of the Channel we may
turn to William the Norman s capture of Exeter in IO67- 1

The ram, the bore, and the mine were the main resources of
the poliorcetic art during our period, but we must mention one
or t\vo engines of lesser importance. Scaling ladders are the
simplest of all the besieger s tools, and the most useless against
a competent defence ; nevertheless a town not unfrequently fell
before an unexpected coup-de-main or a night attack in which
the assailant had no more than ladders to help him. A still
more primitive method was that of heaping up earth fascines or
rubbish of any kind against the lowest part of a hostile wall, and
endeavouring to clamber in over them. Rome itself fell before
this rude expedient in 896, when King Arnulf bade his Germans
lay against the foot of the ramparts their heavy saddles and the
packs of their beasts of burden, and actually succeeded in
entering the Eternal City by scrambling up the heap.-

The movable tower, as distinguished from the mere pent
house destined to shelter a ram, appears at the end and the be
ginning of our period, but seems to be absent during its central
years. Witiges, as we have already had occasion to mention, 3
employed it in vain against Rome in 537. Rut we do not find
it emerging again till the eleventh century. Probably it passed
out of use during the days when fortification was neglected, and
had to be revived when the feudal castle had been produced by
the influence of the Viking and Magyar. It was, at any rate, in

1 See Orderic, iv. p, 510 : " Per plurimos dies obnixe satagit . . . murum subtus
suffodere."

2 Luitprand, Antapodosis, 27. 3 See p. 131.

THE MOVABLE TOWER 135

full employment before the beginning of the Crusades, being
known to William the Conqueror and other competent generals
of his age. 1 The most famous examples of its success, however,
are to be found in the great campaigns of the East, starting from
the capture of Jerusalem in 1099. The tower had a double use:
men posted on its top and armed with missiles overlooked the
defenders of a rampart and shot them down from above, so as
to clear the way for an assault. But it was also quite usual to
fit the tower with a drawbridge, which at a propitious moment
was let down on to the walls and served as a path for a column
of stormers. The tower had all the disadvantages which we have
already seen to be inherent to the penthouse. It was even
heavier to move than that machine: it was equally combustible,
and it was stopped by the slightest ditch, since it could not
advance over uneven ground. Even if the besiegers filled the
ditch with debris, and produced a level at the foot of the walls,
the great weight of the tower often made it sink into the newly-
turned earth, and when once stuck fast it could not be moved
again. We may add that its size and height made it the easiest
of marks for mangonels and petraries. Not unfrequently we
hear of towers battered to pieces by the mere missiles of the
besieged. William of Tyre remarks that those from which the
Crusaders stormed Jerusalem only just served their purpose :
they were so damaged at the moment of the assault that the
chiefs were on the point of ordering them to be rolled back, and
of abandoning the attempt to use them.-

Among the minor tools of early siegecraft the many devices
of twisted hurdle-work deserve mention. These mantlets (plutei,
crates, Jiourdis) were mainly used to shelter the advancing
assailants. They were composed of stakes wattled together
with osiers or other branches, and were generally covered with a
coating of hide. Sometimes a whole storming party would
advance against the walls carrying the mantlets over their heads. s
At other times they were used to protect the smaller siege
engines, which had not penthouses of their own. Sometimes
they were arranged in rows, so as to form a covered way to
enable men to enter the penthouses with safety, or to get close

1 See Guy of Amiens, 1. 699. Ansgar the Staller explains to the Londoners that
"Cernitis oppresses valido certamine muros, Molis et erectae transcendit machina
turres."

- \Villiam of Tyre, viii. 3 See Abbo, i. 220.

i 3 6 THE ART OF WAR IN THE MIDDLE AGES [1000

to the foot of the walls. When set in this fashion, they are often
called by the old Roman names of testndo or vinea. War-bands
who had been long in the field, like the Vikings or the Crusaders,
came to have a great confidence in these light defences, and
grew skilled in the rapid making of them. When the Crusading
armies sat down in front of a Syrian town, we often find the
whole force turning to the construction of a large stock of
mantlets before beginning any serious attack on the place.
They made the leaguer so much less wasteful of life that the
time spent on making them was not thrown away.

The engines for throwing missiles employed in sieges were
the same for assailant and defender. They may be divided
according to the method which they employed for propulsion,
and the missiles which they threw.

There were in the Middle Ages three chief methods of pro
ducing the propelling power required to launch a stone or javelin.
Only two of them, however, seem to have been used in the earlier
centuries with which \ve are now dealing. These were torsion
and tension. The third and later device was the employment
of the counterpoise. By torsion is meant the twisting of ropes
and cords whose sudden release discharged the missile. By ten
sion we mean the mere stretching of the cord, in the same fashion
used to draw the ordinary bow. Both classes were directly bor
rowed from the later Romans. The elaborate details for the
construction of machines given by Vitruvius, and later writers like
Vegetius, Procopius, and Ammianus, explain to us the originals
of most of the machines which were at a later time employed in
the Teutonic kingdoms of W estern Europe. At Constantinople
they continued to be made with the old perfection all through
the Dark Ages : in the lands west of the Adriatic they were small
and rude copies of the Roman originals.

Of the machines working by torsion the best type was the
rnangon, which played the part of heavy siege-artillery. It
consisted of two stout posts joined by a double or quadruple
set of ropes. If a beam is placed between the two sets of ropes,
and drawn back so as to twist them in opposite directions,
a very considerable force is generated. It is utilised either
by making a spoon-shaped hole in the end of the beam
or by attaching a sling to it ; the engineer then places a
missile, e.g. a rock or a ball of lead or stone, in the
spoon or sling, and then suddenly releases the beam.

IOQO] THE MANGON AND BALISTA 137

The ropes, untwisting themselves in a moment, cast the
rock or ball with a high elliptic trajectory. The machine is
difficult to aim, as everything depends on the exact amount of
torsion applied. A wet or dry day, for example, considerably
affects the ropes. But for shooting at large easy marks the
mangon was effective ; it was specially good for what we may
call "bombarding" work, i.e. the casting of missiles at large into
a walled city or an entrenched position. The machine is called
by the name " mangon " as early as 886, where Abbo uses the
word in his account of the siege of Paris. 1 But it is probably
identical with the machine called by the simpler name of sling
(fiuidus, fundibula), which (as we have already had occasion to
mention , \vas in use at a much earlier date. Such no doubt
were the " slings " which were carried by the military train of
Charles the Great. 2 The mangon is the legitimate descendant
of the Roman onager or scorpio described by Ammianus 3 and
Procopius.*

The second class of machines throwing missiles were those
worked by tension, of which we may take the balista as the
type. The balista is a magnified crossbow, as will be seen from
the very clear description of it given by Procopius, when he is
describing the engines used by Belisarius to defend the walls of
Rome in 537. "These machines," he says, "have the general
shape of a bow ; but in the middle there is a hollow piece of
horn loosely fixed to the bow, and lying over a straight iron
stock. When wishing to let fly at the enemy, you pull back the
short strong cord which joins the arms of the bow, and place in
the horn a bolt, four times as thick as an ordinary arrow, but only
half its length. 5 The bolt is not feathered like an arrow, but
furnished with wooden projections exactly reproducing the shape
of the feathers. Men standing on each side of the balista draw
back the cord with little devices \_i.e. winches] ; when they let it
go, the horn rushes forward and discharges the bolt, which strikes
with a force equal to at least two arrows, for it breaks stone and
pierces trees."

In this description Procopius omits only two points : he
neglects to specify what were the " devices " for pulling back

Abbo, i. 364. 2 See p. Si.

3 Ammianus, xix. 7, and xxiii. 4. * Procopius, De Bell. Cott. i. 21.

5 But it threw javelins as well as bolts, and these evidently of great length. See
the passage below from Abbo, about Abbot Ebolus.

138 THE ART OF WAR IN THE MIDDLE AGES [886

the cord, calling them merely W/JMO.! ; we know, however, from
Ammianus, that they were little winches or windlasses which
were wound round and round in order to bring back the cord.
He also omits to state that the cord was usually of twisted gut,
and that when tightened it was caught in grooves or notches
cut in the iron or wooden stock to which the two arms of
the balista were fixed. The machine was then aimed, by
directing the point of the stock at the object which the
engineer wished to strike, and, when good aim had been taken,
the cord was loosed, and sped the missile on its way. 1 Vegetius,
\vho is far shorter on the subject than Procopius, remarks that
the longer the arms of the balista, the harder was the stroke of
the missile which it projected. 2 The bolts thrown by it must
have been formidable things : at the siege of Rome by Witiges,
Procopius saw a mailed Gothic chief, who was struck by a
balista-bolt while mounted in a tree, hang for a long time on
the missile, which, after piercing him, had stuck deep into the
wood. But it cast not only bolts, but long javelins. At the
siege of Paris, Abbo tells us how Abbot Ebolus launched from
a balista a lucky shaft which went through several Danes, who
fell dead pierced by the same missile. The abbot, thinking of
fowls broached on a spit, bade their friends " pick them up
and take them to the kitchen." ;

The balista was, of course, a weapon capable of much more
accurate shooting than the mangon, for its javelins could be
propelled point-blank, and were not hurled with a great curve
like the rocks thrown from the other machine ; it might,
perhaps, be aimed like a modern gun. Hence it was valuable
for accurate shooting at small marks, while the mangon was
more fitted for battering at large ones. The special use of it by
besiegers was to pick off the defenders on the front of wall
which was being attacked. The besieged, on the other hand,
would employ it to play on those of their assailants who were
exposing themselves, especially at men who were out of range of
ordinary arrows or javelins. We shall see that in Abbo s
description of the siege of Paris, the engineers who were

1 Procopius must be read closely with Ammianus here : each supplements the
other. Ammianus does not =peak clearly of the horns of the bow. Procopius omits
the winches and notches.

3 Vegetius, iv. 22: " Quanto prolixiora brachiola habet, tanto spicula longius
Bdittiti"

3 Abbo, i. 110.

1066] THE CROSSBOW 139

directing the construction of the Danish rams \vere slain by a
long shot from a balista while their machines were still very far
from the walls.

The machines of the ninth century, it must also be
remembered, were of very inferior workmanship to their proto
types of the fourth. It is probable that much which was iron
in Ammianus day was wooden in that of Abbo. We doubt
whether the Prankish smiths could make arms for the balista
from iron ; most probably both the arms and the stock were
wooden in the days of the siege of Paris.

There is no doubt that the balista was the parent of the
crossbow of later centuries. The Romans had possessed some
sort of weapon of this kind, but it had so passed out of memory
that the Byzantines of the eleventh century, who preserved so
many other Roman engines, had no knowledge of it. 1 In the
West, on the other hand, it was known and in full use before the
time of the Crusades. William the Norman had " balistantes "
no less than " sagittarii " at Hastings, as Guy of Amiens is
careful to inform us. Nor were the earliest crusaders without
crossbowmen, though they did not at first understand how to
employ them properly against the Turks. The description of
the crusader s arbalest by Anna Comnena is well worth giving,
as it shows an exact correspondence in miniature to the great
balista described by Procopius, with the exception that, owing
to the smallness of the weapon, it can be bent by the force of
the body, and does not need a windlass at the side. " That
hitherto unknown engine, the Tzaggra," she says, " is not a bow
held in the left hand and bent by the right, but can only be
spanned by the bearer stooping and placing both feet against it,
while he strains at the cord with the full force of both arms.
In the middle it has a semicircular groove of the length of a long
arrow, which reaches down to the middje of its stock ; the
missiles, which are of many and various kinds, are placed in
the groove, and propelled through it by the released cord. They
pierce wood and metal easily, and sometimes wholly imbed
themselves in a wall, or any such obstacle, when they have
struck it." : Who was the genius who first conceived the idea of
making a small hand-balista which could be carried and worked
by a single soldier, we are unable to say, nor can we be sure of

1 It was, says Anna Comnena, rols "E\\t]ffi TravreXus ayi/ooi ^ei/ov (x. 8).
- Ibid. x. 8.

140 THE ART OF WAR IN THE MIDDLE AGES [885

the exact date of its appearance probably this revival of the
old Roman manubalista dates back to that darkest of dark ages,
the end of the tenth century.

Of the Trebuchet and other engines working by the use of
heavy counterpoises we shall delay to speak till we reach the
twelfth century. It is by no means clear when they were first
introduced, but apparently they were still unknown in the
centuries (800-1100) with which we are now dealing.

Much confusion is caused to the readers of chronicles
by the fact that the writers of the early centuries of the Middle
Ages use many names for describing the same weapons. All
siege-artillery was either of the type of the mangon, i.e.
relying on torsion, or on that of the balista, i.e. relying on
tension. But they are called indifferently "slings," "catapults,"
" petraries," " machines," " engines," " tormenta," with the most
exasperating vagueness and inaccuracy, by authors who, being
for the most part clergy and not military men, did not fully
understand the principles of the devices which they were
describing. Moreover, confusion is often caused by the fact
that by slight adaptations or changes of shape, the " mangon,"
\vhose proper work was the casting of rocks, might be made
to hurl javelins, and the balista, whose speciality lay in the
accurate propelling of shafts, might be induced to hurl stones.

The best way to gain some idea of the characteristics of a
siege during the Dark Ages, is to investigate the details of a
typical case. Unfortunately, there are very few chroniclers who
give us really good descriptions of such operations. On the
whole, we have a better account of the great siege of Paris in
885-886 than of any other leaguer between the days of Justinian
and the Crusades. Abbo s long poem on the subject is couched
in the vilest Latin, and abounds in the most excruciating false
quantities, but it is very detailed, and on the whole very clear.
As every device of siegecraft known to the Dark Ages was
employed by assailants and defenders, it is well worth while to
give a short sketch of the incidents of those eventful eleven
months.

\Ye have already mentioned that Paris in the autumn of
085 consisted of the old island-city, with the new fortifications
added by Charles the Bald, namely, two bridges crossing the
two branches of the Seine, which encircled old Paris, and
furnished with two bridge-heads. The northern one lay some-

885] THE GREAT SIEGE OF PARIS 141

where about the spot where the tower of the Chatelet afterwards
stood. The southern one must have been somewhere near the
modern Place St. Michel. The bridges were wooden structures,
whose central supports were laid on great piles of stones cast
into the Seine. The bridge-heads were stone towers, but the
northern one was not completed at the moment when the
Danes appeared, having only attained a half or a third of its
destined height. The town was in charge of Odo, count of
the surrounding district, and of its bishop Gozelin. It was
garrisoned by picked men from neighbouring parts of Xeustria
as well as by its own citizens ; among the chief defenders were
Count Ragenar, Robert (afterwards king) the brother of Count
Odo, and Ebolus, Abbot of St. Germain des Pres.

After capturing Pontoise, the Danes appeared in front of
Paris on November 25, 885. They wished to proceed up the
Seine, which was blocked by the two bridges, and sent to offer
terms to Odo and Gozelin, promising to do the city no harm if
their vessels were allowed to pass under the bridges without
molestation. The count and bishop replied in very proper
terms : the Emperor Charles, they said, had placed Paris in
their hands to serve as a bulwark for the rest of Neustria, and
they would be betraying their master if they saved the town
but handed over the bulk of the kingdom to fire and sword.
Siegfried, the Viking commander, returned them the answer
that, as they refused terms, he would take their city by force,
or, if force failed, at least reduce it by famine.

The Vikings at once landed, and made a vigorous attempt
to storm the unfinished northern bridge-head. It failed, but the
defenders were so struck by the weakness of the tower, that
they spent the night in raising it to the full size which it had
been intended to attain, by a hasty superstructure of beams
and planks. Next morning the Danes found it built up to more
than twice the height which it had shown on the previous day.

Seeing that the bridge-head could no longer be stormed, the
besiegers resolved to have recourse to the old Roman device of
sapping its foundation by means of the "bore" or "pick." 1
Preparing mantlets (inusculi), they laid them against the foot
of the tower, and commenced to pull out stone by stone under
cover of these protections. The defenders replied by pouring
boiling oil and burning pitch upon the mantlets, which set them

1 " Qui (Daci) vero cupiunt murum succidere musclis (Abbo, i. 99).

142 THE ART OF WAR IN" THE MIDDLE AGES [885

on fire, and so scorched the men working under cover of them
that they were fain to jump into the river.

The next device of the Danes was an attempt to turn the
use of fire against the defenders. They made a mine under the
tower, probably filling it with combustibles and setting the
mass on fire according to the usual practice. 1 When the mine
fell in, a breach appeared in the base of the bridge-head. The
Vikings tried to enter it, but failed, being overwhelmed by all
sorts of heavy projectiles dropped on them from above. They
then laid combustibles against the door of the tower, to burn it
open ; but a high wind blew the smoke and flame backward, so
that the Ljate stood firm. Meanwhile the defenders brought up
to the tower, and to the parts of the wall of the island-city
which looked out on the tower, many " catapults," i.e. machines
of the balista type casting bolts and darts. These made such
havoc among the Vikings that they finally retired to their ships
with the loss of three hundred men (November 27).

Convinced that the place was not to be taken by a coup-de-
main, the besiegers sent out their bands to ravage the neighbour
hood, and collect a vast store of corn and cattle. They fortified
a camp near the church of St. Germain 1 Auxerrois, with a
foss and stakes, and settled do\vn to beleaguer the city in full
form. Their artificers took some time in preparing three great
rams, each covered by a penthouse of solid wood furnished with
sixteen wheels. The penthouses could hold sixty men apiece
for the working of the rams. When, however, the machines
were wheeled towards the walls, the besieged overwhelmed them
with a hail of missiles, and the two artificers who had designed
them are said to have been both slain by one javelin from a
balista. This disaster to their engineers seems to have
delayed the bringing of the rams into action for some days. 2

January was now far advanced, and the siege had lasted two
months. The Vikings, by no means at the end of their
resources, resolved to try new methods. They prepared a great
number of very heavy mantlets (plutei, or crates, as Abbo calls
them), made of wicker-work, covered with thick coatings of
newly-flayed hides. The main body of the besiegers attempted
to approach the tower under cover of these mantlets, each of
which was capable of concealing from four to six men. Mean
while two smaller parties embarked on their ships and rowed

1 Abbo, i. 133-137- z find. \. 213-215.

886] ASSAULTS ON THE BRIDGE-HEM) 143

up to the bridge, which they tried to climb by mooring their
vessels against its supports.

The assailants on land, having reached the bridge-head under
shelter of the plutei, began to fill up the ditch which surrounded
it. They cast into it clods of earth, boughs, straw, brushwood,
rubbish of all sorts, and ("when they grew excited at their
failure) their store-cattle, and even the bodies of the
unfortunate prisoners whom they had captured in their raids
round the neighbourhood. Meanwhile the besiegers poured a
constant hail of missiles upon them, and slew great numbers ;
but while their attention was thus occupied, the Danes repaired
and brought up the three rams which they had been unable to
utilise at their last assault. The rams were set to batter at
three points of the bridge-head, and began to work considerable
damage among the stones and mortar.

The besieged now put in use a very ancient device, which
had been regularly employed against the ram in Roman times,
letting down large beams with forked teeth, which caught the
ramheads and gripped them, so that they could no longer be
pulled backwards to deliver their stroke. They had also con
structed a number of mangons. 1 The heavy rocks which these
machines cast broke down the thick mantlets whenever they
struck them, and crushed all those sheltered beneath. After
three days of assault, the Danes had lost so heavily that they
withdrew from the walls under cover of the darkness, taking
away such of their mantlets as were intact, but leaving two of
their three rams abandoned and disabled as prizes for the
Franks.

While these unsuccessful attempts were being made upon
the bridge-head, a very exciting struggle had been carried on
around the bridge. The Vikings first tried to take it by assault ;
when beaten off, they had recourse to other measures. Filling
three ships with straw and firewood, they, set them alight, and
towed them up - stream by ropes from the northern bank,
intending to get them under the bridge, and so set it on fire
and break the connection between the island and the bridge
head. Luckily for the besieged, the three vessels all went
aground upon the heaps of stones on which the wooden pillars
of the bridge were laid, and there burned themselves out, or

1 Abbo, 364 : " Machina conficiunt longis lignis geminatis, mangnna quae proprio
vulgi libitu vocitantur."

144 THE ART OF WAR IN THE MIDDLE AGES [886

were sunk by rocks hurled on them from above. The bridge
suffered no harm, and the double assault by land and water
had completely failed. (January 29- February I, 886.)

Four days later, however, an unfortunate accident did for
the Danes what they had been unable to accomplish by their
o\vn hands. Heavy rains swelled the Seine and Marne, and
the furious current which they engendered carried away part of
the northern bridge on the night of the 5th-6th of February.
To add to the misfortune, there were at the moment only twelve
warriors keeping guard in the tower at the bridge-head. Seeing
that the garrison could not be succoured from the city till the
bridge was restored, the Vikings made a sudden and violent
attack on the now isolated tower. They rolled up a cart of
straw against its gate, and set fire to it ; the defenders were too
few to keep them off, while the discharges which the catapults
on the city walls directed against the stormers were distant and
not effective the smoke, we are told, lay about the tower, and
the citizens could not see what was going on. The timber
superstructure of the bridge-head soon caught fire, and the
handful of defenders were forced to evacuate it and take refuge
on the fragment of broken bridge which adhered to the tower.
The Danes offered to spare their lives, professing admiration
for their gallant defence, but no sooner had they laid down their
arms than the treacherous barbarians massacred them one and
all, and flung their bodies into the river. They then proceeded
to throw down the stone foundation of the unfortunate bridge
head. After this success, we should have expected that the
Vikings would have made every effort to get some of their
vessels up-stream through the broken bridge, and then would
have attempted general assaults on the island-city. But they
did nothing of the kind : whether it was that provisions were
running short and required replenishing, or that they were
simply tired of siege operations, they sent the greater part of
their forces off to ravage the land towards the Loire. Their
entrenchments looked so deserted that the defenders thought
that all had departed, and Abbot Ebolus led a sortie to seize
and burn the camp. The vigour with which it was repelled
showed that there were still several thousand Danes lying in
front of the city.

While the siege was thus languishing, Henry Duke of
Saxony appeared on the heights above A lontmartre with rein-

886] HENRY OF SAXONY SLAIN 145

forcements sent by the emperor. The Danes retired into their
camp and took up the defensive, so that the duke was able to
communicate without hindrance with the city, and to throw into
it a large convoy of provisions. The besieged took advantage
of the respite to restore the bridge, and apparently also to
roughly reconstruct the ruined bridge-head. 1 But the siege was
not yet raised : after an unsuccessful attempt to storm the en
trenchments of the Vikings, Henry drew off again, and left Paris
to its own resources (March 886). The besiegers were, however,
sufficiently impressed by the appearance of the relieving force
to transfer their camp from the northern to the southern bank of
the Seine, so as to put the river between themselves and any
force coming from the north. Siegfried, the most important of
the Danish leaders, recommended the raising of the siege, as it
was known that the Emperor Charles was calling together a
large army to carry out the enterprise in which Duke Henry
had failed. The majority refused, however, to follow his advice,
and resolved instead to deliver a general assault on the city
before the emperor should arrive. Early in April they simul
taneously attacked the two bridge-heads, the bridges, and the
island itself, running their boats aground on the narrow shore
at the foot of its fortifications and trying to scale them. They
had no success at any point, and a few days later Siegfried,
followed by a considerable part of the host, took his departure,
after receiving sixty pounds of silver quite a moderate sum
from the besieged, who hoped that he would induce the whole
horde to follow him.

The majority, however, headed by a chief named Sinric,
utterly refused to abandon the siege. They were perhaps
encouraged to persist by the fact that pestilence had broken
out in the crowded city with the return of the warm weather ;
among its victims was Bishop Gozelin, one of the two chief
heroes of the defence. The siege, however, had assumed a
very curious aspect : the Danes being now mainly concentrated
on the south bank of the river, though they kept a corps of
observation opposite the northern bridge-head, the besieged
could communicate in an intermittent way with the open

1 This is nowhere stated by Abbo, but how could Henry have sent the flocks and
herds into Paris without a bridge? Moreover, the "tower, i.e. the bridge-head,
begins again to appear early in the second book of Abbo s poem, and is securely held
by the besieged.

1 46 THE ART OF WAR IN THE MIDDLE AGES [886

country. Sometimes they sent out boats up-stream, sometimes
they ran the blockade in and out of the northern bridge-head.
The fighting died down into skirmishes for egress and ingress
by this route, till in May the Danes tried, without warning or
ostentatious preparations, an attempt at escalade. Three
hundred of them suddenly ran their boats ashore at the foot of
the island wal 1, and swarmed up it with ladders. The head of
the column got within the enceinte at the first rush, but the
defenders, running together from all quarters, were able to
repulse them before their main body could come to their aid.

In the end of June or the beginning of July, Count Odo,
who had slipped out of the city to communicate with the emperor
and gather reinforcements, appeared on Alontmartre with three
thousand men. The Danish corps of observation on the northern
bank tried to intercept him, but he cut his way through them
and re-entered the city with his followers. Soon after the van
of the great army which the emperor had collected from all the
Austrasian and West-German lands came in sight of the city.
Charles the Fat tarried behind at Rheims himself, but sent
Henry of Saxony forward to clear the way to Paris. Less
fortunate than at his first attempt to communicate with the city,
the duke fell into a hidden ditch which the Danes had con
structed in front of their camp, and there perished. The
emperor still holding aloof, the Danes tried one more general
assault. This time they brought up many more engines than
before, and tried to clear the walls of their defenders by incessant
volleys of stones, javelins, and leaden balls cast from a thousand
machines. They then attempted at one and the same moment
to escalade the bridges and the island-wall from boats, and to
burn the northern bridge-head by heaping combustibles against
it. At every point they were repelled after a desperate struggle,
though it seemed at one instant as if they would destroy the
rough wooden fort which had been reconstructed to cover the
northern bridge. At the last moment, when the garrison had
actually been driven out by the smoke, the fire suddenly died
down before the enemy had entered, and the Franks were able
to rush bacK and reoccupy the much-disputed work.

This assault was the last crisis of the siege, which ended very
shortly after, not by the driving away of the Danes by the large
army which had now been gathered against them, but by a dis
graceful treaty. Charles the Fat, instead of attempting to storm

886] THE SIEGE OF PARIS RAISED 147

the Danish camp, offered the Vikings a ransom of seven hundred
pounds of silver and free permission to pass over into Burgundy,
if they would but raise the siege. They accepted his pusillani
mous offer, received the money, and passed southward till they
came to Sens, which city they beleaguered in vain for six
months. Thus, Paris was not relieved by the valour of its
garrison, but by the cowardice of its monarch. Nevertheless,
its gallant defence had no small effect on men s minds. Seeing
the Danes foiled, and the city untaken after so many desperate
attacks, all the people of Neustria were encouraged to resist for
the future.

Two main points of interest strike the reader who studies the
details of this great leaguer. The first is the extraordinary skill
in the technique of siegecraft which the Danes had attained after
sixty years of raiding in the empire. The second, contrasting
strangely with the first, is the fact that they completely failed
to appreciate the necessity of cutting off the communication of
the city with the outer world. A much shorter term of months
must have reduced Paris to surrender if only the assailants had
properly taken in hand the isolation of the fortress.

Turning to the first point, we are amazed to see most of the
tools and engines known to Vegetius and Procopius in full
employment among the wild seamen of the North. The ram,
the machines for hurling missiles, the penthouse, the plutei and
crates, the mine, the use of fire, are all thoroughly understood by
the Vikings. Obviously, the} must have picked them up from
their enemies during the interminable series of raids and sieges
which had begun in the later years of Lewis the Pious. The
Franks are by 885 not a whit more skilled in poliorcetics than
their adversaries.

On the other hand, the general strategy of the siege was
wholly faulty. No proper arrangement of a permanent
covering force was made : any considerable body of relieving
troops which presented itself was able to force its way into
Paris. The succours under Henry of Saxony and Count Odo
had to face some severe fighting in order to pass through the
Danish blockade, but they were neither compelled to engage
in a pitched battle, nor to force lines of circumvallation. During
the first half of the siege the Vikings seem to have neglected
the southern bank of the Seine ; during the second half when
they had moved their camp to St. Germain des Pres the

148 THE ART OF WAR IN THE MIDDLE AGES [886

northern bank seems to have been left \vithout sufficient guard.
All through the long months of the leaguer the defenders were
able to communicate with their kinsmen of the outer world, either
by boat on the Upper Seine or by running the gauntlet between
the outposts of the besiegers. Reinforcements and food were
thrown into the fortress again and again. The Danes should
have blocked the river above the city by a boom, or built boats
upon it to keep the water-way closed. They should also have
been prepared to risk a general engagement with any relieving
force, and not have sent mere detachments against it. Their
position, to compare modern things with ancient, much reminds
us of that of the English and French before Sebastopol in 1855.
A siege may drag on for ever if the assailant only attacks one
side of a fortress, and leaves the other in free communication
with the open country. The Vikings had the additional
difficulty of having only very narrow fronts the two bridge
heads to attack. The river prevented them from making any
really dangerous assaults on the island, whose walls they could
not properly breach by siege-artillery placed on the mainland.
Hence we may fairly say that only famine could have been
relied upon as a certain method of reducing the place, and that
the new methods of fortification introduced by Charles the
Bald thoroughly justified themselves, and proved impregnable
against any mere attack by main force, even when it used the
best siegecraft of the day.

CHAPTER VII

THE LAST STRUGGLES OF INFANTRY THE BATTLES OF
HASTINGS (A.D. ZO66) AND DVRKHACHIUM (A.I&gt;. lOSl)

AS the last great example of the endeavour to use the old in
fantry tactics of the Teutonic races against the now fully-
developed cavalry of feudalism, we have to describe the battle
of Hastings, a field which has been fought over by modern
critics almost as fiercely as by the armies of Harold and
William.

About the political and military antecedents of the engage
ment we have no occasion to speak. Suffice it to say that on
September 25, 1066, Harold Godwineson had defeated and slain
Harold Hardrada and Tostig at Stamford Bridge, after a bloody
struggle, whose details are entirely lost, though \vc know that
both hosts had fought the matter out to the end in the old
fashion of Dane and Englishman, all meeting face to face
on foot, and " hewing at each other across the war-linden," till
the invaders were well-nigh annihilated. On September 28,
William of Normandy and his army came ashore at Pevensey,
unhindered by the English fleet, which after long waiting had
finally been driven from the Channel by want of provisions, 1
and had sailed back to London three weeks before. The
Normans began at once to waste the land, and, since the king
and the field army were far away in the north, they met with
little resistance. Only at Rcmney, as we are told, did the lands-
folk stand to their arms and beat off the raiders. 2

Meanwhile, the news of William s landing was rapidly
brought to Harold at York, and reached him if we may trust
Henry of Huntingdon at the very moment when he was
celebrating by a banquet his great victory over the Northmen. 3

1 Florence of Worcester. A.S. Chronicle, 1066. - William of 1 oictiers, 139.
3 But, according to Guy of Amiens (156), he was returning with his trophies
from the north when the messenger met him.

149

150 THE ART OF WAR IN THE MIDDLE AGES [ro66

The king received the message on October i or October 2 :
he immediately hurried southward to London with all the speed
that he could make. The victorious army of Stamford Bridge
was with him, and the Northumbrian levies of Eadwine and
Morcar were directed to follow as fast as they were able.
Harold reached London on the jth or 8th of October, and
stayed there a few days to gather in the fyrd of the neighbouring
shires of the South Midlands. On the nth he marched forth
from the city to face Duke William, though his army was still
incomplete. The slack or treacherous earls of the North had
not yet brought up their contingents, and the men of the
Western shires had not been granted time enough to reach the
mustering place. P.ut Harold s heart had been stirred by the
reports of the cruel ravaging of Kent and Sussex by the
Normans, 1 and he was resolved to put his cause to the arbitra
ment of battle as quickly as possible, though the delay of a few
days would perhaps have doubled his army. 2 A rapid march of
two days brought him to the outskirts of the Andredsweald,
within touch of the district on which William had for the last
fortnight been exercising his cruelty.

Harold took up his position at the point where the road
from London to Hastings first leaves the woods, and comes
forth into the open land of the coast. The chosen ground
was the lonely hill above the marshy bottom of Senlac, 3 a
place far from all human habitations, and marked to the
chronicler only by " the hoar apple tree " on its ridge, just
as Ashdown had been marked two centuries before by its
aged thorn.*

The Senlac position consists of a hill about a mile long and
150 yards broad, joined to the main bulk of the Wealden Hills
by a sort of narrow isthmus with steep descents on either side.

1 William of Poictiers, 201.

Or even tripled it, says Florence of Worcester. The A.S. Chronicle is more
vague, but to the same effect.

3 This name is only used by Orderic Vitalis (501 A), among the many chroniclers
who describe the battle. But it is substantiated by local documents of a later
date ; and since Santlachc occurs as the name of a tract of abbey land in the Chronicle
of the Foundation of Battle Abbey, there is no reason to doubt that it was the genuine
name of the valley. It is easy to understand that the majority of the writers who
narrate the fight had not heard of this local name, and followed the popular voice in
naming the fight after the town of Hastings, which, though eight miles away, was the
nearest place of importance.

4 Asser, p. 23.

io66] THE POSITION OF SENLAC 151

The road from London to Hastings crosses the isthmus, bisects
the hill at its highest point, and then sinks down into the
valley, to climb again the opposite ridge of Telham Hill. The
latter is considerably the higher of the two, reaching 441 feet
above the sea level, while Harold s hill is but 260 at its
summit. The English hill has a fairly gentle slope towards
the south, the side which looked towards the enemy, but on the
north the fall on either side of the isthmus is so steep as to be
almost precipitous. The summit of the position, where it is
crossed by the road, is the highest point. Here it was that
King Harold fixed his two banners, the Dragon of Wessex, and
his own standard of the Fighting Man.

The position was very probably one that had served before
for some army of an older century, for we learn from the best
authorities that there lay about it, especially on its rear, ancient
banks and ditches, 1 in some places scarped to a precipitous slope.
Perhaps it may have been the camp of some part of Alfred s
army in 893-894, when, posted in the east end of the Andreds-
weald, between the Danish fleet which had come ashore at
Lymne and the other host which had camped at Middleton, he
endeavoured from his central position to restrain their ravages
in Kent and Sussex. 2 No place indeed could have been more
suited for a force observing newly-landed foes. It covers the
only road from London which then pierced the Andredsweald,
and was so close to its edge that the defenders could seek
shelter in the impenetrable woods if they wished, lo avoid a
battle. 3

The hill above the Senlac bottom, therefore, being the obvious

"Crescentes herbae anliquitm aggerein tegebant " (Orderic Vitalis, 501 D).
" Praerupti vnllis et frequentum fossarum opportunitas " (William of Poictiers, 203 D).
Of these one a^ger was in the rear of the English position, and was used against the
Normans in the last moments of the battle. But there was a fowea magna in front of
the English line, according to Henry of Huntingdon, 763 c : " Fugientes [Xormanii]
ad quandam magnam foveam dolose tectam devenerunt, ubi multus eorum numerus
oppressus est." Thisfct ea was well down the slope, and outside the English position.
I think these "frequent ditches" and "ancient earthworks" in an uninhabited place
can mean nothing but the remains of an ancient camp. Both Mr. Archer and Mr.
George pointed this out to me when we were talking over the details of the battle.

2 A.S. Chronicle, 893-894, copied in Ethel weard, Florence, and Henry of Hunting
don. Alfred " encamped as near to them as he might for the wood-fastnesses and the
water-fastnesses, so that he might reach either army, in case it should seek to ravage
the open land."

3 " Mons silvne vicinus crat, vicinaque vallis, et non cultus ager asperitate sui "
(Guy of Amiens, 365, 366).

1 52 THE ART OF WAR IN THE MIDDLE AGES [1066

position to take for an army whose tactics compelled it to stand
upon the defensive, Harold determined to offer battle there.
We need not believe the authorities who tell us that the king
had been thinking of delivering a night attack upon the
Normans, if he should chance to find them scattered abroad on
their plundering, or keeping an inefficient look-out. 1 It was
most unlikely that he should dream of groping in the dark
through eight miles of rolling downs, to assault a camp whose
position and arrangements must have been unknown to him.
His army had marched hard from London, had apparently only
reached Senlac at nightfall, and must have been tired out. More
over, Harold knew William s capacities as a general, and could
not have thought it likely that he would be caught unprepared.
It must have seemed to him a much more possible event that
the Norman might refuse to attack the strong Senlac position,
and offer battle in the open and nearer the sea. It was
probably in anticipation of some such chance that Harold
ordered his fleet, which had run back into the mouth of the
Thames in very poor order some four weeks back, to refit
itself and sail round the North Foreland, to threaten the Norman
vessels now drawn ashore under the cover of a wooden castle
at Hastings. 2 He can scarcely have thought it likely that
William would retire over seas on the news of his approach, 3 so
the bringing up of the fleet must have been intended either to
cut off the Norman retreat in the event of a great English
victory on land, or to so molest the invader s stranded vessels
that he would be forced to return to the shore in order to
defend them.

Harold took one further precaution. He had served a
campaign in the Norman ranks a few years before, on the
occasion of his involuntary visit to Ponthieu, and he thoroughly
knew the Norman tactics. The danger to the English lay, first,
in the rush of the duke s horse ; secondly, in the long shooting of
the duke s archers. To guard against both these perils Harold

1 William of Foictiers, 201 B.

2 Ibid. 201 A. I cannot see why Professor Freeman and other writers have doubted
this statement. The fleet, or some large part of it, must still have been at London in
October.

3 Vet a good authority, William of Poictiers, says that Robert Fitz-Wymara, a
Norman resident in England, sent messengers to the duke to warn him that Harold
was approaching with such a large army that he had better put to sea and return
to Normandy. William, we are told, scnmnilry declined the advice.

1066] HAROLD STOCKADES HIS POSITION 153

directed his men to build a fence of crossed woodwork 1 from the
brushwood in the forest which lay close at their backs. It was
an old Danish device, used two hundred years before, to

1 Here we come to the most vexed point in this most interesting fight. Neither
William of Poictiers, Guy of Amiens. Baldric, Henry of Huntingdon, nor any of the
early minor sources of information, distinctly mention this wicker-work. Can we
trust Wace, who gives an elaborate description of it before the battle and alludes to it
during the course of the engagement ? Wace is an authority of later date than the
others, and wrote some ninety years after the battle. He occasionally makes strange
errors in his narrative (though the earlier writers, it must be remembered, do the same;
and sometimes is guilty of anachronisms, though on the whole he comparatively seldom
clashes with earlier writers in such a way as to show himself absolutely wrong.

Is it likely that Wace, in describing the struggle which was to his audience the battle
par excellence of the last age, would make such a strange error as to describe what was
really a fight on an open hill as an attack on a position which had been entrenched,
even though the entrenching was but slight? On the whole, Wace s general narrative
is so fairly consistent with the earlier sources, that I cannot believe that he made this
great blunder. If it had been the common and ordinary thing for armies to stockade
themselves about 1150-60, though an uncommon thing in 1066, we might have thought
that Wace was committing a mere anachronism. But it was no more unusual at &gt; n&lt;
date than at the other, and I do not see what should have induced him to bring
the wattled barrier into his narrative, unless it existed in the tale of the fight as
it had been told him by his father and others who had talked w ith the victors of the
great battle. In our own day popular tradition is a comparatively feeble thing : the
written word has everywhere supplanted the oral tale : but in the twelfth century the
people s memory was a far more trustworthy thing. I cannot think that Wace, writing
for the grandchildren of the men of Senlac, would have ventured to change so entirely
the character of the engagement.

We can trace in the Roman de Ron the author s knowledge of several of our exist
ing authorities, e.g. of William of Poictiers, Guy of Amiens, and William of Jumieges.
If he had thought the existence of the breastworks inconsistent with their tale, it seems
unlikely that he would have inserted it, for he does not give us the idea of an irrespon
sible inventor of facts, but of one who conscientiously uses the data that come to him,
though he may have to adapt them a little to make them assume a fitting place in his
story.

I conclude, then, that Wace, possibly from some lost chronicle or poem, possibly
only from popular oral tradition, knew of the existence of Harold s wattled breast
works, and mentioned them. His words must imply some kind of wooden barricade
"Fait orent devant els escuz
De fenestres et d altres fuz
Devant els les orent levez
Comme cleies joinz e serrez
Fait en orent devant closture,

Ni laissierent mile jointure." (7815-20.)

The reading fenestres is, as Professor York Powell pointed out to me, possibly a
scribe s blunder for frt sties tresses : if so, the passage translates thus

" They made in front of them shields of wattled ash and of other woods, they
raised them in front of themselves like hurdles joined and set close : they left no open
ing in them but made them into an enclosure." The other main passages referring to
the breastwork are, " d escuz et d ais s avironoent, and " ne doterenl pel ne fosse," in
line 8499.

i54 THE ART OF WAR IN THE MIDDLE AGES [1066

stockade a force against an overwhelming onset of cavalry by
means of breastworks and a ditch. The material for the wattled
hurdles, crates or plutei, as the writers of the time called them,
was plentiful and close at hand. They were intended perhaps
more as a cover against missiles than as a solid protection against
the horsemen, for they can have been but hastily-constructed
things, put together in a few hours by wearied men. In all pro
bability they were no more than four feet high. They were set
along a slight ditch, perhaps a remnant of the ancient camp
which probably lay on the Senlac hill, perhaps a work of the
army itself. The ditch, and the mound made of the earth cast
up from it and crowned by the breastworks, constituted no im
pregnable fortress, but a slight earthwork, not wholly impassable
to horsemen. We must not think either of a six-foot trench or
of massive palisading behind it : such a structure would have
required far more time and exertion than the English had to
spare. The entrenchment, according to Wace, was triple, i.e.
consisted of a centre and two wings, with intervals left between
them. 1

Close behind the breastwork, and ready to hurl javelins or
strike with hand-weapons across it, was ranged the English host
in one great solid mass. 2 Although the Northumbrian and
West - country levies were still missing, the army must have
numbered many thousands, for the fyrd of south and central
England was present in full force, and stirred to great wrath by
the ravages of the Normans. It is impossible to guess at the
strength of the host : the figures of the chroniclers, which
sometimes swell up to hundreds of thousands, are wholly useless.
As the position was about a mile long, and the space required
by a single warrior swinging his axe or hurling his javelin was
some three feet, the front rank must have been some seventeen
hundred or two thousand strong. The hill was completely
covered by the English, whose spear-shafts appeared to the
Normans like a wood, 3 so that they cannot have been a mere
thin line : if they were ten or twelve deep, the total must have

1 " Closre le fist de boen fosse , de treis parz laissa treis entries " (/ . de R. 12106).
Fosst is the technical word for a military trench, and quite distinct horn fosse (feminine),
a ditch.

3 Cuneus, which here, as in most other places, means merely a body in deep order
or column as opposed to line, and does not in the least imply a wedge-shaped array.

3 Guy of Amiens : " Spissum nemus Angligenarum," 421, "silvaque densa prius
rarior efficitur," 428.

1066] THE ARRAY OF THE ENGLISH 155

reached some twenty - five thousand men. Of these the
smaller part must have been composed of the fully - armed
warriors, the king s housecarles, the thegnhood, and the heavily-
equipped soldiery, of whom one had to be furnished by every
five hides of land. The rudely-armed levies of the fyrd must
have constituted the great bulk of the army : they bore, as the
Baycux Tapestry shows, the most miscellaneous arms swords,
javelins, clubs, axes, a few bows, and probably even rude instru
ments of husbandry turned to warlike uses. Their only defensive
armour was the round or kite-shaped shield : body and head
were clothed only in the tunic and cap of everyday wear.

In their battle array we know that the well-armed house-
carles perhaps two thousand or three thousand strong were
grouped in the centre around the king and the royal standard.
The fyrd, divided no doubt according to its shires, was ranged
on either flank. Presumably the thegns and other fully-armed
men formed its front ranks, while the peasantry stood behind and
backed them up, though at first only able to hurl their weapons
at the advancing foe over the heads of their more fully-equipped
fellows.

We must now turn to the Normans. Duke William had
undertaken his expedition not as the mere feudal head of the
barons of Normandy, but rather as the managing director of a
great joint-stock company for the conquest of England, in which
not only his own subjects, but hundreds of adventurers, poor and
rich, from all parts of Western Europe had taken shares. At
the assembly of Lillebonne the Norman baronage had refused in
their corporate capacity to undertake the vindication of their
duke s claims on England. But all, or nearly all, of them had
consented to serve under him as volunteers, bringing not merely
their usual feudal contingent, but as many men as they coulcl
get together. In return they were to receive the spoils of the
island kingdom if the enterprise went well. On similar terms
William had accepted offers of help from all quarters : knights
and sergeants flocked in, ready, " some for land and some for
pence," to back his claim. It seems that, though the native
Normans were the core of the invading army, yet the strangers
considerably outnumbered them on the muster-rolls. Great
nobles like Eustace Count of Boulogne, the Breton Count Alan
Fergant, 1 and Haimer of Thouars were ready to risk their lives

1 Cousin of the reigning sovereign in Brittany.

156 THE ART OF WAR IN 1 THE MIDDLE AGES [1066

and resources on the chance of an ample profit. French,
Bretons, Flemings, Angevins, knights from the more distant
regions of Aquitaine and Lotharingia, even if Guy of Amiens
speaks truly stray righting men from among the Norman
n juerors of Naples and Sicily, joined the host. 1

Many months had been spent in the building of a fleet at
the mouth of the Dive. Its numbers, exaggerated to absurd
figures by many chroniclers, may probably have reached the
.six hundred and ninety-six vessels given to the duke by the
most moderate estimate. 3 What was the total of the warriors
which it carried is as uncertain as its own numbers. If any
analogies may be drawn from contemporary hosts, the cavalry
must have formed a very heavy proportion of the whole. In
continental armies the foot-soldiery were so despised that an
experienced general devoted all his attention to increasing the
numbers of his horse. If we guess that there may have been
ten thousand or twelve thousand mounted men, and fifteen
thousand or twenty thousand foot-soldiers, we are going as far
as probability carries us, and must confess that our estimate
is wholly arbitrary. The most modest figure given by the
chroniclers is sixty thousand fighting men ; 3 but, considering
their utter inability to realise the meaning of high numbers, we
are dealing liberally with them if we allow half that estimate.

After landing at Pevensey on September 28, William had
moved to Hastings and built a wooden castle there for the
protection of his fleet. It was then in his power to have marched
on London unopposed, for Harold was only starting on his march
from York. But the duke had resolved to fight near his base, and
spent the fortnight which was at his disposal in the systematic
harrying of Kent and Sussex. When his scouts told him that
Harold was at hand, and had pitched his camp by Senlac hill,
he saw that his purpose was attained : he would be able to fight
at his own chosen moment, and at only a few miles distance from
his ships. At daybreak on the morning of October 14, William

1 Guy of Amiens, line 259.

- Wace, the latest authority, gives the most reasonable figures. If the vessels had
carried as many men as Viking boats, they might have had sixty thousand men on
board ; but the horses must have taken up half the room, if there were, say, ten
thousand of them.

" William of Poictiers, 199, where the duke says that he would "go on even if he
had only ten thousand men as good as the sixty thousand whom he actually
commanded."

THE ARRAY OF THE NORMANS 157

bade his host get in array, and marched over the eight miles of
rolling ground which separate Hastings and Senlac. When
they reached the summit of the hill at Telham, the English
position came in sight, on the opposite hill, not much more than
a mile away.

On seeing the hour of conflict at hand, the duke and his
knights drew on their mail-shirts, which, to avoid fatigue, they
had not yet assumed, and the host was arrayed in battle order.
The form which William had chosen was that of three parallel
corps, each containing infantry and cavalry. The centre was
composed of the native contingents of Normandy ; the left
mainly of Bretons and men from Maine and Anjou ; the right
of French and Flemings. 1 But there seem to have been some
Normans in the flanking divisions also. 2 The duke himself, as
was natural, took command in the centre, the wings fell
respectively to the Breton Count Alan Fergant and to Eustace
of Boulogne : with the latter was associated Roger of Mont
gomery, a great Norman baron.

In each division there were three lines : the first was com
posed of bowmen mixed with arbalesters : the second was
composed of foot-soldiery armed not with missile weapons but
u ith pike and sword. Most of them seem to have worn mail-
shirts, 3 unlike the infantry of the English fyrd. In the rear was
the really important section of the army, the mailed knights.
We may presume that William intended to harass and thin the
English masses with his archery, to seriously attack them with
his heavy infantry, who might perhaps succeed in breaking the
breastworks and engaging the enemy hand to hand ; but
evidently the crushing blow was to be given by the great force
of horsemen who formed the third line of each division.

The Normans deployed on the slopes of Telhajn, and then
began their advance over the rough valley which separated them
from the English position.

When they came within range, the archery opened upon the

1 (iuy of Amiens, 413 :

" Sed laevam Galli, dextram petiere Britanni,

Dux cum Normannis dimicat in medio. "

This means that the French attacked Harold s left, not that they formed William s
left.

- Robert of Beaumont, a Norman, led a thousand men in the right wing (William
of Poictiers, 202 c).

3 " Pedites firmiores et loricali," as William of Poictiers expresses it.

158 THE ART OF WAR IN THE MIDDLE AGES [1066

English, and not without effect ; l at first there must have been
little reply to the showers of arrows, since Harold had but very
few bowmen in his ranks. The breastworks, moreover, can have
given but a partial protection, though they no doubt served
their purpose to some extent. When, however, the Normans
advanced farther up the slope, they were received with a furious
discharge of missiles of every kind, javelins, lances, taper-axes,
and even if William of Poictiers is to be trusted rude weapons
more appropriate to the neolithic age than to the eleventh
century, great stones bound to wooden handles and launched
in the same manner that was used for the casting-axe. 2 The
archers were apparently swept back by the storm of missiles,
but the heavy armed foot pushed up to the front of the English
line and got to hand-to-hand fighting with Harold s men. 3 They
could, however, make not the least impression on the defenders,
and were perhaps already recoiling when William ordered up
his cavalry. 4 The horsemen rode up the slope already strewn
with corpses, and dashed into the fight. Foremost among them
was a minstrel named Taillefer, who galloped forward cheering
on his comrades, and playing like a jongleur with his sword,
which he kept casting into the air and then catching again. He
burst right through the breastwork and into the English line,
where he was slain after cutting down several opponents. 5
Behind him came the whole Xorman knighthood, chanting their
battle-song, and pressing their horses up the slope as hard as
they could ride. The foot-soldiery dropped back through the

1 Baldric, v. 407 : " Spicula torquentur, multi stantes moriuntur."
- " Lignis imposita saxa" (W. P. 201 i&gt;i. They seem to be represented by the
club-like weapons thrown by some of the English in the Bayeux Tapestry.

: " Festinant parmis galeati jungere parmas, erectis hastis hostis uterque furit "
(Guy of Amiens, 383) ; i.e. the heavy-armed men (galeati) met shield to shield
with the English, and both sides fought furiously with their lances.

4 " Interea, dubio dum pendent proelia marte, : Taillefer and the cavalry came
forward.

5 One would have doubted the romantic episode of Taillefer if it did not occur in
such a good authority as Guy of Amiens. Several later writers give details also.
Guy writes (390-400)

" Interea dubio dum pendent proelk marte

Eminet et telis mortis amara lues
Histrio, cor audax nimium quern nobilitavit,

Agmina praecedens innumerosa ducis
Hortatur Gallos verbis et territat Anglos

Alte projiciens ludit et ense suo,
Incisor Ferri tniinus cognomine dictus," etc.

1066] HASTINGS : THE FIRST ATTACK 159

intervals between the three divisions, as we may suppose and the
duke s cavalry dashed against the long front of the breastworks,
which in many places they must have swept down by their mere
impetus. 1 Into the English mass, however, they could not break :
there was a fearful crash, and a wild interchange of blows, but
the line did not yield at any point. Nay, more, the assailants
were ere long abashed by the fierce resistance that they met ;
the English axes cut through shield and mail, lopping off limbs
and felling even horses to the ground. 2 Never had the
continental horsemen met such infantry before. After a space
, the Bretons and Angevins of the left wing felt their hearts fail,
and recoiled down the hill in wild disorder, the horsemen
sweeping away the foot-soldiery who had rallied behind them.
All along the line the onset wavered, and the greater part of the
host gave back, 3 though the centre and right did not fly in wild
disorder like the Bretons. A rumour ran along the front that the
duke had fallen, and William had to bare his head and to ride
down the ranks, crying that he lived, and would yet win the day,
before he could check the retreat of his warriors. His brother
Odo aided him to rally the waverers, and the greater part of the
host was soon restored to order.

As it chanced, the rout of the Norman left wing was destined
to bring nothing but profit to William. A great mass of the
shire-levies on the English right, when they saw the Bretons
flying, came pouring after them down the hill. They had
forgotten that their sole chance of victory lay in keeping their
front firm till the whole strength of the assailants should be
exhausted. It was mad to pursue when two-thirds of the hostile
army was intact, and its spirit still unbroken. Seeing the
tumultuous crowd rushing after the flying Bretons, William
wheeled his centre and threw it upon the flank of the pursuers.
Caught in disorder, with their ranks broken and scattered, the
rash peasantry were ridden down in a few moments. Their
light shields, swords, and javelins availed them nothing against
the rush of the Norman horse, and the whole horde, to the

1 For a description of the effect of a furious rush of cavalry on a stout abattis see
Kincaid s description of Waterloo. He and his battalion had erected a breastwork
across the road by La Haye Sainte. It was completely swept a-way by two squadrons
of horse who charged through it. (Kincaid s Rifle Brigade, chap. xx. )

2 " Pugnae instrumenta facile per scuta et alia tegmina viam inveniunt " (\V. P.

33)-

3 " Fere cuncta ducis acies cedit " (William of Poictiers, 133).

160 THE ART OF AVAR IN THE MIDDLE AGES [1066

number of several thousands, were cut to pieces. 1 The great
bulk of the English host, however, had not follo\ved the routed
Bretons, and the duke saw that his day s work was but begun
Forming up his disordered squadrons, he ordered a second
general attack on the line. Then followed an encounter even
more fierce than the first. The breastworks were probably
swept away from end to end, and the ditch filled with debris
and the bodies of men and horses ere it slackened. The fortune
of the Normans was somewhat better in this than in the earlier
struggle : one or t\vo temporary breaches were made in the
English mass, 2 probably in the places where it had been
weakened by the rash onset of the shire-levies an hour before.
Gyrth and Leofwine, Harold s two brothers, fell in the forefront
of the fight, the former by William s own hand, if we may trust
one good contemporary authority. 3 Yet, on the whole, the duke
had got little profit by his assault : the English had suffered
severe loss, but their long line of shields and axes still crowned
the slope, and their cries of " Out ! out ! " and " Holy Cross ! "
still rang forth in undaunted tones.

A sudden inspiration then came to William, suggested by
the disaster which had befallen the English right in the first
conflict. He determined to try the expedient of a feigned flight,
a stratagem not unknown to Bretons and Normans of earlier
ages. By his orders a considerable portion of the assailants 4
suddenly wheeled about and retired in seeming disorder. The
English thought, with more excuse on this occasion than on the
last, that the enemy was indeed routed, and for the second time
a great body of them broke the line and rushed after the retreat
ing squadrons. When they were well on their way down the
slope, William repeated his former procedure. The intact portion
of his host fell upon the flank of the pursuers, while those who
had simulated flight faced about and attacked them in front.
The result was again a foregone conclusion : the disordered men
of the fyrd were hewn to pieces, and few or none of them

" Exardentes Xormanni et circumvenientes, millia aliquot insecuta se
momenta deleverunt, ut ne unus quidem superesset" (William of Poictiers, 133).

- William of Poictiers, 202: " Patuerunt tamen in eos viae incisae per diversas
partes fortissimorum militum ferro." This is before the feigned flight.

3 Guy of Amiens.

4 W e cannot say what portion, or what proportion. The Brevis Rclalio says
that it was a " cuneus Normannorum fere usque ad mille equites," and that they were
" ex altera parte " from the duke. But does this mean the right or (he left wing ?

io66] HASTINGS: THE GREAT STRUGGLE 161

escaped back to their comrades on the height. But the slaughter
in this period of the fight did not fall wholly on the English ;
a part of the Norman troops who had carried out the false
flight suffered some loss by falling into a deep ditch, perhaps
the remains of old entrenchments, perhaps the " rhine " which
drained the Senlac bottom, and were there smothered or trodden
down by the comrades who rode over them. 1 But the loss at
this point must have been insignificant compared with that of
the English.

Harold s host was now much thinned and somewhat shaken,
but, in spite of the disasters which had befallen them, they drew
together their thinned ranks, and continued the fight. The
struggle was still destined to endure for many hours, for the
most daring onsets of the Norman chivalry could not yet burst
into the serried mass around the standards. The bands which
had been cut to pieces were mere shire-levies, and the well-
armed housecarles had refused to break their ranks, and still
formed a solid core for the remainder of the host.

The fourth act of the battle consisted of a series of vigorous
assaults by the duke s horsemen, alternating with volleys of
arrows poured in during the intervals between the charges. The
Saxon mass was subjected to exactly the same trial which befell
the British squares in the battle of Waterloo incessant charges
by a gallant cavalry mixed with a destructive hail of missiles.
Nothing could be more maddening than such an ordeal to the
infantry-soldier, rooted to the spot by the necessities of his
formation. The situation was frightful : the ranks were filled with
wounded men unable to retire to the rear through the dense mass
of their comrades,- unable even to sink to the ground for the
hideous press. The breastworks had been swept away : shields
and mail had been riven : the supply of missile spears had given
out: the English could but stand passive, waiting for the night
or for the utter exhaustion of the enemy. The cavalry onsets
must have been almost a relief compared with the desperate
waiting between the acts, while the arrow-shower kept beating
in on the thinning host. We have indications that, in spite of

1 William 01 Malmesbury says that it was a jossatum (i.e. a trench) which the
English avoided because they knew it. It is perhaps the same as Henry of
Huntingdon s "tovea magna " (762 c).

" Leviter sauciatos non permittit evadere sed comprimendo necat sociorum
f densitas " (William of Poicders, 202 D).

1 62 THE ART OF WAR IN THE MIDDLE AGES [1066

the disasters of the noon, some of the English made yet a third
sally to beat off the archery. 1 Individuals, worked to frenzy by
the weary standing still, seem to have occasionally burst out of
the line to swing axe or sword freely in the open and meet a
certain death. 2 But the mass held firm " a strange manner of
battle," says William of Foictiers, 3 " where the one side works by
constant motion and ceaseless charges, while the other can but
endure passively as it stands fixed to the sod. The Norman
arrow and sword worked on : in the English ranks the only
movement was the dropping of the dead : 4 the living stood
motionless." Desperate as was their plight, the English still held
out till evening; though William himself led charge after charge
against them, and had three horses killed beneath him, they
could not be scattered while their king still survived and their
standards still stood upright. It was finally the arrow rather
than the sword that settled the day : 5 the duke is said to have
bade his archers shoot not point-blank, but with a high tra
jectory, so that the shafts fell all over the English host, and
not merely on its front ranks. 6 One of these chance shafts
struck Harold in the eye and gave him a mortal wound. The
arrow-shower, combined with the news of the king s fall, at last
broke up the English host : after a hundred ineffective charges,
a band of Norman knights burst into the midst of the mass,
hewed Harold to pieces as he lay wounded at the foot of his
banners, and cut down both the Dragon of Wessex and the
Fighting Man.

The remnant of the English were now at last constrained to
give ground : the few thousands it may rather have been the
few hundreds who still clung to the crest of the bloodstained

1 William of Poictiers, 202 D, says that there were two sallies of the English
provoked by IS orman feigned flights, in addition to that which followed the first real
flight of the Bretons. " Bis eo dolo simili eventu usi sunt Normanni."

" This is indicated only by Wace, but is eminently probable in itself.

3 William of Poictiers, 202 D : " Fit deindi insoliti generis pugna," etc.

* " Mortui plus, dum cadunt, quam vivi movere videntur 7 (ibid.}.

5 That the arrow-shower alternated with the charges is obvious. The archers
could not shoot while the knights blocked the way. That the arrow was largely used
is proved by William of Poictiers: " Sagittant et perfodiunt Normanni." This
must have been done alternately and not simultaneously. Wace well describes the
dismay caused by the rain of shafts falling from above (13287).

6 Henry of Huntingdon, 763 c. I see no reason to doubt his statement of Harold s
end, corroborated by Wace and William of Malmesbury. The narrative of the
slaughter and mangling of Harold by the four Norman knights, described by Guy of
Amiens, does not really conflict with it.

io66] HASTINGS: FLIGHT OF THE ENGLISH 163

hill turned their backs to the foe and sought shelter in the
friendly forest in their rear. Some fled on foot through the
trees, some seized the horses of the thegns and housecarles from
the camp and rode off upon them. But even in their retreat
they took some vengeance on the conquerors. The Normans,
following in disorder, swept down the steep slope at the back of
the hill, scarped like a glacis and impassable for horsemen, the
back defence, as we have conjectured, of some ancient camp of
other days. 1 Many of the knights, in the confused evening light,
plunged down this trap, lost their footing, and lay floundering,
man and horse, in the ravine at the bottom. Turning back, the
last of the English swept down on them and cut them to pieces
before resuming their flight. The Normans thought for a
moment that succours had arrived to join the English and, in
deed, Edwin and Morcar s Northern levies were long overdue.
The duke himself had to rally them, and to silence the faint
hearted counsels of Eustace of Boulogne, who bade him draw
back when the victory was won. When the Normans came on
more cautiously, following, no doubt, the line of the isthmus and
not plunging down the slopes, the last of the English melted
away into the forest and disappeared. The hard day s work was
done.

The stationary tactics of the phalanx of axemen had failed
decisively before William s combination of archers and cavalry,
in spite of the fact that the ground had been favourable to the
defensive. The exhibition of desperate courage on the part of
the English had only served to increase the number of the slain.
Of all the chiefs of the army, only Ansgar the Staller and Leofric,
Abbot of Bourne, are recorded to have escaped, and both of
them were dangerously wounded. The king and his brothers,
the stubborn housecarles, and the whole thegnhood of Southern
England had perished on the field. The English loss was never
calculated ; practically it amounted to the entire army. Nor is
it possible to guess that of the Normans : one chronicle gives
twelve thousand, - the figure is possible, but the authority is not
a good or a trustworthy one lor English history. But whatever
was the relative slaughter on the two sides, the lesson of the battle
was unmistakable. The best of infantry, armed only with weapons

William of Poictiers, 203-0: " Frequentes fossae et pr.ieroptus vallis."
" Amiquus agger " (Ord. 501 D).

- Annales Altahenses majores, sui&gt; anno 1066.

164 THE ART OF WAR IN THE MIDDLE AGES [1081

for close fight and destitute of cavalry support, were absolutely
helpless before a capable general who knew how to combine the
horseman and the archer. The knights, if unsupported by the
bowmen, might have surged for ever against the impregnable
breastworks. The archers, unsupported by the knights, could
easily have been driven off the field by a general charge. United
by the skilful hand of William, they were invincible.

Yet once more on a field far away from its native land-
did the weapon of the Anglo-Danes dispute the victory with the
Norman lance and bow. Fifteen years after Harold s defeat,
another body of English axemen some of them may well have
fought at Senlac were advancing against the army of a Norman
prince. They were the Varangian Guard the famous Tii /.ixv-
&lt;p6poi of the Emperor Alexius Comnenus. That prince was
engaged in an attempt to raise the siege of Dyrrhachium, then
invested by Robert Guiscard. The Norman army was already
drawn out in front of its lines while the troops of Alexius were
only slowly arriving on the field. Among the foremost of the
emperor s corps were the Varangians, who rode to the battle-
spot, like the thegns of the West, but sent their horses to the
rear when they drew near the enemy. Alexius had entrusted to
their commander a body of light horse armed with the bow,
bidding him to send these first against the enemy, and only to
charge when the cavalry should have harassed and disturbed
Robert s ranks. But Nampites, the Varangian leader, neglected
these orders. When they approached the Norman line, the
English were carried away by their reckless ardour. Before
the Greek army was fully arrayed, 1 and long before the em
peror had designed to attack, they moved forward in a solid
column against the left wing of the Normans. They fell upon
the division commanded by the Count of Bari, and drove it,
horse and foot, into the sea. But their success disordered their
ranks, and Guiscard was enabled, since the main body of the
Byzantine host had not yet approached, to send fresh forces
against them. A vigorous cavalry charge cut off the greater part
of the English: the remainder collected on a little mound by the
seashore, surmounted by a deserted chapel. Here they were
surrounded by the Normans, and a scene much like that of
Senlac, but on a smaller scale, was enacted. After the horse-

1 They, i.Kavov aweaTritrav OL~ aireiplav o^vrcpov /3e/3a5&lt;K&lt;5r, were a considerable
distance from the rest of the army (Anna Comnena, book iv. 6).

io8i] BATTLE OF DYRRHACHIUM 165

men and the archers had combined to destroy the majority of
the Varangians, the survivors held out obstinately within the
chapel. At last Robert sent for fascines and other woodwork
from his camp, heaped them round the building, and set fire to
the mass. The English sallied out, to be slain one by one,
or perished in the flames. Not a man escaped : the whole corps
suffered destruction as a consequence of their misplaced eager
ness to open the fight. 1 Such was the fate of the last important
attempt made by infantry to face the feudal array of the eleventh
century. We shall find, it is true, some instances in the twelfth
century of cavalry being withstood by dismounted troops. But
these were not true infantry, but knights who had sent their
horses to the rear in a supreme moment of peril, and stood firm
to fight out the battle to the end. Well-nigh three centuries
were to elapse before real foot-soldiery, unaided by the cavalry
arm, made another serious attempt to stand up in the open
against the mailed horseman. 2 The supremacy of the feudal
horseman was finally established.

1 Anna Comnena calls the leader of the Varangians " Nampites. This does not
seem to be a true Teutonic name. A military correspondent suggests to me that it
may possibly represent a nickname " Niemecz " or " Nemety " = the German
bestowed on the English chief by Slavonic fellow-soldiers in the Imperial host.

2 I except, of course, attempts such as that of the Danish Ostmen at the battle of
Dublin to withstand Miles Cogan s men (see p. 403). This was a fight on a small
scale in an obscure corner of Europe ; the Scandinavians neglected the cavalry arm
even later than the English. Other cases could be quoted.

BOOK IV

THE BYZANTINES
A.D. 579-1204

167

CHAPTER I

HISTORICAL DEVELOPMENT OF THE BYZANTINE ARMY

IN our first chapter we traced the military history of the
Eastern Empire down to the reign of Justinian, the last
date at which it is possible to discern any continuity of character
between the ancient Roman army and the troops which had
replaced it. For, less than thirty years after the death of the
conqueror of the Goths and Vandals, a complete reorganisation
was carried out, and the last remnants of the old system dis
appeared. It was replaced by a new one whose nomenclature,
tactical units, and methods were as unlike those of Justinian s
day, as the " Palatine " and " Limitary " numeri of Constantine
were to the legions of Trajan or Augustus Caesar. This new
system was destined to survive the shocks of five hundred years
with small change : for all practical purposes the arrangements
of the end of the sixth century lasted down to the end of the
eleventh. Then only did they vanish, dashed to pieces by the
great disaster of Manzikert (1071) even as the old Roman army
had been dashed to pieces by that of Adrianople seven hundred
years before.

Alike in composition and in organisation, the army which
for those five hundred years held back the Slav and the Saracen
from the frontier of the Eastern Empire differed from the troops
whose traditions it inherited. Yet in one respect at least it
resembled the old Roman host -. it was in its day the most
efficient military body in the world. The men of the lower
empire have received scant justice at the hands of modern
historians : their manifest faults have thrown the stronger
points of their character into the shade, and " Byzantinism " is
accepted as a synonym for effete incapacity both in peace and
in war. Much might be written in general vindication of their
age, but never is it easier to produce a strong defence than
when their military skill and prowess are called in question.

169

1 70 THE ART OF WAR IN THE MIDDLE AGES [580

" The vices of Byzantine armies," says Gibbon, " were in
herent, their victories accidental." 1 So far is this sweeping
condemnation from the truth, that it would be far more correct
to call their defeats accidental, their successes normal. Bad
generalship, insufficient numbers, the unforeseen chances of
war, not the worthlessness of the troops, were the usual sources of
disaster in the campaigns of the Eastern emperors. The causes
of the excellence and efficiency of the Byzantine armies are
not hard to discover. In courage they were equal to their
enemies ; in discipline, organisation, and armament far superior.
Above all, they possessed not only the traditions of Roman
strategy, but a complete system of tactics, carefully elaborated
to suit the requirements of the age.

For centuries war was studied as an art in the East, while
in the West it remained merely a matter of hard fighting. The
young Prankish noble deemed his military education complete
when he could sit his charger firmly and handle lance and shield
with skill. The Byzantine patrician, while no less exercised in
arms, 2 added theory to empiric knowledge, by the study of the
works of Maurice, of Leo, of Nicephorus Phocas, and of other
authors whose books survive in name alone. The results of the
opposite views taken by the two divisions of Europe are what
might have been expected. The men of the West, though
they regarded war as the most important occupation of life,
invariably found themselves at a loss when opposed by an
enemy with whose tactics they were not acquainted. The
generals of the East, on the other hand, made it their boast
that they knew how to face and conquer Slav or Turk, Frank
or Saracen, by employing in each case the tactical means best
adapted to meet their opponents method of warfare.

The Byzantine army of the seventh and following centuries
may be said to owe its peculiar form to a reorganisation which it
went through in the last quarter of the sixth century, some
twenty-five years after the death of Justinian. The details of
that reorganisation are preserved for us in the Strategicon? an
invaluable work, which shows us precisely when and by whom

1 Vol. ii. p. 382.

- Nothing better attests the military spirit of the Eastern aristocracy than their
duels ; cf. the cases of Prusianus and others.

;; A work difficult to procure, for its MSS. are very rare, and its only printed edition
is that of Upsala, dated 1664, a book only to be found in a few public libraries.

580] MAURICE S " STRATEGICON " 171

the change was carried out. East - Roman writers of a later
age often erroneously attributed these alterations to the
celebrated warrior-prince Heraclius, the conqueror of Persia
and the recoverer of the True Cross. In reality, the army
with which Heraclius won his battles had already been re
organised by his worthy but unfortunate predecessor, the
Emperor Maurice, whose troubled reign filled the years 582-
602. It is under his name that the Strategicon appears, and by
his hands that it was compiled. There seems no reason what
ever to doubt the attribution of the Strategicon to the Emperor
Maurice. A careful inspection of the chronological data which
are supplied by the book itself shows that it cannot have been
written before 570 or after 600. The Persian king is alluded
to as the chief enemy of the empire, but he is not represented
as a masterful and oppressive neighbour, as would have been
the case in any book written after the Persian invasions of 605-
6-7-8. On the other hand, the Slavs and Avars are declared to
be the hostile powers on the Danube, no mention being made
of Gepidae or Lombards : therefore the latter tribes must have
already vanished from its banks ; i.e. the writer is dealing with a
period after 568. But from the fact that all the fighting with
Slavs and Antae is supposed to take place in the close neighbour
hood of the Danube, and for the most part not on Roman soil,
but beyond the river, we can fairly decide that the great Slavonic
raids of 581-585, which reached as far as Thessalonica and
Thermopylae, cannot yet have begun. The date 570-580 is
rendered still more likely by the fact that the writer does not
speak with the tone and authority of an emperor. He merely
" wishes to turn to the public use the certain amount of military
experience which has come in his way," l and gives advice rather
than commands. A comparison of the preamble of Maurice s book
with that of Leo s Tactica, a work written by a reigning prince,
shows such a complete difference of tone that we feel sure that
Maurice was as yet only a rising general when he penned his
work. He ascended the throne in 582, so the Strategicon
may fairly be placed a year or two earlier. We should imagine
that the work was written nearer to 580 than to 570, from the
fact that an appreciable space of years seems to separate the
writer from the times of Justinian, who only died in 565. For
he alludes to the army as having been for some time in a con-

1 Slrateguon, \.

172 THE ART OF WAR IN THE MIDDLE AGES [580

dition of decay, and as forgetting its old triumphs ; such a
complaint could hardly have been made when the victories of
Taginae and Casilinum (553-555) were still fresh in men s
memories. The decline began in the last few years of Justinian s
time, when (as Agathias tells us) " the emperor having entered
on the last stage of his life seemed to weary of his labours, and
preferred to create discord among his foes or to mollify them
with gifts, instead of trusting to arms and facing the dangers
of war. So he allowed his troops to decline in numbers
because he did not expect to require their services, and the
ministers who collected his taxes and maintained his armies
were affected with the same indifference." x The decay which
began under Justinian spread deeper during the thirteen years
reign of his successor the haughty and incapable Justin II. (565-
578), and may well have reached the disastrous stage described
by Maurice in the latter days of that prince.

But we may venture to determine even more exactly the date
of the Strategicon. When the Emperor Tiberius Constantinus
succeeded Justin II. (578) he carried out a thorough reorganisa
tion of the army, deputing the care of details to two distinguished
officers, Justinian, the son of Germanus, and Maurice himself.
These two colleagues " set right that which was wrong, and
made orderly that which was chaotic, and, in short, reduced
everything to a state of efficiency." - We may therefore con
clude with reasonable certainty that the Strategicon was then
issued by Maurice to serve as the official handbook of the
reorganised host of the Eastern Empire. In that case it may be
ascribed to the year 579&gt; a date which exactly suits all the
internal indications of time of which we have already spoken.

It would seem that the commissioners made many sweeping
changes in the army, for the troops which Maurice describes
are arrayed and named very differently from those of which
Procopius had drawn a picture thirty years before. It is true
that the mailed horse-archer, the ?.a/3a./.^upioi or xoiraroj, 3 as he is
now called, still remains the great power in war, and the stay
and hope of the Imperial host. But a completely new system
of organisation has been introduced both among cavalry and
infantry. Under Justinian there was no permanent unit in the
army larger than the single regiment, the corps which Procopius

1 Agathias, book v. 14. = Theophanes, sub anno 6074.

3 i.e. lancer, from K QVTCX, the long cavalry spear.

