Full text of "A history of the art of war; the Middle Ages from the fourth to the fourteenth century"

"b 



'^ A HISTORY OF 

THE ART OF WAR 



THE MIDDLE AGES 

FROM THE FOURTH TO THE 

FOURTEENTH CENTURY 



BY 

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

FELLOW OF ALL SOULS COLLEGE, OXFORD 



WITH MAPS, PLANS, AND ILLUSTRATIONS 



METHUEN & CO. 

36 ESSEX STREET, W.C 

LONDON 

1898 




^^O Y5IOT8IH A 

^r / ;io T>IA Jim 



u 

Ob 



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 



vi 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 Rerum Gallicarum et 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 Cre9y. 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. Cregy, 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 of 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 the ^^(72fM//-— Indecisive Nature of the Old 
English Wars ........ 63-72 



TABLE OF CONTENTS 



BOOK III 



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"^^ 

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 and 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 Burks 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 Lechfeld (955)— The Magyars 
turned back ........ i 16-125 

Chapter V.— Arms and Armour (800-iioc) 

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 » 



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 

Chapt;er 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 L 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 



Pages 



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 (i 097-1 102) 

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 (iioi and 
1102)— Battle of Jaffa ( 1 102) ...... 268-293 

Chapter IV.— The Tactics of the C^vsAB^s—contijiued 
II. — The Later Battles (i 1 19-1 192) 

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 ( 1 191) — Combat of Jaffa (i 192) . . . 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. — The Fyrd and the Feudal Host — William the Conqueror and 
Knight-service — The Vehis Feoffamentuvi — The Cartae Baronum — 
Scutage— Status of the Knight, and its Changes— The Rise of Mercen- 
aries — The Braban9ons and Crossbowmen .... 357-369 

B. — The 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 . 369-377 



TABLE OF CONTENTS xiii 



Chapter III.— English Battles and their Tactics 
(1100-1200) 



Pages 



attle of Tenchebrai (1106) — Battle of Bremule (1119) — Combat of Bourg 
Theroulde (i 124)— Battle of Northallerton (i 138)— 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 ( 1 1 7 1 )— Surprise of Castle Knock ( 1 1 7 1 )— Characteristics 
of Irish War ........ 378-406 

Chapter IV.— English Battles and their Tactics 
(1200-1272) 

second Battle of Lincoln (121 7) — Combat of Taillebourg (1242) — Battle of 
Lewes (1264) — Campaign of 1265 and Battle of Evesham — Comparison 
of the Merits of Simon de Montfort and Edward i. . . . 407-435 

Chapter V. — Continental Battles 
(1100-1300) 

The Types of Battle and their Variety— Battle of Thielt (i 128)— Battle of 
Legnano (1176) — Battle of Steppes (1213) — Battle of Muret (1213), the 
Greatest Triumph of Unaided Cavalry in the Epoch — Campaign of 12 14 : 
King John's Strategy: Causes of its Failure — Battle of Bouvines (1214) 
— Battle of Benevento (1266) — Battle of Tagliacozro (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 
(1100-1300) 

Scarcity of Stone Fortifications before the Eleventh Century — The Early 
Stockaded Mounds, Burhs and Mottes — 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 W^est — 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 
— Mining: 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) 



Pages 



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 l. — 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 ii. — "King 
Robert's Testament "......- 557-580 

Chapter I L— England and Scotland (1328-1 333) — First Com- 
bination OF Archery and Dismounted Cavalry 

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



BOOK VIII 

THE LONGBOW BEYOND THE SEAS 
Chapter L— The Armies of Edward iil 

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 III. 591-596 

Chapter II.— The Longbow in France: Creqy 

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

Chapter III.— Poictiers, Cocherel, and Auray 

The Effects of Cregy : 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.— Navarette and Aljubarotta 

Arms and Tactics of the Spaniards : "the Genetours" — The Black Prince 
invades Castile and outgenerals Don Henry of Trastamara — Triumph 
of the Archers at Navarette (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 



I. Plans of Daras and Taginae 


Tofacepage 28 


II. Frankish Warriors 


36 


III. The "Themes" of the Eastern Empire in 68c 




and 900 . . . . . 


180 


IV. Byzantine Soldiery . . . . 


186 


V. Byzantine Soldiery 


188 


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, Bremftle, 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 11. . 


510 


XVIII. Seals of Richard i. and Henry III. 


512 


XIX. Thirteenth-century Armour 


514 


XX. Thirteenth-century Armour 


S16 


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 

THE TRANSITION FROM ROMAN TO MEDIEVAL 
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 begar. 
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 which 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 own 
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 unwarlike 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 hold 
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 day 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. For 
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 
provinciae inermes 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 Quadi 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 lUyrian 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 them. 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 da)^ 
of Cannae. It seemed for a moment as if the empire was fated 
to be cut in twain, or even as if some earlier i\laric 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 
lUyrians, 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 ar^ 
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.^ All sorts of cohorts and 
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 

^ 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 cf 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 Gains 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 principal e subsidium} 

But a tendency to increase the proportion of cavalry and light 
infantry, and 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 
filled : " the large majority of recruits insist on enlisting among 
the auxiliaries, where the discipline is less severe, where the work 

^ Vegetius, i. § 2. 



oo] INSTITUTION OF A CENTRAL RESERVE 9 

3 lighter, and where the rewards of good service come quicker 

md are bestowed with a more bountiful hand."^ 

In the Roman army as it was reorganised by Diocletian the 

egionary infantry no longer formed, as of old, the wholly pre- 

Donderant part of the foot-soldiery of the empire, in spite of 
:he 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.^ 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-coniitatenses, 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 

^ Vegetius, ii. § 3. 

- 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. 
Flavia, Valeria, Jovia, Herculea) in the regimental name. See Mommsen, HermeSy 
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 
auxilia in the frontier garrisons, over and above the legions. 



lo THE ART OF WAR IN THE MIDDLE AGES [30c 

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 Falatini 
were at their creation we cannot say ; but by the end of the 
century they appear in the Notitia as a very considerable body, 
comprising twenty-four " vexillations " of horse (regiments of 
five hundred each), 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 scholae 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,^ 
" 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 among 

1 Hermes, 1889. 



o] GROWING IMPORTANCE OF CAVALRY ii 



hem than the born subjects of the empire. The old legionary 

avalry wholly disappears/ and the commands of horse and 

oot are entirely separated. Yet under Constantine and his 

mmediate successors the infantry still remained the more impor- 

;ant arm, though the cavalry was continually growing in relative 

mportance. When we read the pages of Ammianus Marcellinus, 

Are still feel that the Roman armies whose campaigns he relates 

ire the legitimate successors of the legions of Tiberius and Trajan, 

:hough the names of the corps and the titles of the officers are 

50 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 cavalr}^ 

were routed and driven off the field, but the foot-soldiery, though 

their flank was uncovered, formed the testiido^ 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 im_portance. 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 magister peditum and a magister equitwn. 
By the time of the drawing up of the Notitia^ 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 

^ Apparently under Constantine, as tliere are faint traces of it under Diocletian. 



12 THE ART OF WAR IN THE MIDDLE AGES [350 

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,"^ 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 {scj^amasax), as well as the long cutting sword 
{spatha)^ and among some races the di^d^dXy 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 framea 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 
inoj^ale 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 

^ See Tacitus, Annals, ii. 14. 



J78] THE BATTLE OF ADRIANOPLE 13 

:entury 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,^ 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." ^ 

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 

^ 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 [378 

rode off without another effort, followed in disorder by such oi 
the infantry corps on that side of the field as were not toe 
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 w^ere 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 da}' 
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. 



jSo] THEODOSIUS SUBSIDISES THE GOTHS 15 

The battle of Adrianople had completely wrecked the army 
Df the Eastern Empire : Valens had stripped the Persian 
frontier and the whole of Asia to draw together the great host 
tvhich perished with him. His successor Theodosius, on whom 
devolved the task of reorganisation, had to restore the entire 
military system of his realm.^ 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 c>( 
the fifth century. After concluding a peace with the Goths so 
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 : ^ 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 annonae 
foederaticae, which they distributed among their horsemen. 
Thus began the ruinous experiment of trusting the safety of the 
empire to the Foederati, as the Gothic war-bands were now 
called : ^ 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 

^ 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. We find, for example, of Sarmatian horse only Ala vii. 
surviving, Ala I. Armeniorum is missing, and eqttites tertii Parthiz, 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. 

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



i6 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,^ 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 races 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 ofificers 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 

^ 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. 



, 39o] 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.^ 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 Mt/i/ari, is 
preserved a picture of the state of the Imperial army in the 
Western provinces, painted probably in the time of Valentinian 
U:, 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, w^hen 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." 

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

2 I am inclined to hold that the jDe Re Alilitari belongs to the time of 
Valentinian ii., and not, as many good authorities think, to that of Valentinian ill. 
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. 
2 



i8 THE ART OF WAR IN THE MIDDLE AGES [390 

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, they resign themselves 
to the dishonourable alternative of being slaughtered like 
sheep." ^ 

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 

^ Vegetius, i. § 20. 



4oo] GROWING IMPORTANCE OF ARCHERY 19 

battle of Adrianople.^ 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.^ 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. 

^(i) The heat of the day, "Romanes attenuatos inedia sitique confectos, et 
armorum grava7ttibics sarcinis, exurebat." (2) The lines of infantry close, and '* nostri 
occursantes gladiis obtruncant : mutuis \c\\\)\v=. galeae perfringebantitr et loricae,''' 

2 Being dedicated to Theodosius and his two sons as joint Auguni, the work ni; st 
have been written in the years 394-395. 



20 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 Vegetius complained — came to battle with 
their minds 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 
Dignitatum^ 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 ^ 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 
(407-41 1 ), Maximus in Spain (411), and Jovinus and Sebastianus 
on the Rhine frontier (41 1-4 12). 

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 

^ e.g. Leones Seniores, Ursi Valentiniani, promoti braccati seniores, Mauri 
tonantes, etc. 



45o] 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.^ 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. 

^ 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. 
BELTSARIUS AND THE GOTHS 

A.D. 450-552 

T'^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 
prey 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- 

22 



489] ZENO AND THE ISAURIANS ''" 23 

vincials of the southern mountains of Asia Minor.^ 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 

^ Diocletian, however, had raised two Isaurian legions, which appear in the Noiifia. 



24 THE ART OF WAR IN THE MIDDLE AGES [500 

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 chevaux-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 ^ 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, it 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 wc 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 

^ 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. 



53o] THE ROMAN HORSE-ARCHER 'HT ^5 

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 — ■xarocXoyoi is Procopius' name for them 
— of the old standing army ; ^ 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 mediaeval 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. ;b'jiq < iiiOBi ^^imiirj 

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. '\ 10 m'jir.v rjna:; aiii o:t 

" Now our horse-archers 'are very different men. They come 
to the fight culrassed 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 with equal ease whether they are 

^ We hear of numeri still, but no longer of legions — all of them had disappeared. 



26 THE ART OF WAR IN THE MIDDLE AGES [530 

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." ^ 

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 

^ De Bdlo Persico, I. i. 25-40. 



I 530] BELISARIUS' VICTORY AT DARAS HI iiy 

i 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.^ 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 overthrow^n 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," ^ 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.* 

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 
.Helisarius 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 

^ Procopius, De Bella Goflhico, III. i. • , , r 

^ Procopius calls them dopv(popol and viraairiaTai. The usual Latin word for' them 
was Bucce'larii, from Bnccelltini, the ration-biscuit, meaning retainers fed by tlieir lord. 

2 ^dvbov is used by Procopius both for the standard of the regiment, and for the 
regiment itself. 

* e.g., where Ammianus would still talk of the "cohors qtiarta Thracum,''*' Pro- 
copius would call them "that catalogue of Thracians which Bryes led." '' '''■ 



28 THE ART OF WAR IN THE MIDDLE AGES [53c 

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 
fighting 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. 



Htuazes 




Battle of 
PARAS 



AD. 530. 

Roman 



Persian: 



Foot r^ 

Horse p-^ 



Foot 
Horse 



-Bar«smanes 



.^*^' 



Lombaixl^ & Heruli 






% 



.W 






Himwl [lliiiirft>..l HiLj inij mtiwl 



Gothic Horse 



Gothic Foot 

BATTLE OF TAGIN.g:.AD 552 . 



Roman Horse 1^3 
Foot \^ 

Gortiic Horse IH 
Foot C!ll 



^ 



;55] THE BATTLE OF TRICAMERON 29 

n 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 
nfantry on the right, then the similar body from the left, which 
le recalled the moment that the danger on that flank was 
mded. He himself with his bodyguard followed. Charged in 
ilank 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 i\d 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 charged a 



^S] BEUSARIUS AND THE GOTHS 31 

tw minutes later by the Royals and 4th and 5th Dragoon 
juards. 

The Gothic war, the greatest of the three struggles waged by 
ustinian, was essentially a war of sieges and not of battles. In 
he first half of it, indeed, down to Belisarius' capture of Ravenna, 
here was no single general engagement between the Goths and 
he Imperialists. The decisive event of this part of the struggle 
vas the long beleaguering of Rome, from which the Goths retired 
oiled, partly because of their own unskilfulness in siegecraft, 
partly because of the deadly fever of the Campagna, which had 
:hinned their ranks. But if the sieges were the chief events 
n the struggle of A.D. 535-40, there were a good many skirmishes 
md minor engagements which served to display the qualities and 
lactics of the two armies. A glance cast round them shows that 
on both sides the cavalry did almost all the fighting, and would 
^eem to have been the larger half of the host.^ 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, w^ho rode through their 
ranks and put them into confusion. Then the Goths fell on 
them and routed them : the two officers, Principius and Tarmutus, 
who had counselled the sortie, were both slain while trying to 
rally their broken troops.^ 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,* 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 

^ On one occasion we find a force composed of 4500 horse, and only 3000 foot. 

2 Procopius, Be Bell Gott. i. 28. 

3 Ibid, i. 29. ^ Ibid. i. 27. 



32 THE ART OF WAR IN THE MIDDLE AGES [552 

final success which he had ahvays 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 hardly any knowledge 
whatever of archery. 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 Narses, 
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 



y^2] 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 miles 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 IN THE MIDDLE AGES [552 

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 IIl.'s device 
at Cre9y ; still more so does the rest of Narses' 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 Cre^y. For when they reached 
the centre of the semicircle formed by the Roman army, they 



J 



00- 



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,^ 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 

^ iroWCov T€ dvr]K€<TT(i}u KaKwv ej ireipav iXdovres 6\pi re Kai fx6\ii is rCov woXe^iwy 
dipiKOVTo TT}v Trapdra^t-v (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.^ 

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 field 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 Frankish 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 

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



PLATE II. 






^' 



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. 






I 



BOOK II 

THE EARLY MIDDLE AGES 
A.D. 500-768 



CHAPTER I 

I THE VISIGOTHS, LOMBARDS, AND FRANKS 

WHEN we leave the discussion of the mihtary art of 
the later Romans, and pass on to investigate that of 
he Teutonic kingdoms which were built upon the ruins of the 
.Vestern Empire, we are stepping from a region of comparative 
ight into one of doubt and obscurity. If, in spite of our 
possessing military manuals like that of Vegetius, official 
;tatistics such as the Notitia Dignitatum, and histories written 
yy able soldiers like Ammianus and Procopius, we still find 
iifficult points in the Roman art of war, what can we expect 
,vhen our sole literary material in Western Europe consists of 
garrulous or jejune chronicles written by Churchmen, a few 
"ragments of ancient poems, and a dozen codes of Teutonic 
aws ? To draw up from our fragmentary authorities an estimate 
3f 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 



42 THE ART OF WAR IN THE MIDDLE AGES [50c 

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

The Teutonic kingdoms which were founded in the fifti 
century within the limits of the Western Empire were some o 
them established by races accustomed to fight on horseback 
some by races accustomed to fight on foot. All the tribe: 
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 othei 
hand, which had started from the marshes of the Lowei 
Rhine or the moors of North Germany and Scandinavia were 
essentially foot - soldiery ; the Franks, Saxons, Angles, anc 
Northmen were none of them accustomed to fight on horseback 
The sharp division between these two groups of peoples is al 
the more curious because many tribes in each group had beer 
in close contact with the Romans for several centuries, and ii 
might have been expected that all would have learned a similai 
lesson from the empire. Such, however, was not the case : the 
Franks of the fifth century, though their ancestors the Chamav 
and Chatti had been for four hundred years serving the Roman.' 
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 modifiec 
their armament and customs. In the days of the breaking-up 
of the Western Empire the Franks seem no more advancec 



oo] THE VISIGOTHS IN SPAIN 43 

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

(I.) The Visigoths, 600-711. 

We have already spoken of the Goths, and their pre- 
ponderant use of cavalry in war. We have seen the Visigoths 
3f 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. (Not. Or. 
Thebais. 31-53 ; Mesopotamia, 31-33 ; and Phoenicia, 32-37). 



44 THE ART OF WAR IN THE MIDDLE AGES [531 

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 tc 
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 oi 
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.^ 
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 cliente^ 
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 

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



;o] WEAKNESS OF THE VISIGOTHIC ARMIES 45 

nction of adventurers of all sorts, especially that of the thirty 
lousand Foederati in Italy who joined the invader after the 
lurder of Stilicho. Hence in this heterogeneous mass there was 
generally recognised noble blood, such as was to be found 
nong more compact nationalities, like the Lombards, Bavarians, 
r Saxons. The only original distinction came from being 
romoted to official command by the king. But the men who 
ad once been given the appointment of " count " or " duke " 
raw wealthy, acquired lands, and accumulated clients. Their 
escendants in a few generations formed a true nobility based 
n wealth and local influence. The majority of the provincial 
overnors were drawn from their ranks, and they resented in a 
ody the attempts of strong-handed kings to supersede their 
lass in office by the preferment of obscure but loyal members 
f the royal coinitatus. Chindaswinth (641-652) and Wamba 
572-680) tamed them for a short time, but the moment that 
he sceptre passed to weaker hands, the aristocracy asserted 
tself again. At the moment when the monarchy fell in 711, it 
lad become wholly feudalised : the nobles and bishops were 
he real masters of the realm. 

The stream of Spanish annals is such a scanty one that we 
earn very little about the details of the interminable civil wars 
)f the sixth and seventh centuries. Towards the end of the 
atter 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 
;he Visigothic kings were at their wits' end to keep up the 
lumbers of their army ; a notable law of Wamba gives the 
3est proof of it. He orders that " every man who is to go forth 
n our host, duke or count or castellan, Goth or Roman, freeman 
Dr 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 

^ Lex Visigothorum, ix. 2, 9. 



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

thiufads or thousand-men, and centenarii or hundred-men, whos< 
duty was to collect the host each in his own locality. In tim< 
of peace the count and thousand-man were judges and governori: 
like an English ealdorman ; in time of war they took the fiel( 
at the head of the whole levy en masse, Gothic or Roman, o 
their district. Spanish armies, therefore, were often ven ''' 
numerous, but they were disorderly, undisciplined, and generally 
very half-hearted in their service. The masses of provincial 
cared nothing for their ephemeral kings, and thought mucl 
more of propitiating their local despots, the counts. Heno 
rebellious nobles could generally rely on the .service — slacl 
and unwilling though it might be — of the inhabitants of thei 
government. By the seventh century the majority of thes( 
inhabitants had become the " men " of their rulers, who thu 
reached such a pitch of greatness that we find them called 
even in state documents, tyrmini} as if they were independen 
princes. 

The Gothic nobles and their war-bands fought on horseback 
" gaudent equorum praepeti cursu," as Isidore of Seville wrote ii 
6i 5 ;2 though, when necessary, they would dismount. Their grea 
weapon was the lance ; their bodies were covered with harnes 
of ring-mail or scale-armour, and their heads by crested helms 
probably of the same shape as those worn by their neighbour 
the Franks. They bore round shields, swords, and dagger 
{semispatha, scraind). The mace and axe were not unknowi 
to them ; the use of the latter they had learned from the Franks 
and they therefore called it francisca. That defensive armou 
was fairly common may be deduced from the fact that Kin^ 
Erwig (68o) ordered that even of the slaves whom the bishop: 
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.^ The word employee 

^ 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 a 
first to dismiss them as wholly useless repetitions of Roman usage. But this would b 
unjust to the author, who shows that he is not wholly neglectful of the things of hi 
own day by making notes on the scrama-semispatha, and adding a mention of th' 
" secures quas et Hispani ab usu Francorum per derivationem franciscas vocant." It i 
to be noted also that he has no account of the old Roman breast and back harness o 
p'ate under lorica, and only catalogues the mail -shirt of rings and the lorica squatne<. 
of scales. See Etym. xviii. § ii, 13, 18. 

^ Lex Visigothoritni, ix. § 9. 



.o] SIEGECRAFT OF THE VISIGOTHS 47 

r the mail-shirt is zaba, the same which Maurice and Leo use 
r the armour of the Byzantine cavalry-soldier, and not brunia 
yrnie)y the common term of the Franks, Saxons, and other 
eutonic tribes of the North. 