5 8o] THE EAST-ROMAN ARMY IN 580 173

calls a xra&gt;.oyo;, so translating the word numerus, which was
still its official title. Maurice recognises this body, which he
calls an u.f&gt;i9/i.(,c, (i.e. numerus), or more frequently a ray/^a or
Qdvdov, as the base of military organisation ; but he speaks of
the numeri as being formed into larger bodies, brigades and
divisions as we should call them. Six, seven, or eight numeri are
to form a /j.ofpa of two thousand to three thousand men, the
equivalent of a brigade, and three ,ao^&gt;/ are to be united into
a pipot, 1 or division of six thousand or eight thousand men. He
adds that the numerus should be not less than three hundred
or more than four hundred strong, and that moirai should be
formed of an irregular number of numeri, in order that the
enemy should not be able to calculate the exact force opposed
to them by merely counting the number of standards in the line
of battle. Napoleon, it will be remembered, laid down a similar
rule as to his army corps, always taking care that they should
not be of exactly similar force.

A numerus, or " band," or rciy/ta of three to four hundred
strong, is now commanded by an officer called comes or tribtinns.
It is interesting to see how the importance of these names has
shrunk in the fourth century there were only about a dozen
"counts" in the whole empire, and each had ruled a whole
frontier and commanded many cohorts. A tribune in a similar
manner had once been the commander of a whole legion of six
thousand men. Now, however, the two words are used as
homonyms, and applied to a simple colonel. The brigadier in
command of seven or eight bands is now called a (i.oifo.f/ji&gt;c, or, as
a Latin equivalent, a dux (6oD), though the duces of the fourth
century had in precedence and power taken rank below comites.

There is no sign yet in Maurice that the brigading together
of the numeri or "bands" was permanently fixed. He rather
implies that the commander of an army will make it his first
duty to so combine them when war is declared. In this the
army of 580 differs from that of the next century, in which, as
we shall see, a permanent localisation of the regiments and the
constitution of what may be called fixed army corps comes into
being.

The most important change which we trace in the general
organisation of the army by Maurice is the elimination of that
system, somewhat resembling the Teutonic comitatus, which

1 Also called a Spouyyos, a Teutonic name connected with our own word throng.

ry4 THE ART OF WAR IN THE MIDDLE AGES [580

had crept from among the Foederati into the ranks of the
regular Roman army. The loyalty of the soldier was secured
to the emperor rather than to his immediate superior, by
making the appointment of all officers above the rank of
centurion the care of the central government. The commander
of an army or division had thus no longer in his hands the
power and patronage which had made him dangerous. The
men found themselves under the orders of delegates of the
emperor, not of quasi-independent authorities surrounded by
enormous bands of personal retainers. Thus the soldier no
longer regarded himself as the follower of his immediate
commander, but merely as a unit in the military establishment
of the empire.

This reform was rendered all the more easy by the fact that
the barbarian element in the Imperial army was decidedly on
the decrease. The rapid fall in the revenues of the State which
had set in towards the end of Justinian s reign, and which con
tinued to make itself felt more and more under his successors,
had apparently resulted in a great diminution in the number
of Teutonic mercenaries serving in the Roman army. It was
a case, to quote a modern proverb, of "Point a "argent, point de
Suisse" For the foreigner was a more expensive and a more
independent personage than the native soldier, and vanished
when his pay ceased to appear. To the same end contributed
the fact that of the Lombards, Heruli, and Gepidae, the nations
who had formed the majority of Justinian s Foederati, one nation
had removed to other seats, while the others had vanished from
the scene. At last the number of the foreign corps had sunk
to such a low ebb that there was no military danger incurred in
assimilating their organisation to that of the rest of the army.
The barbarian element, as we find it in Maurice s book, appears
under the names of Foederati, Optimati, and Buccellarii. The
former seem to represent the old bands of Teutonic auxiliaries
serving under their own chiefs ; they are apparently spoken of
as invariably consisting of heavy-armed horse. A casual notice in
Theophanes informs us that the Emperor Tiberius Constantinus
found it so hard to keep up their numbers, that he bought all
the Teutonic slaves he could find for sale in and outside the
empire, freed them, and enrolled them as soldiers. The total
number of Foederati was thus brought up to fifteen thousand,
and it was precisely Maurice who was put in command of them,

D] THE EAST-ROMAN AUXILIARY TROOPS 175

vith the title of " Count of the Foederati." The " Optimati "
em to have been the pick of the Foederati : they were chosen
ands of Teutonic volunteers of such personal importance that
each was attended by one or more military retainers, called
Artnati, just as a mediaeval knight was followed by his squires. 1
The Buccellarii, whose name and status has caused much needless
trouble to commentators both in Byzantine and modern times,
were another select portion of the Foederati, who were regarded
as the emperor s personal following they had no doubt done
him homage and regarded themselves as part of his " comitatus " ;
practically they were the barbarian element in the Imperial
Guard, the body which corresponded to the old " Batavian
cohorts " of the first century. The institution, as we have already
had occasion to mention, was of German origin : we find in the
laws of the Visigoths saio and buccellarius used as synonyms for
the oath-bound military dependant whom the Angle or Saxon
would have called a gesith. But it had early been adopted by
the Romans : great captains like Aetius and Belisarius had their
buccellarii just like a Gothic king.

The Teutonic element had thus become comparatively small
in the Imperial army: such as it was, it consisted of the scanty
remains of broken tribes such as the Heruli, Ostrogoths, and
Gepidae, and of stray Lombards who had fled from their king
like the Droctulf of whom we have considerable notice in
Maurice s time. There were also a few " Scythians," i.e. remnants
of the Huns, and Avar refugees who had deserted their lord the
Great Chagan, a habit to which, as we learn from the Stratcgicon,
they were very prone.

Nothing can be more characteristic of the transitional state
of the organisation of the East -Roman army in the day of
Maurice than the extraordinary mixture of Roman, Greek, and
Teutonic words in its terminology. Latin was still the official
language of the empire, and all the drill commands in the
Strategicon are still couched in it ; but we may note that the
Latin is already in a very debased stage, showing signs of
losing or confusing its case endings. 2 Upon the substratum of

1 Procopius mentions a custom which throws light on this. Audoin, the Lombard
king, lent Justinian in 551 for the Gothic war " two thousand noble horsemen and
three thousand five hundred more of meaner rank, who acted as the followers and
attendants of the others " (De Bell. Gott. iv.).

Compare the story of the " Toma fratre " cry, passed down the line of march in
the Slavonic campaign of 587, preserved by Theophanes.

176 THE ART OF WAR IN THE MIDDLE AGES [580

old Roman survivals \ve find a layer of Teutonic words intro
duced by the Foederati of the fourth and fifth centuries such
as bandon for a company of soldiers, drungus (cf. throng) for a
larger body : : burgus, coccoura, betza, and pJmlcus, and similar
words. Finally, we meet with many Greek words, some of
them literal translations of Roman terms for example, apiOfiog
for numerus, some of them borrowed from the old Macedonian
military system by officers of classical tastes, 2 some newly
invented. 8

The whole official language of the empire was, in fact, still in
a state of flux ; the same thing had often two or three names,
one drawn from each tongue. Maurice calls the regiment in
differently /3av3o&gt;&gt;, Tdy/j.a, or api$/j.(,: t and the brigadier ,u.oipa.p%o:,
bfov/ /apici;, or dux. On the whole, however, the Latin holds its
own ; we still find it used for scores of things which in Leo s
Tactica, a work of three hundred years later, have only
Greek names. A very large proportion of the native troops were
still Latin-speaking, all those, in fact, raised in Thrace, Moesia,
and the inner parts of the Balkan peninsula. It was not till
these provinces were overrun by the Slavs, a few years after the
Strategicon was written, that the ancient Roman tongue became
practically a dead language in the Eastern realm. Maurice
seldom or never thinks it worth while to give the Greek
rendering of a Latin technical phrase, while his successor Leo
invariably translates such terms.

One very important military reform which Maurice advocates
deserves especial notice, and serves as a notable sign of the
times. It appears that he was most anxious to break down the
barrier which had been imposed in the fourth century between
the class which paid taxes and that which filled the ranks of
the army. The foreign auxiliaries who had formed such a large
proportion of the army of Justinian were no longer so easily to
be procured, and the tendency to raise more and more native
corps being so strong, Maurice wished to make the empire self-
supporting in military matters, and to recruit the army entirely
from within. " We wish," he writes, " that every young Roman
of free condition should learn the use of the bow, and be con-

1 This curious word is first found in Vegetius, who employs it only for the
confused throngs of a barbarian host.

2 e.g. SL&lt;pa.\ayyta., uira.ffTria Tris, ovpayfa, A6xa"yos.

3 t.g. fioipa and /^pos as technical military expressions.

9 oo] LOCAL MILITARY ORGANISATION 177

stantly provided with that weapon and with two javelins."
Once accustomed to arms, he thought that the provincial would
more easily be induced to enlist. If, however, this was intended
to be the first step towards the introduction of universal military
service, the design was not carried out. Three hundred years
later we find Leo echoing the same words : l " The bow is the
easiest of weapons to make, and one of the most effective. We
therefore wish that those who dwell in castle, countryside, or
town, in short, every one of our subjects, should have a bow
of his own. Or if this be impossible, let every household keep a
bow and forty arrows, and let practice be made with them in
shooting both in the open and in broken ground and in defiles
and woods. For if there come a sudden incursion of enemies
into the bowels of the land, men using archery from rocky
ground or in defiles or in forest paths can do the invader much
harm ; for the enemy dislikes having to keep sending out
detachments to drive them off, and will dread to scatter far
abroad after plunder, so that much territory can thus be kept
unharmed, since the enemy will not desire to be engaging in a
perpetual archery-skirmish."

It is unfortunate that we have no definite information as to
the extent to which this plan for creating a kind of landsturm
apt for guerilla warfare was carried out. That in many districts
of the empire little or nothing came of it we know only too
well. We hear continually of provinces that failed to defend
themselves when they were not furnished with a regular garrison.
On the other hand, there seems to have been some obligation
to provide men for military service incumbent on the themes.
We learn, for example, from a casual reference in Constantine
Porphyrogenitus DC Administrando Imperio that in the time
of his own father-in-law Romanus, " when the emperor wished
to raise Peloponnesian troops for an expedition against the
Lombards, in the days when John the Protospathiarius ruled that
theme, the Peloponnesians offered to give a thousand saddled and
bridled horses and a contribution of one centenar of gold instead
of the levy, and, the offer being accepted, paid it with alacrity.
The Archbishop of Corinth was assessed at four horses, the
Archbishop of Patras at four, the bishops at two horses each,
all protospathiarii resident in the theme at three horses each,
spathiarii at one horse, the richer monasteries at two each, the

1 Tatt. xx. 84.

178 THE ART OF WAR IN THE MIDDLE AGES [900

poorer at a horse for each pair ; while each man liable to serve
personally gave five gold bezants, save very poor men, who were
allowed to give two and a half each ; so the composition was
easily raised." l The unwarlike Greek themes might make such
offers, and pay what the Western Europeans of a later age
would have called a " scutage," but the more martial Asiatic and
Northern themes certainly did not. In many of these border
districts, especially in the later centuries of Byzantine history,
we frequently find the local population turning out in arms. 2
The men of the Armeno - Cappadocian frontier evidently
relied very largely upon themselves for defence. Indeed,
there seem to be traces of a semi-feudal military tenure of
land in the districts in that region, especially in those recon
quered from the Saracen in the tenth and eleventh centuries.
Here military settlers were allowed to establish themselves on
condition of holding their land by the sword. 3 The very curious
and interesting poem of Digenes Akritas? which gives the life of
a border baron on the Cappadocian frontier in the tenth century,
shows us a population of warlike castle-dwelling chiefs sur
rounded by subject villages of their retainers, and waging a
continual war of raids with their Saracen neighbours of Cilicia
and Mesopotamia. They depended on their own strong arms,
and not on the regular garrisons of the themes whose border
they inhabited. In Leo s Tactica we learn from the chapter that
deals with sieges that the government relied on the services of
the citizens whenever a frontier town was besieged, and that they
were distributed to definite posts in the defence. Only if any of
them were suspected of disaffection does the emperor recommend
that they should be refused leave to serve by themselves, and
distributed among the regular companies forming part of the
garrison. The most definite mentions of a generally established
militia in the Asiatic themes are the statements in Cedrenus and
Zonaras that Constantine IX. in 1044 wa s so unwise as to relieve
the provinces of the eastern border of their obligation to keep
up local levies to supplement the Imperial garrison. They had
hitherto been exempted from certain taxes in consideration of

1 Const. Porph. , De Adm. Imfi. cap. 51.

; There seems to have been militia even in the theme of Hellas in 1040, when we
read of the people of Thebes taking arms against the Slav rebels (Cedrenus, 747).

3 The holdings were called KTTJ/WZ.TO. o-TpariwriKa -. they were hereditary, as long as
the military service was paid duly.

4 Edited by Sathas and Legrand, Paris, 1875.

640] DISASTERS OF THE SEVENTH CENTURY 179

this service. Now they were ordered to disband the militia and
in future send money to the central treasury. 1

If universal military service never came into use in the
Eastern Empire, yet Maurice had at least a portion of his
desire fulfilled. From his time onward the rank and file of the
Imperial forces were raised almost entirely within the realm, and
most of the nations contained within its limits, the Greeks alone
excepted, furnished a considerable number of soldiers. The
Armenians, Cappadocians, and Isaurians of Asia Minor, and the
Thracians in Europe, were considered the best material by the
recruiting officer.

The next great landmark in the military history of the
empire after the issue of the Strategical! is the fearful storm
which passed over it in the Persian and Saracen invasions of the
years 604-656. Tiberius Constantinus and Maurice were fairly
lucky in their campaigns, beat back the Persians, and carried
incursions into the land of the Transdanubian Slavs. But
Maurice was unpopular with the army perhaps his cutting down
of the power and importance of the great officers, no less than
his strict discipline and economy, irritated them. He perished
the victim of a mutiny, and the brutal and imbecile Phocas, who
succeeded him, involved the empire in the last and the most
disastrous of its Persian wars. The whole East, from the
Euphrates to the Hellespont, was overrun by King Chosroes,
while at the same time the Slavs and the Chagan of the Avars
moved forward into the European provinces. The empire
seemed on the brink of destruction, and was only saved by the
heroic six years campaign of Heraclius (622-628). But hardly
had the Persian war ended, and the old frontier of the empire been
restored, when the still more fatal Saracen invasion began (633).
In his old age Heraclius saw Egypt and Syria permanently
severed from the empire, and had to reorganise a new military
frontier for his diminished realm along the line of the Taurus.

There was no peace with the Saracen till 659, and for twenty-
six years the whole force of Eastern Rome was concentrated
along its Asiatic border, struggling desperately with the oncom
ing flood of Saracen fanaticism. Either during this long war, or
more probably at its end, when Constans ll. 2 sat on the throne, a
new military organisation of the highest importance was imposed

1 Cedrenus, 790 ; Zonaras, ii. 260.

- Or Constantine IV. , as he should more properly be named.

i8o THE ART OF WAR IN THE MIDDLE AGES [660

on the army and the empire. The old boundaries of the
provinces had been wiped out during the Persian and Saracen
invasions, and all the civil administration was out of gear.
The burden of administration in a time of perpetual martial law
had fallen upon the shoulders of the generals. Recognising
this fact, Constans II. or his son Constantine made a new division
of the lands which still remained unconquered on both sides of the
Bosphorus, using the military organisation of the moment as
the basis of civil as well as of military districts. The forces
serving in Asia Minor at this time consisted (ij of the troops of
the old " diocese " of Oriens, i.e. Syria, now called in Greek
*A&gt;aroAxo/; (2) of the troops of the borders of Mesopotamia and
Armenia, who were generally known as Appwiaxoi ; (3) of the
soldiers of Thrace, brought over into Asia during the stress of
the struggle, and known as Thracesians ; (4) of the surviving
Foederati, now known as the Optimati ; (5) of the native and
foreign halves of the Imperial Guard, known respectively as
the Obsequium and the Buccellarii. During or at the end
of the war these troops were cantoned in various parts of Asia
Minor in separate bodies or army corps, for the long-continued
struggle had rendered permanent their brigading. 1

The new provincial arrangement of the middle of the seventh
century consisted in making these army-corps districts, adopted
first of all only for convenience in the subsistence or mobilisa
tion of the troops, into permanent civil divisions. The com
mander of the army corps became also the governor of the
district and the head of the administration; the "bands" and
" moirai " were permanently fixed down to the posts where they
found themselves. The new geographical divisions and the
army corps both received the appellation of Themes, e eparct.
Their proper names were drawn from the titles of the troops
quartered in each, and were therefore Anatolicon, Armeniacon,
Thracesion, Optimaton, Buccellarion, Obsequium (tyixfa*).
These were the original " themes " of Asia ; shortly afterwards
there was added to them one whose character was similar, but
whose origin was probably naval rather than military ; this was
the Cibyrrhffiot theme, a narrow district reaching along the
southern coast of Asia Minor from Caria to Isauria, and com
prising only the land between the mountains and the sea.

1 I owe the original hint for these paragraphs to Professor Bury s excellent
chapters on the Themes in his History of the Later Roman Empire.

PLATE III.

Probable
Limits of

THETHEMES
nAD 650.

THE "THEMES" OF THE EASTERN UMPIRE IN 650 AND 950

660] THE THEMES 181

Cibyra was a small place, and why it gave its name to the
theme was a constant puzzle to later Byzantine authorities.
Constantine Porphyrogenitus, in his work on the Themes, says
that the name was bestowed in mockery. This is of course
absurd : it is perhaps lawful to conjecture that at the moment
when the new provincial divisions were made, Cibyra was the
chief station of the Imperial fleet which guarded the southern
shore of Asia Minor and the passage into the Aegean. The
district to which it gave its name was purely maritime, and the
isolated coast-plains of which it was composed only com
municated with each other by sea. It was probably, therefore,
the special domain of the fleet, and if there was any regular
cavalry army corps allotted to it, the " bands " told off to protect
it from incursions of the Saracen were probably at the dis
position of the admiral of the Cibyrrhaeot squadron. This, at
least, is made likely by the evidence of a passage in Leo s
Tactica, which bids the general of the Anatolic theme, when
his own theme is attacked by land, to send word to the com
mander of the Cibyrrhaeot fleet, that the latter may land forces
in the rear of the Saracens and devastate Cilicia. 1 By the time
of Constantine Porphyrogenitus, Optimaton, probably on account
of its vicinity to the capital, had no longer any military estab
lishment, and was ruled by a Domesticus, not a general.

Such being the " themes " of Asia, we find that those of
Europe were inferior in number the provinces of the Balkan
peninsula had been so entirely devastated and overrun by the
Slavs in the time of Heraclius, that the whole inland had
passed out of Roman hands. There were probably only three
themes south of the Danube Thrace, Thessalonica, and Hellas ;
to these the other Western possessions of the empire add three
more Sicily, Africa, and the surviving dominions in the empire
in Italy. These last, however, were always called not a theme,
but the Exarchate of Ravenna. Later emperors in the eighth
and ninth centuries subdivided the provinces both of East and
West, till the whole number of themes finally rose to more than
thirty.

Maurice s Strategicon is, of course, too early to give the
themes and the complement of garrison allowed to each. But if

1 arav 8i Sia TIJS 77)5 cK&lt;TTpa.Tfveiv /uAXwffi oi KiXiices flapfiapoi., furivfrit KijS
0i&gt; TrXajfjUoi* (TTpaTriyu), Kal /iera T&y I TT airrbv opofjiwwi fiffTmrr^rui KO.TCL rule
ai ASavftwv -^wplav (Leo, Tactica, cap. xviii. 139).

1 82 THE ART OF WAR IN THE MIDDLE AGES [900

we may follow the Tactica of Leo the Wise, written some two
hundred and fifty years after the theme-system was invented,
the strategos of a theme might usually expect to find himself at
the head of some eight thousand, or ten thousand, or twelve
thousand men, as he is spoken of as commanding two or three
" turmarchs " (or " merarchs," as Maurice would have called them
at an earlier date), the turma running from three thousand up to
five thousand strong. It does not seem, however, to have been
possible for the strategos of a province to mobilise and move
outside of his own district the whole of the troops at his
disposition. Most of the infantry, it seems, were left behind for
garrison duty, and Leo calculates that the average theme should
furnish about four thousand or six thousand picked cavalry, and
not more, when called upon for aid by its neighbours. Nicephorus
Phocas, in his handbook for commanders of frontier themes,
gives five thousand as the total. But this mobilised division was
to consist of troops of the best quality only ; all recruits, weak
and disabled men, and untrained or weakly horses being left
behind at the depots, so that each " turma " would take the field
rather short in numbers, but very compact and fit for hard
service. In one passage, Leo says that the "bands" of the
turma would not muster more than about two hundred and fifty-
six men for this active service.

Just as "theme" meant both the district and its garrison, so
was it with the smaller divisions, each theme being divided up
into districts garrisoned by a " meros " or " turma." So we find
such expressions as that " Cappadocia was a turma of the
Anatolic theme," or that " Cephallenia was a turma of the
theme of Langobardia." Some casual notices in Constantine
Porphyrogenitus s DC Administrando Imperio show us how the
districts were occasionally revised and made into new units.
We read, for example, that, owing to the creation of the new
theme of Charsiana in the days of Constantine s father, Leo, the
author of the Tactica, a large rearrangement was made on
the eastern border. " Charsiana," he says, " was once a turma
of Armeniacon, but when the religious Emperor Leo made it a
theme, then the bands forming the garrisons of Bareta, Balbadon,
Aspona, and Acarcus were transferred from the Buccellarian
theme into the theme of Cappadocia ; and at the same time the
garrisons of Eudocias, St. Agapetus, and Aphrazia were trans
ferred from the Anatolic theme into the Cappadocian theme.

900] ADDITION OF LATER THEMES 183

These seven bands, four originally Buccellarian and three Ana
tolic, made a new Cappadocian turma, called Commata. At the
same time the Buccellarian theme gave up the bands stationed
at Myriocephalon, Hagios Stauros, and Verinopolis to the theme
of Charsiana, these, with other two from the Armeniac theme,
namely the garrisons of Talbia and Connodromus, forming a
new Charsianian turma, called Saniana. The theme of
Cappadocia also gave over to the Charsianian theme the whole
turmarchy of Casa, and the garrisons of Cacsarea and Nyssa." 1
Thus the Charsianian theme was composed of fragments from
the Buccellarian, Armeniac, and Cappadocian army corps, while
Cappadocia was compensated for the large slice taken out of it
by acquiring seven bands from Buccellarion and Anatolicon.
The net result was probably to leave the Buccellarian theme
composed of two turmae instead of three, and Armeniacon and
Anatolicon slightly weakened. All these being now interior
themes, separated from the Saracen frontier by Cappadocia and
Charsiana, they could afford to suffer a reduction of their
garrisons.

By the time that Leo s Tactica and his son Constantine s
work on the governance of the empire were written, there were
some new units of frontier administration in existence which
were smaller than themes, and were purely military in character,
not including any large district, or conferring on their governors
any civil jurisdiction over an extensive region. Such a district
was called a " Clissura," a corruption of the Roman clausura.
It consisted of an important mountain pass with a fortress and
garrison, and was entrusted to a " clissurarch," whose duties one
may compare to those of the " comes littoris Saxonici " of the
fourth century. Some of these " clissuras " comprehended
several passes and a considerable number of garrisons, so that
Constantine doubts in one or two cases whether they ought not
to be raised to the dignity of themes. The command of a
clissura was a splendid opportunity for a young and rising
military officer, as he had an excellent chance of making a name
by repelling the raids of Slav or Saracen, and thus might
ultimately rise to the command of a theme.

1 Constantine Porph. , De Adm. Imp. 50.

CHAPTER II

ARMS AND ORGANISATION OF THE BYZANTINE ARMY

THE extraordinary permanence of all Byzantine institutions
is well illustrated by the fact that the arms and organisa
tion which Maurice sets forth in his Strategicon in 578 are
repeated almost unchanged in the Tactica of his successor Leo
the Wise, written somewhere about the year 900. In particular,
the chapters of Leo which deal with armour, discipline, and the
rules of marching and camping are little more than a rendition
of the similar parts of his predecessor s book. It would not be
fair, however, to the author of the Tactica to let it be supposed
that he was a slavish copyist. Though a mere amateur in military
matters, he reigned for more than twenty years without going out
in person to a single campaign, Leo was an intelligent compiler
and observer. In many chapters of his work the Strategicon is
largely rewritten and brought up to date. The reader is dis
tinctly prepossessed in favour of Leo by the frank and handsome
acknowledgment which he makes of the merits and services of
his general, Nicephorus Phocas, whose successful tactics and
new military devices are cited again and again with admiration.
The best parts of his book are the chapters on organisation,
recruiting, the services of transport and supply, and the methods
required for dealing with the various barbarian neighbours of
the empire. These are the points on which an intelligent war-
minister in the capital could attain full knowledge. The
\veakest chapter, on the other hand, as is perhaps natural,
is that which deals with strategy ; its sections are arranged in
rather a chaotic manner, and form rather a bundle of precepts
than a logical system. Characteristic, too, of the author s want
of aggressive energy, and of the defensive system which he made
his policy, is the lack of direction for campaigns of invasion in
an enemy s country. Leo contemplates raids on hostile soil but

184

9 oo] THE BYZANTINE CAVALRY 185

not permanent conquests ; his main end is the preservation of
his own territory rather than the conquest of his neighbour s.
After reading the book, it is easy to see why the frontiers of the
empire stood still during his reign, though the times were very
favourable for aggression both to East and West. Another
weak point is his neglect to support precept by example ; his
directions would be much the clearer if he would supplement
them by definite historical cases in which they had led to
success. But this he does very rarely ; half a dozen instances
drawn from the campaigns of Phocas, two from the campaign
of Basil I. round Germanicia, a misquoted incident of the
Avaric wars of Justin II. drawn from Maurice s Strategicon?
and a few notes from ancient Greek and Roman history, arc-
all that can be cited. The reader is forced to collect for him
self the data which must have led Leo to arrive at his various
conclusions.

The strength of the East- Roman army in the time of Leo
no less than in the time of Maurice lay in its divisions of heavy
cavalry. The infantry is altogether a subsidiary force, and the
author contemplates whole campaigns taking place without its

2 being brought into action. It seems, in fact, destined rather for
the defence of frontier fortresses and defiles, for the garrisoning
of important centres, and for expeditions on a small scale in
mountainous regions, than for taking the field along with the
horse.

The y.a(3a A/.a.pio: or heavy trooper wore, both in the time of
Maurice and that of Leo, a steel cap surmounted with a small
tuft, and a long mail-shirt reaching from the neck to the thighs. 2

: He was also protected with gauntlets and steel shoes. The horses
of the officers and of the men in the front rank were furnished
with steel frontlets and poitrails ; all had solid well-stuffed
saddles and large iron stirrups an invention which had cropped
up since the fifth century without our being able to say from
whom it had its origin. The trooper was furnished with a light

1 Maurice speaks of a surprise in the campaign near Heraclea which Leo
stupidly misrenders into a campaign of the Emperor Heraclius ! He might have
remembered that Maurice could not possibly have quoted campaigns which took place
twenty years after his death.

Leo concedes that if mail-shirts are not always procurable in sufficient numbers,
it may sometimes be necessary to make shift with scale armour of horn (such as the
ancient Sarmatians wear on Trajan s Column), or even with buff-coats of strong leather
strengthened with thin steel plates.

iS6 THE ART OF WAR IN THE MIDDLE AGES [900

linen surcoat to wear over his armour in hot weather, and with
a large woollen cloak for cold or rainy weather, which was
strapped to his saddle when not in use. His arms were a broad
sword (ffTatf/on), a dagger (vctpa.^piov), a horseman s bow and
quiver, and a long lance (Mvrd.fiov) fitted with a thong towards
its butt, and ornamented with a little bannerole. Some men
seem to have carried an axe at the saddle-bow in addition to
the sword. The tuft of the helmet, the lance-pennon, and the
surcoat were all of a fixed colour for each band, so that the
army may be said to have worn a regular uniform, like its
predecessors of Roman times, and unlike any Western army
that took the field before the sixteenth century.

Byzantine military pictures of a really satisfactory kind, in
which the armour is not affected by the artist having copied
older classical drawings, are not common. It is therefore worth
while to insert here two plates from an eleventh-century MS.,
the Psalter of Theodore of Caesarea, in the British Museum,
where the warriors portrayed are evidently armed exactly as
was the contemporary East-Roman soldier. The MS. being
dated 1066, the soldiery represented in it must wear the same
dress and equipment as the unfortunate army that perished at
Manzikert in 1071. It will be noted that the horsemen do not
in all ways correspond to Leo s description of the cavalry of the
year 900. Their mail-shirts are shorter than we should have
expected, and the tuft on the helms is wanting, unless indeed the
very small ball on the top of the headpiece of the front horseman
in IV. A and of the right-hand foot-soldier in V. C represents it.
These balls, however, look more like small metal knobs. It will
be noted that all the mounted men wear mail-shirts with tunics
below them, and high boots. Their lower arms are unprotected,
but the upper arm of most of them is guarded by the character
istic brassard of narrow metal plates which is seen in most
Byzantine military figures. The horse-archer in IV. B does not
wear this defence, but apparently a sleeveless mail-shirt : the
brassards would have been a hindrance in drawing the bow.
Most of the helms are pointed ; only the horseman in V. C has a
plain round-topped steel cap. The shields are all round and of
moderate size. Several of the cavaliers show their military cloaks
flying behind them. The arms used are lance, bow, axe, and
mace. The last two are to be seen in the group of horsemen
besieging the castle in iv. B. The horses seem to have light

PLATE Jl .

BVZAXTINK ABMOl K A.O. 1066
[Pron A- Pialtrr of Ttoofort o/ Caar ,

o] PERSONNEL OF THE CAVALRY 187

trappings : there is no trace of the frontlets or poitrails of which
Leo speaks in his Tactica.

In some of the provinces, where the use of the bow was not
generally popular, Leo recommends that recruits should be
given two light darts and a shield, until they have been trained
to the practice of archery. This was to be done by giving them
small and weak bows, which were to be progressively changed
for larger and stronger ones as the young soldier grew more
adroit. When skilled in his new weapon, he would have to
abandon the shield, whose employment was incompatible with
the free use of both hands required in shooting.

The Byzantine cavalry-soldier was, like the Roman of the
old republic, a person of some substance and standing. In his
chapter on the raising of troops, Leo writes : " The strategos
must pick from the inhabitants of his theme men who are
neither too young nor too old, but are robust, courageous, and
provided with means, so that, whether they are in garrison or on
an expedition, they may be free from care as to their homes,
having those left behind who may till their fields for them.
And in order that the household may not suffer from the
master being on service, we decree that the farms of soldiers
shall be free from all exactions except the land-tax. For we
are determined that our comrades (for so we call every man
who serves bravely in behalf of our own Imperial authority and
the Holy Roman Empirej shall never be ruined by fiscal
oppression in their absence." 1

The rank and file were recruited parti} from military settlers
holding STpuTiuTixu xT-^/j-ara, but mainly from the ranks of the small
free farmers. Their officers, especially those of the higher ranks,
were drawn from the best families of the Byzantine aristocracy.
"Nothing prevents us," says Leo, "from finding a sufficient supply
of men of wealth and also of courage and high birth to officer
our army. Their nobility makes them respected by the soldiery,
while their wealth enables them to win the greatest popularity
among their troops by the occasional and judicious gift of
small creature-comforts." 2 A true military spirit existed among
the noble families of the Eastern Empire ; houses like those of
Skleros and Phocas, 8 of Bryennius, Kerkuas, and Comnenus, are

1 Tactica, iv. I. - Tactica, iv. 3.

3 The family of Phocas is the most distinguished of the whole Byzantine
aristocracy. It supplied two centuries of notable soldiers, starting from Nicephorus

1 88 THE ART OF WAR IN THE MIDDLE AGES [900

found furnishing generation after generation of officers to the
Imperial army. The patrician left luxury and intrigue behind
him when he passed through the gates of Constantinople, and
became in the field a keen professional soldier.

The whole of the officers and many of the troopers being
men of substance, they brought with them to the campaign a
considerable number of servants and boys some bondsmen,
others free hired attendants. Leo highly approves of this
custom, remarking that when the corps had no camp-followers
many soldiers had to be told off to menial duties and the care
of baggage animals, thus thinning the ranks of the fighting men.
He recommends that the poorer troopers be encouraged to keep
one attendant for every four or five of them, and if possible a
pack-horse to carry such of their baggage as they could not
easily strap to their own saddles. These non-combatants and
baggage animals formed a considerable impediment to the rapid
movement of a cavalry corps, but it was believed that in the
end they justified their existence by keeping the. men in good
physical condition. For when moving in the desert countries
on the frontier, where food for men and fodder for horses were
hard to gather, the troops had largely to depend for subsistence
on their camp-followers, just as an English army in India does
at the present day.

Leo does not give such complete details about the arming
and organisation of the infantry "bands" as about those of the
cavalry. The foot-soldier}- were divided into light and heavy
armed. The former, as in the times of Justinian and Belisarius,
were nearly all archers ; a few provinces where archery was not
practised supplied javelin-men instead. The typical bowman
is described by the Tactica as wearing a tunic reaching to the
knees, and large broad-toed nailed boots. He carried a quiver
with fort}- arrows, and a small round buckler slung at his back,
and an axe at his belt for hand-to-hand fighting. As many as
possible were to be provided with a light mail-shirt : there is no
mention made of helmets, which apparently were not worn by
the archers. Leo only recommends that they shall cut their
hair short, and makes no suggestion about a covering for it.

The heavy-armed foot - soldier, still called scutatiis as in

Phocas, who drove the Saracens from Calabria in 884-887, including the victorious
emperor of the same name, 963-969, and the famous rebel Bardas Phocas, who died
in 989.

PLATE V.

BYZANTINE ARMOUR A.D. 1066
( From (Ae Piolft r of Tluodore of Catfarta]

DO] THE BYZANTINE INFANTRY 189

the days of Justinian, wore a pointed steel helmet with a tuft, a
mail-shirt, and sometimes gauntlets and greaves. He carried a
large round shield, a lance, a sword, and an axe with a cutting
blade at one side and a spike at the other. The shield and
helmet-tuft were of a uniform colour for each band.

In Plate v. will be seen three characteristic figures of foot-
soldiers of the year 1066, taken (like the horsemen described
on p. 1 86) from the Psalter of Theodore of Caesarea. They
wear short mail-shirts above their tunics, and two of the three
also show the characteristic Byzantine brassard on their upper
arms. The third (the left-hand sleeper in V. c) has a short mail
sleeve to his mail-shirt and no brassard. The headdress differs
in each figure : one wears a pointed helm, one a round-topped
helm of classical appearance with a knob at its summit ; the
third has no headpiece at all. It will be noted that thehelmless
man wears mail breeches, unlike any of the other soldiers, horse
or foot, on our plates. One of the two sleepers evidently wears
leather breeches: both have high boots. The spears are long,
the sword short and broad. Two of the shields are circular, in
accordance with Leo s description ; the third is oval, and bears a
device of two coloured bars. Two of the men wear short cloaks
fastened round their necks ; the third is apparently without this
garment.

The infantry, like the cavalry, were followed by a consider
able train of baggage and camp-followers. For every sixteen
men l there was to be provided a cart to carry biscuit, etc., and
a supply of arrows, as well as a second cart carrying a hand-
mill, an axe, a saw, a chopper, a sieve, a mallet, two spades, two
pickaxes, a large wicker basket, a cooking-pot, and other tools
and utensils for camp use. In addition to the car.ts there was
to be a pack-horse, so that when the infantry were forced to
leave the waggon-train behind, for forced marches or other
such purposes, the horses might be able to carry eight or ten
days biscuit with them for immediate use. The two carts and
the pack-horse required at least two camp-followers to drive
them, so that every " band " was followed by a considerable-
body of non-combatants. It will be noted that the contents
of the second cart gave every " century " twenty spades and

1 The "decury," in spite of its name, was sixteen men strong, and not ten. Thus
a century would be about a hundred and sixty men, and three centuries would go to
the " band, making it about four hundred and eighty strong.

THE ART OF WAR IN THE MIDDLE AGES [900

twenty picks for entrenching purposes ; for the Byzantine camp,
like the ancient Roman, was carefully fortified to guard against
surprises.

A corps of engineers (Miteupis (sic) or even Mimopdrof.;)
always marched with the vanguard, and, when the evening
halting-place was settled, marked out with stakes and ropes, not
only the general outline of the camp, but the station of each
corps. When the main body had come up, the carts and pack-
horses, called en masse "Tuldum" (ro rti t.btv), were placed in the
centre of the enclosure, while the infantry bands drew a ditch
and bank along the line of the Mensores ropes, each regiment
doing a fixed amount of the digging. Meanwhile, a thick chain
of pickets was kept far out from the camp, and the men not
engaged in entrenching were kept close to their arms, so that a
surprise was almost impossible, unless the pickets displayed
gross negligence. The carts were often ranged laager-fashion
within the ditch, so as to make a second line of defence. If the
army was not close to the enemy, the majority of the infantry
bands camped outside the fosse, and only the cavalry within it.
But when close to the hostile forces, the whole of the corps both
of horse and foot were placed inside, the infantry taking the
outer posts and the cavalry the inner ones. The object of this
arrangement was, of course, to prevent the cavalry from being
harassed by night attacks, against which they are far more
helpless than infantry, as they have to saddle their horses
before they are of any use.

So perfect was the organisation of the Byzantine army that
it contained not only engineers and military train, but even an
ambulance corps. 1 To each " band " was attached a doctor and
surgeon (Qtfavetvrtti, iurpoi), and six or eight bearers (deputati or
axpifiuHs), whose duty it was to pick up and attend to the
wounded. The deputati were provided with horses furnished
with a sort of side-saddle with two stirrups on the same side,
for carrying the wounded, as well as with a large flask of water. 2
The value attached to the lives of the soldiery is well shown by
the fact that the deputati were entitled to receive a nomisma?
or bezant, for every dangerously wounded man whom they
brought off the field.*

We may now pass on to the tactics of the Byzantine army.

1 Leo, Taetica Const, iv. 6. = Taetica Const, xii. 53.

3 About twelve shillings, or a trifle more. 4 Taetica Const, xii. $ 51.

BYZANTINE TACTICS 191

The first point to observe is that normally the heavy cavalry
form the most important part of the army. Infantry only take
the first place in expeditions among hills and passes where
avalry are obviously useless. In the ordinary operations of war
both arms may frequently be found acting together, but it is
just as usual for cavalry to be working alone, without any
ifantry supports. This partly comes from the inferior reputa
tion of the infantry, but still more from the fact that both in
iurope and in Asia the Byzantines had very frequently to
deal with enemies like the Turks (Magyars), Patzinaks, and
Saracens, whose whole force consisted of horsemen. When
such tribes made an incursion into the empire, the infantry
could not hope to keep up with them. It was quite a normal
thing, when the news of a Turkish or Saracen raid arrived, for
the strategos of the invaded theme to send off all his infantry
to occupy passes in the hills, or fords on great rivers, so as to
block the enemy s retreat ; he would then start with his cavalry
alone to hunt down the raiders. This fact is deducible from
Leo s Tactica, but is still more explicitly stated in the excellent
pamphlet on the defence of the Asiatic border which stands
under the name of the Emperor Nicephorus Phocas. 1

When infantry and cavalry acted together, as would be the
case against an enemy mainly composed of foot-soldiery, e.g.
the Slavs or the Franks, or against a regular invasion of
Saracens as opposed to a mere raid, the usual tactical arrange
ment of the Byzantines was to place the infantry in the centre,
with cavalry on the wings and in reserve behind the line. The
infantry " band " was drawn up sixteen, eight, or occasionally
four deep, with the scutati in the centre and the archers and
javelin -men on the flanks. If expecting to be charged by
cavalry, or to be assailed by a heavy column of hostile foot, the
light troops retired to the rear of the scutati and took refuge
behind them, just as a thousand years later the musketeers of the
sixteenth and seventeenth centuries used to take cover behind
their pikemen. The " band " was taught to fight either in single
or in double line (2/paXayy/a) : to take this latter formation the

1 Niceph. Phoc. iii. I. The strategos is at once, on receiving news of a raid, to
collect his horse and rA jrej urAc oirai ewurwdyew ^jri TTJV ooov Ka6 T\V opuTivovaiv of
iroWjuiot efeXOap. The retreating enemy, heavy with plunder, could be intercepted
easily in the passes by the foot-soldiery, and could be crushed between them and the
pursuing cavalry.

192 THE ART OF WAR IN THE MIDDLE AGES [900

rear ranks (four or eight, according as the band was eight or
sixteen deep in its previous formation) stood still, \vhile the
front ranks moved forward and then halted. 1 In a defensive
battle, the infantry centre of the host was usually drawn up close
to the camp, and protected in the rear by the ditch and waggon-
laager manned by the camp-guard. 2 When, however, the army
had moved out far from its camp to take the offensive, the
infantry were formed in two lines. This formation might be
made either by drawing up a certain number of the battalions
of each brigade (i.e. bands of each driingus] in second line, or
by forming each band into the above-mentioned oifa/.ay/ia.
with an interval of three hundred yards between its front and
its rear half-band. The army was never drawn out in a single
line without reserves ; that order of battle was discouraged by
all Byzantine writers on matters tactical. It was only used as
a last resort when there was a desperate need to produce at all
costs a line equal in length to the enemy s.

Byzantine infantry were accustomed to charge in column
sixteen deep ; the bowmen and javelin - men having retired
behind the scutati, the latter received the command to close up
the ranks (CUXUWSOK), and drew close together, the front rank
locking their shields together, while the second and rear ranks
held their shields aloft over their heads, after the manner of the
ancient Roman testudo. The bowmen in the rear kept up such
a discharge as they best could over their comrades heads. On
getting within a few paces of the enemy, the scutati hurled their
spears, as did the ancient Romans their pi/a, and then fell to
work with sword and axe. It was with these short weapons,
not with the spear, that they were expected to win the day.
Thus a Byzantine infantry division fturma) when charging
would be composed of a number of small columns, with
moderate intervals between them, each composed of from some
two hundred and fifty to four hundred men. 3 The strength ot
the division might be anything between two thousand* and
six thousand strong, and the number of battalions (bands) in it

1 Tattica, vii. 76. = Tactica, vii. 73, 4.

:; An interesting but casual notice in one of the doubtful chapters of the
Tattica (No. xxxiv. ) says that in the Thrakesian theme the bandon was supposed to
be three hundred and twenty strong ; in the theme of Charsiana it was three hundred
and eighty ; in some of the Western themes as much as four hundred.

4 Constantine Porphyrogenitus, quoted above on pp. 182, 183, mentions the
turma of Saniana as only five bands strong.

9 oo] BYZANTINE INFANTRY TACTICS 193

might vary from five to twenty. It was a standing principle
that the divisions should be of unequal sizes, that the enemy
might not be able to calculate the exact force opposed to him
by merely counting the number of divisional standards in the
line. Whether strong or weak, the division advanced in two
lines, of which the first was called the cursores l or fighting line,
the second the defensores - or reserve line. 3

Byzantine infantry would always be covered on the wings by
cavalry when offering battle on any ground where horsemen
could be used. They were not, therefore, obliged to take any
care of their flanks. On the other hand, their rear might possibly
be threatened by hostile cavalry sweeping completely round the
wings of the army. In this case the bands forming the line of
defensores would front to the rear. Or if there was need to keep
watch both before and behind, the individual band would take
the formation we have described above under the name of
bufm t.u. y/iu., and the rear half-band, eight deep, would receive the
order " right about face " (i-zoarpi-^aTi ) and front to the rear,
while the other half-band still kept its original position.

When fighting in hilly country, or in passes and other ground
where cavalry could not be used, the infantry band drew itself
up with the scutati in the middle, and the light troops thrown
forward on either flank, so as to form a kind of crescent-shaped
array. This was especially used for the defence of defiles, when
the heavy-armed men posted themselves across the path, and
the archers and javelin-men endeavoured to line the approaches
to the spot where their comrades were posted, so as to
secure a flanking fire on any enemy endeavouring to force the
road. In forest defiles Leo advises that more reliance should be
placed on the javelin-men, who work best at short ranges : in
rocky defiles, where there was a longer view and a better aim, the
archers would have the preference. 4

Cavalry tactics had been carried to a far greater degree of
elaboration than infantry tactics by the East- Roman army.
The horsemen were, as we have already seen, the preponderant

1 Kovp&lt;rwp(s. oi&vtropes.

3 I infer, though it is nowhere explicitly stated, that the reserve line in a division
or brigade was formed, as a rule, from complete bands, and not from the rear half-
bands of the battalions in the front line, because Leo says, in Jactica, vii. 45, that
a brigadier or divisional general is to tell off his bands into defensores and cursors, and
to be careful that each band gets a fair share of each sort of work.

4 Leo, Tactica, ix. 78.

&lt;3

i 9 4 THE ART OF WAR IN THE MIDDLE AGES [900

arm, and they often in a mixed force equalled or even exceeded
the foot in numbers. 1 When they were in a large majority, Leo
advises that the whole front line should be formed from them,
and the infantry placed in the rear in reserve. This was the
order adopted by Nicephorus Phocas in his celebrated victory in
front of the walls of Tarsus (A.D. 965). 2 Often infantry were
altogether wanting, and the whole army was composed of
cavalry. Both Leo s Tactica and the napa5po,u,r, HO~/. SIJ.W ascribed
to Nicephorus Phocas are very full of directions for this case, and
the most elaborate instructions for the marshalling of a cavalry
host are given by both. They are well worth recording, as
representing the most characteristic development of the Byzantine
art of war.

The main principle of the battle-tactics of the Imperial
cavalry was that the whole force must be divided into (i) a
fighting line, (2) a supporting line, (3) a small reserve behind
the second line, (4) detachments out on the wings, destined
some to turn the enemy s flank, some to protect that of their
own main body. As to the numerical proportions of these four
parts of the host, the front line should average somewhat more
than a third say three-eighths of the whole ; the supporting
line about a third of the whole ; 3 the reserve about a tenth ;
the flanking detachments about a fifth.

As an illustration of such an array Leo gives a practical
example. He supposes that the strategos of an eastern
frontier theme has pursued a large Saracen raiding force and
brought it to bay. Having left behind all weak men and horses,
all recruits, and certain necessary detachments, the general has
with him two weak divisions (titnnae), each composed of two
brigades (dnmgi) of five regiments (banda) each. The individual
band has been weeded down to two hundred or two hundred
and fifty men, but contains only picked troopers. The total of
the host is only about four thousand six hundred men, though

1 John Zimisces in his expedition against the Russians had thirteen thousand horse
and fifteen thousand foot (Leo Diaconus, viii. 4).

2 The centre was formed of Trawicr/poi unrbrcu, behind whom were the infantry,
the wings of cavalry also (Leo Diaconus, iv. 3).

3 TA rpirov irotrbv, says Leo, when laying down his general rule in Const, xii. 29.
But in the practical example which he gives, the supporting line is only thirteen
hundred strong out of four thousand six hundred. In a small army, apparently,
the flanking detachments would be a trifle stronger in proportion than in a laige
one.

THE ARRAY OF A CAVALRY FORCE 195

ic two turmae, if present with their whole effective, would
nount to at least six thousand five hundred or seven thousand.

1. The front rank is to be composed of three bodies each
five hundred strong, i.e. each composed of two bands of two
liundred and fifty men. It is drawn up with the smallest
possible intervals between the bands, so as to present a
practically continuous front. The senior divisional general
(turmarck), the second in command of the whole force, leads the
line : 1 he takes his post in its centre, surrounded by his standard-
bearer, orderlies, and trumpeters. Each of the six bands sends
out to skirmish one-third of its men, all archers : the remainder
are halted till the time for charging comes.

2. The second line is composed of four bands, i.e. one
thousand men. They are not drawn up in continuous line, as
are their comrades in the front, but in four separate bodies a
bowshot apart. The three intervals between the bands are to
serve for the passage of the fighting line to the rear in case it
should be routed. The commander-in-chief, with a bodyguard
of a hundred men and the great battle-flag, takes his position
in the middle of the second line, but is not fixed there ; he may
transfer himself to any point where he is needed. 2 To give an
appearance of solidity to the line, a few horsemen three hundred
are enough are drawn up two deep in each of the intervals
between the four bands 3 (G G G in plan).

3. Behind the second line, not to its rear, but on its
flanks, 4 are placed two bands of two hundred and fifty men each
as a last reserve.

4. On the flank of the fighting line, thrown somewhat
forward, (D) to the right is placed a weak band (two hundred
men), destined to endeavour to turn the enemy s left flank when
the clash of battle comes ; they are called the wepxipdarui.
On the left (E) lies a corresponding band of two hundred men,
who are charged with the duty of preventing any such attempt
on the part of the enemy ; they are called the TrXay/opii/.axEr. It
will be noted that armies are expected to make the outflanking
movement from their own right : this comes from the wish to
get in on the enemy s left side, against his weaponless left arm.

1 xii. 77. - xii. 90. 3 xviii. 147.

4 xii. 30. This point, noted in the general directions for drawing up a cavalry
array, is not repeated in Canst, xviii., where the above-named plan for ordering four
thousand men is to be found.

196 THE ART OF WAR IN THE MIDDLE AGES [900

5. Far out from the whole line of battle, to right and left,
are to be placed two bodies, each of two small bands (or four
hundred men) called the evibpoi or liers-in-wait (F F). They
are intended to make a long circular march, hide themselves in
woods and hollows or behind hills, and come in suddenly and
unexpectedly upon the flank or rear of the enemy.

Thus the whole battle order works out into

Frontline . . . . .6 bands" = 1500 men.

Second line . . . . . 4 bands " = 1000 men.

Third line . . . . . 2 bands " = 500 men.

"firepKepdffTat . . . . . i band = 200 men.

II\a-yio&lt;t&gt;v\a.Kes I band" = 2OO men.

Ec&lt;?3/&gt;oi . . . . . . 4 bands " = 800 men.

General s escort i band" = 100 men.

To rill the intervals in the second line ij bands "= 300 men.

20 " bands " = 4600 men.

I presume that the first turma or division supplied the ten
bands of the front line and the wb^i, while the second turma
furnished the second and third lines and the other small detach
ments. But this is not definitely stated.

The bands are drawn up eight or ten deep, though Leo grants
that this formation is too heavy. With an ideally perfect body
of men he thinks that four deep would be the best forma
tion ; 1 but for practical work with an ordinary regiment he
regards eight deep as the least that a general should allow, and
ten deep as the safest and most solid array.

This order of battle is deserving of all praise. It provides
for that succession of shocks which is the key to victory in a
cavalry combat : as many as five different attacks would be
made on the enemy before all the impetus of the Byzantine force
had been exhausted. The intervals of the second line give full
opportunity for the first line to retire when beaten, without
causing disorder behind. Finally, the charge of the reserve and
the detached troops would be made, not on the enemy s centre,
which would be covered by the second line even if the first were
broken, but on his flank, his most exposed and vulnerable point.
Modern experience has led to the adoption of very similar
arrangements in our own day.

The only point which seems of doubtful value is the arrange
ment of the small detached bodies of men two deep in the

1 xii. 40.

PLATE VI

A BYZANTINE CAVALRY FORCE or TWO TURMAE

IN LINE OF BATTLE

Enemy s Line of Battle

NEIGHBOURHOOD

OF AlMTIOCH

TO illustrate rtie
Campaign oF 1097-38.

900] CRITICISM OF BYZANTINE TACTICS 197

intervals of the second line. Leo intends them to deceive the
enemy s eye, and to give an impression of continuity and
solidity to the array. 1 If the front line is broken, they are to
retire, leave the intervals open, and draw up in the rear of the
second line, and between the two bands of the third line. There
they are to serve as a rallying point for the broken troops from
the front, who will form up on each side of them. But in
practical work this retiring to the rear at the moment when the
remnants of the shattered first line were tumbling in upon them
would be a very hazardous experiment. There would be a great
chance that, instead of the fugitives rallying upon the support,
the support would be carried away by the fugitives, and all go
off the field in disorder. Only the steadiest and coolest troops
could be trusted to carry out the manoeuvre. Still, as we shall
see from the battles which we are about to describe as instances
of Byzantine cavalry tactics, the troops of the empire were quite
capable of rallying and returning to the charge.

CHAPTER III

STRATEGY AND TACTICS OF THE BYZANTINE ARMY

WE have already had occasion to observe that the chapters
on organisation, arms, and tactics in the military writers
of the East-Roman Empire are always more satisfactory than
those which deal with strategy. Gibbon, with his usual sweep
ing contempt, remarks that such works seem to aim at teaching
how to avoid defeat rather than how to achieve victor}-. There
is a certain amount of truth in the sneer, for the main lines of
Byzantine strategy during the greater part of the history of the
empire are somewhat one-sided. They are almost entirely
defensive in their scope, and pay little attention to the offensive.
In this respect they do but reflect the general condition and
needs of those who used them. From 600 to 800, and again
from 1050 to 1453, the rulers of Constantinople were making a
strenuous fight for existence, and not aiming at offensive opera
tions beyond their own borders. Between Heraclius Persian
campaigns (622-28) and Nicephorus Phocas conquest of Cilicia
(964), the East- Roman generals never were able to contemplate
an invasion on a large scale into hostile territory. The tactical
offensive they might often take, but it was always with the
object of preserving or recovering their own lands, not with that
of annexing those of their neighbours. Summed up shortly, the
whole military history of these centuries consists in a struggle
to preserve Asia Minor from the Saracen, the Balkan peninsula
from Slav, Bulgarian, and Turk, 1 and the Italian themes from
Lombard and Frank. Of these struggles the first was far the most
engrossing : when once the pressure was taken off the Eastern

1 i.e. Avar, Magyar, Palzinak : perhaps one ought to include the Bulgarian also
under this name. At least the Byzantine writers often place him in that category.
See Leo, Tactica, xviii.

198

9 oo] BYZANTINE STRATEGY MAINLY DEFENSIVE 199

frontier, owing to the incipient decay of the Abbasside Caliphate
in the middle of the ninth century, the East- Romans suddenly
appear once more as a conquering and aggressive power. Cilicia,
North Syria, and Armenia are overrun, the Balkan peninsula is
reconquered up to the Danube, a vigorous attempt is made to win
back Sicily. Our military text-books, however, belong almost
entirely to the defensive period : 1 an edition of Leo s Tactica
brought up to date by Basil II. would be invaluable ; but
unfortunately it does not exist.

The fact that the main aim of Byzantine strategy was to
protect the empire rather than to attack its enemies accounts
for its main limitations. But it does not explain the whole of
the differences between the military feeling of East and West
during the early Middle Ages. Of the spirit of chivalry there
was not a spark in the Byzantine, though there was a great
deal of professional pride, and a not inconsiderable infusion
of religious enthusiasm. The East-Roman officer was proud of
his courage, strength, and skill ; he looked upon himself as
charged with the high task of saving Christendom from pagan
and Saracen, and of preserving the old civilisation of the empire
from the barbarian. But he was equally remote from the
haughty contempt for sleights and tricks which had inspired
the ancient Romans, and from the chivalrous ideals which grew
to be at once the strength and the weakness of the Teutonic
West. 2 Courage was considered at Constantinople as one of
the requisites necessary for obtaining success, not as the sole
and paramount virtue of the warrior. The generals of the East
considered a campaign brought to a successful issue without a
great battle as the cheapest and most satisfactory consummation
in war. 3 They considered it absurd to expend stores, money,
and the valuable lives of veteran soldiers in achieving by force
an end that could equally well be obtained by skill. They
would have felt far higher admiration for such feats as Marl-

1 The- llapaopofjLT] nti\tnov, which bears Nicephorus Phocas name, is written by an
officer who had seen the rise of the new offensive tactics, but does not know whither they
are about to lead. He is one of the old school, though privileged to see the turning
of the tide, and proud to recognise the changed conditions of war in his own old age.

2 I suppose that Baduila the Ostrogoth, that loyal Christian knight, merciful to
foes, true to his word, guided in all things by his conscience and his love of justice, is
the first chivalrous figure in modern history. Yet he failed before Byzantine fraud and
courage combined.

3 Leo, Const, xx. g 12.

200 THE ART OF WAR IN THE MIDDLE AGES [900

borough s forcing of the lines of Brabant in 1706, with the loss
of only sixty men, or for Wellington s manoeuvring the French
out of the Douro valley in 1813, than for bloody fights of the
type of Malplaquet or Talavera. They had no respect for the
warlike ardour which makes men eager to plunge into the fray,
it was to them rather the characteristic of the brainless
barbarian, and an attribute fatal to anyone who made any
pretensions to generalship. They had a strong predilection
for stratagems, ambushes, and simulated retreats. For the
officer who fought without having first secured all the advan
tages for his o\vn side they had the greatest contempt. Nor
must we blame them too much for such views : fighting with
comparatively small and highly-trained armies against enormous
hordes of fanatical Saracens or savage Turks and Slavs, they
were bound to make skill supply the want of numbers. A
succession of emperors or generalissimos of the headstrong, reck
less type that was common in the West would have wrecked
the Eastern realm in fifty years. The two men who more than
any others brought ruin on the empire were two gallant swash
bucklers who never could keep out of a fray, whether it were
opportune or inopportune, Romanus Diogenes, the vanquished
of Manzikert and the loser of all Asia Minor, and Manuel
Comnenus, the crowned knight-errant who wasted the last
resources of his realm on unnecessary victories in Hungary and
Armenia.

But it must be confessed that there often appear in Byzantine
military history incidents that show something more than a mere
contempt for rashness and blundering courage. Modern generals
have not always been straightforward and honourable in their
observance of the customs of war. 1 But they do not as a rule
proceed to glory in their ingenuity and commit it to paper as a
precedent for the future. There is ample evidence, not only from
the records of chroniclers, but from the chapters of Leo s Tactica,
that the East- Romans felt no proper sense of shame for some of
their over- ingenious stratagems in war. It is with a kind of intel
lectual pride in his own cleverness that the Imperial author
advises that if negotiations with a neighbour are going on, and

1 Napoleon certainly committed breaches of the laws of war as odious as any of
which the Byzantines ever were guilty. None of them ever surpassed those master
pieces of treachery and lying, the seizure of the Vienna bridges in 1805 under pretence
&gt;*f an armistice, and the occupation of the Spanish fortresses in iSoS.

FRAUD AND FORCE 201

it is intended to break them off, the softest words should be re
served to the last day but one, and then a sudden expedition be
launched against the enemy, who has been lulled into a belief in
the certainty of peace. He is quite ready to send bribes into the
hostile camp. He recommends two ancient tricks that were
already a thousand years old in his own day. The first is that
of addressing treasonable letters to officers in the enemy s camp,
and contriving that they shall fall into the hands of the com-
mander-in-chief, in order that he may be made suspicious of his
lieutenants. The second is that of letting intelligence ooze out
to the effect that some important person in the hostile country is
secretly friendly, and adding plausibility to the rumour by spar
ing his houses and estates when raids are going on. 1 Leo is not
above raising the spirits of his own soldiers before a battle by
inventing and publishing accounts of imaginary victories in
another corner of the seat of war. A trick too wel! known in
later as well as in Byzantine times is that of sending parle-
mentaires to the enemy on some trivial excuse, without any real
object except that of spying out the numbers and intentions of
the hostile forces. These and similar things have been tried in
modern times, but they are not now recommended in official
guides to the art of war published under Imperial sanction. 2 It
is only fair to say that the same chapter which contains most of
them (Const, xx.) is full of excellent matter, to the effect that no
plighted treaty or armistice must be broken, no ambassador or
parlementaire harmed, no female captive mishandled, no slaughter
of non-combatants allowed, no cruel or ignominious terms im
posed on a brave enemy. A few precepts of the rather futile
immorality of those which we have instanced above must not be
allowed to blind us to the real merits of the strategical system
into which they have been inserted. The art of war as it was
understood at Constantinople in the tenth century was the only
system of real merit existing in the world ; no Western nation
could have afforded such a training to its officers till the sixteenth,
or we may even say the seventeenth century. If some of its

1 A device as old as the Punic Wars ! Hannibal tried it against Fabius.

2 The most " Byzantine" piece of writing that I can recall in a modern campaign is
Kutusoff s cynical despatch to the Emperor of Russia, avowing the trick which he had
played off on Murat a few days before Austerlitz. " In alleging the conclusion of an
armistice, he wrote, " I had nothing in view but to gain time, and thereby obtain the
means of removing to a distance from the enemy, and so saving my corps." Many
men might have carried out the fraud : few would have openly boasted of it.

202 THE ART OF WAR IN THE MIDDLE AGES [900

precepts leaned a little too much towards the side of fraud, it
may be pleaded that at any rate its methods were more humane
than those prevailing in any other part of the world at the
time.

But we are at present engaged in investigating the efficacy
and not the morality of the military customs of the Byzantines.
A survey of the main lines of the strategy and tactics of their
armies must be our next task.

The generals of the new Rome made it their boast that they
knew how to face and conquer the various enemies of the empire
in East and West, by employing against each the tactical means
best adapted to meet their opponents method of warfare. The
Strategicon of Maurice gives an account of the Persian, Avar,
and Lombard and the methods to be used against them : Leo,
three hundred years later, substitutes for these earlier foes the
Frank and Saracen, the Slav and Turk. His chapter dealing
with them (Const, xviii.) is more detailed and more interesting than
the corresponding passage in his predecessor s work, and deserves
reproduction, alike as showing the diversity of the tasks set be
fore a Byzantine general, and the practical manner in which they
were taken in hand. They serve, indeed, as a key to the whole
art of war as it was understood at Constantinople.

" The Franks and Lombards," says Leo, "are bold and daring
to excess, though the latter are no longer all that they once were:
they regard the smallest movement to the rear as a disgrace, and
they will fight whenever you offer them battle. When their
knights are hard put to it in a cavalry fight, they will turn
their horses loose, dismount, and stand back to back against very
superior numbers rather than fly. So formidable is the charge
of the Frankish chivalry with their broadsword, lance, and shield,
that it is best to decline a pitched battle with them till you have
put all the chances on your own side. You should take advantage
of their indiscipline and disorder ; whether fighting on foot or on
horseback, they charge in dense, unwieldy masses, which cannot
manoeuvre, because they have neither organisation nor drill.
Tribes and families stand together, or the sworn war-bands of
chiefs, but there is nothing to compare to our own orderly
division into battalions and brigades. Hence they readily fall
into confusion if suddenly attacked in flank and rear a thing easy
to accomplish, as they are utterly careless and neglect the use of
pickets and vedettes and the proper surveying of the countryside.

9oo] TACTICS USED AGAINST THE FRANKS 203

They encamp, too, confusedly and without fortifying themselves,
so that they can be easily cut up by a night attack. Nothing
succeeds better against them than a feigned flight, which draws
them into an ambush; for they follow hastily, and invariably fall
into the snare. But perhaps the best tactics of all are to protract
the campaign, and lead them into hills and desolate tracts, for
they take no care about their commissariat, and when their stores
run low their vigour melts away. They are impatient of hunger
and thirst, and after a few days of privation desert their
standards and steal away home as best they can. For they are
destitute of all respect for their commanders, one noble thinks
himself as good as another, and they will deliberately disobey
orders when they grow discontented. Nor are their chiefs
above the temptation of taking bribes ; a moderate sum of
money will frustrate one of their expeditions. On the whole,
therefore, it is easier and less costly to wear out a Prankish
army by skirmishes, protracted operations in desolate districts,
and the cutting off of its supplies, than to attempt to destroy
it at a single blow."

The chapters (xviii. 80-101) of which these directions are an
abstract have two points of interest. They present us with a
picture of a Western army of the ninth or tenth century, the
exact period of the development of feudal cavalry, drawn by the
critical hand of an enemy. They also show the characteristic
strength and weakness of Byzantine military science. On the
one hand, we see that Leo s precepts are practical and efficacious ;
on the other, we see that they are based upon the supposition
that the Imperial troops will normally act upon the defensive, a
limitation which must materially impair their efficiency. Byzan
tine statesmen had long given up any idea of attempting the re-
conquest of Italy; they aimed at nothing more than retaining their
hold on the " Calabrian " and " Langobardic " themes. Hence
come the caution and want of enterprise, the proneness to sleights
and stratagems, displayed|in Leo s chapters, characteristics which
lead the Prankish writers into stigmatising the Past-Romans as
treacherous and cowardly. To win by ambushes, night attacks,
and surprises, seemed despicable to the Prankish mind. These,
nevertheless, were the tactics by which the Pastern emperors suc
ceeded in maintaining their Italian provinces for four hundred
years against every attack of Lombard duke or Prankish
emperor.

204 THE ART OF WAR IN THE MIDDLE AGES [900

The method which is recommended by Leo for resisting the
" Turks " (by which name he denotes the Magyars and the
Patzinaks who dwelt north of the Euxine 1 ) is different in every
respect from that directed against the nations of the West. The
Turkish hordes consisted of innumerable bands of light horse
men who carried javelin and scimitar, but relied most of all on
their arrows for victory. They were " given to ambushes and
.stratagems of every sort," and were noted for the care with
which they conducted their scouting and posted their vedettes.
In battle they advanced not in one mass, but in small scattered
bands, which swept along the enemy s front and around his
flanks, pouring in flights of arrows, and executing partial charges
if they saw a good opportunity. On a fair open field, however,
they could be ridden down by the Byzantine heavy cavalry, who
are therefore recommended to close with them at once, and not
to exchange arrows from a distance. Steady infantry also they
could not break, and foot-archers were their special dread, since
the bow of the infantry-soldier is larger and carries farther than
that of the horseman ; thus they were liable to have their horses
shot under them, and when dismounted were almost helpless, the
nomad of the steppes having never been accustomed to fight on
foot. The general who had to contend with the Turks, therefore,
should endeavour to get to close quarters at once, and fight them
at the earliest opportunity. But he should be careful about his
flanks, and cover his rear if possible by a river, marsh, or defile.
He should place his infantry in the front line, with cavalry on
the flanks, and never let the two arms be separated. Heedless
pursuit by the cavalry was especially to be avoided, 2 for the
Turks were prompt at rallying, and would turn and rend
pursuers who followed in disorder. But a proper mixture of
energy and caution would certainly suffice to defeat a Turkish
host, because in the actual clash of battle they were man for
man inferior to the Imperial Catapkrctcti. These chapters would
have been the salvation of four generations of Western Crusaders
if their chiefs had but been able to read them. Well-nigh every
disaster which the Crusaders suffered came from disobeying some

1 Apparently also the Bulgarians (xviii. 42-44), as he speaks of them as a
Scythian race very like the Turks, and again, of their "differing little or not at all
from each other in their way of life and their methods of war."

: Never let the cursores get more than three or four bowshots from the
is Leo s general rule.

9 oo] TACTICS USED AGAINST THE TURKS 205

one of Leo s precepts from falling into ambushes, or pursuing
too heedlessly, or allowing the infantry and cavalry to become
separated, or fighting in a position with no cover for rear or
flanks. The Byzantines, on the other hand, made on the whole
a very successful fight against the horse-archers who overwhelmed
so many Western armies. It is true that one huge disaster, the
defeat of Manzikert, brought on by the rashness of Romanus iv.,
was perhaps the most fatal blow that the empire ever received.
But, with this and a few other exceptions, the East-Roman armies
gave a good account of themselves when dealing with the Turk.
Alexius Comnenus, though not a genius, was always able to
defeat the Patzinaks ; his son and grandson reconquered from
the Seljouks half Asia Minor, and, even after the Latin conquest
of 1204, Lascaris and Vatatzes held them back. It was not the
horse-archers of the older Turkish tribes, but the disciplined
janissaries of the Ottomans that were destined to give the
coup de grace to the Eastern Empire.

The third group of nations with which Leo deals are the
Slavonic tribes Servians, Slovenes, and Croatians, who inhabited
the north-western parts of the Balkan peninsula. The space
devoted to them is much less than that spent on each of the
other categories of the enemies of the empire. Leo remarks
that since their conversion to Christianity in the reign of his
father Basil, and the treaty in 869 which had made the Dalmatian
and Bosnian Slavs, in name at least, vassals of the empire, they
had given no trouble. They were a nation of foot-soldiers, and
only formidable when they kept to the mountains, where their
archers and javelin-men, posted in inaccessible positions, could
annoy the invader from a distance, or their spearmen make
sudden assaults on the flank or rear of his marching columns.
Such attacks could be frustrated by proper vigilance, while, if
surprised in the plains when engaged in a plundering expedition,
they could be easily ridden down and cut to pieces by the
Imperial cavalry, since they had no idea of discipline and no
defensive arms save their large round shields. Leo gives no
description of the Russians, though they were already beginning
to plague the themes along the Euxine coast. 1 Had he devoted
a chapter to them, we should be the richer by some interesting
details of their early military customs. Sixty years later, when

1 Their first expedition had been in 865, and there was one in Leo s own reign
in 907.

:o6 THE ART OF WAR IN THE MIDDLE AGES [900

they fought John Zimisces, they had adopted the armour and
tactics of their Varangian chiefs, and resembled the Northmen
rather than the Slavs of the South, fighting with shirts of mail,
long kite-shaped shields, and battle-axes, and arraying themselves
in well-ordered columns, which could often beat off cavalry. It
took the most strenuous efforts of the gallant Zimisces and his
chosen horse-guards to break into these stubborn masses, and
the battle of Dorostolon was one of the hardest fought and
perhaps the most creditable of all the victories of the Byzantine
armies (97 1).

The longest and most interesting paragraphs in Leo s
Eighteenth " Constitution " are reserved for the Saracens, and
his description of them can be amplified by details from the very
interesting lit ft napa8p.,a?f n&"/.i,aw, a work written about 980 by a
trusted officer of Nicephorus Phocas, who desired to preserve
his late master s precepts and practice in a literary shape. The
little book is practically a manual for the governors of themes
on the eastern border, giving all the methods to be employed
in repelling Saracen raids, and all the precautions necessary for
the execution of retaliatory invasions of Saracen territory. It is
especially valuable because, unlike the 7*a&lt;:tica of Leo, it gives
lavish historical illustrations and examples, and does not confine
itself to precept.

To deal with the Saracen, the most formidable enemy of the
empire, the greatest care and skill were required. " Of all
barbarous nations," says Leo, " they are the best advised and most
prudent in their military operations." The commander who has
to meet with them will need all his tactical and strategical
ability, the troops must be well disciplined and courageous, if
the " barbarous and blaspheming Saracen " is to be driven back in
rout through the " clissuras " of Taurus.

The Arabs whom Khaled and Amru had led in the seventh
century to the conquest of Syria and Egypt had owed their
victory neither to the superiority of their arms nor to the
excellence of their organisation. The fanatical courage of the
fatalist had enabled them to face better-armed and better-
disciplined troops, as it nerved the Soudanese ten years ago to
face the breechloaders of our own infantry. We, who remember
the furious rush that once broke a British square, cannot wonder
that the troops of Heraclius, armed only with pike and sword,
were swept away before the wild hordes of the early Caliphs.

x&gt;] TACTICS USED AGAINST THE SARACENS 207

It is greatly to the credit of the East-Roman troops and the
louse of Heraclius that Asia Minor did not suffer the same fate
Persia and Spain. But when the first flush of fanaticism had
sassed by, and the Saracens had settled down in their new
homes, they did not disdain to learn a lesson from the nations
they had defeated. Accordingly, the Byzantine army served as
a model for the forces of the Caliphs. " They have copied the
Romans," says Leo, " in most of their military practices, 1 both in
arms and in strategy." Like the Imperial generals, they placed
their confidence in their mailed lancers : they were no longer
the naked hordes of the sixth century, but wore helms, shirts of
chain - mail, and greaves. But the Saracen and his charger
were alike at a disadvantage in the onset : horse for horse and
man for man the Byzantines were heavier, and could ride the
Orientals down when the final shock came.

By the tenth century the Saracens had an art of war of their
own. Some of their military works have survived, though
none, it appears, date back to the times contemporary with Leo.
They had advanced very considerably in poliorcetics and forti
fication ; they had learned how to lay out and entrench their
camps, and how to place pickets and vedettes. But they never
raised a large standing army, or fully learned the merits of drill
and organisation. The royal bodyguards were their only regular
troops ; the rest of the army consisted of the war-bands of chiefs,
miscellaneous bands of mercenary adventurers, or the general
levies of tribes and districts.

Two things rendered the Saracens of the tenth century-
dangerous foes, their numbers and their extraordinary powers of
locomotion. When an inroad into Asia Minor was on foot, the
powers of fanaticism and greed united to draw together even-
unquiet spirit from Egypt to Khorassan. The wild horsemen
of the East poured out in myriads from the gates of Tarsus and
Adana to harry the rich uplands of the Anatolic, Armeniac,
and Cappadocian themes. " They are no regular host, but a
mixed multitude of volunteers ; the rich man serves from pride
of race, the poor man from hope of plunder. They say that
God, who scattereth the armies of those that delight in war, is
pleased by their expeditions, and has promised victory to them.
Those who stay at home, both men and women, aid in arming
their poorer neighbours, and think that they are performing a

1 Tcutica, xviii. 120.

208 THE ART OF WAR IN THE MIDDLE AGES [900

good work thereby. So mere untrained plunderers and ex
perienced warriors ride side by side in their hosts." 1

Once clear of the passes of Taurus, the great horde of
Saracen horsemen cut itself loose from its communications, and
rode far and wide through Cappadocia and Phrygia, burning the
open towns, harrying the countryside, and lading their beasts
of burden with the plunder of a region which was in those days
one of the richest in the world. It was only exceptionally that
the invaders were aiming at serious conquests and halted to
besiege a fortified town. The memory of the awful failures of
the two great hosts that perished before Constantinople in 673
and 718 seems to have been deep impressed in the minds of the
Mohammedan rulers and generals. The two last attempts at
getting a footing beyond the Taurus were those of Haroun-al-
Raschid in 806, and of Al-Motassem in 838. Each, after taking
one considerable town, found such a long and difficult task
before him that he gave up his project and retired. The armies
of their successors, even when counted by scores of thousands,
were aiming at nothing more than vast plundering raids.