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

The Visigoths seem to have had a greater skill in the poliorcetic 
"t than many of their Teutonic kinsmen. Probably it was 
icked up from the East-Romans during the long sieges of 
le haven-towns of South Spain during the reigns of Reccared, 
isibut, and Swinthila, when for a whole generation (580-620) 
le main political object of the kings was to recover the 
orts of Andalusia and Algarve, which the folly of Athanagild 
ad betrayed to the generals of Justinian. We find that the 
'isigoths were acquainted with the funda and balista, which 
irew respectively stone balls and darts, that they used the ram 
iries)^ and aided its work with the pluteus (shelter-hurdle) and 
le musculus for digging into the foundations of walls. In the 
ne siege of which we have considerable details, that in which 
Vamba took Nismes in 673, the ram, the stone-throwing 
lachine, and fire-arrows are described as in use.i 

The end of the Visigoths as a military power was sudden 
nd disgraceful. How far the immediate cause of the loss of 
ae battle of the Guadelete was disloyalty on the part of the 
ounts, or slackness on the part of their subjects in the provincial 
3vies, or a deficiency of properly - equipped fighting men, we 
annot tell. The details of the fatal day are lost; nor have 
/e sufficient notices of any Spanish wars of the previous century 
o enable us to construct a full account of the tactics of the 
/isigothic army. 

(II.) The Lombards, 568-774. 

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

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



48 THE ART OF WAR IN THE MIDDLE AGES [65( 

Empire than any of their brethren during the sixth, seventh 
and eighth centuries, so that we have a certain amount^ 
information bearing on them from Byzantine sources. TImP 
early legends have been preserved by the excellent Paul th( 
Deacon, who also furnishes us with a sketch of their late 
annals, abounding in those picturesque tales which, though the> 
may not be accurate history, are invaluable as giving th( 
manners and customs of the race. In addition we can draw oi 
the information contained in the code of laws drawn up h] 
Rothari (643) and the supplements appended by his successors. 
Like all the races that have ever dwelt by the Middle Danube 
they were essentially a race of horsemen. The primitive folk 
tales recorded by Paul show it very clearly ; on their first actua 
appearance on the stage of history it is equally manifest. Pro 
copius records how they sent to Narses two thousand five hundre< 
horsemen of noble birth, and three thousand of lesser race wh( 
were the attendants and squires of the others. If they dismounte( 
at Taginae to stand the Gothic charge, it was by Narses' order 
the old general had resolved to make his centre solid by placins 
there his steadiest auxiliaries.^ A little later, when they invad' 
Italy on their own account, we read of every king and duke an( 
hero fighting with lance and war-horse at the head of his mer 
One interesting passage in Paul gives us the armament of th 
Lombard knight — helm and mail-shirt, and even greaves, whicl 
last many Western races had not adopted even three centurie 
later.2 In another, we read of the great lance {contus\ s^ 
strong that a Lombard champion, who had pierced a Byzantin 
horseman through the body, actually lifted him from his saddl 
and bore him aloft wriggling on the weapon's point.^ Th 
other great Lombard weapon was the broadsword {spatha 
which seems to have been worn at all times,* not merely whe: 
the warrior was equipped for war. On one occasion only d' 
we hear of a hero fighting with a club, and then only because hi 
lance was not to hand.^ Though acquainted with the bow,^ the; 
do not seem to have used either it or the javelin to any exten 

1 Not, we need hardly repeat, because he wished to prevent troops of doubtfi 
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. 



55o] THE ARMS OF THE LOMBARDS 49 

in war. It was always on lance and war-horse that they placed 
■heir reliance, like the Goths, who had held the plains of 
iVorthern Italy before them. It was always on horseback that 
iheir plundering bands crossed the Alps to ravage Provence and 
Dauphine, faced the Bavarians on the Upper Adige, or pursued 
:he Slovenes of Carinthia when they dared to molest the borders 
3f Friuli. From a passage in the Tactica of Leo the Wise we know 
:hat, when hard pressed and surrounded, the Lombard knights 
vvould turn their horses loose, and fight back to back in a solid 
mass, with spears levelled outwards.^ 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 
:ourage, Argait and his horsemen charged up the steep slope and 
were destroyed by the great stones which the Slavs rolled down 
Dn 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- 
[juently 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 {jneeriuorpJmi), 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.^ 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 

^ Leo Sapiens, Tactica, xviii. § 8i. 2 p^ul. vi. § 24. 

' See Laws of Rothari and Luitprand, passim. 

•* 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 [712 

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 tc 
conquer, though they were only led by the " dukes " {heretoga:, 
or ealdonnen, 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 oi 
a Prankish invasion, aided by the Byzantines, that drove therr 
into combination again, and forced them to crown Authari cU' 
their king. The kingdom thus restored was never so strong as ii 
should have been ; the dukes of Spoleto were in practice, if no1 
in name, independent of it, and those of Benevento hardlj 
acknowledged its supremacy at all. It was only Luitpranc 
(7 1 2—744), who reigned but shortly before the Prankish conquest 
of Italy, that welded the Lombards north of Benevento intc 
a compact state. 

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

^ Domus Nostrae Civitates, Codex Dipl. Long. ii. 334. 



5o] THE LOMBARD HOST 51 

,uch lacking among the Visigoths of Spain. All Lombards 
nail and great were exercitiales (or arimannt)^ bound to turn 
it at the monarch's call to war, like the English ^r^. Many, 
Dth noble and simple, had made themselves the king's " men," 
/ the oath of personal devotion. They were called gaisindi} 
word corresponding of course to the Anglo-Saxon gesith, and, 
<e the gesith, rode in their lord's train, and had their place in 
s hall. The chief of these military retainers were the marpaJiis 
■ constable, the scilpor or shield -bearer, and the banner-bearer 
" the king. The dukes in a similar way kept smaller bands of 
lisindi, but they were never able to make henchmen of the 
hole of the freemen resident in their duchies, as did the counts 
" Visigothic Spain. The number of the Lombards of middle 
rtune was too great to allow of such a usurpation taking place, 
id the king's gastaldus and scJiultheiss (reeve, as the English 
Duld have called him) were present in each duchy, to keep its 
ler in check, and afford protection to any freemen whom he 
ight strive to oppress.^ 

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

(HI.) The Franks, 500-768. 

The Prankish tribes whom Chlodovech had united by the 
'Wer of his strong arm, and who under his guidance overran 
e valleys of the Seine and Loire, were among the least 
^ilised of the Teutonic races. In spite of their long contact 
th the empire, they were (as we have already had occasion to 
sntion) still mere wild and savage heathen when they began 
z conquest of Northern Gaul. The Franks, as pictured to us 

Sidonius Apollinaris, Procopius, Agathias, and Gregory of 
)urs, still bore a great resemblance to their Sigambrian or 
lamavian ancestors whom Tacitus described more than three 
ituries earlier. The words in which Sidonius paints them in 
D are practically identical with those which Agathias used 
)re than a century later, so that even the conquest of Southern 
.ul seems to have made little difference in their military 

^ Paul translates gaisind by salelles, vi. 38. 

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



52 THE ART OF WAR IN THE xMIDDLE AGES [50c 

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 foi 
close combat. He mentions their dense array and their rapid 
rush, " for they close so swiftly with the foe, that they seem tc 
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 wea 
neither mail-shirt nor greaves, and their legs are only protectee 
by strips of linen or leather. They have hardly any horsemen 
but their foot-soldiery are bold and well practised in war. Thej 
bear swords and shields, but never use the sling or bow. Thei 
missiles are axes and barbed javelins (ayywvss). These last ar 
not very long, they can be used either to cast or to stab. Th 
iron of the head runs so far down the stave that very little c 
the wood remains unprotected. In battle they hurl thes 
javelins, and if they strike an enemy the barbs are so firml 
fixed in his body that it is impossible for him to draw th 
weapon out. If it strikes a shield, it is impossible for th 
enemy to get rid of it by cutting off its head, for the iron run 
too far down the shaft. At this moment the Frank rushes ii 
places his foot on the butt as it trails on the ground, and S( 
pulling the shield downwards, cleaves his uncovered adversar 
through the head, or pierces his breast with a second spear." ^ 
The fi'aitcisca or casting axe was even more typically 
Frankish weapon than the barbed angon. Numerous specimer 
have been found in Merovingian graves ; ^ it was a single-blade 
axe with a heavy head, composed of a long blade curved on i 
outer face, and deeply hollowed in the interior. It was careful 
weighted, so that it could be used, like the American tomahaw . 
for casting purposes, even better than for close combat. Tl \ 
skill with which the Franks discharged the weapon just befo 
closing with the hostile line was extraordinary, and its effectiv 
ness made it the favourite national weapon. A shield, swor 
and dagger completed the arms of the warrior : the first-nam( 
was of a broad oval shape, and had a large iron boss and i 
iron rim ; the sword was a two-edged cut-and-thrust weapo 

^ Agathias. 

- One was in the first Frankish monument to which a definite date can be giv< i 
Childeric's tomb at Tournay (481). 



oo] FRANKS AND GAULS 53 

anging from thirty to thirty-six inches in length;, the dagger 
sci^amasax^ semispatha) was a broad thrusting blade of some 
;ighteen inches. 

For some two centuries on from the time of Chlodovech, 
hese were the arms of the Prankish foot-soldiery ; they seem to 
lave borrowed nothing from their Roman predecessors. It is 
:rue indeed that some of the Gaulish levies who served the 
Merovings continued for a space to wear the ancient equipment 
Df 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, eren ^ 
down to the heavy-nailed militaiy 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.^ 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 w^ere 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 Pranks 

^ KoX crrjfiela rk atp^repa iTayofx^voi ovtu is fidxv Kadlffratrrai, Kal crx^A"* t'^^" 
"Poj/j-aLtop ^v T€ rots AWols &ira<rc Kal iv rots vToSrjfxaa-i 8ia(r(bj;'ovcnv {De Bell. Gott. i. 12). 

lAs 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 [55c 

also : " The poor Roman tends to assimilate himself to the 
German, and the wealthy German tends to assimilate himsel 
to the Roman." 1 While the masses in Gaul forgot the ok 
military habits of the empire, and degenerated into disorderl} 
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 show in things civil, so in certain military matters they die 
not remain entirely uninfluenced by the Roman practice. Ir 
the sixth and seventh centuries we find among them the feebk 
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 mentior 
is in Procopius,^ 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 oi 
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.^ 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."^ 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.^ The henchman of Duke Guntram wears a 
breastplate, and is drowned by its weight in a ditch (a.d. 583).^ 
The usurper Gundovald Ballomer is saved by his body armour 
from the stroke of a javelin (A.D. 585).^ In the Saxon war of 
626 we read of both Clothar II. and his son Dagobert wearing 

^ " Romanus miser imitatur Gothum, et Gothus utilis imitatur Romanum." 
'- De Bell. Gott. ii. 25. 

2 So I gather from the account of the single combat of Guntram and Dracolenus in 
Gregor>', v. 

* Gregory, iv. § 18. ^ Ibid. v. § 48. 

6 .Ibid. vi. § 26. • Ibid. vii. § ^Z. 



;oo] PRANKISH ARMOUR 55 

lelm and breastplate (A.D. 626).^ The brunia, which com- 
posed the body armour, was no doubt usually the mail-shirt of 
ings which we find among all Teutonic races in the Middle 
Ages. But scale armour sewn on to a leather foundation was 
iiso known ; it was sometimes of the fish-scale shape, sometimes 
jquare-scaled. In either case it was fixed so that each row of 
scales overhung the one below it, and protected the upper ends 
Df 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, howev^er, 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. II.] 

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.^ 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 Pranks brought no headpiece of their 
own into Gaul ; they had fought bareheaded when they dwelt 

■^ Vita Dagobcrli, § 13. 

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



56 THE ART OF WAR IN THE MIDDLE AGES 



>mH 



on the moors of Toxandria. But they found the late Roi 
helm in full use in their new realm, and there can be no doubl 
that their kings and nobles borrowed it from their subjects 
From the first, as we have seen, the Franks used theii 
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 
i" 539-^ We cannot suppose that when the Gallo-Romar 
Bishop Sagittarius equipped himself in a helm in 574, to fighl 
the Lombards, he put on some newly-invented Frankish 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 ir 
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 ^ — 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.'^ 

The Frankish 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. S3' ^ See p. 54. 

^ The hrttnia is mentioned in the early Ripuarian law, and valued at twelve solidi, 
the helm at six, the sword at seven {Lex Rip. xxxvi. §11). It is more surprising 
to find hainbergae (greaves) mentioned, and valued at six solidi. 

■* See illustration on Plate ir. : the Utrecht Psalter is late, but its drawings arc- 
copied from Merovingian originals. 



)3o] THE FRANKS LEARN HORSEMANSHIP 57 

etinues were habitually fighting on horseback by the end of the 

seventh century, though when pressed or surrounded they 

A'ould still dismount and fight on foot like their ancestors. The 

irst single combat on horseback related to us is that of 

Suntram and Dracolenus in 578. Early instances of the 

ippearance of a considerable body of cavalry are found in the 

irmy of Count Firminus in 567/ and that of Duke Leudigisl in 

584;^ 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 

Frankish 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.^ 

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 Zi^ilpich 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 
[phalanges'] 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." * Obviously this could only 

^ 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. 

" Vtfa Dagoberti, § 13. 

^ " 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). 



58 THE ART OF WAR IN THE MIDDLE AGES [73 

have happened in an infantry fight. Still more interesting i 
the account of the array of the Franks a hundred years later, a 
the all-important battle of Poictiers, where Charles Mart( 
turned back the advancing flood of Saracen horsemen who ha^ 
swept so easily over the debris of the Visigothic monarch} 
" The men of the North," says the chronicler, " stood a 
motionless as a wall ; ^ they were like a belt of ice frozei 
together, and not to be dissolved, as they slew the Arabs wit! 
the sword. The Austrasians, vast of limb, and iron of banc 
hewed on bravely in the thick of the fight ; it was they wh* 
found and cut down the Saracen king." Obviously, therefore, a 
Poictiers the Franks fought, as they had done two hundred year 
before, at Casilinum, in one solid mass,^ without breaking rani 
or attempting to manoeuvre. Their victory was won by th* 
purely defensive tactics of the infantry square ; the fanatica 
Arabs, dashing against them time after time, were shattered t( 
pieces, and at last fled under shelter of the night. But then 
was no pursuit, for Charles had determined not to allow hi: 
men to stir a step from the line to chase the broken foe 
Probably he was right, for an undisciplined army canno 
advance against cavalry without danger, and the Arabs, ever 
when repulsed, were too agile and brave to be allowed th( 
chance of penetrating into the mass. We must conclude 
therefore, that the Frankish chiefs and nobles had all dis 
mounted and fought on foot in the " wall of ice " which thej 
opposed to the fiery onslaught of the Moslem horsemen. SucI 
tactics were, no doubt, exceptional by the eighth century, anc 
adopted only against an enemy all - powerful in horsemen 
Against armies of Saxons, or Frisians, or Bavarians, composec 
wholly or almost wholly of foot - soldiery, the Franks woulc 
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 Frankish knights could have 

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

2 See p. 63. 



7oo] 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 e7i 
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.^ 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 
such 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.^ 
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 F>anks, 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 
Frankish 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 

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

- Gregory of Tours, vi. § 31. ^ Lex Rip. Ixv. § 2. 



6o THE ART OF WAR IN THE MIDDLE AGES [6oc 

ranks of the Gaulish provincials. Great officers of state witt 
Roman names are found early in the sixth century ; by the enc 
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 Prankish king, like al 
Teutonic sovereigns, had his own " men " bound to him by oath 
they were called antrustions, and corresponded to the Englist 
gesithj the Lombard gaisind^ and the Gothic saio. But they dc 
not appear to have been a very numerous body, certainly nol 
one large enough to form the chief element of importance in the 
host, though there were enough of them, no doubt, to furnisl" 
the king with a bodyguard. The Prankish 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 oj 
persons not included in the royal comitatus. Both the Salic and the 
Ripuarian laws value a free Prank at two hundred solidi, but a 
freeman " in the king's trust " at six hundred. That there were 
also Gauls and letes (freedmen) among the antrustions^ is shown b) 
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 king's trust and slain him, nine hundred solidi." ^ Prom the 
ranks of the antrustions were drawn the counts and dukes whc 
headed the Prankish 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 vicar ii 2LX\di centenariiy 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- 

' Lex Sal. xlii. (ed. Hessels ) : "Si quis coUecto contubernio hominem ingenuum in 
domo sua occiderit, si in truste dominica fuit ille qui occisus est, solidos MDCCC 
culpabilis judicelur : solidos DCCCC si quis Romanum vel litum in truste dominica 

occiderit." 



yoo] MILITARY ORGANISATION OF THE FRANKS 6i 

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.^ 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 : ^ " 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. 

^ Gregory of Tours, x. § 3. ' Ibid. ix. § 31. 



62 THE ART OF WAR IN THE MIDDLE AGES [6i. 

Even the kings themselves often found that the hereditary 
respect of their people for the royal blood was insufficient tc 
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.^ As 
might have been expected, the disorderly host was well beaten. 
An example of the opposite form of indiscipline was seen in 
612, when the armies of Theuderich II. 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. 

^ 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 
learly similar to the Franks than to the German tribes who 
Nvandered south. In blood and language, however, they were more 
ikin to the Lombards than to the Franks ; but two or three 
iiundred 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 Notitia} 

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 
gesiths, used to go out to war in their tunics, with undefended 
head and breast, and bearing the broad shield of linden tree 

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



64 THE ART OF WAR IN THE MIDDLE AGES [j 

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

Of the offensive arms of the old English the spear was th( 
most prominent: they were in this respect still in the stage 
which Tacitus had described four centuries back. The mos 
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-headec 
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 seax, 
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 Frankish type, i.e. 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 

arger part still remained in their ancient seats.^ Their chiefs 
vere not the old heads of the entire race, but mere heretogas^ 
eaders in time of war, whose authority had no ancient sanction. 
Mo continental Teutonic State started under such beginnings : 
;he nearest parallel that we can point out is the time when the 
Lombards, after the death of King Cleph, abode for ten years 
A^ithout a king, and pushed their fortunes under thirty inde- 
Dendent dukes. But this condition of things lasted but a few 
/ears in Lombardy, and was soon ended by the outward pres- 
sure from Frank and East- Roman. In Britain it was more than 
'our hundred years before the Danish peril led to a similar 
result. 

The old English kingdoms, therefore, were the small districts 
:arved 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.^ 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 

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

2 This, one must certainly imply from Bede i. 15, and from Xennius. 

5 



66 THE ART OF WAR IN THE MIDDLE AGES [600 

or gesiths, and of other freemen, some of noble blood {eorls)^ some 
of simple blood iceorls). 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 " eorl's 
tun" in the same way^ (laws 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 
Wihtraed ^ ninety years later. 

But the later and larger English kingdoms were of a some- 
what different cast. The picture of Wessex 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.^ Ev^n 
before the coming of Augustine and the introduction of 
Christianity, the English had begun to admit the Romano-Britons 
to terms.^ After a victorious campaign the cities were still 
sacked and burned, but the Celtic country-folk were no longei 
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.^ They even serve in his retinue 
the horse-wealh who rides on his errands is specially mentioned,' 

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

2 Wihtraed's laws, § 5. 

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

** See, for example, Bede's account of the heathen Aethelfrith, "who conquered mor 
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. ^ T^aw 33. 



7oo] THE LAWS OF INI 67 

and King Cynewulf had a Welshman among his gesiths.^ We 
are reminded at once of the Prankish king and his Gallo-Roman 
antnistions on the other sid6 of the Channel.^ 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 hlaford or lord, 
to whom they owe suit and service. If they try to shirk their 
duty to him, heavy fines are imposed on them.^ ^Ve are 
tempted to infer that a large proportion of ceorls v/ere now 
either the vassals of lords or the tribute-paying tenants on royal 
demesne land.^ The king has geneats or landholding tenants, 
who are so rich that they are twelve-hynde and own estates 
even so large as sixty hides.^ 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 valuat^'on. 
Comparing this with the almost contemporary law of the 

^ A.S. Chronicle, a.d. 755 ; but the event related occurred in 784. 

^ See p. 60. 

^ Law 39. * Laws 59, 67, "paying gafol," rent or tribute, to him. 

^ 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.^ 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 : ^ 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.^ 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. 

^ See p. 56. ^ Ancient Books of PVales, Taliessin, xv, 

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



7oo] 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 l^een 
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.^ 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," ^ the " defence wrought with the image of the boar, 
furnished with cheek guards, decked with gold, bright and 
hardened in the fire," ^ 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. 

^ Collectanea Antiqua, vol. ii. " Beo. 1450. 

^ 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^ as did Swiatoslafs Russians before 
Dorostolon. One isolated passage in the Beowulf s'pQd^is of a 
king's war-horse " which never failed in the front when the slain 
were falling." ^ 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,^ 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.* 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. 

^ A.S. Chronicle, Year 1055. 

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

2 A.S. Chronicle under a.d. 547. ^ Bede, iii. 16. 



;oo] 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," ^ i.e. in close 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 ^ 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 w^ar-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 

^ The ' * 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 tesiudo, with which it has been compared. Warriors in Beoivulf 2980 hew 
each other's helms to pieces " over the shield-wall." 

^ The banner is mentioned both in Bede (King Edwin's) and in Beowulf 2<^o6. 



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 



73 



CHAPTER I 

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

rHE accession of Charles the Great serves to mark the 
commencement of a new epoch in the art of war, as in 
nost other spheres of human activity in Western Europe. In 
•ur second book we had to describe the military customs of 
^>ank and Goth, Lombard and Saxon, in separate sections. The 
onquests of Charles combined all the kingdoms of the Teutonic 
Vest into a single State, with the exception of England and the 
)bscure Visigothic survival in the Asturias. Races which had 
litherto been in but slight contact with each other are for the 
uture subjected to the same influences, placed under the same 
nasters, and guided towards the same political ends. The 
escripts of Charles were received with the same obedience at 
?avia and Paderborn, at Barcelona and Regensburg. For the 
irst time since the fall of the West-Roman Empire the same 
organisation was imposed on all the peoples from the Ebro to 
he Danube. The homogeneity which his long reign imposed 
apon all the provinces of Western Europe was never entirely 
ost, even when his dynasty had disappeared and his realm 
lad fallen asunder into half a dozen independent States. In the 
listory of the art of war this fact is as clear as in that of law, 
iterature, or art. In spite of all national divergences, there is 
or 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 
^reat Mayors of the Palace the Franks were showing a decided 
;endency towards the adoption of armour and the development 
3f cavalry service. It is under Charles the Great that this 
:endency receives a definite sanction from the royal authority, 



76 THE ART OF WAR IN THE MIDDLE AGES [^( 

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

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

But Charles the Great undertook offensive wars on a muc 
larger scale than Pepin and Charles Martel. His armies wei 
so far afield, and the regions which he subdued were so broa 
that the old Frankish levy en masse would have been far too slo 
and clumsy a weapon for him. An army of Neustrian ar 
Austrasian infantry could hardly have hunted the Avars on tl 
plains of the Theiss and the Middle Danube. The Franki.' 
realm had been so vastly enlarged that it extended, not as of o 
from Utrecht to Toulouse, but from Hamburg to Barcelon 
To keep this mighty empire in obedience a more quickly-movii 
force was required ; hence Charles did his best to increase tl 
number of his horse-soldiery. It was also incumbent on him 1 
raise the proportion of mailed men in his host : against tl 
well-armoured Lombard and Saracen, and later against tl 



CHARLES THE GREAT AND THE LOMBARDS 77 



1 rse-bowmen of the Avars, troops serving without helm and 
lie were at a great disadvantage. 

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

In the first half of his reign Charles issued a good deal of 

ilitary legislation for his newly-conquered Lombard subjects. 

;e imposed upon them the Prankish regulations on military 

irvice, which made the fine for neglecting the king's " ban " 

xty solidi, — the old Ripuarian valuation of the offence, — and 

le penalty for desertion in the field, " which the Franks call 

^resliscs" death, or at least to be placed at the king's mercy 

oth for life and property.^ It is interesting to find in the 

.ombardic Capitulary of y^6 that the Lombards who are to 

kvear obedience to the royal mandates are defined as cavalry 

ne and all, being described as " those of the countryside, or men 

f the counts, bishops, and abbots, or tenants on royal demesne, 

r on Church property, all who hold fiefs, or serve as vassals 

nder a lord, all those who come to the host with horse and arms, 

hield, lance, sword, and dagger." * The possession of this mass 

•f Lombard horsemen was of the greatest importance to Charles 

n his wars with the Avars. Nearly all the fighting against 

hese wild horse-bowmen was done by the Lombards, under 

^epin, the king's son, whom he had made his vicegerent in 

^taly. It was a Lombard host which in 790 pushed forward 

nto the heart of Pannonia, beat the Avars in the open field, and 

;tormed their camp. The slow-moving Austrasians meanwhile 

lad only wasted the Avaric borders as far as the Raab. A few 

y^ears later it was again the Lombard horsemen who practically 

nade an end of the Avaric power : under Pepin and Eric Duke 

3f Friuli they captured the great " Ring," or royal encampment 

of the Chagan, hard by the Theiss, and sent its spoils, the 

^ Cap. Mm. § 7 : " Ut bauga et bruniae non dentur negociatoribus. " 

2 Cap. Aquisg. § 7. ^ Cap. Ticinense, § 3. ■* Cap. Langobardiae of 786, § 7. 



78 THE ART OF WAR IN THE MIDDLE AGES [8c 

accumulation of two centuries of plunder, to deck the halls ( 
Aachen. The Avars never raised their heads again, and fe 
into decrepitude. If he had led only Prankish infantry levie 
Charles would never have been able to subdue this race of noma 
horsemen : the numerous Lombard knights, however, could bot 
pursue them and ride them down when caught. It is interestin 
to note how the strong domineering spirit of the great kin 
inspired his new subjects to undertake and carry out an adver 
ture which their own kings had never been able to achieve, fc 
the Avar raids had been a scourge to Friuli and Lombard] 
" Austria " for two centuries, and no remedy had been foun 
against them. 

The chief military ordinances of Charles the Great ai 
five rescripts dating from the later years of his reign — th 
Capitulare de Exercitu Proniovendo of 803, the Capitula-t 
Aquisgranense of 805, the later edicts issued from the same cit 
in 807 and 813, and the Capitulare Bononiense of 81 1. All thes 
deserve careful study. 

The first of them, the edict of 803, is directed towards tb 
substitution of a smaller but better-armed force for the ol 
general levy. It ordains that the great vassals must take to th 
field as many as possible of the retainers whom they hav 
enfeoffed on their land {homines casati). A count may lea\ 
behind only two of his men to guard his wife, and two moi 
to discharge his official functions. A bishop may leave onl 
two altogether.^ Secondly, a new arrangement is made as t 
the field service of all Franks holding land. Everyone wh 
owns four mansV^ ox over, must march himself under his Ion 
if his lord is serving on the expedition, — under his local cour 
if the lord be busy elsewhere. To every man who owr 
three mansi there shall be added another who has but one, an 
these two shall settle between them for the service of one ma 
properly equipped : if the wealthier goes himself, the poorc 
shall pay one- fourth of his equipment ; if the poorer goes, th 
wealthier shall be responsible for three- fourths. Similarly, a 
men owning two mansi are to be arranged in pairs : one is t 
march, the other to provide half the equipment. And so, agair 
holders of one jnansus are to be arranged in groups of four : on 
will go forth, the other three will each be responsible for om 



1 Cap. de Exercitu Proniovendo, § 4. 

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



I 



5] THE CAPITULARIES OF CHARLES THE GREAT 79 

iirth of his equipment.^ The local counts are charged to see 
at all men holding- a mansus or more are placed in one of these 
cups : those found unenrolled are to be heavily fined for 
irking the ban.^ Thus we see that the service of the ill- 
med poor is lightened, and that of the well-armed rich strictly 
iforced. The general result would be a decrease in numbers, 
it a rise in average personal efficiency, in the host of the 
aim. 

The Capitulare Aquisgranense of 805 is intended to supple- 
ment the ordinance of 803. It orders that every man having 
velve mansi must come to the host in a mail-shirt : anyone 
ho has such armour and fails to bring it to the host is to forfeit 
oth the byrnie and any beneficimn that he may hold from 
ic king.^ The fine for neglecting the ban, or failing to be 
iiroUed in one of the contributary groups established in 803, is 
3 be half a man's substance ; — three pounds of gold for anyone 
olding land or chattels to the value of six pounds, thirty 
Dlidi for a man owning three pounds, and so forth.* The 
rohibition against selling arms outside the realm is re-enforced, 
ncl it is enacted that no man shall carry weapons within his 
'wn district in time of peace : " if a slave is found with a spear, 
t shall be broken over his back."^ 

The bulk of the army consisting of men owning less than 
welve mansi, it is obvious that the minority only were as yet 
urnished with armour. All the men of the contributory groups 
ire 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 
jven less than the one mansus which was laid down as the base 
)f 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 
3f half a mansus, or possessors of ten solidi in chattels, are to 
contribute to equip one of themselves.^ 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. 

^ Cap, de Exerc. Promov. § i . ^ /^^/^^ § 2. 3 Cap. Aquisg. 805, § 6. 

■^ Ibid. § 19. 5 Ibid. § 5. 6 jjjid^ go7^ § 2. 



8o THE ART OF WAR IN THE MIDDLE AGES [Sn 

The ordinance for the Frisians is quite different. The counts 
all holders of a royal " beneficium," and all who serve on horse 
back (caballai'ii omnes), are to march out whenever the ban \\ 
proclaimed ; of the commons {pauperiores) every six men are tc 
join in equipping one warrior for the host. There is unfortunatel} 
no statement of the limits of the class which served as mountec 
men ; we should have been glad to learn its character. No 
improbably it may have consisted of the holders of twelve 
mansi, and the personal retainers of the great vassals anc 
officials. 

For the inner discipline of the host the Capitulai^e Bononienst 
(8ii) is very important. We learn from it that those whc 
arrived late at the muster were punished by being compelled t( 
abstain from wine and flesh for just so many days as they hac 
fallen behind the appointed time.^ Anyone found drunk ir 
camp was to be deprived of wine till the campaign was ended. 
Every holder of a " beneficium " who deserted his comrades ii 
the hour of need, either from cowardice or from private feud 
was to forfeit his holding to the crown.^ The provision of fooc 
which each man was to bring to the host is defined as beinj 
three months' rations ; it consisted, as we learn from a late 
document, of flour, bacon, and wine.^ The three months were tc 
count from the border, with certain relaxations in favour o 
those coming from afar. Thus anyone coming from beyond th< 
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 beyonc 
the Rhine going east may only count from the Elbe, anc 
a dweller beyond the Loire going south may only count fron 
the Pyrenees.^ The Capitulare Bononiense is very clear on the 
necessity for providing as many fully-armed men as possible : i 
enacts that if any bishop or abbot finds that he has more byrnie: 
in store than he has to contribute men to the host, he must no 
let them lie idle, but at once inform the king of their existence. 
It also lays great stress on the necessity of all retainers follow 
ing the host even when their lord is not present : if he neglect 

1 Cap. Bon. § 3- ' Jbid. § 6. ^ //./^^ § ^_ 

* Cap. Aquisg. 813, § lo. Cf. also the curious story about Charles and th 
drunken guards in the Monk of St. Gall, book ii. 

6 Cap. Bon. § 8. 

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



13] THE CAPITULARIES OF CHARLES THE GREAT 8i 

3 forward them to the local count, he must pay the fine that 
bey have incurred by slighting the hereban} 

The section on rations in the Capitulare Bononiense can be 
upplemented by a clause of the edict De Villis Dominicis, which 
lys down the rule that cars such as follow the host should each 
le able to contain twelve bushels of corn, or twelve small barrels 
•f wine, and that each car should be furnished with a leather cover 
•ierced with eyelet holes, and capable of being turned into a 
)ontoon by being sewed together and stuffed (with hay?). 
^ach cart was to carry a lance, a shield, a bow and quiver — pre- 
umably to equip the driver in time of need.^ 

Last of the military decrees of Charles the Great comes the 
Zapittilare Aquisgra7tense of 813, which contains several im- 
)ortant notices. It provides that the count, when his men are 
nustered, must see that each has a lance, a shield, a bow, two 
)owstrings and twelve arrows. No one is for the future to 
ippear carrying a club alone ; the most poorly-armed men must 
it least have a bow. The stress laid on the bow in this document 
md in the Capitulare de Villis Dominicis is important. The 
veapon was practically new to the Franks, and the attempt to 
nake it universal was probably due to experience in war against 
he Avars,^ the only neighbours of the empire who made much 
ise of the weapon. Another clause provides that all the " men " 
obviously the household men) of counts, bishops, and abbots 
nust have both helm and mail-shirt. We get from section 10 
)f this document a glimpse at the existence of a military train : 
)n the royal cars are to be pickaxes, hatchets, iron-shod stakes, 
Davises, rams, and slings (obviously machines, not merely hand- 
dings). The king's marshals are to provide stones suitable for 
:asting from th^-so^ fundibuli. 

On all these documents the best commentary is the summons 
,vhich calls Fulrad, Abbot of Altaich, to the royal host in 806. 
(t is worth quoting at length. " You shall come to Stasfurt by 
:he Weser on May 20," writes the king, " with your ' men ' 
Drepared to go on warlike service to any part of our realm that 
vve may point out ; that is, you shall come with arms and gear 
md all warlike equipment of clothing and victuals. Every 
Norseman shall have shield, lance, sword, dagger, a bow and a 

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

2 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. 

6 



82 THE ART OF WAR IN THE MIDDLE AGES [8o( 

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 ab]( 
to hold out for six. On your way you shall do no damage t( 
our subjects, and touch nothing but water, wood, and grass 
Your men shall march along with the carts and the horses,^ anc 
not leave them till you reach the muster-place, so that they ma} 
not scatter to do mischief. See that there be no neglect, as yoi 
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 hi 
retainers for general service. It is noteworthy that all Fulrad' 
followers are expected to appear on horseback ; there is n( 
mention of any foot-soldiery, or directions as to their equipment 
It is not definitely stated that all the abbot's horsemen are t< 
appear in mail ; the summons being dated before the laws o 
807 and 813, it naturally contains no such order. Any o 
Fulrad's men who had twelve mansi would have been bound t( 
serve in a byrnie by the edict of 805, but compulsion is not ye 
put upon the rest. The command to bring the bow is to b 
compared with the contemporary attempt to make the infantr 
adopt the same weapon. In neither case did the experimen 
succeed. The very large quantity of provisions and the heav^ 
entrenching tools must have made the waggon train very cumber 
some. It was evidently contemplated that the camp migh 
have to be fortified, in order to protect the mass of baggage 
it is for this purpose that the iron-shod stakes and the spade 
are required. Charles is also, as the last clause of the summon 
shows, very anxious to avoid the cardinal vice of the old Mero 
vingian hosts — the plundering of the districts through which th« 
troops had to march before reaching the frontier. Hence th 
very heavy load of rations which Fulrad is directed to bring wit] 
him. If the train made the army slow to assemble and slow t( 
move, it at any rate enabled it to carry on operations even in ; 
hostile or a devastated district for several months, long after th 
date at which a Merovingian expedition would have commencec 
to starve and then to disband. 

When all the royal commands were carried out under th 

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



;oo] THE BURGS OF CHARLES THE GREAT 83 

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

But perhaps the most important of all Charles' innovations is 
lis systematic use of fortified posts. When a district had done 
lomage and given hostages and tribute, he did not evacuate it as 
lis predecessors would have done, and leave it free to revolt 
Lgain at the first opportunity. He selected a suitable position, — 
L hill by a riverside was his favourite choice, — and there erected a 
)alisaded and ditched " burg," in which he left a garrison. Each 
)ost was connected with the next, and with its base on the old 
rontier, by a road. Charles and his officers at last acquired a 
'Cry considerable skill in the laying out of entrenchments ; it 
vas unfortunate for the empire that his successors neglected the 
irt, till a long series of Danish invasions compelled them to learn 
t again. Probably the most ambitious work of entrenchment 
vhich was undertaken in his reign was the grea.t circumvallation 
ound Barcelona, which was constructed in 800 by the king's son 
_.ewis and the levies of Aquitaine and Septimania. The army 
ay around the town for the whole winter of 800-801, hutted and 
jirt by a double trench and palisade, to guard against sorties 
"rom within and diversions from without. The works were so 
efficient that the Moorish garrison, after a gallant resistance, was 
tarved out and forced to surrender. The bui^gs of Charles were 
ndeed a very successful expedient: it was seldom that they 
vere taken ; that of Eresburg only fell by treachery in ^^6^ 
hough that of Karlstadt seems to have been fairly stormed by 
he desperate assault of the Saxons (778). The use of these 
ortifications was a new lesson in the art of war for Western 
iurope ; the Teutonic nations hitherto had never even fortified 
heir own camps, much less had they thought of employing the 
ipade 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 day. 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 tc 
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 wit! 
another detachment is converging on the enemy from a dififereni 
base. Eager that they should have the sole credit of the 
victory which they supposed to be in their hands, they bade thei) 
men snatch up their arms, "and hastened on as if they were 
about to pursue a beaten army, not to fight an intact one, each a. 
fast as his horse zvould go} so that they came all in disorde; 
against the Saxons, who stood ranged in front of their camp. 
The reckless attack was beaten off, and four counts, two miss 
doininici, and more than twenty other persons of account, fel 
" with many of their men, who chose to follow them to the deatl 
rather than to survive them." If these words do not imply tha 
the whole of Geilo's and Adalgis' forces were cavalry, they mus 
at least mean that so large a proportion of them were horsed tha 
the counts hoped to win without the aid of their infantry, whicl 
in such a mad onslaught must have been left miles behind. 
The latter, in all probability, is the real meaning of the passage 
and the desperate courage of the Prankish horsemen is to b 
accounted for by the fact that they were the henchmen an( 
enfeoffed retainers (Jiomines casati) of the counts, whom the; 

^ " Prout quemque velocitas equi sui tulerat, unus quisque eorum sumn 
festinatione contendit " {Ann. Einh. 782). 

^ The army had been raised in Thuringia and among tlie Franconian district 
where we should expect to find more foot than horse. 



ii 



ij THE STRATEGY OF CHARLES THE GREAT 85 

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

Of the order of Prankish hosts in battle, i.e. whether the horse 

tood on the wings or in front of the foot-soldiery, we are equally 

iiiable to speak with certainty. Whether there was any larger 

mit in the assembled army than the count and his local follow- 

ng we are never informed. That the host marched in divisions 

vith a rearguard and vanguard may be deduced from the 

iccount of the disaster of Roncesvalles, where the rear (" ii qui 

lovissimi agminis incedentes, praecedentes subsidio tuebantur"^) 

vere so far from the main body that they were cut to pieces 

before their comrades could return to help them. A march in 

)arallel columns over open country can probably be traced in one 

)f the Avaric campaigns of 791 and the Saxon campaign of 804. 

Perhaps the most scientific disposition of forces recorded in 

ill the wars of Charles occurs in a campaign at which he was not 

limself present — the invasion of Catalonia in 800-801. On this 

occasion his son Lewis, who held the command, while under- 

:aking the siege of Barcelona with one-third of his forces, placed 

mother third, under William Count of Toulouse, some leagues 

•vest of the town to act as a covering army, while he himself 

A^ith the remainder took post nearer his base of operations in 

R.oussillon, ready to aid either of the other fractions that might 

'equire his help. The Caliph of Cordova advanced from 

Saragossa, but found the covering army so strongly posted that 

le 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 MH^DLE AGES [8i/ 

he describes King Desiderius and his henchman Ogier the Dam 
watching the long column of the invading army draw near. A.' 
each body comes into sight, the king asks whether his rival anc 
the main host have not now appeared. Ogier replies again anc 
again that Charles is not yet at hand — the numerous warrior: 
that have passed by are but his vanguard. At last the plaii 
grows dark with a still migntier column than any that have ye 
drawn near. " Then appeared the iron king, crowned with hi; 
iron helm,^ with sleeves of iron mail on his arms, his broad breas 
protected by an iron byrnie, an iron lance in his left hand, hi. 
right free to grasp his unconquered sword. His thighs wer< 
guarded with iron mail, though other men are wont to leavi 
them unprotected that they may spring the more lightly on thei 
steeds. And his legs, like those of all his host, were protectee 
by iron greaves. His shield was plain iron, without device o 
colour. And round him and before and behind him rode all hi 
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 ii 
every quarter reflected from iron. ' Iron, iron everywhere,' crie( 
in their dismay the terrified citizens of Pavia." ^ 

The interest in this description of ninth-century armour i 
that we learn that the short byrnie, not reaching below the hip.' 
was usual not only in the day of the great emperor, but in that c 
his great-grandson, Charles the Fat, to whom the Chronicle of S1 
Gall was dedicated. Greaves {ocreae, bainbergae) were evidentl; 
in full use when the description was written, but the thighs wer 
generally unprotected. That the sleeve is spoken of apart froE 
the byrnie as if it was a separate piece of armour is notable. Th 
description is borne out by a passage in the will of Coun 
Eberhard of Frejus, who in 837 leaves a helm with a hauberk 
a byrnie, one sleeve, and two greaves. Probably the sleev 
{manica) was only needed for the right arm, the left bein; 
guarded by the shield. 

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

^ Does "ferrea cristatus galea " imply that the helmet was a crested one, like thos: 
in contemporary Frankish drawings in MSS. ' 
2 Monachus Sangallensis, ii. § 26. 



32] GROWTH OF FEUDALISM 87 

rganisation introduced by Charles was to some extent kept up 
lay be deduced from an edict of Lewis and his son Lothar, 
ated 828, which orders the counts to inquire accurately whether 
11 the smaller landholders are properly enrolled in contributary 
roups for service in the host, such as had been instituted in 803.^ 
vnother document issued by Lothar at Pavia in 832 for his sub- 
ingdom in Italy, recapitulates the prohibition against selling 
lail outside the kingdom, and restates the old regulation that the 
older of twelve mansi must come to the host wearing a byrnie. 