When the Saracens had passed the defiles of Taurus, they
pushed on for some days at an almost incredible speed, for their
Da ggage was all laid on camels or sumpter beasts, and their foot-
soldiery were either provided with horses of some sort or taken
up on the cruppers of the cavalry. 2 They made for the district
that they had marked out for plunder, and trusted to arrive in
such haste that the natives would not have had time to gather
in their property and take shelter within walled towns.

Now was the time for the Byzantine general to show his
mettle. If he was a competent commander, he would have had
regular outposts, relieved every ten or fifteen days, to watch the
passes. The moment that these were driven in, they would take

Leo here adds, xviii. 129: "And would that we Christians did the same.
For if all of us, both soldiers and those who have not yet borne arms, could agree to
strengthen our hearts and go forth together, if ever) man armed himself, and the
people gave their money to equip such a. host, and their prayers to help it, then
marching against that race which blasphemes our Lord and God, Christ, the King of
all, we should obtain victory. For the Roman armies being increased manifold, and
furnished liberally with all weapons of war, and abounding in military skill, and
having heaven as their aid, could not fail to crush the barbarous and blaspheming
Saracen." This surely is the spirit of the Crusader, appearing two hundred years
before its time.

2 Toi&gt;r St irefous aiTav Qtpovffiv f, i&lt;j&gt; iWuv iSlwv oxov/rfvovs, rj fnria&cv TUV K a8a\-
(xviii. 115).

9 oo] HOW TO DEAL WITH SARACEN RAIDS 209

the tidings to the chief town of the theme, and to the nearest com
manders of bands and turmae. While the main body of the cavalry
of the theme concentrated under the strategos at a central point,
it would be the duty of the turmarch into whose district the
raid had come, to collect the nearest two or three bands in haste,
and to hang on to the skirts of the invading force at all costs.
For even a small observing force compels the invaders to move
cautiously, and to abstain from letting their men straggle for
plunder. Meanwhile, all the disposable foot -soldiery of the
theme would be hurried off to seize the mouths of the passes
by which the enemy would probably return. These were not
so numerous but that a competent officer might make some
provision for obstructing them all. 1

To ascertain the enemy s route and probable designs, the
commander of the theme must spare no pains. The turmarch
charged with following the raiders ought to be sending him con
tinual messages ; but in addition, says Leo, " never turn away
freeman or slave, by day or night, though you be sleeping or
eating or bathing, if he says that he has news for you." Success
is almost certain if continual touch with the enemy is kept up ;
the most disastrous consequences may follow if he is lost. When
the strategos has concentrated all or most of his regiments, he
makes with all speed for the district which the raiders are
reported to have reached. If they are in comparatively small
numbers, he must endeavour to fight them at once. If they are
too strong for him, he must obstruct their way by all means
which do not expose him to an open defeat. If there are fords
or defiles on their path, he must defend them as long as possible ;
he must block up wells and obstruct the roads with trenches.
Above all, he must endeavour either to cut off all raiding parties
that leave the enemy s camp, or if these are too strong to
adopt the opposite course, and storm the camp in their absence.
By such devices he may either worry them into returning, or
else detain them long enough to allow of the arrival of the
mobilised troops of two or three neighbouring themes. When a
sufficient force has accumulated, open battle can be tried. But
these Saracen invasions in force ("Warden-Raids," if we may
borrow a phrase from the similar expeditions of our own

1 All this is from Nicephorus Uepi UapaSpofj.^ HoX^/uou. cap. i. I. The chapter
is really excellent ; it might be used on the Indian north-west frontier to-day, so
practical is it.

210 THE ART OF WAR IN THE MIDDLE AGES [900

Borderers) were of comparatively unfrcquent occurrence, and it
was not often necessary to " set all the rest of the themes of the
East marching," each with its picked corps of four thousand or
four thousand five hundred cavalry. If needed, however, Leo
states that thirty thousand cavalry of the best quality could be
collected in a moderate space of time. A most perfect instance
of such a concentration had taken place in A.D. 863 (though Leo
does not mention it l ), when a great Saracen army under Omar,
the Emir of Malatia, had been completely surrounded and
absolutely exterminated by the skilful and simultaneous appear
ance of no less than ten contingents, each representing a theme. 2
The more typical Saracen inroad, however, was on a smaller
scale, and only included the warriors of Cilicia and Northern
Syria, assisted by casual adventurers from the inner Moham
medan regions. To meet them the Byzantine commander
would have no more than the four or five thousand horsemen
of his own theme. When he came up with them, they would
probably turn and offer him battle : nor was their onset to be
despised. Though unequal, man for man, to their adversaries,
the Saracens were usually in superior numbers, and always
came on with great confidence. " They are very bold when
they expect to win : they keep firm in their ranks, and stand
up gallantly against the most impetuous attacks. When they
think that the enemy s vigour is relaxing, they all charge together
in a desperate effort." If this, however, failed, a rout generally
followed, " for they say that all misfortunes come from God, and
if they are once well beaten, they take it as a sign of divine
wrath, and altogether lose heart." Their line once broken, they
have not discipline enough to restore it, and a general sauve qui
peut follows. Hence a Mussulman army, when routed, could be
pursued d routrance? and the old military maxim, Vincesedm
nimis vincas, was a caution which the Byzantine officers could
disregard.

In the actual engagement with the Saracen foe, the tactics

1 Perhaps because the reigning emperor was Michael III., whom Basil I. (Leo s
father) had murdered,

- Having sacked Amisus and ravaged Paphlagonia and Galatia, Omar found his
way home blocked by the contingents of the Anatolic, Obsequian, and Cappadocian
themes ; at the same time those of the Buccellarian, Paphlagonian, Armeniac, and
Colonean themes encompassed him on the north ; and that of the Thracesian theme,
strengthened by European troops of the Macedonian and Thracian themes, closed in
on the west. The Saracens were absolutely exterminated.

3 Nic. Phoc. xxiv. 10.

9 oo] SARACEN METHODS OF WAR 211

recommended were those of the double line, with flank-guards,
reserve, and outlying detachments to turn the enemy, which we
have described in the section dealing with the organisation
of the Byzantine army. The Saracens were accustomed to
array themselves in one very deep line, which Leo calls a
solid oblong (TEr^dywvov -/.,; sTi,aj}xj wapoTaJjiv). Their cavalry were
practically the sole force that gave trouble, the foot being a
mere rabble of plunderers, which would never stand. Their
only useful infantry were composed of Ethiopian archers, but
these, being wholly destitute of defensive armour, could never
face the Byzantine footmen. In battle the single heavy line
of the Orientals must under ordinary circumstances give way
before the successive charges of the three Byzantine lines. The
generals of the East had already discovered the great precept
which modern military science has claimed as its own, that " in
a cavalry combat the side which holds back the last reserve
must win." They were equally masters of the fact that this
last reserve should be thrown in on the flank rather than on the
front of the enemy. It was not, therefore, without reason that
the author of the napaopo/^ exclaims that " the commander
who has five or six thousand of our heavy cavalry and the help
of God needs nothing more." l

It would sometimes, however, happen that the Saracens were
not caught on their outward way, and that the forces of the
Byzantine general only closed in on them as they were retreating. 2
Loaded with booty, the raiders would be constrained to move
far more slowly than on their advance ; their camps, too, would
be filled with captured herds and flocks, laden waggons, and
troops of prisoners. In this case Xicephorus Phocas recom
mended a night attack, to be delivered by infantry or dismounted
cavalry. " Send three infantry bands, ranged a bowshot apart,
to charge into each flank of their camp," says the emperor,
"assail the front a little later with your main body of foot, and
leave the rear, where lies the road to their own land, unattacked.
In all probability the enemy will instinctively get to horse, and
fly by the only way that seems to lead to safety, leaving their
plunder behind them." 3

1 Nic. Phoc. Preface, s IS-
- Nic. Phoc. xvii, 15.

3 ei 52 aviiflri \\j6Tjvai. T ! qv irapaTa^tf, 01 eavruv affLVraroi vai avciriarpoqiOL yevbticvoi
MO"V IV fftiiOf/nai eAa Voi O i) (xviii. 116).

2i2 THE ART OF WAR IN THE MIDDLE AGES [5

But success was most certain of all if the invaders could
caught while retreating through the passes of Taurus. If t
infantry of the theme had succeeded in reaching the defiles a
posting themselves there before the retreating enemy arrivi
\vhile at the same time the pursuing cavalry pressed them
the rear, the Saracens were lost. Wedged in the narrow ro;
with their line of march mixed with countless waggons a
sumpter-beasts laden with spoil, they were quite helple
They could be shot down by the archers, and would not stai
for a moment when they saw their horses, " the Pharii whc
they esteem above all other things," struck by arrows from
distance ; for the Saracen, when not actually engaged in clc
combat, would do anything to save his horse from harm. 1

The most noted instance of a victory of this kind was tb
won in 963 by Leo Phocas, brother of Nicephorus, over t
hosts of Seif-ed-dauleh ben Hamdan, Emir of Aleppo. Thou;
he had with him only the forces of his own theme of Charsian
Leo captured or slew the whole of the Saracen army, recoven
much plunder, liberated many thousands of Christian prisone
and bore off in triumph the standard and the silver can
equipment of the emir. Mohammedan historians confess t
greatness of the disaster, though the} reduce the number
their slain to three or four thousand. 3 Se ff-ed-dauleh hims&lt;
escaped with three hundred men only, by climbing an almc
impracticable precipice. His ruin is ascribed by Abulfeda
the fact that he had dared to return to Cilicia by the same pa:
that of Maghar-Alcohl, by which he had entered into the Rom;
territory. It is interesting to find the very methods which L
describes in 900 used sixty years after with perfect success-
sufficient proof that the emperor was not altogether undeservii
of his name of " the Wise."

Many other points of interest may be gathered from tl
chapters of Leo and of Nicephorus Phocas. Cold and rail
weather, we learn, was distasteful to the Oriental invader ;
times when it prevailed he did not display his ordinary firmne
and daring, and could be attacked with great advantage. Mui
might also be done to check his progress by delivering a vigoro
counter-attack into Cilicia or Northern Syria, the moment th
the Saracen was reported to have passed north into Cappadoc

1 Leo, xviii. 135. - Xic. Phoc. Preface, 15.

3 Jemaleddin, p. 134; Abulfeda, ii. 469.

So] SUCCESSFUL ADVANCE AGAINST THE SARACENS 213

or Charsiana. On hearing of such a retaliatory expedition, the
Moslems would often return home to defend their own borders. 1
This destructive practice was very frequently adopted, and the
sight of two enemies each ravaging the other s territory with
out attempting to defend his own was only too familiar to
the inhabitants of the borderlands of Christianity and Islam.
Incursions by sea supplemented the forays by land. " When
the Saracens of Cilicia have gone off by the passes, to harry the
county north of Taurus," says Leo, " the commander of the
Cibyrrhaeot theme should immediately go on shipboard with all
available forces, and ravage their coast. If, on the other hand,
the Cilicians have sailed off to attempt the shore districts of the
Imperial provinces, the clissurarchs of Taurus can lay waste the
territories of Tarsus and Adana without danger."

All through the tenth century the Saracens were growing less
and less formidable foes, owing to the gradual dropping off of
the outlying provinces of the empire of the Abbassides, who by
the end of the period were masters of little more than the
Euphrates valley, and were dominated even in their own palace
by their Turkish guards. The Byzantine realm, on the other
hand, under the steady and careful ministers who served
the Macedonian dynasty, was at its very strongest. For a
hundred and fifty years after the accession of Basil I., the empire
was always advancing eastward, and new themes were continu
ally being formed from the reconquered territory. There is a
great difference of tone between the language which Leo, writing
about 900, and the author of the Hc/.pabpo/^, writing about 980,
use concerning the Saracen enemy. To the former they are
still the most formidable foes of the empire ; the latter opens
his preface with the words : " To write a treatise on frontier
operations may seem at the present day no longer very
necessary, at least for the East, since Christ, the one true God,
has in our day broken and blunted the power of the sons of
Ishmael, and cut short their raiding. . . . But I write neverthe
less, thinking that my experience may be useful, because I was
an eye-witness of the commencement of our successes and of
the application of the principles which led to them. Through
the use of these principles I have seen small armies accomplish

1 The author of the Hapa!po/j.-fi speaks of this device, quoting it as a good piece
of counsel given by Leo, and gives as example an occasion when the siege of Misthea
was raised by means of a retaliatory raid against Adana (xx.).

214 THE ART OF WAR IN THE MIDDLE AGES [980

great feats. What once, when the Saracens of the border were
strong, seemed impossible to a whole Roman army, has been of
late carried out by a single good general with the forces of a
single theme. By the use of these principles I have seen a force,
though too small to face the enemy in open fight, yet defeat
his purpose, and preserve our borders unravaged. The system
was first, as far as I know, utilised in modern times by Bardas
Caesar, 1 who foiled the Saracens of the Tarsiot border not once
but ten thousand times, and erected countless trophies over them.
Constantine Melei nos, strategos for many years in Cappadocia,
won magnificent successes by using these principles. 2 But
Nicephorus Phocas, that prince of immortal memory, accom
plished by their use feats that defy description and enumeration.
He it was who bade me write down the system, for the use of
future generations. And this I do with the more readiness
because it can be applied not only to the eastern border, but to
the western, as I (who have served most of my time on the
latter) can state from my own experience."

By the end of the tenth century the Byzantines were habitu
ally taking the offensive against the Saracens, and, instead of
seeing Cappadocia or Phrygia ravaged, were themselves pushing
their incursions almost to the gates of Damascus and Bagdad.
The conquest of Cilicia by Nicephorus Phocas was but the first
of a series of advances which promised ultimately to restore to
the empire the frontier that it had held in the days of Justinian.
Antioch was conquered, the Emirs of Aleppo and Tripoli were
made tributary, and kept in that position for sixty years. Even
after the death of Basil II., the greatest soldier of the Eastern
realm, the Imperial borders continued to advance eastward :
Edessa was captured in 1032, and a new theme was established
in Mesopotamia. The whole of Armenia was annexed in 1045,
and Constantine IX. might have boasted that his provinces ex
tended farther to the East than those of any of his predecessors
since Trajan.

1 This, I suppose, was the unfortunate Bardas Caesar who was murdered by his
nephew Michael III. in 866. There had been some great victories in his day,
notably that over Omar (see p. 210), and he is said to have devoted much attention
to military affairs, but it is surprising to find him given such a marked place by the
author of the Hapaopo,u.tf. Did his exploits inspire the sections on border warfare in
Leo s Tactical

" There were several good generals of this name. I suppose this to he the one
who ruled Cappadocia about 960 A. D.

THE COMING OF THE SELJOUKS

2I 5

But at the moment when the East - Roman boundaries

cached their largest extent, the new foe was at hand who was

deal the fatal blow from which the empire was never wholly

recover. The disastrous day of Manzikert (1071) is really

tie turning-point in the history of the great East - Roman

ealm.

CHAPTER IV

DECLINE OF THE BYZANTINE ARMY IO7I-I2O4

THOUGH the internal condition and administration of the
empire had been steadily deteriorating since the death of
Basil ii. (1024), it cannot be said that its army showed any decline
till the very day of Manzikert. Indeed, as we have already seen,
the Imperial frontier continued to advance down to the moment
of that disaster, and the first advance of the Seljouks was met
without wavering. For some years the Turks had no higher
aim than to win booty by sudden inroads into Asia Minor. Of
their raiding bands some were turned back, and some cut to
pieces ; but their numbers were so great that the line of defence
could not be held everywhere, and on different occasions Caesarea,
Iconium, and Chonae fell into their hands. No lodgment, how
ever, was made in the empire, and the fact that the decisive
battle was fought so far east as Manzikert, in farther Armenia,
hard by the Lake of Van, shows that the hold of the government
on its frontier provinces was not yet shaken.

The Seljouks of Alp Arslan were in tactics just like the Turks
whom Leo the Wise had described a century and a half before.
They only differed from the Patzinaks and other Western tribes
of the same blood by their enormously superior numbers. Xo
such formidable invasion had befallen the empire since the days
of Leo the I saurian, and to meet it there sat on the Byzantine
throne a gallant hot-headed soldier with a doubtful title and
many secret enemies. Romanus Diogenes had been lately raised
to the purple by his marriage with Eudocia, the widow of
Constantine XI., and reigned as colleague and guardian of her
young son Michael. He knew that he was envied and hated
by many of his equals, who had aspired to fill the same place:
hence he was nervously anxious to justify his elevation by
military success, as his great predecessors, Nicephorus Phocas

216

i] BATTLE OF MANZIKERT 217

and John Zimisces, had done. He was in the field for almost
the whole of the three uneasy years for which he reigned (1068-
71); and if energy and ceaseless movement could have driven off
the Seljouks, he must have been successful. But he was a bad
general, easily distracted from his aims, and too quick and rash
in all his actions.

In the spring of 1071 Romanus collected a very large army,
at least sixty thousand strong, and betook himself to the extreme
eastern corner of his dominions, with the intention of meeting
the Turks at the very frontier, and recovering the fortresses of
Akhlat and Manzikert, which had fallen into their hands. He
had retaken the latter place, and the former was being besieged
by a detached division of his army, when the main host of the
Seljouks came upon the scene. It was a great horde of horse-
archers, more than a hundred thousand strong, and full of confid
ence in its victorious Sultan. The tactics which Romanus should
have employed were those laid down in Leo s manual to beware
of ambushes and surprises, never to fight with uncovered flanks
or rear, to use infantry as much as possible, and never to allow the
army to get separated or broken up. Romanus violated all
these precepts. His first brush with the enemy was a disaster
on a small scale, caused by pure heedlessness. When a small
body of Turkish cavalry came forward to reconnoitre the Imperial
camp, it was furiously charged by a rash officer named Basilakes,
who commanded the theme of Theodosiopolis : he drove it
before him till he lost sight of his master, and fell into an ambush,
where he and all his men were killed or captured. A division
which Romanus sent to support them found nothing but the
bodies of the slain.

With this warning before him, the emperor should have acted
with all caution : perhaps, indeed, he intended to do so till his
rashness ran away with him. He drew up his host in front of
his camp with great care. The right wing was composed of the
cavalry from the easternmost themes Cappadocia, Armeniacon,
Charsiana, and the rest, under Alyattes, strategos of the Cap-
padocian theme. The left wing, under Nicephorus Bryennius,
was formed of the drafts of the European themes. In the centre
was the emperor, with his guards and the regiments of the
metropolitan provinces. A very strong rear line, composed of the
mercenary cavalry (which included a regiment of Germans and
also some Normans from Italy) and the levies of the nobles of

2i8 THE ART OF WAR IN THE MIDDLE AGES [1071

the eastern frontier, 1 was placed under Andronicus Ducas, a
kinsman of the late Emperor Constantine. He was unfortunately,
though a good officer, a secret enemy of Romanus.

Alp Arslan had been so moved by the news of the size and
splendour of the army which was moving against him, that on
the morning after the skirmish in which Basilakes had been
captured, he sent an embassy offering peace on the terms of
n fi possidetis. He would withdraw and undertake to make no
further invasions of the empire. Romanus was probably right in
refusing to negotiate, for Turkish promises could not be trusted.
He told the ambassadors that the first condition of peace must
be that the Sultan should evacuate his camp, retire, and allow it
to be occupied by the Imperial forces. Alp Arslan would not
consent to sacrifice his prestige, and the armies were soon in
collision. The Turks, after their usual manner, made no attempt
to close, or to deliver a general attack on the Imperial host.
Large bodies of horse-archers hovered about and plied their
bows against various points of the line. The Byzantine cavalry
made such reply as they could, but, their skirmishers being out
numbered, suffered severely in the interchange of arrows, and
many horses were disabled. Both the emperor and his troops
grew angry at the protraction of this long random fight, and in
the afternoon Romanus gave orders for the whole line to advance.
He was, however, sufficiently master of himself to see that the
distances were observed, and that the reserve division kept its
place accurately, so as to prevent any attack from the rear. For
some hours the host drove the Turks before them, inflicting, how
ever, little loss, as the enemy refused to make a stand anywhere ;
they even passed over the site of the Sultan s camp, which had
been evacuated and emptied of all its contents some hours before.
As the dusk came on, Romanus halted : his men were tired and
thirsty, and he had left his camp insufficiently garrisoned, so that
he was anxious to return to it, lest it might be surprised in his
absence. Accordingly, he gave orders to face about and retire.
Then began the disasters of the day : the order to retreat was
not executed with the same precision in all the divisions of the
host ; those on the flanks received it late, did not understand
its cause, and, when they wheeled about, did not keep their
dressing with the centre. Gaps began to appear between several

1 These are, I suppose, the eralpoi and ri&gt; O.PXOVTIKUV of which Eryennius speaks
in his account of the battle.

toy i] BATTLE OF MANZIKERT 219

of the corps. The Turks, according to their custom, commenced
to close in again when the army commenced its retreat. They
molested the retiring columns so much that Romanus at last
gave orders to face about again and beat them off. The whole
front line carried out this order, but the reserve under Andronicus
did not : out of deliberate malice, as most of the authorities allege,
this treacherous commander refused to halt, and marched back
rapidly to the camp, observing that the day was lost, and the
emperor should fight out his own battle. To lose the rear line,
and to be left without any protection against circling move
ments on the flanks, was fatal. The Turks began to steal round
the wings and to molest the fighting line from behind : they
particularly concentrated attention on the right wing, which,
trying to face both ways, fell into disorder in the twilight, and
at last broke up and fled. The victors at once fell on the flank
and rear of the centre, where the emperor made a gallant defence,
charged repeatedly both to flank and rear, and held his own.
But the European troops in the left wing had got divided from
the centre, and, after fighting a separate battle of their own, gave
way, and were driven off the field. Thus left isolated, Romanus
encouraged his men to stand their ground, and held out till
dark, when the Turks broke into his column and made a dread
ful slaughter. The emperor s own horse was killed beneath him :
he was wounded and taken prisoner, with many of his chief
officers : the whole centre was cut to pieces, and not a man of it
escaped.

Thus Romanus Diogenes, like Crassusof old, paid the penalty
for attacking a swarm of horse-archers in a open rolling country,
where he had cover neither for his flanks nor for his rear. It is
only fair to say that he would have in all probability brought
home his army without any overwhelming loss but for the abomin
able misconduct of Andronicus Ducas. When encompassed by
the Turks on the open plain, he was not nearly so helpless as the
Romans had been at Carrhae : his force, being all cavalry, was
capable of fairly rapid movement, and a sufficiently large propor
tion of the men were armed with the bow to enable him to make
some reply to the Turkish arrows. Still, by his inconsiderate
pursuit of the enemy he had placed himself in a radically false
position : it is useless for heavy troops to pursue swarms of light
horse, unless they are able to drive them against some obstacle
a river or a defile, which prevents farther flight. In this case the

220 THE ART OF WAR IN THE MIDDLE AGES [1071

Turks could retire ad infiiiitum, while the Byzantines, continually
moving farther from their camp and their stores, were at last
brought to a standstill by mere fatigue. Their retreat was bound
to be dangerous ; that it was disastrous was the fault of Ducas,
not of his master. We shall see in our chapter on the Crusades
that the details of Manzikert show a striking similarity to those of
several later battles in which the chivalry of the West had to
face the same Turkish tactics.

The empire had suffered other defeats as bloody as that of
Manzikert, but none had such disastrous results. The captivity
of Romanus Diogenes threw the nominal control of the realm
into the hands of his ward, Michael Ducas, who, though he was
only just reaching manhood, displayed the character of a pedant
and a miser. His reign of seven years was one chaotic series of
civil wars: half a dozen generals in corners of the empire
assumed the purple ; and Romanus, after his delivery from prison,
tried to reclaim his crown. Meanwhile, the Seljouks flooded the
plateau of Asia Minor, almost unopposed by the remnants of the
Imperial army, who were wholly taken up in the civil strife. No
man of commanding talents arose to stem the tide, and ere long
the horse-bowmen of Malekshah, the son of Alp Arslan, were seen
by the yEgean and even by the Propontis. The Turkish invasion
was a scourge far heavier than that of the Saracens. While the
latter, when bent on permanent conquest, offered the tribute as
alternative to the "Koran or the sword," the Seljouks were mere
savages who slew for the pleasure of slaying. They were bar
barous nomads, who had no use for towns or vineyards or arable
land. They preferred a desert in which they could wander at
large with their flocks and herds. Never, probably, even in the
thick of the Teutonic invasions of the fifth century, was so much
harm done in ten short years as in Asia Minor during the period
1071-1081. By the end of the latter year the flourishing themes
which had been for so long the core of the East-Roman realm
had been reduced to mere wastes. Thirty years after Manzikert,
when the armies of the Crusaders marched from Nicaea to
Tarsus, right across the ancient heart of the empire, they nearly
perished of starvation in a land of briars and ruins.

It seemed for a time quite probable that the fall of Constan
tinople might put the crown to the misfortunes of the empire, for
the would-be Caesars who were contending for the throne left
the Seljouks alone. Both Michael vn. and his foe, the usurper

1079] DISAPPEARANCE OF THE ARMY OF THE EAST 221

Nicephorus Botaniates, actually bought the aid of Turkish
auxiliaries by formally surrendering whole provinces. In 1080
the barbarians even seized Nicasa, thus obtaining a footing on
the Propontis, and almost within sight of the gates of the capital.

In this chaos the old Byzantine army practically disappeared.
The regiments which had fallen at Manzikert might in time
have been replaced, had the Asiatic themes still remained in
the hands of the empire. But within ten years after the fall of
Romanus IV. those provinces had become desolate wastes : the
great recruiting-ground of the Imperial army had been destroyed,
and the damage done was irreparable. So wholly had the army
of the East been cut off, that in 1078 Michael Ducas, by collecting
all the scattered and disbanded survivors of the old corps from
the Asiatic side of the Bosphorus, and supplementing them with
recruits, only obtained a division of ten thousand men, the so-
called " Immortals," with whom the future emperor, Alexius
Comnenus, made his first great campaign. 1 Yet, only ten years
before, the Asiatic provinces had shown twenty-one themes, or a
standing army of at least a hundred and twenty thousand men.

The European themes were, no doubt, not so thoroughly dis
organised ; we find some of their old corps surviving into the time
of the Comneni. But even here great havoc was made by the
ten years of endemic civil war, from 1071-1081, and by the revolts
of the Servians and Bulgarians.

After Manzikert, indeed, we find foreign mercenaries always
forming both a larger and a more important part of the Imperial
host than in the flourishing days of the Macedonian dynasty.
Franks, Lombards, Russians, Patzinaks, Turks, were enlisted in
permanent corps, or hired from their princes as temporary
auxiliaries. It is no longer the old Byzantine army which we
find serving under Alexius Comnenus and his successors, but a
mass of barbarian adventurers, such as the army of Justinian had
been five hundred years before. The old tactics, however, still
survived : the generals were the same if the troops were changed.
A concrete example may be quoted to show the old methods still
prevailing.

In A.D. 1079 Nicephorus Botaniates, who sat on a most

1 /3d(ri\ei S MixaTjX i5wv r6 TTJS EaJas ffrpaTevfjLa aTraf 7)677 ^xXeXoiTTos, us vTro^ipi
TWV foi pKuv yfvbfifi ov, typ&i TUTt (is oiov Tf aTpaTfVfM KaraffTrjirat Vf6\fKTOv, xai
Tims TUV CK TTJS Atrias SuKnnipfVTuii rai iirl fiiaffu Sov\fv6vruiv &lt;rv\\{yui&gt;, 6wpa.K&s
fvfSvf KO.I 8vpeoi&gt;s {SLdov, etc. etc. (Nic. Bry. iv. 4).

222 THE ART OF WAR IN THE MIDDLE AGES [1079

uneasy throne at Constantinople, sent against the rebel Nicephorus
Bryennius his general Alexius Comnenus, whom he had lately
made " Domestic of the Scholae," i.e. commander of the Imperial
Guard. Nearly all the European provinces had fallen away to
Bryennius, and as Asia had been overrun by the Turks as far as
Nicxa and the Fropontis, the ruler of Constantinople was not
able to put into the field so large an army as the insurgents.

The armies,both wholly composed of cavalry, met at Calavryta,
hard by the river Halmyrus. Comnenus, as the weaker of the two,
waited to be attacked, and chose a position with a comparatively
narrow front, apparently where a road crossed the slope of a hill :
on the left of his position were some hollows, screened from the
eyes of those approaching from the plain by a rise in the ground.
Comnenus drew up his main body, composed of the " Immortals"
whom Michael Ducas had organised, and a body of Prankish
mercenaries, across the road. He hid his left wing in the hollows,
ordering them to keep wholly out of sight till the enemy should
have passed them, and then to charge in upon Bryennius right
flank. His right wing, composed of garrison troops strengthened
by a considerable force of Turkish mercenaries all horse-archers
was placed under the command of Catacalon ; it was in
military terminology " refused," and ordered to devote its whole
attention to preventing the enemy from turning the flank of the
main body. Thus, to use the technical terms of Leo s Tactica,
Comnenus had wopoi or uKipxipa.ara.1 on his left wing, and
T?.ay/opi/.ax; on his right.

Bryennius, on the other hand, came on with his host divided
into three parallel columns. The right wing, five thousand
strong, was led by his brother John, and contained the cavalry
of the theme of Thessaly and the veteran remnants of the old
army of Italy, which had long served under John Maniakes
against the Normans and Saracens. The left wing, under
Tarchaniotes, three thousand strong, was composed of Mace
donian and Thracian regiments. The centre, led by the usurper
himself, was also formed from Macedonian and Thracian corps,
strengthened by a picked body of ap%otri; local nobles and their
followers. But Bryennius intended to strike his chief blow with
a body of Scythian (Patzinak) horse detached from his main
army and moving a quarter of a mile to its left, with orders to
turn the right of Alexius line, serving in fact, as Leo would
have said, as i&lt;jrepxepu.arai.

1079] BATTLE OF CALAVRYTA 223

When the rebel army came level with the hollows where the
Imperialist left was concealed, the hidden troops suddenly issued
forth and charged John Bryennius in flank, while Comnenus
and his main body rode down upon the usurper s own central
division. Both these attacks failed : John Bryennius wheeled to
his right in time, and beat off the attack of the troops in ambush.
Nicephorus Bryennius defeated the squadrons of the Immortals,
and drove them off the field, while the Prankish mercenaries
who formed the remainder of Comnenus centre were wholly
encompassed by the rebels, 1 and cut off from the possibility of
retreat. Meanwhile, on the extreme right of the Imperialist
army, the garrison troops under Catacalon had been charged
and routed by Bryennius flanking force of Patzinak horse.
The victorious barbarians went oft" in wild pursuit of the
fugitives, and seem to have overlooked the other corps on the
Imperialist right, the Turkish auxiliaries, who found themselves
left without an enemy in sight. 2 When the Patzinaks returned,
they began plundering their own employer s camp, instead of form
ing up to aid him in an engagement as yet by no means ended.

Alexius Comnenus had extricated himself with difficulty
from the mele e in the centre, and retired over the brow of the
hill, where he at once halted and began endeavouring to rally
his broken troops. During the combat he had charged into the
personal escort of the usurper, and had chanced to come upon
the squires who led the second charger of Bryennius, adorned
with purple housings and a gold frontlet, and carried the two
swords of state which were always borne on each side of an
emperor. Alexius and those with him had the fortune not only
to capture these insignia, but to cut their way out of the tumult
without losing them. Displaying the horse and the swords to
his routed troopers, Alexius proclaimed that he had slain
Bryennius. Encouraged by this fiction, a considerable body
formed up around him, and at the same time the Turks from
the left wing came up and placed themselves at his disposition.

Without delay Comnenus determined to attempt a second

1 I suppose by the wheeling in of Tarchaniotes men, who must have outflanked
Comnenus line considerably to the right, as the army of Bryennius was stronger by
far than that of the Imperialists.

- Probably the Patzinaks charged the extreme right corps, and so did not come
into contact with the one which lay nearer the Imperialist centre. Or possibly, as
one account of the fight might imply, the Turks were only just arriving on the field
when Catacalon was routed.

224 THE ART OF WAR IN THE MIDDLE AGES [1079

attack. He placed two bodies of the rallied troopers under
cover to right and left, and with part of the Turks and the
" Immortals " came down the hill again towards the site of the
first engagement. The victorious rebels were in some disorder:
many had dismounted to plunder the slain, and with them were
mixed their camp-followers, now fleeing from the Patzinak
marauders, who were beginning to plunder the tents. Bryennius
himself and the centre division were surrounding the Franks of
the Imperialist army, who, when they had been cut off, had
dismounted, and offered to surrender. The commanders of
these mercenaries were standing on foot before Bryennius and
doing homage to him just as Alexius came down the hill for his
second charge.

Though much surprised by the return of the enemy to the
fight, Bryennius and his men came boldly forward. Alexius
set his Turks to skirmish, and bade them empty their arrows
into the disordered rebels before he made any endeavour to
close ; he wished to fight a cautious battle, avoiding any general
charge. As the enemy advanced, he retired before them slowly
till he had reached the point far up the hill where he had left his
ambush. When he saw the flanks of Bryennius exposed to the
lateral attack, he halted, faced to the front, and charged. At the
same time the concealed troops, dashing out " like a swarm of
wasps," attacked the rebels on both flanks. Already much
disordered, and with hundreds of horses disabled by the Turkish
arrows, the squadrons of Bryennius could not face the charge,
but broke and fled. The rebel chief himself, with a small body
of devoted followers, refused to give ground, fought to the last,
and was finally dragged from his charger and taken prisoner. 1

The battle of Calavryta was fought in the time of the
Byzantine decadence which set in after Manzikert : there were
many raw troops in both armies, 2 and a large proportion of
foreign auxiliaries not drilled or disciplined after the traditional
methods of the Imperial army. Nevertheless, the incidents of
the fight show the main characteristics of the system which

1 Most of the details of this interesting fight came from Anna Comnena, who has,
for a lady, a very fair grasp of things military. No doubt she accurately put down her
father s account of his doings, and we are really reading Alexius versions of his fight.
Deducting the Homeric diction and the far too hairbreadth scapes of the narrator,
they are very favourable specimens of Byzantine military annals.

2 Alexius complained that the majority of the Immortals were recruits xSts rf TO!

i Kai 56paTa.

THE ARMY UNDER THE COMNENI 225

prevailed during the better days of the empire. Both generals
endeavour to win by flank attacks, Bryennius by an open one,
Comnenus by a sudden sally from an ambush. The horse-
bowmen Turks on one side, Patzinaks on the other are used
to prepare the way for the general charge. The troops have
enough discipline to rally around their unbroken reserve and
return to the charge within a very short time. Anna Comnena
most unfortunately forgets to tell us whether the corps fought,
according to the old rule, in a double line, with cursores and
defensores properly divided, and with a reserve. Nor does her
spouse, Nicephorus Bryennius, whose account tallies almost
exactly with hers, give us any more help on this point, though
he is careful to compliment his grandfather and namesake, the
usurper, on his military reputation.

The numerous contemporary chronicles which describe the
reigns of the three able Comneni, Alexius, John, and Manuel
(1071-1 180), show us that the old military organisation based
on the themes was never again restored. For the future the
Imperial army was a very haphazard and heterogeneous body.
When the western third of Asia Minor was reconquered by
Alexius and John, it was not divided up again into army-corps
districts. The Comneni, indeed, were centralisers, and preferred
to manage affairs from headquarters rather than to trust their
forces to the strategi of the themes. They preferred to raise
bodies of troops for general service rather than to localise the
corps. A dangerous proportion of the army was for the future
composed of foreign mercenaries : the earlier emperors had
enlisted Franks, Russians, and other aliens in considerable
numbers, but they had never made them the most important
part of the host. They had always been outweighed by the
regular cavalry of the themes. The Comneni, however, found
native troops hard to raise, now that the old Asiatic recruiting-
ground was gone, and they had also learned, from their contact
with the Normans of Robert Guiscard and with the knights of
the first Crusade, a great respect for Western valour. Prankish
adventurers were easy to enlist, they were less likely to rebel in
favour of pretenders than the native soldiery, and they had
proved at Dyrrhachium and many other fields that, man for
man, they could ride down the East-Roman troopers. Hence
Alexius I. and his descendants enlisted as many Western
mercenaries as they could get together. Nor was this all : the
15

226 THE ART OF WAR IN THE MIDDLE AGES [1204

Franks were not suited for light cavalry service, but the Turks,
Patzinaks, and Cumans excelled in it. To supplement the
Western spear the Comneni called in the Eastern bow.
Thousands of horse-archers hired from the nomad tribes rode
in their hosts. The native corps began to take quite a secondary
place : l they felt it, and resented it. In proportion as they were
despised, they grew less confident in themselves, less efficient,
and less daring.

The Comneni achieved many splendid feats of arms at the
head of their mercenary bands. They reconquered half Asia
Minor from the Seljouks, subdued the Franks of Antioch, and
routed the Magyars beyond the Danube. But they never built
up a real national army. When the strong hand of Manuel was
removed, and the wretched Angeli sat upon the Imperial throne
(11851204), the military machinery of the empire went to
wrack and ruin. The weak and thriftless emperors Isaac II.
and Alexius in. were neither able to find money to pay their
troops nor to maintain their discipline. A state which relies for
its defence on foreign mercenaries is ruined when it allows them
to grow disorderly and inefficient : in times of stress they mutiny
instead of fighting. Such was the fate of the empire in 1204:
when the Franks were actually breaking into the city, the
defenders struck for higher pay and refused to charge. The
city fell, and the old Byzantine military organisation passed
away.

1 There seems to have been some revival of local native forces during the
existence of the empire of Nicaea (1204-61). We hear of militia in Bithynia under
Lascaiis and Vataties, and their disbandroent by Michael Palaeologus is said to
have been one of the causes of the successful advance of the Ottoman Turks
(Pachy meres, i. 129).

BOOK V

THE CRUSADES
1097-1291.

227

CHAPTER I

INTRODUCTORY

BY the end of the eleventh century the supremacy of the
mailed horseman was firmly established all over Western
and Central Europe. In many countries infantry had practically
disappeared as a force that counted for anything in the day of
battle ; in all it had ceased to be the more important arm.
Only in nations of the remoter North and East the Irish,
Scandinavians, and Slavs did it still preserve its ancient
importance.

The three enemies who had threatened Christendom in the
ninth and tenth centuries had now been beaten off. The
Magyars had been pushed back to the line of the Leitha ; they
were now converted, and had become members of the common
wealth of Christian Europe. Instead of forming an impassable
barrier between Germany and Constantinople, they now offered
a free line of communication down the Danube. The Moors
had been driven out of Sicily and Sardinia instead of plaguing
Italy with their inroads, they were now busy in defending their
own African shore from the raids of the Genoese, Pisans, and
Normans. It seemed for a time as if the last-named of these
three maritime powers would actually eftect a lodgment south
of the Mediterranean. 1 In Spain, too, the balance had turned
definitely in favour of the Christians ; Toledo had fallen in
1085, and %vith its fall had ended the Moorish domination in
the central parts of the Iberian peninsula.

Lastly, the third and most formidable of the enemies of
Christendom had at last begun to slacken in their assaults.

1 The landmarks in the history of the struggle of the Italians and the Moors are
the expulsion of the latter from Sardinia in 1016 and from Sicily in 1060-91, the raids
on Bona and El-Mahadieh in 1064 and 1087. The last Moorish attacks on Italy had
only ceased early in the century, Pisa having been sacked in ion.

229

230 THE ART OF WAR IN THE MIDDLE AGES [1096

Scandinavia was now converted ; the fiercest of its Viking
hordes had found new homes for themselves in England,
Normandy, and Ireland, and were no longer seeking whom
they might devour. Harold Hardrada s raid of 1066, the last
of the great assaults of the Norsemen on their neighbours of
the South, had ended in utter defeat and disaster. Sweyn the
Dane, a few years later, had failed to make the least impression
on the new Norman kingdom of England. The peoples of the
North were just about to sink into the comparative obscurity
which covers them during the later half of the Middle Ages.

Free from external dangers for the first time since the days
of Charles the Great, the European nations were themselves
able to think of taking the offensive. The two all-impoitaAt
data which governed their enterprises, were, firstly, that a free
land route down the Danube to the borders of the Byzantine
Empire had become available since the conversion of the
Magyars; secondly, that the Italian states of Venice, Genoa,
and Pisa had lately developed war-navies strong enough to
guarantee a free passage for expeditions aiming at the Levant.
Down to the year looo the only naval powers in the Medi
terranean had been the Byzantines and the Moslems. The
whole face of affairs was changed by the appearance of the
Italian republics as a third party in the strife for supremacy
at sea.

Even before the preaching of the first Crusade there were
signs that Western Christendom was about to bestir itself and
take the offensive. The steady advance of the Germans against
trie Slavs of the East, the attacks of the Genoese and the
Sicilian Normans on Africa, were signs of the coming movement.
But no one could have foreseen the shape which the advance of
the European nations was to take. Swayed by a sudderi
religious impulse, they threw themselves upon the Levant, and
began the long struggle for the dominion of the Eastern
Mediterranean which was not to end till the fall of Acre in

I29I. 1

With the causes of the Crusades we are not concerned ; nor
are 1 their religious, social, or commercial aspects our province:
It is with their military side alone that we have to deal a

1 In a w-ay we might say that the last effects of the Crusades were not over till
the Turks evicted the Venetians from Cyprus (1571*, Crete (1669) and the Morea
(1715).

1096] STATE OF THE LEVANT IN 1096 231

subject sufficiently vast and varied to fill many volumes if
we had space to descend into detail.

Stated broadly, the problem which was started in 1096, and
lasted till 1291, was whether feudal Europe, with the military
customs and organisation whose development we have been
tracing, would prove strong enough to make a permanent
lodgment in the East, or perchance to make good the whole of
the ancient losses which Christendom had suffered at the hands
of the Saracen and Turk from the days of Heraclius to those of
Romanus Diogenes.

The state of the Moslem powers of the Levant in 1096 was
on the whole favourable for the assailants who were about to
throw themselves upon Syria and Asia Minor. .It had seemed
in the early days of the Turkish invasion, and soon after the
fatal day of Manzikert, that a single great empire might establish
itself in Western Asia under the house of Alp Arslan. But no
such result had followed the conquests of the Seljouks. At the
moment when the first Crusaders crossed the Bosphorus, the
Sultanate of Roum had separated itself from the main body of
the Turkish Empire, petty princes governed Aleppo, Antioch,
Damascus, and Mesopotamia, and the Fatimite sovereigns of
Egypt were still clinging to the southern parts of Palestine.
The political situation was most favourable for the assailants ;
a few years earlier they would have found their task far harder,
and the heroic courage which habitually saved them from the
consequences of their incredible lack of strategy and discipline
might have failed to accomplish the conquest of Western Syria.
Fighting against jealous and divided enemies, they only just
succeeded in conquering Jerusalem and Antioch. Opposed by
a single monarch wielding all the resources of Asia Minor and

o J^&gt;

the Levant, they would probably have failed on the threshold,
and never have seen the Taurus or the Orontes.

The first crusading armies displayed all the faults of the feudal
host in their highest development. They were led by no single
chief of a rank sufficient to command the obedience of his com
panions. Neither emperor nor king took the cross, and the crowd
of counts and dukes, vassals of different suzerains, had no single
leader to whom obedience was due. If a mediaeval king found
it a hard matter to rule his own feudal levies, and could, never
count on unquestioning obedience from his barons, what sort of
discipline or subordination could be expected from a host drawn

232 THE ART OF WAR IN THE MIDDLE AGES [1096

together from all the ends of Europe ? It is perhaps more
astonishing that the Crusaders accomplished anything, than
that they did not accomplish more than their actual achieve
ments. When we realise the nature of the numerous and unruly
council of war which directed the army that took Jerusalem, we
are only surprised that it did not meet with more disasters and
fewer successes. Yet this host was superior to most of the other
crusading expeditions in the efficiency of its fighting men, the
high character of its leaders, and the care that had been devoted
to its organisation. To understand the general aspect of the
crusading armies, we must remember all the unfortunate hordes
that perished obscurely in the uplands of Asia Minor and left
no trace behind.

CHAPTER II

THE GRAND STRATEGY OF THE CRUSADES

T OOKED at from the most general point of view, the
\_s Crusades, as a whole, may be said to have had two main
objects. The first was to relieve the pressure of the Turks on
Constantinople, which had been so dangerous ever since the day
of Manzikert. The second was to conquer the Holy Land and
restore its shrines to the custody of Christendom. Both of these
purposes were to a certain extent accomplished : the Turkish
frontier in Asia Minor was thrust back many scores of miles, and
nearly two centuries elapsed before the Seljouk Sultans were
able to recover their lost ground. Jerusalem was stormed, and
for ninety years remained in the hands of the Franks. But
these ends were achieved in the most wasteful manner, by the
most blundering methods, and at the maximum cost of life and
material.

One of the main causes of the disasters of all the crusading
armies was a complete lack of geographical knowledge. A
cursory glance at the itineraries of the various expeditions
shows that the majority of them were chosen on the most
unhappy principles, and were bound to lead those who adopted
them into grave peril, if not to utter destruction. \Ve must not
blame the men of the eleventh and twelfth centuries overmuch
for their errors : to a great extent they were inevitable in face
of their utter want of geographical information concerning the
countries of the Levant. Any misdirection was possible in days
when the whole available stock of information in the West con
sisted of garbled fragments of the ancient Roman geographers,
reinforced by a certain amount of oral information gathered
from merchants and pilgrims. The Franks could hardly be
expected to have any knowledge concerning the Eastern waters :
the Byzantines and Saracens had for many centuries divided

233

234 THE ART OF WAR IN THE MIDDLE AGES [1096

the control of the seas beyond Sicily, and the navies of the
Italian republics were but just beginning to trespass on them:
beyond Constantinople there was no accurate knowledge avail
able. The land routes were even more uncertain than those of
the sea. The road to the Bosphorus across Hungary and
Servia had only become practicable after the conversion of the
Magyars to Christianity ( 1000-6 1). 1 It had not yet been adopted
as a channel for commerce or a route for pilgrimages. Beyond
Constantinople there was only such information to be obtained
as the Greeks could give. This information was not always
honestly purveyed : the Byzantine emperors had purposes of
their own to serve, and often sent the pilgrim hosts on itineraries
which suited themselves rather than those which were best
adapted for the purposes which the Franks had in view. Yve
need not believe the constant complaints of the Western
chroniclers that the Comneni deliberately guided the pilgrims to
destruction, out of jealousy and treachery. But Alexius and
Manuel, if not John, were quite capable of serving their own
ends by despatching the invaders of Asia Minor on routes which
were not the best available. When the Crusaders had gone on
their way and beaten off the Turks, the emperor followed behind,
somewhat after the manner of the jackal, and seized what he
could. The recovery of Lydia and Mysia was undoubtedly due
to the first Crusade, and that of Northern Phrygia and Galatia
to the Crusade of 1 101.

It is only fair, however, to notice that in the case where de
liberate misdirection by the Greeks seems on the face of things
most probable, a deeper inquiry shows that the Crusaders them
selves were to blame. When, in 1101, Raymond of Toulouse
and the Lombards marched by the incredibly round-about
way of Ancyra-Gangra-Amasia, we might have suspected that
Alexius had recommended it to them in order that he might
follow in their rear and reoccupy Galatia, as indeed he did.
But both Raymond d Agiles on the side of the Franks, and
Anna Comnena on that of the Byzantines, assert that the un
happy choice was made by the Crusaders themselves. Anna
adds that her father pointed out to them the madness of their
attempt to march on Bagdad through the mountains of Armenia,
and that they utterly refused to listen to him. It was not his

] King Stephen placed Hungary under the papal supremacy in 1000. But the
last pagan rising was not put down till 1061, in the reign of King Bela I.

iioo]THE CRUSADFlRS 1 IGNORANCE OF GEOGRAPHY 235

fault if, after recovering Ancyra for the empire, they were starved
and harassed in the trackless lands beyond the Halys, so
that only a few thousands of them finally struggled back to
Sinope. It must also be remembered that the Byzantines them
selves, though they had all the old Roman road-books, and
elaborate data for the distances in their own lost " themes " in
Asia Minor, were not able to give accurate information concern
ing the present condition of the land. The Turks had wrought
so much damage in the last twenty years, burning towns, filling
up cisterns, and extirpating the population of whole districts,
that the old information concerning the interior had no longer
its full value. Routes easy and practicable before 1070 were
broken and desolate by 1097. The many perils which the
Comneni suffered in their own campaigns in inner Asia Minor
are sufficient proof that their information as to the land was no
longer reliable.

It would be unfair, therefore, to attribute to wilful misdirection
on the part of the Greeks the whole of the misadventures of the
Crusaders in Asia Minor. The larger part of their troubles were
of their own creation, and came from carelessness, presumption,
improvidence, and selfishness. Even when put upon the right
road, they were apt to go astray from blind conceit or want of
discipline. This comes out most clearly from the fact that
many crusading expeditions miscarried in Hungary or the
Slavonic lands just to the south of the Danube, before they ever
reached Constantinople. For an elaborate example of a wrong-
headed choice of routCj nothing can be more striking than that
which Raymond of Toulouse and the Provencals selected in
1096. In all South-Eastern Europe there is no district more
destitute of roads and more inhospitable than the Illyrian coast
line. But Raymond chose to march from I stria to Durazzo
through the stony valleys and pathless hills of Dalmatia,
Montenegro, and Northern Albania, among the wild Croats and
Morlachians. It is surprising that he was able to bring half his
following to Uurazzo: he must have failed altogether had not his
expedition been by far the best equipped and the most carefully
provisioned of all those which set out for the first Crusade.

For the pilgrimage to Syria there were two great alternatives
open the land voyage by Constantinople and the sea voyage
direct to the Levant. The latter was in every way preferable
when once the sea routes had been surveyed. But at the time

236 THE ART OF WAR IN THE MIDDLE AGES [1097

of the first Crusade it was practically unknown : only the
adventurous sailors of Venice, Pisa, and Genoa attempted it :
the French, Burgundians, Provencals, Germans, and Lombards
all preferred the longer road by Constantinople. Even in later
times the landsmen s horror of the water drove a majority of the
Crusaders to shun the voyage by sea : all the greater chiefs of
the second Crusade, and Frederic Barbarossa among the leaders
of the third, persisted in taking the land route. The first great
expeditions made by sea by any save the Italian powers were
those of Philip Augustus and Richard of England in 1190. But
from that time onward the advantages of the direct voyage to
Palestine seem to have been recognised, and all the later
Crusaders preferred it. It was obviously better to arrive fresh
and unwearied at Acre or Tyre, rather than to run the thousand
risks from Hungarian, Greek, and Turk which threatened all
who marched by land.

(A) THE LAND ROUTES THROUGH ASIA MINOR.

Since, however, the majority of the early Crusaders were
unaware of the superiority of the sea route, and chose to make
Constantinople their basis for the march on Jerusalem, we must
begin by pointing out the strategical aspects of their under
taking. In 1097 almost the whole of Asia Minor was in the
hands of the Seljouks: the Emperor Alexius held little more
than Chalcedon, Nicomedia, the Mysian coast-region, and a few
isolated towns on the Black Sea, like Sinope and Trebizond.
The Turks were established on the Sea of Marmora : they had
chosen Nicaea, only twenty-five miles from its shore, as their
capital. All the inland plateau of Asia Minor was in their
hands, and all the coast-line also, save the few Byzantine sea
ports and a patch or two in Cilicia, where Armenian mountain-
chiefs maintained a precarious independence.

If Alexius Comnenus had been able to direct the crusading
army at his own good pleasure, he would have used it to clear
Bithynia, Lydia, and Phrygia of the Seljouks. If the Franks,
on the other hand, had been entirely their own masters, they
would have marched straight across Asia Minor to the Cilician
gates, and made Antioch their first halting-place. But since
neither party could disregard the wishes of the other, a kind of
compromise was concluded : the Crusaders took Nicsea for
Alexius, and then went on their way. The reduction of the

197] THE ROUTES OF ASIA MINOR 237

jrkish capital was of inestimable advantage to the emperor :
Constantinople could breathe freely when the Seljouks were dis
lodged from the stronghold almost in sight of its walls which
they had been holding for the last fifteen years. With this
Alexius had to be content for the present. Murmuring bitterly
that they had been restrained from plundering and occupying the
city, the Crusaders moved forward into Phrygia. The route
across Asia Minor which they adopted was, except in some small
details, the right one. Their successors in later years would have
been wise if they had always adhered to it.

The great peninsula consists of a high central plateau sur
rounded by a number of small coast-plains. For those who wish
to march from west to east there is no good road either along
the Euxine shore or the shore of the Sea of Cyprus. On the
north the mountains of Paphlagonia and Pontus, on the south
those of Lycia and Isauria, come down to the water s edge at
many points, and cut the practicable route in so many places,
that it is for all intents and purposes impassable for an army.
No traveller in his senses would attempt to use the coast-roads.
The inland roads, one of which he must choose, are practically
three in number. Two of them suit those who start from Nicaea,
the third those whose base is Sardis, Miletus, or Ephesus. This
last was not available for the Crusaders of 1097 ; they had no
wish to make the long detour along the ^Egean, through Mysia
and Lydia, which would have brought them to Sardis or any of
the other suitable starting-points for the march to Philadelphia-
Philomelium-Iconium-Tarsus. There remained for their choice
the two other routes, one of which passes north, one south, of the
great Salt Lake of Tatta (the Tuz Gol of the Turks) and the little-
known region of the Axylon l which lies around it. The southern
route is that which they chose : it runs by Dorylseum, Philo-
melium, Iconium, and Heraclea-Cybistra to the Cilician gates. 2
The northern and the longer way leads to the same pass by

1 Mr. Hogarth informs me that the Axylon does not deserve its well-known
reputation for barrenness and desolation.

2 Why Godfrey of Bouillon and the larger half of the crusading host diverged from
the obvious route by Heraclea, the Cilician gates, and Tarsus, and only sent Baldwin
and Tancred upon it, it is hard to discover. But they undoubtedly took the extra
ordinary and circuitous road by Nigdeh, Caesarea-Mazaca, Coxon (Cucusus-Goeksun),
and Marash, and suffered severely from privations in the Anti- Taurus while crossing
the Doloman Dagh, between Coxon and Marash. Probably they were attracted by the
friendly Armenian population of Eastern Cappadocia.

238 THE ART OF WAR IN THE MIDDLE AGES [1097

Tatiaeum, Ancyra, Caesarea-Mazaca, and Tyana, Both were
good Roman roads, and had been kept in order by the Byzantines
down to the disastrous year 1071. Now, however, the land lay
desolate : bridges were broken, cisterns empty, and for many
stages the whole population had been slain or driven off by the
Seljouks. There were no insuperable natural obstacles on either
road : the two perils to the Crusaders were starvation and the
chance of being wearied out and brought to a stand from ex
haustion by the incessant attacks of the Turkish horse-archers.
More fortunate than any of their successors, the hosts of Godfrey
and Bohemund opened their march by inflicting a crushing defeat
on the enemy, who was so utterly awed that he held off, and did
not venture to harass the marching columns for many weeks.
They moved by Philomelium, Antioch-in-Pisidia, and Iconium,
with no let or hindrance. It was not till they reached Heraclea-
Cybistra that they again met the Turks in arms, and then they de
feated them with ease. Though unmolested by the Seljouks, the
Franks suffered dreadfully from want of stores and forage. This
was unavoidable in a desolate land, for the Western armies of that
age had no proper conception of commissariat arrangements :
they depended mainly on the districts they passed through ; and
if the countryside was barren, they were bound to suffer. The
trouble was made far worse by the long and useless train of
non-combatants of both sexes which the crusading host dragged
behind it. If they had endured many privations in Christian
regions like Hungary and Bulgaria, it was obvious that the
passage through Asia Minor was bound to be accompanied by
terrible loss of life. Nevertheless, the greater part of the host
struggled through, some to Marash, others to Tarsus, where they
could rest and recruit themselves for a space among the friendly
Armenian population of Cilicia.

On the whole, therefore, the passage of the first Crusaders
through Asia Minor may be described as fairly successful when
their difficulties are taken into consideration. Far otherwise was
it with their successors of 1 101. The miscellaneous bands under
Sweyn the Norseman, Archbishop Anselm of Milan, William of
Poictiers, Stephen of Blois, and Eudes of Burgundy, all fared far
worse. Some were wholly destroyed, others were turned back
with the loss of nine-tenths of their numbers ; of the remainder a
few stragglers only succeeded in pushing their way to Tarsus and
Antioch. The causes of their disasters are sufficiently obvious :

uoi] THE CRUSADE OF noi 239

they showed even less discipline than their predecessors, and
they had formed a wholly erroneous conception of the easiness of
their task from the comparative immunity enjoyed by Godfrey
and Bohemund s army during its passage. They were so purled
up with the idea of their own invincibility that they declared
their intention of " crossing the mountains of Paphlagonia and
forcing their way into Khorassan, in order to besiege and take
Bagdad." l It was in pursuit of this mad design that the majority
of their host started off on the route Ancyra-Gangra- Amasia,
which, if they had been able to pursue it to the end, could only
have stranded them in the mountains of Armenia. After a
terrible march among the highlands of Fontus, 2 where the foot-
soldiery died by thousands of weariness and starvation, and the
cavalry were almost entirely dismounted, the Lombards and Pro
vencals were brought to a standstill by the army of Mohammed
ibn Danishmend, Emir of Cappadocia, whose light troops hovered
around them day after day, cutting off their stragglers and for
aying parties. When the Turks thought the Crusaders sufficiently-
exhausted to fall an easy prey, they offered them battle at a
place named Maresh (or Marsivanj, somewhere in the neighbour
hood of Amasia. The combat was indecisive, but on the follow
ing night Raymond of Toulouse, the man of greatest note in the
host, fled away by stealth and deserted his comrades. Others
hasted to follow his example, and, in the disorderly retreat which
then set in, Danishmend cut the whole army to pieces, with the
exception of a few thousands who succeeded in distancing their
pursuers and finding shelter in the Greek fortress of Sinope.

Meanwhile, the smaller division of this band of Crusaders,
who had refused to take the unwise route along the northern
edge of the plateau of Asia Minor, had been reinforced by
William Count of Nevers and a large band of French pilgrims.
They then marched fifteen thousand strong 3 by the long but not
irrational line of Ancyra-Iconium-Heraclea. All the way
from Iconium to Heraclea they were encompassed by the hordes
of Danishmend and Kilidj-Arslan, fresh from their victory over
the Lombards at Maresh. Harassed incessantly, day and night,

1 Albert of Aix, viii. p. 7. Cf. the identical statement in Anna Comnena,
book xi. 8.

- \Ve get from Anna only the fact that they had crossed the Halys ; the Frankish
chroniclers thought they were still in " Flagania," i.e. Paphlagonia.

3 Albert of Aix, viii. p. 29.

240 THE ART OF WAR IN THE MIDDLE AGES [nor

by the enemy, and suffering horribly from thirst, they were
reduced to the most pitiable condition when they reached Heraclea
and had the passes of the Taurus in sight. Then the Turks,
fearing that their prey was about to escape them, closed in and
offered battle. In a long straggling fight between the city and
the foot of the Taurus the Christian army was gradually broken
up and shot down in detail. Seven hundred knights, who at last
abandoned their unhappy foot-soldiery l and took to the hills,
got off in safety over one of the minor passes of the Taurus, and
reached Germanicopolis in Cilicia, where they took shelter with
the Byzantine garrison. William of Nevers himsel f finally reached
the same spot with only six companions. The rest of the fifteen
thousand Franks had been slain ; the Parthian tactics of the
Turks had not been frustrated by any such happy chance as that
which saved Bohemund and Robert of Normandy at Dorylaeum. 2
A very similar fate befell a large body of Aquitanian
Crusaders, led by their duke, William of Poictiers, who had started
shortly after the departure of the Count of Nevers from Constanti
nople. This host, a much larger one than either of those which
preceded it, followed the same route as Godfrey and Bohemund
had taken four years before. They had little trouble from the
Turks till they reached Iconium, and were successful in taking
and pillaging the towns of Philomelium and Salabria. 3 But at
Iconium their provisions gave out, and they learned of the
destruction of the army of the Count of Nevers. Nevertheless,
they resolved to press forward, and soon found themselves beset
by Kilidj-Arslan and Danishmend. Their immunity from attack
hitherto had only been secured by the fact that the division of
Nevers was eight days ahead of them, and had attracted all the at
tention of the Seljouks. The fifty-five miles between Iconium and
Heraclea proved as fatal to the Aquitanians as it had been to
their predecessors. The want of water was their ruin, 4 and when
they approached the river near Heraclea they broke their order
and pushed forward without any thought save that of slaking
their thirst. Some were across the stream, some on its banks,
some still straggling up from the rear, when the Turks closed in

1 Albert of Aix, viii. 30.

2 See the account of this battle on pp. 271-274.

3 This place, not far from the great Tuz Gol lake, must have been taken by an
expedition sent out from Iconium, as it does not lie on the itinerary Nicsea-Iconium.

4 Robert the Monk, book iii., tells us how Godfrey of Bouillon avoided this danger
by taking water with him.

] DISASTERS IN ASIA MINOR 241

from all sides and began pouring in their arrows. The Crusaders
were too scattered to form a line of battle or oppose any regular
resistance. After a certain amount of righting, those who were
not utterly surrounded, or who could cut their way through the
enemy, turned their faces towards the Taurus, and fled as best
they might. Most of the leaders and a certain number of the
mounted men were able to reach the hills, and straggled into
Tarsus in small parties. The wretched infantry, as was always
the case in these unhappy battles of 1 101, were wholly destroyed.

When the wrecks of the hosts of the Lombards, the Count of
Nevers, and William of Poictiers, had finally gathered themselves
together at Antioch in the spring of 1102, they only amounted
to ten thousand men. This small force marched along the
Syrian coast and took Tortosa. No other profit came to
Christendom from the waste of three armies, which are said to
have amounted at their setting forth to more than two hundred
thousand men. Their failure, as it is easy to see, came from
three causes : in the case of the Lombards from an impossible
itinerary ; in that of the Counts of Nevers and Poictiers from their
absolute ignorance of Turkish methods of warfare and their
insufficient supply of provisions and water. The route taken by
the two counts was the best available, and no blame can be
laid upon the chiefs for adopting it. But they were almost doomed
to failure from the first by the number of useless mouths which
they took with them. A heavy train and a multitude of non-
combatants made the army slow, when speed was necessary to
prevent the food running out and to cross the many waterless
tracts. Even, however, if the provisions had held out, and the
armies had been in fair fighting trim, it is doubtful whether they
would have succeeded in discomfiting the Seljouks. None of
the leaders had the least notion of the proper method of resisting
the Turkish tactics. They had no idea of using infantry and
cavalry in combination, and wished to do all the work with
their mounted men alone. Hence they were bound to fail : only
a steady infantry largely armed with missile weapons could
have saved them, and such a force they did not possess.

We have still to consider three more great expeditions across
Asia Minor those of Louis of France and the Emperor Conrad
in 1148-49, and that of Frederic Barbarossa in 1 190.

Between the opening of the twelfth century and the second
Crusade the political geography of Asia Minor had been pro-
16

242 THE ART OF WAR IN THE MIDDLE AGES [1148

fotindly modified by the conquests of the Comneni. Profiting by
the blows which the Crusaders had dealt the Seljouks, Alexius
and John II. had thrust forward their frontier far inland, and
reoccupied the western third of Asia Minor. . Their line of posts
ran far into Phrygia, passing by Dorylseum, Philadelphia, and
Laodicea. They had also recovered the whole southern coast of
the peninsula, as far as Cilicia. The Sultans of Roum, thus
pressed back into the interior, had made Iconium their capital
instead of the lost Nicasa. It was just possible to march from
Constantinople to Tarsus without leaving Christian soil, though
to use such a route entailed an intolerably long itinerary. A
chronicler of the second Crusade thus describes the situation,
showing a geographical knowledge very unusual in his class: 1
" From the Bosphorus [or the Arm of St. George, as it was then
called] there are three roads to Antioch, unequal in length and dis
similar in their merit. The left-hand road is the shortest : if there
were no obstacles in the way, it would take no more than three
weeks. After twelve marches it passes by Iconium, the Sultan s
residence, and five days after that it enters Cilicia, a Christian
land. A strong arm} , fortified by the faith and confident in its
numbers, might despise its obstacles ; but in winter the snows
which cover the mountains are very terrible." This is the old
route of the first Crusaders by Dorylseum, Iconium, Heraclea, and
the Cilician gates. " Secondly, there is the road most to the
right, which is better in some ways, as supplies are to be had
all along it. But those who use it are delayed by two things the
long gulfs cutting up into the coast-line, and the innumerable
rivers and torrents to be crossed, all dangerous in winter, and as
bad as the Turks and the snows on the first route." By this road
Odo means the long, circuitous passage by Pergamus to Ephesus,
and thence along the Carian, Lycian, Pamphylian, and Isaurian
coasts to Seleucia. " The middle road, "continues our chronicler,
" has less advantages and also less drawbacks than either of the
other two. It is longer and safer than the first, and shorter but
poorer and less safe than the second." The middle route of
Odo is the line by Pergamus, Philadelphia, Laodicea, Cibyra,
Attalia, and thence by the Cilician coast, to which Louis VII.
and the French Crusaders committed themselves in the winter
of 1148-49. The Emperor Conrad and the Germans took the
" left-hand road," i.e. the short and dangerous line through the

1 Odo of Deuil, book v.

1148! DEFEAT OF THE EMPEROR CONRAD 243

midst of the Turkish territory, which passes by the gates of
Iconium.

The fates of the t\vo expeditions were not xvholly dissimilar,
though the Germans fared much worse than the French. Both
failed more by their own mistakes than by the difficulties which
lay in their way. Conrad started from Nicsa, with guides lent
him by the Emperor Manuel Comnenus. He only took with
him supplies for eight days, a wholly inadequate provision when
we reflect that he had much more than two hundred miles to
cover, and that he was forced to accommodate his pace to that of
his baggage train. The Turks allowed him to advance into the
heart of Phrygia without resistance ; but when he was somewhere
near Philomelium, and was still some seventy or eighty miles from
Iconium, his food-stores were completely exhausted. His army
was involved in the spurs of the Sultan Dagh, which cut across
the road at this point : seeing themselves starving and in a
desolate and difficult country, the Germans accused their guides
of treachery. When threatened, the Greeks absconded, and
apparently fled to the Turkish Sultan. Hearing of the bad
state of Conrad s army, Masoud at once determined to close in
and attack them. Then began one of those long running fights
such as had ruined the pilgrim hosts of 1101 a stage or two
farther to the east. The Germans, in spite of all the warnings
of previous Crusades, had no provision of crossbowmen l to keep
off the Turks, while their cavalry had so suffered for want of
forage that those knights who still bestrode horses could hardly
spur them to a trot. Conrad determined to turn back, and was
pursued for many scores of miles by the Seljouks, who regularly
cut off the devoted rearguards which he detached to cover his
retreat, and gleaned thousands of starving stragglers every day.
At last the harassed Germans reached Nicaea, and could once
more obtain provisions ; but their past sufferings had been so
great that thirty thousand men are said to have died of dysentery,
cold, and exhaustion after reaching the shores of the Propontis. 2
As a military machine the army was ruined ; the greater part
of the survivors drifted back to Germany, and the emperor took
only a few thousand men by sea to Palestine out of the seventy
thousand who had set out with him.

Louis of France, seeing that the greater part of Conrad s

1 This is especially remarked upon by Odo of Deuil, book v. p. 343.

2 Odo of Dexiil, book v. p. 347.

THE ART OF WAR IN THE MIDDLE AGES [1149

disasters had come from want of food and forage, was confirmed
in his design of keeping as far as possible within the borders of
the Byzantine Empire, where supplies would be procurable.
Accordingly, he marched through Mysia and Lydia by Prusias
(Broussa), Pergamus, Smyrna, and Ephesus. He kept his
Christmas feast in the valley of the Cayster, a few miles from
Ephesus, and then proceeded to move up the Maeander towards
Laodicea. His cautious route had hitherto kept his army free
from all trouble, and, as he was still within Byzantine territory,
he reckoned on a quiet march. But the Turks, hearing of his
advance, had resolved to cross the border and attack him. Near
Antioch-on-Masander they opposed the advance of the French
as they were fording the river, and at the same time attacked
them in flank and rear. But Louis troops were fresh and in
good order, and a vigorous charge of the French knights swept
the Seljouks away ; they gave no trouble for some days, so that
the army arrived safely at Laodicea, the border town of the
Byzantine Empire. Here their troubles began. Louis had pro
posed to fill up his stores at Laodicea before beginning the
difficult march through the mountains of Pisidia to Attalia.
This region, full of small towns in the old Roman days, had
been harried bare by the Seljouks. There was hardly an
inhabited village on the route, which turned out to be no less
than fifteen days in length, though the French had calculated
on taking a much shorter time to traverse it. But the governor
of Laodicea refused to sell any provisions to the Crusaders
from treachery, according to the French chroniclers, but more
probably because he dared not exhaust his stores when the
Turks were known to be in the immediate neighbourhood.

It was accordingly with a very insufficient stock of food that
the French marched past Laodicea and started on their way by
the pass between the Baba Dagh and the Khonas Dagh which
leads up into the highlands. On the second day after leaving
Laodicea their disasters began. The army was marching with a
proper advance guard and rearguard, the baggage and non-com
batants in the centre. The whole occupied many miles of route.
At the difficult pass of Kazik-Bel (three thousand eight hundred
feet above the sea level), the van, under Geoffrey de Rancogne and
Amadeus Count of Maurienne, the king s uncle, was ordered to
seize and hold the exits of the defile till the whole army had
passed. But, preferring to spend the day comfortably in the plain

1 149] LOUIS VII. IN ASIA MINOR 245

t&lt;
Themisonium (Kara - Eyuk - Bazar), the commanders of the
advance guard descended from the heights and pushed on
several miles to encamp in the valley. The Turks had been
hiding near the mouth of the defile, and, when Geoffrey and
Amadeus had passed on, burst out upon the unprotected train
of beasts of burden and unarmed pilgrims who were struggling
through the pass. Shooting down from the more elevated points
on the helpless crowd, they wrought great slaughter, and pre
cipitated many into the ravine which winds at the bottom of
the pass. The king hurried up from the rear with a small body
of his retainers, but, since he had not his crossbowmen with
him, 1 he could make no reply to the arrow-shower from above.
Presently the Turks came down upon the confused mass and
attacked them at close quarters. Louis himself had to fight for
some time alone, with his back against a rock, and owed his life
to his swordsmanship. At last the tardy return of the advance
guard took off some of the pressure, and when night fell the
Turks drew off, and the whole of the French armament struggled
down into the plain. They had lost most of their stores, thousands
of horses, a great part of the unfortunate non-combatant pilgrims,
and not a few knights of note.

It was generally agreed that the blame of the disaster rested
upon the careless commanders of the van, and Geoffrey of
Rancogne would have been hung but for the fact that Count
Amadeus, who shared his responsibility, was the king s uncle.
When the host was reassembled, Louis, with a prudence and
self-restraint seldom shown by the crusading chiefs, declared
that he would hand over the future conduct of the march to
experienced hands. The Grand Master of the Templars,
Everard des Barres, accompanied the host, and many veteran
knights of the Order with him. The king consigned to them
the regulation of the army, and a certain Templar named Gilbert
marshalled it for the rest of the way to Attalia. They moved
for the remaining twelve days of the march with a vanguard of
mounted men, and rearguard of bowmen, strengthened by all
the knights who had lost their horses. So successful was the
new commander that four attacks of the Turks were beaten off
with ease and considerable slaughter of the infidels. Even at
the difficult passage of the two branches of the Indus (near
Cibyra) the army suffered no harm, for Gilbert had the Turks

1 Odo of Deuil, book vi. p. 363.

2 4 6 THE ART OF WAR IN THE MIDDLE AGES [1149

driven away from the strong positions flanking the ford before
he would allow the army to cross.

But if the enemy did little harm with his arrows, the want
of forage for the horses, and the gradual exhaustion of the in
sufficient stores which remained for the men, ruined the efficiency
of the army. For the last week of their march the French were
living almost entirely on horseflesh, and a few days more would
have reduced them to absolute starvation. On arriving at
Attalia, the king held a council of war and abandoned his
intention of proceeding any farther by land. It was, as men
said, forty days march to Antioch if they followed the Cilician
shore, and all through difficult roads like those they had already
passed over. On the other hand, it was but three days by sea
to Syria if the wind was fair. So, hiring ships from the Greeks,
the king and his knights and nobles passed over to Antioch,
The winds, as it chanced, were contrary, and the voyage took
three weeks instead of three days, but all reached their goal in
safety. It was otherwise with the unhappy infantry; there had
not been ships enough to take more than a small proportion of
them, and they remained behind for months under the walls of
Attalia, starving after they had spent their last deniers in
buying food from the Greeks at very exorbitant rates. At last
some eight thousand of them, headed by a few knights, resolved
that anything was better than longer waiting, and started off by
the coast road to cut their way to Tarsus. They forced the
passage of the Cestrus, but the Eurymedon, the next river along
the coast, proved unfordable, and on its banks they were attacked
and cut to pieces by the Turks. Of the survivors some entered
the Greek service, others turned Moslems in despair, "for the
Turks, cruel in their kindness, gave them bread and took from
them the true faith " ; the majority, however, died of disease or
famine in the neighbourhood of Attalia.