The time of Lewis being one in which the central power was 
apidly growing weaker, and the independence of the local counts 
.rowing more marked, we cannot doubt that the mailed and 
lorsed retainers of these notables must have been continually 
•rowing in numbers and importance as compared with the 
narmoured infantry of the local levies. The perpetual civil wars 
/hich occupied the later years of Lewis' reign are so full of 
udden desertions and inexplicable changes from side to side on 
he part of large bodies of troops, that we see that the self- 
nterest of the counts has become of more importance than the 
general loyalty of their subjects. Docile obedience to the royal 
)an has been replaced by the most open treason. Owing to the 
:mperor's foolish liberality to his sons, the realm had four rulers 
.t once, and ambitious nobles could cloak their private schemes 
)y pretending to adhere to one or other of the rebellious young 
dngs. When the will of the local ruler became of more import- 
mce than that of the nominal head of the empire, the day of 
eudalism was beginning to draw nigh. Already in the time of 
Zharles the Great we find the counts accused of pressing hardly 
ipon the smaller freemen, exacting from them illegal impositions 
md services — misdemeanours against which the capitularies 
declaim again and again. Under weak rulers like Lewis and 
lis sons the evil was perpetually growing worse. At the same 
:ime, the other characteristic sign of feudalism, in its social as 
apposed to its political aspect — the commendation of an ever- 
-growing proportion of the smaller landholding classes to their 
greater neighbours — was steadily going forward. Probably the 
leavy 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- 

^ See Cap, Papiense, 832, § 15. 



88 THE ART OF WAR IN THE MIDDLE AGES [84c 

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

But the tendencies towards feudalism in the State, with th( 
corresponding tendency towards the depreciation of the nationa 
levies of foot-soldiery, would have been comparatively slow in it; 
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 th( 
direct result of the incursions of the Northmen. The lesse: 
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 th( 
Scandinavians. 



CHAPTER II 

THE VIKINGS (80O-9OO) 

HOSTILE relations between the peoples of the North and 
the Prankish kingdom had begun three centuries 
efore, on the day when Theudebert of Ripuaria slew Hygelac 
le Dane, the brother of the hero Beowulf, on the Frisian shore 
;i5). But it was seldom that Frank and Dane had met; the 
arrier of independent Saxons interposed between the two 
aces had always kept them apart. Down to the time of 
liarles the Great the Scandinavian peoples were mainly engaged 
1 obscure wars with each other. They are seldom heard of in 
he North Sea. But at last the Frankish power, with its wealth, 
:s commerce, and its Christian propaganda, swept over Saxony 
nd moved on its boundaries to the Eider. It was within a very 
e\v years of Charles' first conquest of Saxony that the Vikings 
Wickings, men of the shallow fiords that face the Cattegat and 
:'kager Rack) made their first appearance on the scene as 
erious disturbers of the peace of Western Europe. Perhaps 
he first seeds of trouble were sown when Witikind the Saxon 
led before the sword of the Franks and took refuge in Jutland ; 
ve need not doubt that he told his Danish hosts terrible tales 
)f the relentless might, the systematic and irresistible advance 
)f the iron king of the Franks. The danger was now at their 
ioors — 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 



I 



Norseman were struck by the latter. Strangely enough, 
earliest recorded Danish raids were not aimed against the realr 
of Charles the Great, but at more distant lands. The isolatet 
piracy of the " three ships from Herethaland " which burne* 
Wareham in Dorsetshire in 789 ^ is the first note of the appear 
ance of the Scandinavians on the offensive. Four and five year 
later two small fleets burned the rich abbeys of Lindisfarne an( 
Wearmouth on the Northumbrian coast. In 795 the Dane 
appeared so far west as Ireland, and destroyed the monasterie 
of Rechru on Dublin Bay. It was only in 799, ten years afte 
the descent on Wareham, that the first recorded raids of th 
Vikings on Prankish territory are noted. In that summer the; 
are said to have landed and made havoc both in Frisia and ii 
Aquitaine : the ever- watchful Charles was soon on the spot, an( 
ordered a fleet to be built to guard the narrow seas and th 
coast of Neustria. But the only serious trouble which th 
empire suffered from the Danes was a daring invasion of Frisi; 
by the warlike king Godfred in 810. With two hundred ships ii 
his train, Godfred overran the Frisian Isles and extorted fron 
their inhabitants a large tribute. He spoke in his hour o 
triumph of visiting the emperor at Aachen, but one of his owi 
men murdered him not long after, and his nephew and successo 
Hemming at once made peace with the Franks and sailec 
home ; the Danes were not destined to see Aachen till seventy 
six years later. The peace which Hemming promised was il 
kept, and several small raids on the northern coast of th< 
empire are recorded between 810 and 814. But these were al 
trifling matters. It was not till the reign of Lewis the Pious tha 
the Viking raids began to grow serious. During the later year 
of Charles, the favourite sphere of activity of the Vikings wa 
Ireland, where, from 807 onward, they were making sad havo« 
of the whole coast-line, and harrying one by one the ricl 
monasteries which lay along its bays and islands. 

During their first tentative raids the Scandinavians had no 
yet learned their own strength, nor were they such practisec 
marauders as they afterwards became. It is strange enough 
however, to see how suddenly they asserted themselves as a nev 
military power. At first they were sailing in unknown seas 

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



o] THE SHIPS OF THE VIKINGS 91 

id their ships were but long, light, undecked vessels, that 
;emed unfitted to face the wild Atlantic. That such craft, less 
lan twenty years after their first appearance in the North Sea, 
lould be risking their slight frames in rounding the rocky 
lores of Donegal and Kerry, is the most astounding proof of 
le wonderful seamanship of the Vikings. The boats were 
^sentially rowing, not sailing vessels ; their masts could be and 
ften were unshipped; they were only used when the wind set fair. 
or their propulsion the Viking ships relied on their oars, from 
m to sixteen a side, though a larger number was employed 
hen boat-building had become more scientific, in the tenth and 
leventh centuries : even a second tier of oars seems to have been 
ccasionally used in these later times. The prows and sterns were 
oth high and curved. The former were often fashioned into the 
ragon-shaped figure-heads which are so famous in the sagas. 
here was no helm, but the ship was steered by a long oar 
ashed near the stern, as is a Shetland sixern of to-day. The 
arly Viking vessels probably carried from sixty to a hundred 
nen — only the larger constructions of the tenth century could 
ontain as many as two hundred. 

The Danes, Swedes, and Norsemen of the year 800 were in 
. state of society very much resembling that in which their 
\nglian and Saxon kinsmen had come to Britain three hundred 
ears before. The raiders were not compact tribal bodies, but 
var-bands of adventurers enlisted under the banner of some 
loted leader, who was, as often as not, a mere warrior of renown, 
lot a member of one of the old royal houses. There are few 
J examples in the early Viking age of hosts commanded by the 
lational king, though the first notable raid — that which King 
jodfred 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,^ the main body must have been wholly unmailed. 

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



92 THE ART OF WAR IN THE MUDDLE AGES [81 

After gold and silver, helms and mail-shirts were the form ( 
plunder which the raiders most yearned for. This did n( 
endure for long: in less than two generations the Northmen ha 
armed themselves from the spoils of their enemies, and their ow 
smiths too had begun to essay the armourer's art. So essenti; 
was mail to the professional Viking, whose hand was again: 
every man, whose sole occupation was war, that by 850 or 9c 
it was the rule, and not the exception, in their hosts. Their bod 
armour seems to have been exactly of the Prankish model ; tY 
helm, however, was pointed and often furnished with a nasal, ui 
like the old semi-classical shape which had prevailed among tY 
Franks down to the ninth century.^ The shield was at fir: 
round, like those of most of the other Teutonic races ; it was onl 
in the tenth century that it took the kite-shape familiar to i 
in the Bayeux Tapestry and other contemporary works of ar 
Shields were often painted red or some other bright hue, and, hun 
on the bulwarks of the war-ship when the warriors were at se: 
produced lines of brilliant colouring along the gunwale. 

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

^ The helm with nasal, however, was probably known to the Franks in the nint 
century ; it was most likely the "helmum cian direct 0^^ of the Ripuaxian Code. 



;ol EARLY RAIDS OF THE VIKINGS 93 

orth were proud of their skill with the arrow ; it was held an 
)nourable weapon by them, while among their enemies it was 
e mark of the poorest military classes. Readers of the sagas 
ill remember the marksmanship of Olaf Tryggeveson and his 
:nchman Einar, and the celebrated shot with which King 
iagnus slew Earl Hugh the Proud on Menai Strait. 

It was only some time after their appearance in western 

aters that the Vikings acquired a complete ascendency over 

le peoples of the older Teutonic realms. They were at first 

uitious, attempting no ravage deep in the land, but absconding 

tor the plunder of some one seaboard town or abbey. The 

ranks, Irish, and English seem to have been more angered than 

nrified by the first raids, and several times caught and destroyed 

3nsiderable bodies of the invaders.^ But the fleets grew larger, 

le raiders more daring and better armed, their knowledge of 

le strong and weak spots of the line of defence more perfect. 

vbout forty years after the first plunderings in England, and 

lirty after the first assault on the Franks, Western Europe 

egan to awake to the fact that the Northmen were beginning 

D be no mere pest and nuisance, but a serious danger to Christ- 

ndom. The landmarks of this period are the first serious inva- 

ion of the interior of Ireland by a great host under Thorgils 

832), the plunder of the rich haven of Dorstadt and the famous 

athedral city of Utrecht among the Franks (834), and the 

rection of the first fortified Viking camp in England on the isle 

)f Thanet in 851. The invaders were beginning to grow so 

iumerous and so daring that it was obvious that some new 

neasures must be taken if their progress was to be checked. 

Among the faction-ridden tribes of Ireland it was hopeless 
o look for union or skilfully-combined resistance. More might 
lave been hoped from the English and the Franks. But the 
:ontemporary political situation of neither of those peoples was 
"avourable. 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 (S^S), 

' e.^^. the Northumbrians destroyed in 794 the band that had sacked Wearmouth. 
In 811 the Irish defeated a host in Ulster, and in 812 another in Connaught. 



94 THE ART OF WAR IN THE MIDDLE AGES [85 

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

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

^ 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 seerr 

to have been subject to Mercia : the defeat of the latter by the King of Wessex brougl 

about that rising of the East Anglians in which two kings of Mercia, first Beornwul 

then Ludica, perished. 



;o 



DECAY OF THE CAROLINGIAN EMPIRE 95 



le foolish tenderness of the emperor and the unfihal ambition of 
s children may have supplied the formal cause of disruption, 
> essential cause was the desire for independence on the part of 
le subject nationalities. In all the realm the Austrasians were 
le only people who consistently stood up for the cause of union 
id imperialism. The civil wars of the sons of Lewis had begun 
. 830, and for some time the ever-thickening Viking raids 
;emed to the statesmen of the empire tiresome diversions, 
istracting them for the moment from the all - important 
uestions whether Lewis should subdue his children or lose his 
irone, and whether his youngest son Charles should or should 
Dt obtain the kingly crown along with his brothers. Lewis 
ied in 840, after having seen the Danes cut deep into Frisia 
id push daring raids up the Meuse and the Loire. After his 
isappearance from the scene the civil wars only became more 
instant and more chaotic: the bloody battle of Fontenay (541) 
here the might of Austrasia was for ever broken, settled the 
ite of the empire. It was to split up permanently into inde- 
endent national kingdom.s, and never again was one sovereign 
ill to sway all the military force of the West, from Hamburg to 
Barcelona, for a common end.^ 

Now, from some points of view it might appear quite probable 
lat three or four compact national kingdoms would be better 
)le to cope with the Vikings than the vast but somewhat 
nwieldy empire of Charles the Great. But the dynastic 
iterests of the Carolingian house were still too strong to 
How real national States to develop themselves. Each king 
as snatching at his brother's or cousin's provinces, in a vague 
ope of reconstituting the empire for his own benefit. It was 
ot till the male line of the eldest son of Lewis the Pious died 
ut in Italy (875), and that of his second son in Germany (911), 
iiat those intermittent projects of reunion died out. As long 
s they lasted they were wholly evil : while Charles the Bald 
.as getting himself crowned at Metz or Rome, while Wido was 
verrunning Burgundy, or Carloman and Arnulf devastating the 
.ombard plain, the Dane and Saracen and Magyar were tearing 
heir realms to pieces behind their backs. Kings immersed in 
mperial politics could not find time to discharge the simple 
iuty of superintending the local defence of their own coast and 

* ^ Charles the Fat, though king of Germany, West Francia, and Lombardy, never 
uled in the Burgundies, so the above statement is Hterally correct. 



96 THE ART OF WAR IN THE MIDDLE AGES [85 

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

We must now investigate the tactics of the Northmen, an( 
the various expedients which their English and Frankisl 
adversaries employed against them. By the middle of th 
ninth century the invaders had increased into a formidabl 
multitude: their expeditions had been so fortunate that th 
whole manhood of Scandinavia had thrown itself into th 
Viking career. The Northmen were now members of old war 
bands contending with farmers fresh from the plough — vetera: 
soldiers pitted against raw militiamen. They were far bette 
provided with arms than their adversaries : the helm and byrni 
seem to have become universal among them, while the Englis 
/j/rd and the Prankish local levies were still mainly compose 
of unarmoured men. Only the thegnhood on this side of th 
Channel, and the counts and their retainers on the other, wer 
sufficiently well equipped to be able to face the invaders ma 
to man. With anything like equal numbers the Vikings wer 
always able to hold their own. But when the whole country 
side had been raised, and the men of many shires or countship 
came swarming up against the raiders, they had to beware les 
they might be crushed by numbers. It was only when a fleet c 
very exceptional strength had come together that the Northme 
could dare to disregard their opponents, and offer them battl 
in the open field. Fighting was, after all, not so much thei 
object as plunder, and, when the landsfolk mustered in ovei 
whelming force, the invaders took to their ships again and saile 
off to renew their ravages in some yet intact province. The 
soon learned, moreover, to secure for themselves the power c 
rapid locomotion on land : when they came to shore they wouL 
sweep together all the horses of the neighbourhood, and mov 
themselves and their plunder on horseback across the land. T 
fight as cavalry they did not intend : it was only for purpose 
of swift marching that they collected the horses. The firs 
mention of this practice in England comes in the year S6( 
when "a great heathen army came to the land of the Eas 



I 



66] THE FORTIFIED CAMPS OF THE VIKINGS 97 

\ngles, and there was the army a-horsed." ^ Curiously enough, 
t is in the same year that we first hear of the Danes in the 
^>ankish realm ^ trying the same device. Their base of opera- 
ions, however, was of course their fleet, and such excursions 
ilways ended in a swift return to the boats. It was only when 
L waterway was not available that the raiders dared to cut them- 
selves adrift from their vessels. As a rule, their method was to 
vork up some great stream, sacking the towns and abbeys on 
:ach shore of it ; when they got to the point where it was no 
onger navigable, or where a fortified city stretching across both 
3anks made further progress impossible, they would moor their 
diips or draw them ashore. They would then protect them with 
1 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 Neustria 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 861. 
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. 

^ Annales Bertinenses, p. 84: "Nortmanni circiter quadringenti de Ligeri nun 
caballis egressi, commixti Britonibus Cenomannis civitatem [Le Mans] adeunt." 

7 



I 



98 THE ART OF WAR IN THE MIDDLE AGES [891 

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 
Ethelred 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 89 1 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 theii 
camp and forced to fight without protection, they would draw 
themselves up in the best position they could find, on a steef 
hillside, as at Ashdown (871) or Ethandun (SyS), or behind £ 






Ti] TACTICS OF THE VIKINGS 99 

tream ; they formed their shield-wall/ and fought the matter out 
o the end. On many occasions, when broken in the open by the 
:harge of the Prankish horse, they would retire behind the 
learest cover, — a village, as at Saucourt (881) ; a church, as at 
orisarthe (866); a large building, as in the fight in Frisia in 873, 
—and there hold out till they either beat off the enemy, were 
hemselves cut to pieces, or at nightfall were able to abscond. 

Nothing shows better the stubbornness of the Danes than 
he way in which they often by a desperate rally repaired a lost 
:)attle. At the great fight in front of York in 868 they were 
horoughly beaten by Osbert and Aella, and forced back on 
;he 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.^ The same had 
ilmost 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 
)ii 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 
iiard 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 

^ 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. 



loo THE ART OF WAR IN THE MIDDLE AGES [900 

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

We must now turn to a consideration of the methods bj 
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 o 
Egbert had to face was that the half-armed local levies of th( 
fyrd or the ban were insufficient to cope with the invaders. Th( 
Prankish counts and the English ealdormen made many i 
gallant attempt to beat off the raiders : sometimes they wen 
successful, but much more frequently they suffered a disastrou: 
defeat. The Vikings were too well - armed, too wary, toe 
experienced in every shift of war, to be adequately faced bj 
the raw militia opposed to them. Some more efficient bod} 
of troops had to be improvised to meet them, some system o 
defence devised to keep them from overrunning the opei 
country. Down to the ninth century the Frankish towns 
unless they had old Roman walls, were not provided with an^ 
systematic protection ; the English were even more exposed 
for such of them as had the Roman circumvallation had allowec 
it to moulder away ever since the first conquest,^ while thos 
which had arisen since Roman days had never been fortifiei 
at all. 

1 Vork, for example, the greatest centre of Northern Britain in Roman days, vvj 
in 867, in the words of Asser {sub ann. 867) imperfectly protected, for "non enii 
tunc ilia civitas firmos et stabilitos muros eo tempore habebat " ; therefore tl 
Northumbrians were able "murum frangere" by a rush — to hew down a palisai 
suppose. Canterbury seems to have had walls rather early, however. 



I 



CHAPTER III 

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

THE military history, therefore, of the ninth century shows 
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 two 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. Portifications 
are good things in their way, but they need trustworthy 
garrisons. The most elaborate entrenchments serve no end — 
as King Lewis of West Prankland found in 881 — 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.^ 

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, 

^ Annales Berlinenses, 881 : "Quod magis ad munimentum paganorum quam ad 
auxilium Christianorum factum fuit, quia ipse rex Hludovicus invenire non potuit cui 
illud castelluni ad custodiendum committere posset." 

101 



I 



102 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. We have 
already seen that by the year 800 both the Frankish and the 
English realms possessed an aristocracy, originally dependent 
on the kings, and wholly official in character — a "nobility oj 
service," to use the phrase that we have already had so many 
occasions to employ. On the Continent it now included no1 
only actual holders of countships or great offices about the 
court, but large numbers of persons, both lay and clerical, whc 
held " beneficia," feudal grants of land, from the king. Each o: 
these counts and vassi of various sorts had his bands of persona' 
followers, landed or unlanded, homines casati^ or sub-tenant' 
with holdings of various size. The vassal-class was steadil) 
growing: a family which had once held office and receivec 
grants of " beneficia " did not drop back into the ranks of th( 
ordinary freemen. The class, too, was already tending tc 
encroach on its poorer neighbours ; the counts were using theii 
official position, the holders of " beneficia " their less legal bu 
equally efficient powers of bringing pressure to bear on tht 
smaller men. Above all, the Church was extending it; 
boundaries on every side so rapidly, that, as early as 831, Lothar 
the son of Lewis the Pious, began special legislation against th( 
handing over of land to the " dead hand." When the hideou: 
distress caused by the Danish invasions came to aid the alreadj 
existing tendency towards feudalisation, the result was easy t( 
foresee. By the end of the tenth century the vast majority o 
the smaller freemen had passed under the control of thei 
greater neighbours, either by voluntary commendation, or as th( 
result of deliberate encroachment. 

Nor were the Danish invasions less powerful in hastening 
the development of the other side of feudalism, the establishmen 
of the counts and dukes as hereditary local potentates, wh( 
practically could no longer be displaced by the crown. Ther« 
was an obvious convenience during the time of trouble in lettinj 
the son succeed to the father's government ; none would knov 
so well as he the needs and capacities of the district in whicl 
he had been brought up. Moreover, there was danger, in thos« 
days of incessant dynastic war, in the attempt to remove \ 
powerful noble from his father's post ; he might at once transfe 



36o] CREATION OF THE GREAT FIEFS 103 

lis allegiance to some other member of the Carolingian house. 
Charles the Bald and his short-lived successors habitually 
oought respite from the peril of the moment by letting the son 
succeed to his progenitor's office. In the next generation, the 
jounties 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, ^6j ; in Anjou, 
870 ; in Gascony, 872 ; in Burgundy, 2)y'j ; in Auvergne, ^2)6, 
In East F>ancia, 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 
Saxonum 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 Liineburg 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 Prankish 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 



I04 THE ART OF WAR IN THE MIDDLE AGES [866 

fought, as the chronicler remarks, " equestri praeHo quia uterque 
Gothus esset."^ 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," ^ because it is not usual for the Frankish 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 
Austrasia to snatch territory from his nephew Lewis, he is made 
to exclaim that "his army was so great that their horses woulc 
drink up the Rhine, so that he might go over dry-shod." ^ 

The definite date at which we may set the permanent 
depression of the infantry force in West Francia, is in S66 
From this year dates the celebrated clause in the Edict of Pitres 
in which Charles orders that every Frank who has a horse, or ie 
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," ^ and no one ir 
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 tc 
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 of ter 
thousand horse and six thousand foot. The chronicler Richer 



^ Vita Hhidovici, § 33. - Ann. Field. 891, 

3 Ann. Fuld. 876. ■* Edict of Pitres, 2. 26. 



i 



8] THE ADVANTAGES OF FEUDALISM 105 

ho tells of this levy, calls the cavalry milites, as opposed to the 
ot-soldiery,/^</^"/^j.^ This is the first indication of the use of 
e word miles, the warrior /^r excellence, for the mounted soldier, 
few years before, it would have been applied to all fighting 
en ; we now see it starting on its way to become the designa- 
3n of the knight of the later Middle Ages. By the time that 
le tenth century has arrived, the infantry in West Francia 
;em wholly to have disappeared ; in such battles as the bloody 
3ld of Soissons, where King Robert was slain, both armies, 
ithout exception, seem to have been composed of mounted 
len. 