It might have been thought that the fate of the armies of
Conrad and Louis would have finally demonstrated that the land
route to Syria was inferior to that by sea. Yet one more great
expedition passed over the central plateau of Asia Minor, and
(unlike its predecessors ever since 1 101) succeeded in reaching
its goal. This army, however, was commanded by an experi
enced soldier, and adopted all the precautions which had been
neglected by the ordinary crusading hosts ; yet even Frederic
Barbarossa nearly failed from the force of hunger, though he

U9i] FREDERIC BARBAROSSA S CRUSADE 247

beat the Turkish hosts in every encounter. The great emperor
took in the first half of his march (March-April 1 190) a route
not very unlike that which had been followed by Louis vil.,
keeping well inside the Byzantine border in Mysia and Lydia.
He passed by Philadelphia and Tripoli s into the valley of the
Maeander, and reached Laodicea. But from this point he did
not turn south like the French king, but set his face due east,
and moved by the great Roman road which passed by Apamea
and the Pisidian Antioch to Iconium. This was the main artery
of the communications of the central plateau, and it is curious to
find that no other crusading army had tried it. The Turks
closed round Frederic and attacked him at the sources of the
Ma.-ander, near Apamea, but were beaten off with great loss
(April 30). They returned to the charge in the passes of the
Borlu Dagh, near Sozopolis, but only to receive a second check
(Ma} r 2). By this time, however, famine, the most trusty ally of
the Turks, was beginning to make itself felt in the German host,
and the horses were dying in large numbers from lack of forage
the enemy having burned the grass in all directions. On
reaching the lake of Egirdir the stores were running so low that
Frederic resolved to quit the direct but desolate route to
Iconium by Carallis, " the royal road on which the Emperor
Manuel Comnenus had been wont to march." l Swerving from
it, he crossed the Sultan Dagh by a difficult bridle path, and
came down into the fertile plain of Philomelium thus falling
into the route which the first Crusaders under Godfrey and
Bohemund had taken. The Germans found some resources
here, but had at once to fight for their lives the Turkish
armies, no longer pent up in the hills, were operating in one
of the great rolling plains, which best suited their tactics of
circumventing the enemy. For twelve days, from the 4th to
the 1 6th of May, the army was slowly forcing its way over the
seventy-five miles which separate Philomelium from Iconium.
They had to march in order of battle, with a front in every
direction and the impedimenta in the centre. The rear, the
point of greatest danger, was brought on by the Dukes of Suabia
and Meran and the Margrave of Baden, with a great force of
archers and a bod}- of dismounted knights. There was always
danger lest the rear, facing about to defend itself from an attack,
should get separated from the main body, and so the Turks

1 See the Efistola de Mortc Frederici, p. 346.

24 8 THE ART OF WAR IN THE MIDDLE AGES [1191

might slip in between. On one occasion this did occur, and a
vast amount of baggage was lost. The knights themselves
suffered little ; " many were wounded, but few slain," for their
coats of mail effectively kept out the Turkish arrows. But their
horses, not yet armed in steel like those of later times, suffered
terribly. By the 1 3th of May there were only six hundred
effective chargers left, and the majority of the knights were
serving on foot. Nevertheless, the Seljouks were always beaten
off. Twice they ventured to close in, on May 6 and May 13,
and on each occasion they were well punished for their audacity :
in the first fight three hundred and seventy -four chiefs and
emirs and six thousand horsemen fell before the weapons of the
Germans. On May 16 the army reached Iconium, wearied and
almost starving; there it got food and plunder from the summer
palaces of the Sultan outside the walls. After resting them
selves for a day, part of the host made a front against the
Turks, while the remainder stormed the town with unexpected
ease, and obtained such an ample store of food that the danger
of starvation was at an end. " The place was as big as
Cologne," and full of all manner of riches, which the Germans
plundered at their leisure for five days. The Sultan Kilidj-
Arslan 1 was now brought to such a depth of discouragement
that he began to treat with the emperor. He promised the
Germans a free road to Cilicia if they would depart at once,
and gave twenty of his chief emirs as hostages. This was better
fortune than any crusading army had experienced before, and
the emperor accepted the terms. He marched, not by the usual
route of Heraclea and the Cilician Gates, but by Laranda,
Karaman and the pass which leads to Seleucia-by-the-sea.
Here the army arrived, without having suffered any further
molestation, save from an earthquake which inspired it with
great fear. On the very day of his arrival at Seleucia, Frederic
Barbarossa was, by the most unlucky of chances, drowned while
bathing in the Calycadnus (June 10, 1190). His army, deprived
of its leader, but now safe, " after six weeks of constant march
ing and starving," 2 took its way through Christian territory to
Antioch, where it arrived in safety.

Having now surveyed all the Christian invasions of Asia
Minor, we can legitimately draw our general conclusions as to
their characteristics.

1 Not Malek Shah. See Boha-ed-din, p. 272. - Ep. de Morte Frederici, 350.

M9i] THE CAUSES OF DISASTER 249

Our first deduction must be couched in the form of a testi
monial to the very efficacious nature of the Seljouk methods of
warfare. The Turks had deliberately established a broad belt
of wasted and uninhabited territory between themselves and the
Byzantine border. Moreover, when a Christian army passed
through their dominions, they did not hesitate to destroy their
own crops and sacrifice their villages. The cattle were driven
into the hills, the corn burned, the very grass in the valleys fired.
Consequently, every crusading host which crossed Asia Minor
suffered horribly from famine. Of all the causes of failure this
was the most obvious.

A thoroughly disciplined regular army, with an organised
waggon-train, could no doubt have triumphed over this system
by bearing its own food with it. But the Franks were a mixed
multitude, with little or no organisation, always clogged in their
progress by the hordes of non-combatants, largely paupers, whom
they dragged with them. Against such foes the Turkish system
was most efficacious. We may, indeed, express our wonder that
Godfrey and Frederic Barbarossa struggled through in spite of
all opposition. That the Crusaders of 1101 and 1148 failed is
less a matter of surprise.

The second among the main causes of the disasters of the
crusading armies was that ignorance of geography on which we
have already had to dilate. When men could dream of finding
their way to Bagdad and Khorassan through Paphlagonia and
Pontus, or deliberately consider the advisability of adopting the
route from Constantinople to Tarsus by the Carian, Lycian, and
Cilician coast-line, they might meet with any kind of disappoint
ment. Concerning this topic we need not enlarge the history
of the individual expeditions forms a sufficient commentary
on it. We need only add that over and above mere want of
geographical knowledge we must allow for the effect of minor
ignorances that, for example, of climate. The extreme heat and
cold of the plateau of Asia Minor in summer and winter respect
ively was a fact for which the Crusaders made no allowance.
What could have been more mad than for Louis VII. to choose
the months of January and February for his excursion through
the Pisidian mountains? The torrents were at their full, the
winter rains were destructive of stores and tents, and the snow
was lying on the higher slopes of the hills.

Third among the causes of the failures of the Crusaders we

250 THE ART OF WAR IN THE MIDDLE AGES [1099

must place their own want of providence, discipline, and self-
control. Even the best-behaved of their armies were, by the con
fession of their own chroniclers, terribly addicted to riot and
plunder. Their interminable quarrels with the Greeks mostly
arose from their own fault. That there existed a very considerable
jealousy and ill-will on the part of Byzantines no one can dispute,
but the conduct of the pilgrims was so bad that we cannot wonder
at the resentment they provoked. Their want of discipline was as
well marked as their proneness to plunder : deliberate disobedience
on the part of officers was as common as carelessness and
recklessness on the part of the rank and file. This was always
the case in feudal armies : in the East the fault was seen even
more clearly than elsewhere. Most notable of all is the evident
inability of the Franks to learn from the unhappy experiences
of their predecessors. The thousands of veterans who drifted
back from the East did not succeed in teaching their successors
to observe the precautions appropriate to Turkish warfare. Fifty
years after the first Crusade, Conrad in. and Louis vil. com
mitted exactly the same mistakes as the contemporaries of
Godfrey and Bohemund. They marched without caution ; they
did not properly combine infantry and cavalry ; they had not
provided themselves with the necessary proportion of men armed
with missile weapons such as the bow and arbalest ; their stock
of food was always running short. It seemed that the art of
learning by experience hardly existed in the military circles of
the West. The description of the faults of the Frank as a soldier
which Maurice wrote in 580, and Leo the Wise repeated in 900,
might still be utilised almost word for word in describing the
Crusaders of 1 150.

(B, THE STRATEGY OF THE CONQUEST OF SYRIA.

The primary impulse of the men of the first Crusade was
religious, not strategic. Their end was to recover Jerusalem, not
to establish a sound military base for the ultimate conquest of
the whole of Syria. There were those among the Prankish
leaders who saw that it was dangerous to march from Antioch to
Jerusalem, leaving hostile towns to right and left, and sacrificing
the connection with their only base ; but they were overruled by
the majority, whose ruling desire was to get possession of the
Holy Places. We must not, therefore, criticise the campaign of
1099 as if it had been carried out on logical military lines.

THE GEOGRAPHY OF SYRIA 251

It was only when Jerusalem had fallen, and the Crusaders had
determined to establish a permanent feudal state in Palestine,
that strategical considerations came to the front.

When Godfrey was crowned, the new kingdom consisted of
nothing more than the towns of Jerusalem and Jaffa. Whether
Bohemund, isolated at Antioch, and Baldwin in his distant county
of Edessa, would ever truly become the vassals of their theo
retical suzerain was most uncertain. The future of the Franks
in Syria was not settled for many years : indeed it was not till
about 1125 that any general conclusions as to the new states
could be formulated.

Before passing on to consider the military history of the
conquest, it is necessary to understand the general strategical
aspect of Syria. It may be divided into four narrow zones
running from south to north, one behind the other. The first of
these the shore consists of a series of coast-plains of very
vaiying size and width ; they are cut off from each other by
mountains running down to the water s edge, like Carmel, the
spurs of Lebanon, and the " Black Mountains " by Antioch.
Most of these level coast-tracts are narrow, but the southmost of
them, the celebrated plain of Sharon, is larger than the rest, and
averages fifteen miles in breadth. Occasionally, too, the coast-
plain runs inland up a river valley, as in the plain of Esdraelon
just north of Carmel, and in the valley of the Orontes near
Antioch. In the central districts of the Syrian shore, however,
about Tripoli and Beyrout, it is exceptionally narrow and much
broken up.

The second zone of territory comprises the mountainous
upland overhanging the coast-plain. This region consists of the
spurs of three main chains the mountains of the Ansariyeh (the
C asius of the ancients) in the north, Lebanon in the centre, and
the mountains of Ephraim and Juc!a;a in the south. The two
former are lofty ranges rising at some points to eleven thousand
feet above the sea level ; the last has a broader and less well-
defined crest, and seldom rises to a greater height than three
thousand feet. The spurs and shoulders of all these chains con
tain many fertile and populous tracts.

The third zone consists of the deep-sunk valleys of three
great rivers the Orontes, Leontes ^Litanyj,and Jordan. The two
former find their way to the sea the first by a gap between the
mountains of the Ansariyeh and the Black Mountains ( Ahmar

252 THE ART OF WAR IN THE MIDDLE AGES [1099

Dagh), the second by a much narrower defile just north of Tyre.
But the Jordan, whose course is mostly below the level of the
Mediterranean, falls into the Dead Sea, a sheet of water with no
exit. The Orontes and Leontes have broad and fertile valleys,
while that of the Jordan is a narrow, precipitous, and marshy
defile, only to be crossed at a limited number of points. The
deep depression through Central Syria formed by these three
streams and by the Dead Sea is continued yet farther south by
the gorge of the Wady-el-Arabah, which runs down to the eastern
head of the Red Sea, and to the port of Elath or Akabah.

Beyond the valley " hollow Syria," as the ancients called it-
is the high-lying eastern plateau, in some places flat, in others
mountainous. It runs into the Great Desert, and is itself
barren in many parts. But it contains man} fruitful and well-
watered districts, such as those around the great cities of
Aleppo and Damascus.

Syria as a whole is eminently defensible : the sea and desert
cover it on three sides the west, east, and south ; on the north
the Amanus and the Euphrates give an excellent and well-
marked frontier. But the Crusaders never got possession of the
whole country : they only held the coast, the greater part of the
mountain, and certain regions of the central valley. The larger
half of the latter and the whole of the eastern plateau remained
unconquered. It was for this reason that the kingdom of
Jerusalem was always in a precarious position. A chain of
Mohammedan states always shut it out from expanding to the
eastward and reaching its natural boundary.

The cause of this anomaly is not hard to find. The crusad
ing states were never really strong enough to complete the con
quest of Syria : they would not even have succeeded in subduing
the whole of the coast if they had been forced to rely on their
own resources and could have counted on no external aid. But
the great Italian republics were deeply interested in the conquest
of the Syrian shore. It was of high importance to their
commerce that the whole of the ports of the Levant should be
in Christian hands. Hence they co-operated with the greatest
zeal in the sieges of the coast-cities : they and not the kings of
Jerusalem were really the conquerors of the whole coast-plain.
The Venetians were the real captors of Sidon (i 1 10) l and Tyre
(1124). The Pisans gave assistance to the Prince of Antioch at

1 Largely aided by King Sigurd of Norway on this occasion.

25] WEAKNESS OF THE CRUSADING STATES 253

Laodicea (1103) and to Count Bertram at Tripoli (1109); they
were also present at the siege of Beyrout (i 1 10). The Genoese
were still more energetic : to them were due the falls of Caesarea
(1101), Tortosa (1102), Acre (1104), Giblet (1109), Beyrout
(i 1 10). Casual aid was often given to the kings of Jerusalem by
other crusading fleets, such as those of the Englishmen Harding
and Godric, and the Norse king, Sigurd the Jerusalem-farer
(1109-10). But it was mainly by the aid of the Italians that
the Syrian coast became Christian.

Inland, the aid of these all-powerful allies was not available.
Their interests did not bid them equip armies to conquer
Damascus or Aleppo. Hence it was with their own weak feudal
levies alone, aided by occasional hosts of Western pilgrims, that
the kings of Jerusalem and princes of Antioch carried on their
wars with the emirs of the inland. The military resources of
the Prankish states were more than modest : the largest army
that they ever put into the field was one of thirteen hundred
knights and fifteen thousand foot, 1 a number only obtained by
collecting every available man and leaving the towns and castles
almost ungarrisoned. Larger numbers were of course assembled
when a crusading host from the West was present ; but the help
of the pilgrims was transient : they always returned home after
a short sojourn in the Holy Land. As a rule, the domestic
forces of the Syrian Franks seldom took the field more than six
or seven thousand strong. Often, when the fate of the kingdom
was at stake, the numbers of the royal host were still smaller.
Baldwin I. had only two hundred and forty knights and nine
hundred footmen at Jaffa in 1 101 to face the whole force of Egypt.
At Ramleh, when he had unwisely left his infantry behind, he
actually gave battle with no more than three hundred knights
as his whole army, and was utterly defeated. Some years later
he considered seven hundred horse and four thousand foot
enough to face the united forces of the emirs of Syria. But
perhaps the most extraordinary of all the expeditions of the
Syrian Franks was a raid into Egypt in 1118, in which no more
than two hundred and sixteen knights and four hundred infantry
took part. They advanced within three days march of Cairo,
and actually returned safely to Palestine. 2

1 To withstand Saladin s invasion of 1183. William of Tyre calls it the largest
host he had ever heard of among the Franks of Syria (xxii. p. 448).
- Albert of Aix, xii. p. 205.

254 THE ART OF WAR IN THE MIDDLE AGES [1127

Want of numbers, then, was the real cause of the failure of
the Franks to conquer inner Syria. That they ever succeeded
in establishing themselves firmly on the coast, and in holding
many districts of the mountain zone, must be attributed to the
divisions of the Moslems. As long as the interior lands were
divided between three or four independent emirs, the Crusaders
not only held their own, but actually advanced their frontiers.
Down to the rise of Zengi, the first prince who began to unite
the emirates, the Franks were slowly but surely occupying the
cities of the Infidel. 1 Nothing, indeed, could have been more
opportune than the fact that, in the early years of the twelfth
century, Damascus, Aleppo, Kayfa, Mosul, Mardin, were in the
hands of different families, all bitterly jealous of each other, and
sometimes even ready to ally themselves with the Christian if
thereby they might do their neighbours an ill turn. 2 This fact it
was which enabled a few hundred Prankish knights to ride
roughshod over Syria for some twenty years, till in 1127 Zengi
took up the governorship of Mosul, The interesting picture of
the state of the land in this year given by the Moslem chronicler
who wrote the history of the Atabegs 3 is well worth quoting.

"At the moment when Zengi appeared, the power of the
Franks extended from Mardin and Scheikstan in Mesopotamia
as far as El-Arish on the frontier of Egypt, and of all the
provinces of Syria only Aleppo, Emesa, Hamah, and Damascus
were still unconquered. Their bands raided as far as Amida in

1 The dates of the changes of dynasty in the emirates are all-important for
understanding ihe history of the Crusades. They are as follows :

Alfpfo. Held by the house of Tutush-ibn-AIp-Arslan, 1094-1117.

Held by Il-Ghazi of Mardin and his nephew Soliman. 1117-1123,
Held by Balak-ibn-Bahram, 1123-1125.
l&gt;y Il-Borsoki and his son Massoud, 1125-1128.
Surrendered to Zengi, 1128.
Damascus. Held by Dukak the Seljouk, 1095-1103.

Held by Toktagin and his house, 1103-1154.
Surrendered to Nur-ed-din, son of Zengi, 1154.

Mosul. Held by Kerboga, 1096-1102; by Jekermish, 1102-1107; by
Javaly, 1107-1108; by Maudud and his nephew Massoud, HoS-
1113 ; by Il-Borsoki, 1113-1127. Taken over by Zengi, 1127.
: The strange battle of Tel-basher in 1108 is worth notice. Tancred of Antioch
and Joscelin, Lord of Tel-basher, had quarrelled. So had Ridwan of Aleppo and
Javaly of Mosul. Each allied himself with a stranger against his own co-religionist,
and in the fight Frank fought with Frank, and Turk with Turk. Tancred and Ridwan
were victorious. Albert of Aix and William of Tyre both allude to the story.

3 The Turkish, deputies or generals of the great Seljouk Sultan, who ruled as
practically independent princes in Syria and Mesopotamia.

WIDEST EXTENSION OF THE FRANKS 255

the province of Diarbekir, and in that of El-Jezireh [Upper
Mesopotamia] as far as Xisibis and Ras-Ain. The Mussul
mans of Rakkah and Haran [Carrhae] were exposed to their
oppression and the victims of their barbarous violence. All
the roads to Damascus except that which passes by Rahaba
[Rehoboth] and the desert were infested by their plundering
parties. Merchants and travellers had to hide among the rocks
and the wilderness, or to trust themselves and their goods to the
mercy of the Bedouins. Things were growing worse and worse
and the Christians had begun to impose a fixed blackmail on
the surviving Moslem towns, which the latter paid to be quit of
their devastations. . . . They took a regular tribute from all the
territory of Aleppo as far as the mill outside the garden-gate
only twenty paces from the city itself. Then Almighty God,
casting his eyes on the Mussulman emirs and noting the contempt
into which the true faith had fallen, saw that these princes were
too weak to undertake the defence of the true religion, and
resolved to raise up against the Christians a man capable of
punishing them and exacting a due vengeance for their crimes." 1
At this moment, when the progress of the Franks was
abruptly stopped by the rise of Zengi, we may pause to define
the limits of their conquests. The kingdom of Jerusalem held
all the coast from Beyrout to Ascalon. The latter town was
still in the hands of the Fatimite princes of Egypt, and gave
them a good base for invasions of the Holy Land by the route
of El-Arish and Gaza. But the Egyptian dynasty was in a
decaying condition, and its armies seldom crossed the desert.
Indeed, Prankish raids on the Delta were more common than
attacks pushed by the Moslems into Palestine. Eastward, the
boundary of the Latin kingdom was the Jordan, save that the
strong castle of Paneas (Banias), placed beyond the head waters
of that river, gave it a watch-tower to observe Damascus. The
realm had also another outpost towards the East and South.
In 1116 Baldwin I. had resolved to push his frontier towards the
Red Sea, so as to cut the great caravan route from Damascus to
Egypt through the desert. He had executed the fatiguing
march to the head of the Gulf of Akabah, and there had
established the castle of Ailath (Elim-Elath) at its northernmost
point (1117). This stronghold communicated with Palestine by
means of two other castles, Montreal (Schobek) near Petra in

1 Quoted in Michaud s BibliothZque des Croisa.de ,, vol. iv. p. 61.

256 THE ART OF WAR IN THE MIDDLE AGES [112

the centre of the Edomite desert, and Kerak in the land of
Moab. The fief of Montreal-Kerak or of " the land beyond
Jordan " was one of the four great baronies of the Latin kingdom.
It formed such a dangerous outpost, and its position was so
forbidding, lost as it was in the desert, that we are surprised
to find that the Franks held it from 1116 till 1187, the year of
the fall of Jerusalem. 1 As long as it survived, it made the
communication between Damascus and Egypt very precarious:
Moslem caravans had to pay blackmail to its lords, or suffer
untold danger of starvation and misdirection in passing by
stealth between the three fortresses in the wilderness. Military
communication between the Fatimites and the rulers of Damascus
was equally hard ; armies marching through the sands and rocks
of Idumea were always exposed to sudden attacks from these
garrisons. They were such thorns in the side of Islam that
repeated attempts were made to capture them, all of which
failed even when Saladin himself took the matter in hand.
They only fell with the fall of the Latin kingdom, and Kerak
actually held out longer than Jerusalem.

North of Kerak the frontier of the Franks was guarded by a
chain of castles watching the defiles which lead down to the
fords of the Jordan. The line was composed of Paneas, Beau
fort, Chateau-Neuf, Safed, Castellet, and Beauvoir. South of the
last-named, where the valley of the Jordan is most deep and
rugged, there seems to have been a gap left, the natural defences
being apparently too formidable to require strengthening.

Stretching along the coast from Beyrout northward lay the
county of Tripoli, the weakest of the four crusading states. Its
rulers never succeeded in pushing inland through the passes
of Lebanon or getting a lodgment in Ccele-Syria. They only
possessed the series of narrow coast - plains round the strong
cities of Markab, Tortosa, Tripoli, and Giblet, together with the
spurs of the mountains above and between them. The great
chain of Lebanon, however, gave a strong frontier for defence.
In commanding positions, watching the few practicable passes
through the range, were the inland castles of Montferrand, Krak,
and Akkar. Weak for offence, but strong for resistance, the
county of Tripoli preserved its mountain boundary far into the
thirteenth century.

1 Kerak fell in 1188 only, but Elath had been recovered by the Moslems in 1170,
and Reginald of Kerak had failed to retake it in 1183-84.

1 1 27] PRINCIPALITIES OF ANTIOCH AND EDESSA 257

The principality of Antioch, on the other hand, had not such
advantageous frontiers. Extending far up the valley of the
Orontes, it had no natural obstacles to divide it from the Moham
medans of Aleppo. Hence the boundaries of Frank and Turk
were always fluctuating. Sometimes the Christians held Athareb,
a fortress close up to the walls of Aleppo : sometimes the
Infidels were at the gates of Antioch. The strongly-fortified
capital was the one solid centre of resistance which the Franks
possessed in Northern Syria: Athareb, Harrenc, and the other
fortresses to the east were ahvays changing hands. But the
splendid Byzantine walls of Antioch, which had held Godfrey
and Bohemund at bay for so many months, were impregnable
when held by a Christian garrison, and the city was never taken
till 1268. All its Eastern dependencies had fallen many years
before.

The county of Edessa may almost be called an Armenian
rather than a Prankish state. The number of Crusaders who
settled in it was small, and its sovereigns, unlike their neighbours
farther south, depended mainly on their Armenian subjects to
fill the ranks of their armies. It would have been a fortunate
thing for the rulers of Antioch and Jerusalem if they too could
have recruited their infantry from among the native Christian
population. But the Syrians were a far less warlike race than
the Armenians, and gave little or no military aid to their masters.
From a strategical point of view it was no doubt a mistake for
the Franks to push into Mesopotamia when North Syria was
still unsubdued. Surrounded on three sides by the emirs of
Mosul and Aleppo and the Danishmend princes of Eastern
Cappadocia, Edessa was always in danger. The county con
sisted of a few strongly-fortified places the capital, Turbessel,
Ravendal, and Hazart, with an indeterminate and ever-varying
territory around them. It had no natural boundaries, and, being
so weak in military resources, was bound to fall whenever a
strong prince should arise and unite against it the resources
of the neighbouring Mohammedan districts. The rise of Zengi
implied the disappearance of the county : it vanished after main
taining a precarious existence for less than fifty years.

It had survived so long merely because the rival dynasties

at Aleppo, Mosul, Mardin, and Kayfa had never united to crush

it. At best it was no more than a useful outwork to protect the

flank of the principality of Antioch, an outwork so distant, so

258 THE ART OF WAR IN THE MIDDLE AGES [1127

weak, and so exposed that there was no hope of permanently
retaining it. Edessa would have fallen long before if it had not
been repeatedly saved by the intervention of its neighbours to
the south. Tancred and King Baldwin I. led armies from
Antioch and Jerusalem to save it : without their aid it must
have succumbed in mo, or perhaps even in 1104. It would
undoubtedly have been better for the general defence of
Syria if the first conquerors of the land had seated themselves
at Turbessel rather than at Edessa, and contented themselves
with holding only the districts west of Euphrates : they
might then have made the great river their boundary, and
served as efficient guardians of the marches of North-Eastern
Syria.

The extension of the Frankish dominion ceased immediately
on the appearance of Zengi. The only important conquest
made after the year 1127 was that of Ascalon, taken from the
Fatimite Sultan of Egypt by Baldwin in. in 1153. Before the
end of the long reign of the great Atabeg, the balance had begun
to turn definitely in favour of the Moslems. The great mark of
the change was the destruction of the northernmost crusading
state, the county of Edessa, by Zengi s hand, in 1144. The
union of Mesopotamia and Northern Syria under Zengi s rule
completely checked the expansion of the Frankish dominion
inland. There remained the three surviving Christian states
the kingdom of Jerusalem, the principality of Antioch, the
county of Tripoli, forming a long straggling strip of territory
along the coast, much cut up by mountains, and nowhere much
more than fifty miles broad. They had no good land communi
cations with each other, and depended for their union solely on
the maritime predominance of the Italian republics.

One chance only of triumph remained to the Franks the
possibility of the arrival of a new crusading host from the West
sufficient to enable them once more to take the offensive. It
was obvious that the strength of the Latin states of Syria
unassisted would not even suffice to preserve themselves. For
one moment in 1 149 it appeared as if this chance might come
into realisation. Deeply stirred by the news of the fall of
Edessa, the nations of the West sent out the great hosts of
Conrad III. and Louis vii. on the second Crusade. Only the
broken wrecks of these expeditions ever reached Palestine, but
even these were numerous enough to encourage the King of

H49] LAST ATTEMPTS AT CONQUEST IN SYRIA 259

Jerusalem to make a bold push forward. The great campaign
of 1 149 was made upon the right lines, and a systematic
attempt was made to break the long belt of Mussulman territory
in its centre by the capture of Damascus. All other Christian
attacks on that great city were mere raids : this was a deliberate
advance, intended to bring about its permanent subjection.

If the great city had now fallen, the line of Mohammedan
states would have been cut in two, Egypt would have been
definitely severed from Aleppo and Mesopotamia, and the
fatal combination of the northern and southern Moslems
under Saladin could never have taken place. At all costs the
Crusaders should have endeavoured to break the line which links
Mosul, Aleppo, Emesa, Hamah, Damascus, and Bozrah with
the road to Egypt. But so far were the Syrian Franks from
appreciating the fact, that there is good authority, both Christian
and Mohammedan, for stating that the king and barons of
Jerusalem were very slack in pushing the attack on Damascus,
just because it seemed more likely to profit their French
and German auxiliaries than themselves. Anar, the Vizier of
Damascus, is said to have sent secret letters to King Baldwin III.
to point out to him that the capture of the place would perhaps
benefit some of his fellow-Christians, but would do himself no
good ; on the other hand, the strong fortress of Paneas by the
sources of the Jordan should be restored to him if the siege
was raised. Anar swore also that if Baldwin would not consent
to depart, he would deliver Damascus to their common enemy,
Nur-ed-din of Aleppo, the son of Zengi, rather than let it cease
to be part of Islam. 1 It is certain that the King of Jerusalem
pressed the leaguer slackly, and at last departed homeward, to
the great disgust of the emperor and the other pilgrim princes
from the West. Thus ended the one serious attempt of the
Franks to establish themselves in inner Syria and carry their
frontier up to the desert.

The fact that Zengi s dominions were divided up among his
sons (Nur-ed-din taking Syria and Seyf-ed-din Mesopotamia;,
so that for a time the unity of command was lost, and the Franks
obtained a respite, did not lead to any permanent change in the
fortunes of the crusading states. The King of Jerusalem turned

1 See Ibn-Alathir on p. 96, vol. iv. of the Bibl. dcs Croisades. Cf. also William
of Tyre, book xvii. pp. 14, 15, who says that the Count of Flanders was to be made
prince of Damascus by the Westerns, which the Syrian Franks would not endure.

2 6o THE ART OF WAR IN THE MIDDLE AGES [1163

aside to make a series of attempts to conquer Egypt, when his
eyes should have been fixed on Damascus and Aleppo. The
danger at his gates should have engrossed his attention, and no
distant enterprise should have been undertaken till the frontiers
of the kingdom of Jerusalem were safe. Four great invasions
of Egypt took place between 1163 and 1168, and more than
once King Amaury seemed on the point of succeeding. By
adroitly taking part in the war between the Egyptian vizier
Shawir and Shirkuh, the general of the Syrian prince Nur-ed-
din, he obtained a free entry into Egypt, and occupied many
towns as the ally of Shawir. For a short time a Prankish
garrison actually held Cairo in the name of the Fatimite caliph,
and defended it against the Turks and Syrians of Shirkuh.
But Amaury s position in Egypt was always precarious, because
he had continually to be keeping an eye on his own realm in
Palestine, exposed in his absence to the raids of Nur-ed-din s
governors in Damascus and Coele-Syria. It was bad strategy
to strike at the Nile while Jerusalem and Antioch still had an
enemy encamped only a few score miles from their gates. It
was the consciousness of the danger of his own realm that
ahvays kept Amaury anxious and preoccupied during his
Egyptian campaign. He had always, so to speak, to " keep one
eye behind him " : a demonstration on Jerusalem by Nur-ed-
din might bring him back from Cairo at any moment. This is
the true reason why he lost the fruits of successful campaigns, by
allowing himself to be bought off by great sums of money.
Hence it came that he levied great fines from Egypt, and for
several years received a regular tribute from Shawir, but never
made a firm lodgment in the land. At last, the most unhappy
contingency for the Franks came to pass. Shirkuh murdered
Shawir, and seized Egypt for his master Nur-ed-din (1169).
Syria and Egypt were at last united in the hands of a single
prince, for the Fatimite caliph did not long survive his vizier,
meeting, like him, a bloody end at the hands of Nur-ed-din s
lieutenants (i 17 1). 1 Amaury made one last invasion of Egypt
after the fall of his ally Shawir, leaguing himself with the
Byzantine emperor, Manuel Comnenus. But the Greek fleet
and the Prankish army lay long before Damietta, and failed to

1 So at least say the Frankish historians. Saladin s biographers either pass over
the event without details, or say that El-Adid died a natural death. See the
Mohammedan authorities quoted in the Bibliothtquc des Croisadcs, iv. 147.

187] THE RISE OF SALADIN 261

take it. Presently came the news that Nur-ed-din \vas in the
field, and harrying the borders of the kingdom of Jerusalem. At
once Amaury raised the siege and hurried home to protect his
own dominions. For the future the Franks were never able to
make another offensive move.

The union of Syria, Mesopotamia, and Egypt ought to have
brought about the instant ruin of the kingdom of Jerusalem.
That the state survived for nearly twenty years more was due to a
lucky chance. Yussuf Salah-ed-din (Saladin), who succeeded his
uncle Shirkuh as the lieutenant of Nur-ed-din in Egypt, proved
a disloyal vassal, and did not combine his power effectively with
that of his master. He did not openly break with the Syrian
prince, but played his own game, and not that of his suzerain.
Hence it was only when Nur-ed-din had died (1172) and
Saladin had overrun and annexed the dominions of his late
master s sons ( 1179-83), that all the Moslem states from the
Tigris to the Nile were really united under a single ruler.

The day of doom for the kingdom of Jerusalem was now at
hand. Saladin s realm surrounded the crusading states on all
sides, and when he threw himself upon them their fall was
sudden and disastrous. At the great battle of Tiberias (Hattin)
in 1 187, the Prankish host was exterminated ; Jerusalem fell in a
few months, and after its fall fortress after fortress dropped into
Saladin s hands, till little remained to the Crusaders save Tyre,
Tripoli, and Antioch. That these small remnants of the
Christian states escaped him was due to the third Crusade.
Richard of England and Philip of France failed to retake
Jerusalem, but they recovered Acre and most of the coast-towns
of Palestine. Richard inflicted a crushing blow on Saladin at
the battle of Arsouf (1191), and shortly after the Franks and
Moslems came to an agreement, which saved for Christendom a
wreck of the kingdom of Jerusalem. The inland was lost, but
the long narrow coast-slip from Antioch to Jaffa was preserved.
Saladin died shortly afterwards (1192), and his dominions broke
up ; his sons and his brother El-Adel each kept a portion.
This disruption of the Ayubite realm was the salvation of the
Syrian Franks ; their hold on the coast-region of the Levant
was to endure for yet another hundred years. But the kingdom
of Jerusalem (it might more appropriately have been called the
kingdom of Acre) was now a mere survival without strength to
recover itself. It might have been stamped out at any moment,

262 THE ART OF WAR IN THE MIDDLE AGES [1291

if a leader of genius had arisen among the Mohammedans
and united again all the resources which had been in Saladin s
hands. But the unending civil wars of the Ayubites gave a
long lease of life to the decrepit Prankish realm. Strange as it
may appear, the Christians were even able to recover the Holy
City itself for a moment. Jerusalem was twice in their hands
for a short space once in 1229, when the Emperor Frederick II.
got possession of it once in 1244. On each occasion the
reconquest was ephemeral it marked the weakness of the
Saracen, not the recovered strength of the Frank. But along
the coast the thin line of ports was firmly held ; strengthened
by all the resources of the scientific combination of Eastern and
Western fortification, they long proved impregnable. The sea
was always open to bring them food and reinforcements ; the
Italian maritime powers were keenly interested in their survival
tor commercial reasons. Hence it was that the banner of the
Cross still waved on every headland from Laodicea to Jaffa till the
thirteenth century was far spent and the house of the Ayubites
had vanished. The end of the kingdom of Jerusalem only drew
near when the new and vigorous dynasty of the Bahri Mamelukes
had once more united Egypt and Syria. Then at last came the
doom of the Prankish realm, and one after another the ports of
the Levant yielded before the arms of the great Sultans, Bibars,
Kelaun, and Malik-el-Ashraf. Acre the last surviving strong
hold fell after a two months siege in May 1291. The only
wonder is that it had survived so long ; had Saladin s life been
protracted for ten years, the end would have come nearly a
century earlier. But in the thirteenth as in the twelfth century
the dissensions of the Mohammedans were the salvation of the
Franks.

As an example of the importance of the sea-power in the
Middle Ages, we may note that the long survival of the coast
fortresses of Syria would have been wholly impossible if any of
the Eastern powers had possessed a competent navy. But the
Genoese and Venetians completely dominated the waters of the
Levant, and the Prankish ports could only be attacked on the
land side. Even when they had fallen, the Mamelukes made
no attempt to use them as the base for the creation of a war-
navy. They sank to mere fishing villages when they fell back
into Mohammedan hands, and never appeared again as military
ports. Hence it came to pass that the insular kingdom of

I2i8] JOHN DE BRIENNE IN EGYPT 263

Cyprus, the last foothold of the Franks in the Levant, endured
for more than two centuries after the fall of Acre. It was only
lost to Christendom when there arose at last a Moslem power
which built a great fleet and determined to expel the Italian
galleys from the Levant. The Ottoman Turks overran the
island in 1571, and then only did the maritime domination of
the Franks in Eastern waters come to an end.

Before dismissing the subject of the grand strategy of
the Crusades, we have still to deal with two 1 considerable
diversions executed by the Franks outside the limits of
Syria during the thirteenth century diversions rendered
possible by their complete possession of the command of the
sea. We refer to the two invasions of Egypt in 1218-20
and 1249-50 those of John de Brienne and St. Louis. There
was more to be said in favour of these expeditions than for
those which King Amaury carried out in 1 163-69. At the
earlier date there was still a kingdom of Jerusalem which
needed protection, and to take away its garrison for a
campaign on the Nile was dangerous. Things were much
changed in the thirteenth century : the kingdom had shrunk to a
few coast-fortresses, which were, for the most part, self-sufficing,
and could take care of themselves. Its defence, therefore, had
become much more easy : if during the Egyptian expedition
the governors of Damascus or Jerusalem should march on Acre
or Tyre, the cities could be trusted to hold out for many months.
They had the sea at their backs and could count on the aid of
Venice and Genoa. Moreover, the attack on Egypt was to be
made, not by the home levies of the barons of Palestine, but by
great crusading forces from the West. Nothing, therefore, was
risked in Palestine over and above the ordinary danger from
the inland.

Egypt was a tempting prey rich above other lands, peopled
by an unwarlike race, and ruled by a monarch depending for
his military resources not on his born subjects, but on mercenary
bands of Turks, Kurds, Syrians, and Arabs. Egypt and Syria,
too, were divided between different branches of the Ayubitc

1 The expedition of St. Louis to Tunis has no bearing on the general history of
the Crusades, and was inspired by a religious, not a military object it being supposed
that the ruler of Tunis might be converted to Christianity !

264 THE ART OF WAR IN THE MIDDLE AGES [1218

house in 1219: El-Kamil reigned at Cairo, El-Muazzam at
Damascus ; and though they were not unfriendly to each other,
yet two rulers can seldom combine their efforts to act like one.
The conquest of Egypt, regarded as an enterprise wholly
independent of the defence of Palestine, presented both in
1219 and in 1249 many attractions. A commander of genius
might probably have accomplished it with the forces led by
either John or Louis. It is more doubtful whether the land
could have been held when once subdued ; but, at least, the
experiment was worth making.

But if the problem was not an impossible one, it was one
which required to be solved according to the general rules of
strategy. Egypt must always be "grasped by the throat" by a
bold march on Cairo, and for a march on Cairo there are only
two practicable routes. It is absolutely necessary to avoid getting
entangled in the countless canals and waterways of the Delta.
The first of the two alternative routes is to land near Alexandria,
to keep west of the westernmost branch of the Nile, as did
Bonaparte in 1798, and to march by Damanhour and Gizeh.
The drawbacks of this route are that its first two or three stages
are through desert, and that it brings the invader opposite to
Cairo, with the Nile still interposed between him and his goal.
The crossing of the main stream in face of the enemy, when the
army has pushed so far inland, might prove very perilous. The
second and far preferable route is to start near the ancient
Pelusium and march by Salahieh and Belbeis on Cairo, keeping
east of the easternmost branch of the Nile. This brings the
invader directly on to the capital ; he has no canals or water
ways to cross, and the distance he has to cover is no more than
a hundred miles. Here also the main difficulty to be faced is
that the first two stages are through desert country. Egypt has
always been invaded by this line ; it was followed by Cambyses,
Alexander the Great, Antiochus Epiphanes, Amru, and Selim I.
Lord Wolseley only diverged from it in 1882 because he was
able to utilise the Suez Canal, and so shorten his land march by
forty miles. This route was well known to the Franks ; Amaury
had used it in 1168, taking Belbeis, and actually laying siege to
Cairo, which he might have captured if he had not allowed himself
to be bought off by an enormous war-indemnity. It is therefore
most astonishing that both John de Brienneand St. Louis neglected
this obvious and easy line, and chose instead to land at Damietta.

i22o] FAILURE OF JOHN DE BRIENNE 265

The road from that place to Cairo leads through the very midst
of the Delta, over countless canals and four considerable branches
of the Nile. Across it lie a dozen strong positions for the defend
ing army. It is not too much to say that the invasion of Egypt
by this line is bound to fail, if the masters of the country show
ordinary vigour and intelligence. The fates of the two Prankish
expeditions are a sufficient commentary on the wisdom of their
leaders. John de Brienne only took Damietta after a siege of
eight months ; his troops were already much exhausted when he
advanced into the Delta ; they were brought to a stand by the
line of the Ashmoun Canal, behind which lay the army of the
Sultan El-Kamil. They made several unsuccessful attempts to
break through, and were already despairing of success when
they learned that the land between them and their base at
Damietta had been inundated ; the Nile was rising, and the
Egyptians had cut the dikes. They hastily retreated towards
Damietta ; but the waters were out everywhere, the Sultan
followed hard behind them, and, to save themselves from starva
tion or drowning, the Crusaders had to come to terms. El-
Kamil granted them a free departure, on condition that they
should evacuate Damietta (August 1221).

Far worse was the fate of St. Louis when he tried the same
route in 1249. Considering how John de Brienne had fared, we
can only marvel that he ventured to choose the same road. He
started with somewhat better fortune than his predecessor, for
Damietta fell into his hands after a very slight engagement with
the Moslems. But he then wasted no less than six months in
waiting for stores and reinforcements ; all this time was employed
by the Sultan in increasing his army and in preparing obstacles
for the march of the French. When, in November 1249,
King Louis did at last begin his advance, he was promptly
checked by the same bar which had ruined John de Brienne,
the impassable Ashmoun Canal, defended by the Egyptian
army. Time after time the bridges and causeways which he
strove to construct were swept away by the military machines
of the enemy. At last Louis got across by night with his cavalry
at a deep ford practicable only for horsemen ; the infantry could
not follow. The Egyptians were for a moment surprised, but
the king s brother, Robert of Artois, threw away all chance of a
victory by charging rashly into the streets of Mansourah with
the van long ere the king and the main body had come upon

266 THE ART OF WAR IN THE MIDDLE AGES [1250

the field. He and the whole of his division were cut to pieces,
and when Louis arrived he only succeeded in forcing his way to
the neighbourhood of Mansourah at the cost of half his knights.
At last, however, he worked his way to the bank opposite his
own camp, and his infantry were able to finish the causeway at
which they had long been labouring, and so to join him. The
French thus obtained a lodgment beyond the Ashmoun, but
the success had cost them so dear that they could advance no
farther. They lingered near Mansourah for some months, unable
to move forward and unwilling to turn back, till at last famine
and pestilence broke out, and compelled them to abandon the
invasion. But the Egyptians had broken the road between them
and Damietta, and as they straggled northward they were cut
to pieces in detail in a long running fight extending over several
days. At last the king was surrounded and taken prisoner, and
soon after the few surviving wrecks of the army laid down their
arms. They could not even make terms for themselves, as John
de Brienne had done in 1221, and the greater part of the captives
were put to death in cold blood by the Egyptians. 1

As a comment on King Louis strategy we need only point
out that, even if he had successfully forced the passage of the
Ashmoun when he first reached it, he would yet have had to
pass three broad branches of the Nile and numerous canals, all sus
ceptible of easy defence, before reaching Cairo ! Nothing but the
entire want of geographical knowledge in those mapless days
can explain the madness of the Crusaders in twice selecting
the utterly impracticable route Damietta-Mansourah-Benha-
Cairo, when it was open to them to use the easy and obvious
road by Salahieh and Belbeis. Apparently they were attracted
by the port and fortress of Damietta, which seemed to offer an
excellent base and storehouse, while there was no town at all
in the tract east of the ancient Pelusium, the proper starting-
place for the descent. There was nothing else to account for
the preference : one landing-place was as open as another to an
armament in full command of the sea, and the coast east of
Felusium, though shallow inshore, does not present any real
obstacle to the approach of vessels of such light draught as were
those of the thirteenth century. A careful examination of the
Government Survey maps of the Delta seems to show that east of
Pelusium and its marshes there is a sandy shore, with sufficient

1 For a more detailed account of Mansourah, see pp. 340-347.

1250] FAILURE OF ST. LOUIS IN EGYPT 267

depth of water for light vessels to get close in. The region is
so remote from the military centres of Egypt that no local
resistance need have been feared.

We may fairly say, therefore, that the two great invasions of
Egypt in the thirteenth century failed mainly because they were
undertaken with insufficient geographical knowledge, and con
ducted along an impossible route. That they would have had
a fair chance of success if they had been more wisely directed,
is best shown by the fact that the Moslem historians one and all
assure us that their compatriots had completely lost heart after
the first successes of the Christians. In I22O El-Kamil actually
offered to surrender Jerusalem, Tiberias, Giblet, Ascalon,
Nazareth, and Laodicea, if the Crusaders would but restore
Damietta and return home. In 1249 Damietta was evacuated
almost without the striking of a blow, and the army which
mustered behind the Ashmoun was in great disorder and deep
depression. If forced to fight not covered by a broad water
course, but in the open country about Salahieh or Belbeis, it
would certainly not have held its ground.

It was the same utter want of geographical knowledge which
had ruined the Provencal Crusaders of 1101, and the French
host of Louis vil. in 1 248, that brought to such disastrous ends
the two formidable expeditions which endeavoured to subdue

Egypt.

CHAPTER III

THE TACTICS OF THE CRUSADERS

Section I. The Early Battles and their Tactics : Dorylceum,
Antioc/i, A sea/on, Ramleh.

THE Western countries which contributed the largest propor
tion of warriors to the first Crusade were precisely those
in which cavalry were at the time most predominant France
and Aquitaine, Lotharingia, Western Germany, and Italy both
Lombard and Norman. In each of the contingents which
marched out in 1096 to join the great host which mustered at
Constantinople, the horsemen were considered the main combat
ant force. If foot-soldiery followed by tens of thousands, it was
not because their lords considered them an important part of
the line of battle, but because the same religious enthusiasm had
descended upon the poor as upon the rich, and all were equally
bent on seeking the path to the Holy Sepulchre. It was evident,
too, that infantry would be required for sieges, the service of the
camp, and the more onerous and less attractive labours of war.
So little, however, were they esteemed, that in the first general
engagement in which the grand army of the first Crusade engaged
the battle of Dorylasum the foot-soldiery were left behind by
the tents, and the horsemen alone drew up in the line of battle. 1
Nor did the infantry even prove competent to keep the camp
safe they did not prevent the flanking parties of the Turks
from entering it and massacring hundreds of the non-combatants
committed to their care.

The Crusaders then were accustomed only to one development
of tactics the shock-tactics of heavily-armed cavalry. They
regarded infantry as fit at best to open a battle with a dis-

1 See p. 271.

io 9 7] THE CRUSADERS AND THE TURKS 269

charge of missiles, before the serious fighting began, or to serve
as a camp-guard.

Ranged to oppose them, however, they found enemies of whom
the most formidable were the Turks, a race long accustomed to
defeat by their Parthian tactics the most powerful and the best
disciplined heavy cavalry of the day that of the East-Roman
Empire. The other Moslem powers who still employed the
older methods of Saracen war, such as the Egyptian Fatimites,
were far less dangerous to the Crusaders. They like their
predecessors described by Leo and Nicephorus Phocas still
depended on the impact of their mailed horsemen, who were
individually inferior to the Byzantine trooper, and still more so to
the Prankish knight. But the Turkish horse-archers were the
foe who were destined to prove the main danger to the Crusaders,
as they had long been to the emperors of Constantinople. It was
they who were to teach not only the first invaders of the East,
but every army that followed them, many a bitter lesson.

We have already recapitulated in an earlier chapter the
canons which the masters of military science in the Byzantine
Empire had drawn up for use in campaigns against the Turks.
They were, put shortly, (i) always to take a steady and sufficient
body of infantry into the field ; 1 (2; to maintain an elaborate
screen of vedettes and pickets round the army, so as to guard
against surprises; 2 (3) to avoid fighting in broken ground
where the enemy s dispositions could not be descried ; 3 (4) to
keep large reserves and flank-guards ; 4 (5) to fight with the rear
(and if possible the wings also) covered by natural obstacles,
such as rivers, marshes, or cliffs, so as to foil the usual Turkish
device of circular attacks on the wings or the camp-guard ; 5
(6) always to fortify the camp ; (j) never to pursue rashly
and allow the infantry and cavalry to get separated after a
first success. 6 With the necessity of all these precautions well
understood, the Byzantines had yet suffered many disasters
at the hands of the Turks. How was it to be expected that
the Crusaders would fare, to whom some of these precau
tions would have seemed impossible, some ignominious, all
unfamiliar? As a matter of fact, they knew nothing of them,
since they utterly despised the Greeks and their methods of
warfare, disdained to learn anything from them, and took

1 Leo, Tactica, xviii. 63. - Ibid. 68. 3 Ibid. 64.

J Ibid. 71. 5 Ibid. 73. Ibid. 74.

270 THE ART OF WAR IN THE MIDDLE AGES [1097

nothing but guides and money from the emperor. 1 In their
first campaign they were as successful in violating every one
of these rules as if they had committed them to memory for
the express purpose of not carrying them out.

The hordes under Peter the Hermit and Walter the Penniless
which first crossed the Bosphorus can hardly be called an army.
Even in the eyes of their own countrymen they scarcely counted
as a military force, since they comprised but a very few mounted
men. When they were destroyed by the Seljouks near Nicaea,
they are said to have numbered only five hundred horse to
twenty-five thousand foot : 2 they had lost many thousands on the
way by the hands of the Greeks and Bulgarians, but it is certain
that in these earlier disasters the infantry had suffered infinitely
more than the cavalry, so that the original force must have
shown a still larger preponderance of men on foot.

The great army which started from Constantinople in May
1097 was a very different host. According to Western ideas, it
was a most formidable instrument of war. Many rich counts
and dukes and their well-equipped retainers served in its ranks.
Its numbers are given as high as a hundred thousand horse and
six hundred thousand foot figures impossible in themselves, but
showing a proportion between the two arms which was infinitely
more suited to the practice of the day than that which had pre
vailed in the unfortunate horde of Walter the Penniless.

Yet this great host came very near to suffering a complete
disaster in its first serious conflict with the Turks. After laying
siege to Nicsea and repelling with success the attempts of Sultan
Kilidj-Arslan-ibn-Soliman to relieve it, they forced the place to
surrender. On June 27 they started forth to march into the
interior of Asia Minor, following the great Roman road which
leads by Dorylasum, Philomelium, and Iconium to Tarsus. The
countryside was wholly desolate : " Romania, a land once rich
and excellent in all the fruits of the earth, had been so cruelly
ravaged by the Turks, that there were only small patches of
cultivation to be seen at long intervals." 3 Food for man and
horse was difficult to procure, and it was perhaps to cover a

Save, indeed, Raymond of Toulouse, who borrowed some " Turcopoles," i.e.
cavalry taught to act as horse-archers after the Turkish fashion, for his second
expedition. But he got no use out of them, except to escort his flight (Albeit of
Aix, viii. p. 19).

- William of Tyre, book i.

3 Fulcher of Chartres, chap. v.

1097] BATTLE OF DORYL.-EUM 271

greater space for foraging, and not out of mere carelessness, 1 that
the army split into two columns, marching parallel to each other
at a distance of some seven miles. The right-hand corps was
composed of the followers of Godfrey of Bouillon, Raymond of
Toulouse, Hugh Count of Vermandois, and most of the French
and Lotharingian contingents. The left column included
Bohemund and Tancred with the Sicilian Normans, Robert of
Flanders, Robert of Normandy, and Stephen of Blois. They
seem to have been fairly equal in size and composition.

Battle of Dorylccu m, July I, 1097.

After debouching from the Bithynian mountains, the Crusaders
found themselves in a broad upland plain, watered by the
Thymbres, a tributary of the Sangarius. It was a rolling
country, destitute of strong positions, and very well suited to the
peculiar tactics of the Turks. Flying parties of their light horse
soon began to hover around the advancing columns, but the
crusading leaders did not take the obvious precaution of draw
ing together, or at least arranging to keep in close touch. On
sighting the enemy they merely contracted their straggling line
of march and kept vedettes out to prevent a surprise. On June
30 they camped some miles on the north side of the Thymbres,
and not very far from the ancient and ruined town of Dorylaeum.
On the 1st of July the left division, with which we are most con
cerned, moved forward to resume its route, and had marched for
about an hour when its scouts reported the approach of the
Turks in huge numbers. Bohemuncl, to whom the other chiefs
had committed the general charge of the host, ordered the tents
to be pitched and the baggage unladed by the side of a reedy
marsh - which gave a certain amount of cover, and deployed his
men in front of it. The infantry were left to guard the impedi
menta, 3 the cavalry alone drew up in line of battle. The camp
was not fully pitched, nor the squadrons completely ranged in
order, when swarms of Turks suddenly appeared from all
directions, pressing in on the flanks and rear of the army as well

1 Fulcher (chap, v) says that the parting was accidental, owing to the divergence
of one column at a cross-road, and the failure to get into touch again. Albert of Aix
says that it was deliberate, and ordered for the reason stated above. William of Tyre
says that it was uncertain whether it was accidental or not.

2 "Juxta quoddam arudinetum " (Fulcher, v.).

3 Gesta Francornm, 6 : " Pedites prudenter et citius extendunt tentoria, milites eunt
viriliter obviam iis [Turcis].

1272

THE ART OF WAR IN THE MIDDLE AGES [1097

as upon its front. The Sultan had gathered all his available
forces, and, though too late to relieve Nicaea, trusted to avenge
himself on its conquerors by a battle in the open field. The most
distant Seljouk hordes of Asia Minor had now had time to join
him, and his host was enormous the Crusaders estimate it at
from a hundred and fifty thousand to three hundred and fifty
thousand strong. What struck the Franks with the greatest
surprise was that every man was mounted : the whole horde was
composed of horse-archers, and not a foot-soldier was visible.

In a few minutes the Crusaders found themselves enveloped.
The Turks pressed in from all quarters at once ; some appeared
in the rear and cut to pieces many belated parties who had not
reached the camp at the moment when the fight began ; l others
threatened the flanks ; the majority advanced against the
Prankish line of battle. But they were not drawn up in any
regular array or order : in loose swarms they kept riding along
the crusading line and discharging their flights of arrows into
the masses of heavy-armed cavalry. There was no main body
which the Franks could charge, and Bohemund, lest his men
should fall into disorder, refrained from ordering a general
advance, hoping that the enemy might ere long close with him.
But they showed no intention of doing so, and fresh hordes
were continually pressing up, emptying their quivers, and then
sweeping off to the flanks. At last the Crusaders grew restless
and angry : many bands from various parts of the line broke
out and dashed to the front. But they could not reach the
Turks, who rode off at their approach, overwhelming them with
showers of arrows and slaying their horses by scores the
mail-clad men suffered much less than might have been expected.
But when they turned to make their way back to the line, the
enemy closed around them, cut off the stragglers, and destroyed
many of the parties wholesale. Seeing the little profit that the
sallies brought them, the Crusaders soon desisted from attempting
to drive off the enemy, and contented themselves with closing
their ranks and standing firm. But this passive policy only
made them a more helpless prey to the Turks, whose arrows
fell so thickly among the crowd that the line began to grow loose
and disordered. This unequal combat, in which the Franks
suffered heavy loss and the Turks little or none, went on for
-several hours. At last the host grew more and more unsteady,

1 Raymond d Agiles, i.

1097] DORYL^UM: THE HOUR OF PERIL 273

and instinctively began to fall back towards the camp, the flanks
especially giving ground and closing in towards the centre, so
that the whole tended to become a clubbed mass instead of an
orderly line of battle. But there was no help in the camp ; while
the main battle was going on, many bands of Turks had assailed
it from the rear, and had broken in among the disorderly infantry
who had been charged with its defence. They were already
pillaging the tents and slaying the non-combatants, priests,
servants, and women, whose screams rose loud above the tumult
as the cavalry fell back towards the encampment. At the
approach of the horsemen the Turks in the rear stopped their
plundering and drew off, thinking that the Crusaders were re
turning to drive them away. " But," as an eye-witness remarks,
" what they thought was a deliberate move on our part was really
involuntary, and the result of despair. For, crushed one against
another like sheep penned up in a fold, helpless and panic-stricken,
we were shut in by the Turks on every side, and had not the
courage to break out at any point. The air was filled with shouts
and screams, partly from the combatants, partly from the multi
tude in the camp. Already we had lost all hope of saving our
selves, and were owning our sins and commending ourselves to
God s mercy. Believing themselves at the point to die, many
men left the ranks and asked for absolution from the nearest
priest. It was to little purpose that our chiefs, Robert of
Normandy, Stephen of Blois, and Bohemund kept striving to beat
back the Infidels, and sometimes charged out against them. The
Turks had closed in, and were attacking us with the greatest
audacity." 1

Everything portended an instant and terrible disaster, when
suddenly the face of the battle was changed in a moment.
Messengers had been sent off earlier in the day to seek the right-
hand column, whose exact position seems to have been unknown
to the leaders of the left-hand corps. They had at last found it,
encamped some six or seven miles away. 2 On receiving the news,
Duke Godfrey, Raymond of Toulouse, and the other chiefs armed
and mounted, and spurred off for the battlefield, with all the
horsemen of their host. They sent before them some swift riders
to warn Bohemund of their approach. The infantry remained be
hind to guard the tents.

The Turkish Sultan seems to have altogether neglected to

1 Fulcher, i. 5. - Albert of Aix.

274 THE ART OF WAR IN THE MIDDLE AGES [1097

reconnoitre the march of Godfrey s division, or, at any rate, had
forgotten to pay any heed to its possible arrival on the field.
The Crusaders, as they pushed on towards the fight, found no one
in their way, and at last, topping the ridge which bounded the
valley where the conflict was raging, saw the whole battle at their
feet. They had, in the most fortunate manner, come in upon
the left flank and rear of the Turkish host, which had now closed
in upon Bohemund s camp and was contracted into a small space.

Godfrey saw that the most splendid opportunity for a sudden
attack on the flank and rear of the Turks was in his hands, if he
struck hard at once, before his arrival had been seen and provided
against. Sending back orders to those behind to gallop in at
full speed, he himself dashed at the Turks with the head of his
column, fifty knights of his own following. The Sultan and his
bodyguard were visible, stationed on a hillock behind the centre
of the Turkish semicircle. Godfrey charged straight at them,
and his impetuous assault from the rear seems to have been the
first notice of the change in the battle that reached the enemy.
The rest of the Crusaders of the right column came riding in at
full speed behind him, each band crossing the ridge by the
shortest cut it could find Raymond of Toulouse on the left
next the camp of Bohemund, the Count of Flanders to the
centre behind Godfrey, the Bishop of Puy by a distant hillside
and through a gap in the ridge which brought him to the rear
of the Turkish centre.

The Infidels had no time to form a front, before they realised
that a new army was in the field. Thousands of Christian horse
men were dashing in upon them, rolling up their left flank, and
striking their centre from the rear. They hardly attempted to
rally, though the Franks in their hasty deployment and hurried
advance must have come in upon them in considerable disorder. 1
Struck by a simultaneous impulse of panic, the whole Turkish
host swept off the field in wild rout : only the Sultan s bodyguard

1 The ground over which the right column reached the field was mountainous
(Baldric of Dol; Guibert of Nogent. See Delpech, ii. 153). I conclude, therefore, that
they cannot have marched in line : they had started off in haste, and no doubt the rear
must have straggled far behind the head of the column. As a sudden blow was ab
solutely necessary, there cannot have been any time for them to deploy into a regular
order of battle. If Godfrey had waited to do so, the Turks would have got off. It
seems certain, therefore, that each contingent came over the ridge at the point nearest
and most convenient to itself, the Count of Toulouse far to the left, so as to join
Bohemund and the left column in their final attack.

log?] DORYL^UM: VICTORY OF THE CRUSADERS 275

held out for a few minutes to allow their master to get a fair
start in the flight. The victory was made more crushing by the
fact that Bohemund s tired troops delivered a desperate charge
the moment that their friends appeared in the rear of the enemy.
Thus the Turkish left wing was caught between the two Christian
hosts, and suffered severely ere it could get off.

The victorious Crusaders pursued the defeated foe with the
greatest energy, prevented them from rallying, seized their richly
stored camp, and finally scattered them to the winds. Kilidj-
Arslan did not dare to offer battle again during the many weeks
occupied in the march through the interior plateau of Asia
Minor. The panic among his followers had been so great that
they continued flying at full speed long after the victors had
stayed their pursuit. When the Crusaders resumed their march,
they found the roadside, for three days journey from the field,
strewn with the horses which the Turks had ridden to death in
the wild flight, " although the Lord alone was now pursuing
them." 1

The losses on both sides had been less than might have been
expected. The Turks had only suffered in the last ten minutes
of the battle, when their left wing was caught between the two
Christian divisions. The Franks of Godfrey s host had not
suffered at all : those of Bohemund s column had been under
the arrow-flight for five hours, but their armour helped them,
and more horses than men had been slain. We need not be
surprised to hear that the victors had lost only four thousand
and the vanquished only three thousand men. Much the largest
share of the Christian loss fell upon the wretched foot-soldiery,
who had been massacred among the tents. 2

Dorylaeum can only be called a victory of chance. The
Crusaders had deserved defeat by their careless march in two
disconnected columns. How utterly unknown the locality of
the two divisions was to each other is best shown by the fact
that it took five hours 3 for Godfrey s succours to reach
Bohemund, though there were only six or seven miles between
them. Evidently the greater part of this time must have been
wasted while Bohemund s messengers, sent off when the Turks

1 Fulcher, i. 5. Raymond d Agiles, 239.

- Figures taken from William of Tyre a late authority, though a very capable one.
3 Fulcher gives five or six hours as the duration of the engagement, and also
remarks that the messengers reached Godfrey very late : (chap. v. ).

276 THE ART OF WAR IN THE MIDDLE AGES [1097

first threatened an attack, were vainly searching for the right
column. A body of men numbered by tens of thousands, and
carrying with it an enormous train of baggage, is not a hard thing
to find, if only its general direction is known. We can but con
clude, therefore, that the two divisions must have completely lost
touch with each other, and have marched quite at haphazard.

The left column would obviously have suffered a terrible
disaster if the succours had not appeared at the right time and
in the most effective position. The Franks were wholly unable
to cope with the unexpected form of the Turkish attack. They
made no attempt to use their infantry in conjunction with their
cavalry, either by setting those armed with missile weapons to
return the hostile showers of arrows, or by employing those
armed with long weapons spears and the like to serve as a
refuge and shield for the cavalry in the intervals between its
charges. Probably in their untrained state the foot-soldiers
would have been unable to discharge either function very
effectively we have seen that they were not even able to defend
the camp. But for want of them Bohemund and his colleagues
condemned themselves to fight that most hopeless form of battle
in which cavalry endeavour to act on the defensive and to hold
a position. This course was almost as dangerous as the one
which they avoided that of making a general charge with
unprotected flanks into the midst of the great circle of Turkish
horsemen. To wait and receive the enemy s shafts without
being able to reply to them could only retard disaster, and not
avert it. As a matter of fact, after five hours of endurance the
Franks had recoiled to their tents in a disorderly mass, and
were about to break up and suffer massacre when their comrades
came to their aid.

Undeserved as the victory had been, it yet gave the Crusaders
a free passage through Asia Minor. They were not again
obliged to fight a pitched battle till they had arrived at Antioch.
By the time that the siege of that place had been formed, the
condition of the army had greatly changed. The privations
which it had been forced to endure on its long march had fear
fully thinned its ranks. The infantry had fallen by the way in
tens of thousands : the cavalry had lost the greater part of its
horses. For the Western chargers could not stand the heat,
and the forage provided for them was both insufficient in
quantity and different in form from that to which they were

1097] THE SIEGE OF ANTIOCH BEGUN 277

accustomed. In the winter of 1097-98 there are said to have
been less than a thousand left in the Christian camp fit for
service. The whole army would have been dismounted if it had
not been for one or two lucky captures which furnished them
with a quantity of Syrian horses won from the enemy.

With the long siege or rather blockade- of Antioch we
have not much to do. The military machines of the Franks
proved wholly unable to deal with the splendid walls of the
city a legacy from Justinian. For many months the Crusaders
lay encamped in a secure triangular position between the
Orontes and the city wall, blocking three of the gates on the
east and north-east, but leaving free ingress and egress to the
enemy through those which led to the north-west and north.
At this rate the leaguer might have gone on for ever the
besieged only began to be inconvenienced when, five months
after they had arrived before the place, 1 the Franks built a tower
to command the western gate, 2 and a sort of tete-du-pont (if we
may use the term in an unusual sense) to block the exit from
the Bridge-Gate, where the city ran down to the bank of the
Orontes. After this the Turks were straitened for supplies of
food, and especially for forage for their horses, but they were
not thoroughly enclosed, as they could still get in and out at
nights by posterns, and never lost their communications with
their friends without. Meanwhile, the Christians were suffering
quite as much as their adversaries : they had drained the
immediate neighbourhood of supplies, the parties which they
sent out to plunder at a distance were repeatedly cut off by
the Turks, and though they succeeded in getting in touch with
the sea at the port of St. Simeon, where a Genoese flotilla had
come to anchor, their communication with it was often inter
rupted and always hazardous. Famine reigned in the camp all
through the winter and early spring, and men and horses died
off like flies.

It was fortunate for the Franks that the two most serious
engagements during the siege were fought in places where the
Turkish methods of fighting could not easily be employed.

The first fight was the more important one. The emirs of
Syria had gathered an army, variously estimated at from twelve
thousand to twenty-eight thousand strong, to raise the siege, or

1 The siege began October 2lst. The new works were not begun till February.
3 The gate of St. George.

278 THE ART OF WAR IN THE MIDDLE AGES [1098

at least to harass the besiegers. Hearing of its approach, the
crusading chiefs determined to make a bold stroke at it before
it closed in upon them. The Turkish force had met at the
town of Harenc (Harim), sixteen miles east of Antioch. Their
best way of communicating with the place was by advancing
through the open ground north of the Orontes and the Crusaders
camp, and so coming in upon the Bridge-Gate. But this route
had one dangerous defile upon it. About seven miles east of
Antioch, there is a place where the great lake of Begras at its
southern end approaches within a mile of the Orontes : 1 the
road passes through this narrow neck of land. This was the
point at which the Crusaders resolved to intercept the relieving
army : the neighbourhood of their camp was now well known to
them, and Bohemund had noted this spot not only as giving a
narrow front where superior numbers would not avail, but also
as affording opportunities for a surprise, for the approach was
hilly, and there were man} dips in the ground where a consider
able force could lie hid.

The Franks could only put into the field seven hundred
well-mounted men : their horses had fallen into such bad con
dition that only that number could be found fit to face a short
night march and a battle to follow. With this small band
Bohemund, to whom the command had been given for the day,
marched out under cover of the darkness, and, " passing over
seven valleys and seven ridges," - took post close to the narrow
neck between the lake and the river. At dawn the Turks were
seen advancing, with a swarm of horse-archers thrown out in
front to cover their main body. When the whole were in the
defile, the Crusaders, having formed a line of five small squadrons,
with a sixth in reserve, galloped in upon this vanguard. The
Turks yielded after a smart skirmish, and fell back in disorder
on their main body. If there had been room and time for the
Infidels to deploy, 3 the Crusaders must have been crushed, but

1 The distance was apparently much shorter in 1098 than now ; probably the
marshy southern end of the lake is drying up and receding.

- Raymond d Agiles, p. 253.

3 Raymond d Agiles and \Villiam of Tyre agree on this. The latter says : " Com-
primentibus eos locorum angustiis, hinc lacu inde fluvio licentiam evagandi inhibente
ad consuetas discurrendi artes et sagittandi habilitatem discurrere non dabatur."
William of Tyre, however, does not seem fully to have grasped the topography
when he speaks of the Turks as having "crossed the river during the night at the
upper bridge." There is no river between Harenc and the battle-spot. The only
stream between the Bridge-Gate and Harenc is the Iferin (Labotas), the river which

1098] COMBAT OF HARENC 279

the Turks were caught still massed, and with the lake and river
close on each flank. The van was thrown in upon the rest of
the host in helpless rout, the main body was so crushed and
cramped in the confined ground that they could not scatter or
outflank the Crusaders, and though they made some attempt
to bear up against the charge, yet, when Bohemund and his
reserve were thrown into the fight, they slackened in their
resistance and strove to fly. But flight was not easy, with the
waters so close on each side, and no less than two thousand
horsemen were slain or drowned. The Franks pursued
vigorously, and captured the town of Harenc and the whole
of the enemy s baggage before nightfall (Feb. 8, 1098).

The second fight was of a still simpler description. The
garrison made a sally in force from the Bridge -Gate, and
crossed the Orontes to operate in the plain beyond it. Promptly
attacked, with the river at their backs, they could neither deploy
into their usual crescent - shaped formation, nor practise the
alternate advances and retreats which formed the basis of their
system of tactics. Crushed back against the water by vigorous
charges, they were badly beaten, and in struggling back to the
gate, which had been shut behind them by a foolish inspiration
of the Emir Baghi-Sagan, they suffered heavily, and many
hundreds were drowned or slain (March 1098).

Antioch fell by treachery on June 4, IO98. 1 It obviously
could not have been taken by force, and that it could have been
reduced by starvation is very improbable, as its communications
with the open country were straitened rather than cut off. The
very day of its fall the vanguard of a great relieving army
appeared in the vicinity. Not only the nearer princes of Syria,
but the more distant powers of Mesopotamia and Persia, had
combined to rescue Baghi-Sagan from his assailants ; their host
was headed by Kerboga, the Emir of Mosul, and was reckoned
at one hundred and fifty thousand or two hundred thousand
strong. In a few days the newly-arrived army overran the

drains the lake, and this lies considerably to the Orontes west of the defile between the
lake and the Orontes. Therefore the Crusaders passed it, but not the Turks. If the
narrow neck had been west of where the Iferin falls into the main river, we might
suppose that this was the stream which the Turks crossed. But the fact being the
reverse, William must be wrong. Apparently he was making some confusion with
the Iron Bridge over the Orontes six or seven miles east of the camp.

1 For a description of the walls of Antioch, their topography, and the Crusaders
entry, see chap. vii. of Book VI.

280 THE ART OF WAR IN THE MIDDLE AGES [1098

plain of Antioch, and forced the Crusaders to keep within their
old camp and the newly-captured city. The position of the
Franks was dangerous, as the citadel was still holding out.
Shems-ed-Dowleh, the son of Baghi-Sagan, and the wrecks of
the garrison had sought refuge in it when the place fell. They
had to be watched, and their sallies were only restrained by the
erection of forts on the precipitous heights leading up to their
place of vantage.

Battle of Antioch, June 28, 1098.

The position of the Crusaders, therefore, was hardly altered
for the better. Though they had taken Antioch, they were
themselves practically besieged by Kerboga. After waiting for
more than three weeks, during which things went from bad to
worse, and the famine which had made the winter so miserable
broke out for a second time, the Prankish chiefs saw that they
must fight or perish. They accordingly resolved to sally out
from the city by the Bridge-Gate and attack the Turks, whose
main body lay encamped in the plain to the north of the Orontes.
On this occasion they resolved to combine horse and foot in
their line of battle. It was absolutely necessary to make the
experiment : when the mounted men had dwindled to a very few
thousands, 1 they could no longer suffice to cope with the vast
army of Kerboga. There were many hundreds of knights of
approved valour who had lost their chargers, and it would have
been absurd to leave them out of the fight. If they marched
on foot, they would serve to give confidence and steadiness to
the untrained and untrustworthy infantry. 2 The infusion of
mailed men of approved courage and high rank would naturally
diminish the tendency to panic and disorder which made the
Western foot-soldiery of that day so helpless before the enemy.
Accordingly, the greatest care was taken to bring the infantry
into fighting trim : it was divided into small bodies placed under
competent leaders, and in all probability sorted according to the
character of the arms it bore. We hear most about the archers
and arbalesters, though there must have been thousands who
were not armed with these missile weapons. But for fighting

1 William of Tyre s number of one thousand and fifty is incredibly small. We
know that on one occasion and another the Crusaders had captured more than two
thousand chargers from the enemy.

2 Albert of Aix, iv.

1098] ANTIOCH: SORTIE OF THE CRUSADERS 281

enemies like the Turks, who placed their whole confidence in
their arrows, troops armed with long-range weapons would be
especially valuable. We have already had occasion to remark
more than once that the foot-archer is the most efficient check
on the horse-archer, because he can carry a larger weapon with
a longer range. Probably Western archer} , save in some few
districts, was not very efficient, yet it would still be of much
avail against the Turk. Of course, however, it was not by the
arrow that the crusading chiefs intended to win. The infantry-
were to be mere auxiliaries in the fight, and the charge of the
mailed horsemen was to deal the decisive blow. The battle
order was to consist of lines of infantry with small bodies of
cavalry in the rear of each, the former to open the fight, the
latter to end it.

On Monday, June 28, the army was drawn up in the streets
of Antioch, corps by corps, with the van lying just inside the
Bridge-Gate, and ready to sally out when the signal should be
given. It is most difficult to make out the exact disposition of
the various divisions ; various chroniclers give almost every 7
number between four and thirteen for them. Of the two really
good authorities, Raymond d Agiles and the Gesta Fmncorum,
the first gives eight, the second six. 1 But Raymond adds the
curious statement that " the princes had arranged eight
corps, but when we had got outside the city, with every man
able to bear arms put into the ranks, we found there were
five more corps, so that we fought with thirteen instead of the
original eight." 2 Comparing the elaborate list of names in each
division which two or three of the chroniclers give, we find that
there is little or no dispute about the first four and the last two
of the corps, but that in the middle of the line we have a difficulty
in reckoning the bodies formed by the Burgundian, South-French,
and Provencal contingents. In these parts of the army, which
were led by Godfrey of Bouillon and Adhemar, Bishop of Puy,
some reckon only two large masses, some four, some as many as
seven smaller ones. The general result of our investigation
seems to be that though the original intention had been to com
pose the centre of two corps of Lorrainers and Burgundians, and

1 Fulcher of Chartres gives four, the Cesla six (as also many chroniclers who
copy the Gesta), Anselm of Ribeaumont and Orderic Vitalis seven, Raymond eight,
(or thirteen), Gilo nine, Albert of Aix and William of Tyre twelve.

Raymond d Agiles, p. 287.