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

The other expedient which the Franks used against the 

^ " Odo congregari praecepit milites peditesque : quibus collectis in decern millibus 
aquitum peditum vero sex millibus erat," etc. (Richer, § 81). 

^ 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 

Northmen was the systematic and elaborate fortification] 
points of vantage. The deHberate adoption of this policy 
laid down in the same Edict of Pitres (866), which we hav 
already had to quote for its importance in the development c 
cavalry. But the actual scheme had been begun as early as 86: 
It had occurred to Charles the Bald that the Danish fleets migh 
be kept from running up the rivers by erecting at favourable spot 
fortified bridges, through which they would be unable to fore 
their way up stream. Pitres, some miles higher up the Sein 
than the Viking stronghold on the isle of Giselle, was the chie 
point which he pitched upon. Here he began to build a grea 
bridge with tetes-du-pont at either end ; it took some years t 
complete, and the Danes still dashed through its unfinishe 
centre when they chose. He therefore constructed anothe 
less ambitious bridge higher up, at Trilbardou, and by means c 
it blocked the return of the raiders. After trying to brea 
through in vain, Weland, the Northmen's chief, gave up hi 
prisoners and plunder, on condition of being allowed to dro 
down stream under the bridge unmolested.^ The great structur 
at Pitres was finished in 866, and smaller ones at Auvers an 
Charenton-le-Pont were erected to guard the Oise and Marnc 
as additional precautions. Most important of all, Charles mad 
the island-city of Paris throw bridges across to the norther 
and southern banks of the Seine. These structures wer 
destined to have more influence on the future of the Vikin 
invasions than any of the new buildings down stream. For th 
weak point of the plan was that the new bridges require 
garrisons, and that a permanent force to hold them was hard t 
find. A city like Paris could find men to man its own defence: 
but isolated fortifications, like those at Pitres, required specie 
bodies of troops, which were not always at hand. Apparentl) 
they were broken through during the civil wars at the end c 
the reign of Charles. At any rate, we find the West Franks i 
885 devoting all their attention to building, as a substitute fc 
them, a new fortification at Pontoise. When the Danes cam 
up the Seine for the great siege of Paris, they had first to destro 
this obstruction. It made a creditable resistance, but, gettin. 
no succour from without, was compelled to surrender.- Thei 
pushing up to Paris, the invaders began the eleven montb 
beleaguering of the place. Paris had been more than once i; 

^ Annales Bertinenses, 862. - Annals of. St. Vedast, 885. 



,oo] FORTIFIED BRIDGES 107 

/iking hands before Charles the Bald fortified it,^ but now its 
lew defences enabled it to make a very different resistance. 
ts gallant defenders, Odo and Bishop Gozelin, held it against 
:very attack, though the Emperor Charles the Fat gave them 
ittle or no help. It is true that the Danes ultimately succeeded 
n getting up the river, by laboriously dragging their vessels 
icross the flat sl^ore round the southern bridge-head.^ But they 
:ould not take the place, and were at last glad enough to receive 
I bribe and depart, leaving Paris free [886]. This successful 
iefence was almost as great a landmark in the history of West 
^Vancia as the victory of Ethandun in England, or that of 
.ouvain in Austrasia. 

The Danish ravages in Germany are of little importance 
ifter the year 900 ; in the Western realm they con- 
inued much later, but were never so threatening again as 
;hey had been in the years before 8S6. For the future, the 
Prankish victories are almost as numerous as those of the 
Northmen. The fights of Montfaucon (888), Montpensier (892), 
md Chartres (911), are all worthy of notice. They show that 
:he F>anks were now no longer wont to shirk the ordeal of 
Dattle, as they had been thirty years before, but fought 
ivhenever they had the chance. As often as not they 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 SS6, or Sens in S8y, 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 
])laces with all that was worth carrying away. The refuges and 
strongholds were now numerous enough to afford shelter to the 

' It had been plundered in 845 and 856. 
Me^z Annals, 888, and Abbo. See pp. 14 1-6 for a detailed narrative of the siege. 



io8 THE ART OF WAR IN THE MIDDLE AGES 



I 



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

It was the increasing difficulty and barren results of the 
raids in France which led the Danes of Rolf in 911 to come t 
the same bargain with Charles the Simple which the Danes c 
Guthrum had made with Alfred of Wessex in 878. When th 
king offered them a great Danelagh (as the English woul 
have called it), reaching from the river of Epte to the Wester 
Sea, Rolf and his followers accepted the bargain, and agreed t 
draw together, settle down, and make a peace with the Frank; 
Contrary to what might have been expected, the settlement wa 
on the whole a success from the point of view of Charles th 
Simple. Gradually all the other Danish bands, leaving th 
Loire and the Garonne mouths, gathered in to settle along wit 
Rolfs men. Like Guthrum in England, Rolf in Normandy wa 
a more faithful vassal than might have been expected, and eve 
sent his bands on several occasions to help the king againe 
native rebels. It was only when Charles had fallen into th 
deadly snare of Count Herebert of Vermandois that th 
Normans were turned loose again on the land (928). Th 
Franks proved now well able to defend themselves, and Kin 
Rodolf cut to pieces at the battle of Limoges (929) the ho.^ 
that tried to open once again the old route of the raiders int 
Aquitaine. From the time of William Longsword onward, th 
Normans appear no longer as heathen invaders from withou 
but as unruly vassals within. By the year 1000 they may fc 
most purposes be regarded as assimilated to their neighbour 
and Normandy is but the most important fief of the Frenc 
crown. 

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

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



do] the ENGLISH THEGNHOOD 109 

J force of horsemen when the Viking raids began ; Ecgbert's 

my was in this respect wholly unlike that of Charles the 

reat. There was no question of reinforcing the cavalry arm 

1 England, for no such force existed. But in other respects 

e find the Prankish methods reflected, with some variations, 

n this side of the Channel. If Wessex had no mailed horse- 

len to serve as models for the reorganisation of the whole host, 

le had heavily-armed foot-soldiery. The "gesithcund man 

olding land," as Ini would have called him, the "thegn," as 

le laws of Alfred name him, was practically equivalent to the 

assiis or holder of a beneficium of the Continent. As among 

le Franks the tendency of the ninth century was to drive all 

len into the feudal hierarchy, — the more important freeholders 

ecoming vassals, the less important serfs, — so in England the 

middle classes tend to be divided in a similar way. The richer 

eorls are absorbed into the thegnhood, the poorer sink into 

abjection to their greater neighbours. In the laws of Alfred 

: is easy to detect the fact that the free middle class is far less 

)rominent than it had been even in the time of the laws of Ini.^ 

There were already "hlafords" and dependants in the day of 

he elder code ; by the day of the later they must have been 

he most important part of the population. How the change 

ame about may be gathered from the two important but 

I monymous documents of the early tenth century, the one 

lealing with Weregelds, the other with " The People's Ranks 

md Laws," printed on pp. 79-8 1 of Thorpe's Early English Laws. 

■ n the Weregeld document the first draft states that " if a ceorl 

:hrive 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 utware, he is nevertheless a ceorl. But if his son and 

ion's son so thrive that they have so much land afterwards, 

;he offspring shall be of ' gesithcund ' race, and the weregeld 

2000 thrymsas."'^ 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] he is sit/tcund, etc. etc." ^ 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 

^ See Alfred's Laws, i and 37, particularly the latter. 

^Weregeld Document, 9, 10, 11. 

'Weregeld Document, 2nd version, 9, 10, 11. 



no THE ART OF WAR IN TPIE MIDDLE AGES [9. 

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

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

The chief charge of the thegnhood was, of course, the du 
of following the host in full mail whenever the king took t 
field. At all costs it was intended to raise the proportion 
well-armed men in the army to a maximum. It is worth notii 
that we find, in the " Ranks and Laws " document, sub-tenar 
holding under a " hlaford " who have reached the assessment 
wealth necessary to qualify for gesithcund rank : though n 
directly sworn to the king, they are yet reckoned part of t 
thegnhood, being called " medial thegns." ^ 

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

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

1 Ranks and Laws, § 2. ^ Ranks and Laws, § 6. 

^ A phrase to be found in Canute's Heriot-law, Leges C. § 72. 
^A.S. Chronicle, 894. 



)7] THE VIKINGS TURNED BACK in 

England being far behind tliat of the Continent, — they seem 
. have been very effective in checking ravages. In a few cases 
)lid masonry seems to have been used — for example, in patching 
D the Roman wall of London, which Alfred " Aonorifice restaur- 
vit^ in 887. Alfred's warlike daughter Ethelflaed followed his 
sample in this respect at Chester in 907, where her rude repairs 
m still be discerned among the Roman masonry. Canterbury, 
)0, had walls very early. But it was mainly by stake and foss 
[ concentric rings, enclosing water-girt mounds, that Alfred and 
is children protected their frontier. Edward the Elder worked 
gainst the Danelagh with such strongholds in a most systematic 
ay. His first line of burhs was to guard his own border, but 
radually he and his sister Ethelflaed pushed forward a second 
ne of forts of offensive purpose. These ii:i7nyj6ij.ara^ as a Greek 
ould have called them, were built opposite every Danish town, 
nd furnished with garrisons to contain the sallies of the inhabit- 
ats and hold down the neighbourhood. Hardly one fell in 
venty years of war, so ineffectual were the siege operations of 
le Danes. 

It would seem that the system by which the burhs were 
laintained was somewhat like that which Henry the Fowler ^ 
stablished in Germany a few years after Edward had begun, his 
y^stem of fortification. To each burh was allotted a certain 
umber of hides of the surrounding region, and all the thegns 
isident in that district were responsible for the defence of the 
tronghold. Each of them was bound to keep within the palisade 
f the burh a house, which he must either inhabit himself, or fill 
ath a trustworthy representative able to bear arms in his stead, 
.^hus the original inhabitants of the burhs were a race of warriors, 
hough in later years, when the land settled down into quiet, and 
Dwn houses grew to be valuable property, the thegn might let 
lis tenement to a merchant or craftsman whose primary occupa- 
ions were not warlike. But in the early ninth century the burh- 
aen were essentially military in their pursuits. It would seem 
hat the cnihten-gilds, as we find them at Cambridge, London, 
nd elsewhere, were the original association of the settlers, who, 
oming in from all sides to hold reconquered land, had no 
ommon local tradition, and had to start new bonds of unity 
imong themselves.^ 

^Asser, 887. 2 See p. 120. 

^ All these suggestions I get from Professor Maitland's invaluable Doiiiesday Book 



112 THE ART OF WAR IN THE MIDDLE AGES [89 

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

Far better, however, than any mere fortification of th 
inland was the third great plan which Alfred adopted for brin^ 
ing his Danish wars to a successful conclusion. He began t 
build a strong fleet, able to contend at sea with the Vikings. I 
the very first years of his reign he had seen that this was th 
one really effective way of keeping the coast secure. As earl 
as ^j6, long before the peace of Wedmore, he gathered a fe 
ships and chased off a small raiding squadron.^ After he ha 
gained some leisure by the peace with Guthrum, he kept coi 
tinually enlarging this force ; by 885 he had apparently son- 
dozens of ships afloat, though not enough to cope with the mai 
Viking fleets.* Later, as the Chronicle tells us, he built " lor 
ships that were full nigh twice as long as others ; some had sixt 
oars, some more ; they were both swifter and steadier, and ah 
higher than others, and they were shaped neither as the Frisia 
nor as the Danish vessels, but as it seemed to himself that the 
might be most useful." The first successful doings of the ne 
squadron are recorded under the year 897. The nucleus of 
well-built fleet was perhaps the most precious legacy of all th; 
Alfred left to England ; his son steadily increased it. In 91 

and Beyond. The " Burgal Hidage " which he gives in full, seems to belong tc 
period early in Edward's reign, when the reconquest of Mercia and Essex was ji 
commencing. It has very full details of the division of all the shires south of Tham 
into districts depending upon burhs, but becomes incomplete as we advance into t 
regions which were beginning to be reconquered from the old enemy. There the syste 
was but just being built up. 

^ A.S. Chronicle, 896. ^ gee p. 106. 

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



)05] CAMPAIGNS OF EDWARD THE ELDE:R 113 

Edward was able to send out some hundred ships to guard 
he coast of Kent ; twenty years later the navy was so large and 
;o well practised, that ^thelstan, Alfred's grandson, was able to 
:oast up the whole eastern shore of Britain unresisted, to invade 
:he domains of Constantine, King of the Scots.^ The Danes of 
N^orthumbria were in rebellion at the time, but they were 
evidently unable to launch any squadron large enough to molest 
lis armament. 

Among the Franks, then, mailed cavalry and systematic 
fortification, among the English, mailed infantry, well-built 
ourhs, and a fleet, ultimately succeeded in curbing the raids of 
:he Northmen. It must not be forgotten, however, that to a 
:ertain extent this triumph of the defensive over the offensive 
uas 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 hcTmes 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 

^ A.S. Chronicle, 933. 



114 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 yean 
later caused all the English Northmen to take King Edward " tc 
father and lord," after he had gradually subdued East Anglic 
and the " Five Boroughs " [924]. 

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

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



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.^ 
Their glorious end was to fall to the last man fighting round the 
Dragon banner of VVessex, 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 nsXs-/.v(, 



^ In the battle against Macbeth there were slain " Osbern 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 
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,^ 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 theii 
previous homes on the Bug and Dnieper, came flooding through 
the passes of the Carpathians into the valleys of the Theise 
and Danube. The Avars had long sunk into nothingness, anc 
the Slavs who had succeeded them on the Middle Danube seerr 
to have been perfectly helpless before the invaders. So thf 
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. As 
the chronicler Regino observed, " no man could stand against 
them if their strength and their perseverance were as great as 
their audacity." ^ But they were incapable of besieging a walled 
town, or of standing firm in the shock of hand-to-hand fighting. 
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 Frankish 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." 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 

^ Regino, 889, i. 600. 

2 " Missos illorum sub dolo ad Baioarias pacem optando, region em illam ad explo- 
randum transmiserunt " [Ajtn. Fttld. 900). 



ii8 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 w^as an isolated and a w^hoUy insufficient 
precaution, but the only one w^hich 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, wath 
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.^ 

Pitched battles, however, were rare in the Hungarian wars, 
for the raiders were more set on plunder than fighting. Nor 
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 

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



)54] THE GREAT RAIDS OF THE MAGYARS 119 

vould come suddenly galloping out of a wood on to the prey : 
)nly the smoke and the nightly sky red with flames showed 
vhere each of their troops had been." ^ 

It was their rapid movement, far swifter even than that of 
he Danes, which alone made the Magyars formidable. The 
vide sweeps which some of their expeditions made far exceed 
n length any Viking raid. The most formidable of all were 
hose 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 
rom the Ardennes, and returned across Franconia to the Danube. 
[n 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, 
md, 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.^ 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 
niade 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 (955).^ 

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 

^ Ekkehard, c. 52. ^ Flodoard Amt. 924. 

^ For this raid see Witikind, iii. § 30, and Cont. Regino, 954. 



I20 



THE ART OF WAR IN THE MIDDLE AGES 



[92 



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

It was not till the advent of Henry the Fowler (or Henry th 
Builder, as contemporaries more wisely called him) that an 
check was set to the Magyars by either of the necessar 
expedients. Henry from his first accession showed himself 
far more powerful prince than his unfortunate predecessor 
Conrad of Franconia and Lewis the Child ; but it was not ti 
he had been five years on the throne that he found leisure t 
devise a system of defence against the invaders. Having i 
924 concluded a truce with them, on the ignominious terms c 
paying a large " Magyargeld " (if we may coin the word b 
analogy from " Danegeld " ), he set to work to garnish th 
frontier with new fortresses. In Saxony and Thuringia h 
made every ninth man of the agrarii milites — i.e. all men i 
the countryside liable to the ban in time of need — remove int 
a walled place. He set the whole population to work da 
and night to build these strongholds, and to construct house 
inside them : these being finished, he settled that each nint 
man should dwell therein, and take care of the eight neighboui 
ing houses which his companions were to occupy in time of wa 
while the eight were to pay the indweller in return one-third c 
the net products of their lands.^ All the legal and festal mee* 
ings of the district were to take place inside these new fortifie 
places, so as to induce the population to haunt them as much a 
possible. Among these foundations were Merseburg, Quedlir 
burg, Goslar, Nordhausen, Grona, and Pohlde. Henry als 
compelled the abbeys to wall themselves in, and repaired th 
fortifications of the older centres of population which date 
back to the burgs of Charles the Great. At first the ne^ 
strongholds were little more than thinly-inhabited places c 
refuge, but ere long most of them became real towns. Th 
founding of Merseburg, the easternmost and the most expose 
bulwark of Saxony, deserves a special notice. Henry peopled 
by sparing the life of every " strong thief" that he caught, o 
condition that he should go to dwell at Merseburg and receive 
^ All this is told very elaborately in Witikind, i. 35. 




^Ss] THE BATTLE ON THE UNSTRUT 121 

rrant of land in its environs. Strangely enough, this " legio 
:ollecta a latronibus," as the chronicler calls them,^ did very well 
n their new settlement, and, like Romulus' robber band, made 
:heir 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- 
Dack. We are unfortunately without information as to the 
neans 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 Weser. 
We only know that when the next Magyar raid came, in 933, it 
found North Germany for the first time possessed of " milites 
squestri praelio probatos," 2 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), 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 

^ Witikind, ii. 3. 2 /^^-^^ n 33. 



122 THE ART OF WAR IN THE MIDDLE AGES 

other in the midst of a great civil war, and with the traiton 
co-operation of the rebels. For the greater part of the ec 
years of his reign (936-955) the realm was fairly free from raids^ 
we 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.^ Saxony, safe 
behind its new line of fortresses, seems to have held its owr 
without difficulty .- 

The great Magyar invasions of 954 and 955 were a lasl 
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.^ 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 tc 
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 Swabia, with a thousand Franconians, and the same number oi 
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 intc 
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 themselve.' 
(Witikind, ii. § 14). 

^Gerh. V. Oudalr. §12. 



5] THE BATTLE OF THE LECHFELD 123 

sed of cavalry, and each mustering about one thousand men. 
iree " legions " were Bavarian, two Swabian, one Franconian, 
iC Bohemian ; the eighth was composed of the king's personal 
llowing and of picked men from the other divisions ; it was 
mewhat larger than the rest. The army was small compared 
;th that which had accompanied Otto on his invasion of France 
946, when (as he boasted) " thirty-two legions had followed 
m, every man wearing a straw hat," — for in the summer heat 
e Germans had marched unopposed through Champagne with 
eir helms at their saddlebows, and the peaceful headgear of 
raw shading their brows.^ 

On hearing of the king's approach, the Hungarians hastily 

ised the siege of Augsburg, and drew themselves up on the 

'oad and level Lechfeld, a region very well adapted for the 

-actice of their usual Parthian tactics. Otto, however, moved 

) meet them through broken ground which was unsuitable for 

leir manoeuvres, and then camped by the side of the Lech. 

[e drew up his army in a single line of corps, his own chosen 

and in the centre, on its right the three Bavarian " legions " and 

lat of the Franconians, on his left the two Swabian divisions. 

^he Bohemians, whether because their loyalty was doubted or 

ecause they were considered less solid troops, were placed behind, 

1 charge of the baggage. They were a camp-guard, not a reserve. 

The Magyars soon came in sight — a confused weltering mass 

f hundreds of small troops ; the German chronicles say that they 

'ere a hundred thousand strong, and, however exaggerated the 

gure may be, they no doubt many times outnumbered Otto's 

ost. They had crossed the Lech far sooner than had been 

xpected. Their first manoeuvre was characteristic : while some 

'f them threatened the German front, a great body slipped off 

o the left, apparently unseen, and suddenly fell upon Otto's 

amp. The Bohemians left there on guard were routed after a 

hort struggle. The Magyars then suddenly changed their 

lirection, and charged in upon the rear of the two Swabian 

:orps of the king's left wing. Taken by surprise by this attack 

"rom an unexpected quarter, the Swabians were defeated, and 

Iriven towards the German centre : Otto then sent the Franconian 

:orps from his right wing to aid them. Led by Duke Conrad, a 

^ Witikind, iii. § 2. The straw hat was a specially Saxon head-dress for summer 
-ear. See the passage from Rather of Verona, quoted in Pertz's edition of Witikind, 
'• 451- 



24 THE ART OF WAR IN THE MIDDLE AGES 






lately pardoned rebel who had to win back his reputation f( 
loyalty, the Franconian horse charged with such a fierce shoe 
that the Magyars were completely routed, and fled in disord 
to join their main body. Otto meanwhile, with his own divisic 
and the Bavarians, had been watching and containing the rest 
the Magyars. When he saw the horde which had turned his flar 
crushed by Conrad, he hastily rearranged the disordered le 
wing, and ordered a general charge of his whole line.^ 

The Magyars, dismayed by the disaster which had befalk 
their detached corps, made a poor resistance. They we 
indeed wholly incapable of standing up to the Germans man ' 
man : their horses were smaller, and very few of them wore ar 
defensive armour.- After letting fly a few volleys of arrows, the 
wheeled off and fled. Many were overtaken and slain, for the 
horses were fatigued by the first fight ; more were drowned 
the Lech, for its farther bank was steep, and they could n' 
readily climb the slippery slope ; they had easily descended it ; 
they attacked, but found it almost impossible to mount c 
their retreat. 

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

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

^ Thietmar is apparently wrong in making the battle last two days ; in Witikind t 
whole of the fighting takes place on St. Lawrence's Day, August 10. 

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

3 Wit. iii. 49. 



, 5] THE MAGYARS TURNED BACK 125 

ls a desperate final rally. Just as in the history of the Otto- 
m assaults on Christian Europe we place the real moment of 
jatest danger during the siege of Vienna in 1529, not during 
at in 1683, so the most threatening time of the Magyar attack 
IS undoubtedly in 933, when they had never yet received a 
eck of importance, and not in 955, when they had already been 
it and turned back many times by Otto and Otto's generals. 

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



CHAPTER V 

ARMS AND ARMOUR (80O-I lOO) 

WE have seen that down to the time of Charles the Grei 
there had been comparatively little alteration in tt 
character of arms and armour since the days of the first found; 
tion of the Teutonic kingdoms in the fifth century. In the nint 
century, however, we find a gradual change coming over the out( 
appearance of the warriors of Christendom. Not only do 
much greater proportion of them wear defensive arms, but t\ 
arms themselves begin to change in appearance. All the alter.- 
tions are in the direction of securing greater protection for tt 
wearer. The short byrnie reaching to the hips and the ope 
Prankish helm seem to have been regarded as insufficient again; 
the Danish axe and the Magyar arrow. 

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

^ Vol. V. p. 67. But their date is much later than Viollet supposed. 

126 



)oo] THE HAUBERK 127 

leadplece no longer protected the skull. The lower edge of 
.he hauberk was sometimes tucked under the upper edge of the 
)yrnie 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 
.enth and eleventh centuries. The poorer men had only the 

I ihort mail-shirt, the richer supplemented it by the hauberk. We 
ind clear traces of its use in incidents such as that at the battle 
)f Soissons in 923, where King Robert, to make himself known, 
' pulled out his long beard from under its covering," ^ that the 
memy might see it. So, too, Duke Conrad at the Lechfeld 
•eceived a mortal arrow-wound in the throat, because, overcome 

, oy the heat, he had loosened his hauberk to take the air in 

: :he moment of victory.^ 

The next step in the development of this piece of armour 

hvas 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 
:ill the end of the eleventh or beginning of the twelfth century. 
Most of the warriors of the Bayeux Tapestry wear mail-shirts not 
oined to their hauberks, for in several representations of byrnies 
lot in actual use we see that they have no hoods. When in the 

I :welfth century hauberk and byrnie became one, the name of the 

! brmer was often used to cover the whole suit — a fact which has 

I :aused much confusion to those who, knowing the term in this 
ate use, have not seen that it was at first a mere cheek-guard 
langing from the helm. 

The helm itself changed entirely in shape in the ninth century. 
The open crested Prankish helmet with its peak disappears, and 
s superseded by a crestless conical headpiece. The latter shape 
s better for turning off sword or axe blows, but it is probable 
:hat it came in not merely for that reason, but because it could 
3e 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 

^ " Barbam obvelatam detegii, 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. 



128 THE ART OF WAR IN THE MIDDLE AGES 






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

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

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

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

From this short sketch it can easily be seen that the warric 
of 1050, with conical helm and nasal, hauberk covering his eai 
and throat, and long mail-shirt reaching below the knee, Wc 
entirely different in appearance from the Carolingian fightin 
man, who still preserved a certain resemblance to the late-Roma 
soldier. He was also, it must be owned, more effectively arme( 
if less sightly to look upon. The covering of ring-mail was nc 
yet growing so heavy as to incommode or fatigue the wearer. 

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

^ Helmum cum directo occurs in the Ripuarian Laws. " Richer, i. 46. 



looo] AXE AND SWORD 129 

rider,^ 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 
Dn 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 scutum 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 1 171 
the Danes who fought Strongbow's Normans at Dublin had the 
round red shield which their ancestors had carried three hundred 
y^ears before.^ 

Offensive arms did not alter their shape nearly so much as 
defensive during the years 800-1 100. The double-handed axe, 
IS we have already seen, was introduced by the Danes, and 
idopted 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 
Df Poictiers' description of the battle of Hastings, as hurled by 
the English at the oncoming Normans.^ The sword grew 
decidedly longer, and had by 1050 received a rounded point 
nstead 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 
vvas 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 
t most. The Danish blood of the Normans accounts for the 
arge proportion of archers whom they employed at Hastings. 
Neither the Germans, the English, nor the French seem to have 
:aken to it kindly. Abbot Ebolus, the defender of Paris in 886, 
is the only notable archer among these peoples who occurs to my 

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

" Giraldus Camb., Exp, Hib. i. § 2i: '* Clipeis quoque rotundis et rubris, ferro 
nrculariter munitis." 

3 W. P. 201. 



I30 THE ART OF WAR IN THE MIDDLE AGES [hoc 

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

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



CHAPTER VI 

SIEGECRAFT — A.D. 80O-IIOO 

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 
listory of any other province of the military art. When we read 
the account of Witiges' siege of Rome in 537, of the beleaguering 
Df Gundovald Ballomer in Comminges in 585, of Wamba's 
:apture 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 
Df 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 — 
vve find Witiges and Godfrey of Bouillon relying on exactly the 
iame methods. When the rude expedients of striving to fill the 
iown-ditch and swarm up the wall on ladders do not avail, the 
besieger in each case falls back on two main resources. The 
Dne is that of breaching the fortress with rams, the other that of 
:learing 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 [ 

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 



8oo] 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 (teretriis^ terebrus^ terebrd) 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 musculus, the mouse,^ because 
its object was to gnaw a round hole in the lower courses of the 
rampart. 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)-- 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 vulpes 
is less commonly used for it : ^ in this case, as in that of musculuSy 
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 

^ 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." 

^ Albert of Aix uses it in his account of the siege of NicKa, 1097. 



I 



134 THE ART OF WAR IN THE MIDDLE AGES [80c 

reminding them of the oven of everyday life. Mines were o: 
course very effective against places built on soft soil : the diggen 
could work undisturbed by the storm of stones and darts froir 
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 o 
it, so that its exact locality and direction could not be easil} 
discerned, he had a very fair chance of success. For an earl} 
example of the mine in use on this side of the Channel we ma} 
turn to William the Norman's capture of Exeter in 1067.^ 

The ram, the bore, and the mine were the main resources o 
the poliorcetic art during our period, but we must mention one 
or two engines of lesser importance. Scaling ladders are th( 
simplest of all the besieger's tools, and the most useless agains 
a competent defence ; nevertheless a town not unfrequently fel 
before an unexpected coup-de-main or a night attack in whicl 
the assailant had no more than ladders to help him. A stil 
more primitive method was that of heaping up earth fascines o 
rubbish of any kind against the lowest part of a hostile wall, anc 
endeavouring to clamber in over them. Rome itself fell befon 
this rude expedient in 896, when King Arnulf bade his German: 
lay against the foot of the ramparts their heavy saddles and th( 
packs of their beasts of burden, and actually succeeded ii 
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 centra 
years. Witiges, as we have already had occasion to mention, 
employed it in vain against Rome in 537. But we do not fine 
it emerging again till the eleventh centur}\ Probably it passec 
out of use during the days when fortification was neglected, anc 
had to be revived when the feudal castle had been produced b} 
the influence of the Viking and Magyar. It was, at any rate, ii 

^ See Orderic, iv. p. 510 : '* Per plurimos dies obnixe satagit . . . murum subtu 
suffodere." 

^ Luitprand, Antapodosis^ § 27. ^ See p. 131. 



looo] 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.^ 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, hotcrdis) 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.^ 
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 

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

2 Wilham of Tyre, viii, s ggg Abbo, i. 220. 



136 THE ART OF WAR IN THE MIDDLE AGES [looo 

to the foot of the walls. When set in this fashion, they are often 
called by the old Roman names of testudo 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 we 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 Western 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 
mangon, 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. 



ooo] THE MANGON AND BALISTA 137 

lie ropes, untwisting themselves in a moment, cast the 
Dck or ball with a high elliptic trajectory. The machine is 
ifficult to aim, as everything depends on the exact amount of 
orsion applied. A wet or dry day, for example, considerably 
ffects the ropes. But for shooting at large easy marks the 
aangon was effective ; it was specially good for what we may 
all " bombarding" work, i.e. the casting of missiles at large into 
. walled city or an entrenched position. The machine is called 
)y the name " mangon " as early as 886, where Abbo uses the 
/ord in his account of the siege of Paris.^ But it is probably 
dentical with the machine called by the simpler name of sling 
fundus^ fundibuld)^ which (as we have already had occasion to 
nention) was in use at a much earlier date. Such no doubt 
vere the '' slings " which were carried by the military train of 
Zharles the Great.^ The mangon is the legitimate descendant 
)f the Roman onager or scorpio described by Ammianus ^ and 
r*rocopius.* 

The second class of machines throwing missiles were those 
vorked by tension, of which we may take the balista as the 
ype. The balista is a magnified crossbow, as will be seen from 
he very clear description of it given by Procopius, when he is 
iescribing the engines used by Belisarius to defend the walls of 
^ome in 537. "These machines," he says, "have the general 
;hape of a bow ; but in the middle there is a hollow piece of 
lorn loosely fixed to the bow, and lying over a straight iron 
;tock. When wishing to let fly at the enemy, you pull back the 
>hort strong cord which joins the arms of the bow, and place in 
:he horn a bolt, four times as thick as an ordinary arrow, but only 
lalf its length.^ The bolt is not feathered like an arrow, but 
'urnished with wooden projections exactly reproducing the shape 
Df the feathers. Men standing on each side of the balista draw 
Dack the cord with little devices \i.e. winches] ; when they let it 
^o, the horn rushes forward and discharges the bolt, which strikes 
.vith 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 

1 Abbo, i. 364. 2 See p. 81. 

^ Ammianus, xix. § 7, and xxiii. § 4. ^ Procopius, De Bell. Gott. i. 21. 

^ 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 [88 

the cord, calling them merely [Mrr/avai ; we know, however, fror 
Ammianus, that they were little winches or windlasses whic 
were wound round and round in order to bring back the core 
He also omits to state that the cord was usually of twisted gu 
and that when tightened it was caught in grooves or notche 
cut in the iron or wooden stock to which the two arms c 
the balista were fixed. The machine was then aimed, b 
directing the point of the stock at the object which th 
engineer wished to strike, and, when good aim had been takei 
the cord was loosed, and sped the missile on its way.^ Vegetiu 
who is far shorter on the subject than Procopius, remarks the 
the longer the arms of the balista, the harder was the stroke ( 
the missile which it projected.^ The bolts thrown by it mu.' 
have been formidable things : at the siege of Rome by Witige 
Procopius saw a mailed Gothic chief, who was struck by 
balista-bolt while mounted in a tree, hang for a long time o 
the missile, which, after piercing him, had stuck deep into th 
wood. But it cast not only bolts, but long javelins. At tli 
siege of Paris, Abbo tells us how Abbot Ebolus launched froi 
a balista a lucky shaft which went through several Danes, wh 
fell dead pierced by the same missile. The abbot, thinking ( 
fowls broached on a spit, bade their friends " pick them u 
and take them to the kitchen." ^ 

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

^ Procopius must be read closely with Ammianus here : each supplements t 
other. Ammianus does not speak clearly of the horns of the bow. Procopius om 
the winches and notches. 

^ Vegetius, iv. § 22: " Quanto prolixiora brachiola habet, tanto spicula longi 
mittit." 

^ Abbo, i. 1 10. 



io66] THE CROSSBOW 139 

li reeling the construction of the Danish rams were slain by a 
^ong shot from a baHsta 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- 
t\pes 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.^ 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 middle 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 

^ It was, says Anna Comnena, rots "EWrjai TraureXQs dyvoovfievov (x. 8). 
- /hW. X. 8. 



I 



140 THE ART OF WAR IN THE MIDDLE AGES [88 

the exact date of its appearance — probably this revival of th 
old Roman manubalista dates back to that darkest of dark age; 
the end of the tenth century. 

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

Much confusion is caused to the readers of chronicle 
by the fact that the writers of the early centuries of the Middl 
Ages use many names for describing the same weapons. A] 
siege-artillery was either of the type of the mangon, u 
relying on torsion, or on that of the balista, i.e. relying o 
tension. But they are called indifferently " slings," " catapults, 
" petraries," " machines," " engines," " tormenta," with the mos 
exasperating vagueness and inaccuracy, by authors who, bein: 
for the most part clergy and not military men, did not full; 
understand the principles of the devices which they wer 
describing. Moreover, confusion is often caused by the fac 
that by slight adaptations or changes of shape, the " mangon. 
whose proper work was the casting of rocks, might be mad 
to hurl javelins, and the balista, whose speciality lay in th 
accurate propelling of shafts, might be induced to hurl stones. 

The best way to gain some idea of the characteristics of 
siege during the Dark Ages, is to investigate the details of . 
typical case. Unfortunately, there are very few chroniclers wh< 
give us really good descriptions of such operations. On th 
whole, we have a better account of the great siege of Paris ii 
885-886 than of any other leaguer between the days of Justiniai 
and the Crusades. Abbo's long poem on the subject is couchec 
in the vilest Latin, and abounds in the most excruciating fals 
quantities, but it is very detailed, and on the whole very cleai 
As every device of siegecraft known to the Dark Ages wa 
employed by assailants and defenders, it is well worth while t< 
give a short sketch of the incidents of those eventful elevei 
months. 

We have already mentioned that Paris in the autumn o 
885 consisted of the old island-city, with the new fortification 
added by Charles the Bald, namely, two bridges crossing th 
two branches of the Seine, which encircled old Paris, an( 
furnished with two bridge-heads. The northern one lay some 



585] THE GREAT SIEGE OF PARIS 141 

vhere about the spot where the tower of the Chatelet afterwards 
tood. The southern one must have been somewhere near the 
nodern Place St. Michel. The bridges were wooden structures, 
vhose central supports were laid on great piles of stones cast 
nto the Seine. The bridge-heads were stone towers, but the 
lorthern one was not completed at the moment when the 
Janes appeared, having only attained a half or a third of its 
lestined height. The town was in charge of Odo, count of 
he surrounding district, and of its bishop Gozelin. It was 
garrisoned by picked men from neighbouring parts of Neustria 
IS well as by its own citizens ; among the chief defenders were 
Hount Ragenar, Robert (afterwards king) the brother of Count 
Ddo, 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 
:erms to Odo and Gozelin, promising to do the city no harm if 
:heir vessels were allowed to pass under the bridges without 
Tiolestation. The count and bishop replied in very proper 
lerms : the Emperor Charles, they said, had placed Paris in 
:heir hands to serve as a bulwark for the rest of Neustria, and 
they would be betraying their master if they saved the town 
out 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." ^ 
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 

^ " Qui (Daci) vero cupiunt miirum succidere musclis" (Abbo, i. 99). 



142 THE ART OF WAR IN THE MIDDLE AGES [88 

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

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

Convinced that the place was not to be taken by a coup-dt 
main, the besiegers sent out their bands to ravage the neighboui 
hood, and collect a vast store of corn and cattle. They fortifie 
a camp near the church of St. Germain TAuxerrois, with 
foss and stakes, and settled down to beleaguer the city in fu 
form. Their artificers took some time in preparing three grcc 
rams, each covered by a penthouse of solid wood furnished wit 
sixteen wheels. The penthouses could hold sixty men apiec 
for the working of the rams. When, however, the machine 
were wheeled towards the walls, the besieged overwhelmed thei 
with a hail of missiles, and the two artificers who had designe 
them are said to have been both slain by one javelin from 
balista. This disaster to their engineers seems to hav 
delayed the bringing of the rams into action for some days.^ 

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

1 Abbo, i. 133-137. ^ !bid. i. 213-215. 



,] ASSAULTS ON THE BRIDGE-HEAD 143 

p to the bridge, which they tried to climb by mooring their 
essels against its supports. 

The assailants on land, having reached the bridge-head under 
lelter oi Xkv^ plutei, began to fill up the ditch which surrounded 
:. They cast into it clods of earth, boughs, straw, brushwood, 
ubbish of all sorts, and (when they grew excited at their 
lilure) their store-cattle, and even the bodies of the 
nfortunate prisoners whom they had captured in their raids 
Dund the neighbourhood. Meanwhile the besiegers poured a 
onstant hail of missiles upon them, and slew great numbers ; 
Lit while their attention was thus occupied, the Danes repaired 
nd brought up the three rams which they had been unable to 
itilise at their last assault. The rams were set to batter at 
hree points of the bridge-head, and began to work considerable 
lamage among the stones and mortar. 

The besieged now put in use a very ancient device, which 
lad been regularly employed against the ram in Roman times, 
etting down large beams with forked teeth, which caught the 
amheads and gripped them, so that they could no longer be 
)ulled backwards to deliver their stroke. They had also con- 
tracted a number of mangons.^ The heavy rocks which these 
nachines cast broke down the thick mantlets whenever they 
itruck them, and crushed all those sheltered beneath. After 
hree days of assault, the Danes had lost so heavily that they 
vithdrew from the walls under cover of the darkness, taking 
uvay such of their mantlets as were intact, but leaving two of 
;heir three rams abandoned and disabled as prizes for the 
Franks. 

While these unsuccessful attempts were being made upon 
^he 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 

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



144 THE ART OF WAR IN THE MIDDLE AGES 



I 

[88' 



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

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

While the siege was thus languishing, Henry Duke c 
Saxony appeared on the heights above Montmartre with rein 



386] HENRY OF SAXONY SLAIN 145 

"orcements sent by the emperor. The Danes retired into their 
:amp and took up the defensive, so that the duke was able to 
:ommunicate without hindrance with the city, and to throw into 
t a large convoy of provisions. The besieged took advantage 
)f the respite to restore the bridge, and apparently also to 
■oughly reconstruct the ruined bridge-head.^ But the siege was 
lot yet raised : after an unsuccessful attempt to storm the en- 
renchments of the Vikings, Henry drew off again, and left Paris 
.0 its own resources (March 886). The besiegers were, however, 
sufficiently impressed by the appearance of the relieving force 
:o transfer their camp from the northern to the southern bank of 
:he Seine, so as to put the river between themselves and any 
brce coming from the north. Siegfried, the most important of 
:he Danish leaders, recommended the raising of the siege, as it 
vas known that the Emperor Charles was calling together a 
arge army to carry out the enterprise in which Duke Henry 
lad failed. The majority refused, however, to follow his advice, 
md resolved instead to deliver a general assault on the city 
oefore the emperor should arrive. Early in April they simul- 
:aneously attacked the two bridge-heads, the bridges, and the 
sland itself, running their boats aground on the narrow shore 
it the foot of its fortifications and trying to scale them. They 
lad no success at any point, and a few days later Siegfried, 
bllowed by a considerable part of the host, took his departure, 
ifter receiving sixty pounds of silver — quite a moderate sum — 
rom the besieged, who hoped that he would induce the whole 
lorde to follow him. 

The majority, however, headed by a chief named Sinric, 
itterly refused to abandon the siege. They were perhaps 
encouraged to persist by the fact that pestilence had broken 
)Ut in the crowded city with the return of the warm weather ; 
imong its victims was Bishop Gozelin, one of the two chief 
leroes of the defence. The siege, however, had assumed a 
^ery curious aspect : the Danes being now mainly concentrated 
m the south bank of the river, — though they kept a corps of 
)bservation opposite the northern bridge-head, — the besieged 
:ould communicate in an intermittent way with the open 

^ This is nowhere stated by Abbo, but how could Henry have sent the flocks and 
lerds into Paris without a bridge? Moreover, the "tower," i.e. the bridge-head, 
>egins again to appear early in the second book of Abbo's poem, and is securely held 
)y the besieged. x 



146 THE ART OF WAR IN THE MIDDLE AGES [88 

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

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

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



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, they 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 without 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. Theii 
position, to compare modern things with ancient, much remind.' 
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 communicatior 
with the open country. The Vikings had the additiona 
difficulty of having only very narrow fronts — the two bridge- 
heads — to attack. The river prevented them from making anj 
really dangerous assaults on the island, whose walls they coulc 
not properly breach by siege-artillery placed on the mainland 
Hence we may fairly say that only famine could have beer 
relied upon as a certain method of reducing the place, and thai 
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. I066) AND DYRRHACHIUM (A.D. I08i) 

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 we 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,^ 
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 Romney, as we are told, did the lands- 
folk stand to their arms and beat off the raiders.- 

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.^ 

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



T50 THE ART OF WAR IN THE Mn)DLE AGES [1066 

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 7th 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. But Harold's heart had been stirred by the 
reports of the cruel ravaging of Kent and Sussex by the 
Normans,^ 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.^ A rapid march o 
two days brought him to the outskirts of the Andredsweald 
within touch of the district on which William had for the lasi 
fortnight been exercising his cruelty. 