282 THE ART OF WAR IN THE MIDDLE AGES [1098

the left of two corps of Aquitanians and Provencals, yet on getting
intothe plain these two grand divisionswere re-formed respectively
into three and four brigades. If we can trust Raymond d Agiles,
it was an inspiration of the moment, caused by the fact that the
numbers of these contingents had been underrated in the council
of war which drew up the order of battle. 1

Summing up our authorities, we may conclude that the
probable order was as follows: (i) North-French, under Hugh,
brother of the King of France ; (2) Flemings, under their Count
Robert; (3) Normans, under Robert, son of William the Conqueror.
These three divisions formed the right wing, and headed the
column. The next to issue from the gate were the contingents

o o

(4) (5) (6), three corps of Lorrainers, Burgundians, and Mid-
French, under the general command of Duke Godfrey the other
two bodies in this division seem to have been under Reginald
Count of Toul, and Hugh Count of St. Pol. The whole was
destined to form the centre in the fight. Next were (7) (8) (9) (10)
four corps of Provencals, Aquitanians, and West-French, under
the general command of Bishop Adhemar, the three other leaders
in this wing (the left) being Raimbaud Count of Orange, Isoard
Count of Die, and Count Conan the Breton. 2 Finally, fn)
Tancred and Gaston de Be arn, with Apulians and Gascons; and
(12 ) Bohemund, with the main body of the Normans of Italy and
Sicily. The last-named corps was to form a reserve division
behind the others, and to guard the rear when all should have
defiled over the bridge and into the plain. 3

The only useful notice which we have concerning the number
of men in each division is Albert of Aix s statement that Duke
Godfrey s own corps consisted of no more than two thousand
men, horse and foot all told. Albert grossly exaggerates the
weakness of the Franks in all his account. But Godfrey s corps
may have been smaller than the rest we are told at least that

1 The original design, according to Raymond, was to make four grand divisions
(i) North-French, P lemings, and Normans ; (2) Lorrainers and Burgundians; (3)
Aquitanians and Provencals; (4) Sicilian and Apulian Normans (Raymond, p. 283).
Each grand division was composed of "duo ordines duplices," i.e. two corps in
two lines, one of foot and one of horsemen. So there were to be eight corps
in all.

" Raymond of Toulouse should have shared the command of this wing with
the bishop, but was left behind in Antioch to observe the citadel with two hundred
knights. He was too sick to ride that day.

3 All this array is given with reservations ; there may be, and probably are, faults
in it. But the divergences of the chroniclers only allow us to give probabilities.

PLATE I ll

SIEGE AND BATTLL

A.N.T-l.QC;Mk

fflct I097-Junel098

1098] ANTIOCH : ADVANCE OF THE CRUSADERS 283

Bohemund s corps was much larger, 1 Yet it would be hazardous
to put the full force of the army which marched out at more
than from twenty-five thousand to thirty thousand men, of
which one tenth, perhaps, may have been mounted. We know
that the divisions in the front line covered, when deployed, a
front of over two miles. Allowing for intervals between the
corps, this would require twenty-one thousand foot-soldiers six
deep ; the formation is not likely to have been thinner than that
depth, as the infantry were known to be unsteady, and could not
have been trusted to stand firm if arrayed only in three or four
ranks. Adding a few thousands more for Bohemund s corps and
the cavalry, we may reach thirty thousand altogether.

Kerboga s camp lay to the north-east of Antioch, under the
hills which rise abruptly two miles beyond the Orontes. The
Crusaders were resolved to march straight upon it, after crossing
the bridge and deploying into line. Thus their front would lie
east-north-east, with the Orontes close to their right flank and
the hills close to their left. It was arranged that as each corps
passed the bridge it should deploy in order on the plain beyond,
the van halting immediately that it had crossed and forming
close to the river, the centre prolonging the line northward,
and the left which would have far the longest space to march ^
reaching to the foot of the hills. The danger of this plan lay
in the possibility that Kerboga might let one or two corps pass,
and then fall upon them while the rest were struggling out of
the gate and on to the bridge. If he had done so, the fate of the
Crusaders might have been like that of Earl Warrenne s army at
Cambuskenneth, 2 the van might have been battered to pieces
before the main body could force its way to the front. But the
Emir preferred to let the whole Christian army march out into
the plain, where he hoped to have room to outflank and surround
them in the usual Turkish fashion. 3 " The farther they come out
the more they will be in our power " are said to have been his
words. 4

1 Albert of Aix, iv. 47. But Albert much overstates the misery of the Crusaders,
says that many knights rode to battle on asses, and that there were only two hundred
horses in the army. He was not an eye-witness, and his informants exaggerated
grossly.

2 See chap. i. of Book vn.

3 Albert of Aix, not an eye-witness, and William of Tyre, writing a century later,
say that Kerboga sent out a corps of archers to hold the ground just across the bridge,
and prevent the Franks deploying. No good authority mentions such a move.

4 Gesta Franconiin, xxix. 3.

284 THE ART OF WAR IN THE MIDDLE AGES [1098

It was only when corps after corps came pouring over the
bridge, and it became evident that the Christians were far more
numerous than he had supposed, and might when drawn up fill
the whole breadth of the plain, and prevent any turning move
ment, that Kerboga roused himself and put his army in motion.
Apparently, the divisions of Hugh, the two Roberts, and Godfrey
were already in line, and that of Bishop Adhemar was beginning
to take ground to their left, when the Emir endeavoured to throw
his right wing across the level ground at the foot of the hill where
the remaining Christian corps were intending to draw up. Fifteen
thousand horse, filing along the foot of the hills, succeeded in
getting round the flank of the Crusaders and placing themselves
perpendicularly to the still incomplete left wing. These are
said to have been the Turks of Kilidj-Arslan of Roum, and
Ridwan of Aleppo. 1 The corps of Bishop Adhemar and the
three which followed it had the greatest difficulty in fighting
their way into line with the centre and right. 2 But they
succeeded in doing so, and thereby cut the army of Kerboga in
two, the detached corps under the Sultan of Roum becoming
completely separated from the main force. 3 Hence the battle
consisted of two independent fights one between the main
Christian army and the Turkish centre and left, the other
between the detached right wing of the Infidels and the Christian
reserve under Bohemund. For the latter prince, seeing the
fatal consequences which might ensue if Kilidj-Arslan attacked
Godfrey and Adhemar in the rear, hurried forward and deployed 4
his corps facing westward, with their backs to the main body.
His position must have been parallel with the divisions of
Adhemar and Godfrey, i.e. behind the left centre of the main
army. Godfrey, according to some of our sources, hastily sent
the corps of Reginald of Toul to assist in keeping off the attack
from the rear.

In the main battle the Crusaders won a complete victory

1 But this we have only from two secondary chroniclers, William of Tyre and his
authority, Albert of Aix.

- Raymond d Agiles, p. 286 : " We had to strive hard in the space at the foot of
the hills, as the enemy was trying to envelop us, and had their largest corps in front
of us."

3 " Denique divisi sunt Turci : una pars ivit contra mare ; alii steterunt contra nos."

4 Ralph of Caen compares the Christian army so arrayed to the snake of the fable
which had a head at each end, or to a monster with two faces, and specially mentions
that Bohemund "turned his back to his friends, and his face to his enemies"
(pp. 169, 170).

1098] ANTIOCH : VICTORY OF THE CRUSADERS 285

with astonishing ease. Kerboga was a bad general, and his
colleagues, the Emirs of Damascus and Aleppo, were mistrustful
of him and of each other. Moslem historians tell us that at the
moment of action a great body of Turcoman auxiliaries, with
whom Ridwan of Aleppo had been tampering, treacherously took
to flight and threw the whole line into confusion. It is certain,
at any rate, that when the Christian armies advanced in steady
line, with archers in front and knights behind, the Turks retired
from their first station towards their camp. There they again
made a front, but there was no further chance of putting their
usual tactics into play, since the Franks filled the whole plain from
the river to the hills, and could not be outflanked. Their first
retreat had some semblance of order, but when pressed again the
Infidels broke up more and more, and finally fled at full speed, the
cowardly Kerboga at their head. They made off by the road
between the Orontes and the lake of Antioch, abandoning their
camp and the masses of unfortunate camp-followers to the sword
of the Franks. " No man of rank fell," says Kemal-ed-din, "but
there was a horrid slaughter of our foot auxiliaries, grooms, and
servants." 1

The combat in the rear had been much more serious. The
Turks of Roum and Aleppo fell with fury upon Bohemund s
corps, where the infantry threw themselves into a dense circle
and did their best to hold firm. They were in great danger,
exposed to the Turkish arrows and attacked at intervals by
parties who abandoned their usual tactics and charged in with
the sword. The corps of Reginald of Toul when it came up was
also assailed with great vigour, and suffered heavy loss : accord
ing to some authorities, nearly the whole of its infantry was cut
to pieces. But presently the Turks saw their own main army
flying, and knew that the battle was lost. Apparently, too, the
victorious Crusaders detached more troops to help Bohemund.
Firing the grass to cover their retreat, 2 the Infidels made off west
wards towards the sea, and left the corps of Bohemund and
Reginald maltreated, but still holding firm. The diversion had
utterly failed because of the cowardly conduct of Kerboga and
the main army.

1 See the quotations from Kemal-ed-din, Abulfeda, and Abulfarag in Michaud s
Hibliothique des CroisaJes, iv. 9.

2 We need not believe the unlikely story about the smoke signals concerted
between Kerboga and his lieutenants.

286 THE ART OF WAR IN THE MIDDLE AGES [1098

The battle of Antioch filled both Frank and Turk with
wonder. The Christians marvelled at their own victory : with
inferior numbers and men debilitated by famine and the heat of
the Eastern sun, they had swept the Infidels before them in a
single desperate charge. They attributed their success wholly
to supernatural causes : the Holy Lance borne before Bishop
Adhemar, they said, had turned the enemy to flight, and the
hosts of heaven, headed by St. George and St. Demetrius, had
been seen co-operating in the victory, " clothed in white, riding
on white horses, and bearing white banners before them." The
Moslems attributed the victory of the few over the many, the
famished over the well-nurtured, to the inscrutable will of Heaven,
desiring to chastise the emirs for their sins.

To those in search of more earthly explanations the meaning
of the fight is obvious enough. The Turks had fought once
more in a space too confined for their usual tactics : the right
wing of the Franks rested on the river, and could not be turned.
Their left wing, the point in real danger, broke through the hordes
sent to surround it and got in touch with the hills. When both
flanks were protected, they had only to execute a straightforward
charge, and the Turks must choose between the hand-to-hand
combat, which they always disliked, and flight. They chose the
latter alternative, and the day was won. If the rear had not been
guarded by Bohemund and Reginald of Toul, a disaster might
well have occurred ; but while the attack on the rear was held in
check, the main Turkish army could do nothing.

The lessons of Dorylsum and Antioch should have remained
fixed deep in the minds of both Christian and Moslem, but we
shall see that only the keenest minds on each side suspected the
meaning. Both parties for the next hundred years frequently
repeated their original blunders the Turks that of fighting in
cramped ground, the Franks that of failing to combine horse
and foot in due proportions.

Battle of Ascalon, August 14, 1099.

There was no general engagement of importance beside the
battle of Antioch during the conquest of Syria. The rest of the
history of the year 1098-99 consists of a series of sieges, with
which we shall have to deal when treating of the siegecraft of the
early Crusaders. It was not till August 1099 that another battle
in the open field was fought, and this time the enemy was not the

BATTLE OF ASCALON 287

Turk, but the Fatimite ruler of Egypt, El-Mustali Abul-Kasim
Ahmed. The Egyptians had been in possession of Palestine at the
moment of the arrival of the Crusaders, and it was from them that
Jerusalem had been wrested. Shortly after it had fallen (August
1099 , El-Mustali sent his general, El-Afdal, with a large army
to drive off the Crusaders and recover the Holy City. The
forces of El-Afdal were unlike those with which the Crusaders
had hitherto had to contend. They resembled the old Saracen
armies with which the Byzantines had so often fought : there
were many thousand infantry, all black Soudanese, armed with
bows and iron maces (or flails, ; while the cavalry consisted partly
of Moorish and Bedouin light horse, partly of mailed troopers of
the Caliph s regular army. All of these were spearmen, and not
archers like the Turks. Having long been at war with the
Turkish princes of Syria, El-Mustali had no help to expect from
them. But there seem to have been a few mercenaries of
Turkish blood in his ranks. The whole army is estimated at
the usual vague figure of three hundred thousand by the crus
ading writers. It may possibly have reached in reality some fifty
thousand or sixty thousand in all. 1

The Franks marched out from Jerusalem on August 13,
with five thousand horse and fifteen thousand foot. 2 The knights,
it will be observed, were all remounted since their victorious
march through Syria, having found Arab horses for themselves
to replace their lost chargers. Hence the proportion of cavalry
to infantry is far larger than it had been at Antioch. When
they arrived in the neighbourhood of the enemy, they feared to
be surprised and surrounded on the march, and formed the army
in nine small corps, each composed of foot and horse. These
corps marched three abreast, so that whether attacked in front,
rear, or flank there would always be three divisions to face the
shock, three to sustain them, and three more in reserve.- So far,
however, were they from suffering from any such danger, that
they themselves surprised and captured the flocks and herds of
El-Afdal s army, which were grazing, under the guard of three
hundred men, in a valley some miles north of Ascalon.

1 The Moslem Ibn-Giouzi says no more than twenty thousand. This is probably
an understatement. Perhaps it only includes the Caliph s regular troops.

- So say the Princes in their letter to the Pope. The usually trustworthy Raymond
gives the number as twelve hundred knights and nine thousand foot only.

3 Raymond of Agiles, p. 388.

288 THE ART OF WAR IN THE MIDDLE AGES [1099

The fugitives soon brought the news to the Egyptian vizier,
who prepared to fight next morning. He took up his position
on the shore north of the town of Ascalon, with his left wing
resting on the sea and his right on the hills, which here run some
two miles inland from the water. In his rear was the town with
its orchards and plantations, and the camp pitched outside the
Jerusalem gate. He placed his Soudanese archers in the front
line, his regular cavalry behind them. On the right a corps of
Bedouins were to endeavour to encircle the enemy s flank : on
the left the sea rendered any such attempt impossible.

On August 14 the Crusaders came in sight, marching down
the sandy plain between the water and the hills, which gradually
broadens as it approaches Ascalon. When they came into the
neighbourhood of the Egyptians, they proceeded to deploy into
line from the order of march in nine columns which they had
hitherto kept. Apparently the front three columns, under
Robert of Normandy, halted, while the second line, under
Raymond of Toulouse, took ground to their right next the sea,
and the rear line, under Godfrey of Bouillon, filed off to the left
and took post towards the hills. 1 The whole nine corps thus
came up into a single line, and no reserve was left behind : in
each corps the infantry were formed in front, the cavalry to
the rear.

When the two armies were within bowshot, the Soudanese
opened fire on the Crusaders, " falling on one knee to shoot,
according to the custom." 5 At the same time the whole Saracen
army struck up a horrible din of trumpets and nakers to daunt
the Christians, and the Bedouin squadrons rode out to the right
to encircle the left flank of the enemy. The opening of the
fight by the Infidels is described by one good authority as
resembling " a stag lowering his head and extending his horns
so as to encircle the aggressor with them ; " 3 but there can have
been no attempt to do this on the western flank, where the sea
was too close to allow of any such manceuvre.

The turning movement was easily stopped by Duke Godfrey,
who charged with his knights and easily rode down the light-

1 This deployment seems certain from the words of the Cesla Fratuorum, xxxix.,
which say that Raymond fought on the right, Godfrey on the left, and all the others
between them : it names Robert of Normandy, Tancred, and Robert of Flanders as
among those who commanded in the centre, but says that " alii omnes " were there also.

Albert of Aix, vi. s Fulcher of Chartres, xix.

PLATE VI II.

ASCALOX.

Au!J.14.109.

V-

*v q

Q r,

. #"

9 P P 9

P P

D P D D
O

O D

Robert
Fulcoy

BATTLE or HAB.

Infantry

The

An lio chelae
Barons

Rons of
Tripoli

Christians E-

Turks P 6 P 6

BATTLE OF ASCALON 289

armed Arabs. At the same time, a general advance was made
all along the line, the Christian cavalry charging before the
Soudanese had time to discharge their bows more than once. 1
In every quarter the Egyptian foot were rolled back on to their
horse, and the whole army fell at once into complete confusion.
They seem to have made a very poor resistance, and the
Crusaders penetrated everywhere through their line. Robert
of Normandy slew the vizier s banner-bearer, Tancred charged
right into the Moslem camp, Raymond of Toulouse hurled the
hostile left wing into the sea. Some of the Egyptians got into
the town, others fled away to the south, some even swam out to
their fleet, which lay moored off the shore. But thousands were
slain on the field, many more crushed to death as they tried to
force their way in at the crowded gates, and a considerable
number were drowned. For some hours after the fight ended, the
Crusaders were hunting down fugitives who had concealed them
selves in the orchards or even climbed up trees to hide in their
tops. 2 They captured the hostile camp with vast spoils, and
narrowly escaped slaying or taking the Emir El-Afdal. The
victory was a far more crushing one than either Antioch or
Dorylaeum, for the enemy had not so good an opportunity of
getting off, and suffered much more severely. His wretched
infantry were completely cut to pieces.

Obviously the Egyptians were an enemy to be treated far
more unceremoniously than the Turks. They tried to face
the heavy cavalry of the Crusaders with less efficient horsemen,
armed only with the spear, and their infantry were in no wise
superior to that of the Franks. Hence in an open field they
were sure to be beaten, even though their numbers were largely
superior, as undisciplined Asiatic armies have usually been when
they meet Europeans under competent leaders. The Crusaders
came to hold the Egyptians in such contempt that they neglected
the most common precautions against them, and would attack
them if they were but one to ten, and even in most unfavourable
ground. This rashness was chastised a few years later at the
battle of Ramleh, where King Baldwin suffered heavily at the
hands of the despised foe.

1 Albert of Aix, who was not an eye-witness, gives an unintelligible account of the
fight : I follow the Cesta, checked by Fulcher and Raymond.

2 This is mentioned by the Arab chronicler Ibn-Ghiouzi as well as by several of
the Christian writers, e.g. Albert of Aix.

2 9 o THE ART OF WAR IN THE MIDDLE AGES [1101

Battles of Ramleh, September 7, 1 101, and May 1 102.
It had taken the Vizier El-Afdal two years to recover from
the shock which the defeat of Ascalon had imprinted on his
mind. But in the autumn of noi he sent out a new army to
invade Palestine: Ascalon, still in Egyptian hands, served as
a base for the operation of the host. Baldwin I. had now
succeeded his brother Godfrey, who had only worn the crown
of Jerusalem for a year. His little kingdom consisted of
nothing more than his capital and the three seaports of Jaffa,
Arsouf, and Caesarea : the last two he had only just subdued by
the aid of a Genoese fleet in the summer of uoi. Baldwin all
his life through was a rash and reckless leader, one of the
typical Prankish generals on whom the Byzantine authors pour
so much scorn. The Egyptian troops were not so strong as
they had been at Ascalon, but still very numerous: Fulcher of
Chartres estimates them at eleven thousand horse and twenty-
one thousand foot; 1 the Moslem chroniclers state that they
were led by the Emir Saad -ed - Dowleh. Baldwin, however,
resolved to march against them with the scanty force that he
could collect in Jerusalem at a few days notice. He would not
wait for outlying parties of his own followers, much less sit still
for weeks while reinforcements should be summoned from
Antioch or Edessa. The Egyptians having moved out from
Ascalon, Baldwin left Jerusalem and marched down to Jaffa on
September 5. The Egyptians did not come to meet him there,
but pushed in between the king and his capital, marching to
Ramleh a point equidistant from Jerusalem and Jaffa. Thither
Baldwin followed them with two hundred knights, sixty sergeants,
hastily mounted on borrowed horses to swell the number of his
cavalry, and nine hundred infantry. He divided this little army
into six corps, each containing both horse and foot, and marched
recklessly into the midst of the Egyptian host, who had been
warned of his approach, and had formed up with a front far out
flanking the Crusaders on both wings. Baldwin and his little
band plunged in among them "like fowlers into the midst of a
covey of birds." Of the exact order of the Franks we have no
further particulars than the vague statement of Fulcher, that
they were "arrayed according to the rules of military art."
Even the simple critics of the twelfth century, however, were

1 rulcher, chap. xxvi.

noi] FIRST BATTLE OF RAMLEH 291

ready to grant that Baldwin s attack was made with a rash
disregard of possible dangers. 1

It seems that when the lines were a thousand paces apart 2
the knights put spurs to their horses and, leaving the foot-soldiery
behind, dashed at the Egyptians. Only Baldwin himself, with
one of the six corps of cavalry, forty or fifty riders at the most.
remained in the rear with the infantry. When the Christians
charged, the Egyptian host folded in its wings and fell upon
the Crusaders on all sides, attacking the infantry no less than
the horse. The two right squadrons of the knights were taken in
flank, 3 and completely rolled up, so that hardly a man escaped.
The other three were swallowed up among the multitude of the
Infidels, and seemed likely to succumb also, when Baldwin and
his small reserve of horsemen dashed into the thickest of the
fight and gave the necessary impulse to the surging mass. The
Egyptian centre broke and fled, and presently their victorious
left wing also quitted the field. While the battle was being settled
by the cavalry fighting, the infantry in rear had been beset on
all sides by the horsemen at the extreme wings of the Egyptian
host. They were very roughly handled, so that Fulcher acknow
ledges that " while the Christians were victors in front, they
came off the worst in the rear." If Baldwin s victory had been
delayed a few minutes, the infantry would probably have been
entirely broken up and cut to pieces. As it was, the success had
been so dubious that a body of five hundred Arabs from the
victorious left wing of the Egyptians had ridden up to the walls
of Jaffa, displaying the shields and helms of the crusading
l:nights whom they had slain, and had shouted to the garrison
that Baldwin and all his host had perished. These troops were
returning, ignorant of the rout of their main body, when they
rode by accident into the midst of the Christians and were
mostly cut down.

The losses in Baldwin s army were very heavy. Eighty
knights had fallen a third of the whole cavalry: no doubt they
nearly ail belonged to the two squadrons which had perished at
the opening of the battle. 4 A much larger number of the.

1 " Minus caute," says Ekkehard in his Hierosolymila, 2 Lkkehard.

J Ekkehard says that one squadron only was cut to pieces by a flank attack j
Fulcher (a better witness) that two were destroyed. Albert of Aix, exaggerating
fearfully, makes four perish, and says that the king won the battle with his own forty
knights alone.

4 Fulcher, p. 125.

292 THE ART OF WAR IN THE MIDDLE AGES [1102

infantry had also been cut down. The loss of the Egyptians is
put as high as fifteen thousand men an impossible figure ; they
probably did not suffer much more than their adversaries. The
Moslem historians give no details, but allow that their chief Saad-
ed-Dowleh was left on the field killed by a fall from his horse,
as had been prophesied to him long before by an astrologer. 1
The whole fight had only occupied a short hour. 2

After having been within an ace of destruction in September
1 101, it is astonishing to find that Baldwin repeated his reckless
tactics in an aggravated form on the very same battlefield, only
eight months after. 3 In May 1 102 the Egyptians once more broke
forth from Ascalon and marched on Ramleh, where they pitched
their camp. Baldwin set out against them with his military
household, without waiting for any reinforcements from the out
lying towns of his little kingdom. He picked up at Jaffa a band
of pilgrim knights, survivors of the unfortunate Crusade of 1 101,
who were just embarking to return to France : they were led by
Stephen of Blois and Stephen Count Palatine of Burgundy. This
gave him no more than two hundred lances ; nevertheless, he
marched straight on Ramleh, believing (it is said) that the enemy
were only a raiding party some eight hundred or a thousand
strong: 4 as a matter of fact they were a whole army, about as
large as that which had been beaten on the same spot in the
previous year. 5 "It was pride and presumption that led the
king," says Fulcher, " not to wait for more troops, not to move
to the battle in proper military order, not to listen to any warn
ing, not to wait for his foot-soldiery, and not to stop in his march
until he saw before him, and far closer than he wished, such a
huge multitude of the enemy." With no more than his t\vo
hundred knights Baldwin rode straight at the hostile centre,
hoping to repeat his exploit of the previous year. But the odds
were too great, and this time he had no infantry with him to
protect his rear and take off some of the pressure. The Franks
were engulfed in the hostile mass, and slain off almost to a man.
Baldwin and a few more cut their way out of the melee, but the

1 See the chronicle quoted on p. 17 of vol. iv. of the Bibliothtque dcs Croisades.
- Fulcher.

3 In all that follows I have taken Fulcher as guide, not Albert of Aix, who varies
hopelessly from Fulcher s tale, and was not, like the fonner, on the spot.

4 Fulcher, chap, xxvii. p. 135.

5 Ibid. : Twenty thousand horse and ten thousand foot, the latter all Soudanese "
numbers grossly exaggerated, we r.ecd not doubt.

iio2] FINAL DEFEAT OF THE EGYPTIANS 293

Counts of Blois, Burgundy, and Vendome, and more than a
hundred and fifty knights, were left dead on the field. It was
possible to despise too much even an Egyptian army, and the
king had to learn that headlong courage of the most desperate
kind is not enough to compensate for a disparity of numbers in
the proportion of a hundred to one.

After several narrow escapes, Baldwin reached Arsouf, and
from thence sailed to Jaffa in the galley of Godric, an English
adventurer. There he received reinforcements which would have
reached him in time for the battle if he had only consented to
wait a few days eighty knights from Galilee under Hugh of
St. Omer, ninety from Jerusalem, and a considerable body of
infantry. Some weeks later there arrived a great pilgrim fleet
of two hundred sail from England and Germany, under Harding
the Englishman, and the Westphalians Otto and Hademuth.
The crews landed armed, and with their aid Baldwin felt strong
enough to march out of Jaffa to face the Egyptians once more
in the open field. This time he had learned his lesson, and
combined his cavalry and his infantry. The foot-soldiery, no
less than seven thousand strong, owing to the reinforcements
from the fleet, were armed mainly with bow and arbalest, and
kept the enemy s horse at bay, while the knights, a thousand
strong, charged out again and again whenever the Egyptians
tried to close, and beat back every attack. At last the Infidels,
finding they could make no impression on the Franks, rode off,
abandoning their camp to the spoiler. They do not seem, how
ever, to have lost any very great number of slain : the estimate
of the crusading chroniclers is only three thousand a very modest
number compared with their usual figures. The victory was
indecisive, but it saved Palestine, while a defeat would have made
an immediate end of the Latin kingdom.

We should have been glad to have had more particulars as
to the service of the English in this fight. They must have
been present in considerable numbers, but none of our native
chroniclers tell of Godric and Harding unless, indeed, the
former is the Godric of Finchale who afterwards became a
hermit and a saint.

CHAPTER IV

THE TACTICS OF THE CRUSADERS continued

Section II. Tactics of tJie Later Battles : Victories at Hctl\
Hazartli, Marj-es-Safar, Arsouf, Jaffa.

A S our task is not to write a history of the kingdom of
.ly. Jerusalem and its wars, but to indicate the main military
tendencies of the crusading age, we must not attempt to give in
detail each of the numberless fights of Frank and Moslem, but
only to comment on such of them as show features of import
ance. Speaking in general, we may say that the same points
of interest which we have observed at Dorylsum, Antioch,
Harenc, Ascalon, Ramleh, and Jaffa, are to be found repeated
in all the fights of the twelfth century.

Against the Turk the Crusaders were generally successful if
they took care (i) to combine their cavalry with a solid body of
infantry armed with missile weapons ; (2) to fight on ground
where the Infidel coulcl not employ his usual Parthian tactics of
surrounding and harassing his enemy ; and (3) to avert the
danger of starvation by carrying a sufficient store of food.
Against the non-Turkish Moslems, such as the Egyptians, the
Crusader was far more certain of success; he had only to use
the common military precautions, and he might fairly count on
victory. The battles of the Franks with these less formidable
foes sometimes remind us of the early English battles in India,
where the few striking boldly at the many were so often
victorious in spite of ever) disadvantage. The one all-important
canon which had to be observed was that there must be
infantry on the field to serve as a support and rallying point for
the cavalry. If the foot-soldiery seldom won the battle, they
always made the winning of it by the knights possible.

If, on the other hand, the Frank chose to advance recklessly

234

1 1 19] BALDWIN II. IN NORTH SYRIA 295

into unknown ground in desolate regions, where he could be
surrounded, harassed, and finally worn out, or if he allowed his
class-pride to lead him astray, and left his infantry behind, he
was liable to suffer terrible disasters.

We have selected as examples of typical victories of the
Crusaders the battles of Hab (1119), Hazarth (1125), Marj-es-
Safar (i 126), Arsouf (i 191), Jaffa (1192). As instances of defeat
brought about by neglect of first principles, we may take the
fights of Carrhae (1104;, Tiberias (1187), Acre (1189), and
Mansourah (1250).

Battle of Hab, August 14, 1119.

On the 2/th of June 1119, Roger, Prince of Antioch, had
fallen with many of his knights in the disastrous fight of Cercp.
The victor, Il-Ghazi, Emir of Mardin, began to overrun the
whole principality of Antioch. To rescue it from the Infidels,
Baldwin II. of Jerusalem, with his vassal Pons Count of Tripoli,
hastened up from the south. The troops of Edessa also made
their way to join their suzerain, and when the wrecks of the
Antiochene army had united themselves to the host it counted
seven hundred knights and several thousand foot. Baldwin
advanced to relieve Zerdana, a castle to which Il-Ghazi had laid
siege. It fell before his arrival, but he was unaware of the fact
on the day of the battle. Il-Ghazi had also been joined by
reinforcements : his rival, Toktagin of Damascus, had agreed to
sink his private enmity, and had brought up a large contingent
of his o\vn riders, and some more levies from Emesa. The
Infidels mustered in all some twenty thousand horse : of foot
there is no mention ; the Turkish emirs generally depended on
their horse-archers alone. 1

Advancing by Hab towards Zerdana, Baldwin drew up his
army before daybreak in a less simple order of battle than was
usual among the Crusaders. The front line was formed by
three corps, each consisting of a body of cavalry supported by a
body of infantry, " that each arm might protect the other."
Behind the centre of this line was Baldwin himself, with the
knights of his household drawn up in three corps ; on his right
was the Count of Tripoli with his vassals ; on his left Robert
Fulcoy, lord of Zerdana, with the barons and knights of

1 All this comes from Gautier the Chancellor. \Villiam of Tyre, Fulcher, and
the rest are vague, and speak at second-hand.

296 THE ART OF WAR IN THE MIDDLE AGES [1119

Antioch ; another party of Antiochenes seems to have been
detailed as a rearguard, if our chief source, Gautier the
Chancellor, can be trusted. 1 The squadrons of Pons and
Robert were placed not parallel to, but somewhat to the rear
of, the front line, in order that they might defeat attempts to
turn its flanks, while the king could strengthen it if the main
pressure of the enemy was thrown upon its centre. Whether
by chance or design, this order bears a striking resemblance to
that which the Byzantine Leo the Wise advocates for use
against the Moslem. A comparison of the plan on p. 195, with
the sketch of the battle of Hab on p. 290, makes this clear in a
moment. The only difference is that Baldwin had infantry,
perhaps two thousand or three thousand strong, behind his first
line of horse, while Leo is describing the order of a division of
cavalry unassisted by any foot-soldiery. The nine squadrons,
each about eighty strong, were three in the first rank, three in
the second, one on each flank, and one behind. Il-Ghazi and
Toktagin seem to have hoped that they might be able to surprise
the Franks at daybreak, but when the sun rose Baldwin s
little host was already advancing in good order, and all the
war-cries and din of trumpets and nakers with which the
Infidels burst in upon it were completely thrown away.
Il-Ghazi resolved, therefore, to use the ordinary Turkish tactics,
and advanced in a half-moon, lapping round both flanks of the
Christian army. He himself, with the Mesopotamians, was on
the right, while Toktagin, with the men of Damascus and
Emesa, held the left. The Turks were well aware that the
greatest danger to themselves lay in the combination of infantry
and cavalry by the Christians. Il-Ghazi had therefore resolved
to do his best to overwhelm the front line of the enemy, and
prepared a desperate assault on Baldwin s centre, where all the
foot-soldiery were collected. They and the three squadrons of
knights in front of them were very fiercely assailed ; 2 the

1 This is the only way of construing " acie comitis Tripolitani a dexteris podta,
aciebus baronum a sinistris et post : jussu regis quibusque suo loco positis " (Gautier,
p. 460). If the Antiochenes had all been on the left of Baldwin in one body, we
should have had acie, not aciebus. Bongars prints the colon before et, but evidently
it should be before jussti, making no good sense if introduced after sinistris.

- That the knights were in front of the infantry and not behind, is shown by
Gautier s wording: " Turci, ambitiosi manum pedestrem prosternere, qua gravius
refrenabantur, cum hanc praccedentibus aciebus, et acies hac protegi videbant, vi
maxima . . . arcubus brachiis immissis, strictis ensibus, nostros percutiunt," etc.
The word fraeffJciifilnjs is conclusive (p. 461).

^I9] BATTLE OF HAB 297

irsemen were driven back on the foot, and the latter attacked,
not with the usual arrow-shower of the Turks, but by vigorous
charges home with lance and sword. The Prankish footmen,
when the knights were driven off, proved unable to bear up
against the Mesopotamians. Armed with missile weapons to
withstand the Turkish bo\v, they were less fitted for close
combat. They fought well, but began to fall into disorder, and
lost heavily.

Meanwhile, the fortune of battle on the wings had been
evenly divided ; on the right Il-Ghazi s men had assailed and
beaten back the Count of Tripoli, whose whole corps was
finally driven in and thrown on to the flank of Baldwin s own
division in the second line. On the left, however, Robert Fulcoy
and the Antiochenes had charged the men of Damascus with
such vigour that they had completely scattered them, and
driven them off in confusion. Robert might have won the day
by promptly charging the hostile centre from the flank. But
no such idea entered into his head ; his main desire was to
relieve his own castle of Zerdana, whose fall had not yet reached
his cars. 1 Accordingly he pursued the Damascenes for a space,
and then rode straight for Zerdana without making any further
attempt to join in the battle. He and his corps were absent
from the field during the remainder of the engagement.

Il-Ghazi s men on the other flank made no such mistake, but
closed in on Baldwin s second line. The fight now became
very confused ; the van and right wing of the Franks were
driven in on their centre in a disorderly mass, and it remained
to be seen whether the king would be able to save the day with
his reserve. Time after time he charged out with his knights
and drove off each swarm of Turks as it pressed in to complete
the victory. Whether the attack threatened front or flank or
rear, he and his chivalry were always at the point of danger.
Again and again the cry of " Holy Cross ! " and the impact of the
heavy squadron of men-at-arms drove back the Infidels from
their prey. 2 Towards evening Il-Ghazi gave up the struggle
and rode off, leaving Baldwin in possession of the field.

1 So Kemal-ed-din, who seems very well informed. Gautier the Chancellor
imagines that the news had already readied the Christians, which is improbable.
Robert would not have acted so if he had been aware of it (p. 460).

" " Rex, virili audacia fretus, qua parte hostium turmas magis vigere comperuit,
illic exclamando Sanctae Crucis protectionem et auxilium, velocissime irruit,
perfidos prostravit et in dispersionem impulit," etc. (Gautier, p. 461).

298 THE ART OF WAR IN THE MIDDLE AGES [1119

As he retired, the lord of Mardin came into collision with
the corps of Robert Fulcoy, returning tardily to join the king
after they had discovered that Zcrdana was already in the hands
of the Infidels. The Antiochene knights, marching in disorder
and without proper precautions, were easily dispersed by the
Turks, and Robert himself, falling from his horse in the flight,
was made prisoner. He was put to death by Toktagin in cold
blood some days after the battle : it is impossible to say
that his fate was undeserved, as his selfish abandonment of
his comrades at the outset of the battle merited the heaviest
punishment.

Baldwin, unaware of this disaster, held the field till night
and then retired to Hab. He returned next morning to bury
his own dead and strip those of the enemy. As the Turks had
entirely disappeared, he with justification regarded himself as
victor. The battle had in truth been indecisive ; but as the
enemy made no further advance against Antioch, the end for
which it had been fought \vas achieved. The losses had been
very heavy: Baldwin counted a hundred knights and seven
hundred footmen among the slain, and many more were dis
persed and did not rejoin for several days. The Turks had lost
from two thousand to three thousand horse.

The incidents of this battle, in which the fortune of the day
W.TS for a long time so equally divided, remind us of those of
Mont l hery, and Gautier s account of the flight of each side
may well stand beside the well-known passage in Commines.
" Our fugitives," he writes, "fled to Hab, to Antioch, and even
as far as Tripoli, reporting that the king and the whole army
had been exterminated. On the other hand, those of the Turks
who had been driven off the field (by our left wing) poured
into Aleppo, swearing that Il-Ghazi and Toktagin and all the
Turkomans had been slain to the last man." 1 If Baldwin
could claim that he had held the fie d at sunset, Il-Ghazi could
display as trophies one of the royal banners of the Latin
kingdom, torn from the king s squire who bore it, and
Robert Fulcoy and many other noble prisoners. That, after
massacring thirty of them, he then returned to Mardin to raise

1 In face of Gautier s explicit statement?, it is impossible to believe Kemal-e&lt;l-din s
allegation that at nightfall the Turks pursued the Christians to the gates of Hab. If
any of them did follow, it must have been at a safe distance, and as scouts rather
than pursuers.

1 1 25] BATTLE OF HAZARTH 299

more troops instead of pursuing his campaign, is a sufficient
proof that the claim of victory which he made was a very
empty one. But it seems to have deceived his chronicler,
Kemal-ed-din, from whose pages we should never gather
that Baldwin also could declare himself the conqueror in
the strife. The events of the succeeding months plainly
showed who was the real victor. Il-Ghazi returned home ;
Baldwin kept the field, and retook in the autumn Zerdana
and most of the other castles and cities which the Infidels
had captured after the death of Prince Roger.

This battle of Hab or Danit has many points of interest. It
shows us the Crusaders adopting for the first time a much more
complex order of battle than the simple line of infantry sup
ported by cavalry which they had displayed at Antioch, Ascalon,
and Ramleh. Baldwin, instructed by his many battles with
the Turk while he was but Count of Edessa, had employed
as king the fruits of his experience. The Turks, too, have learned
much : they no longer trust entirely to the bow, but charge
home vigorously \\ith sword and lance. They have come to
see that the Prankish foot-soldiery with their missile weapons
are even more dangerous to them than the knights, and devote
most of their energy to clearing away the infantry, not en
deavouring to shoot them down, an attempt in which Turks
seldom succeeded, owing to the inferiority of the horseman s
bow to the arbalest, but to ride over them with the lance.
That they succeeded on this occasion was apparently due to
Baldwin s mistake in drawing up his three squadrons of knights
in front of and not behind the infantry of the centre.

For a further development of the tactics of both sides, we
must advance a few years, to the battles of Marj-es-Safar and
Hazarlh.

Battle of Hazarth, June u, 1125.

At Hazarth, which was fought on June 11, 1125, Baldwin
seems to have returned to the simple order of battle of the
days of Antioch and Ascalon. He drew up his army in
thirteen small corps, each consisting of infantry and cavalry.
As there were eleven hundred knights and two thousand foot,
the squadrons must each have been about eighty strong (much
the same as at Hab), and the infantry divisions have mustered
somewhat over a hundred and fifty. These thirteen bodies were

300 THE ART OF WAR IN THE MIDDLE AGES [1126

divided into a centre and two wings : the right was composed
of the troops of Antioch, the left of those of Edessa and
Tripoli ; the centre, 1 the strongest of the three divisions, was
formed of the king s own vassals from Palestine. Presumably
the wings contained each four and the centre five corps, but
neither Fulcher nor William of Tyre, our two authorities,
definitely state the fact.

Il-Borsoki, the opponent of Baldwin, arrayed his fifteen
thousand horse in twenty-one corps, and pressed forward to
attack the Prankish infantry we have no mention of his
attempting any encircling movement after the usual Turkish
fashion. The interchange of missiles had gone on for some
short time, and close fighting had begun, when Baldwin gave
orders for a general charge of the cavalry. 2 The Infidels stood
firm for a moment, but, when the knights burst in among them,
lost heart, broke, and fled. Two thousand of them fell, while
the Christians only lost twenty-four. The proper combination
of infantry and cavalry had secured an almost bloodless victory.

Battle of Marj-es-Safar, January 25, 1126.

In the following year the Turks for the first time put foot-
soldiery in the field. They had evidently realised at last that
the combination of the two arms was more effective than their
own horse-archery. In January 1126 King Baldwin had crossed
the Jordan and advanced toward Damascus, harrying the land
far and wide, in revenge for a similar raid which Toktagin had
directed against Palestine in the preceding autumn. Against
him came forth the Atabeg chief and his son, bringing with
them not only their riders, but chosen youths trained to spring
up armed behind the horsemen, who, when the enemy drew
near, descended and fought on foot : for so they hoped to
disorder the Franks by attacking them with infantry on one
side and cavalry on another." 3 The Jehad had been preached

1 I do not think we are justified in concluding from Fulcher s (chap. Ixii.) calling
Baldwin s own corps " densior et posterior" that he was in a second Jine. Probably
only "last and largest" is meant. William of Tyre evidently read it so when
he wrote "in medio dominum regem," and not pone or fast. Fulcher says that
"Baldwin charged, bidding the rest follow, for they dared not commence the fray
before he gave the word." If he was in a second line, this would have been
impossible.

"The bows had been bent and the drawn sword was being used at close
quarters," says Fulcher (chap. Ixii.).

* Fulcher, last words of chap. Ixxi.

1 126] BATTLE OF MARJ-ES-SAFAR 301

in Damascus and its subject towns, and many thousands of un
trained citizens went out on foot to fight for Islam.

The armies met at Marj-es-Safar, not far from Damascus,
on the 25th of January, the day of the conversion of St. Paul.
Baldwin drew up his men in twelve corps, each containing
both infantry and cavalry, "that the two arms might give each
other the proper support." 1 The Damascenes were not in any
very great numerical superiority, save in the number of their
irregular foot-soldiery ; the Christian chroniclers confess that
the two armies were not very unequal, and do not ascribe
the usual vast preponderance to the enemy. But whether it
was that they were fighting close to their capital to protect
their own homes and families, or whether it was the unwonted
assistance of infantry which helped them, it is certain that they
made a much fiercer stand than usual. It was one of the
stiffest, though not the most bloody, fights in which the Franks
had engaged for many years. 2 Fulcher allows that for a space
the battle seemed going against Baldwin ; the arrow-shower
was too bitter, and " no part of body or limb seemed safe against
the shafts, so thickly did they fly." The host recoiled for a long
space, and it was only by a desperate rally in the afternoon that
it saved itself and resumed its advance. "But our king bore
himself well that day, as did all his knights and vassals, and
Almighty God was with them." 3 At dusk the Turks fled, and
the day was won. Two thousand Damascene horse and an
innumerable number of the Infidel foot had fallen ; of the
Christians twenty-four knights and eighty infantry only were
slain. William of Tyre, in his rather unsatisfactory narrative
of this battle, says that the Christian foot, fired by the example
of the king and his knights, charged the enemy at the supreme
moment along with the cavalry, and that they did most damage
to the Turks by shooting their horses, so that the dismounted

1 "Ordinatae sunt tarn militum quam peditum acies duodecim, ut ab alterutrn.
corroboretur caterva, si necessitas admoneret " (Fulcher, cap. Ixx. ). This can only mean
that foot might help horse and horse foot, not that each of the twelve corps might
help the other. It is hardly necessary to point out that allerttler can only be used
of two, not of many ; but I have seen several accounts of the battle by modern
authors where this simple rule of Latinity is neglected.

- William of Tyre is of course wholly in error when he calls it the most
dangerous and doubtful fight since the foundation of the realm (xiii. 18). At
Hab, only seven years before, the Christian losses were eight times as great and the
result far more uncertain.

3 Fulcher, cap. Ixx.

302 THE ART OF WAR IN THE MIDDLE AGES [1126

Infidels fell easy victims to the pursuer. 1 But it is not easy to
make out whether the infantry, as he conceived the fight, were
behind or in front of the knights. For, on the one hand, he
makes the foot-soldiers " pick up and carry back to the baggage
their wounded comrades, and set on their feet again those who
had been overthrown ; " while, on the other, they are said to shoot
the Turkish horses, so that the riders " fall into the hands
of their companions who follow behind." The first statement
seems to indicate that the knights had already charged over
the ground which the infantry were crossing; the second that
they were following behind them. But William is not always
happy in following his authorities for battles that took place
before his own day, and his picture here is decidedly confused.
In all probability the action began with the infantry in the first
line, and the cavalry in support. When it grew hot, the cavalry
must have charged out to the front, and in the final advance
the foot-soldiery must have been following in the wake of the
knights to complete the victory rather than preceding them.
It is a pity that we have not any detailed account of the
battle from Moslem sources ; if it existed, we might clear up
its difficulties, as we can those of the fight at Hazarth, by the
comparison of the two hostile chroniclers.

There are many Christian successes worth recording in
the years between Marj-es-Safar and the fall of Jerusalem in
nSj. But as they are not of any special tactical importance,
presenting merely the same features that we have already
noted, they may be passed over without any detailed narra
tion. The defeats of this period are more interesting than the
victories : notes on several of them will be found in the
succeeding chapter, where we treat of the causes of the many
failures of the Franks.

The battle which must next arrest our attention is the last
of the great triumphs of the Christians, and the most notable, as
it was won over the finest general whom the Infidels ever owned,
the great Saladin himself, commanding the most powerful and
most formidable if not the largest host which the Moslems
ever put into the field. The Christians, too, were in far larger
force than ever before in any battle of the Holy Land. It is

1 " Equis hostium sauciandis omnem dabant operam, eorumque sessore? sul-
sequentibus sociis parabant ad victimam " (\V. T. xiii. . 18). Thi?, I presume,
means shooting rather than stabbing the horses.

1191] RICHARD I. AT ACRE 303

fortunate that we have excellent accounts of the fight from both
sides, and that its topography can be easily ascertained. Every
detail of it is well worth study.

Battle of Arsouf, September 7, 1 191.

After a siege of nearly two years, Acre had been recovered
by the Franks on July 12, 1191. The garrison had laid down
its arms and surrendered to the kings of France and England,
after having protracted its defence to the last possible moment.
Saladin had done his best to succour the place, and delivered
perpetual assaults on the camp of the besiegers, but all to no
purpose. Seeing that there was no hope of relief, and that Acre
must fall by assault in a few days, the Emirs Karakush and
Mashtoub opened the gates, after promising that they would
induce the Sultan to pay two hundred thousand bezants as
ransom for the garrison, and also to restore the True Cross and
fifteen hundred Christian prisoners, the survivors of the disaster
of Tiberias, who were in chains at Damascus and elsewhere..

For some weeks after the fall of the great fortress, the
Christians remained encamped in and around Acre, while
Saladin still observed them from his camp on the mountain to
the east. The delay was caused partly by the exhaustion of
the victors, partly by the necessity for repairing the shattered
walls of the city, partly by the protracted negotiations concern
ing the ransom of the garrison. Meanwhile, Philip of France
took his way homeward rmidst the curses of the whole army,
swearing that on his return he would be a quiet and peaceful
neighbour to the dominions of the King of England. " How
faithfully he kept that oath is sufficiently notorious to all men, for
the moment that he got back he stirred up the land, and set Nor
mandy in an uproar." 1 He left the bulk of his army in the camp
under the Duke of Burgundy and Henry Count of Champagne.

The attempts to come to an agreement with Saladin failed
hopelessly. Into the ugly story of the massacre of the Turkish
garrison, when their ransom was not forthcoming, we need not
enter. On Tuesday, August 20, Richard and the Duke of
Burgundy beheaded the two thousand six hundred unfortunate
captives, and all chance of peace was gone. Two days after, the
crusading army set out upon its march.

1 Jliiiernrium Regis Ricardi, iii. 22 : " Quam vero fidelitcr hoc stcterit jrra-
mcnto satis inr.otuit universis," etc.

304 THE ART OF WAR IN THE MIDDLE AGES [1191

Richard had as his objective Jerusalem, whose recovery was
the main end of the Crusade. But to move directly from Acre
on the Holy City is impossible. The mountains of Ephraim
interpose a barrier too difficult to be attempted when an
alternative route is possible. For a march on Jerusalem the
best base is Jaffa, and to that place Richard resolved to trans
fer himself and his army. He accordingly arranged that the
host should march along the great Roman road beside the sea
by Haifa, Athlit, Caesarea, and Arsouf, while the fleet should
advance parallel with it, and communicate with it at every
point where it is possible to get vessels close to the shore. This
co-operation was all-important, for the army was lamentably
deficient in means of transport, and depended on the ships for
its food. So few were the beasts of burden, that a great part
of the impedimenta had to be borne on the backs of the
infantry, who loaded themselves with tents, flour-bags, and
miscellaneous necessaries of all kinds. Nearly half of them
were employed in porter s work, and thereby taken out of the
ranks when the host began to move forward. No food was to
be found on the way, for Saladin had already ravaged the shore,
and dismantled Haifa, Caesarea, and Arsouf.

It was obvious that the Crusaders would be harassed by
Saladin the moment that they started on their march. The
temptation to assail a host strung out in one thin column along
many miles of road would certainly draw the Turks down from
their strongholds in the hills. Richard had therefore to provide
an order of march which should be convertible at a moment s
notice into an order of battle. His front, rear, and left flank
were all equally liable to assault. Only the right would always
be covered by the proximity of the sea.

In view of this danger the king made the best disposition
possible. Next the sea moved the beasts of burden and the
infantry employed to carry loads. Inland from them were the
cavalry, distributed into compact bands and spaced out at equal
intervals all along the line of march. Inland again from tlie
cavalry were the main body of infantry, marching in a con
tinuous column, and so covering the whole eastern flank of the
army. Though the contingents were placed so close that no
gaps were left between them, they were for purposes of organisa
tion divided into twelve bodies, to each of which there was
attached one of the cavalry corps, which marched level with it.

u 9 i] RICHARD MARSHALS HIS ARMY 305

Thus there were twelve divisions of foot and twelve of horse ;
these smaller units were united into five main corps, of which
the exact composition is not easy to ascertain. The Templars
and the Hospitallers, who knew the country well, and had in
their ranks many " Turcopoles," i.e. horse-bowmen armed like
the Turks and specially fit to cope with them, took the van and
the rear, the two points of greatest danger, on alternate days.
With the centre division of the army moved the royal standard
of England fixed on a covered waggon drawn by four horses,
like the carrochio which the Milanese had used at Legnano a
few years before. The order of the various corps was, as we
gather, somewhat varied on different days. On one occasion
Richard and his own military household took the van, but
usually he reserved for himself no fixed station, but rode
backward and forward along the line of march with his house
hold knights, carefully supervising the movement of the whole
and lending aid wherever it was required. The heat was great,
September being not yet come, and the king was determined
not to harass the army by long stages. Accordingly he moved
very slowly, using only the early morning for the march, and
seldom covering more than eight or ten miles in the day.
Moreover, he habitually halted on each alternate day, and gave
his men a full twenty-four hours (or even more) of rest. Thus
the host took as much as nineteen days to cover the distance
of eighty miles between Acre and Jaffa. It is \vcll worth
while to give Richard s itinerary, in order to show the care
which he took of his troops.

7 liin-silay, August 22. From the neighbourhood of Acre to the river Eclus [2 miles].

Friday, Au^it;l 23. The army crosses the Belus [2 miles].

Saturday, August 24. Rest in camp and preparations for march.

Sunday, August 25. To Haifa [n miles].

Monday, August 26. Rest at Haifa.

Tuesday, August 27. From Haifa to Athlit, round the shoulder of Mount

Carmel [12 miles].

ll cilncsJar, August 28. Rest in camp.

Thursday, August 29. Rest in camp. The fleet arrives and lands stores.
Friday, August 30. From Athlit to El-Melat [Merla] [13 miles].
Saturday, August 31. From El-Melat to Caesarea [3 miles]. The fleet lands

stores and reinforcements.
Sunday, September I. From Caesarea to the "Dead River " [Nahr Aklidar]

[3 miles].

Monday, September 2. Rest in camp.
Tuesday, September 3. From the Dead River to the "Salt River" [Nahr Isken-

deruneh] [7 miles].

306 THE ART OF WAR IN THE MIDDLE AGES [1191

Wednesday, September 4. Rest in camp.

Thursday, September 5. From the Salt River through the Forest of Arsouf lo

Rochetaille [Nahr Falaik] [10 miles].
Friday, September 6. Rest in camp.

Saturday, September 7. From Rochetaille to Arsouf Battle [6 miles].
Sunday, September 8. Rest in camp at Arsouf.
Monday, September 9. From Arsouf to the Nahr-el-Aujeh [6 miles].
Tuesday, September 10. Nahr-el-Aujeh to Jaffa [5 miles]. The fleet lands fresh

stores.

Throughout the march the army was incessantly worried by
the attacks of the Turks, especially on the 25th and 3Oth of
August and the 1st and 3rd of September. The respite on the
26-/-8-9th was due to the fact, that while Richard had hugged
the coast from Haifa and gone round the shoulder of Mount
Carmel, Saladin had struck across country, passed the hills
farther east, and come down on to the neighbourhood of
Caesarea, before the Crusaders, moving slowly and on a longer
road, had drawn near the place. From August 30 to Sep
tember 7, on the other hand, he was always within a few
miles of them, waiting for his opportunity to dash down from
the hills if they exposed themselves. The author of the
Itinerarium gives an interesting description of the Turkish tactics
during these days : -

" The Infidels, not weighed down with heavy armour like
our knights, but always able to outstrip them in pace, were a
constant trouble. When charged they are wont to fly, and
their horses are more nimble than any others in the world ; one
may liken them to swallows for swiftness. When they see that
you have ceased to pursue them, they no longer fly but return
upon you ; they are like tiresome flies which you can flap away
for a moment, but which come back the instant you have stopped
hitting at them: as long as you beat about they keep off: the
moment you cease, they are on you again. So the Turk, when
you wheel about after driving him off, follows you home without
a second s delay, but will fly again if you turn on him. When
the king rode at them, they always retreated, but they hung
about our rear, and sometimes did us mischief, not unfrequently
disabling some of our men " (Itin. iv. 8). 1

1 NOTE ON THE BATTLE OF ARSOUF.

In my account of this fight I have followed the Itinerarium, Boha-ed-din, and
King Richard s letter to the Abbot of Clairvaux in Hoveden. All thcse^three
accounts fit into each other admirably. On the other hand, the narrative of Benedict

u 9 i] RICHARD S MARCH TO ARSOUF 307

Saladin, in keeping up this incessant skirmish along the
flank of the crusading host, was not merely endeavouring to
weary it out. Though he only showed small bands hovering
about in all directions, often but thirty or fifty strong, he was
always waiting close at hand with his main army. He kept it
hidden in the hills, hoping that the Franks would some day be
goaded into making a reckless charge upon his skirmishers. If
they would only break their line by a disorderly advance, he
would pounce down, penetrate into the gap, and sweep all before
him. King Richard, however, kept his men in such good order
that in the whole three weeks of the march they never gave the
Sultan the opportunity that he longed for. The king himself
and his meinie would occasionally swoop out upon bands that
came too close, but the main order of march was never broken.
Only on one occasion, on the first day of the march from the Belus
(August 25), did the Turks get a chance of slipping in while the
rearguard was passing a defile, and then the Crusaders closed
up so quickly that the assailants had to fly, after accomplishing
nothing more than the plunder of a little baggage. Boha-ed-
din s account of the Crusaders march is as well worth quoting
as the note on the Turkish attack which we have cited from
the Itinerarittin. He is describing the events of Saturday,
August 31.

" The enemy moved in order of battle : their infantry
marched between us and their cavalry, keeping as level and
firm as a wall. Each foot-soldier had a thick cassock of felt,
and under it a mail-shirt so strong that our arrows made no im
pression on them. They, meanwhile, shot at us with crossbows,
which struck down horse and man among the Moslems. I noted
among them men who had from one to ten shafts sticking in
their backs, yet trudged on at their ordinary pace and did not fall
out of their ranks. The infantry were divided into two halves :
one marched so as to cover the cavalry, the other moved along

of Peterborough is absolutely irreconcilable with them. He makes much of the
fighting turn on the crossing of a river by the Christian army, and puts the engage
ment on the l6th instead of the 7th of September. It is satisfactory to know that
his story is rendered wholly impossible by the topography of the place. For a mile
north of the Nahr-el-Falaik the road is bordered by the impassable swamp of the
Birket-el-Hamadan. North of this again it runs over flat sand dotted with salt-water
ponds, and with the forest running do\vn into it. This will not do for the battlefield
as described by the Itinerarium and Boha-ed-din. On the other hand, the country
south of the Nahr-el-Falaik suits the narrative excellently. See my map, carefully
reduced from the i-inch-to-the-mile Ordnance Survey of Palestine.

3o8 THE ART OF WAR IN THE MIDDLE AGES [1191

the beach and took no part in the fighting, but rested itself.
When the first half was wearied, it changed places with the
second and got its turn of repose. The cavalry marched between
the two halves of the infantry, and only came out when it
wished to charge. It was formed in three main corps: in the
van was Guy, 1 formerly King of Jerusalem, with all the Syrian
Franks who adhered to him ; in the second were the English
and French; in the rear the sons of the Lady of Tiberias 2
and other troops. In the centre of their army there was visible
a waggon carrying a tower as high as one of our minarets, on
which was planted the king s banner. The Franks continued to
advance in this order, fighting vigorously all the time: the
Moslems sent in volleys of arrows from all sides, endeavouring
to irritate the knights and to worry them into leaving their
rampart of infantry. But it was all in vain : they kept their
temper admirably and went on their way without hurrying
themselves in the least, while their fleet sailed along the coast
parallel with them till they arrived at their camping-place for
the night. They never marched a long stage, because they had
to spare the foot-soldiery, of whom the half not actively
engaged was carrying the baggage and tents, so great was their
want of beasts of burden. It was impossible not to admire the
patience which these people showed : they bore crushing fatigue,
though they had no proper military administration, and were
getting no personal advantage. And so they finally pitched
their camp on the farther side of the river of Caesarea." 3

From the 29th August to the 6th September, Saladin had
been perpetually seeking an opportunity for delivering a serious
attack. But the caution and discipline which Richard had im
posed upon his army foiled all the hopes of the Infidel. It be
came evident that, if the Christians were to be stopped before
they reached Jaffa, a desperate attempt must be made to
break in upon them, in spite of their orderly march and firm
array. Saladin resolved, therefore, to try the ordeal of battle in
the ground between the Nahr-el-Falaik (the river of Rochetaille)
.and Arsouf. There was every opportunity for hiding his host

1 This account of the distribution of the Christians does not tally with the
Jtinerarium, and is probably wrong. Boha-ed-din calls Guy " Geoffrey" by a curious
error.

- Barons of the party among the Syrian Franks who opposed King Guy and
wished to recognise Conrad.

3 Boha-ed-din, p. 252, in the Chroniqururs Cricntaux.

ngi] ARSOUF: PREPARATIONS FOR BATTLE 309

till the moment of conflict, for in this district one of the few
forests of Palestine, the " Wood of Arsouf," runs parallel to the
sea for more than twelve miles. It is a thick oak wood covering
all the lower spurs of the mountains, and reaching in some
places to within three thousand yards of the beach. Two days
of Richard s itinerary (the 5th and 7th of September) ran
between this forest and the sea. He was not less conscious
than Saladin of the advantage which the cover would give to
an enemy plotting a sudden attack. Accordingly he warned
the army on the 5th that they might have to fight a general
engagement on that day, and took every precaution to prevent
disorder. 1 But the Turks held back, and the first half of the
forest was passed in safety. On the 6th September the
Crusaders rested, protecting their camp by the large marsh
which lies inland from the mouth of the Nahr-el-Falaik ; this im
passable ground, the modern Birket-el-Ramadan, extending for
two miles north and south, and three miles east and west, covers
completely a camp placed by the river mouth.

On the 7th the English king gave orders to move on : the
day s march was to cover the six miles from the Nahr-Falaik to the
dismantled town of Arsouf. The road lies about three-quarters
of a mile inland from the beach, generally passing along the slope
of a slight hill : between it and the foot of the wooded mountains
there was an open valley varying from a mile to two miles in
breadth. The forest on the rising ground was known to conceal
the whole of Saladin s host, whose scouts were visible in all
directions.

On this day Richard divided his army into twelve divisions,
each consisting of a large body of infantry and a small squadron
of knights. 2 The foot-soldiery formed a continuous line, with
the crossbowmen in the outermost rank. The impedimenta
and the infantry told off to guard them moved as usual close
to the sea. The order of the march of the twelve divisions is
not clearly given to us ; we know that the first consisted on this
day of the Templars, with their knights, Turcopoles, and foot-
sergeants. The next three consisted mainly of Richard s own
subjects Bretons and Angevins forming the second, Poitevins
(under Guy, the titular King of Jerusalem) the third, and
Normans and English the fourth : the last-named corps had
charge of the waggon bearing the great standard. Seven corps

1 Itincrariiim, iv. 16. 3 Ibid. iv. 17.

3 io THE ART OF WAR IN THE MIDDLE AGES [1191

were made up from the French, the barons of Syria, and the
miscellaneous small contingents from other lands. Lastly, the
Hospitallers brought up the rear. Probably the French con
tingents were divided into four " battles," under (i) James
d Avesnes, (2) the Count of Dreux and his brother the Bishop
of Beauvais, (3) William des Barres and William de Garlande,
(4) Drogo Count of Merle. Henry Count of Champagne was
charged with the duty of keeping out on the left flank to watch
for the breaking forth of the Turks from the woods. The Duke
of Burgundy, the commander of the French host, rode by
Richard s side up and down the line, keeping order and ready
to give aid wherever it was wanted. The whole twelve corps
were divided into five divisions, but it is not stated how they
were thus distributed. Some of the five must have included
three, some only two, of the brigaded bodies of horse and foot.

Saladin allowed the whole Christian host to emerge from
the camp and proceed some little way along the road before he
launched his army upon them. While threatening the whole
of the long line of march, he had resolved to throw the main
weight of his attack upon the rearguard. Evidently he hoped
to produce a gap, by allowing the van and centre to proceed,
while delaying the rear by incessant assaults. If the Hospital
lers and the divisions next them could be so harassed that they
were forced to halt or even to charge, while the van still went
on its way, it was obvious that a break in the continuous wall
of infantry would occur. Into this opening Saladin would have
thrown his reserves, and then have trusted to fighting the battle
out with an enemy split into at least two fractions and probably
more. He had, as we shall see, wholly underrated the prudence
and generalship of King Richard, and was preparing for himself
a bloody repulse.

The Crusaders were well upon their way when the Moslems
suddenly burst out from the woods. In front were swarms of
skirmishers both horse and foot black Soudanese archers, wild
Bedouins, and the terrible Turkish horse-bowmen. Behind were
visible deep squadrons of supports the Sultan s mailed Mame
lukes and the contingents of all the princes and emirs of Egypt,
Syria, and Mesopotamia. The whole space, two miles broad,
between the road and the forest, was suddenly filled with these
imposing masses. "All over the face of the land you could see
the well-ordered bands of the Turks, myriads of parti-coloured

PLATE IX.

Battle of

ARSOLTF.

Sept 7 1191.

1 Crusaders
5 Baggage Train
3 Turkish Horse

.irnii^hei 3

6 * Sk

AA Crusaders Infanlry

and Baggage
B. Tempters
C Angevins.
Poiteviiis
E English Nonnans

with the Standard
f. Ho.s p
C.C Infanlry.
HH Turkish Skirmishers
I.I Saladin s Main Army

1191] ARSOUF: SALADIN ATTACKS RICHARD 311

;

banners, marshalled in troops and squadrons ; of mailed men
alone there appeared to be more than twenty thousand. With
unswerving course, swifter than eagles, they swept down upon
our line of march. The air was turned black by the dust that
their hoofs cast up. 13efore the face of each emir went his
musicians, making a horrid din with horns, trumpets, drums,
cymbals, and all manner of brazen instruments, while the troops
behind pressed on with howls and cries of war. For the Infidels
think that the louder the noise, the bolder grows the spirit of the
warrior. So did the cursed Turks beset us before, behind, and
on the flank, and they pressed in so close that for two miles
around there was not a spot of the bare earth visible; al! \vas
covered by the thick array of the enemy." 1

While some of the Turks rode in between the head of the
army and its goal at Arsouf, and others followed the rearguard
along the road, the majority closed in upon the left flank and
plied their bows against the wall of infantry and the clumps of
horsemen slowly pacing behind it. The pressure seems to have
been hardest upon the rear, where the right wing of the Turks
delivered a most desperate attack upon the squadron of the
Hospitallers and the infantry corps which covered them. The
French divisions opposite the Turkish centre were less hardly
pressed; the English, Poitevins, and Templars in the van, though
constantly engaged, were never seriously incommoded.

In spite of the fury of the attack, the Crusaders for some
time pursued their way without the least wavering or hesitation.
The crossbowmen gave the Turks back bolt for bolt, and
wrought more harm than they suffered, since their missiles were
heavier and possessed more penetrating power than those of the
enemy. The cavalry in the centre of the column rode slowly
on, though their horses soon began to suffer from the incessant
rain of arrows. Many knights had to dismount from mortally
wounded chargers, and to march lance in hand among the foot.
Others picked up crossbows, stepped into the front rank of the
infantry, and revenged themselves by shooting down the Turkish
horses. 2

The slow march southward went on for some time ; the
infantry held firm as a wall, and no opportunity was given for
the enemy to break in. Saladin, seeing that he was making no
progress, flung himself among the skirmishers, followed only by

1 Itiiiet-ariunt, \v. 1 8. ~ Ibid.

312 THE ART OF WAR IN THE MIDDLE AGES [1191

two pages leading spare horses, and continued to urge his men
on and to press them closer in upon the Prankish foot The
stress soon became very severe in the rear division of King
Richard s host, which was exposed to a double fire from flank
and rear. Some of the crossbo\vmen began to waver, but the
majority held firm, forced though they were to walk backwards
with their faces to the pursuing enemy, 1 for, when they turned
for a moment to move on, the Turks rushed in so fiercely that
there was grave danger that the corps of the Hospitallers might
be broken up. " They had laid their bows aside, and were now
thundering upon the rearguard with their scimitars and maces
like smiths upon anvils."

The Grand Master of the Hospitallers repeatedly sent
forward to the king, asking leave to charge. The horses were
being shot down one by one, he complained, and the knights
could no longer endure this passive kind of battle, in which they
were struck themselves, but not allowed to strike back. Richard
returned the reply that the rear was on no account to break
their order : he had settled that there should be a general charge
of the whole line when he bade six trumpets blow; before the
signal no one must move. His design was evidently to get the
whole Turkish army committed to close combat before he rode
out upon it. At present the rear alone was seriously engaged :
the van and centre were only being harassed from a distance.
Moreover, there would be great advantage in waiting till the
van had reached Arsouf, whose gardens and houses would give
good cover for its flank when the moment for the decisive
charge came.

In obedience to these orders, the Hospitallers endured for
some time longer, but they were growing restive and angry as
horse after horse fell, and man after man was disabled by
arrows in the parts of his body which the armour did not fully
protect. Presently the whole rear division lurched forward in
disorder and joined the French corps which was marching
immediately in front of it. At last, just when the head of the
army had reached the outskirts of Arsouf, the patience of the
rear was wholly exhausted. Ere the king had bade the six
trumpets sound, but (as it would seem) only just before the
moment that he would have chosen, the Hospitallers burst
forth. The ringleaders in this piece of indiscipline were two of
1 Hint rar iii7it, iv. 10, p. 264.

ARSOUF: THE CRUSADERS CHARGE 313

their leaders, their marshal and a notable knight named
Baldwin de Carron, who suddenly wheeled their horses, raised
the war-cry of St. George, and dashed out through the infantry
upon the Infidels. Those immediately about them followed;
then the French divisions ranged next them took up the
movement. It spread all down the line, and Richard himself,
seeing the die cast, was constrained to allow the cavalry of the
van and centre to follow up the attack. To the Saracens it
bore the appearance of a preconcerted movement. " On a
sudden," says Boha-ed-din, " we saw the cavalry of the enemy,
\vho were now drawn together in three main masses, brandish
their lances, raise their war-cry, and dash out at us. The
infantry suddenly opened up gaps in their line to let them pass
through." 1 Thus the attack of the Crusaders was delivered in
echelon, the left (/&gt;. the rear) leading, the centre starting a
moment after, and the right (i.e. the van) a little later than the
centre.

The Turks did not endure for a moment the onset of the
dreaded knights of the West. The sudden change of the
crusading army from a passive defence to a vigorous offensive
came so unexpectedly upon them, that they broke and fled with
disgraceful promptness. Nothing can be more frank than
Boha-ed-din s account of the behaviour of his master s host. 2
"On our side," he says, "the rout was complete. I was myself
in the centre: that corps having fled in confusion, I thought to
take refuge with the left wing, which was the nearest to me;
but when I reached it, I found it also in full retreat, and making
off no less quickly than the centre. Then I rode to the right
wing, but this had been routed even mere thoroughly than the
left. I turned accordingly to the spot where the Sultan s body
guard should have served as a rallying-point for the rest. The
banners were still upright and the drum beating, but only
seventeen horsemen were round them.

In the northern end of the battle, where the Hospitallers and
the French corps immediately in front of them were already in
close contact with the foe at the moment of the charge, a
dreadful slaughter of the Infidels took place. The rush of the
Crusaders dashed horse and foot together into a solid mass,
which could not easily escape, and the knights were able to
take a bloody revenge for the long trial of endurance to which

1 Boha-ed-din, p. 258, in Chroniqucurs Orien .aux. 2 Ibid. p. 259.

THE ART OF WAR IN THE MIDDLE AGES [1191

they had been exposed since daybreak. Before the Moslems
could scatter and disperse to the rear, they had been mown
down by thousands. In the centre and the southern end of the
battle the Turks had an easier flight, since their pursuers were
not so close. Here the contact and the slaughter must have
been much less. We know from the author of the Itinerarium
that the English and Norman knights who formed the fourth
division, counting from the van, never reached the flying enemy,
though they followed in echelon the movement of the rear and
centre corps. 1 The same was probably the case with the other
three corps of the van, for King Richard, in his letter to the
Abbot of Clairvaux, states that only four of his twelve divisions
were seriously engaged, and that these four alone really de
feated the whole host of Saladin. 2

Having pursued the Turks more than a mile, the Crusaders
halted and began to re-form there was no rash pursuit like that
which had so often ruined the Franks in earlier fields. Those
of the Infidels who still kept their heads, ceased to fly when
they were no longer pursued, and turned to cut off the scattered
knights, who had pushed far to the front, and were now riding
back to fall into line with their comrades. Of these some few
were cut off and slain among them James d Avesnes, a
notable knight, who had commanded one of the rear divisions
of the line of march. Among those of the Turks who rallied
most quickly and came back first to the fight was Taki-ed-din,
Saladin s nephew, with the seven hundred hoisemen who
followed his yellow banner.

When the Christian line was once more in order, Richard
led it on to a second charge ; the Turks broke again and made
no stand. Yet when the king cautiously halted his men, after
sweeping the enemy backward for another mile, there was still
a considerable body which turned back and once more showed
fight. A third and final charge sent them flying into the forest,
which was now close at their backs. Here they dispersed in all
directions, and made no further attempt to resist. Richard,
however, would not pursue them among the thickets, and led
back his horsemen at leisure to Arsouf, where the infantry had
now pitched their camp.

That evening many of the foot-soldiery and camp-followers
went out to the field of battle, where they stripped the dead

1 Itimrarium, p. 272. - Letter printed in Iloveden, Rolls Series, iii. 131.

ri 9 i] ARSOUF: RICHARD VICTORIOUS 315

and found much valuable plunder, since the Turks, like the
Mamelukes in later days, were wont to carry their money sewed
up in their waist-belts or under their clothing. They reported
that they had counted thirty-two emirs among the slain, arid
more than seven thousand of the rank and file. 1 Boha-ed-din
names as the most prominent of the Moslems who had fallen
Mousec, the prince of the Kurds, and two emirs named Kaimaz-
el-Adeli and Ligoush. 2 Among the Christians, James of
Avesnes was the only man of distinction who was slain: their
total loss was under seven hundred men.

So ended this important and interesting fight, the most
complete and typical of all the victories of the Franks over
their enemies. The old morals of the earlier engagements are
once more repeated in it. With a judicious combination of
horse and foot, and a proper exercise of caution, the Crusader
might be certain of victory. But we note that Richard, though
new to the wars of the East, shows far more self-restraint,
wisdom, and generalship than any of his predecessors. He
could have driven off Saladin at any time during the day, but
his object was not merely to chase away the Turks for a
moment, but to inflict on them a blow which should disable
them for a long period. This could only be done by luring
them to close combat ; hence came the passive tactics of the
first half of the day. The victory would have been still more
effective, as the author of the Itinerarinm remarks, 3 if the charge
had been delivered a little later. But the precipitate action of
the marshal of the Hospitallers caused it to be made a moment
earlier than the king had intended. Nevertheless, the results of
the fight were very well marked. Saladin reassembled his army,
but he never dared close in upon his enemy again : he resumed
his old policy of demonstrations and skirmishes. As Boha-ed-
din remarks, the spirit of the Moslem army was completely
broken. Recognising that he could not hold the open country
against the Franks, the Sultan at once dismantled all the
fortresses of Southern Palestine Ascalon, Gaza, Blanche-
Garde, Lydda, Ramleh, and the rest. He dared not leave
garrisons in them, for he was fully aware that his men would

1 Jtiiitrariuui, p. 275. " Boha-ed-din, p. 260.

3 iv. 19: " Quodsi [mandatum regis] fuisset observatum, universi illi Turci
fuissent intercept! et confusi : praedictorum vero militum nimia properatio cedebat
in deuimentum universi negotii" (p. 258).