Harold took up his position at the point where the roac 
from London to Hastings first leaves the woods, and come.' 
forth into the open land of the coast. The chosen grounc 
was the lonely hill above the marshy bottom of Senlac,^ c 
place far from all human habitations, and marked to th( 
chronicler only by " the hoar apple tree " on its ridge, jus 
as Ashdown had been marked two centuries before by it: 
aged thorn.* 

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

^ William of Poictiers, 201. 

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

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

* Asser, p. 23. 



o66] THE POSITION OF SENLAC 151 

^he road from London to Hastings crosses the isthmus, bisects 
he hill at its highest point, and then sinks down into the 
alley, to climb again the opposite ridge of Telham Hill. The 
atter is considerably the higher of the two, reaching 441 feet 
bove the sea level, while Harold's hill is but 260 at its 
ummit. The English hill has a fairly gentle slope towards 
he south, the side which looked towards the enemy, but on the 
lorth the fall on either side of the isthmus is so steep as to be 
ilmost precipitous. The summit of the position, where it is 
li'ossed by the road, is the highest point. Here it was that 
-ling Harold fixed his two banners, the Dragon of Wessex, and 
lis own standard of the Fighting Man. 

The position was very probably one that had serv^ed before 
or some army of an older century, for we learn from the best 
uithorities that there lay about it, especially on its rear, ancient 
3anks and ditches,^ in some places scarped to a precipitous slope. 
Perhaps it may have been the camp of some part of Alfred's 
irmy 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.^ 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 Aridredsweald, 
and was so close to its edge that the defenders could seek 
shelter in the impenetrable woods if they wished to avoid a 
battle.-^ 

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

^ " Crescentes herbae antiqmtm aggerem tegebant" (Orderic Vitalis, 501 d). 

" Praerupti vallis et frequentzim fossarum opportunitas " ("William of Poictiers, 203 d). 

Of these f)ne agger 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 fovea magna in front of 

i the English line, according to Henry of Huntingdon, 763 c : " Fugientes [Normanii] 

I ad quandam magnam foveam dolose tectam devenerunt, ubi multus eorum numerus 

i oppressus est. " This fovea 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. 

^ A.S. Chronicle, 893-894, copied in Ethelweard, 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." 

* " Mons silvae vicinus erat, vicinaque vallis, et non cultus ager asperitate sui " 
(Guy of Amiens, 365, 366). 



152 THE ART OF WAR IN THE MIDDLE AGES [io6 

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

Harold took one further precaution. He had served 
campaign in the Norman ranks a few years before, on th 
occasion of his involuntary visit to Ponthieu, and he thoroughl; 
knew the Norman tactics. The danger to the English lay, firs1 
in the rush of the duke's horse ; secondly, in the long shooting c 
the duke's archers. To guard against both these perils Harol( 

^ William of Poictiers, 201 B. 

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

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



o66] HAROLD STOCKADES HIS POSITION 153 

irected his men to build a fence of crossed woodwork ^ from the 
rushwood in the forest which lay close at their backs. It was 
n old Danish device, used two hundred years before, to 

^ Here we come to the most vexed point in this most interesting fight. Neither 
N'illiam of Poictiers, Guy of Amiens, Baldric, Henry of Huntingdon, nor any of the 
irly minor sources of information, distinctly mention this wicker-work. Can we 
list Wace, who gives an elaborate description of it before the battle and alludes to it 
uring the course of the engagement ? Wace is an authority of later date than the 
ihers, and wrote some ninety years after the battle. He occasionally makes strange 
rrors in his narrative (though the earlier writers, it must be remembered, do the same) 
ad sometimes is guilty of anachronisms, though on the whole he comparatively seldom 
'a^hes 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 ^/le battle 
■xcellence of the last age, would make such a strange error as to describe what was 
jally a fight on an open hill as an attack on a position which had been entrenched, 
ven though the entrenching was but slight ? On the whole, Wace's general narrative 
■ -SO fairly consistent with the earlier sources, that I cannot believe that he made this 
reat blunder. If it had been the common and ordinary thing for armies to stockade 
lemselves about 1150-60, though an uncommon thing in 1066, we might have thought 
iiai Wace was committing a mere anachronism. But it was no more unusual at one 
ate than at the other, and I do not see what should have induced him to bring 
:ie wattled barrier into his narrative, unless it existed in the tale of the fight as 
; had been told him by his father and others who had talked with the victors of the 
reat 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 
)eople's memory was a far more trustworthy thing. I cannot think that Wace, writing 
'. or the grandchildren of the men of Senlac, would have ventured to change so entirely 
f he character of the engagement. 

We can trace in the Roman de Ron the author's knowledge of several of our exist- 

ng authorities, e.g. of William of Poictiers, Guy of Amiens, and William of Jumieges. 

! f he had thought the existence of the breastworks inconsistent with their tale, it seems 

■ inlikely that he would have inserted it, for he does not give us the idea of an irrespon- 

ible inventor of facts, but of one who conscientiously uses the data that come to him, 

, hough he may have to adapt them a little to make them assume a fitting place in his 

tory. 

I conclude, then, that Wace, possibly from some lost chronicle or poem, possibly 
mly from popular oral tradition, knew of the existence of Harold's wattled breast- 
vorks, 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 nule jointure." (7815-20.) 

The ttdiding fenestres is, as Professor York Powell pointed out to me, possibly a 
icribe's blunder iox fresne: tresses : if so, the passage translates thus — 

"They made in front of them shields of wattled ash and of other woods, they 
aised them in front of themselves like hurdles joined and set close : they left no open- 
ng in them but made them into an enclosure. " The other main passages referring to 
he breastwork are, " d'escuz et d'ais s'avironoent," and " ne doterent pel ne fosse," in 
ine 8499. 



154 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, abates or phitei, 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 tc 
spare. The entrenchment, according to Wace, was triple, i,e, 
consisted of a centre and two wings, with intervals left between 
them.^ 

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.^ Although the Northumbrian and 
West -country levies were still missing, the army must have 
numbered many thousands, for the fyrd of south and centra] 
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 seventeer 
hundred or two thousand strong. The hill was completely 
covered by the English, whose spear-shafts appeared to the 
Normans like a wood,^ so that they cannot have been a mere 
thin line : if they were ten or twelve deep, the total must have 

^ '* Closre le fist de boen fosse, de treis parz laissa treis entrees " {R. de R. 12106). 
Fossi is the technical word for a military trench, and quite distinct hcmx fosse (feminine). 
a ditch. 

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

^ Guy of Amiens : *' Spissum nemus Angligenarum," 421, "silvaque densa priu- 
rarior efficitur," 428. 



o66] THE ARRAY OF THE ENGLISH 155 

cached some twenty - five thousand men. Of these the 
mailer part must have been composed of the fully -armed 
arriors, the king's housecarles, the thegnhood, and the heavily- 
quipped soldiery, of whom one had to be furnished by every 
ive hides of land. The rudely-armed levies of the fyrd must 
lave constituted the great bulk of the army : they bore, as the 
^ayeux Tapestry shows, the most miscellaneous arms — swords, 
avelins, clubs, axes, a few bows, and probably even rude instru- 
nents of husbandry turned to warlike uses. Their only defensive 
Limour was the round or kite-shaped shield : body and head 
vere clothed only in the tunic and cap of everyday wear. 

In their battle array we know that the well-armed house- 
arles — 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 
3n either flank. Presumably the thegns and other fully-armed 
nen formed its front ranks, while the peasantry stood behind and 
racked them up, though at first only able to hurl their weapons 
it the advancing foe over the heads of their more fully-equipped 
ellows. 

We must now turn to the Normans. Duke William had 
undertaken his expedition not as the mere feudal head of the 
oarons of Normandy, but rather as the managing director of a 
^reat 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 
I duke's claims on England. But all, or nearly all, of them had 
I consented to serve under him as volunteers, bringing not merely 
: their usual feudal contingent, but as many men as they could 
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,^ and Haimer of Thouars were ready to risk their lives 

^ Cousin of the reigning sovereign in Brittany. 



156 THE ART OF WAR IN THE MIDDLE AGES [106 

and resources on the chance of an ample profit. Frencl 
Bretons, Flemings, Angevins, knights from the more distar 
regions of Aquitaine and Lotharingia, even — if Guy of Amien 
speaks truly — stray fighting men from among the Norma 
conquerors of Naples and Sicily, joined the host.^ 

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

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

^ Guy of Amiens, line 259. 

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

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



o66] THE ARRAY OF THE NORMANS 157 

)ade his host get in array, and marched over the eight miles of 
oiling ground which separate Hastings and Senlac. When 
hey reached the summit of the hill at Telham, the English 
)osition came in sight, on the opposite hill, not much more than 
. mile away. 

On seeing the hour of conflict at hand, the duke and his 
alights drew on their mail-shirts, which, to avoid fatigue, they 
lad not yet assumed, and the host was arrayed in battle order. 
The form which William had chosen was that of three parallel 
orps, each containing infantry and cavalry. The centre was 
omposed of the native contingents of Normandy ; the left 
nainly of Bretons and men from Maine and Anjou ; the right 
)f French and Flemings.^ But there seem to have been some 
\^ormans in the flanking divisions also.- The duke himself, as 
vas natural, took command in the centre, the wings fell 
espectively to the Breton Count Alan Fergant and to Eustace 
jf 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 
:omposed of foot-soldiery armed not with missile weapons but 
vith pike and sword. Most of them seem to have worn mail- 
ihirts,^ unlike the infantry of the English fyrd. In the rear was 
;he 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 
lis 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 
)f horsemen who formed the third line of each division. 

The Normans deployed on the slopes of Telham, and then 
3egan their advance over the rough valley which separated them 
rom the English position. 

When they came within range, the archery opened upon the 

^ Guy 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 
eft. 

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

^ " Pedites firmiores et loricati," as William of Poictiers expresses it. 



158 THE ART OF WAR IN THE MIDDLE AGES [io( 

English, and not without effect ; ^ at first there must have be( 
little reply to the showers of arrows, since Harold had but vei 
few bowmen in his ranks. The breastworks, moreover, can ha^ 
given but a partial protection, though they no doubt serve 
their purpose to some extent. When, however, the Normai 
advanced farther up the slope, they were received with a furioi 
discharge of missiles of every kind, javelins, lances, taper-axe 
and even — if William of Poictiers is to be trusted — rude weapoi 
more appropriate to the neolithic age than to the elevem 
century, great stones bound to wooden handles and launche 
in the same manner that was used for the casting-axe.^ Tl 
archers were apparently swept back by the storm of missile 
but the heavy armed foot pushed up to the front of the Englij 
line and got to hand-to-hand fighting with Harold's men.^ The 
could, however, make not the least impression on the defendei 
and were perhaps already recoiling when William ordered i 
his cavalry.* The horsemen rode up the slope already stre\\ 
with corpses, and dashed into the fight. Foremost among the 
was a minstrel named Taillefer, who galloped forward cheerir 
on his comrades, and playing like a jongleur with his swor 
which he kept casting into the air and then catching again. Y 
burst right through the breastwork and into the English lin 
where he was slain after cutting down several opponent; 
Behind him came the whole Norman knighthood, chanting the 
battle-song, and pressing their horses up the slope as hard ; 
they could ride. The foot-soldiery dropped back — through tl 



i 



^ Baldric, v. 407 : " Spicula torquentur, multi stantes moriuntur." 
2 *' Lignis imposiia saxa" (W. P. 201 d). They seem to be represented \i\ 
club-like weapons thrown by some of the English in the Bayeux Tapestry. 

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

^ " Interea, dubio dum pendent proelia marte," Taillefer and the cavalry cai 
forward. 

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

" Interea dubio dum pendent proelia 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 mimus cognomine dictus," etc. 



io66] 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.^ 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.^ 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,^ 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 V/illiam 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 

^ 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 Ilaye Sainte. It was completely sivepi away by two squadrons 
of horse who charged through it. (Kincaid's Rijle Brigade^ chap. xx. ) 

^ " Pugnae instrumenta facile per scuta et alia tegmina viam inveniunt " (W, P. 

^ " Fere cuncta dacis acies cedit " (William of Poictiers, 133). 



i6o THE ART OF WAR IN THE MIDDLE AGES [io6 

number of several thousands, were cut to pieces.^ The grea 
bulk of the English host, however, had not followed the route 
Bretons, and the duke saw that his day's work was but begui 
Forming up his disordered squadrons, he ordered a secon 
general attack on the line. Then followed an encounter eve 
more fierce than the first. The breastworks were probabl 
swept away from end to end, and the ditch filled with d^bri 
and the bodies of men and horses ere it slackened. The fortun 
of the Normans was somewhat better in this than in the earlie 
struggle : one or two temporary breaches were made in th 
English mass,2 probably in the places where it had bee 
weakened by the rash onset of the shire-levies an hour before 
Gyrth and Leofwine, Harold's two brothers, fell in the forefror 
of the fight, the former by William's own hand, if we may truj 
one good contemporary authority.^ Yet, on the whole, the duk 
had got little profit by his assault: the English had suffere 
severe loss, but their long line of shields and axes still crowne 
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 b 
the disaster which had befallen the English right in the firs 
conflict. He determined to try the expedient of a feigned fligh' 
a stratagem not unknown to Bretons and Normans of earlie 
ages. By his orders a considerable portion of the assailants 
suddenly wheeled about and retired in seeming disorder. Th' 
English thought, with more excuse on this occasion than on th- 
last, that the enemy was indeed routed, and for the second tim 
a great body of them broke the line and rushed after the retreat 
ing squadrons. When they were well on their way down th 
slope, William repeated his former procedure. The intact portioi 
of his host fell upon the flank of the pursuers, while those wh< 
had simulated flight faced about and attacked them in froni 
The result was again a foregone conclusion : the disordered mei 
of the fyrd were hewn to pieces, and few or none of then 

^ " Exardentes Normanni et circumvenientes, millia aliquot insecuta s 
momento deleverunt, ut ne unus quidem superesset" (William of Poictiers, 133). 

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

^ Guy of Amiens. 

* We cannot say what portion, or what proportion. The Brevis Relatio say 
that it was a "cuneus Normannorum fere usque ad mille equites," and that they wer 
" ex altera parte " from the duke. But does this mean the right or the left wing ? 



o66] HASTINGS: THE GREAT STRUGGLE i6i 

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

Harold's host was now much thinned and somewhat shaken, 
)ut, in spite of the disasters which had befallen them, they drew 
ogether their thinned ranks, and continued the fight The 
;truggle was still destined to endure for many hours, for the 
nost daring onsets of the Norman chivalry could not yet burst 
nto the serried mass around the standards. The bands which 
lad been cut to pieces were mere shire-levies, and the well- 
irmed housecarles had refused to break their ranks, and still 
"ormed a solid core for the remainder of the host. 

The fourth act of the battle consisted of a series of vigorous 
issaults by the duke's horsemen, alternating with volleys of 
irrows poured in during the intervals between the charges. The 
Saxon mass was subjected to exactly the same trial which befell 
;he British squares in the battle of Waterloo — incessant charges 
Dy 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 

^ William oi 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 "lovea magna " (762 c). 

^ " Leviter sauciatos non permittit evadere sed comprimendo necat sociorum 
densitas " (William of Poictiers, 202 d). 



i62 THE ART OF WAR IN THE MIDDLE AGES [io6f 

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

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

^ William of Poictiers, 202 D, says that there were ^zco sallies of the Engli.' 
provoked by Norman feigned flights, in addition to that which followed the first re 
flight of the Bretons. *' Bis eo dolo simili eventu usi sunt Normanni." 

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

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

^ " Mortui plus, dum cadunt, quam vivi movere videntur " [ibid.). 

^ That the arrow-shower alternated with the charges is obvious. The arche 
could not shoot while the knights blocked the way. That the arrow was largely us( 
is proved by William of Poictiers: '■^ Sagittant et perfodiunt Normanni." Th 
must have been done alternately and not simultaneously. Wace well describes tl 
dismay caused by the rain of shafts falling from above ( 1 3287). 

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



io66] HASTINGS: FLIGHT OF THE ENGLISH 163 

lill turned their backs to the foe and sought shelter in the 
riendly forest in their rear. Some fled on foot through the 
rees, some seized the horses of the thegns and housecarles from 
he camp and rode off upon them. But even in their retreat 
hey took some vengeance on the conquerors. The Normans, 
bllowing in disorder, swept down the steep slope at the back of 
he hill, scarped like a glacis and impassable for horsemen, — the 
Dack defence, as we have conjectured, of some ancient camp of 
3ther days.i Many of the knights, in the confused evening light, 
blunged down this trap, lost their footing, and lay floundering, 
nan and horse, in the ravine at the bottom. Turning back, the 
ast of the English swept down on them and cut them to pieces 
Defore resuming their flight. The Normans thought for a 
moment that succours had arrived to join the English — and, in- 
ieed, 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 tor 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 

MVilliam of Poictiers, 203 d: " Frequentes fossae et praeruptus vallis." 
" Antiquus agger" (Ord. 501 d). 

^ Annales Altahenses majoresy sub anno 1066. 



i64 THE ART OF WAR IN THE MIDDLE AGES [io8 

for close fight and destitute of cavalry support, were absolutel 
helpless before a capable general who knew how to combine th 
horseman and the archer. The knights, if unsupported by th 
bowmen, might have surged for ever against the impregnabl 
breastworks. The archers, unsupported by the knights, coul 
easily have been driven off the field by a general charge. Unite 
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 th 
Norman lance and bow. Fifteen years after Harold's defea 
another body of English axemen — some of them may well ha\ 
fought at Senlac — were advancing against the army of a Norma 
prince. They were the Varangian Guard — the famous n?Xsx 
<popot — of the Emperor Alexius Comnenus. That prince wi 
engaged in an attempt to raise the siege of Dyrrhachium, the 
invested by Robert Guiscard. The Norman army was alread 
drawn out in front of its lines while the troops of Alexius wei 
only slowly arriving on the field. Among the foremost of i\ 
emperor's corps were the Varangians, who rode to the battl< 
spot, like the thegns of the West, but sent their horses to \\ 
rear when they drew near the enemy. Alexius had entrusted 1 
their commander a body of light horse armed with the bo^ 
bidding him to send these first against the enemy, and only ■ 
charge when the cavalry should have harassed and disturbc 
Robert's ranks. But Nampites, the Varangian leader, neglecte 
these orders. When they approached the Norman line, tl 
English were carried away by their reckless ardour. Befo 
the Greek army was fully arrayed,^ and long before the er 
peror had designed to attack, they moved forward in a sol 
column against the left wing of the Normans. They fell up( 
the division commanded by the Count of Bari, and drove 
horse and foot, into the sea. But their success disordered the 
ranks, and Guiscard was enabled, since the main body of tl 
Byzantine host had not yet approached, to send fresh fore 
against them. A vigorous cavalry charge cut off the greater pa 
of the English: the remainder collected on a little mound by tl 
seashore, surmounted by a deserted chapel. Here they we 
surrounded by the Normans, and a scene much like that 
Senlac, but on a smaller scale, was enacted. After the hors 

^ They, 'Uavov dir^aTrjaav 5c atrupiav d^vrepov ^e^adiKdres, were a considera 
distance from the rest of the army (Anna Comnena, book iv. § 6). 



o8i] BATTLE OF DYRRHACHIUM 165 

nen and the archers had combined to destroy the majority of 
he Varangians, the survivors held out obstinately within the 
hapel. At last Robert sent for fascines and other woodwork 
rom his camp, heaped them round the building, and set fire to 
he mass. The English sallied out, to be slain one by one, 
)r perished in the flames. Not a man escaped : the whole corps 
offered destruction as a consequence of their misplaced eager- 
less to open the fight.^ Such was the fate of the last important 
ittempt made by infantry to face the feudal array of the eleventh 
;entury. We shall find, it is true, some instances in the twelfth 
entury of cavalry being withstood by dismounted troops. But 
hese were not true infantry, but knights who had sent their 
lorses to the rear in a supreme moment of peril, and stood firm 
o fight out the battle to the end. Well-nigh three centuries 
vere to elapse before real foot-soldiery, unaided by the cavalry 
irm, made another serious attempt to stand up in the open 
igainst the mailed horseman.^ The supremacy of the feudal 
lorseman was finally established. 



^ Anna Comnena calls the leader of the Varangians " Nampites." This does not 
eem to be a true Teutonic name. A military correspondent suggests to me that it 
nay possibly represent a nickname — *'Niemecz" or * * Nemety " = the German — 
)estowed on the English chief by Slavonic fellow-soldiers in the Imperial host. 

- 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 
cale in an obscure corner of Europe ; the Scandinavians neglected the cavalry arm 
;ven later than the English. Other cases could be quoted. 