3 16 THE ART OF WAR IN THE MIDDLE AGES [1191

not hold firm : the fate of the defenders of Acre and the result
of the fight of Arsouf were always before their eyes, and they
would not have maintained themselves for long. How well
founded was this fear, became sufficiently evident from the one
exception which Saladin made to his rule. He left a force in
Darum, the last fortress of Palestine on the way to Egypt.
Richard made a dash against it with the knights of his own
household alone, a force inferior to the garrison in number.
Yet so half-hearted had the Moslems grown, that the king
stormed the place in four days. The Turks surrendered the
citadel on the bare promise of life, though, if they had shown a
tithe of the courage of the garrison of Acre, they would certainly
have been able to hold out for weeks, if not for months. 1

Arsouf therefore gave the Franks the whole coast-land of
Southern Palestine. After repairing the walls of Jaffa, to
serve them as a basis for the attack on Jerusalem, they were
free to resume the offensive. But the jealousies and divisions
in the host ruined the campaign which had begun so brilliantly,
and, though there were several gallant feats of arms performed
during the stay of Richard in Palestine, the Holy City was
never recovered, and the war ended in a treaty which did
no more than confirm the Syrian Franks in the possession of
the coast-region which the English king had reconquered for
them.

One fight, little more than a skirmish in itself, deserves
mention as illustrating Richard s methods of war. This was
the engagement of August 5, 1192. While the king had
returned to Acre with his army, Saladin had descended to the
coast and endeavoured to retake the newly-fortified town of
Jaffa. The garrison had been driven into the castle, and was
on the point of surrendering, when Richard hastily returned by
sea with eight vessels only and saved them (August i). The
Turks were driven off for the moment, but, learning that their
enemies were very few in number, came down at daybreak on
the 5th of August to surprise the Christian camp. Richard
had with him only fifty-five knights and two thousand infantry,
the latter largely Genoese and Pisan crossbowmen drawn from
the ships which had brought him. Warned in time that seven
thousand horse, all Mamelukes and Kurds, were swooping down
upon the sleeping camp, he promptly proceeded to get his men

1 Ttintrarium, p. 356.

1192] COMBAT OF JAFFA 317

in order. He composed his front line of infantry armed with
spears, who knelt down with one knee fixed in the sand, and
with the points of their weapons levelled at the height of a
horse s breast. Behind stood the crossbowmen, one in each
interval between two spearmen : it was this soldier s duty to
discharge as fast as possible the arbalests handed to him by
another, who stood behind him, bending and loading each as
it was handed back. Thus there was no intermission in the
discharge. The Turks swept down, band rapidly following band,
against the front of the Christian line, but never dared to close.
Each squadron swerved and passed away without daring to
rush on the spears ; they did little harm with their arrows, but
suffered far more from the constant rain of arbalest bolts which
beat upon them. When they were all in disorder, Richard boldly
charged out upon them, though no more than fifteen of his
knights were horsed. He cut right into their midst, and then
hewed his way back again, saving by his personal valour the
Earl of Leicester and Ralph of Maulcon, who had been sur
rounded and were nearly made prisoner.-. The fight lingered
on for some hours after the surprise had failed, but when the
king brought up some small reserves from the fleet (he left only
five men on each galley) the enemy fled, leaving seven hundred
men and fifteen hundred horses dead upon the field. Of the
Crusaders only two men had fallen, so secure had their order
of battle kept them I 1

1 All this from the excellent account in I. ii;t&gt;-arinm, vi. g 21-24.

CHAPTER V

THE GREAT DEFEATS OF THE CRUSADERS CARRHAE,
HARENC, TIBERIAS, ACRE, MANSOURAH

HAVING now given fair typical instances of the methods
by which the Franks won success in the interminable
campaigns which followed the establishment of the Latin States
in Syria, it remains that we should show in the same fashion
the manner and causes of their defeat. With those which were
the inevitable consequences of strategical blunders we have dealt
in our chapter on Strategy. It is with tactical errors that we
are now concerned. As illustration we have chosen four battles.
Carrhae (1104) will show the result of careless pursuit and the
neglect of the proper precautions required in Turkish warfare.
Tiberias (1187) displays a complicated series of blunders the
neglect of commissariat arrangements, the choice of unsuitable
ground, the imperfect reconnoitring of the enemy, and (most
important of all) the fatal results of dividing the infantry and
cavalry. The battle in front of Acre (i 190) proves that a victory
practically won might be turned into a defeat by the want of a
guiding hand and neglect of the most rudimentary discipline.
Mansourah (1250) points out that a fault originating in bad
strategy may logically lead to bad tactics, and illustrates as well
the normal want of discipline in all Western hosts.

The battle of Carrhae may be taken as an example of the
manner in which even the most practised veterans of the first
Crusade could fail when they neglected obvious precautions
and fought on unfavourable ground. In the spring of 1104,
Bohemund, now for the last six years Prince of Antioch, and
Baldwin of Bourg, Count of Edessa, resolved to make a bold
push into Mesopotamia. The Turks had lately threatened
Edessa ; in retaliation the princes formed a project for seizing
and garrisoning the strong town of Harran (Carrhae), the frontier

318

iio 4 ] BATTLE OF CARRHAE 319

post of the Moslems. It was close enough to Edessa to be a
troublesome neighbour, only twenty-five miles separated the
two places, while at the same time it was a favourable point
to serve as a base for further progress eastward. Baldwin
called in to his aid his cousin Joscelin, to whom he had granted
a great lordship west of the Euphrates, round the town of
Turbesel. Bohemund brought with him his kinsman Tancred,
the hero of so many exploits in the first Crusade. The oppor
tunity seemed fair, for by systematic ravagings Baldwin had
ruined the countryside round Carrhae, and knew that the place
was straitened for provisions. Moreover, the two Turkish
princes who ruled in Mesopotamia, the Atabeg Sokman ibn-
Urtuk of Kayfa, and Jekermisch the successor of Kerboga in
the emirate of Mosul, were engaged in bitter strife with each
other.

At the head of what passed for a considerable army among
the Syrian Franks, the allied princes marched on Carrhae and
formed the siege. The place, as Baldwin had known, was ill
stored, and ere long the famished citizens began to treat for a
surrender. But while the terms were being disputed, a relieving
army came in sight : Sokman and Jekermisch had come to
terms in face of the common danger, and had combined their
forces to save Carrhae. The former brought to the field seven
thousand Turkish horse-archers ; the latter, three thousand Kurds,
Bedouins, and Turks. They had resolved to threaten an attack-
on the Christian camp, and to throw a convoy into the city
while the besiegers attention was distracted. Their success
was far greater than they could have hoped : when the Franks
saw them, they formed in three "battles," each composed of
horse and foot, and marched out to attack them. Bohemund
held the right, Tancred the centre, Baldwin and Joscelin the
left, in the Christian host. When the Franks advanced, the
Turkish princes applied the ordinary stratagems of their race :
they retreated into the broad plain eastward of the city,
harassing the advancing enemy with their arrows. Old
soldiers like Bohemund and Baldwin should have known better
how to deal with such tactics, but with inexcusable rashness
they pursued the Turks into the rolling sandy plain till they
had got twelve miles east of Carrhae. The Turks, still falling
back, crossed the river Chobar, and the Crusaders rapidly
followed them. Men and horses were growing fatigued, the

320 THE ART OF WAR IN THE MIDDLE AGES [1104

infantry were tired to death, and, when the afternoon was far
spent, Bohcmund at last gave the signal to halt, and ordered
his host to encamp for the night, not dreaming that the enemy
was likely to suddenly take the offensive. This was the moment
for which the Turks had been waiting. When they saw the
Franks falling out of line, dismounting, and taking off their
arms, they suddenly came charging in with loud shouts and
dashed among their enemies, using the sword ns well as the
arrow. Baldwin s division was caught wholly unprepared, and
ridden down before it had time to re-form ; both he and his
cousin Joscelin were taken prisoners, and with them Benedict,
Archbishop of Edessa. The camp and all its stores fell into
the hands of the Turks. Tancred, more cautious than the
Count of Edessa, had not allowed his men to disperse, and was
able to rally them and form up on a hill a mile or two behind
the camp ; here Bohemund joined him with the main body of
the Christian right wing, which had been disordered, but not
wholly destroyed. 1 The two princes waited to be attacked, but
the Turks only demonstrated against them ; they had no inten
tion of closing, and were well satisfied with their partial victor}-,
and eager to share the plunder they had taken. When night
fell, the Franks found themselves in evil plight : they had lost
not only their camp, but all their provisions ; horses and men
alike were famished and exhausted after the long day s march
in the sandy plain. Nevertheless, the princes resolved to renew
the combat next morning, and bade the starving army prepare
for a second battle. But the Franks were demoralised : under
cover of the darkness their foot-soldiery melted away towards
the fords of the Chobar, drove off the guard which had been
placed there to stop desertion, and made off towards Edessa.
When the flight of the greater part of the infantry was observed,
many knights stole away after them, and Bohemund and
Tancred ultimately found themselves deserted by all save the
mea of their own military household. It was impossible to
await the dawn and the Turkish advance, so the princes followed
their panic-stricken host towards the ford. It was fortunate
that the enemy kept a bad watch, or the whole Christian army
might have been destroyed in detail. But the Turks were

1 So Ralph of Caen ; the Arab Ibn-Ghiouzi says" that Tancred was at some
distance from Baldwin, on the other side of a hill, and that the Count of Ede=sa was
routed before his ally could come up to help him.

1 104] BATTLE OF CARRHAE 321

spending the night in a hot dispute ; Sokman s men had been
plundering the Prankish camp while Jekermisch s troops had
been observing Tancred s rallied division. On their return at
dusk, the Mosulite horsemen demanded their share of the prey,
and Jekermisch seized the person of Baldwin, the chief of the
captives, who had been placed in Sokman s tent. The Turks o
Kayfa drew their swords to resent this insult to their master,
and a general combat would have followed had not Sokman
succeeded in appeasing his men, and at the same time bought
off Jekermisch by a promise to divide the spoil fairly. 1

Meanwhile, the Christians got a long start, and were all over
the river and straggling back towards Edessa before the day
dawned. They were, of course, pursued the moment that their
departure was ascertained, and many stragglers were cut off ;
the main body, however, reached the city in safety. 2 But the
blow had been a heavy one : more than half the army was
missing, 3 and the Christians were thrown upon the defensive
for some years. It is astonishing that the Turks did not
make more profit from their victory, but, after besieging Edessa
in vain for fifteen days, they dispersed and returned to their
homes.

It is strange to find that the Crusaders were routed on the
same field where the younger Crassus and his fifteen hundred
Gallic horsemen were cut to pieces by the Parthian archers be
fore the eyes of his father the Triumvir nearly eleven centuries
before. That cavalry from the far West armed with the lance
should strive again on that sandy plain with the Turanian horse-
bowmen, and should succumb again, was one of the most curious
coincidences of history. The march of the Triumvir and his
legions among the evasive Parthians suggests somewhat the
advance of Baldwin and Bohemund, but the Roman was worse

1 Ibn-Alathir says (see Michaud, BibliothZqite ties Croisades, iv. 19) that Sokman
exclaimed, "Islam will have no joy from this victory if we quarrel after it. I will
rather lose my spoil than let the Christians taunt us with folly."

2 See in Ralph of Caen, 281, 282, the story of the flight, especially the comic tale
of Archbishop Bernard, who, " when no one was pressing, thought he had behind him
hosts of Turks with bended bows and drawn swords," and cut off his palfrey s tail to
flee the faster.

3 Ibn-Alathir no doubt exaggerates when he says that twelve thousand Franks
were slain or taken, and that Tancred got away with six knights only. But the
importance of the disaster is vouched for by William of Tyre s statement that "in
no battle of the East down to our own day were so many strong and valiant men slain,
and never did a Christian army fly so shamefully" (x. no).

322 THE ART OF WAR IN THE MIDDLE AGES [1187

off than the Franks. He was fighting, as it were, blindfold,
against a foe whose tactics were wholly unknown to him ; while
the veterans of Dorylaum and Antioch were experienced in
Turkish wiles, and ought never to have been caught unprepared
Their failure to observe common precautions was all the more
inexcusable, and if their host got off more cheaply than the
unfortunate followers of the two Crassi, it was by good luck and
not by their deserts. 1

Battle of Tiberias, July 4, 1 187.

Disastrous as was the battle of Carrhae, it cannot compare
either in its scale or in consequences with the great fight eighty
years later which gave Jerusalem to the Infidel. The battle of
1104 did not even destroy the single principality of Edessa:
that of 1 187 was the great turning-point in the whole history of
the Crusades, since it entirely deprived the Crusaders of their
hold on inner Syria, and left them for the future masters of
nothing more than a narrow strip of coast-land.

In 1187 Saladin, after having cut short the borders of the
Christians in many quarters, resolved to risk an attack on the
centre of their strength, by a direct invasion of the kingdom of
Jerusalem. He first despatched a considerable force to execute
a raid into its northern parts : it was put in charge of Modhaffer-
ed-din, Prince of Edessa and Haran, who crossed the Jordan,
harried the hill-country of Galilee, and cut to pieces at the bloody
encounter of Saffaria (May i) the knights of the Temple and
the Hospital, who had come forth against him with more zeal
than discretion, before any succours could reach them. His safe
return emboldened the Sultan to ride forth in person.

In June he gathered all his disposable forces from Egypt,
Syria, and Mesopotamia at Ashtera in the Hauran. There were
ten thousand mailed Mamelukes of his regular army, beside the
innumerable contingents of his provinces : the total may have
amounted to some sixty or seventy thousand men. On June
26 he led them down to the vicinity of the Jordan, and
encamped at Sennabra, close to the bridge of El-Kantara, which
crosses the river a mile south of the point where it issues from

1 We find that there were men in Latin Syria learned enough to observe the
coincidence. William of Tyre remarks that "this was that same Carrhae wheie
Crassus the Dictator (!) had his celebrated mouthful of the Parthian gold for which
he had been so greedy" (W. T. book x.).

ti:87] SALADIN INVADES GALILEE 323

the Sea of Galilee. Three days later he passed the stream and
advanced into Christian territory. His first aim was to capture
the town of Tiberias, the capital of the principality of Galilee.
Posting his main army on the hills east of that place, he sent a
corps to lay siege to it. The town yielded with unexpected ease,
but the garrison and their mistress, the Countess of Tripoli, with
drew into the castle, a strong fort overhanging the water, which
was capable of holding out for many weeks.

Meanwhile, the Christians were assembling in great strength.
Modhaffer-ed-din s raid had seriously disturbed them, and, when
they heard that Saladin was concentrating his army in the
Hauran, they had resolved to draw together in full force. King
Guy summoned in all his barons and knights ; the military Orders
put all their available men into the field, thinned though their
ranks had been by the disaster at Saffaria. The towns sent
contingents even larger than they were bound to furnish. The
Count of Tripoli, who had only lately reconciled himself to his
suzerain, did his best to atone for past disloyalty by bringing
the full levy of his county to the muster. The True Cross was
fetched out from the Church of the Hoi} Sepulchre and sent to
the front, in charge of the Bishop of Lydda. The castles and
cities of Southern Palestine were left with garrisons of danger
ously small numbers. By this concentration, the Franks were
able to assemble twelve hundred knights, many hundred
Turcopoles or mounted bowmen equipped in the Turkish fashion,
and eighteen thousand foot, 1 the largest force that they had ever
put into the field save that which had been mustered for the
abortive campaign of 1 184- 2 Their meeting-place was the village
of Saffaria the spot where the disaster to the Templars had
occurred seven weeks before. It lies in a well-watered upland
valley, three miles north of Nazareth and seventeen east of Acre.
From thence to Tiberias is sixteen miles, by a road passing
across one of the most desolate and thinly-peopled districts in
the Galilean hills. 3 The time was the hottest month of the
summer, and Saladin s raiders had burned the villages and
destroyed the wells all around. They had even defiled the
Church of the Transfiguration on the summit of Mount Tabor.

1 So Ralph of Coggeshall, the best authority for the campaign, p. 218.
On that occasion they had raised what William of Tyre calls the largest host ever
seen in the kingdom (xxii. p. 448).

* There are only two small villages, Toron and Lubieh, on the road.

324 THE ART OF WAR IN THE MIDDLE AGES [1187

There was long talk and hot disputing at Saffaria as to
whether the army should march to the relief of the castle of
Tiberias. The Franks had mustered in such full force that they
could never hope to raise a larger army. Saladin had placed
himself in a position where defeat would mean ruin, since he
had the broad Sea of Galilee at his back, and his retreat either
north or south would be through difficult and dangerous denies.
On the other hand, it was hazardous to risk the whole resources
of the kingdom in a single fight. If the army at Saffaria was
beaten, there were no reserves left on which it could fall back.

The Count of Tripoli, the most experienced warrior in the
Christian host, took the side of caution. He pointed out that if
they did not march against Saladin, the Sultan would be forced
to march against them, since he could not long abide in the
desolate country round Tiberias. His only other alternative
would be to return to Damascus, a course which he certainly
would not consent to take when his pride had risen so high
and when his army was so strong. It would suit the policy of
the Christians to be attacked at Saffaria, where they had a good
position, plenty of food, and an ample supply of water. Saladin,
on the other hand, would arrive with an army tired out by a
fatiguing march and discouraged by the distance from its base ;
for the Turks must fight, knowing that they had no shelter
nearer than Damascus, and with the lake and the Jordan at
their backs. Raymond added that he, if anyone, should feel
interested in the preservation of Tiberias, since his own wife
and children were being beleaguered in the citadel ; nevertheless,
he advised that a waiting policy should be adopted, and the
responsibility of the initiative thrown on the enemy. If the
Christian army marched over the mountains, it would have to
fight when worn out by thirst and heat ; it was far better that
the Infidels should have these disadvantages on the day of
battle. 1

Unfortunately the advice of Raymond was ill received.
His enemies whispered that he was the king s enemy, and that
his cowardly counsel was that of a deliberate traitor. The
majority of the barons voted that it would be shameful to
abandon the garrison of Tiberias. The king assented, and on
Thursday, July 3, the army marched out from Saffaria lightly

1 Ralph of Coggeshall, p. 222, here agrees wonderfully well with the Moslem
chronicler, Ibn-Alathir.

t[8y] THE FRANKS MARCH ON TIBERIAS 325

quipped, and leaving all its impedimenta behind in the camp,
he order of march is not very clearly stated ; but we know
tat the Count of Tripoli, as the chief vassal of the Crown
resent, led the van, while the Templars brought up the rear.
The king, with his military household, and with another corps
told off to the defence of the True Cross, was in the middle.
How many divisions the whole army contained we are not
told, nor is it explicitly stated that each consisted of horse and
foot combined, though this must almost certainly have been
the case.

The Franks had marched about nine or ten miles, when
they began to be surrounded by swarms of Turkish skirmishers.
Saladin did not display his main force, but enveloped their army
with a cloud of horse-bowmen, whose orders were to make the
march slow and painful. By the time that the host drew near
the deserted village of Marescalcia, 1 it was terribly weary and
harassed. Only some six miles now separated it from the
town of Tiberias and the lake. 2 The van, which had pushed
down into the lower ground and was still advancing, was within
three miles of the water. But between the weary Crusaders
and their goal lay the hills of Tiberias, a range rising to about
one thousand feet above sea level : the northern point, Kurn-
Hattin, is eleven hundred and ninety-one feet high. Behind the
crest of these hills the ground falls suddenly towards the deep-
sunk hollow of the Sea of Galilee. Tiberias itself is no less than
six hundred and fifty-three feet below the level of the Mediter
ranean. All along the range the Turks were arrayed, and it
was necessary for the army to cut its way through them by
one of the two passes which cross at its lowest points the
depressions called the Wady-el-Muallakah and the Wady-el-
Hammam.

Tired as the army was, there was an absolute necessity that
it should push on, for there was no water available for three
miles around, and men and horses were already perishing of
thirst. The Count of Tripoli sent back to King Guy, begging
him to hasten the advance at all costs, as the day was drawing
on, and the lake must be reached ere nightfall if the army was

1 Probably the modem Lubieh.

- The distance is grossly understated in Coggeshall, who says that there were
only three miles between Marescalcia and the lake, and that the van under the
Count of Tripoli was actually only one mile from Tiberias (p. 223).

326 THE ART OF WAR IN THE MIDDLE AGES [1187

to be preserved. But the king and his counsellors were dis
heartened, and no longer possessed the courage to order a final
assault upon the heights where the Turks clustered so thick.
Moreover, the Templars in the rear were sending messages to
say that they were so hard pressed that they had been forced to
halt, and could not keep up with the advance of the column in
front of them. Harassed and tired out, the king ordered the
whole army to halt and encamp where it stood, on the hillside
near Lubieh. The command was a fatal mistake ; it would
have been wise to push on at all costs to Tiberias : if this was
not done, a lateral movement of only three miles northward
would have brought the host to the perennial stream in the
\Vady-el-Hammam, where the whole army could easily slake its
thirst, and four miles more would take them to the lake.
Fearing, however, that the Templars would be cut off if any
further advance was made, and shirking the attack on the
formidable bodies of Turks holding the hilltops, Guy bade the
trumpets sound for halt and encampment. Raymond rode
back to join the main body, exclaiming, "Alas, alas, Lord God!
the war is ended ; we are all delivered over to death, and the
realm is ruined." 1

That night the Franks camped, huddled together around
the royal standard on the hillside. There was little food and
hardly a drop of water in the host : even sleep was impossible,
for the T \irks came close in under cover of the darkness, and
kept up a constant shower of arrows into the camp. They also
fired the dry grass to windward of the Crusaders, so that
stifling clouds of smoke were drifting over it all night. "God

o o o

fed the Christians with the bread of tears, and gave them to
drink without stint of the cup of repentance, till the dawn of
tribulation came again." 2 The Saracens were not much more
easy in their minds than their enemies: with the lake at their
back and the formidable Christian army still intact, they had
many qualms of spirit when the fight was renewed on the
morning of Friday, June 4. 3

King Guy had once more ranged his army in order, with the
same divisions as he had drawn up on the previous day the
Count of Tripoli in front, the military Orders in the rear.
Swerving from his original route, he now ordered the march to

1 Ralph of Coggeshall, p. 223. Ibid. p. 224.

3 Boha-ed-din, p. 94.

PLATE X.

the

Localities of
BATTLE of
TIBERIAS

Valley of
Suhel el Butluu

LI L EE

( 6SO F belo* Sea Letel)

BATTLE OF ACRE.

A King Guy B Conrad of Montferrat.

C Lewis of THurin^ia D Master ol (h* Templars .

f. Kuj-dj * E^Yptiiina F,Me*oroliTuaiis G 5yn0n*

1187] TIBERIAS: DESTRUCTION OF THE INFANTRY 327

be directed towards the Wady-el-Hammam and the village of
Hattin, aiming at the nearest water, and no longer taking the
shortest way to Tiberias. Saladin had now brought up his
whole host, which encircled the Christians on all sides, though
the thickest mass lay across the road to the lake. The
Crusaders moved forward for some distance, and were about to
join in close combat, when the king detected great unsteadiness
in his infantry. They had been told off to the various corps of
cavalry, and were directed to form line in front of them, "that
the two arms might give each other the proper support, the
knights protected by the arrows of the foot, and the foot by the
lances of the knights." l At the moment of close combat,
however, the greater part of the infantry, after wavering for a
moment, shrank together into one great mass, and, swerving off
the road to the right, climbed a hill (probably Kurn-Hattin)
which lay to that flank, and formed in a dense clump on its
summit, deserting the horsemen on the road below. 2 The king
sent messenger after messenger to them, imploring them to
come down and play their part in the battle. The only answer
which they returned was that they were dying of thirst, and
had neither will nor strength to fight. Already despairing of
the event of the day, but determined to push on as long as it
was possible, Guy ordered the knights to advance towards the
lake. But ere long the Templars and Hospitallers in the rear
sent to him to say that they were so hard beset that they could
not move forward any more, and must succumb if not strongly
reinforced. "Then the king, seeing that the infantry would not
return, and that without them he could not prevail against the
arrows of the Turks, ordered his men to halt and pitch their
tents. So the battles broke up, and all huddled together in a
confused mass around the True Cross." 3

It was not, however, the whole of the Christian knighthood
which gave way to this impulse of despair and fell into a
passive defensive which was bound to prove fatal in the long-
run. The Count of Tripoli and the van division, seeing the ruin
behind them, and finding the Turks already stealing in between
them and the king s corps, resolved not to return, but to cut

1 Ralph of Coggeshall, p. 224.

; " Conglobati sunt in unum cuneum, et veloci cursu cacumen excelsi monies,
relinquentes exercitum, malo suo ascenderunt " (ibid. p. 225).
3 Ibid. p. 225.

328 THE ART OF WAR IN THE MIDDLE AGES [1187

their way through the Moslems and seek refuge in flight. "The
battle is hopelessly lost ; let every man save himself if he can," 1
cried Raymond, and, forming his corps in a close body, he
charged the Turks immediately in front of him, aiming no
longer at the lake, but at the hills to the north-west. His
desperate assault burst right through the circle of horse-archers,
and he, with his comrades, Balian of Nablous and Reginald of
Sidon, and the whole of their retainers, got safely away to the
north. The Moslem chroniclers say that Saladin s nephew,
Taki-ed-din, who commanded in this part of the field, made no
serious effort to check or pursue them, because he judged that
it would be more profitable to let them go, for their departure
enfeebled the Christian army by a third, and left the remainder
a more certain prey to Saladin. It is permissible to suspect
that the plea was an afterthought, and that the Turks were in
truth cowed by the sudden charge of these desperate men.

Meanwhile, all had gone to ruin in the rear. While one
swarm of Moslem horse beset the confused mass of knights
huddled together around the king s banner and the True Cross,
the rest turned to assault the infantry. The wretched fugitives
on the hill were too exhausted to offer any real resistance. The
first charge of the enemy split up their ill-compacted ranks ;
some were ridden down, some were cast by the impact over
the cliff at the back of the hill, and met their death in the fall.
The majority threw down lance and arbalest and held out their
hands to the conquerors. The Turks slew many, and accepted
the rest as captives.

The fate of the king and his knights was no less dismal.
They held out for a long time, though neither victory nor
retreat was any longer possible. Encompassed on all sides by
the dense swarm of Turks, they could only stand to be shot
down. At last, though their horses were reduced to the last
pitch of fatigue, and though they themselves had drunk their
last drops of water on the previous night, the whole or part of
the host resolved to make one more push for liberty. They
might perhaps cut their way through to safety, as the Count
of Tripoli had done a few hours before. A Mohammedan
chronicler 2 has preserved a good account of this last charge,

1 "Qui potest transire transeat, quoniam non est nobis praelium." A perfect
mediaeval rendering of " Sauve qui peut." (Ralph of Coggeshall, p. 225.)
a Ibn-Alathir.

1187] TIBERIAS: THE FRANKS SURRENDER 329

which he drew from the memory of an eye-witness, Saladin s
son, Malek-el-Afdal, who first drew sword at the battle of
Tiberias. The prince rode by his father s side at the head of the
Sultan s reserve, behind the circle of skirmishers which was
besetting the Crusaders.

" When the king of the Franks and his knights," said Malek-
el-Afdal, " found themselves pressed together on a hillock on
the side of the hill of Kurn-Hattin, I was with my father. I
saw the Franks make a gallant charge at those of the Moslems
who were nearest them, and drive them back close to the spot
where we stood. I looked at my father and saw that he was
deeply moved ; he changed colour, grasped his beard in his
hand, and moved forward crying, Let us prove the devil a
liar! 1 At these words our men precipitated themselves upon
the Franks, and drove them back up the hillside. I began
myself to be overjoyed, and to cry, They fly ! they fly ! But
the enemy presently came back to the charge, and for a second
time cut their way to the foot of the hill ; when they were
again driven back, I began to cry afresh, Tltey fly ! tliey fly !
Then my father looked at me and said, Hold your tongue, and
do not say that they are really routed till you see the king s
tent fall. Shortly after we saw the tent come down ; then my
father dismounted, prostrated himself to the earth in thanks to
God, and wept tears of joy."

When the second attempt to pierce the Moslem circle had
failed, and all hope was gone, we are told that in their despair
the Franks dropped from their exhausted horses, cast down
their lances, and threw themselves sullenly upon the ground.
The Turks ran in upon them and took them captives without
another blow being struck. To their great surprise, they found
that very few of the knights were seriously hurt ; their mail-
shirts had protected them so well from the arrow-shower that
few were badly wounded and hardly any slain. Thirst and
exhaustion had brought them down, rather than the shafts or
scimitars of the conquerors. On the other hand, there was
hardly a horse that was not sorely hurt, and not one that could
have carried his rider out of the battle. The poor beasts were
utterly worn out by two days deprivation of water and forage.

In the corps which thus surrendered with the king were all

1 Meaning, I suppose, that as God had promised victory to the True Believers,
any thought of defeat must be an inspiration from Satan.

330 THE ART OF WAR IN THE MIDDLE AGES [1187

the great barons of Palestine save those who had got off in
company with the Count of Tripoli. They included the king s
brother Amaury, Constable of Jerusalem, the Marquis of
Montferrat, 1 Joscelin, titular Count of Edessa, Reginald of
Chatillon, lord of Kerak and Montreal, Humphrey of Toron,
Hugh of Tiberias, 2 Hugh of Giblet, the Bishop of Lydda, the
Master of the Hospitallers, and many scores of knights of
wealth and name. Few persons of any note had fallen the
Bishop of Acre, who had borne the Holy Cross throughout the
battle, is the only magnate reckoned among the dead.

That evening Saladin held a review of the prisoners. He
kindly entreated King Guy and most of the barons, but he
called out and slew with his own hand Reginald of Chatillon,
who had earned his hate by breaking a truce and by plundering
some pilgrims to Mecca who had passed by his castle of
Montreal. He also bade his bodyguard slay off-hand all the
knights of the Temple and Hospital who had fallen into his
hands. Xot content with this, he proclaimed throughout his
host that any private soldier who had captured any member of
the military Orders must give him up. For each knight so
surrendered he paid the captor fifty dinars, and then sent the
prisoner to join his comrades in death. More than t\vo hundred
Templars and Hospitallers were thus slain in cold blood.
Saladin looked upon them as the professed and professional
enemies of his faith, and never gave them quarter. When we
remember that he had committed such atrocities, we need not
blame too bitterly misdeeds on the other side such as Coeur de
Lion s massacre of the garrison of Acre.

Few victories have brought in their train more important
results than that of Tiberias : within a few months the whole of
the kingdom of Jerusalem save a few coast-fortresses was in the
hands of Saladin. The realm had been drained dry of men to
supply the army which perished on the hillside of Hattin, and
its towns and castles fell helplessly before the Moslem for sheer
lack of defenders. Places that had braved the assaults of the
Infidel for eighty years opened their gates at the first summons,

1 Boniface, father of the more celebrated Conrad of Montferrat, who figures in
the third Crusade.

2 I suppose that the "son of the Lady of Tiberias," named by Boha-ed-din, is-
this Hugh, eldest son of the lady, who had by now married as second husband
Raymond of Tripoli.

ii88] THE SIEGE OF ACRE BEGUN 331

because there were none but clerks and women left within them.
Jerusalem itself surrendered after a siege of only twelve days.
A few remote castles like Kerak and Montreal had been left
better garrisoned, because they lay in the extreme limit of the
kingdom, and some of these held out till 1 188. Montreal, endur
ing the extremities of famine, did not surrender till May 1189.
But in the main body of the realm, Tyre, whither the sad
survivors of Tiberias had retired, was the only stronghold of
first-rate importance which remained in Christian hands.

Such were the consequences of the overhaste of King Guy,
and of his determination to cut his way to the relief of Tiberias
without having taken account of the character of the country
side in which he was to fight. We may safely say that if he had
taken more care about supplies, and especially about his provision
of water, and had carefully planned out his itinerary, he might
have reached his goal. The Saracens were in a very uncomfort
able position, with the lake at their backs and no place of refuge
near ; one more such push as the Count of Tripoli had advised
on the evening of the first day would probably have led to
their withdrawal. But a much more easy alternative would
have been to have encamped in some well-watered spot, such as
Saffaria, and awaited the retreat of Saladin. The Sultan must
have soon retired for want of provender (and especially of
fodder) in the wasted country about Tiberias, and he could not
have dared to disperse his army for foraging purposes in the
face of the Christian host, while it remained intact and con
centrated in front of him. The whole battle, therefore, was
unnecessary, and the details of Guy s bad generalship are
comparatively small blunders when compared with the enormous
initial mistake of fighting at all.

Battle of Acre, October 4, 1 189.

When, only two years after the fatal day of Tiberias, we once
more find the Christians capable of contending on equal terms
with Saladin, it is of course due to the arrival of reinforcements
from the West. The exhausted remnant of the Syrian Franks
could have done nothing. When King Guy was freed from
captivity in 1188, and set himself to gather forces for the
recovery of some foothold in his lost realm, it took him a year
to collect seven hundred knights and nine thousand foot, and
these were not for the most part his own vassals (though Tripoli

332 THE ART OF WAR IN THE MIDDLE AGES [1189

and Antioch lent him some succour), but early arrivals from
among the men of the West who had taken the Cross when the
news of the capture of Jerusalem reached Europe. Guy was
not even in possession of Tyre, the one important city of his
realm which still remained in Christian hands. His rival and
brother-in-law, Conrad of Montferrat, shut its gates and refused
to admit him.

It was, therefore, an act of no small daring when, on August
28, 1189, Guy and his little army boldly challenged the power
of Saladin by marching on Acre and encamping before its
walls. The siege began as a blockade and nothing more, for
the Turks were able to pass in and out of the place at will.
But gradually the crusading contingents began to drop in one
by one from the West, and, less than a month after the siege
began, nearly forty thousand men were assembled around Acre.
On September 14 they engaged in a bloody and indecisive
fight with a relieving army which Saladin in person had led to
the succour of the garrison ; the Sultan succeeded in throwing
a large convoy into the city, but failed in his design of driving
off the besiegers. This encouraged the Crusaders, whose
numbers were still growing every day, to attempt a counter-
stroke. They first completed the investment of Acre by extend
ing their pickets from sea to sea across the neck of land on
which the city stands. Then, after having shut off the garrison
from the army without, they resolved to offer battle in the open
by marching upon the Sultan s camp.

The crusading host lay in a semicircle round Acre, with
the king s pavilion pitched on "Mount Turon " (Tel-el-Fokhar),
a low hill ninety feet high, which lies about fourteen hundred
yards from the walls. The Turkish army formed a much larger
semicircle, separated from the Franks by an interval of about
two miles. Its central rallying-point was the hill of Ayadieh,
rising two hundred and fifty feet above the plain : here Saladin
himself lay. His subordinates stretched out to right and left,
watching the whole of the plain from the river Belus (Nahr-en-
Namein) on the south to the sea on the north. That the armies
engaged were really very large, and that the chroniclers for once
cannot be very far mistaken in the numbers that they give, is
best shown by the fact that the length of the Prankish lines
must have been more than two miles, and the front covered by
the Sultan s host no less than three miles.

1189] ACRE: THE FRANKS ADVANCE 333

Descending from Mount Turon into the plain of Arab-el-
Ghawarneh, which stretches away to the foot of the hill of
Ayadieh, the Crusaders formed themselves in four corps. The
first (counting from the right) was commanded by King Guy,
and consisted of the Hospitallers, the king s own following, and
the French Crusaders under the Count of Dreux and the Bishop
of Beauvais. In the second corps were the Archbishop of
Ravenna and Conrad of Montferrat, with the greater part of the
Italian Crusaders and such of the barons of Palestine as adhered
to Conrad l in his feud with King Guy. In the third was Lewis,
Landgrave of Thuringia, with the greater part of the German
contingents and the Pisans under their archbishop. In the
fourth marched the Templars, under their Master, Gerard of
Rideford, the Counts of Bar and Brienne with the Crusaders
from Champagne 2 and the smaller part of the Germans

Geoffrey of Lusignan, the king s brother, and James of
Avesnes remained behind in the camp with a reserve. 3 They
had to watch the city, whose investment had to be relaxed when
the army took the field. Apparently the space from Mount
Turon northward to the sea was no longer observed, nearly a
mile being left open ; only the eastern face of the wall was
covered by the camp, the northern face was free.

In each of the four marching divisions of the Christian host
the proper disposition of horse and foot was carried out. The
bowmen and arbalesters formed a long continuous first line :
behind them marched the knights in close order. The whole
host fronted north-east, and set its face towards the Sultan s tent,
plainly visible on the hill of Ayadieh. The line looked very
formidable and strong : the chroniclers give its numbers at
four thousand horse and a hundred thousand foot figures
from which some deductions may be made.

On seeing the Christians moving forth from their camp,
Saladin had promptly drawn up his host in front of them. The
army reached from the sea to the Belus, with a semicircular
front of more than three miles : the centre was somewhat refused,
the wings somewhat thrown forward. The array of the various

1 Conrad had been temporarily reconciled to King Guy, and had lately come to
help him in the siege : with him had arrived the Archbishop of Ravenna.

"Catalauni," as the letter of Theobald given in Ralph de Diceto calls them.
Henry of Champagne himself came later to Acre, but the Counts of Bar and Brienne,
both Champenois, were already in the field.

3 Probably Syrian Franks and Netherlanders.

334 THE ART OF WAR IN THE MIDDLE AGES [1189

contingents is elaborately set forth by Boha-ed-din : to the south,
next the river, were the garrison of Egypt, the old troops of
Shirkuh ; next to them were the followers of Modhaffer-ed-din,
lord of Edessa and Haran ; beyond these the contingent of
Sinjar in Mesopotamia; next were the whole of the tribes of
Kurdistan, under their great emir, Mashtoub. These four
corps formed the left wing. The centre consisted of the Sultan s
bodyguard and the Mesopotamia!! troops from Diarbekr, Mosul,
and Hisn-Kayfa. The Sultan himself, his son Malek-el-Afdal,
and his nephew Malek-ed-Dafer, were here in command.
The right wing, which lay towards the sea, was composed of the
Syrian contingents, headed by Saladin s nephew Taki-ed-din,
Prince of Emesa.

When the Christians began to advance into the plain, they
soon found that the intervals between the four corps in their line
of battle were growing greater. This was necessarily the case
when they marched out from a comparatively narrow position
into a wide plain whose whole breadth was held by the enemy.
When they began to extend their front to make it equal to that
of the Turks, each step farther forward brought about a wider
separation between the centre and the wings. This was a
disastrous fact for the Franks, whose main chance of victory lay
in their being able to keep a well closed-up line. In the actual
fighting, as we shall see, this was so far from being the case that
three separate engagements were fought by the left wing, the
right wing, and the two centre divisions.

The first contact occurred in the northern part of the field,
where the Master of the Templars faced the Syrian contingents
of Taki-ed-din. After a few minutes the Infidels began to
give ground : Boha-ed-din assures us that the movement was
voluntary, and that the Prince of Emesa was desirous of drawing
away the Christian left wing from the main body by his retreat,
Whether this was so or not, it is at any rate certain that Saladin,
seeing his right wing retiring, sent to its aid heavy reinforcements
from his centre. These succours enabled the Syrians to retake
the offensive, and the Templars had to re-form their line on a hill
lying toward the sea (probably the rising ground now known as
Kisr-el-Hammar). Here the battle stood still for some time
without marked success on one side or the other.

Meanwhile, Saladin s despatch of troops from his centre
towards his right had been observed by the Franks, and the

1189] ACRE: THE FRANKS ROUTED 335

two central divisions of the Christian host, led by Conrad of
Montferrat and Lewis of Thuringia, delivered a fierce assault
on the Sultan s main body. They marched at a moderate pace
with the infantry in front shooting hard, till they came in
contact with the Mesopotamian troops from Diarbekr and Mosul.
When the lines closed, the knights passed through intervals
opened out for them by the foot-soldiery, and crashed into the
Turkish ranks. The Infidels could not stand the shock : their
line was broken, and they fled in wild confusion toward their
camp on the hill of Ayadieh. Saladin could not rally them, and
many of the fugitives were so panic-stricken that they rode
without drawing rein as far as Tiberias, or even Damascus.
Following the routed Turks, the two divisions of the Prankish
centre stormed up the hill and plunged into the camp. It would
have been hard to keep them in order among the tents and
other obstacles which broke their line ; but, as a matter of fact,
no one made any attempt to restrain them. Horse and foot
scattered themselves through the encampment and turned, some
to slaughter and some to plunder. The Sultan s own pavilion
was sacked and cast down, three of his body servants being slain
therein. Some of the Franks turned to cutting down the camp-
followers, others burst into the sutlers quarter and plundered
the market. No one made any attempt to prevent the routed
Turks from rallying, or to take in flank the still intact wings of
Saladin s army.

Meanwhile, King Guy and the right wing of the Franks
seem not to have come to a decisive engagement with the
Kurds and Mamelukes of Saladin s left. Neither Western nor
Eastern writers give any clear account of the movements in this
part of the field. It seems likely, however, from a passage in
Ibr-Alathir, that the Moslems were somewhat outflanking the
Christians, since the latter had partly followed the advance of
their centre. Lest the enemy might use the opportunity and
get between him and the camp, the king may probably have
held back.

By the most untiring personal exertions Saladin at last
succeeded in gathering together a great part of his routed centre
somewhere at the western foot of the Ayadieh hill. His officers
besought him to lead them to storm their lost camp, but he
refused, and bade them wait till the Franks should leave it, and
then to charge them when their backs were turned to the

336 THE ART OF WAR IN THE MIDDLE AGES [1189

Moslems. Before long the Germans and Lombards began to
evacuate the hill, some burdened with plunder, others wishing to
re-form on the open ground and then to go to the help of the
king or the Templars. The retreat was executed in great dis
order, and not without panic : many thought that some disaster
had happened in the rear to account for the fact that their
comrades were tramping down hill. The author of the Itinerarium
tells us that in one part of the field a knot of Germans, running
to catch an Arab horse which had broken loose, were supposed
by the rest to be flying, and caused a senseless rush to the rear.

When the Christians were trooping in disorderly masses back
to the plain, Saladin suddenly let loose his rallied Mesopotamian
horsemen upon them. The results of this charge were decisive:
the scattered bands of Crusaders were caught wholly unprepared ;
they had no time to form up and defend themselves, but were
hurried back across the plain by the shock of the Turkish
horsemen. In utter rout some fled toward King Guy s corps,
some straight to the camp. Saladin followed, slaying the hind
most and easily driving all before him. The crusading right
wing seems to have made some attempt to rescue the fugitives,
and Guy himself is said to have saved the life of his old enemy,
Conrad of Montferrat, by hewing out a passage for him when he
had been surrounded by the pursuers. 1 But the king and the
Hospitallers could not restore the battle, and were themselves
thrust back towards the camp by the rushing mass of pursuers
and pursued. Apparently the Turkish left wing tried to push
itself between the Franks and their place of refuge, 2 and, though
it failed to cut off their main body, its movements must have
hastened the retreat. The flight only ceased when James of
Avesnes and Geoffrey de Lusignan led the reserve out of the
camp and covered the flight of the disorderly crowd of horse and
foot to their tents. Saladin halted below Mount Turon, and
would not allow any attempt to be made to storm it : he dreaded
the strength of the Franks when acting on the defensive.

Meanwhile, a separate battle had been fought on the hillside
to the north by Taki-ed-din and the Master of the Templars.
We have already mentioned that, after the first shock, the fight
had come to a standstill in this quarter, owing to the reinforce
ments which Saladin had sent to his nephew. A second acces
sion of forces to the Moslems settled the fate of the combat.

1 Itintrarium, p. 71, cap. xxx. 2 Ibn-Alathir.

1189] ACRE: THE CAUSES OF DEFEAT 337

Seeing the Christians engaged in the battle and paying no heed
to the town, the garrison of Acre sallied out five thousand strong,
from the northern gate, that most remote from Mount Turon. 1
Then, taking a circuitous route, they came out upon the rear of
the Frankish left, and fell upon the Templars and the Champenois
while the latter were hotly engaged with Taki-ed-din. The
intervention of this new corps broke the spirit of the Crusaders.
They gave up all for lost, and merely strove to cut their way
back to their camp. Being beset in front and rear, it was only
a portion of them who succeeded. Eighteen knights of the
Temple fell, and their Grand Master, Gerard, was captured, and
beheaded by Saladin s orders. Andrew of Brienne, the brother of
the Champenois count, was also slain, and forty knights more.
So great was the slaughter in this part of the field that the numbers
of the fallen in the Christian left wing far exceeded those lost
by the right and the centre. 2 Thus ended in defeat a battle
which might under proper guidance have led to the complete
discomfiture of the relieving host. The Franks had risked much
by engaging in the vast plain of El-Ghawarneh, where their
corps were certain to get separated the one from the other.
Nevertheless, the misbehaviour of the Sultan s centre put the
victory into their hands. If, instead of falling on the camp, and
there wasting a precious hour, Conrad and the Landgrave had
turned to take the Turkish wing-divisions in the flank, the
Infidels could not possibly have escaped a dreadful disaster.
Taki-ed-din s corps might have been hurled into the sea, and the
Kurds and Egyptians thrust into the marshes of the Belus, if
either of them had delayed a moment too long before taking to
flight. But when the battle was really won, the leaders and the
led were equally incapable of using their advantage. The men
turned to pillage, and we have no proof that any of their
officers thought of calling them off or conducting them to
another part of the field. Hence the Sultan, with his usual
ability, was able to rally his men, and snatch a victory out of
the jaws of defeat.

1 Itiiierarium, p. 70, and letter of Theobald and Peter Leo in Ralph de Diceto.

2 Boha-ed-din (p. 145) took great pains to make out the sum of the Christian
losses. He considered the number of seven thousand, that which was generally
accepted in the Sultan s camp, as exaggerated. But having questioned the officer who
had been charged to make away with the Christian corpses on the northern part of
the field, he was told that four thousand one hundred had been carted off. He
therefore estimated the losses of the right and centre at less than three thousand.

338 THE ART OF WAR IN THE MIDDLE AGES [1189

Our only wonder is that he did not utilise his success for a
further assault on the Franks. But he had a wholesome dread
of the enemy when acting on the defensive, and (as we are told)
his own army was in the greatest disorder. Not only the
Crusaders, but the Turkish camp-followers had turned to pillag
ing the tents on the hill of Ayadieh, and for the whole day afte*
the fight, as we read, the troopers were occupied in seeking their
lost goods and extracting them from the plunderers. When a
few hours were past, the Christians, whose losses had been far
less than might have been expected, only the left wing had
really suffered much slaughter, were safe in their camp, and more
angry than afraid. When the Sultan held back, they were so far
from being cowed that their next move was to run a line of
circumvallation from sea to sea, and actually seal up the garrison
of Acre within its walls.

As to losses, we have no good account of those suffered by
the Moslems. The contemporary letter of Theobald and Peter
Leo to the Pope estimates them at fifteen hundred horsemen, a
not improbable figure. Boha-ed-din names as slain the Kurdish
Emir Modjelli and a few more chiefs, together with about a
hundred and fifty persons of no importance. Considering the
rout of the centre, these numbers are wholly improbable, and
cannot be accepted. On the other hand, the Christian sources
give the loss of the Crusaders at fifteen hundred only, 1 naming
Andrew of Brienne and Gerard the Grand Master as the only
notable men among the slain. These figures are equally incred
ible, especially in face of Boha-ed-din s statement as to the
counting of the corpses. 2 On the whole, we may perhaps guess
that each side made a better estimate of its enemy s losses than
its own, and put them at fifteen hundred Turks to seven thousand
Franks.

Battle of Mansourah, February 8, 1250.

In our chapter on the Strategy of the Crusades we have
already had occasion to mention the battle of Mansourah as the
ill-fought end of an ill-planned advance into Egypt. We pointed
out the madness of a march across the canals and waterways of
the Delta, and showed how the campaign was certain to end in
a check, owing to the numerous and strong defensive positions
which were in the hands of the Egyptian army.

1 ftinerarium, p. 72. 2 See p. 337.

1249] ST. LOUIS IN EGYPT 339

St. Louis started on his adventure under much more favour
able circumstances than his predecessor King John of Jerusalem
had met thirty years before. The Crusaders of 1219 had only
secured themselves a basis of operations by the capture of
Damietta after besieging the place for a year. Their strength
was exhausted before theyeven started on their march up-country.
By an extraordinary chance St. Louis in 1249 took the town
without striking a blow. All Egypt was in disorder owing to the
mortal sickness of Sultan Malek-Saleh, 1 and there was no single
strong hand at the helm. When the troops who had been told
off to oppose the landing of the French were beaten back, and
retired towards the interior, the corps which had been selected
to garrison Damietta evacuated the place in a panic and fled
after the rest. 2 It was to no purpose that the Sultan roused
himself from his sick-bed to order fifty of their officers to be
hanged : the strong city had passed into the hands of the
Crusaders, and gave them a secure starting-point and place of
arms : it was full of stores and in perfect order, since there had
been no occasion to batter its walls with siege engines (June
6, 1249).

Having begun so well, it was incumbent on the French king
to utilise his first success and push forward while the enemy
were still panic-stricken. It is therefore with nothing less than
astonishment that we hear that King Louis waited nearly six
months at Damietta before he began his march on Cairo. The
circumstances explain, but do not excuse, this halt : a large part
of the armament had been blown into the Syrian ports by a
contrary wind, and it was thought necessary to await its
appearance : the summer was at its height, and the Nile flood
was rising over Lower Egypt, so that the face of the land was
well-nigh covered with water. These would have been good
reasons for delaying the attack on Damietta till the approach of
the cold weather and the sinking of the flood : it was obviously
the worst possible month for an advance when the heat was at
its greatest and the country most water-logged. Undoubtedly
June was a bad season for the invasion, but, having once begun,
the French were bound to go on : their delay merely enabled

1 The Sultan was dying of a malignant ulcer in his thigh, which contemporary
rumour ascribed to his having Iain on a poisoned mat spread for him by one of his
slaves.

a Makrizi in the Bibliothique &lt;/ Croisades, iv. 42.

340 THE ART OF WAR IN THE MIDDLE AGES [1249

the Sultan to organise his resistance with a clear knowledge of
the route which his enemies must take. There had been a fear
ful panic at Cairo when the news of the fall of Damietta arrived,
but the long quiescence of the Franks enabled the Egyptians to
recover their self-possession and bethink them of the best means
of defence. 1

It was not till October that the last contingents of the French
army straggled in from Syria : they had brought with them a
number of the barons of the Holy Land, who placed themselves
under the Count of Jaffa. 2 There was some discussion when the
whole host was mustered as to whether it should not be trans
ferred to Alexandria, 3 and attack Egypt from that side. This
plan was supported by Peter of Brittany and many other barons,
and had its advantages, for the march into Egypt from Alexandria
presents far less difficulties than that from Damietta. But it
must have begun with a second disembarkment and a toilsome
siege. When the king s brother, Robert Count of Artois,
explained that those who wish to kill the snake strike at its
head, 4 and voted for an immediate advance on Cairo along the
Damietta branch of the Nile, he carried the king and the council
with him, and the hopeless march began.

On the 2oth of November 6 the army commenced its march,
moving slowly forward past Fareskour, Scharemsah, and Fara-
moun, while the flotilla advanced parallel with it on the Nile.
A few miles after Faramoun was passed, the advance came to
a standstill (igth December), when four weeks had been occupied
in advancing fifty miles. The check was caused by the fact
that the king found in front of him the first formidable water
course which cuts the way from Damietta to Cairo. At the
town of Mansourah the Damietta branch of the Nile divides
itself into two parts : the one flows down to Damietta, the other
turns east and falls into the swamps of Lake Menzaleh. It was
in front of the latter that the Christian army found itself stopped;
this second waterway, which the natives call the Ashmoun
Canal, lay across its path. Behind it the whole levy of Egypt
was massed ; the Sultan had taken post there when Damietta

1 Jemal-el-din in the Bib. des Croisades, iv. 451, 452.

2 John of Ibelin. He had himself been with the king at the first landing (Join-
ville, p. 215).

3 By sea, I presume : not even the French barons can have dreamed of marching
over three branches of the Nile and the whole breadth of the Delta.

4 Joinville, p. 219. 5 William of Nangis, p. 374.

1249] THE FRENCH REACH MANSOURAH 341

fell, knowing that it was the first strong defensive position which
the French must attack. Just as the critical moment was
approaching, his old malady carried him off in the last week of
November, and he had been dead some time when St. Louis
reached Mansourah. His widow and his ministers, however, kept
his death secret, and orders were still issued in his name. The
real charge of the defence of Egypt fell to the Emir Fakr-ed-
din, the commander of the army, on whom it was agreed to
confer dictatorial powers. Meanwhile, swift messengers were
sent to seek Malek-Saleh s son and heir, Turan Shah, who was
far away at Hisn-Kayfa in Mesopotamia. Till he should arrive
the Sultan s death was concealed from his subjects.

The French army now found itself at the point of a narrow
tongue of land, an "island" as Joinville calls it, between the
main branch of the Damietta Nile and the Ashmoun Canal.
It was necessary to force the passage of one or the other of
these waterways ; and, both because it was smaller and because
it covered the direct road to Cairo, the king chose the Ashmoun
as his objective.

Opposite him lay the tents of the Egyptian army, stretching
for two or three miles along the farther bank. In their midst
rose the walls of Mansourah, and outside its western gate the
palace of the Sultan. The place was but thirty years old ; in
1220, after he had beaten John of Brienne on this same ground,
the Sultan El-Kamil had built a new city to commemorate his
victory. The strategical exigencies of the roads of the Delta
had placed St. Louis and Malek-Saleh in exactly the same
position as was occupied by their predecessors during the fifth
Crusade.

The Egyptian army was now composed of better stuff than
had been the case in 1 220. It was Malek-Saleh who had first
organised the celebrated corps of the Circassian Mamelukes
which was to dominate Egypt for the next six centuries. The
mercenary troops of his predecessors had been mainly Kurds
and Syrians, but he had learned the military worth of the men of
the Caucasus, and had been steadily buying Circassian slaves
for many years and incorporating them in his guard. The eight
or ten thousand Mamelukes formed the core of his host: to
support them were arrayed the horsemen of the Bedouin tribes
and the general levy of Egypt, who had marched out at the
exhortation of their mollahs and imams to save Islam. These

342 THE ART OF WAR IN THE MIDDLE AGES [1250

formed a great mass of troops, both horse and foot, but were of
little military value. The whole brunt of the war fell upon the
heavily-armed and well-mounted Mameluke horsemen.

Seeing the Egyptians clustering so thick around Mansourah,
St. Louis resolved not to make any attempt to throw his army
across the canal by means of his boats, but to build a solid
causeway and so darn up the channel and cross on foot. Accord
ingly he set his foot-soldiery to cast earth into the Ashmoun on
a broad front ; the causeway advanced a few yards, but soon
the discharge of missiles from the opposite bank became so
deadly that the work was stopped ; the king saw that the earth-
bearers must be protected, and therefore built along the
incomplete dam two " cats," i.e. covered-ways or penthouses,
under shelter of which he trusted that the workers might complete
their task. The "cats" were protected by two high wooden
towers called " belfreys " placed at the water s edge. To batter
down these protections the Egyptians soon set their military
machines to work, and sixteen perrieres and mangonels hurled
large stones or barrels of combustible matter at the covered-
ways and wooden towers. The French replied by setting up
against them eighteen similar engines, and the two parties shot
at each other across the river for some days.

As long as the " cats " were safe the causeway could advance,
and the labourers succeeded in filling up the ted of the canal for
more than half its breadth. But on the other side the Egyptians
began cutting away the bank, and, the force of the current
aiding them, they succeeded in keeping the Ashmoun open.
"In one day they undid what it had taken us three weeks to
accomplish," says Joinville, "for all our work in stopping the
channel was useless when they enlarged it on the other side." 1

Meanwhile, Fakr-ed-din threw a detachment across the canal
lower down its course, and sent them to fall on the rear of the
French camp : they were, however, beaten off with some loss by
the king s brothers, the Counts of Anjou and Poictiers (December
25, 1249). This was but a diversion: the real centre of the
fighting was the causeway ; here the matter finally went ill with
the French. By hurling barrels of Greek fire at the belfreys and
"cats," the Infidels finaily succeeded in setting them in flames.
Nothing could be done till they were rebuilt with ship-timber
which the king bought for the purpose. But only a few days

1 Joinville, p. 221.

PLATE XL

Rosttta

Scbennytic
ya\

I .im H
Mouth

To Illustrate the Crusades.

of Ou. JCiru&gt;r
and/Watf.7&gt;cotirses o-re omitted Olzeh

A. Place of S Louis I)yXe &: Engines.
B. Place of the Egyptian Engines.

NEIGHBOURHOOD

MANS OUR AH.

124-9-50

1250] MANSOURAH : THE FRENCH AT THE FORD 343

after the new engines had been erected, they were again burned by
the same means as before.

A deep discouragement now pervaded the French host: it
seemed that they had been brought to a complete standstill.
But a few days later the Constable Humbert of Beaujeu dis
covered a Copt or a renegade Mussulman 1 who told him that
four miles to the east of Mansourah there was a ford over the
Ashmoun, deep and difficult indeed, but quite practicable for
cavalry (Feb. 7, 1250).

The army had now been stranded for nearly two months in
front of Mansourah, and Louis felt that he must leave no device
untried, even though it were as dangerous as that of crossing a
deep ford in face of the enemy and without any possibility of
aid from his infantry. He accordingly resolved to attempt the
passage on the next morning.

During the night of the 7th - 8th February his disposi
tions were made. The Duke of Burgundy and the barons of
Palestine with their knights were to remain behind in the camp,
and take charge of the great mass of foot-soldiery. When the
king should have reached and captured the Egyptian machines
which commanded the half-built causeway, they were to complete
it in all haste and cross over to join their leader.

Meanwhile, Louis himself, with his three brothers, Charles of
Anjou, Robert of Artois, and Alphonso of Poictiers, and the
main body of the horsemen, was to march to the ford and pass
it at daybreak. When they were on the southern bank they
were to push along it to the Egyptian camp, burst into it, and
capture or destroy the engines at the causeway before the enemy
should recover from his surprise.

We have no complete account of the array of the cavalry
corps which marched to the ford. We know, however, that the
Templars, under their Grand Master, William de Sonnac, rode
first, and that the van division included also the followers of
Robert of Artois, Peter Duke of Brittany, John Count of
Soissons, Raoul lord of Coucy, and the small English con
tingent which William Longsword, the titular Earl of Salisbury, 2

1 Joinville, p. 220, calls him a Bedouin, so does William of Nangis. But some of
the Mohammedan writers call him a Copt.

- Henry III. had refused to give him his father s earldom, and conferred a pension
on him instead. But William was nevertheless called earl by most of his contempor
aries.

344 THE ART OF \VAR IN THE MIDDLE AGES [1250

had brought to the Crusade. They had with them all the
king s mounted crossbowmen. In the second battle among the
Champenois was John of Joinville, who has left us our best
account of the campaign ; unfortunately he has omitted to give
us the complete list of those who marched with him. Charles of
Anjou was probably commander of the corps. 1 The king and
his household knights, with his brother Alphonso of Poictiers
and Henry Count of Flanders, rode in the third division. Louis
had issued strict orders that no knight should straggle from his
corps, and that the three battles should keep close together;
the van was not to advance till all three had passed the ford. 1

The Egyptians kept a careless watch along the canal, and
though the ford was only four miles from their camp, at the
village of Sahnar, the French reached it unobserved. The van
division crossed, not without some difficulty, for the bottom was
muddy and the opposite bank scarped and slippery : a few
knights lost their footing and were drowned.. When they were
already over, a body of three hundred Arab horse appeared, but
promptly took to flight when the Count of Artois charged them ;
they saw that the passage was lost, 3 and rode off to warn their
comrades.

Flushed with this trifling success, Robert of Artois forgot
his brother s orders, and began to move off in pursuit. The
Master of the Temple rode up to him and besought him to stop,
but the hot-headed count would not listen to his remonstrances,
and spurred off towards the Egyptian camp. Thinking that he
would be shamed if he abandoned his place in the van, the
Master unwillingly followed, and after him all the other con
tingents of the van battle. 4

Count Robert rode so hard and so re: that he came

hurtling into the eastern end of the Egyptian camp almost as
soon as the flying Bedouin whom he was chasing. He found
the Infidels in a state of disarray and unpreparedness. which
reflects little credit on their commander. The horses were not

- : gather from the fact that he rescued Joinrille before the king and the third
corps had reached the field (Joinville, p. 226).
- Rothelin MS., p. 602.

3 Toinville, p. 224. They appeared when Toinville himself was crossing, i.e. after
the van had passed.

4 Toinville tells a curious tale of a deaf knight who was pulling the count s bridle
and shouting " Forward and at them ! " at the top of his voice all the time that the
Master was pleading for delay.

1250] MANSOURAH: COUNT ROBERT S CHARGE 345

saddled nor the men armed. The French rode through the
camp, slashing right and left and driving all before them, till
they came to the place where were the perrieres and mangonels
which commanded the unfinished causeway. They wrought
great slaughter, and killed the Emir Fakr-ed-din himself, fresh
from his bath and without his coat-of-mail, as he rode up and
down trying to rally his men. Hitherto Robert s haste had not
done any irreparable harm : if he had halted and taken post
among the machines to guard the spot till the infantry should
complete the work, he would almost have justified his reckless
charge. For if he had waited till the second and third battles
had crossed the narrow ford, the enemy would have had ample
time of warning, and would not have been surprised in their
camp.

But the fiery count was now to take the fatal step which
ruined the whole enterprise. Seeing the Egyptians in hopeless
disarray, he imagined that he had gained the day with his own
division alone, and thought of nothing but pursuit and slaughter.
After a very short breathing space, he ordered a second advance
towards the town of Mansourah, into which many of the fugitives
were pouring. The Master of the Temple again besought him
to pause and await his brother s coming, and William of Salis
bury added his remonstrances to those of William of Sonnac.
Count Robert replied with inexcusable discourtesy, telling the
Templar that the military Orders loved to protract the war for
their own ends, and did not really wish Christendom to triumph,
lest their own occupation should be gone. 1 Then, turning to the
Earl of Salisbury, he flung in his face the old taunt about
" Englishmen with tails " and the curse of cowardice that rested
on them. " I shall go this da}- where you will not dare to keep
level with the tail of my horse," replied Salisbury, and, replacing
his helmet and lowering his lance, he rushed forward with the
rest to meet his fate. 2

1 Artois language to the Templar, as reported by Matthew Paris (v. 149), deserves
record as showing the suspicion which the Crusaders entertained of the military Orders.
" O antiqua Templi proditio ! Hoc est quod diu praecinimus augurio, quod terra tola
Orientalis jamdiu fuisset adquisita nisi Templi et Hospitalis fraudibus nos seculares
impediremur. Timent autem Templarii et eorum complices quod si terra juribus subdatur
Christianis, ipsorum expirabit (qui amplis reditibus saginantur), dominatio. Hinc est
quod fideles ad negotium crucis accinctos %-ariis inficiunt potionibus, et Saracenis con
foederati proditionibus interficiunt."

2 Matthew Paris makes a bad error in placing this altercation after instead of
before the irruption into the town of Mansourah.

346 THE ART OF WAR IN THE MIDDLE AGES [1:50

The Egyptians were still so discouraged that Artois and his
followers were able to penetrate within the walls of Mansourah
and to ride through the town, cutting down the fugitives; some
of the knights even emerged at its western gate, and almost
reached the Sultan s suburban palace. But they were scattered
in the streets and separated one from another, so that the
impetus of their charge and the advantage of combined action
were lost. The Egyptians fled into the houses and flung darts
and tiles upon the knights as they galloped up and down the
narrow lanes. Presently the troops from the camps west of the
town, who had not shared in the panic of the rest of the Moslem
army, began to pour into Mansourah. They found the French
scattered in small bands, some intent on plunder and some on
slaughter, but all unprepared to receive a fresh attack. Hence
the new-comers won an easy success over the Christians : many
were slain in the streets, others hunted out of the town and
cut down in the open. The only route which the fugitives
could take lay through the eastern camp of the Egyptians,
where the Mamelukes were now rallying and getting into battle
order. Hence it is not surprising to find that nearly the whole
of Artois corps was annihilated. He himself was slain in the
town, and his surcoat with the royal French lilies was exhibited
to the Moslems as a proof that the King of the Franks had
fallen. With Robert there died William Longsword, the Master
of the Temple, the lord of Coucy, and man) barons more.
Joinville tells us that three hundred knights perished, besides
the sergeants and horse-arbalesters who accompanied them. 1
The Temple alone lost two hundred and eighty horsemen of
various ranks. The Moslems say that fifteen hundred French
were cut off in all, 2 and the figure is very probably correct.
Only a few scattered bands escaped, among whom were the
Duke of Brittany and the Count of Soissons.

Meanwhile, during the hour which Artois had wasted by his
mad charge, the remainder of the French cavalry had been
gradually crossing the Ashmoun. Joinville, who was in the
front of the middle corps, seems to have followed Count Robert
at a distance, before the king was well over the ford. At any
rate, he saw, when he reached the Egyptian camp, that some of
the enemy were already rallying, having retired from the tents
into the open fields where they were drawing up in line of battle.

1 Joinville, p. 224. * Makrizi.

1250] MANSOURAH: THE MAIN BATTLE 347

The seneschal charged the nearest squadron, but was soon swept
back to the edge of the canal by the advance of the mass of the
Infidels, whom he estimated at about six thousand horse. He
and his followers only saved their lives by retiring into a ruined
house, where they maintained themselves, fighting on foot in the
doorway, till Charles of Anjou and the main body of the second
corps came up and delivered them by driving off their assailants.

Soon after, King Louis himself and the rear division came
upon the scene of battle. They were at once assailed by the
Mamelukes, who were now rallied and in good order. A fierce
struggle began in the outskirts of the camp, and was maintained
for many hours. The Mamelukes poured a constant rain of
arrows into the ranks of the French, and Louis was compelled
to charge them again and again before he could resume his
advance towards the all-important spot where the half-finished
dam lay. It was absolutely necessary to reach it, in order that
the infantry might have their chance of joining the horse. But
being continually attacked on their left flank, the French could
not advance as they wished, but were always having to face
southward to beat off the Mamelukes. Seeing their enemy
growing weary, and noting that hundreds of the knights were
dismounted owing to the loss of horses under the rain of arrows,
the Mamelukes at last threw their bows over their backs and
charged down with mace and scimitar upon the king. Louis
was hard pressed, and some of his followers lost heart and
plunged into the Ashmoun to swim back to the Christian camp.
But he persisted in his original plan of advancing to the cause
way, and at last came level with it.

Then the French infantry, throwing earth, planks, fascines,
broken military engines, and all manner of miscellaneous rubbish
into the unbridged half of the canal, succeeded in making a rough
but sufficient bridge over the gap. The arbalesters and pikemen
began to pour across the crazy structure by thousands. Humbert
of Beaujeu, the constable, at once drew up the first crossbowmen
that arrived so as to cover the harassed cavalry. They opened
a destructive fire upon the Mamelukes, and the battle took a
new turn.

The moment that the Egyptian leaders Bibars, who twenty
years later became Sultan, is chiefly named among them saw
that the French infantry were entering on the scene, they ordered
their horsemen to draw back. Retiring out of bowshot, they still

348 THE ART OF WAR IN THE MIDDLE AGES [1250

maintained a threatening attitude. The king might now have
advanced, but his knights were so thoroughly tired out and
harassed that he refrained from doing so. He contented him
self with ordering the infantry to construct a large circular tete-
du-pont covering a considerable space of ground on the farther
bank of the canal. The work was easily and rapidly finished
by using as materials the woodwork of the captured Egyptian
machines.

Thus King Louis had acquired a solid lodgment on the
southern side of the watercourse which had so long held him
in check. But he had failed to defeat the Egyptian army, which
still watched him at the distance of no more than a few hundred
yards, and was rather encouraged than abashed by the results
of the day s fighting. The losses of the French had been so
much greater than those of their adversaries that the Moslems
regarded themselves as the successful party. Louis had lost, as
far as can be calculated, nearly half his cavalry and a still greater
proportion of his horses. The real meaning of the battle was
sufficiently shown by the fact that three days later 1 the
Egyptians assumed the offensive, and vigorously attacked the
tete-du-pont, while the French stood entirely upon the defensive,
and even after beating off the assault made no further attempt
to advance. The invaders had lost their impetus and their
desire to push on : not long after we find them thinking of
retreat. The battle, though it had ended in the crossing of the
Ashmoun Canal, had so exhausted the Crusaders that they
despaired of the result of the campaign. We cannot call it
anything but a check and a disaster. 2

Such were the main features of the fight of Mansourah, the

1 The battle had been fought on Shrove-Tuesday, and the Moslem attack on the
French lines followed on the first Friday in Lent.

- Joinville s interesting personal adventures after the king had come upon the
field are well worth reading, but evidently had no important influence on the fortune
of the day. He had been employed to ride on to Mansourah to look for the Count of
Artois, who was said to be yet alive, but got involved in a long skirmishing encounter
with a body of Egyptians on and about a little bridge which crossed a brook running
into the Ashmoun from the south. lie succeeded in detaining opposite him a body of
the enemy who would otherwise have gone to aid in the attack on the king. But
their arrival would not have turned the event of the day indeed, these were Egyptian
rabble, not Mamelukes, as many of them were on foot, and they pelted Joinville and
his men with clods and shot at them with fire-arrows instead of charging in. His pp.
227-228 are of great interest, but we could wish that they contained more details about
the king s main fight with the Mamelukes.

1250] MANSOURAH: THE CAUSES OF FAILURE 349

last of the great pitched battles of the Crusades. It displays,
even more clearly than the other engagements with which we
have dealt, the absolute interdependence of cavalry and infantry
in the Christian hosts when dealing with the formidable horse-
archers of the East For want of men armed with missile
weapons (all the mounted crossbowmen had been slain along
with Robert of Artois) the king and his chivalry were on the
very verge of destruction. They were saved the moment that
their infantry succeeded in getting across the canal and joining
them. Without that succour they would probably have been
destroyed to the last man, for they had been cut off from their
retreat to the ford, and the watercourse at their back proved
impassable to such fugitives as attempted to cross it.

It is curious to note that the Mohammedan writers grasped
much more clearly than the Christian the fact that the tardy
arrival of the French infantry turned the engagement into a
drawn battle, and that their earlier appearance would have made
it a decisive victory for St. Louis. Joinville 1 and William of
Nangis 2 mention the coming up of the crossbowmen indeed, but
keep all their interest and admiration for the king s feats of
personal valour. It is left for Jemal-ed-din and Makrizi to
observe that " if the first division of the Christian cavalry had
held out " (i.e. if Artois had remained by the engines instead of
plunging into Mansourah), "and if the whole of the Christian
infantry had been engaged, Islam would have been ruined," 3 and
that " if the French infantry could have joined their cavalry, the
defeat of the Egyptians and the loss of the town of Mansourah
would have been inevitable." * Blinded by chivalrous enthusiasm
and class-pride, the French chroniclers omit to draw the moral
which to the Moslem writers was obvious.

The separation of horse and foot while St. Louis was making
his turning movement was unfortunate, but absolutely necessary.
We cannot blame the king for it, as he had no other alternative
before him. All the more must the gravest censure fall on

1 " It happened that towards evening the king s constable, Humbert de Beaujeu,
brought us the foot -arbalesters, who drew up in front, while we dismounted. Incon
tinently the Saracens went off and left us in peace," says Joinville a very inadequate
account of the crisis of the day, when whole pages have been devoted to individual
exploits.

2 " Nostri usque ad horam nonam graves sustinuerunt impetus. Tandem balistariorum
subsidio multis Saracenorum vulneratis . . . nostri campum obtinuerunt " (p. 374).

s Jemal-ed-din, p. 459. * Makrizi, p. 548.

350 THE ART OF WAR IN THE MIDDLE AGES

Robert of Artois for his mad charge into Mansourah in direct
disobedience to his brother s orders. If he had only halted
among the Egyptian engines opposite the French camp, and
held his ground there till the infantry could complete the cause
way, and till his brother could arrive with the main body of
the horse, the day would have gone well for Christendom. The
king did his best to detain him, sending ten knights to bid him
halt and wait, 1 but Robert, in deliberate defiance of his chief,
chose to make the second mad charge, which lost the day and
ended his o\vn rash career. Even the leader of a feudal army
could not have rationally expected to see his plans wrecked by
such a piece of wanton and wicked indiscipline. .

1 Joinville, p. 224.

BOOK VI

WESTERN EUROPE

FROM THE BATTLE OF HASTINGS TO THE
RISE OF THE LONGBOW

CHAPTER I

INTRODUCTORY

IN studying the Crusades we have seen the military art of the
nations of Western Europe at its best and its worst-
Nowhere are more reckless displays of blind courage, or more
stupid neglect of the elementary rules of strategy and tactics
to be found, than in the great expeditions to the Levant. On
the other hand, we have also had to observe among the more
capable leaders of the crusading armies a far higher degree of
intelligent generalship than was usual among their contem
poraries in the West. If the Crusades of 1101 and 1147 are
decidedly more distressing to the critic than the average wars of
France, England, or Germany.there are also battles and campaigns
, such as that of Arsouf which show very favourably beside
those of the lands nearer home. Many of the Crusaders seem to
have been at their best when facing the new problems of the
East. Richard Coeur de Lion at Acre, Arsouf, and Jaffa rises
far above his ordinary level : we find ourselves wondering how
the very capable general of 1190-91 can on his return waste so
much energy and ability to no purpose in the wretched peddling
French wars of 1 194-99. We may add that the great Frederic I.
of Germany never shows to such good effect in his home cam
paigns as in the conduct of his expedition through Asia Minor.
Many of the lesser figures of the Crusades, including the good
Godfrey of Bouillon himself, are obscure and undistinguished in
the wars of their native lands, and only show the stuff that is in
them Avhen they have crossed the high seas.

The worst military errors of the Christians in the East came,
as we have seen, from their gross ignorance of the conditions of
warfare in Syria or Asia Minor, and of the tactics of the enemies
with whom they had to deal. At home leaders and led alike
were safe from such dangers, since they knew the military-
23

354 THE ART OF WAR IN THE MIDDLE AGES [tioo

character and usages of their neighbours, and had some rough
idea of the geography, climate, and productions of their neigh
bours territory. But if this knowledge preserved them from
certain dangers, it seems, on the other hand, that in the familiar
border wars of the West the best qualities of a commander were
often not developed. It is new and unforeseen dangers and
difficulties that test most adequately the stuff that is in a man.

When we turn from the history of the Crusades to consider
the contemporary history of the Art of War in Western Europe,
the first thing that strikes us is the comparatively small influence
which the great campaigns in the Levant seem to have had upon
the development of strategy and tactics at home. Tens of
thousands of barons, knights, and sergeants came back as
veterans from the East, and one would expect to see the
lessons which they had learned in fighting the Turk and Syrian
perpetually applied to the wars of their native countries. Yet
it is by no means easy to point out obvious instances of such
application of new principles of war, save in the provinces of
fortification and of arms and armour. In strategy and tactics
it is difficult to detect from a broad survey much direct influence
flowing from the Crusades.

We may take as the clearest example of this the entire
neglect by the Western nations of the most important tactical
lesson of the Crusades. \Ve have shown by a score of examples
that the one great principle which settled the fate of wars with
the Turk was that generals who properly combined infantry and
cavalry in their line of battle were successful, and that generals
who tried to dispense with the support of foot-soldiery always
failed disastrously. The fact that the combination of the two
arms is better than simple reliance on one had been shown at
Hastings long ere the Crusades began, but the lesson was even
more clearly visible in the details of such fights as Antioch or
Ascalon as compared with the disasters of 1 101 or the narrow
escape from destruction at Dorylaeum.

We should expect, therefore, to find that the return home of
the warriors of the first Crusade would be followed by the
development of a rational use of infantry and cavalry in close
alliance and interdependence. But we find little of the kind :
over the greater part of Western and Central Europe the
cavalry arm still maintains its exclusive predominance, and
infantry is still despised and distrusted. In Italy, it is true, the

i zoo] THE PREPONDERANCE OF CAVALRY 355

workings of the experience of the Crusades are to be recognised
in the sudden growth of the popularity of the crossbow, and
probably also in the increased importance of the civic infantry.
But in the only other parts of Europe where foot-soldiery show
to any effect England and the Netherlands we are dealing
with an old Teutonic survival, not with any new development.

In many of the twelfth-century battles of Western Europe,
when by some rare exception we do find combatants on foot
entrusted with a principal part in the fight, we discover on
closer inquiry that they are not ordinary foot - soldiery, but
knights who have dismounted in order to carry out some
desperate duty. We are, in short, merely witnessing a recurrence
to that ancient habit of the Teutonic races which Leo the Wise
had described two hundred years before. 1 Such instances are
to be found on the part of the English and the Normans at
Tenchebrai 2 (lioC), and again at the first battle of Lincoln 3
(1146), where both King Stephen and the rebel earls dis
mounted the pick of their knights to form a solid reserve. The
same is the case in the English army at Bremule (ing), and at
the battle of the Standard 4 (1138;, where the Yorkshire knights
left their horses and joined the yeomanry of the fyrd in order to
stiffen the mass when it was about to be assailed by the wild
rush of the Scots. The Emperor Conrad s German chivalry
behaved in a similar way at the chief combat during the siege
of Damascus in 1148.

Such expedients, however, are exceptional. On the other
hand, we not unfrequently find battles in which neither side
brought any foot-soldiery to the field, such as Thielt (1128),
Tagliacozzo (1268), and the Marchfeld (1278). Cases where one
side had no infantry whatever in the battle line are still more
numerous. Such are Bremule (1119), Legnano (1176), Muret
(1274)-

When true infantry are engaged on both sides, it is rare
to find them actually settling the fate of the day. Generally
they are only used as a very subsidiary force, employed merely
for skirmishing and not for the decisive charge. The main
exceptions to this rule arc to be found, as we shall have to show
later on, in Italy and the Netherlands. But if the infantry
in most battles had no great part in the winning of the day,
they were often the chief sufferers in a defeat. As a rule, those

1 See p. 202. J See p. 379. 3 See p. 392. * See p. 386.

356 THE ART OF WAR IN THE MIDDLE AGES [1250

of the beaten army were fearfully mishandled by the knights of
the victorious side. When the day was won, the infantry of the
vanquished party were nearly always cut to pieces in the most
ruthless manner, while their countrymen of the knightly classes
were not slaughtered, but reserved for ransom.

The mailed horseman, then, maintains his place as the chief
factor in battle down to the end of the thirteenth century, and
the main features of the two hundred years from Hastings
onward are the feudal knight and the feudal castle. We shall
have to note that while tactics and strategy make comparatively
small and slow progress in these t\vo centuries, the art of forti
fication grows very rapidly. Between the simple castle of the
time of William I. and the splendid and complicated fortresses
of the end of the thirteenth century there is an enormous gap.
The methods of attack made no corresponding advance, and by
1300 the defensive had obtained an almost complete mastery
over the offensive, so that famine was the only certain weapon
in siegecraft. It is not till the introduction of cannon and gun
powder in the fourteenth century that the tables begin to be
turned.

In chapter iii. of Book III. we dealt with the origin and
evolution of the feudal knight and the feudal castle. We have
now to treat of their further developments.

CHAPTER II

THE ARMIES OF THE TWELFTH AND THIRTEENTH

CENTURIES

Section A. England.

WE have first to concern ourselves with the knighthood of
Western Europe and its tactics. Fortress- building
and siegecraft, though equally important in their influence on
the general history of the period, must take the second place.
An English writer is inevitably forced to illustrate the period
mainly from English military history, but we shall conscien
tiously endeavour to point out all the details in which continental
practice differed from that in use in our own island.

The Norman Conquest brought about a complete change in
the military organisation of England : under William the
Bastard the system of raising the armed force of the realm, the
tactics that it employed, and the weapons that it used, were all
alike transformed. For the next two hundred years the
Norman castle and the Norman horseman were to be the main
features in the military history of England.

The kings continued to call out the fyrd on occasion, but
they never treated it as the chief part of their host : it was
indeed mainly employed when the feudal levy of the realm was,
for some reason or another, not to be wholly trusted. William
Rufus summoned the fyrd once for real active service, and once
as a mere means of getting money. It was employed in the
first year of his reign for the sieges of the castles of the barons
who had rebelled against him under the pretence of supporting
his brother Robert. Infantry were always required for siege-
work the knights would have resented the hewing and digging,
and a large force of pioneers was needed. The second occasion
on which we hear of the mustering of the old national host was

357

35 THE ART OF WAR IN THE MIDDLE AGES [1181

when Ralph Flambard taught the king to turn a dishonest
penny by a new device. Rufus called the shire-levies to
Hastings, nominally for a campaign in Normandy (1094). They
came to the number of twenty thousand, each bearing the ten
shillings which the shire was bound to provide for him.
William took the money from them and then told them that
they might disperse, as they were not needed. 1 Henry I. also
used the fyrd early in his reign, in circumstances much like
those which had forced his brother to employ it. Robert of
Belesme and his fellows were in rebellion, and had manned and
stored their castles. Large forces were needed for siegework,
and Henry called upon the English, who came gladly, and, as
Orderic tells us, greeted him, after the surrender of the
enemy s great stronghold at Bridgenorth, with the joyful cry,
" Rejoice now, King Henry, and say that you are truly lord of
England, since you have put down Robert of Belesme and
driven him out of the bounds of your kingdom." 2 Still later the
shire-levies were raised by Stephen for the battle of the
Standard, 3 and by Henry II. to put down the great feudal rising
of 1174. The Assize of Arms of 1181 shows us how miscel
laneous and heterogeneous was their armament : even when
providing for the improvement and reorganisation of the force,
the king does not dream of enforcing uniformity, and the poorer
classes are allowed to come to the muster armed with nothing
better than swords, knives, and darts. There is evidently a wish
to assimilate the wealthier men to the armament of the
mercenary Brabancon pikemen whom Henry was employing in
large numbers at the time, as the sheriffs are directed to see that
persons owning sixteen marks of chattels are to bear mail-shirt,
steel cap, shield, and spear.

But alike for foreign expeditions and domestic wars, the
Norman and Angevin kings relied mainly on the masses of
mailed horsemen provided by their feudal vassals. Still armed,
like their fathers at Hastings, with the long mail-shirt, the peaked
helmet with its nasal, and the kite-shaped Danish shield, the
Norman knights were the flower of the chivalry of Europe,
whether they served in their own land, in the conquered realm
of England, in the new kingdom which they had built up in
Apulia and Sicily, or in the Crusades of the far East.

1 Florence of Worcester, sitl anno 1094. = Ord. Vit. xi. 3,

3 Richard of Hexham, c. 321.

noo] THE "OLD ENFEOFFMENT" 359

William I. had divided up the greater part of the soil of
England among new holders. Only about a fifth stayed with
the old Saxon owners, and such of them as survived were
compelled to surrender their land to the king, and receive it
back from him saddled with the duties of the continental vassal.
We have seen 1 that "knight-service" and "castle-ward" were
ideas not altogether unfamiliar before the Conquest, and that
the obligation of every five hides of land to send a mailed
warrior to the host was generally acknowledged. Theoreti
cally, it would seem, the old notion that the five hides must
provide a fully-armed man was remembered : the man, however,
for the future was to be a horseman instead of a foot-soldier.
But William, in distributing the burdens of military service
among his tenants, seems often to have dealt loosely and
liberally with the old system, frequently letting off his vassals
with less men than their acreage should have called for.
"Beneficial hidation," the counting by favour of four or five
hundred acres as if they were but a mere hundred and
twenty, was as prevalent in military arrangements as in
taxation. It was specially frequent when Church lands were
being dealt with ; e.g. we know that the Abbey of Ramsey had
seventy hides, and should therefore have provided fourteen
knights, but it was let off with an assessment of four only
Nor was this favour confined to ecclesiastical estates alone :
some lay tenants-in-chief got off very easily, though the
majority were obliged to supply their proper contingent.

It has been clearly shown of late, by an eminent inquirer
into early English antiquities, that the hidage of the townships
was very roughly assessed, and that the compilers of Domesday
Book incline towards round numbers. 2 Five-hide, ten-hide,
or twenty-hide townships are so common that there was little
difficulty in apportioning the military service due from the
tenants-in-chief who owned them. Hence there was not so
much difficulty from fractions as might have been expected. If
estates had been assessed with absolute accuracy in acres and
yards, nearly every landholder would have been responsible for
eccentric fractions of a knight, over and above the units which
his manors gave when their extent was divided by the normal
five hides. But estates were not accurately measured and

1 See pp. in, 112.

* See Professor Maitland s Domesday Book, etc. , passim.

360 THE ART OF WAR IN THE MIDDLE AGES [1166

assessed, and so the knights of "the old enfeoffment," 1 as
William s arrangement \vas entitled, are generally found in
round numbers: the fractions which occur are for the most part
quite simple ones.

The landholder, knowing his servitium debitnm according to
the assessment of the veins feoffamentum of the Conqueror, had
to provide the due amount of knights. This he could do in two
ways : he might distribute the bulk of his estate in lots roughly
averaging five hides to sub-tenants, who would discharge the
knight-service for him, or he might keep about him a household
of domestic knights, like the housecarles of old, and maintain
them without giving them land. Some landholders preferred
the former plan, but some adhered, at least for a time, to the
latter. But generally an intermediate arrangement prevailed:
the tenant-in-chief gave out most of his soil to knights whom
he enfeoffed on five-hide patches, but kept the balance in
doininio as his private demesne, contributing to the king for the
ground so retained the personal service of himself, his sons, and
his immediate domestic retainers.

An interesting series of documents, just a century later than
the Conquest, survives, and can be used to show what the
barons had been doing with their land during the three genera
tions which had elapsed since the first assessment. These are
the Cartae Baronum of Il66, 2 a series of answers given by the
tenants-in-chief to Henry II. in response to certain inquiries
which he made from them. The king demanded a statement
as to the number of knights whom each tenant-in-chief owned
as sub- tenants, how many were under the "old enfeoffment"
of William I. and how many of more recent establishment, and
a.lso whether the lord provided his due contingent wholly by
means of sub-tenants, or was accustomed to contribute the
personal service of himself and his household for land held in
demesne. It is interesting to find that the answers show that
the majority of the baronage had given away the larger share
of their estates, but still kept a certain amount in demesne for

1 I think that there is no doubt that Mr. Round in his Feudal England has
proved that we may be reasonably certain that the vetus feoffamentwn really runs
back to the Conqueror, and was a formal distribution. The other view, that it was
irregularly and gradually established under Rufus and Henry I., seems less probable.
On the other hand, Mr. Round s " Constabularies of Knights" are not convincing.

3 The Cartae Baronum are printed in cxtenso in Heame s Liber Ni;er Scactarii.
They are unfortunately incomplete, and do not cover nearly the whole of England.

u66] THE "CARTAE BARONUM " 361

which their own personal service was due. The smaller men,
responsible only for the service of one or two knights, had
not usually enfeoffed sub-tenants, but served themselves. At
first a few great landholders, mostly abbeys, had refrained as
far as possible from cutting up their estates into sub-tenancies,
on account of the financial advantages of keeping land in
demesne. But this plan had the corresponding disadvantage
of compelling the abbot to keep up a household of idle knights,
who drank and roistered about the abbey precincts, and made
themselves an intolerable nuisance. 1 Thus the house was usually
driven, even if unwilling, to give the knights their fiefs in order
to get them away from headquarters. Where, as in the case
of Ramsey, the abbey was very lightly assessed for knight-
service, the proportion of its land which it would have to
distribute to fulfil its servitium debitum would not necessarily
be a large one. But though economy dictated the enfeoffing
of as few knights as possible, nepotism, the curse of the
mediaeval monastery, often drove abbots to give land to their
own needy kinsmen, so that not unfrequently it was found that
a house had created far more sub-tenants than it required. In
such cases the "due service" was sometimes obtained by
making the body of enfeoffed knights undertake to send as
many of themselves to the host as was necessary ; 2 a private
arrangement settled who was to go on each individual expedi
tion.

In the twelfth century the hard-and-fast rule that five hides
ought to make a knight s fee came gradually to be disregarded.
In some cases a liberal lord gave his sub-tenant a good deal more
than the normal holding ; in other cases knights were enfeoffed
on a good deal less occasionally on patches no larger than two
hides. Thus we can find a tenant describing his holding as
" pauperrimum," and grumbling at its counting as a fee at all.
But such cases, in spite of their numbers, were theoretically
abnormal, and the notion which connected five hides with the
knight survived down to the time of Henry n. In the Cartae

1 Liier Eliensis, 2"]$.

3 At Ramsey " Homines faciunt quattuor milites in communi in servitium
domini regis, ita quod tola terra abbatiae communicata est cum iis per hidas ad
praedictum servitium faciendum ; " i.e., though only four knights are required (a very
small contingent from seventy hides), the abbey has not designated four particular
patches to discharge its knight-service, but all the tenants, as well as the abbey
demesne land, club together to " make " four knights for the host.

362 THE ART OF WAR IN THE MIDDLE AGES [1166

Baronum we get a good example of this : Roger de Berkeley
owed two knights and a half on the "old enfeoffment ": giving
more details than his fellows generally supply, he explains as

follows : 1

" The first knight is thus made up

Michael holds one hide

William Fitz-Baldwin, two hides 1 _ ^ hides

Helyas de Boivill, one and a half hides

Hugh de Planta, half a hide

and from these you have an entire knight.
" For making up the half knight-

Ralph de Yweley holds half a hide

The wife of Ralph Cantilene, half a hide

The wife of Richard Gansell, three virgates (J hide)

Roger de Albamara, one virgate (J hide)

Simon de Coverley, one virgate

The Prior of Stanley, one virgate ,,

= two and a
half hides,

and here you have half a knight.

" For making up another knight, Walter de Holecombe,
Gerard, and Reginald de Albamara hold between them ten
hides, but deny their full obligation and say that they do me
service only for one virgate each. From them you can make
up a knight, and so you have two and a half knights enfeoffed."

Roger s argument in the third paragraph is hard to follow:
cither the figures in the text have got corrupted, or he thinks
his disputed claim to ten hides will be compounded for half its
value, and that Walter, Gerard, and Reginald will do one
knight s service between them. However this may be, the
first two paragraphs of his answer amply show that he conceived
five hides to be the proper and normal allowance of land
which should provide a knight. He concludes his "Carta"
with a list of his demesne land, which shows that (unlike most
of his fellows) he had let to sub-tenants only the smaller part
of his ancestral estates.

As a rule, no one except a very great baron with plenty of
house-room in his castle cared to have many domestic knights
dwelling with him throughout the year. Most of the holders of
middle-sized estates had carved the greater portion of them
into knights fees, and only kept in demesne as much as they
themselves and their sons could do service for.

1 Hearne, Liber Niger Scaccarii, p. 165.

n66] THE DANGERS OF SUBINFEUDATION 363

There was always a great deal of trouble in keeping the
sub-tenants up to their work. In times of civil strife, a tenant-
in-chief might rebel, or might remain loyal. If he rebelled,
some of his vassals would try to save themselves from confisca
tion at the king s hands by refusing to join in the rising.
Such indeed was the bounden duty of the English sub-tenant, ever
since the Conqueror at the great moot of Salisbury had impressed
upon the English knighthood the fact that their allegiance was
primarily due to the Crown, and not to their immediate lords.
On the other hand, when the tenant-in-chief adhered to the
king, it was not unusual for some of his knights to slip into the
rebel camp : if the rising succeeded, they would have every
chance of shaking off their lord and freeing themselves for
the future from the service that they owed him. In Stephen s
reign, when anarchy prevailed for well-nigh a score of years,
the relations of countless lords and vassals had been confused :
disputed claims to overlordship were found on every side.
Many of the answers of the barons of 1166 show that they
were not quite certain as to all their own rights and possessions.
They qualify their statements with clauses to the effect that
they have replied to the best of their knowledge and belief, or
note (like Roger of Berkeley quoted above) that some of their
sub-tenants deny their obligations. The clerical tenants are
specially bitter against spoilers who have robbed them of
homage, or compelled them to enfeoff knights contrary to their
will. We are surprised to find such a respectable person as the
great Chancellor Roger of Salisbury reported as an oppressor
by the Abbey of Abbotsbury in Dorset. 1

The importance of King Henry s inquest of 1166 was
twofold. It not only gave him the information that he
required as to the proper maintenance of the debitum
servithim due under the "old enfeoffment " of the Conqueror,
but showed him how many more knights had been planted out
since that assessment. Having possession of this valuable
information, he was able to demand for the future, when raising
aids and scutages from his tenants-in-chief, payment not
merely for the theoretical number of knights whom they owed,
but for the real number which they actually possessed. This

1 Liber Niger, p. 76 : "Cum Rogerus episcopus habuit custodiam abbatiae, duas
hidas apud Atrum, ad maritandam quandam neptem suam dedit Nicolao de Meriet,
contradicente conventu."

364 THE ART OF WAR IN THE MIDDLE AGES [1166

gave a welcome relief to the Treasury, as in many instances the
"old enfeoffment" had been as \ve have already mentioned -
very lax and liberal, and did not adequately represent the re
sources of the land.

The Carlae Baronum are unfortunately incomplete: if they
had all been preserved, \ve should have been able to say both
what was the number of knights due from the whole of England
under the " old enfeoffment " of the eleventh century, and
what was the number of knights fees actually existing in 1166.
A careful and ingenious calculation has been worked out by
supplementing the Cartae from other sources, which makes it
clear that the full feudal force of England was well over four
thousand five hundred knights, but little, if at all, over five
thousand. Of these the Church fiefs supplied about eight hundred,
the lay tenants-in-chief between four thousand and four thousand
two hundred. 1 These modest figures contrast most strangely
with the vague numbers given by contemporary chroniclers,
who were so far from appreciating the actual size and resources
of the land that they often state that England could supply
thirty thousand or even sixty thousand 2 knights for the king s
service. The whole fyrd of foot-soldiery added to the knight
hood would probably not have reached the latter figure.

We must be careful, when dealing with the knight of the
eleventh and twelfth centuries, to clear away from our minds the
chivalrous connotation of the same word in the fourteenth or
fifteenth century. The knight of William the Conqueror s army
was not necessarily nobly born, nor had he gone through the
elaborate ceremonial of admission to the knightly order which
prevailed three centuries later. He was simply a soldier who
fought on horseback, and who received from the king, or from
one of the king s tenants-in-chief, a patch of land on condition that
he should do mounted service in return for it. The original
knights of the "old enfeoffment" were a mixed multitude of
many races drawn from many different stations in life : some
were the kinsmen of great Norman barons, others were military

1 Mr. Round s calculations on this point in his Feudal England, pp. 289-293, are
most valuable and convincing. The result is certainly surprising, and shows most
clearly the extraordinary want of appreciation of large figures in the thirteenth-
century chroniclers, and even in Government officials who ought to have known
better.

2 Swereford in the Liber Rubeus says thirty-two thousand ; Ordericus Vitalis is
responsible for the still more monstrous sixty thousand.

1 1 66] THE EARLY KNIGHTS 365

adventurers who had drifted in from all parts of the Continent.
Into this heterogeneous body were incorporated the remains
of the old English thegnhood, all the lucky survivors who had
been permitted to "buy back their land" from the king by
paying him a fine and doing him homage on feudal conditions
after his coronation. English-speaking men applied to this
newly-formed and miscellaneous class of military tenants and
sub-tenants the word " cniht," which had been used before the
Conquest for the military dependants of the great landholders. 1
It was really equivalent to the cliens, serviens, or famulus of the
Continent, and has the same original meaning of subordination
and subservience. But names chance on different histories in
different countries ; and while " knight" became in England the
equivalent of miles, the name serviens came across the Channel
some generations later, in the form of " sergeant," to express a
class of men distinctly below the knightly rank. It is curious
to note that in Germany knecJit, starting with much the same
meaning as " knight " in the eleventh century, gradually came
to denote persons of a more and more inferior status, sinking
to mean combatants who were not of noble blood, 2 and finally
denoting mere servants and attendants of the army.

It will help us to realise the modest status of many of these
"knights "of the Norman period, if we remember that a sub-tenant
with a fewhundred acres of land would probably have been called by
a chronicler of the time of Henry I. a " miles," by a chronicler of
the time of John or Henry ill. a " sergeant," 3 and by a chronicler

For a picture of pre-Conquest "knights" in England, see the interesting
description of the rights and duties of the "radknights" of Bishop Oswald of
Worcester, which Professor Maitland has worked out in his Domesday Book and
Beyond," pp. 305-311.

3 The word Edetknecht was invented to denote the non-knightly combatants of
good birth ( = English esquire), and then knecht without the prefix came to distinctly
imply want of birth.

3 That "sergeant" originally means not a professional soldier, nor a knight s
attendant, but a landed military dependant \vho is not a knight, is well shown by
the letter of Geoffrey Ferland, Sheriff of Leicester and Rutland in 1216, giving
" the names of all the knights anJ sergeants domiciled in his district who have
adhered to Louis of France " (Rymer, 144), Another good example is John s writ
of 1213, to call out the full feudal levy : " Rex vicecomiti tie X. salutem, etc.
Summonecomitesbarones milites et omnesliberos homines et servientes, dequocumque
teneant,utsintapud Doveram cum armiset equis," etc. (Rymer, I. no). The "armis
et equis" clause shows that we are dealing with mounted men, and the "de quo-
cumque teneant " that we are dealing with sub-tenants and not merely small tenrnts-
in-chief.

366 THE ART OF WAR IN THE MIDDLE AGES [1150

of the time of Edward ill. a " squire " (armigcr or scutifer). The
condition of the three men would have been much the same, but
the name changed thrice. By 1350 the title of knight had come
to be restricted to persons of some importance, and we often find
large bodies of men commanded by mere esquires in the wars of
Edward III. The reigns in which the change first made itself
felt were those of Henry III. and Edward I., whose repeated
attempts to make holders of knightly fees take up the knightly
title by the writs of " distraint " are well known. 1 But the
attempt did not succeed, and ere long we find the king conceding
that even the parliamentary knights of the shire may be persons
who have not actually received knighthood, because that in
many counties there cannot be found sufficient competent
persons who have taken up the required status.

Before proceeding to investigate the character of the battles
of the twelfth and early thirteenth centuries, we must take
note of one marked feature of the early Plantagenet reigns
the prominent place taken in their military affairs by mercenary
troops. From the time of Stephen onward, we perpetually find
the feudal levies of the realm supplemented by great bodies of
professional soldiers, nearly all foreigners. There had for a
long time existed a large floating body of adventurers in
Western Europe : from them William the Conqueror had drawn
no small proportion of the host that fought at Hastings. The
original Norman conquerors of Apulia had belonged to this
class no less than the Varangian Guards of the Eastern emperors.
During the early Norman reigns we not unfrequently find
mention of stipendiarii militcs in England, 2 but it is not till
the time of Stephen that we begin to find them appearing in
great force and forming a prominent feature in the host.
Stephen, deserted by the greater part of the baronage, supplied
the place of the missing contingents by bringing over great
bodies of Flemings and Braban$ons, under leaders such as
William of Ypres and Alan of Dinan. Henry II. and Richard I.
kept up the system : without the aid of a permanent army
they could not have maintained their long wars over sea. For
sieges in Normandy and Aquitaine the service of the English
feudal levy would have been almost useless to them. Its
forty days would have ended almost before it could arrive at

1 Especially in xix. Henry in. and in vi. Edward I.

2 See Florence of Worcester, sub anno 1085.

ii73] THE ORIGIN OF SCUTAGE 367

the distant seat of war. Moreover, a feudal host was untrained,
undisciplined, disorderly, and sometimes disloyal. The mer
cenaries, on the other hand, were trained professional soldiers,
who served with fidelity as long as they were regularly paid,
and had no wish to cut the war short by an intempestive
return to their homes. Hence for foreign service Henry
and Richard preferred the steady squadrons of mercenaries
who kept the field all the year round, to the short and uncertain
aid of the knighthood of England. To repel a Scottish foray
or to carry out an expedition into Wales, on the other hand,
the servitiurn debituin of the English tenant-in-chief was still
exacted. Such campaigns were short, and cost less if carried
out by the levies of the border shires. Henry II., therefore, very
seldom brought over his mercenary bands to England : the
only occasion when they appeared in force on this side of
the Channel was to aid in suppressing the feudal rebellion of
U73-74- 1 this campaign they met their likes in battle, for
the rebel Earl of Leicester had enlisted a great body of Flemish
rentiers, and was righting at their head when he was taken
prisoner at Fornham.

When the king did not wish to call out the feudal levy of
England, he was accustomed to exact from all the exempted
knights a scutage. By this arrangement the holder of a fief
compounded for his personal service by paying a fixed sum for
every shield (scutum) that he should have brought to the host.
The usual sum raised was 26s. Sd. two marks which seems
to represent forty days service at Sd. a day, the normal pay of
a knight in the twelfth century. The individuals from whom
the servitiuni debitum was due seem to have been allowed the
choice of attending in person or paying the scutage. 1 If the
campaign was near at hand, the majority would appear in arms ;
if it was distant, only a few mainly the larger tenants would
follow the host.

1 The whole body of feudal tenants do not seem to have been so prone to accept
the alternative of composition as might be inferred from the chroniclers. For example,
as Mr. Round has shown, Robert cle Monte tells us that in 1159 King Henry took
with him "capitales barones suos cum paucis, solidarios vero milites innumeros" ; but the
scutage figures show that the sum received was ,1714, i.e. the money representing 1280
knights, not more than a third of the number liable to serve from the lay fiefs, so that
not only the great barons must have followed the king, but some two-thirds of the smaller
men also (Feudal England, p. 280). The reason advanced for the king s preference
of a scutage is obviously not the right one. In reality he wanted the money to pay
mercenaries.

368 THE ART OF WAR IN THE MIDDLE AGES [1215

Scutage appears as a recognised institution under Henry I., 1
but it was his greater grandson who made it normal and
customary. By the end of his reign the bulk of the rural
knights had grown into the habit of compounding instead of
going on wearisome expeditions to Poitou or Aquitaine, over
the stormy seas so hateful to the mediaeval mind. The pay
ment of scutage became the rule, and the hiring of mercenary
horsemen with the proceeds of this imposition gave the
king a more permanent and trustworthy army than he could
otherwise have kept together. It was mainly at the head of
these professional soldiers that Henry II. and Richard Coeur
de Lion fought out their weary and uninteresting French
campaigns.

John, because he was more hated by his subjects than his
father and brother had been, was still more prone than they
to employ mercenary troops. No small part of his unpopularity
in England came from the fact that after he had been driven out

of Normandy in 1204 he brought back with him the horde of
foreign adventurers who had followed his unlucky standard
on the Continent. They were, as] might have been expected,
very undesirable guests : the barons resented the favour which
the king showed to the leaders unscrupulous ruffians, for the
most part, like Fawkes de Breaute. The common people
suffered from the plundering propensities which the mercenaries
had picked up on the Continent. To the hatred they won from
rich and poor alike, the adventurers owe their dishonourable
mention in the Great Charter. The king is forced to promise
to dismiss all the " alienigenos milites et balistarios et servientes
stipendiarios " who " venerunt cum armis et equis ad nocumentum
regni." 2 A special clause names several of the leaders who were
condemned to banishment Gerard of Athies, Philip of the
Mark, Englehard de Cigognes, Guy de Cancelles, and others. 3
As everyone knows, John slipped easily out of the obligation
the mercenaries were not expelled, and formed the best
part of the army with which the king fought his unfortunate
campaign of 1215. The troopers of Fawkes de Breaute", and
also his crossbowmen, are specially mentioned as having done
good service, early in the reign of Henry in., at the second
battle of Lincoln. It is not till the reign of Edward i. that

1 See the proofs in Mr. Round s Knight-service, pp. 268, 269.

2 Magna Carta, clause 51. 3 Ibid. 50.

noo] CONTINENTAL ARMIES 369

foreign mercenaries cease to form a prominent part in the armies
of the Plantagenets.

Section B. The Continent.

During the twelfth and thirteenth centuries the armies of
the English kings differed less from those of the sovereigns of
the Continent than at any other period in history. The Norman
influence had assimilated the military forces of our island to
those of the rest of Western Europe. The chief points of
difference worth noticing are, firstly, that in England there was
never such a clear line of division between the various classes
of feudal tenants as elsewhere ; and, secondly, that shire-levies
of foot-soldiery, the lineal descendants of the fyrd, though
occupying a very secondary place in war, are yet much more im
portant than the infantry of most continental districts. Only in
the Netherlands and to a certain extent in Italy do foot-soldiery
come prominently to the front. In other regions the mercenary
crossbowmen are the only dismounted men who receive much
mention, till we come to the attempt of Philip Augustus to
turn the levies of the French communes to account.

The normal army of an emperor or a French king was com
posed of the same elements as those with which our Norman or
Angevin monarchs took the field a mass of mounted feudal
tenants and sub-tenants, often supplemented by a certain propor
tion of mercenary horsemen and crossbowmen. Occasionally we
find civic militia in the field it develops in Italy and the Low
Countries long before it is found elsewhere. Very rare is the
appearance for any practical purpose of the foot-levies of the
countryside, which the feudal lords could as a last extremity
drag out to battle.

In the eleventh century the important part of a continental
army consists of all the warriors holding fiefs, either directly
from the Crown or as sub-tenants, on condition of doing service
on horseback. The chroniclers often speak of the whole mass
of them as " inilites," whether they be small men or great, but a
careful inquiry into the character of the body shows that it is
not homogeneous. \Vhen we find phrases like " miles privii
ordinis" or "miles gregarius" \ve see that within the body
of inilites there are class distinctions. The highest rank is
composed of free vassals of noble blood holding considerable
fiefs : this is the only class which retains the knightly style in
-4

370 THE ART OF WAR IN THE MIDDLE AGES [1150

subsequent ages, but the name miles in 1080 (abroad as in
England x ) is far more vague, and covers far more persons than
it does in 1180 or 1280.

Below these milites primi ordinis are a number of other
horsemen, some of noble but more of non-noble blood. Some
are the king s personal retainers, serving him as minor officials or
guardsmen : a twelfth-century German chronicler would probably
call them " minister iales" an English or a French chronicler
" scK icntes regis." Much more numerous are the personal
retainers of the barons, bishops, and abbots, whether enfeoffed
or not enfeoffed on land. These " men " of the king or of
the tenants -in -chief are sometimes styled milites grcgani,
milites ignobiles, milites plebei, or milites mediocris nobititatis.
They are also found with names which differentiate them more
clearly from the knights of higher rank,, and point to their sub
servient and dependent condition e.g. satellites^ seri icntes,
dientes, famuli. As a rule, they served on lighter horses, and
wore less complete armour than the knightly vassals. Down to
the thirteenth century the} much exceeded in numbers the
nobler and more heavily-armed horsemen. 2

When in the later twelfth century the title miles becomes
strictly confined to the upper ranks of the military class, seri iens
(sergeant) is the most usual term for the horsemen of lower
status. In France it grew to be the only recognised name for
them. In Germany it was not so common, sariant (the German
form of the word) being used indifferently along with other
appellations, such as scutifer, armiger, strator. These twelfth
and thirteenth-century servicntcs or sciitifcri are not to be confused
with the squires of the fourteenth and fifteenth centuries who
were the personal attendants of a knight. In the earlier age the
knight had no mounted follower. His armour-bearer accom
panied him on foot, and was not necessarily a combatant at all.
The " sergeants " were often formed into separate corps, apart
from the knights, and used for the purposes for which light
cavalry are required ; or, again, they were placed in the less
important parts of the battle-array. Not unfrequently we find
sergeants placed in the front line to open the combat, while the
knighthood is held in reserve to deal the decisive blow. We

1 See p. 440.

- e.g. we shall see that at Legnano the emperor s host comprised five hundred
knights to fifteen hundred sergeants. See p. 442.

i2oo] KNIGHT AND SERGEANT 371

shall find Philip Augustus employing this arrangement in his
right wing at Bouvines (I2I4). 1 Frederic II. did the same at
Cortenuova in 1237. But it \vas by no means the regular rule
to separate the lighter and the heavier horsemen. It was more
common to compose each of the divisions of an army of
sergeants, " stiffened " by the admixture of a certain proportion
of knights, as did, e.g., the elder Montfort at Muret (I2I3). 2

A further complication is introduced into the nomenclature
of the military class when, in the twelfth century, the word
miles has its meaning still further changed by the spread of
the new idea of chivalry. When the notion is introduced that a
knight must be solemnly invested with the arms and insignia of
the knightly rank by his feudal superior or some other personage
of importance, and must not call himself miles till he has
been so honoured, there necessarily comes into existence a class
of holders of knightly fiefs who have not yet received the
knightly name. A young baron with very large estates may
serve for some time before earning the title. On the other
hand, a warrior of approved courage, whether of noble or non-
noble blood, may receive knighthood from king or duke for
some notable feat of arms. Thus a baron not yet knighted
was often followed to war by vassals who had attained the rank
to which he was still aspiring.

Hence, in the later twelfth or in the thirteenth century, when
\veexamine the composition of that part of the personnel of a
feudal host which does not consist of knights, we find quite a
large variety of classes represented in it. We may notice i)
young holders of knightly fiefs who have not yet received the
knightly title ; (2) men of knightly blood, holding small fiefs,
who, on account of poverty (or some such other reason) do not
intend to take up the honour ; (3) younger sons of barons and
knights, who have no land and therefore cannot afford to aspire to
knighthood (this was a class out of which the mercenary cavalry
were very largely recruited) ; (4) various degrees of persons of
non-knightly blood enfeoffed on land by their lords. The first
three sections are men of the knightly class, but not knights :
the last is the one to which the title of sergeant properly belongs.
A cross-division is made by the fact that a wealthy sergeant
may sometimes succeed in providing himself with a heavy war-
horse and the full panoply of mail, while poor members of

1 See p. 471. 2 See p. 453-

372 THE ART OF WAR IN THE MIDDLE AGES [1250

classes 2 and 3 may be serving in incomplete armour and on
inferior chargers.

In the later thirteenth century we find the three latter classes
tending to melt together, and to be considered as all equally
forming part of the military aristocracy, so that most of the
sergeants ultimately became " noble." Though not knights, they
form the lower ranks of the knightly caste. It is easy to under
stand that when the knightly title became restricted to a com
paratively few individuals of the knightly houses, and when the
poorer members of them were continually serving along with
the richer sergeants, the latter should ascend a step on the
social ladder. It was more natural that the sergeants should
advance to a better status, than that the brothers and younger
sons of the holders of knightly fiefs should descend to a lower
one. So by the fourteenth century the French noblesse and
the German Add have extended their ranks so as to include
classes which two hundred years earlier would not have been con
sidered to belong to the nobly-born. The term sergeant passes
out of use as meaning a feudal horseman of the lower rank, 1 and
armies are reckoned not as containing milites and servientes,
but by the number of "helms" or " barded horses " that they
muster. No one now stops to inquire whether the warrior who
wears the full panoply and rides a heavy charger has or has not
received the knightly spurs and girdle. He is an equally
efficient member of the host, whether he bears the knightly
name or not. The general body of the feudal horsemen who
have not won their spurs are now called squires (/cuyers,knechte^
aniiigeri], or men-at-arms.

It is, of course, impossible for an army to dispense altogether
with light cavalry ; they are needed for purposes of foraging and
reconnoitring. In this capacity the place once held by the
servientes is occupied in the fourteenth century mainly by
mercenaries, but partly also by the incompletely armed servants
of the knights and squires, who brought with them to the host
a certain number of mounted attendants (valets amids, Diener).
There were, however, to be found light horse who were neither
mercenaries nor mere dependants of the men-at-arms. Such
troops certainly existed in England ; we recognise them in the

1 Remaining in use, however, as we shall see later on, for certain individuals, e.g.
the king s personal retinue of " sergeants-at-arms," employed by him for various
small official duties. It also survives in occasional use for foot-soldiery.

1250] CONTINENTAL MERCENARIES 373

pauncenars and hobilars of the Calais muster-roll of Edward
III. (I347). 1 On the Continent, too, they appear as panzer ati
or rentier in Germany, as Jianbergeons in France.

No account of the armies of the twelfth and thirteenth
centuries would be complete without mention of the mercenary
cavalry. We have already seen that in England they occupy a
very prominent place in military history, and the same is the
case on the Continent. From the days of the Norman adven
turers who ousted their unfortunate employers from Apulia and
Benevento, the mercenary is always intermittently in evidence.
Robert Guiscard and William the Conqueror were able to
recruit them by the thousand, and in most continental wars we
find them serving side by side with the emperor s or king s
liege vassals. Their bands would include a much smaller pro
portion of knights and a much larger proportion of combatants
of lesser status than did the normal feudal host. The knights
who left their fiefs to follow the career of adventure were
naturally not so numerous as the smaller men. The bulk of a
mercenary band would be composed of the landless younger
sons of sub-tenants, mixed with adventurers of lower birth who
had taken to the profession of arms from love of fighting or
from the wish to escape from villeinage. Whatever the origin
of these mercenary horsemen, all who were not knights were
commonly known as " sergeants," the escaped villein no less than
his better-born companion. At first it was more common to
buy the service of mercenaries by the gift of land, but by the
twelfth century there was enough money in circulation to enable
kings and emperors to retain the hired horsemen in service by
the regular payment of a daily, monthly, or yearly salary.
This was in every way better for the employer : the enfeoffed
mercenary was generally a bad and turbulent subject (we need
only recall to the English reader such instances as Fawkes de
Bre"aute), while the adventurer hired for a fixed term could be
duly discharged when he was no longer needed.

The mercenary bands were increasing in importance all
through the period with which we are now dealing. Only local
wars could be conducted by the regular feudal levy ; all long
and distant campaigns and all large schemes of conquest
required the co-operation of hired soldiery. Kings with a wide
and scattered empire, like Henry II. of England, were necessarily

1 See p. 366. Brady, vol. iii., Appendix.

374 THE ART OF WAR IN THE MIDDLE AGES [1308

driven to employ them. Adventurers in search of a realm, like
Charles of Anjou in the succeeding century, naturally relied
upon them. Long-continued wars hardened them into compact
masses, till by the end of the thirteenth century we find the
condottiere system coming into existence noted mercenary
chiefs have collected huge bodies of men numbered by the
thousand, and hawk their services about from court to court.
The first 1 of these hosts of free-companions which comes into
prominence is the " great company " of Roger de Flor, formed
from the discharged mercenary bands of the King of Aragon,
turned loose when Peter ended his long struggle for Sicily with
Charles II. of Anjou. Roger s horde was strong enough to
shake the whole Levant, to bring the Byzantine Emperor
Andronicus to his knees (1308), and to carve out for itself a new
home in the duchy of Athens.

Turning to the continental foot-soldiery, we find that we
need not in the twelfth century concern ourselves greatly with
France or Germany ; the Netherlands and Italy are the two dis
tricts which demand our attention. Closely akin to the English,
the inhabitants of Flanders, Brabant, and the neighbouring regions
had, like their kinsmen on this side of the water, taken late to
horsemanship. Unlike England, the Netherlands had never
been conquered and divided up by any invader, and it seems
likely that their steady infantry descends directly and without
a break from the times of the Carolingians. The growth of an
indigenous feudal cavalry in the duchies and counties of the
Low Countries did not entirely extinguish the foot-soldiery, as
was the case in most other regions. As early as 1 100 we have
notices of Netherlandish infantry armed with the pike which
enjoyed a reputation far above that of the foot-levies of other
countries. 2 In the earliest cases they are called guidons the
same word, it will be remembered, which Wace uses for the
English axemen at Hastings. 3 We may guess that the mailed

1 We can perhaps hardly count Stephen s Flemish captain William of Ypres or
Richard- Cceur de Lion s follower Mercadier as real condottieri, as it does not seem
that they hawked about already formed bands for service, but rather that they gathered
and kept together new corps at the king s expense.

- In I 106 the Annals of Hildesheim, 3. 1 10, mention that the Duke of Brabant sent
to aid the Archbishop of Cologne " quoddam genus hominum qui vocantur Gelduni,
viri bellatores et strenui, et nimis docti ad praelia."

3 Wace, 12927 :

" Geldons Engleiz baches portoient
E gisarmes ki bien trancheoient. "

r[2i4] THE INFANTRY OF THE NETHERLANDS 375

mercenary infantry armed with the pike which the Conqueror
employed in that same fight were largely Flemings.

Later in the twelfth century we find these pikemen serving
in all the wars of the Low Countries along with the feudal
cavalry of their lords, and, ere long, pushing abroad as mer
cenaries. They generally appear under the name of Brabanons,
which becomes a technical term for mailed mercenary foot-
soldiery : English and French kings and Roman emperors are
all found employing them ; they appear in the Italian wars of
Frederic Barbarossa, the French expeditions of Henry Planta-
genet, and the victorious campaigns of Philip Augustus. The
last fight in which we note them taking a prominent part is
Bouvines, where a small body of them 1 in the service of the
Count of Boulogne did far the best service performed by any
foot-soldiery in the allied army. In the thirteenth century the
Flemings and Brabangons do not keep their place as mercenaries,
the crossbowman, rather than the pikeman, is the typical hired
foot-soldier of that age ; but in their native land they continue
to serve as before, and the mailed militia of pikemen is still
reckoned a notable part of the host. We may see their usual
tactics at Steppes (1212;,- and read of their greatest triumph at
Courtray (1302). In the thirteenth and fourteenth centuries,
the civic levies of the Flemish cities are the most prominent
exponents of such methods of combat.

The Netherlandish infantry had little mobility or initiative.
They fought in heavy masses, and could not manoeuvre. But for
purely defensive tactics they were formidable : the weapons of the
pikemen were much longer than the knightly lance, and if onlv
the mass held firm it was extremely difficult to break into it.
But since it could not easily advance or change its front, it
could not unaided win a battle: at the most it could only
repulse its enemy. To be actively successful it must be helped
by mounted men : when the pikes have checked the foe, the
onset of horsemen is required to break him and pursue him.
For use in combination with cavalry the pikeman is inferior to
the man armed with missile weapons : he can only harm his
adversary at the moment of contact, while the archer or cross-
bowman can keep up a continuous discharge as long as the

1 " Homines de Braibanto, pedites quidem, sed in scientia et virtute bellandi
equitibusnon inferiores" (Gen. Com. Fland. in Bouquet, xvii. 567 c.).

2 See p. 444.

376 THE ART OF WAR IN THE MIDDLE AGES [1200

enemy is within a hundred or a hundred and fifty yards of
him.

Roughly speaking, we may say that these early pikemen
could give valuable assistance in winning a battle, but could
not gain it by themselves. They could supply a rallying-point
for the cavalry, or bear the brunt of the fight while the latter were
re-forming ; they could oppose a long passive resistance, but
had little or no active power. If we ever find them taking the
main part in a victory, peculiar local circumstances must be the
explanation ; e.g., at Courtray the fearful slaughter of the French
chivalry was caused by the fact that they fought with a deep
marshy ditch in their immediate rear, so that they could not
easily retreat. Usually attempts of the Netherlander to fight
without the aid of horsemen only brought them disasters like
Cassel and Roosebeke.

In Italy, where foot-soldiery had never been prominent since
the old Roman days, their reappearance is intimately connected
with the rise of the great towns. Just before the age of the
Crusades, the cities of Northern Italy were beginning to start
on their career of municipal independence, and had practically
freed themselves from their counts and bishops. We have
already noted the vigour with which they flung themselves,
first into the struggle to expel the Moorish pirates from the
central Mediterranean, and then into the more distant Crusades
of the Levant. 1 Seafarers like the Venetians, Genoese, and
Pisans naturally developed into foot-soldiery. It is as cross-
bowmen that they appear at every siege and battle in Syria
during the twelfth century. Of all the peoples of Europe, none
had such skill in the use of the arbalest : after winning a high
reputation as marksmen in the battles of the East, we find these
Italian foot-soldiers, and especially the Genoese, passing north of
the Alps as mercenaries, and fighting in the French service at
Courtray, Sluys, and Crecy.

While the inhabitants of the seafaring towns were mainly
skilled in the use of the crossbow, the civic militia of the inland
cities was chiefly composed of pikemen. The army of an
important municipality like Milan or Verona consisted of a
mass of infantry, backed by a certain proportion of horse. For
the Lombard states owned a not inconsiderable amount of
cavalry, provided partly by the nobles of the countryside, who

1 See pp. 252, 253.

1 200] ITALIAN FOOT-SOLDIERY 377

had been more or less willingly incorporated in the civic body,
partly by the richer burgesses of the city, the local patrician
families. Every town of importance could put in the field some
hundreds of mailed horsemen, while Milan mustered more than
two thousand. But the bulk of the hosts of the Italian munici
palities consisted of the infantry serving under the banners of
their quarters or parishes. (At Milan the division of the city
was into " gates.") They were well equipped with pike, steel cap,
and mail-shirt, and, when properly led, showed great solidity in
the field.

The Italian infantry never attempted, as did the Flemish
more than once, to dispense with the assistance of cavalry. They
always worked in company with the horsemen of their cities, and
made no pretensions to be self-sufficient. When pitted against
an enemy who used mounted men alone, or only brought
inefficient and ill-armed foot-soldiery to the field, they often
turned the scale in favour of their own side. As a typical fight
of this description, we shall narrate the battle of Legnano, 1
where the steadiness of the Milanese foot saved the day, by
allowing the routed Lombard horse time to rally and resume
the charge.

1 See p. 442.

CHAPTER III

ENGLISH BATTLES AND THEIR TACTICS, IIOO-I2OO.

TencJiebrai (\ 106) Eremftle (i 1 19) Northallerton (i 138)
Lincoln (1141) Battles in Ireland ( 1 1 69-7 1 ).

IT has been often observed that the period of the completcst
supremacy of cavalry in the West, the twelfth century, was
not a period of great battles. There are more important fights in
England in the open field during the sixteen years of the Wars
of the Roses, or the six years of the Great Rebellion, than in
the whole century between iioo and 1200. The same is the
case on the Continent, though in not quite such a notice
able degree. The main reason of this was, that the develop
ment of fortification during the century was so enormous, that it
was more profitable for the weaker side to take the defensive
behind strong walls than to fight in the open. Hence the
century is pre-eminently one of sieges rather than of pitched
battles. Henry i. s victories of Tenchebrai and Bremule were
very small affairs, in which only a few hundred knights took
part. The long civil wars of Stephen and Matilda abound with
sieges, but only supply the two battles of Northallerton and
Lincoln. All the long French wars of Henry II. do not give us
a single first-rate engagement in the open ; the skirmish of
Fornham and the surprise of Alnwick are the only fights in his
reign that we need notice. The same is the case with the long
bickering of Richard I. and Philip Augustus along the Norman
and Poitevin borders. It is hardly too much to say that between
Lincoln (1141) and Bouvines (1214) no English troops were
present at an engagement of first-rate importance in Western
Europe. If it had not been for the distant crusading battle of
Arsouf (1191), we might have said that there was no really great
battle in the whole period in which they were engaged.

378

nc6] BATTLE OF TENCHEBRAI 379

For the most part, these unimportant conflicts of the
twelfth century were both simple and short. Another
notable point about them was, that they were accompanied
by very little effusion of blood, save when some luckless
infantry had been dragged into the field by one side or the
other : in that case there was often cruel butchery in the
pursuit ; otherwise the knights gave each other quarter, and
the main loss of the defeated side consisted of prisoners and
not of slain.

Battle of Tenchebrai, September 28, 1 1 06.

Henry I. of England had invaded the lands of his brother
Robert and overrun most of the duchy of Normandy. He was
beleaguering Tenchebrai, a castle of the Count of Mortain, when
the duke resolved to make a desperate attempt to raise the
siege. Gathering all the forces that he could muster, he
marched on Henry s camp and offered battle ; he was very
inferior in the number of his knights, but had brought a mass
of ill-armed peasantry and citizens with him. Possibly his
experience in the Crusades had given him the idea that the
knight and foot-soldier should be combined in the line of battle ;
but he evidently did not know how to turn his notion to profit
able account. Finding himself outnumbered and outflanked, he
dismounted his knights and put them at the head of the unsteady
infantry. The army formed three corps ; the right was led by
William of Mortain, the centre by the duke, the left wing by
Robert of Belesme, the rebel whom Henry had expelled from
England six years before.

The king s army consisted wholly of mounted feudal levies ;
but, seeing that his brother had ordered his knights to fight on
foot, Henry also bade a great portion of his host to send away
their horses, in order that he might oppose a mass of equal
solidity to the duke s columns. 1 The whole of the English and
Normans were dismounted and formed into three corps, placed
under Ralph of Bayeux, Robert of Mellent, and William of
Warenne. The first -named faced William of Mortain, the
second the duke, the third Robert of Belesme. But Henry
commanded his vassals from Maine, under their count, Helie of
la Fleche, and his auxiliaries from Brittany, to keep their horses

1 "Rex namque et dux et acies caeterae pedites erant ut constantius pugnarent "
(Henry of Huntingdon, 235).

380 THE ART OF WAR IN THE MIDDLE AGES [1106

and to ride off and to take position on his right wing, at some
distance from the main body.

The battles of the king and the duke clashed together with
equal courage, and stood locked for a short time in close conflict.
Then William of Mortain drove back Ralph of Bayeux and
Henry s left wing for some space, 1 while the centre and right of
the king s army held their ground. But immediately after,
He lie of Maine led his horsemen against the flank and rear of
the Norman left wing. At the first shock Robert of Belesme s
corps broke up, then that of the duke, then that of Count
William. The horsemen rode in among the fugitives and cut
down two or three hundred of the unmailed Norman infantry.
But the knights were mostly admitted to quarter: only a few
escaped, 2 the rest, four hundred in all, were taken prisoners.
Waldric, one of Henry s chaplains, was the captor of Duke
Robert, for which unclerical feat he was soon after made bishop of
Llandaff. With Robert were taken William of Mortain, Robert
d Estouteville, William de Ferrers, William Crispin, and all the
chief nobles of Normandy. We are somewhat surprised to find
in their company Eadgar the Atheling, who had broken his old
friendship with Duke Robert some time before, but had returned
to his side to share his day of misfortune. 3 Robert of Belesme,
who fled too early for his own good fame, 4 was the only man of
note in the duke s army who got away.

The whole fight had not occupied an hour, and not a single
knight on Henry s side had been slain. We have to turn to
Italian chronicles of the fifteenth century to find such a blood
less fight followed by such great results for the victory of
Tenchebrai gave King Henry the whole duchy of Normandy.
He had used horse and foot combined, against isolated infantry,
and had been properly rewarded for his adherence to his father s
example at Hastings. 5 It is curious to see that it was the

1 " Consul \ViIIeImus aciem Anglorum de loco in locum turbans promovit " (tbiJ.).

2 William of Jumieges, p. 573.

3 Anglo-Saxon Chronicle, sub anno 1 106. The king shortly released him, though
he condemned the others to perpetual bonds.

4 Orderic Vitalis, 701.

5 Matthew Paris (writing a hundred and fifty years after the fight) thinks that
Henry s " English and Normans on foot" are a different body from the three corps
under Ralph of Bayeux, Robert of Mellent, and William de \Varenne. This is an
error, produced by misunderstanding Orderic s "Primam aciem, etc. . . . Rex autem
Normannos et Anglos pedites secum detinuit, Cenomannos et Britones longius in campo
posuit." The three corps are the pedites.

it 19] LOUIS VII. INVADES NORMANDY 381

brother who had stayed at home, and not the brother who had
been to the far East, that had best realised the military meaning
of the experience of the first Crusade.

Battle of Breinide (Brenville) August 20, 1119.

King Henry s second battle in Normandy was an even
shorter and simpler affair than the battle of Tenchebrai ; it
hardly deserves, indeed, to be called anything more than a
skirmish. It only lasted a few minutes, and the total number
of men engaged on both sides was less than a thousand in all.

Louis VI. of France had invaded Normandy, to endeavour to
place on its throne his young protege, William Clito, the son of
Robert, who had now been languishing for many years in Cardiff
Castle, and was well-nigh forgotten. William, a clever and bright
lad of eighteen, was now old enough to take the field in person
along with his champion. They had crossed the frontier, and a
few trusty old adherents of Robert had joined their standard,
but the great bulk of the barons of the duchy stood firm in
their allegiance to King Henry.

Since castles and cities kept their gates closed, the invasion
dwindled down into a series of mere plundering raids. Based
on the town of Les Andelys, Louis and his knights rode out,
harrying the countryside, and pushing useless forays as far as
the neighbourhood of Rouen and Evreux. Meanwhile, King
Henry came upon the scene with a small army : he had a few
English with him, but the bulk of his force was composed of
the native feudal levies of Normandy ; he took post at Noyon-
sur-Andclle, intending to cover the duchy from the destructive
inroads of the French. On the 2oth of August, the smoke rising
from the burning barns of the monks of Bucheron 1 showed
clearly to Henry that the French were out upon one of their
habitual forays. He marched for the scene of destruction, with
the five hundred knights who were around him, and soon came
into sight of the scattered outriders of Louis. When the French
king heard that his enemy was at hand, he swerved aside to
meet him, in spite of the advice of his wiser counsellors, who
pointed out to him that he had but four hundred horsemen
with him, and that the force of the Normans was unknown.
Without taking any military precautions, or even drawing up a
definite plan of battle, Louis galloped off to attack Henry.

1 The name of the place on which the abbey of Noyon-sur-Andelle was built.

382 THE ART OF WAR IN THE MIDDLE AGES [1119

Meanwhile, the English king had seen the foe approaching,
and found ample time to draw up his host. He followed the
same general arrangement that had served him so well at
Tenchebrai. The majority of his knights were directed to dis
mount, and to send their horses to the rear. Only one hundred
kept their saddles. The exact details of the marshalling of their
host are not certain : of our three primary authorities Suger on
the French side, and Henry of Huntingdon and Orderic Vitalis
on the Norman no two agree. Suger tells us that the host was
drawn up with the horse in front and the dismounted knights
in a second line. 1 Henry of Huntingdon says that the king
made three battles the first of mounted Norman knights, the
second consisting of his private military household, headed by
himself, also mounted, the third on foot under his sons, Robert 2
and Richard, which was strongest of the three. 3 Orderic states
that there were a hundred knights under the king s son Richard
on horseback, while the rest of the Normans fought on foot
around the king, who was himself dismounted. 4 He does not
mention whether the horse were in front line or reserve, and
might be understood rather to imply the latter, as in his account
the first attack of the French seems to be directed against the
dismounted knights. 6 But since Suger and Huntingdon agree
in putting the horsemen in front, and Orderic does not actually
contradict them, we must not press his wording, and may Con
clude that Henry placed his infantry (if one may call them such)
behind his cavalry. Apparently he drew out the small body of
horse to allure the French to attack, and kept his strong force
of dismounted knights somewhat out of sight. 6 The one fact

Suger, p. 45: "(Henricus) milites armatos, ut fortius committant, pedites
deponit." Then the French charge, and "primam Normannorum aciem fortissima
manu caedentes a campo fugaverunt, et priores equitum acies super armatos pedites
repulerunt."

2 The famous Earl of Gloucester of the civil wars of Stephen s day.

3 "Rex Henricus in prima acie proceres suos constituerat : in secunda cum
propria familia eques ipse residebat : in tertia vero filios suos cum summis viribus
pedites collocaverat " (Henry of Huntingdon, p. 241).

4 Orderic says that Ricardus filius regis et c milites equis insidentes ad bellum
parati erant : reliqui vero pedites cum rege in campo dimicabant " (p. 722).

5 There were no more than the five hundred knights present on Henry s side.
The "grand pleinte de sergeants" whom the Grands Creniques de France introduce
are an invention, as can be seen by carefully comparing them with Suger.

6 This, I fancy, is what Suger means when he says that Henry speculatus regis
Francorum improvidam audaciam militum acies in eum dirigit : incentiva ut in euro
extraordinarie insiliat, ponit : milites armatos pedites deponit."

nig] BATTLE OF BREMOLE 383

on which our authorities arc hopelessly at issue, is that Orderic
says that the horse were commanded by the king s sons, while
Henry of Huntingdon says that they were led by the " proceres
Normannorum," i.e. the Counts of Eu and Warrenne, and that the
royal bastards led the infantry reserve. We cannot hope to
reconcile them on this point. The French can hardly be said
to have had any battle-array at all : l they rode up in disorder
in three troops, of which the first was headed by the Xorman
rebel William Crispin, and contained only eighty horsemen ; the
second (mainly composed of the knights of the Vexin) was
headed by Godfrey of Serranz, Bouchard of Montmorenci,
Ottomond of Chaumont, and Guy of Clermont ; the third was
led by the king and his seneschal, William de Garlande. Henry
of Huntingdon, however, speaks of the first two squadrons as if
they formed a single corps, and says that they had been placed
by King Louis under the orders of the young duke, William
Clito which is likely enough in itself, but conflicts with the
other authorities.

Orderic and Henry of Huntingdon agree in stating that
William Crispin charged first, and won a certain amount of
success : this success was, as we learn from Suger and Henry,
that he scattered and drove off the hundred horsemen whom the
English king had placed in front of his line. But then, plung
ing recklessly in among the serried ranks of the column of
dismounted knights, his men were surrounded, torn from their
horses, and made prisoners. He himself cut his way to Henry
and dealt him a severe blow, which was only prevented from
being fatal by the strength of the king s mail coif. But his
horse was killed under him, and Roger de Bienfaite threw him
down and captured him, saving him with difficulty from being
slain by the angry knights of the king s household. 2 When the
first French squadron was already practically disposed of, the
second charged in with equal courage, made the Norman phalanx
reel for a moment, but soon shared the fate of Crispin s men,
nearly all being surrounded, pulled down, and taken prisoners. 3

1 Rex autem, nullum pr.nelii constituere dignatus apparatum, in eos indiscrete
evolat " (Suger, 45).

- Cf. Orderic and Henry of Huntingdon : the latter says that William got two
fair cuts at the king s head, the former speaks of only one.

3 Suger speaks of the Vexin knights as being in the first charge : " Priores qui
manum applicuerunt Velcassinenses primam Normannorum acietn ... a campo
fugaverunt."

384 THE ART OF WAR IN THE MIDDLE AGES [1119

Seeing this disaster, the knights about King Louis advised
him to retreat: he turned his rein, and then his whole corps
broke up and fled in hopeless panic. The victorious Anglo-
Normans called for their horses, mounted, and pursued the
fugitives as far as the gates of Andelys. King Louis was so
closely chased that he had to spring from his charger and plunge
into a wood on foot Thence he escaped by devious paths, and
was led to Andelys by a friendly peasant. His horse and his
banner fell into the hands of the conqueror. A hundred and
forty knights were captured, but only three slain in the battle :
" for they were clothed from head to foot in mail, and because
of the fear of God and the fact that they were known to
each other as old comrades, there was no slaughter." x Of the
leaders of the t\vo front squadrons of the French no one
escaped captivity save William Clito. All the rest were made
prisoners.

The conflict of authorities on minor points does not prevent
us from having a very clear idea of the military significance of
Bremule. Disorderly charges of cavalry, unaided by either
infantry or archers, avail nothing against a solid mass of
well-armed knights on foot. Louis, seeing the Anglo-Norman
host in such good order, could only have had a chance of
success by dismounting some of his own knights, or by
bringing men armed with missile weapons into the field,
to harass the column of his adversaries. But he thought of
nothing but of sweeping them from the ground by a desperate
charge, and received the reward of his rashness in a crushing
defeat.

The records of an insignificant skirmish, which occurred a
few years after Bremule and would have escaped notice but for
its tactical interest, 2 suffice to show that the combination of
archery with the mounted arm was not wholly forgotten in the
Norman school of war. The memory of Hastings must always
have kept it alive. In 1124 Waleran Count of Mellent was in
rebellion against King Henry, and had drawn his kinsmen,
Amaury Count of Evreux, and Hugh of Neuchatel, into
his plot. But the royal forces were too much for him ; most
of his castles fell, and he and his knights became wanderers
on the face of the land. He had been raiding near Bourg

1 Orderic, p. 722.

- M. Delpech must have the credit of bringing it into notice.

ii24] COMBAT OF BOURG TH&ROULDE 385

The roulde, and committing horrid atrocities on the peasantry, 1
when he found himself intercepted by a body of three hundred
of the king s mercenary troops who had drawn together from
the neighbouring garrisons. They were headed by the chamber-
Iain William of Tankerville, and Ralph of Bayeux. 2 The
pursuers were superior in numbers, but they knew that Count
Waleran was in a desperate state of mind, and that his followers
were the best knights in Normandy. Instead of attacking, they
resolved to place themselves across the road and offer battle in
a defensive posture. Of the horsemen, part dismounted and
formed a solid mass, the rest remained on their steeds ; but
Ralph and William had with them not only knights, but also
bowmen, and, what is more surprising, mounted bowmen. We
should not have known of their existence but for the explicit
mention of them in William of Jumieges, for Orderic Vitalis,
the other narrator of the fight, does not mention the fact that
they were horsed. 3 Probably they were mercenaries, who had
been furnished with a mount in order that they might be able
to move rapidly along with the knights when pursuit was
needed. There were forty of them in the party; these men
Ralph and William placed on the left of their force, but thrown
forward en faience, so that they would take in flank any body of
men which charged up the road. 4 They were posted on the
left, in order that they might shoot at the unshielded right sides
of the rebels. Probably they dismounted in order that they
might use their bows to better effect. Waleran of Mellent
might have turned back and escaped by the way that he had
come. But, as his adversaries had calculated, the desperate
count had no such intention. He harangued his companions and
bade them ride down the pack of " mercenaries and rustics " 5 who
dared to block the way. He himself, with forty knights of his
meinie, headed the charge ; the rest, under the Count of Evreux,

1 His pleasing habit was to cut off one foot of the peasants who fell into his
hands (Orderic, p. 740).

a Orderic and William of Jumieges speak as if Ralph had been in command, but
Henry of Huntingdon and William of Malmesbury mention Tankerville only.

3 William, p. 576: "Denique catervis more pugnantium, necnon et equitibus
sagiltariis (quorum inibi exercitus regis maximam mukitudinem habehal) in dextra
parle hostium praemissis . . . clamor utrinque attollitur." \Villiamisatiresomeand
confused author, but can hardly have gone wrong on a point like this.

* "In prima enim fronte quadraginta archhenentes catallos occideiunt, et ante-
quam ferire possunt sunt dejecti" (Orderic, p. 74)-

" Gregarios et pagenses milites."

386 THE ART OF WAR IN THE MIDDLE AGES [1138

followed. But when they came level with the archers, the
latter let fly at their horses, and brought down nearly the
whole of them by a few well-directed volleys. The second
squadron suffered the same fate, and then the king s troops
advanced and took prisoners the whole party, for some were
pinned to the ground under their slain horses, and the others
were too heavily weighted by their mail, and too bruised and
shaken to get off rapidly. Eighty knights in all were captured,
including Waleran himself, and his nephews, Hugh of Neuchatel
and Hu r h of Montfort. The Count of Evreux would have
suffered the same fate had he not fallen into the hands of an
old friend, who collusively allowed him to escape.

This skirmish, exceptional in so many of its details,
distinctly reminds us of the tactics which Edward ill. was to
employ at Crecy two hundred years later. To receive a cavalry
charge by a body of dismounted men-at-arms flanked by
archers, while a mounted reserve remains behind to gather the
fruits of the day, argues a high degree of soldierly skill on the
part of the victorious commanders. Horsed archers are rarely
found in Western Europe in the twelfth century : they were no
doubt the predecessors of the mounted crossbowmen of the
time of John and Henry ill. Such troops were called into
existence by the need of having men armed with missiles, who
could keep up with the cavalry in their rapid marches against
raiders. Foot-bowmen could not have intercepted Waleran s
raid : but if provided with mounts of some sort, they might
reach the field ; they would then leave their horses and
join the knights, who had also sent their chargers to the
rear.

Batik of NorthaUcrton, August 22, 1138.

The celebrated " Battle of the Standard " differs in character
from the other fights which we have been investigating, in that
the enemy was not the mailed and mounted chivalry of France,
but the wild hordes of Celtic tribesmen from beyond the Tweed.
\Ve might have expected that the commanders of the Yorkshire
levies would have endeavoured to turn their superiority in
horse to good effect against the disorderly masses of Highlanders
and Galwegians ; but as a matter of fact they dismounted every
rider, as Robert of Normandy had done at Tenchebrai, and
the sole cavalry charge of the day was delivered by the small

1138] NORTHALLERTON : ARRAY OF THE ENGLISH 387

body of knights of English and Norman descent who served in
the Scottish host.

A short account of the battle will suffice, since neither side
showed any tactical insight or attempted any new device.
King David of Scotland had crossed the Tweed with a great
horde of Highlanders and Galloway men arrayed in their clans.
He led also the more orderly levies of the English-speaking
Eastern Lowlands, and many mailed knights of the exiled
English families who had removed to Scotland \vith Eadgar
Atheling, or of the Norman settlers who had drifted in somewhat
later. The Scots harried Northumberland and Durham with
great ferocity, slaying the priest at the altar, and the babe at its
mother s breast. Hence the Yorkshiremcn looked upon the
war as a crusade against savages, and marched out under the
banners of their saints, St. Peter of York, St. John of Beverley,
and St. Wilfrid of Ripon, all of which, together with that of St.
Cuthbert of Durham, were placed on a chariot and borne in
the midst of their host. The large majority of the English
consisted of the feudal levy and the fyrd of Yorkshire ; but
Stephen had sent some small succours from the south under
Bernard Baliol, and among the barons present we detect a few
who had brought their contingents from shires south of the
Humber, such as Derby and Nottingham. 1 The chief person
present was the young William of Albemarle, but Walter 1 Espec,
Sheriff of Yorkshire, seems to have shared the command with
him. They drew up their whole force in one deep line along a
hillside on Cowton Moor near Northallerton, with the chariot
bearing the standards in the rear of their centre. The knights
all dismounted and served on foot with the shire - levies,
apparently forming a mailed front line behind which the
half-armed country-folk arrayed themselves. There were a
considerable number of archers among the Yorkshiremen, who
are said to have been " mixed " with the spearmen. Presumably
they stood in the mass and shot over their friends heads, down
the slope, for there is no statement that they took position
either on the flank or in front of the main body. Some of
the elder knights formed a sacred band in reserve around the
Standard : among them stood the commanders of the host,
Albemarle and L Espec. 2

1 See John of Hexham, p. 262, for the men from Derby and Notts.
1 Richard of Hexham, p. 322 ; Aelred of Rivaulx, p. 343.

3 88 THE ART OF WAR IN THE MIDDLE AGES [1138

The King of Scots had a far larger army than his adversaries :
the total of twenty-six thousand men ascribed to him is probably
not very much over the real figure. But in mailed knights
and in archers he was comparatively weak : the vast majority
of his host were "Highland kerne" and Picts of Galloway,
armed with nothing more than a dart, a target, and broad
sword. Seeing the solid mass of the English awaiting him on
foot, David resolved to assail them with their own tactics, and
ordered his knights to dismount and form the head of the
attacking column, while his archers were to advance along with
them. The rest of the host was to follow, and to try to break in
when the knights made a gap in the English front. 1

But David had forgotten to reckon with the pride and
headlong courage of his Celtic subjects : they refused to let the
Lowland knights strike the first blow. The leaders of the
Galloway Picts claimed that they had an ancient right to take
the front place, and the Highlanders refused to give precedence
to the Norman and English strangers. 2 When the king
persisted in his design, Malise Earl of Stralhern, one of the
chiefs from beyond the Forth, angrily exclaimed, " Why trust
so much, my king, to the goodwill of these Frenchmen ? None
of them, for all his mail, will go so far to the front as I, who
fight unarmoured in to-day s battle." At this the Norman, Alan
Percy, cried, "That is a big word, and for your life you could
not make it good." The earl turned on him in wrath, and so
hot an altercation burst out between the Highlanders and the
Southern knights, that the king in despair withdrew his first
order of battle, and granted the Galloway men the foremost
place.

In the second scheme the Scots were drawn out in four
masses : as far as we can follow Aelred of Rivaulx s description
of the array, the Gahvegians were in the centre of the front
line, somewhat in advance. The two wings were formed, the
right by the king s son, Henry, with the greater part of the
knights of the Lowlands and the levies of Strathclyde and
Teviotdale, the left by the English of Lothian combined with
the West Highland clans of Lorn, Argyle, and the Hebrides.

1 " Placuit ut quotquot aderat militum armatorum et sagittarii cunctum praeirent
exercitum, quatenus armati armatos impeterent, milites congrederentur militibuF,
sagittae sagittis obviarent" (Aelred, p. 342).

3 Ibid. 342.

NORTHALLERTON: THE GALWEGIANS CHARGE 389

King David was in reserve, with the men of Moray and the
Eastern Highlands : he also kept about him as a bodyguard a
few of his modest contingent of mailed knights. 1

When the Scots drew near the hillside where the English
were arrayed, Robert Bruce, a Yorkshire baron, who held also
the lordship of Annandale in Scotland, rode down to the
hostile army and tried to induce the king to consent to terms
of peace. But the young knights about David s person taunted
Robert as a traitor, so that he had to withdraw, solemnly
disavowing his feudal allegiance for Annandale ere he went.

A moment later the Galloway men dashed at the English
centre, raising a terrible shout of " Albanach, Albanach ! " Their
wild rush made the fyrd waver for a moment, but the knights
rallied and sustained the common folks, and restored the line
without a moment s delay. 2 The Galwegians soon came back
to the charge: they shivered their light darts on the serried
line of shields which the Yorkshire men opposed to them, and
then laid on with their claymores. But they could not break
in a second time, and in the intervals between their charges
the archery galled them sorely. Yet they furiously returned,
"many of them looking like hedgehogs with the shafts still
sticking in their bodies," 3 to make one last bid for victory.

At this moment Prince Henry and his corps moved in
upon the English left wing. He and his few scores of knights
led the charge on horseback, the mass of Strathclyde men
following on foot. The charge was fairly delivered, and the
gallant prince with his horsemen hewed their way right through
the line of the Yorkshire men till they came out at the back of
the mass, scattering disorder all around them. Henry then
saw the horses of the enemy, held by the grooms of the English
knighthood, a short way to the rear. He rode on to seize
them, thinking that the infantry of his corps would penetrate into
the entry that he had made, and reckoning the battle as won.

1 Richard of Ilexham, whose account of the Scottish array is incomplete, only
says that the Galwegians were in front, the king and a bodyguard of English knights
in the mid-battle, while the clans were around him, "cetera barbaries circumfusa
etat " (p. 322).

* "Galwegensium cuneus tanto impetu irruit in australes, ut primes lancearios
stationem deserere compelleret, sed vi militum iterum repulsi in hostes animum et
vigorem resumunt " (Aelred, p. 345).

1 Videtes ut ericium spinis, sic Galwegensem sagittis undique circumseptum
nihiloniinus vibrare gladium, et caeca amentia proruentem nunc hostem caedere, nunc
ininem aerem cassis ictibus verberare" (Aelred, p. 345).

39 THE ART OF WAR IN THE MIDDLE AGES [1138

Herein he was sadly mistaken : he wasted but a few minuted
in dashing at the horses, but those few minutes were the crisis
of the day. The English closed up the gap through which he
had cut his way, and drove back the Strathclyde men who
strove to thrust themselves into it. Meanwhile, in the centre
the fire and fury of the Galwegians was used up : leaving their
chiefs Donald and Ulgerich dead on the field, they dispersed
and fled. On the Scottish left wing the men of Lothian and
Lorn behaved far worse : their leader (his name is not given)
being slain by an arrow in the first clash of spears, they made
no second charge, and retired tamely to the rear. King David
now ordered his reserve of Highlanders to advance, and sprang
off his horse to lead it forward. But, seeing the disasters in
the front line, the fickle Celts began to melt off to right and
left, and David soon found himself alone with his small body
guard of English and Norman knights. It Avas hopeless to
proceed, so he bade his standard - bearer turn back, and with
drew to a neighbouring eminence, where there presently
assembled round him the wrecks of his host. The mass looked
so formidable that the Yorkshiremen dared not attack it, but
waited till it began to retreat. Then they followed at a
distance, slaying stragglers and taking many knights prisoners.

Prince Henry, having (as we have seen) worked his way to
the very rear of the English line, was left in a position of
desperate danger when the Scottish host broke and retired.
He saved himself by a ready stratagem : he wheeled and faced
to the north, then, bidding the few knights around him throw
off their badges l and mingle with the advancing line of the
enemy, he pushed on unobserved along with the English, and
gradually passed through them. When safely in advance of
their foremost ranks, he moved off at a moderate pace, so as not
to awaken suspicion, and finally got clear away, rejoining his
father by a circuitous route on the third day. The Scots
suffered very heavily in the fight, though the ten thousand or
eleven thousand dead of which the chroniclers speak are only
one more instance of the usual mediaeval inability to deal with
high figures. It is more credible that of two hundred knights

" Projectis itaque signis quibus a caeteris dividimur, ipsis nos hostibus inferamus,
quasi insequentes cum iis." What were the signal Probably not coats-of-arms,
which were only just coming into use, but some common token which the Scots were
all wearing to distinguish them from the English.

1138] NORTHALLERTON : THE SCOTS RETREAT 391

whom Henry led to the charge fifty were captured, and so
many slain and wounded that only nineteen came back un
touched with horse and arms. The prince himself had cast
off his mail -shirt when the battle was over, 1 refusing to be
burdened with it in the long ride across the moors which lay
before him ere he could rejoin his father. The slaughter
among the chiefs had been very heavy in all the front divisions
of the Scottish host : only the king s corps, which behaved so
tamely, had got off fairly unscathed.

Of the English, only one knight, the brother of Ilbert de
Lacy, had fallen ; but a considerable number of the half-armed
fyrd had been trampled down in the first rush of the Galwegians
and in the desperate charge of Prince Henry.

Thus ended the Battle of the Standard, a fight of a very
abnormal type for the twelfth century, since the side which had
the advantage in cavalry made no attempt to use it, while that
which was weak in the all-important arm made a creditable
attempt to turn it to account by breaking into the hostile flank.
The tactics of the Yorkshiremen remind us of Harold s arrange
ments at Hastings, even to the detail of the central standards
planted on the hill ; but they had this advantage over the
Saxon king, that they were well provided with the archery in
which he had been deficient. David s plan of attack was not
unwise, but he was ruined by the Celtic pride and Celtic fickle
ness of his followers. If his two hundred knights could have
opened a gap, and the fierce Galwegians could have thrown
themselves into it, the fortune of the day might have been
changed. But wild rushes of unmailed clansmen against a
steady front of spears and bows never succeeded : in this
respect Northallerton is the forerunner of Dupplin, Halidon
Hill, Flodden, and Pinkie. The most surprising incident of the
fight is the misconduct of the English - speaking spearmen of
Lothian on the Scottish left wing : it was not usually the wont
of the men of the Lowlands to retire after a single onset and
when there was no pursuit. Possibly they had no great heart
in the Celtic crusade against England, and were discontented
at the king s subservience to the Highlanders. It is certain
that during the retreat the Lowlanders and Highlanders fell
out and came to blows, each accusing the other of cowardice

1 Aelred; p. 346. He gave it to a peasant by the way, saying, " Accipe quod
mihi est oneri, tuae consulat necessitati."

392 THE A XT OF WAR IN THE MIDDLE AGES [1141

and treachery, 1 " so that they came home not like comrades
but like very bitter enemies."

First Battle of Lincoln, February 2, 1 141.

When we turn to the battle of Lincoln, we find ourselves on
more familiar ground, and recognise the old tactics of Tenchebrai
and other Anglo-Norman fields. Unfortunately we have for this
important fight no account of such merit as Aelred of Rivaulx s
excellent narrative of the Battle of the Standard.

In the winter of 1140-41 the barons of the West and the
Welsh border were up in arms against King Stephen, and had
sworn allegiance to his rival, the Empress Matilda. Among the
many strongholds which they had taken was the very important
castle of Lincoln. The king marched against it in the depth of
the winter, and seized the city (whose inhabitants were friendly
to him), while the rebels retired into the castle. He lay before its
walls for a month, during which space the Earls Ralph of Chester
and Robert of Gloucester were collecting an army with which
they purposed to raise the siege. On the first of February 2 their
approach was reported to the king ; his counsellors advised him
to refuse a battle, and to call in his adherents from the south,
since he had but a small force with him. But Stephen despised
his enemy, and announced his intention of fighting at once. To
get at him the earls had to cross the flooded Fossdike, 3 and a
guard had been set upon the fords to keep them at bay. But
on the morning of February 2 Ralph and Robert forced the
passage, though the water was deep and the marshes dangerous :
the corps which Stephen had set to observe them was easily
brushed away.

Hearing of their approach, the king drew up his army in front
of the walls of Lincoln. In the absence of any precise indica-

1 " Rex, recollectis suis qui sparsim de pugna, non ut consortes, sed potius sicut
hostes inimicissimi fugerant, obsidionem apud Carbarn corroboravit. Nam Angli
et Scoti et Picti, quocunque casu se inveniebant, alios mutuo vel trucidabant vel
vulnerabant vel saltern spoliabant, et ita a suis sicut ab alienis opprimebantur "
(Richard of Hexham, p. 323). Angli of course means the Lowlanders, Scoti the
Highlanders, and Picti the Galloway men.

2 Stephen took ihe town "circa natale domini" (December 25), and was still
before the castle on February I, when the enemy appeared.

3 A channel cut from the Trent to the Witham in the time of Henry I., which
protected the south - west front of the city. This must be the stream, not the
Trent, as some chroniclers put it. I am glad to find that on this point I agree with
Miss Norgate s Angevin Kings.

U4i] LINCOLN: KING STEPHEN S ARRAY 393

tion of the battle spot, we have to put the following facts
together in order to identify it. (i) The earls forded the
Fossdike somewhere west of Lincoln. (2) They fought with it
at their backs, so that defeat meant disaster ; i.e. they faced
north or north-west (3) The routed cavalry of Stephen s host
escaped into the open country, not into the town ; i.e. they were
drawn up so as to give a free flight to the north. (4) The infantry
fled into the town, which was therefore quite close. Probably
the battlefield lay due west of the city, and the Royalists
apparently faced south or south - west. Stephen used the
tactics which his uncle Henry I. had employed at Bremule:
the greater part of his knights were ordered to dismount
and fight on foot around the royal standard ; with them
were incorporated some infantry of the shire-levy, mainly
composed of the citizens of Lincoln. 1 In front of this mass
of dismounted men were drawn up t\vo small " battles " of
horsemen ; that on the left was headed by William of Albe-
marle, whom the king had made an earl for his services
at Northalle,rtor., and by William of Ypres, a mercenary captain.
That on the right was under a multitude of chiefs Hugh Bigot
Earl of Norfolk, William Earl of Warrenne, Simon of Senlis Earl
of Northampton, Waleran Earl of Mellent, 2 and the mercenary
Alan of Dinan, whom the king had created Earl of Richmond.
But these great names represented no great following; several
of them were pseudo-comites, men whom the king had made
earls in title, though their power and estates did not justify
the promotion ; 3 it was said that they had no more connection
with the counties whose names they bore than that of receiving
the third penny of the shire-fines. The rest had come to Lincoln
without their full seri itium debitum of knights, "as if to a
colloquy, and not to a battle." 4 The two squadrons between them
only mustered a very few hundred knights.

The rebel earls likewise drew up their host in three main
corps. One was headed by Ralph of Chester, the second division
by Robert of Gloucester, the third was composed of the numerous

1 We get this fact from the speech of Earl Ralph in Henry of Huntingdon.
Recapitulating the king s forces, he says: "Gives Lincolnienses, qui slant suae urbi
ptoximi, in impetus gravedine ad domos suas transfngere videbitis " (p. 269).

2 The vanquished rebel of the skirmish of Bourg Theroulde (see p. 384).

1 "Paucosenim militessecunvfc/; ^/af/iv/comitesadduxerant"(Gervase, p. 1354).
4 " Persuaserunt Seniores regi congregare exercitum, sese enitn inermes ad regis
colloquium occurrisse, non ad praelii precinctum profitentes."

394 THE ART OF WAR IN THE MIDDLE AGES [1141

Midland knights and barons whose estates Stephen had declared
confiscated for rebellion ; the chroniclers call this corps the
" array of the disinherited." Robert had also brought with him
from the marches a body of Welsh light infantry under t\vo brothers
named Meredith and Cadwallader ; these wild levies, " courageous
rather than formidable," 1 as the chronicler calls them, were thrown
out on the flank of the front line. Ralph of Chester and his
knights dismounted and formed the reserve, incorporating with
themselves (just as Stephen had done) the remaining infantry
of their host. 2 In the front line the " disinherited " faced Bigot,
Mellent, Alan, and the other earls, while Robert and the Welsh
were opposite Albemarle and William of Ypres. The number
of horsemen on the two sides was about equal; 3 the king had
the advantage in foot-soldiery.

The first clash came when the cavalry divisions of the front
line charged. On the one wing the " disinherited " completely
broke and scattered the five earls, whose whole squadron was in
a moment either slain, captured, or in wild flight* On the other
flank William of Albemarle and William of Ypres came into
collision with Gloucester s knights and the Welsh light infantry.
The Royalists rode down the Welsh and drove them to take
shelter with the Earl of Chester and the barons reserve. But when
they were assailed at once by Gloucester s horse and Chester s
mailed foot, they gave way, and the two Williams fled in rout
as prompt and complete as that of the earls in the other wing.
None of the beaten Royalist horse made any attempt to rally:
looking back on the field, William of Ypres observed that " the
battle was lost, and that they must help the king some other
day," and continued his flight.

Then the whole army of the rebel earls concentrated their
efforts on the king s column of infantry ; apparently Chester and
his dismounted knights charged it in front, while the "disin
herited " and Gloucester beset it on the flanks and rear. The
Royalists made a gallant resistance, but at last the mass was
broken up ; those who could sought refuge within the gates of

1 " Audacia tnagis quam armis instruct! " (Huntingdon, 268).

2 " Animosam legionem Cestiensium peditutn " (Orderic, 769).

3 See Baldwin s speech in p. 272 of Henry of Huntingdon : " Nobis numerus in
equitibus non inferior, in peditibus confertior." This is more probable than Orderic s
"bostes nimia multitudine peditum et Wallensium praevaluenint " (769).

4 " In ictu oculi dispersi sunt, et divisio eorum in tria devenit : alii namqne occisi
sunt : alii capti ; alii aufugerunt" (Henry of Huntingdon, 273).

PLATE XII.

TENCHEBRAI

SEPT 28. 1106.

E3 Normans

3 English .

Robert Ml,.
JfcllmJ- dc Wuorn

Count
Hclie

BREMULE

AUC 30.1119
King Louis

W Crisp in

Robert &

Itench
Enjhah

Kutg Henry

NORTHALLERTON

AuG.22.//38.

Kin*Pl)ivid

I Lotfiian

fellow,

3col&gt;

English

1 st Battle of
.LINCOLN

Royalists
Rebels

2 " a Battle of

LINCOLN

MAY 19 IZM.

6 Th^Blockcd (j.-ue.

C Place o

and death of the
Count of Perchc

D CaAtlc PhsttTIl

E The Royalist Army

LINCOLN. THE KING CAPTURED 395

Lincoln, where the foe promptly pursued them and cut them up
in the streets. But Stephen and his truest followers stood firm
by the standard, and held out long after the rest of the fighting
was over. The king fought till his sword was broken, and then
used a Danish two-handed axe which a citizen of Lincoln
slipped into his hand. 1 His terrible strokes long held the rebels
at bay, but at last a final rush swept down his faithful band, and
he himself was thrown to the ground by William de Caimes, a
powerful knight, who caught him by the helmet and dragged
him over. With him were captured Bernard Baliol, Roger de
Mowbray, William Fossart, William Peverel, William de Clerfait,
Baldwin Fitz-Gilbert, Richard Fitz-Urse, and many other gallant
knights and barons. 2

The first battle of Lincoln is a perfectly normal and typical
thirteenth-century engagement. Each side used the same tactics
of a front line of horse and a reserve of dismounted knights :
the Welsh light infantry on the rebel flank are the only unusual
feature, and they had no influence whatever on the event of the
day. Probably they were South Welsh archers, intended to gall
the flank of the Royalist horse by a cross-fire, like the bowmen
at Bourg TheVoulde in 1124. Putting them aside, we see that
the battle was lost because Stephen s cavalry were so dis
comfited that they could not rally behind the reserve and
return to the fight. When they had left the field, the king s
fate was sealed : like his uncle Robert at Tenchebrai, he found
that infantry unsupported must fail before horse and foot
combined.

Of the reign of Henry 11. even more than of the rest of the
twelfth century is the statement true that the age was one of
sieges rather than of battles. All through his reign the king was
fighting hard, yet he was never present at an engagement of
first or even second-rate importance in the open field. Only twice
was he even on the edge of a great battle once at the raising
of the leaguer of Rouen in 1174, and once when, in 1187, he lay
by Chateauroux with a great host, while Philip of France on the
other side of the Indre was drawing out his arm} day after day,
and offering to fight if the Anglo-Normans should endeavour to

1 John of Hexham, p. 269.

8 For the list see John of Hexham, p. 269. He is by far the most full in
enumeration.

396 THE ART OF WAR IN THE MIDDLE AGES [1173

pass the river. Both kings were prudent, and would not risk
the passage, and finally they made a truce instead of settling
their quarrel with the sword.

In the troublous years 1173-74, when Henry s enemies
were in arms on all sides, and half England was overrun by the
rebels, there were two engagements of high political importance,
but neither takes rank as a real battle or gives us any interest
ing tactical features. The disaster of William the Lion at
Alnwick was a curious instance of a great invasion stopped by
the chance encounter of a few hundred knights. The King of
Scots had invaded Northumberland with an army not less than
that which his grandfather led to the Battle of the Standard.
He lay before Alnwick with part of his force, while the rest were
raiding far and wide in the valleys of the Tyne and Tees. Mean-
while,Robertd Estouteville,the Sheriff of Yorkshire.had mustered
the shire-levies of the great county, and the loyal barons of the
north had gathered to his aid. They resolved to march towards
Alnwick, but cautiously, since they knew that the Scots out
numbered them fourfold. In the long march from Newcastle
to Alnwick the knights outrode the weary infantry. On the
morning of June 13, 1174, they found themselves close to the
beleaguered castle, but a heavy fog lay over the face of the land,
and it seemed reckless for four hundred knights to try to pick
their way between the besiegers camps in the darkness. They
attempted the dangerous feat, and were rewarded by an unex
pected prize. When they had ridden some miles, the fog cleared,
and Alnwick was seen close at hand ; but closer still was a small
body of mailed men riding at leisure round the castle. It was
King William and a party of his knights : the rest were out
raiding or scattered in distant camps. The king at first
thought the English were some of his own host, and cantered
unsuspiciously toward them. Only when he was too close to
flee did he recognise the hostile banners: seeing his danger, he
cried, " Now shall we see who is a true knight," 1 and, levelling his
lance, rode at the Yorkshiremen. This foolish feat of chivalrous
daring had the natural result : his horse was slain, and he
himself and all his companions were captured. His host broke
up and retired in confusion into Scotland the moment that the
disastrous news got abroad. Thus a great invasion was foiled

1 " Modo apparebit quis miles esse invenit " (William of Newbury, 185).

ii73] BATTLE OF FORNHAM 397

by a trifling skirmish, in which less than five hundred knights
took part.

Of the fight of Fornham (October 17, 1173), the other blow
which crushed King Henry s enemies, we could wish that we had
better details. The rebel Earl of Leicester was marching across
Suffolk from Framlingham towards his own county with eighty
knights and three thousand Flemish mercenaries, horse and foot,
whom he had imported to strengthen his rebellion. To inter
cept him, the Constable Humphrey de Bohun and the Earls of
Arundel and Cornwall marched to Bury St. Edmunds with a few
loyal knights and three hundred of King Henry s stipendiary
horsemen. The shire-levy of Suffolk and Cambridge joined
them in great force, for the Flemings had made themselves
hated by their cruel ravages in Xorfol.k. They were reported to
have sung to each other,

" Hop, hop, NVillcken, hop ! England is mine and thine,"

and the fyrd came out readily against them, though many were
armed with nothing better than flails and pitchforks. 1 The
host of the Constable outnumbered the rebels fourfold, but, as
Ralph de Diceto remarks, if only properly armed men counted,
the earl had far the more formidable following. 2 De Bohun,
following, caught him as he was passing a marsh near Fornham,
and, falling upon him suddenly, discomfited the rebels in a few
moments. Apparently the whole fight was a surprise, for the
Flemings seem to have found themselves in a helpless plight,
and Leicester and his knights fled early. 3 The infuriated
peasantry gave no quarter, and thrust the foreigners into bog
and ditch till more were drowned than slain with stroke of
sword. 4 Only a very few survived to share the captivity of the
earl and his high-spirited countess, who had gone through the
campaign at her husband s side. Such a rout of trained soldiers
by raw levies led by a few hundred horsemen, can hardly be
accounted for save by the hypothesis that the rebels were
surprised in a place where cavalry could not act freely : they

1 Matthew Paris, Hist. Angl. 381.

2 Ralph de Diceto, 377 : " Si milites regis militibus comitis conferantur regalium
numerus militiam comitis excedet in quadruplum. Si vero capita capitibus, si
armatorum copiam aequa lance quis colligat, multo plures erant cum coroite quatn ex

adverse.

3 " In ictu oculi victus est comes et captus (Hoveden, 307).
Jordan Fantosme, p. 294, line 1091.

398 THE ART OF WAR IN THE MIDDLE AGES [1169

allowed themselves to be attacked by the Royalists, made no
attempt to take the offensive, and hardly stood for a moment.
If the ground had been firm and open, they must surely have had
the better of the fyrd.

The English in Ireland, 1169-75.

We have, as it chances, a far better knowledge of another
set of Anglo-Norman fights than of those of the great rebellion
of 1 173-74. The Expugnatio Hiberniae and the invaluable Song
of Dermot and t/ie Earl 1 enable us to form a very clear notion
of the tactics and strategy by which a few hundred knights of
the Marches of Wales subdued within the space of five years
the better half of Ireland. Of all the many conquests of the
Normans in East and West, this was perhaps the most astonish
ing, for the resources of the invaders were weaker even than
those of the conquerors of Naples and Sicily, and the Irish
dwelt in one of the most difficult and inaccessible regions of
Europe.

Ireland in 1169 was one vast expanse of wood, bog, and
mountain, in which the tracts of open land were few and far
between. Between every tribal settlement lay difficult passes
over marshes or between woods and rocks. The natives, if
fickle and ill compacted, were not wanting in wild courage, and
had in their long wars with the Danes evolved a system of
defensive warfare which was well adapted to the character of
their country. On every trackway which led from district to
district there were well - known positions which the tribesmen
were wont to fortify with considerable skill. In the bogs they
dug trenches across the road and erected stockades on the farther
side, so that the passage was almost impassable for horsemen.
In the forest tracts they "plashed the woods," i.e. cut down the
underwood and wattled it together in abattis across and along
side of the roads, so that those who tried to force their way
through found themselves beset on flank and front by unseen
enemies, who could only be reached by hewing down the
screen of thick boughs. The Song of Dermot and the Earl
is full of descriptions of barriers of these two kinds : the
account of the pass of Achadh-Ur (Freshford in Kilkenny)
may serve as an example. This was a passage between the

1 I have of course used Mr. Orpen s excellent edition of 1892.

1169] THE ENGLISH IN IRELAND 399

river Nuenna and steep wooded hills. Mac-Donnchadh, king
of Ossory

" Un fosse fist jeter aitant
Haut e large roist e grant,
Puis par afin richer
E par devant ben herdeler,
Pur defendre le passage
Al rei Dermod al fer corage."

" He bade his men throw up a trench high and wide, steep
and large, and to strengthen it at the back with stakes and in
front with hurdles, in order to dispute the passage of King
Dermot the stout-hearted" (lines 1013-19).

Whenever the English marched out, the Irish "plashed the
woods and dug across the roads" (line 1 595), and it was hard to get
from place to place " on the hard field and by the open ground."
Such tactics were most distressing to invaders accustomed to win
by the ponderous charge of mailed cavalry across the unenclosed
fields and hillsides of England or Normandy. Yet, as we shall
see, they succeeded in triumphing over these difficulties, and
firmly established themselves in the conquered land.

The weak point of the Irish was their want of defensive
armour and their inability to stand firm in the open. If once
the enemy could close with them, and catch them far from the
shelter of stockade and trench, they were easy to deal with, for
they dreaded above all things the impact of the mailed horse
man, and had never learned to stand fast, shoulder to shoulder,
and beat off the charge of cavalry. Neither they themselves
nor their old enemies the Danes were accustomed to fiszht on

horseback, and they were utterly cowed by the Norman knight
and his reckless onset. Their arms, indeed, were very unsuited
to resist cavalry : only the Scandinavian settlers of the coast-
towns and a few of the chiefs of the inland wore mail ; the rest
came out " naked " to war. As one of their own bards sang

" Unequal they engaged in the battle,
The foreigners and the Gaedhil of Teamhair ;
Fine linen shirts on the race of Conn,
And the foreigners in a mass of iron." 1

Nor were the offensive arms of the Celts very suitable for
repelling cavalry ; they carried two darts, a short spear, and

1 Poem of Gilla Bhrighde M Conmidhe, quoted by Mr. Orpen in Dermot and the
Earl, p. 268.

400 THE ART OF WAR IN THE MIDDLE AGES [1169

large-headed axes wielded by one hand, but had no long pikes
nor any skill in archery. 1 They hurled darts and stones at
close quarters from behind their stockades and fosses, but could
not keep off their enemy by the distant rain of arrows. In short,
they were formidable while skirmishing in woods and bogs, but
easily to be routed in the open.

The Anglo-Norman leaders soon learned to adapt their
tactics to those of the enemy. They had to avoid, as far as
possible, fights in woods or bogs, and to lure the enemy into the
clear ground. If this was impossible, and if the Irish stood firm
behind their defences, the only courses open were either to essay
surprises and night attacks the Celts habitually kept a very
poor watch or to gall the defenders with arrows from a dis
tance. Fortunately for themselves, the knights of the Welsh
March had close to their hand the very associates most suited to
aid them in such difficulties. The men of South Wales were the
most skilled of all the inhabitants of Britain in archery, and drew
the longest and the strongest bows. It was by their aid that the
invaders were accustomed to triumph over the Irish hordes.
None of the barons who won Ireland ever marched forth without
a large provision of bowmen, and after a time they habitually
mounted them, in order that they might be able to keep up with
the knights in every chance of war, and might not be left behind
in rapid advances or pursuits. Giraldus Cambrensis in his
Expugnatio devotes the best part of a chapter to explaining the
advantage which the Welsh archers gave to the invaders, and
urges the leaders of his own day to enlarge the proportion of
Welsh in their bands, 2 on account of their lightness and swift
ness, which enabled them to follow the Irish into heavy or moun
tainous ground, where the mailed men could pursue only slowly
or not at all. A few descriptions of battles will show how the
Anglo-Normans contrived to deal with their adversaries.

Battle on the Dinin, \ 169.

Dermot of Leinster, with his allies, Robert Fitz-Stephen and
Maurice de Prendergast, had executed a successful raid into
the lands of his enemy MacDonnchadh, King of Ossory. They
had with them three hundred knights and archers of Wales, and

1 Topographia Hilerniae of Giraldus Camb. p. 151.

- See his Expugnotio t book II. chapter xxxviii. : "Qualiter gens Hibernica
expugnanda sit."

1169] BATTLE OF THE DININ 401

a thousand of Dermot s followers from Hy-Kinselagh (County
Wexford). On their return they had to cross a defile between
wood and water, in the valley of the Dinin. The Irish were march
ing first, under Donnell Kavanagh, King Dermot s son ; behind
were the king himself and his Anglo-Norman allies. When the
pass was reached, the men of Ossory were found stationed there in
great force, under their king. The spot was dreaded by the men
of Kinselagh, for three times had the army of Leinster been routed
there within King Dermot s reign. When they found themselves
attacked, they lost heart at once, and fled into the woods :
Donnell Kavanagh only brought forty-three of his followers back
to his father s side. The English were at the bottom of the marshy
valley, in a place where they could not easily resist an attack,
and a move onward to seize the well-manned pass seemed
equally hopeless.

Maurice de Prendergast at once proposed a retreat from the
valley and the woods up to the high open ground from which
the army had descended in order to attempt the pass. If the
men of Ossory should follow them, as was likely, it would be
possible to turn upon them where neither trees nor marsh pro
tected them from the charge of the Norman horse. His advice
was promptly carried out ; the Anglo-Normans retired up the
hillside with every sign of hurry and dismay. When they began
to approach the end of the wood, they dropped forty archers
under a certain Robert Smiche (Smithe ?) by the wayside, with
orders to hide in a thicket till the Irish should have passed
by, and to fall on their rear when the opportunity came.

The precipitate retreat of the invaders had the effect that
Prendergast had hoped. MacDonnchadh and " all the pride of
Ossory " came out in haste from their impregnable position, and
followed them across the valley and up the hill. They passed
the ambush without noticing it, and swept out into the open
ground. When they had left the wood some way behind, they
were surprised to see the Normans turn and form line of battle.
Before the meaning of the movement was realised, the knights
charged in among them, the archers and sergeants following
close behind. The Ossory men were six or seven to one, their
numbers are given at from seventeen hundred to two thousand *

1 In line 659 the author of Dermot and the Earl calls them "mil e set scent,"
but in 718 "par aime erent ii millers." Neither figure seems too high, considering
the usual exaggeration of the mediaeval poet.
26

402 THE ART OF WAR IN THE MIDDLE AGES [1169

strong, but they could not stand for a moment against the
impact of the mailed horsemen. They were broken and scattered
in all directions with great slaughter: whether the ambush of
archers fell upon the fugitives with much effect we are not told,
but the cowardly men of Hy-Kinselagh emerged from the woods
where they had been skulking, and hunted the fugitives for some
distance. They brought back two hundred and twenty heads
no quarter was given in Irish war and laid them at King
Dermot s feet. To the horror of his auxiliaries, the brutal king
was seen 1 to pick out the head of one of his special enemies,
and to tear with his teeth the nose of the fallen chief.

The feigned retreat which won the battle of the Dinin was an
old Norman device, whose most famous example was seen at
Hastings. Without its use the army of Dermot and Fitz-Stephen
must have been crushed in the valley between the marsh and
the wood, where no cavalry charge would have been possible.

The next two engagements which we must notice were both
fought close to the walls of Dublin, 2 which had fallen into the
hands of the English in the autumn of 1170, its Danish lord,
Haskulf Thorgilson, having been expelled and driven to seek
refuge in the Western Isles. Richard de Clare, the famous
" Strongbow," was now at the head of the invaders, and had laid
claim to the whole kingdom of Leinster, since the death of his
father-in-law, King Dermot, in May 1171. It was only a fort
night after his accession that a Viking fleet cast anchor in Dublin
Bay. Haskulf had sought aid from the Scandinavian settlers
in Man, Orkney, and the Hebrides, and had gathered a fleet of
sixty sail to restore him to his lost possessions. His auxiliaries
were led by an adventurer named John "the Madman " or "the
Furious," 3 a famous "Baresark," who had won much glory in
the wars of the North. The Norsemen landed, ten thousand
strong, or even more, according to the estimate of their enemies,
which must be wholly futile : Orkney and Man could not have
supplied half that number of warriors. They formed up on the

1 Giraldus, Expiignatio, \. 4. The author of Dermot and the Earl does not give
this discreditable trait of his hero s conduct.

2 It is strange to find that Giraldus and the author of Dermot differ as to the
order of the two sieges : Giraldus puts the Danish siege in May and the Irish siege
in June, while the poet makes the Danish siege so late as September, three months
after Roderick s.

Joannes " Insanus " or "Vehemens" or " Le Wode " in Giraldus (p. 264).
The Song of Derntot calls him Jean le Devc (from Jesver, to go mad).

BATTLE OF DUBLIN 403

shore and marched toward the city in a solid column, all clad
in mail-shirts and bearing their Danish axes on their shoulders.
This was a host very different from the hordes of naked Irish
with whom the invaders had hitherto had to cope, and far more
formidable.

Battle of Dublin, May 1171.

Miles Cogan was in command of Dublin in the absence of
his master, Earl Richard. He had with him about three
hundred mounted men, 1 besides archers and sergeants on foot,
probably fifteen hundred men in all, if the infantry bore to the
cavalry the proportion that was usual in the bands with which
the Anglo-Normans overran Ireland. Miles came out at first
into the open, with his archers and spearmen in front and his
knights in second line. But he was unable to break into the
Viking ranks, and was forced back against the eastern gate of
Dublin (St. Mary s Gate or Dame s Gate). Foreseeing that this
might occur, he had previously detached his brother, Richard
Cogan, with thirty knights, to issue from the town by its western
gate (Newgate), fetch a compass around the walls, and fall on
the rear of the enemy. The main body of the English was
barely holding its own about the east gate when a shout from
the back of the Viking host told them that the diversion had
begun. Richard and his knights had made a desperate charge
into the rear ranks of the Norsemen. " When John the Wode
scented the noise of those behind and the shouting, he departed
from the city, he wished to succour his friends who were left
behind ; John and his meinie, ten thousand strong or nine (I
know not which), departed from the city to succour their com
panions in the rear." 2

The diversion, trifling as it was, had checked the Norse attack,
and in the confused movement towards the rear the solid column
had been broken up, and gaps showed in it. Miles and the main
body of the English, horse and foot, threw themselves upon the
mass. The knights succeeded in penetrating into the heart of
the column, and wrought so much damage among the Vikings
that they began to retire in disorder towards their ships. John
the Wode refused to fly, and fought with astonishing strength
and courage ; he struck one knight such a fearful blow with his
two-handed axe that he hewed off his thigh in spite of hauberk

1 Song of Dermot, line 2384. " Sottg of Derm ot, lines 2375-80.

404 THE ART OF WAR IN THE MIDDLE AGES [1171

and mail breeches, and slew some nine or ten English before he
was himself cut down. Haskulf Thorgilson was taken prisoner l
at the same time, and the Vikings ere long fled in complete
rout. Some Irish levies of uncertain loyalty under one Domnahl
Macgille Moholmog 2 had been watching the fight from afar,
ready to turn against whichever side had the worst of the
encounter. When they saw the Norsemen break up, they rushed
down and aided in the slaughter of the fugitives. Two thousand
were slain and five hundred drowned on the beach before the
survivors succeeded in thrusting their galleys out to sea and
getting into the offing. 3

Surprise of Castle Knock, July 1171.

Only a month after the Vikings had been beaten, another
army appeared under the walls of Dublin. This time it was
Roderic O Connor, the high-king of all Ireland, with sixty
thousand men levied from all the clans of the island. They
encamped around Dublin in four separate bodies the high-king
and his men of Connaught at Castle Knock ; Macdunlevy and
the clans of Ulster at Clontarf the site of Brian Boroihme s
old victory ; O Brien of Thomond at Kilmainham ; and Murtough
M Murrough with the men of Leinster at Dalkey. Earl Richard
had by this time returned to his capital and taken over the
command from Miles Cogan, but he was in despair at the over
whelming strength of the array which O Connor had brought
out against him, and did not dare to stir from the walls. After
a siege of six weeks, famine began to threaten the garrison.
"The measure of wheat was sold for a silver mark, and the
measure of barley for half a mark." * Nor was there any hope

1 He was beheaded after the fight. He had been reserved for ransom,
but so angered his captors by his haughty answers that they slew him (Giraldus
p. 265).

The Song of Dermot tells us that Miles Cogan, knowing Domnahl s fickleness,
had bade him stand afar off and strike in against the losers. " If these men be dis
comfited, then you shall aid us with your force to overthrow them. But if we be
recreant, then you shall aid these men to cut us to pieces and slay us." To this
the Irishman readily consented (lines 2300-2310).

3 The Song of Dermot says that two thousand Norsemen escaped, two thousand
were slain, and five hundred drowned. This would give a total of four thousand five
hundred for their army a far more probable figure than the nine thousand or ten
thousand given above, or the impossible twenty thousand which is also attributed to
the Vikings.

4 Song of Dermot, lines 1825-30.

ii7 1] STRONGBOW S SORTIE FROM DUBLIN 405

of bringing in provisions by water, for Guthred, King of Man,
was lying in the bay with a Viking fleet the relics, no doubt, of
the armament of John the Wode.

Richard endeavoured, therefore, to make peace with King
Roderic, offering to hold Leinster as his vassal and do fealty to
him. But O Connor replied that he might hold the three to\vns
of the Ostmen, Dublin, Waterford, and Wexford, but not a foot
more. These terms appeared so hard to Earl Richard that he
resolved to hazard a sortie, in spite of the desperate odds against
him. On the very afternoon of the abortive negotiations he
marshalled the forces which could be spared from garrisoning
the ramparts, and marched out against the camp at Castle
Knock (five miles from Dublin) in three small columns. Each
was composed of forty knights, sixty mounted archers, 1 and a
hundred sergeants on foot. 2 Miles led the first, Raymond Le
Gros the second, and the earl himself the third. They hurried
at full speed from the west gate and reached the camp of the
men of Connaught before the alarm was given. The Irish were
caught entirely unprepared ; they were lounging about their
cabins and huts, and the king himself was in his bath. They
had surrounded their encampment with a stockade, but no one
was in arms to guard it. The invaders broke in easily at three
points, and rode through the lanes between the huts, hacking
and hewing at every band that strove to concentrate against
them. In a few minutes the fight was over, for the Irish broke
up and ran off with disgraceful alacrity, the king, all naked
from his bath, leading the flight Fifteen hundred were slain,
while the English only lost one single sergeant. On hearing
of Roderic s defeat, the Irish in the other three camps dispersed
and went homeward, and the siege was raised (July 1171).

Thus ended a fight which bears a strong similarity to
another sortie made by an English garrison from Dublin, five
hundred years later. Colonel Michael Jones in 1649 was be
leaguered like Earl Richard by a vastly superior host dispersed
in several distant camps. Like the earl, he hazarded a sortie
against one of the hostile corps, and was successful in surprising
and dispersing it. And when Ormond s men had been routed

1 That the archers were mounted seems to follow from the correction of satellites
cquestres" for "arcarii" in the later texts of Giraldus, I. xxiv.

2 Giraldus makes the first two columns led by only twenty and thirty knights
respectively, and says that Raymond rode before instead of after Miles Cpgan.

406 THE ART OF WAR IN THE MIDDLE AGES [1180

at Bagotsrath, the other Irish divisions dispersed and retired
without fighting. 1 The rebels of 1649 were as divided in their
counsels and as chary of giving each other prompt aid as the
levies of 1 171.

The three battles which we have thus set forth give us the three
main tactical devices by which the Normans won their victories
the feigned retreat, the flank attack by horsemen, and the sudden
surprise. After three years of fighting, the Irish were so cowed
that they habitually retired to wood or bog when the invaders
advanced, and never fought save in night surprises or behind im
pregnable stockades and ditches. These defensive tactics handed
over the open country to the conquerors, who forthwith secured
it by erecting castles everywhere, structures against which the
Irish could seldom prevail indeed, a castle, when once completed,
never fell save by treachery. On the other hand, the Anglo-
Normans were almost equally incapable of mastering the woods
and bogs in which their enemies took refuge. Hence came
that unhappy division of the island, destined to last for four
centuries and more, in which the natives held out in their
fastnesses, while the invaders dominated the open land each
levying unending war on the other, yet neither able to get the ad
vantage. The land could make no progress, and in the sixteenth
century .the natives were as barbarous as in the eleventh, while
the invaders had almost sunk to their level, instead of advancing
in civilisation parallel to the English and the other nations of
Western Europe. The wars of Elizabeth s day in Ireland
exhibit the " mere Irish " absolutely unchanged from their
ancestors of the twelfth century : their primitive tactics, their
arms, their plashed woods and wattled stockades are absolutely
the same as those of the days of Strongbow. Except that some
of their chiefs had learned to ride 2 to battle, we see no change.

1 Ormond was caught in bed just as Roderic O Connor was caught in his batTi
by the sortie party.

2 And that as early as the fourteenth century, as is shown by the description of
the Irish by the captive squire in Froissart, xxii. p. 429.

CHAPTER IV

ENGLISH BATTLES OF THE THIRTEENTH CENTURY

Second Battle of Lincoln (1217) Taillebourg (1242)
Lewes (1264) Evesham (1265)

AS we have already had occasion to remark, the wars of
Richard I. and John with Philip Augustus were singularly
unfruitful in battles. Bouvines is the one first-class engagement
in the whole generation ; and though there were English troops
mainly mercenaries fighting at that most decisive field, it
cannot be called an English battle. Salisbury and Hugh de
Boves were only present as the emperor s auxiliaries, and had
little to do with the conduct of the campaign or the marshal
ling of the host for combat. We have therefore dealt with
Bouvines among continental and not among English battles.

It is not till the second battle of Lincoln (May 19, 1217)
that we come upon another field well worthy of notice, were it
only for the strange fact that it was a cavalry fight fought in
the narrow streets of a town perhaps the most abnormal and
curious form of engagement which it is possible to conceive.
The Whitsuntide of 1217 found the barons who had espoused
the cause of Louis of France engaged in the siege of the castle
of Lincoln. They were in possession of the town, but the
castle was denied to them by Nicola de Camville, the castellan s
widow, who maintained the stronghold by the help of a small
garrison under a knight named Geoffrey de Serland.

Lincoln lies on a hill sloping down southward towards the
river Witham. On the high ground lie the castle, at the north
west angle of the town, and the minster, more to the east. The
streets run down to the water, which is crossed by a bridge (then
known as Wigford Bridge) leading to the suburb of St. Peter s-

407

4 o8 THE ART OF WAR IN THE MIDDLE AGES [1217

at-Gowts, beyond the Witham. The besiegers lay within the
walls, and pressed the siege by battering the south and east
sides of the castle with perrieres. They had shaken part of the
curtain, and hoped to see the battlements crumble within a few
days.

The Royalist army mustered at Newark under William
Marshall, the Earl of Pembroke : he had with him the Earls
of Chester, Salisbury, and Derby, and the greater part of the
barons who had remained loyal, as also Fawkes de Breaut^
and the remnant of King John s mercenaries, horse and foot.
Altogether they mustered four hundred and six knights, with
three hundred and seventeen crossbowmen and a considerable
number of foot-sergeants. 1 They marched from Newark north
westward when they heard of the straits to which the castle
was reduced, and slept on the night of the iSth at Torksey and
the neighbouring village of Stow, some nine miles from
Lincoln. From thence they ascended the high ground along
which the Roman road (Ermine Street) runs, and moved
cautiously toward the north front of the city. This route gave
them a chance of communicating with the castle, unless the
enemy should choose to fight at a considerable distance from the
walls. The host was marshalled in four " " battles," the first led
by the Earl of Chester, the second by the Marshal, the third
by the Earl of Salisbury, the fourth by Peter des Roches, Bishop
of Winchester, the most unpopular but the most able of the late
king s foreign favourites. The crossbowmen under Fawkes de
Bre"aute moved a mile in front of the knights. The baggage
with a guard of infantry followed, the same distance in the rear
of the four corps of cavalry. 3

Second Battle of Lincoln, June 19, 1217.

The besiegers of Lincoln received timely warning of the ap
proach of the relieving army, and sent out Saher de Quincey, Earl
of Winchester, and Robert Fitzwalter to reconnoitre the advancing
columns. They soon returned with the report that the Royalists

1 So the Song of William the Marshal, 16264-8. Matthew Paris (p. 1 8) says four
hundred knights and two hundred and fifty crossbowmen, as also " multi servientes
qui vices rnilitum possent pro necessitate implere. "

-Matthew Paris (p. 19) says seven "battles," but the Song of William tfa
Marshal is so clear and full that it would be dangerous to refuse to follow it and to
choose the later authority.

3 Matthew Paris, p. 19.

THE ROYALISTS APPROACH LINCOLN 409

seemed somewhat weaker than themselves, and that it would be
advisable to attack them in the open, far from the city, in order
that they might not be able to communicate with the garrison
of the castle. The estimate was not far wrong, as the besieging
army counted six hundred and eleven knights and a thousand
foot-sergeants, 1 a force decidedly superior to the Marshal s host.
But the Count of Ferche, who commanded the French contingent
in the rebel army, 2 insisted on going forth in person to take a
second view of the enemy, before committing himself to a battle.
Mistaking the distant baggage-guard and its column of sumpter-
beasts and waggons for an integral part of the Royalist army, he
came back with a firm belief that he was largely outnumbered,
and insisted on keeping his men within the walls of the city, and
taking the defensive. 3 This line of tactics seemed to promise
absolute security, since it appeared impossible that the very
modest host of the Earl Marshal would be able to do serious
harm to the rebels, when the latter were covered by the strong
fortifications of Lincoln. The storming of a city or castle by
main force and without a long preparatory leaguer was an almost
unknown thing in thirteenth - century warfare. Accordingly
the barons continued their operations against the castle, and set
their machines to play upon its walls with redoubled energy.
The only precaution which they took against the relieving army
was to tell off detachments to guard the four gates by which
the Marshal might attempt to enter the city, the north gate
which lay immediately opposite him, the east gate and Potter s
gate on the right flank, and the Xcwland gate on the extreme
left between the castle and the river \Vitham. It cannot have
escaped the notice of the commanders of the baronial army
that their tactics allowed free communication between the castle
and the Royalists, and that it was possible for the Marshal to
enter the castle and sally forth into the town by the great gate
in its eastern curtain. But this exit was well guarded by the
detachment told off to operate against the castle, and such a

1 William the Marshal, 16336-9.

5 The chroniclers only preserve the names of three of the French chiefs in the
host, though the French contingent seems to have been strong. These are the
Count of Perche, the Marshal Walter of Nismes (Matt. Paris, p. 20), and the Chatelain
of Arras (Song of Wiltia.ni the Marshal, 16607).

3 Matthew Paris, p. 19. He says that the barons had left many standards with
the baggage-guard, and that their appearance misled the count into taking it for a
reserve corps in the rear of the Royalist line of battle.

410 THE ART OF WAR IN THE MIDDLE AGES [1217

sally on a narrow front appeared to present no very great
danger. Any transference of troops from the relieving army
into the castle must take place under the very eyes of the
defenders, and could be easily provided against by a corre
sponding shifting of their own forces.

When William the Marshal and his host approached Lincoln,
they were somewhat surprised to find that the enemy would
not come out to meet them in the open. Drawing up at a
cautious distance from the city, they proceeded to communicate
with the castle. John Marshall, the earl s nephew, swept round
the north-west corner of the place with a small party, and
entered the castle by its postern gate. He learned that the
garrison were reduced to great straits, and bore back the
message to his uncle. On leaving the postern he was pursued
by a party of rebel knights who issued from the Newland gate
to chase him, but outrode them and reached the main arm} in
safety. 1

The Marshal then resolved to send into the castle Bishop
Peter, who was renowned for his good military eye, that he
might decide whether the proper course of action would be to
throw troops into the castle and sally forth from it, or to attack
the gates and the city. The bishop made a rapid survey of
the place, and fixed his main attention on the point where the
castle joined the north-west front of the town wall. Here there
lay, quite unguarded, and close under the castle, so as to be
swept by its fire, an old blocked-up gate, on which the
barons had set no guard. 2 He bade a party of the garrison
steal out and tear down the stones which closed the gate, so
as to make an opening in this unguarded front. Meanwhile, he

1 Song of William the Marshal , 16438-40.

" Une vielle porte choisi
Qui ert de grand anuquite
Et qui les murs de la cite
Joigniet avec eels del chastel,
Mes el fut anciennement
Close de piere e de cinient.
Quand li evesques ont veiie,

La fist abbatre et trebuchiet
E que J ost veist et seiist
Que seure entree i eiist " (16509-17).

This gate must have been that generally known as Westgate ; it must have been
rendered comparatively useless when the castle- building destroyed the north-west
houses of the town, and was temporarily blocked up.

DE BREAUTE S SORTIE 411

rapidly returned to the Marshal, and advised him to throw part
of his men into the castle and make a sally from it, but to direct
his real attack on the blocked postern, which would soon be
opened again, and on the north gate of the city.

The Marshal therefore sent into the castle Fawkes de
Breaute and all his crossbowmen, who ran to the walls and
opened a fierce fire on the party of the enemy which was
observing the castle gate. Many of the horses of the rebels
were slain, and the whole body thrown into confusion. Fawkes
then sallied out with his troops and made a vigorous attack on
the besiegers, but they were too many for him, and he was
beaten back into the castle with loss. 1 He himself was for a
moment a prisoner in the enemy s hands, but was rescued by a
party which turned back to save him.

While this assault was being delivered from the castle, the
Marshal and the main body of his host had drawn near to the
northern wall of the city, probably somewhat masked from the
rebels view by the houses of the suburb of Newport. 2 Apparently
the attention of the defenders had been so distracted by the
sally of Fawkes de Breaute, that they had not noticed that the
postern in the north-west wall had been broken open. At any
rate, when the Royalists made a simultaneous dash at this entry
and at the north gate, they succeeded in penetrating within the
city at the breach, though not at first at the more obvious and
better-guarded point. 3 A party headed by John Marshall, the
earl s nephew, broke right into the streets, and assailed the
detachment of the rebels who were busied with repulsing the
sally from the castle. They took the enemy s engines in flank
and killed their chief engineer, just as he was placing a stone in
his perriere to cast at Fawkes de Breaute"s men. 1 Having thus

1 Matthew Paris, p. 21.

2 That they were among houses seems to follow from line 16600 of the Song of
William the Marshal, where the earl before charging says

" Attendez mei a ces-t ostal
Tant que j ai mon helme pris."

3 The assailants (line 16657)

"Entra sis filz en la cite

Pur la breque o plante des suens."

But from Matthew Paris we gather that they succeeded in forcing the north gate
later on, as he says, "Januis tandem civitatis licet cum difficultate confractis,
villam ingrediens," etc. {p. 21). Probably this was done after the attention of the
rebels was distracted by the successful entry at the blocked gate.

4 Line 16633.

THE ART OF WAR IN THE MIDDLE AGES [1217

won an entry into the place, the earl pushed his men through
the breach into the streets as fast as he was able. They could
not advance with ease, for the barons had rallied and massed
their forces against the assailants, who were obliged to advance
on a narrow front down the tortuous lanes of the town, and
could not deploy. A fierce jousting took place in all the north
western streets of Lincoln, and it was only by very vigorous
righting that the Royalists were able to win their way forward.
Their foot-soldiery slipped in among them, shooting or ham
stringing the horses of the French and the rebels.

At last the whole of that part of the city which lay near
the castle was occupied. The enemy fell back, part along the
high ground towards the cathedral and the north-east quarter
of the place, part down the broad street leading to the bridge at
Wigford and the south gate. In the open space before the
minster the Count of Perche rallied the best knights of the
baronial army, and made head for some time against the
Marshal and the main column of the Royalists. At last his men
gave way, and he himself was surrounded ; he was offered
quarter, but "would not yield to any traitor Englishmen," 1 and
was slain by a thrust which pierced the eyehole of his helm.
After his fall the rebels lost heart and rapidly gave ground,
some flying by the east gate, others southward towards the
river and the bridge. At both exits there was soon a crowd
massed in hopeless confusion, the passages being too narrow to
allow so many fugitives to pass out at once. The south gate
had a swing door, which closed automatically after each passer
by pushed it open ; 2 the east gate is said to have been jammed
on a frantic cow which got mixed with the horsemen. 3 Hence
the pursuers were able to make prisoners of an enormous
proportion of the rebel knights and barons. About four hundred
in all out of the six hundred and eleven who had engaged in the
battle were captured. They included three earls, Saher de Quincey
of Winchester, Henry de Bohun of Hereford, and Gilbert de
Gand of Lincoln. Among the other captives were several of the
twenty-five signatories of the Great Charter. The slaughter, on

1 "Juramento horribili affirmavit quod se Anglico alicui nequaquam redderet
qui propri regis proditores fuerunt" (Matt. Paris, pp. 21, 22).

- Matt. Paris, p. 22.

3 This was the east gate ; the poem of William the Marshal describes it as "the
one that leads" " dreit a 1 Hospital," i.e. St. Giles Hospital, founded by Remigius,
outside the east gate (line 16943).

iz I?] THE MARSHAL VICTORIOUS 413

the other hand, had been small, though the wounds were many.
The victors lost but one knight, a certain Reginald le Croc ;
of the vanquished, only the Count of Perche and one other
knight are recorded as slain, though many of the foot-soldiery
on both sides perished,

It must be confessed that the details of the "Fair of
Lincoln," as the battle was called in jest, do not give us a
very high idea of the tactical accomplishments of cither side.
The arrangements made by the rebels were ill conceived and
carelessly carried out. Their neglect to watch the blocked
gate is most extraordinary, and, even when it was forced, they
might have had a good chance of victory if they had barricaded
the streets and fought on foot, instead of endeavouring to
expel the Royalists by cavalry charges.

To the victors the only praise that we can give is that they
knew how to utilise a false attack in order to distract attention
from the real one. Bishop Peter must apparently take more
credit for the plan adopted than the Marshal ; the poem written
in praise of the latter ascribes the idea to the Churchman, and
only the execution of it to the earl a piece of evidence
conclusive as to the attribution of the design, for William s
encomiast would certainly have claimed the glory for his
hero had he been able to do so. The details of the fighting
after the breach was once forced show nothing but hard blows ;
we have no evidence that the crossbowmen were used in the
street fighting, as they well might have been, or that the enemy
were evicted by flanking movements by side streets. All
apparently was done by vigorous jousting down the main
thoroughfares and in the open space by the minster.

Nearly fifty years elapsed before Englishmen fought another
battle on English soil, and we shall see, when we pass on to
investigate Lewes and Evesham, that the art of war had moved
on considerably in the interval. But there is no material for us
to use in filling up the gap save the insignificant battle of
Taillcbourg, where the imbecile Henry III. allowed himself to
be defeated by Louis IX., a general whose strategy we have
learned not to admire in studying the campaign of Mansourah. 1
On Taillebourg we need not waste much attention. Stated
shortly, the gist of the battle was as follows :

Henry, with sixteen hundred knights, seven hundred picked

1 See p. 339.

4H THE ART OF WAR IN THE MIDDLE AGES [1242

crossbowmen, and the general levy of the towns of Guienne, lay
on one bank of the Charente near Taillebourg. The army was
almost wholly composed of his continental vassals ; only eighty
English knights were present. Louis, with a much larger force,
appeared on the other side : the river was broad and swift, and
there appeared to be no means of crossing save the bridge, where
Henry set a strong guard. Relying on the safety of his position,
he kept no proper watch on the enemy. St. Louis determined
to risk an attempt to force the passage, and prepared for that
purpose a number of large boats. He then vigorously attacked
the bridge, and at the same time threw across a body of
crossbowmen, dismounted knights, and sergeants by means of his
vessels. The guards of the bridge, fearing to be attacked behind
by the newcomers, gave ground, and so allowed the main body of
the French to evict them from the passage they were sent to guard.
When King Henry saw the bridge lost, he did not make any
attempt to fall on the small part of the French army which had
crossed, but drew off and sent his brother Richard to ask for a
truce. It was granted, and under cover of it he withdrew at night
fall with shameful haste, abandoning his camp and baggage.

A capable commander would have had his army in order,
would not have been caught off his guard, and would have fallen
on the French van when it had passed the bridge, and over
whelmed it before the main body could come to its aid. Such
were the tactics employed in a similar case by Wallace at the
battle of Stirling Bridge. 1 But Henry was the most helpless
and imbecile of leaders, and threw away his chances in the most
faint-hearted manner. At the moment that he sent to ask for a
truce, the number of French who were over the river did not
amount to a tithe of his own army, yet he parleyed instead of
charging. 2 If Louis had not listened to his demand, he would
probably have given the signal for flight at once, and would
have got off in even worse plight than was actually the case.

Lewes and Evesham show a distinct advance in the art of
war, which we may fairly set down to the influence of Simon
de Montfort, who, though not a general of the first class, had at

i See p. 563.

3 Joinville says that there were "not one hundred part as many" French troops
over the bridge as Henry mustered. Matthew Paris conceals the facts of the dis
graceful skirmish in a way not creditable to his veracity, when we consider what a
capable writer he was and how fully he tells the rest of the campaign.

1*64] KING HENRY TAKES NORTHAMPTON 415

least a quick eye and a wide experience. He had been brought
up on the traditions of Muret and the rest of his father s victories.
He had himself seen several campaigns both on the Continent
and in the East. Though not an innovator, he was a capable
exponent of the best methods of his own generation. But it is
only as a tactician that he shines : strategy is nowhere apparent
in his campaigning, and in 1265 he was hopelessly outgeneralled
by the young Prince Edward. We shall see that he relied, like
all his predecessors, on the force of cavalry ; the infantry count for
nothing in his battles. He triumphed, when opposed by the
incapable Henry ill., because he possessed decision, rapidity of
movement, and a cool head. But it was only in the fight of
Lewes that his abilities shone out : in the preceding campaign
he does not show to much more advantage than his incompetent
opponent.

Far otherwise is it with the victor in the campaign of Evesham.
Here we shall see Edward showing a real mastery of strategy as
opposed to mere tactics. When we study his operations in 1265,
we shall be quite prepared to find him, thirty years later, presid
ing at the inauguration of a new epoch in war at the bloody
field of Falkirk. But in his youth he was still, as regards
tactics, employing the old methods which he had learned from
Montfort as his teacher.

Battle of Lewes, May 14, 1264.

Down to the day of battle the operations which led up to the
fight of Lewes show all the characteristic incoherence and in
consequence of a mediaeval campaign, and do no credit to either
of the parties concerned. King Henry had raised a considerable
army in the Midlands, while the baronial party had made itself
strong in London, but had also seized and garrisoned the im
portant towns of Northampton, Leicester, and Nottingham.
The king resolved to subdue the three midland centres of revolt
before undertaking any further operations. Northampton fell
with unexpected ease, owing to the treachery of the monks of
St Andrew s Priory, who admitted the royal troops through a
passage into their garden. This was a severe blow to the
barons, for some of their chief leaders were made prisoners,
including Simon the Younger, the second son of the great Earl
Simon, his kinsman Peter de Montfort, and fifteen barons and
bannerets more (April 5, 1264).

4i 6 THE ART OF WAR IN THE MIDDLE AGES [1264

A few days later (April 11) Leicester was sacked, and
Nottingham, the spirit of whose defenders was shaken by the
disaster at Northampton, surrendered at the king s summons
(April 13). Having thus cleared the eastern Midlands of
enemies, Henry should at once have marched on London with
his victorious army. The fall of the capital would have settled
the fate of the war, and, in spite of all the efforts of De
Montfort, the spirits of his followers were sinking low. Simon
himself had started to relieve Northampton, and had reached St.
Albans when the news of disaster reached him. He immediately
fell back and prepared to defend the city. Finding, however,
that the king showed no signs of striking at London, and had
marched northward, the earl resolved to make a rapid stroke at
Rochester, the one Royalist stronghold in the neighbourhood of
the capital. He stormed the bridge, penetrated into the town,
and drove the garrison within the walls of the castle (April 18).
He captured its outworks, but the massive strength of Gundulfs
Norman keep was too much for such siege appliances as the earl
could employ. The garrison, under John de Warrenne, the Earl
of Surrey, held their own without difficulty.

Meanwhile, the king had received news of the siege, and
left the Midlands. He should undoubtedly have risked all other
objects, and thrown himself upon London. The mere news of
his having turned southward was enough to draw Simon and
his host back from Rochester to defend the capital (April 26).
The earl merely left a few hundred men stockaded in front of
the gate of the keep to hold the garrison in check a thing easily
done, because the narrowness of the exits of a Norman castle
rendered sallies very difficult.

But, instead of striking at London, King Henry merely sent
forward his son, Prince Edward, with a small cavalry force, to see
if the city was in a state of defence, 1 and then committed the
extraordinary error of coasting round it by a vast circular march.
Returning down the Watling Street, he struck off it by St. Albans,
passed the Thames at Kingston, hastily rushed across Surrey
by way of Croydon, and arrived at Rochester on April 28.
The blockading force was easily driven off, and the few prisoners
made were cruelly mutilated.

This huge flank march had no merit but its swiftness.
Prince Edward and the mounted part of the royal army

1 See Annals of Dunstable.

1264] HENRY III. MARCHES ON LEWES 417

marched from Nottingham to Rochester a hundred and fifty
miles in five days, 1 and the infantry were not very far behind.
The pace, however, had told heavily on the Royalists : many of
the horses were ruined when the prince arrived at Rochester, and
the foot-soldiery had left thousands of stragglers on the way.

As it turned out, the king s hurried movement had no
adequate object. Having relieved Rochester, he might again
have turned towards London, though with less advantage, since
he was now separated from it by the broad reaches of the Lower
Thames. But this did not enter into his plan of operations :
he marched instead against Tunbridge, a great castle of the
Earl of Gloucester, and when it fell with unexpected ease (May
i) moved still farther from London, with the object of over
awing the coast-towns. 2 But the barons of the Cinque Ports
had sent their fleet and their armed force to sea, and Henry
obtained nothing but a few hostages from Winchelsea and
Romney. His next move was still more inexplicable he
pushed westward between the Weald and the sea, and marched
by Battle and Hurstmonceaux to Lewes. No object seems to
have been served by this turn, save that of placing himself in
the midst of the estates of his brother-in-law and firm supporter,
De Warrenne. It had the disadvantage of putting the almost
trackless forest of the Weald between himself and London, and
of causing his army much discomfort as they threaded their
way through the wood-tracks for the men of Kent and Sussex
cut off his stragglers and plundered his baggage, and a detach
ment of Welsh archers, whom Montfort had sent forward from
London, are said to have molested the rear of the host. The
king s object is impossible to fathom, more especially as we are
told that he feared that his] enemies would strike at Tunbridge
when he had marched off, and therefore garrisoned that castle
with a very large force ; no less than twenty bannerets and
many of his foreign men-at-arms are said to have been left
there.

De Montfort and the barons, however, had no intention of
wasting their time in sieges when they could strike at the main
objective, the king s army. Having collected every available
man, and armed a great body of the citizens of London, they
marched across Surrey, plunged into the paths of the Weald,
and did not halt till they had reached Fletching, a village and

1 Wykes. 1264, 4. 3 Knighton. * Wykes. 1264, 5.

418 THE ART OF WAR IN THE MIDDLE AGES [1264

clearing nine miles north of Lewes (May 6th-ioth). From
thence they addressed proposals for peace to the king, dated
with prudent vagueness " in bosco juxta Lewes." They must
have known well enough that Henry would refuse them, after
his late successes at Northampton and Tunbridge, and on re
ceiving his angry reply prepared for instant action. Although
he had the smaller force, Simon was resolved to take the
initiative, trusting to his own skill, the greater enthusiasm of
his supporters, and the king s well-tried incapacity in war.

The town and castle of Lewes lie at a point where the line
of the South Downs is cut through by the river Ouse. To the
east of the place the steep sides of Mount Caburn rise directly
above the water, hardly leaving room for the suburb of Cliffe
along the river-bank. To the west of the Ouse there lies a
mile and a half of gently-undulating ground before the ascent
of the Downs begins. In this comparatively level spot lies the
town of Lewes, flanked to the north by De Warren ne s castle
on its lofty mound, to the south by the great Cluniac Priory of
St. Pancras, including within its precinct-wall some twenty acres
of ground. The Ouse in the thirteenth century was still a
tidal river as far north as Lewes, and at high water the south
wall of the priory and the southern houses of the town looked
out on a stretch of mingled pools and mud-banks which formed
an impassable obstacle.

North and east, therefore, Lewes is protected by the river,
and on the south by this tidal marsh, but to the west it had
no protection but the castle and the town wall. If an enemy
approached from that side, the king s army would have either
to stand a siege, or to retire behind the Ouse, or to come out
and fight at the foot of the hills.

On this side the main range of the Downs descends rather
gently towards the river, not with a uniform slope, but in three
spurs separated by slight valleys. The road from Fletching to
Lewes passes over the easternmost of these spurs by the hamlet
of Offham, and by this path would have been the shortest
approach from the barons camp. But Simon had wisely re
solved not to come down a road cramped between the hills
and the river. Marching at early dawn on May 14, he
turned off the road north of the Downs, and ascended them at
a hollow slope called the Combe, four miles from Lewes. 1 This

Blaaw and Prothero seem undoubtedly right on this point of topography.

1264] LEWES: SIMON S ARRAY 419

he was able to do quite unmolested, as King Henry had made
no proper arrangements for keeping an eye on his adversaries.
He had not sent out any reconnaissance towards Fletching, and
the sole precaution that he had taken was to place on the
previous day a small party on a high point of the Downs
to keep watch. No measures had been taken to relieve the
watchers on the I3th, and, being tired and hungry, they
slipped back into Lewes to rest themselves, leaving a single
man on guard. This individual lay down under a gorse-bush,
and was caught sound asleep by the first of De Montfort s men
who climbed the slope. Thus the earl was able to put his whole
force in array on the ridge of the Downs before the Royalists
had the least idea that he was within two miles of them. Simon
had spent the previous day and night in distributing his men into
corps, and assigning the position of each on the march and in
battle-line a task which, as the chroniclers tell us, no other
man in his raw army was competent to discharge. 1 Now he
had full leisure to see that his exact intentions were carried
out, and to settle the smallest details of the marshalling.

Owing to the disasters at Northampton and Nottingham,
the barons army was much smaller than might have been
raised by the full levy of the party, for many of their most
important leaders were prisoners in the king s hands. 2 The
estimate of forty thousand men given by several chroniclers
as Simon s force is one of the hopeless and habitual exaggera
tions of the mediaeval scribe. But, small though the army was,
it was divided not into the usual three battles, but into four.
There is no doubt that the fourth, which was led by the earl
himself, was a reserve corps placed behind the others, but none
of the chroniclers expressly state this fact. It can be inferred,
without any danger of doubt, from the circumstance that the
three first-named battles of Simon s army each engaged with
one of the three bodies which formed the king s left, right, and
centre, and that the earl s division came later into the fight
than the other three.

As arrayed on the Downs before descending to battle, the

1 Rishanger, p. 31.

" Including Simon de Montfort the Younger, Peter de Montfort and his sons
Peter and William, Adam of Newmarch, one of the greatest of the barons of the
Welsh border, Baldwin Wake, William de Furnival, all captured at Northampton,
\ViIliam Bardolf, captured at Nottingham, and the young Earl of Derby who had
been taken in his own castle of Tutbury.

420 THE ART OF WAR IN THE MIDDLE AGES [1264

baronial army was drawn up as follows : On the right or
southernmost wing were Humphrey de Bohun, the eldest son of
the Earl of Hereford, John de Burgh (the grandson of the
great Justiciar, Hubert de Burgh), and De Montfort s two sons,
Henry and Guy. In the centre was Gilbert de Clare, the
young Earl of Gloucester, with John Fitz-John and William
de Montchensy, two of the most vigorous members of the
baronial party. The third or northern wing was composed of
the numerous infantry of the Londoners, and of a body of
knights commanded by Nicholas de Segrave, Henry de
Hastings, John Giffard, and Hervey of Borham. The earl s
reserve corps lay behind the centre ; the horsemen in it con
sisted of his own personal retainers, the foot were probably
Londoners, as they were commanded by Thomas of Pevelsdon,
an alderman of the city, who had always been one of Simon s
most sturdy adherents.

Deployed in this order, and probably with the knights of
each division in front and the infantry behind, Simon s forces
halted just as the bell-tower of Lewes Priory came in sight, to
engage for a moment in prayer, after a short address from their
leaders. Scattered over the slope of the Downs were small
parties of the grooms of the Royalists, grazing their lords
horses, for forage had failed in Lewes. They caught sight of
the baronial host as it came down the hill, and fled back to the
town to rouse their masters. Simon s host followed close at
their heels, leaving on the upper ridge of the hill such small
impedimenta as they had brought with them, the chief of which
was the earl s chariot, 1 to which he had bound his great banner,
after the manner of the Milanese at Legnano or the Yorkshire-
men at our own Battle of the Standard. Inside the carriage
were three (or four) citizens of London whom Simon had
arrested for opposing him, and was determined to keep in safe
custody. The banner and baggage were left in charge of a
guard of infantry under William le Blound, one of the signatories
of the agreement for arbitration which had ended so unhappily
at Amiens. 2

1 Simon had broken his leg in the previous year, and was forced to use this
carriage for many months.

2 Of the twenty-four laymen who signed for the barons party in 1263, the
following were at Lewes : Earl Simon, Ralph Basset, William le Blound, Humphrey
de Bohun, John de Burgh, Hugh Despenser, John Fitz-John, Henry de Hastings,
Henry de Montfort, William de Montchensy, Nicholas de Segrave, Robert de Ros,

PLATE XIII.

LEWES. May 14.1264-
EarlSimon sArnry. Ki m* Henrys Army

a Horse ^m&lt; r C3 Horse C^DFoot

A-CuyandHenry DEarlSimon. F KineHenry

dtMontfort
B Gloucesier E.TheChanor G. Richard of

C u Cornwall
Segpave and the Londoners n Prmce ELd*3rd

The EVE SHAM
CAMPAIGN

1265.

1264] LEWES: THE KING S ARRAY 421

The king and his followers had barely mounted and armed
and issued from the town of Lewes, when they saw the baronial
army coming down upon them. But they had just time to
form up in three " battles " before the conflict began. Knighton
informs us that the king had originally organised his troops
into four corps (like Earl Simon), but that the whole of the
fourth division had been left behind to garrison Tunbridge, so
that the Royalists had no reserve. 1 Perhaps Henry might
have told off other troops to play that part had he been granted
time to think. But he was completely taken by surprise, and
considered himself lucky to be able to form any battle-order at
all. His right division was led by his heir, Prince Edward, who
was accompanied by his foreign half-uncles, William de Valence
and Guy de Lusignan, as also by the Earl of Warrenne and
Hugh Bigot the Justiciar. The centre was under the command
of Richard of Cornwall, King of the Romans, brother to King
Henry ; with him was his son Edmund, and three great Anglo-
Scottish barons, Robert de Bruce, John Baliol, and John
Comyn, who had come to join the Royalists with a large body
of light-armed infantry from north of Tweed. In this division
also were John Fitz-Alan and Henry de Percy. The left or
southern wing was commanded by the King of England himself
under his dragon-standard : ~ in his company was the Earl of
Hereford, whose eldest son was serving in the very division of
the baronial host which was about to bear down upon his
father. All accounts agree that the Royalists outnumbered the
forces of Simon, especially in their array of fully-armed knights,
though we cannot believe the exaggerated statement that the
king had fifteen hundred men-at-arms on barded horses
(dextrarii coperti) and the barons only six hundred.

Geoffrey de Lucy, John de Vesey, Richard de Vipont fourteen in all. Simon
junior de Montfort, Peter de Montfort, Adam of Newmarch, Baldwin Wake,
William Marshall, had been captured at Northampton ; William Bardolf at
Nottingham. Richard de Grey was holding Dover Castle. Nothing is known as to
the whereabouts of Walter de Colville and Robert de Toeny.

1 H. Knighton, p. 247 of Rolls Series edition.

There are some difficulties in the array of the Royalists, as in that of the baronial
host. On the whole I am compelled to conclude that Earl Richard led the centre,
and the king the southern wing. I imagine that the position of the king on the left
must have been due merely to the hurry and haste of the muster. Being encamped in
the priory, he drew up in front of it. For by all mediaeval military etiquette he
should have led the right or centre, and not taken the post of least honour. But there
was no time to rearrange the host, and each body fell into line as best it could.

422 THE ART OF WAR IN THE MIDDLE AGES [1264

When the Royalists had got into order, the castle lay
behind Prince Edward, the exit from the town of Lewes behind
Richard of Cornwall, and the priory at the back of the king s
own wing. Before they had advanced more than a few hundred
yards from the town, the baronial army charged down upon
them. There seems to have been little or no preliminary
skirmishing, the battle commencing with a sharp shock all
along the line, starting from the northern wings of each host,
who met the first. This came from the fact that the Londoners
on the baronial left had a shorter space to cover before contact
took place : some of the chroniclers observe that they were so
much in advance that the Royalists supposed that they were
trying to outflank the castle and the division of Prince Edward.
There is at any rate no doubt that the first clash of arms started
on this wing. It was unfavourable to the baronial party : the
knights who followed Segrave, Hastings, and GifTard were
broken by the furious charge of the prince. GifTard was taken
prisoner ; Hastings turned his rein too soon for his own good
repute ; l their horsemen were flung back on the Londoners, and
threw them into woeful disorder even before Edward s knights
dashed into the wavering mass. A moment later the whole
left wing of Simon s host broke up and dispersed, the knights
flying northward between the river and the Downs, the infantry
northwestward up the steep slope, where they thought that the
Royalist horsemen would find it hard to follow. Prince Edward
had an old grievance to settle against the Londoners, for the
insults which they had heaped on his mother in the preceding
year. He urged the pursuit furiously, and forgot entirely the
battle that was raging behind him in the centre and left of his
father s army. The fugitives suffered fearfully from his fierce
chase : sixty horsemen are said to have perished in striving to
ford the Ouse ; hundreds of the men of London were cut down
as they fled along the slopes and then towards Offham and the
woods behind. The prince did not stay his hand till he was
three miles from the battlefield, and quite out of sight of Lewes,
which was hidden from him by the corner of the Downs. Then,
at last rallying his men, he remounted the slope to return to his
father ; but on his way he caught sight of Earl Simon s chariot
and its great banner, standing isolated at the head of the slope,

1 " Paene primus H. de Hastings, audaciae fonnidinem anteponens, e proelio
fugit" (Wykes. 1264, 6).

1264] LEWES: SIMON VICTORIOUS 423

under the protection of Le Blound and the baggage-guard. The
Royalists jumped to the conclusion that Simon was still in his
chariot, not knowing that his broken leg was long since healed,
and that he was fighting hard on his horse in the valley below.
They therefore wheeled aside and furiously attacked the baggage-
guard. Le Blound and his men made a gallant resistance, but
were at last overwhelmed and cut down. Then shouting, " Come
out, Simon, thou devil," 1 the prince s knights broke open the
chariot and hewed to pieces the unhappy hostages who were
confined in it, before they could explain that they were the earl s
foes and not his friends. 2 Disappointed of their prey, Prince
Edward and his men at last set forth to return to their main body.
But meanwhile complete victory had crowned the arms of
Earl Simon in the southern part of the field. The Earl of
Gloucester in the baronial centre had after severe fighting
broken the line of Richard of Cornwall s division, captured most
of its leaders, including Percy, Baliol, Comyn, and Bruce, and
forced Richard himself to take refuge with a few followers in a
windmill, where he defended himself for a space while the tide
of battle rolled past him towards the town. It is probable that
Earl Simon threw his reserve into action against the northern
flank of the king s own corps, when he saw that the line was
giving way : at any rate, the Royalist left broke up soon after
the centre had failed. The king s horse was killed under him,
but he was dragged off by his household and carried into the
priory, where all who could followed him. But the greater
part of his centre and left wing had been thrust southward by
the successful advance of the barons, and found themselves with
the marshy ground, half covered by water at the full tide,
behind them. Some tried to escape by swimming over, but
the mud sucked them in, and next day scores were found at
the ebb, drowned in their saddles, with their drowned horses
still between their legs, lodged fast in the slime. 3 Others slipped

1 Chron. de Mailros, 1264, I.

2 Some of the Royalist chroniclers call the chariot a " vas dolositatis," and say
that Simon hung his banner on it and placed it on the height specially to distract the
enemy from the main battle. This is most improbable : he would certainly not have
exposed to certain death Le Blound, one of his most trusted followers, and the whole
affair was (no doubt) a mere chance.

3 Chronicle of Lanercost. This authority has some graphic touches given on the
authority of an eye-witness, but is mostly vague and erroneous ; e.g. it says that the
barons formed only three battles and that one of them was led by Hugh le Despenser.

424 THE ART OF WAR IN THE MIDDLE AGES [1264

through the streets of Lewes and got over the bridge ; a good
many took refuge with the king in the priory ; a certain number
were slain, but the majority laid down their arms and were
granted quarter by the victorious barons. These prisoners were
soon joined by King Richard, who, after being blockaded in his
mill for some time, and much scoffed at by his besiegers, had to
come out and surrender himself to a young knight named Sir
John Beavs.

While the barons were battering at the castle gate, and
shooting arrows tipped with burning tow against the priory to
set it on fire, Prince Edward and the victorious Royalist right
wing came into sight on the slopes of the Downs. They rode
hastily on to the field at about two o clock in the afternoon, and
the prince resolved to recommence the fight. But when the
baronial host came swarming out of the town against them,
the large majority of Edward s followers lost heart : the t\vo
Lusignans, Earl Warrenne, and Bigot the Justiciar, with five
hundred knights at their back, turned their reins and rode off.
The prince himself, with a few faithful followers, charged and
cut his way as far as the priory, which he entered and so was
able to join his father. But it was clear by nightfall that they
would be unable to make a long defence, and with great wisdom
Henry and his son sent to ask for peace from the barons. Thus
came about the celebrated " Mise of Lewes," by which the king
laid down his arms, gave up his son as hostage, and agreed to
abide by terms to be settled by arbitration.

The battle had not been so bloody as many mediaeval fights:
the estimate of the losses runs from twenty-seven hundred to
four thousand, the better authorities inclining to the smaller
figure. The captives were far more numerous than the slain :
among the latter are named only two men of importance on
each side ; on that of the king, William de \Vilton was slain, and
Fulk Fitzwarren drowned in the marsh : the barons had to
lament a Kentish banneret named Ralph Heringot, and William
le Blound, the commander of the baggage-guard.

It will be observed from the above narration that Lewes was
essentially a cavalry battle : the infantry seem to have had little
or no influence on its fate ; we only hear of them as suffering,
not as inflicting losses. It is especially curious that we have
no mention whatever of the employment of archery on either
side. One chronicler praises the slingers in the baronial

1264] LEWES : A CAVALRY BATTLE 425

army, another mentions crossbowmen, but of archery there is
no word, though the Assize of Arms of 1252 had named the bow
as the yeoman s special weapon. In the whole campaign we
only once hear of the use of that arm when the king on
his march to Lewes was molested in the woods by Simon s
Welsh bowmen, and drove them off with some loss. It is
obvious that the supremacy of cavalry was still well-nigh un
checked, and that the proper use of infantry armed with missile
weapons was not yet understood.

The main interest of the fight is