BOOK IV 

THE BYZANTINES 
A.D. 579-1204 



167 



I 



CHAPTER I 

HISTORICAL DEVELOPMENT OF THE BYZANTINE ARMY 

IN our first chapter we traced the miHtary 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 " mimeri 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 (lO/i) 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 



lyo THE ART OF WAR IN THE MIDDLE AGES [580 

" The vices of Byzantine armies," says Gibbon, " were in- 
herent, their victories accidental."^ 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 oi 
war, not the worthlessness of the troops, were the usual sources oi 
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 theii 
enemies ; in discipline, organisation, and armament far superior 
Above all, they possessed not only the traditions of Romar 
strategy, but a complete system of tactics, carefully elaboratec 
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. Tht 
young Prankish noble deemed his military education complete 
when he could sit his charger firmly and handle lance and shiek 
with skill. The Byzantine patrician, while no less exercised ir 
arms,2 added theory to empiric knowledge, by the study of th( 
works of Maurice, of Leo, of Nicephorus Phocas, and of othe 
authors whose books survive in name alone. The results of th( 
opposite views taken by the two divisions of Europe are wha 
might have been expected. The men of the West, thougl 
they regarded war as the most important occupation of life 
invariably found themselves at a loss when opposed by ai 
enemy with whose tactics they were not acquainted. Th' 
generals of the East, on the other hand, made it their boas 
that they knew how to face and conquer Slav or Turk, Fran] 
or Saracen, by employing in each case the tactical means bes 
adapted to meet their opponents' method of warfare. 

The Byzantine army of the seventh and following centurie 
may be said to owe its peculiar form to a reorganisation which i 
went through in the last quarter of the sixth century, som 
twenty-five years after the death of Justinian. The details c 
that reorganisation are preserved for us in the Strategiconf a 
invaluable work, which shows us precisely when and by whor 

1 Vol. ii. p. 382. 

2 Nothing better attests the miUtary spirit of the Eastern aristocracy than the 
duels ; cf. the cases of Prusianus and others. 

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



i 



So] MAURICE'S " STRATEGICON » 171 

le change was carried out. East - Roman writers of a later 

ge often erroneously attributed these alterations to the 

elebrated warrior-prince Heraclius, the conqueror of Persia 

nd the recoverer of the True Cross. In reality, the army 

,'ith which Heraclius won his battles had already been re- 

rganised by his worthy but unfortunate predecessor, the 

Lmperor Maurice, whose troubled reign filled the years 582- 

02. It is under his name that the Strategicon appears, and by 

is hands that it was compiled. There seems no reason what- 

\er to doubt the attribution of the Strategicon to the Emperor 

Jaurice. A careful inspection of the chronological data which 

re supplied by the book itself shows that it cannot have been 

vTitten before 570 or after 600. The Persian king is alluded 

o as the chief enemy of the empire, but he is not represented 

.8 a masterful and oppressive neighbour, as would have been 

he case in any book written after the Persian invasions of 605- 

y-'j-Z. On the other hand, the Slavs and Avars are declared to 

)e the hostile powers on the Danube, no mention being made 

)f Gepidae or Lombards : therefore the latter tribes must have 

dready vanished from its banks ; i.e. the writer is dealing with a 

oeriod after 568. But from the fact that all the fighting with 

Slavs and Antae is supposed to take place in the close neighbour- 

lood of the Danube, and for the most part not on Roman soil, 

3Ut beyond the river, we can fairly decide that the great Slavonic 

aids of 581-585, which reached as far as Thessalonica and 

Thermopylae, cannot yet have begun. The date 570-580 is 

'endered 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," ^ 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- 

^ Strategicon^ i. 



172 THE ART OF WAR IN THE MIDDLE AGES [58c 

dition of decay, and as forgetting its old triumphs; such i 
complaint could hardly have been made when the victories o: 
Taginae and Casilinum (553-555) were still fresh in men'j 
memories. The decline began in the last few years of Justinian'.' 
time, when (as Agathias tells us) " the emperor having enterec 
on the last stage of his life seemed to weary of his labours, anc 
preferred to create discord among his foes or to mollify then 
with gifts, instead of trusting to arms and facing the danger: 
of war. So he allowed his troops to decline in number: 
because he did not expect to require their services, and th( 
ministers who collected his taxes and maintained his armie: 
were affected with the same indifference."^ The decay whicl 
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 describee 
by Maurice in the latter days of that prince. 

But we may venture to determine even more exactly the dat< 
of the Strategicon. When the Emperor Tiberius Constantinu 
succeeded Justin II. (578) he carried out a thorough reorganisa 
tion of the army, deputing the care of details to two distinguishec 
officers, Justinian, the son of Germanus, and Maurice himselt 
These two colleagues "set right that which was wrong, an( 
made orderly that which was chaotic, and, in short, reducec 
everything to a state of efficiency." ^ We may therefore con 
elude with reasonable certainty that the Strategicon was thei 
issued by Maurice to serve as the official handbook of th 
reorganised host of the Eastern Empire. In that case it may b 
ascribed to the year 579, a date which exactly suits all th 
internal indications of time of which we have already spoken. 

It would seem that the commissioners made many sweepin 
changes in the army, for the troops which Maurice describe 
are arrayed and named very differently from those of whic' 
Procopius had drawn a picture thirty years before. It is tru 
that the mailed horse-archer, the '/.a.[3aXXdpiog or -/.ovrdrog,^ as he i 
now called, still remains the great power in war, and the sta; 
and hope of the Imperial host. But a completely new syster 
of organisation has been introduced both among cavalry an 
infantry. Under Justinian there was no permanent unit in th 
army larger than the single regiment, the corps which Procopiu 

^ Agathias, book v. 14. - Theophanes, sud anno 6074. 

^ i.e. lancer, from k6»'tos, the long cavalry spear. 



So] THE EAST-ROMAN ARMY IN 580 173 

alls a ■/.araXoyog, SO translating the word numerus, which was 
till its official title. Maurice recognises this body, which he 
alls an dpid,a6g {i.e. numerus), or more frequently a rdy/Ma or 
rhdov, as the base of military organisation ; but he speaks of 
he numeri as being formed into larger bodies, — brigades and 
ivisions as we should call them. Six, seven, or eight numeri are 
o form a fioTpa of two thousand to three thousand men, the 
quivalent of a brigade, and three >j.oTpai are to be united into 
. .agpoc,^ or division of six thousand or eight thousand men. He 
dds that the numerus should be not less than three hundred 
ir more than four hundred strong, and that moirai should be 
ormed of an irregular number of numeri, in order that the 
nemy should not be able to calculate the exact force opposed 
o them by merely counting the number of standards in the line 
)f battle. Napoleon, it will be remembered, laid down a similar 
ule as to his army corps, always taking care that they should 
lot be of exactly similar force. 

A numerus, or "band," or ray/xa of three to four hundred 

trong, is now commanded by an officer called comes or tribinms. 

t is interesting to see how the importance of these names has 

! ;hrunk — in the fourth century there were only about a dozen 

I 'counts" in the whole empire, and each had ruled a whole 

: rontier and commanded many cohorts. A tribune in a similar 

nanner had once been the commander of a whole legion of six 

:housand men. Now, however, the two words are used as 

lomonyms, and applied to a simple colonel. The brigadier in 

:ommand of seven or eight bands is now called a iMoipap-xjag, or, as 

a Latin equivalent, a dux (3oSg), though the duces of the fourth 

:entury 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 

^ Also called a 5P01J770S, a Teutonic name connected with our own word throng. 



174 THE ART OF WAR IN THE MIDDLE AGES [58 

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

This reform was rendered all the more easy by the fact the 
the barbarian element in the Imperial army was decidedly 
the decrease. The rapid fall in the revenues of the State whic 
had set in towards the end of Justinian's reign, and which cor 
tinned to make itself felt more and more under his successor 
had apparently resulted in a great diminution in the numb( 
of Teutonic mercenaries serving in the Roman army. It Wc 
a case, to quote a modern proverb, of ^'- Point d' argent^ point i 
Suisse!' For the foreigner was a more expensive and a mo] 
independent personage than the native soldier, and vanishe 
when his pay ceased to appear. To the same end contribute 
the fact that of the Lombards, Heruli, and Gepidae, the natioi 
who had formed the majority of Justinian's Foederati, one natic 
had removed to other seats, while the others had vanished fro 
the scene. At last the number of the foreign corps had sur 
to such a low ebb that there was no military danger incurred 
assimilating their organisation to that of the rest of the arm 
The barbarian element, as we find it in Maurice's book, appea 
under the names of Foederati, Optimati, and Buccellarii. Tl 
former seem to represent the old bands of Teutonic auxiliari' 
serving under their own chiefs ; they are apparently spoken 
as invariably consisting of heavy-armed horse. A casual notice 
Theophanes informs us that the Emperor Tiberius Constantin 
found it so hard to keep up their numbers, that he bought < 
the Teutonic slaves he could find for sale in and outside tl 
empire, freed them, and enrolled them as soldiers. The tot 
number of Foederati was thus brought up to fifteen thousan 
and it was precisely Maurice who was put in command of thei 



8o] THE EAST-ROMAN AUXILIARY TROOPS 175 

7ith the title of " Count of the Foederati." The " Optimati " 
sem to have been the pick of the Foederati : they were chosen 
lands of Teutonic volunteers of such personal importance that 
ach was attended by one or more military retainers, called 
Irmati^ just as a mediaeval knight was- followed by his squires.^ 
^he Buccellarii, whose name and status has caused much needless 
rouble to commentators both in Byzantine and modern times, 
/ere another select portion of the Foederati, who were regarded 
s the emperor's personal following — they had no doubt done 
im homage and regarded themselves as part of his " comitatus " ; 
tractically they were the barbarian element in the Imperial 
juard, the body which corresponded to the old " Batavian 
ohorts " of the first century. The institution, as we have already 
lad occasion to mention, was of German origin : we find in the 
aws of the Visigoths saio and buccellarius used as synonyms for 
he oath-bound military dependant whom the Angle or Saxon 
v^ould have called a gesith. But it had early been adopted by 
he Romans : great captains like Aetius and Belisarius had their 
)uccellarii just like a Gothic king. 

The Teutonic element had thus become comparatively small 
n the Imperial army: such as it was, it consisted of the scanty 
emains of broken tribes such as the Heruli, Ostrogoths, and 
jepidae, and of stray Lombards who had fled from their king 
—like the Droctulf of whom we have considerable notice in 
Vlaurice's time. There were also a few " Scythians," i.e. remnants 
)f the Huns, and Avar refugees who had deserted their lord the 
jreat Chagan, a habit to which, as we learn from the Strategicon, 
hey were very prone. 

Nothing can be more characteristic of the transitional state 
)f 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 
anguage 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 
osing or confusing its case endings.^ Upon the substratum of 

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

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



176 THE ART OF WAR IN THE MIDDLE AGES [58c 

old Roman survivals we find a layer of Teutonic words intro 
duced by the Foederati of the fourth and fifth centuries — sucl 
as bandon for a company of soldiers, drimgus (cf. throng) for ? 
larger body : ^ burgus, coccoiira^ betza^ and phulcus, and similai 
words. Finally, we meet with many Greek words, some 
them literal translations of Roman terms — for example, api^i^o 
for numerus, — some of them borrowed from the old Macedoniar 
military system by officers of classical tastes,^ some newl) 
invented.^ 

The whole official language of the empire was, in fact, still ii 
a state of flux ; the same thing had often two or three names 
one drawn from each tongue. Maurice calls the regiment in 
differently (Sdvdov, rdyfia, or apt&fMoc^ and the brigadier ixoipapyji., 
bpoxjyyaptog^ ov dux. On the whole, however, the Latin holds it 
own ; we still find it used for scores of things which in Leo' 
Tactica, a work of three hundred years later, have onl 
Greek names. A very large proportion of the native troops wer 
still Latin-speaking, all those, in fact, raised in Thrace, Moesi; 
and the inner parts of the Balkan peninsula. It was not ti 
these provinces were overrun by the Slavs, a few years after th 
Stj-ategicon was written, that the ancient Roman tongue becam 
practically a dead language in the Eastern realm. Mauric 
seldom or never thinks it worth while to give the Gree 
rendering of a Latin technical phrase, while his successor Le 
invariably translates such terms. 

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

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

^ e.g. 8i.(pa\ayyla, vTaaTriaTTjs, ovpaydi, X6xct70S. 

^ £.£-. fioipa and fjJfios as technical military expressions. 



Qo] LOCAL MILITARY ORGANISATION 177 

:antly provided with that weapon and with two javelins." 
)nce accustomed to arms, he thought that the provincial would 
lore easily be induced to enlist. If, however, this was intended 
) be the first step towards the introduction of universal military 
.n'vice, the design was not carried out. Three hundred years 
iter we find Leo echoing the same words : 1 " The bow is the 
asiest of weapons to make, and one of the most effective. We 
lerefore wish that those who dwell in castle, countryside, or 
3wn, in short, every one of our subjects, should have a bow 
f his own. Or if this be impossible, let every household keep a 
ow and forty arrows, and let practice be made with them in 
hooting both in the open and in broken ground and in defiles 
nd woods. For if there come a sudden incursion of enemies 
ito the bowels of the land, men using archery from rocky 
round or in defiles or in forest paths can do the invader much 
larm ; for the enemy dislikes having to keep sending out 
letachments to drive them off, and will dread to scatter far 
.broad after plunder, so that much territory can thus be kept 
inharmed, since the enemy will not desire to be engaging in a 
)erpetual archery-skirmish." 

It is unfortunate that we have no definite information as to 

he extent to which this plan for creating a kind of landsturm 

Lpt for guerilla warfare was carried out. That in many districts 

)f the empire little or nothing came of it we know only too 

veil. We hear continually of provinces that failed to defend 

hemselves when they were not furnished with a regular garrison. 

3n the other hand, there seems to have been some obligation 

; o provide men for military service incumbent on the themes. 

I iVe learn, for example, from a casual reference in Const antine 

I ?*orphyrogenitus' De Administi-ando Imperio that in the time 

■ )f his own father-in-law Romanus, " when the emperor wished 

I :o raise Peloponnesian troops for an expedition against the 

I Lombards, in the days when John the Protospathiarius ruled that 

;heme, the Peloponnesians offered to give a thousand saddled and 

Dridled horses and a contribution of one centenar of gold instead 

i )f the levy, and, the offer being accepted, paid it with alacrity. 

The x^rchbishop of Corinth was assessed at four horses, the 

Archbishop of Patras at four, the bishops at two horses each, 

ill protospathiarii resident in the theme at three horses each, 

spathiarii at one horse, the richer monasteries at two each, the 

1 Tact. XX, § 84. 
12 



178 THE ART OF WAR IN THE MIDDLE AGES [90c 

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 waj 
easily raised." ^ 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.- 
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 ol 
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.^ The very curious 
and interesting poem of Digenes Akritas^^ which gives the life oi 
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 v/hose border 
they inhabited. In Leo's Tactica we learn from the chapter that 
deals with sieges that the government relied on the services oi 
the citizens whenever a frontier town was besieged, and that they 
were distributed to definite posts in the defence. Only if any oi 
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 was so unwise as to relieve 
the provinces of the eastern border of their obligation to keef 
up local levies to supplement the Imperial garrison. They had 
hitherto been exempted from certain taxes in consideration oi 

^ Const. Porph. , De Adm. Imp. cap. 51. 

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

' The holdings were called KTrj/nara o-TpaTiuTLKa : they were hereditary, as long at 
the military service was paid duly. 

^ Edited by Sathas and Legrand, Paris, 1875. 



54o] DISASTERS OF THE SEVENTH CEJNTURY 179 

:his service. Now they were ordered to disband the militia and 
n future send money to the central treasury .^ 

If universal military service never came into use in the 
Eastern Empire, yet Maurice had at least a portion of his 
iesire fulfilled. From his time onward the rank and file of the 
imperial forces were raised almost entirely within the realm, and 
nost 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 
rhracians in Europe, were considered the best material by the 
ecruiting officer. 

The next great landmark in the military history of the 
iinpire after the issue of the Strategicon is the fearful storm 
\ hich passed over it in the Persian and Saracen invasions of the 
.ears 604-656. Tiberius Constantinus and Maurice were fairly 
ucky in their campaigns, beat back the Persians, and carried 
ncursions into the land of the Transdanubian Slavs. But 
Maurice was unpopular with the army — perhaps his cutting down 
)f the power and importance of the great officers, no less than 
lis strict discipline and economy, irritated them. He perished 
he victim of a mutiny, and the brutal and imbecile Phocas, who 
succeeded him, involved the empire in the last and the most 
lisastrous of its Persian wars. The whole East, from the 
^^uphrates to the Hellespont, was overrun by King Chosroes, 
vhile at the same time the Slavs and the Chagan of the Avars 
noved forward into the European provinces. The empire 
eemed on the brink of destruction, and was only saved by the 
leroic six years' campaign of Heraclius (622-628). But hardly 
lad the Persian war ended, and the old frontier of the empire been 
estored, when the still more fatal Saracen invasion began (633). 
n his old age Heraclius saw Egypt and Syria permanently 
evered from the empire, and had to reorganise a new military 
rentier for his diminished realm along the line of the Taurus. 

There was no peace with the Saracen till 659, and for twenty- 
ix years the whole force of Eastern Rome was concentrated 
ilong its Asiatic border, struggling desperately with the oncom- 
ng flood of Saracen fanaticism. Either during this long war, or 
nore probably at its end, when Constans II.^ sat on the throne, a 
lew military organisation of the highest importance was imposed 

^ Cedrenus, 790 ; Zonaras, ii. 260. 

' Or Constantine iv., as he should more properly be named. 



i8o THE ART OF WAR IN THE MIDDLE AGES [66c 

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 force.^ 
serving in Asia Minor at this time consisted — (i) of the troops o 
the old "diocese" of Oi^iejiSy i.e. Syria, now called in Greet 
'AvaroX/;to/; (2) of the troops of the borders of Mesopotamia anc 
Armenia, who were generally known as 'A^/A^i//a%o/; (3) of th( 
soldiers of Thrace, brought over into Asia during the stress o 
the struggle, and known as Thracesians ; (4) of the surviving 
Foederati, now known as the Optimati\ (5) of the native anc 
foreign halves of the Imperial Guard, known respectively aj 
the Obsequmm and the Buccellarii. During or at the enc 
of the war these troops were cantoned in various parts of Asii 
Minor in separate bodies or army corps, for the long-continuec 
struggle had rendered permanent their brigading.^ 

The new provincial arrangement of the middle of the seventl 
century consisted in making these army-corps districts, adoptee 
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 th( 
district and the head of the administration; the "bands" anc 
" moirai " were permanently fixed down to the posts where the} 
found themselves. The new geographical divisions and th« 
army corps both received the appellation of Themes, ^g.aaTct 
Their proper names were drawn from the titles of the troop 
quartered in each, and were therefore Anatolicon, Armeniacon 
Thracesion, Optimaton, Buccellarion, Obsequium {h-^r/.m^ 
These were the original " themes " of Asia ; shortly afterward 
there was added to them one whose character was similar, bu 
whose origin was probably naval rather than military ; this wa: 
the Cibyrrha^ot theme, a narrow district reaching along th( 
southern coast of Asia Minor from Caria to Isauria, and com 
prising only the land between the mountains and the sea 

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



PLATE III. 




Probable 
Limits of 
TH E THEMES 
AD. 650. 




'V^. 7^,9. "<Sebaste 



ikion n ■ ^-C,-- -^■"^-. -'t?. 



Ops 



> *■■''.-^^Vana' 

^«C0> v^r^'j^ ^? o JSeleucil 




THE " THEMES " OF THE EASTERN EMPIRE IN 650 AND 95O 



56o] THE THEMES i8i 

Cibyra was a small place, and why it gave its name to the 
:heme was a constant puzzle to later Byzantine authorities. 
Constantine Porphyrogenitus, in his work on the Themes, says 
-hat the name was bestowed in mockery. This is of course 
ibsurd : it is perhaps lawful to conjecture that at the moment 
vvhen the new provincial divisions were made, Cibyra was the 
:hief station of the Imperial fleet which guarded the southern 
^hore of Asia Minor and the passage into the Aegean. The 
district to which it gave its name was purely maritime, and the 
solated coast-plains of which it was composed only com- 
municated with each other by sea. It was probably, therefore, 
uhe special domain of the fleet, and if there was any regular 
:avalry army corps allotted to it, the " bands " told off to protect 
t 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 
Tacticay which bids the general of the Anatolic theme, when 
lis own theme is attacked by land, to send word to the com- 
mander of the Cibyrrhaeot fleet, that the latter may land forces 
n the rear of the Saracens and devastate Cilicia.^ By the time 
3f Constantine Porphyrogenitus, Optimaton, probably on account 
Df its vicinity to the capital, had no longer any military estab- 
ishment, 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 
massed out of Roman hands. There were probably only three 
"hemes south of the Danube — Thrace, Thessalonica, and Hellas ; 
:o these the other Western possessions of the empire add three 
iTiore — Sicily, Africa, and the surviving dominions in the empire 
in Italy. These last, however, were always called not a theme, 
3ut the Exarchate of Ravenna. Later emperors in the eighth 
md 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 

5rav 5^ 5id r^s 7^75 e'/ccrr/jareuetJ' ^AXwci o\ KiXiKes ^dp^apoi, /xrjv6r}S t^ Ki^vpaidTrj 
rod TrXuitfiov crrpar7;7y, /cat fiera tu)v utt' avTov dpofjubpwv elcnrnrT^rit} Kara tCjp Tapdwv 
Kol 'Adaveluv x^ptwi/ (Leo, Tactica, cap. xviii. § 139). 



! 



i82 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 tc 
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, sc 
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 De 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 Buccellariarj 
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 



i 



9oo] 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 Caesarea and Nyssa." ^ 
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. 

^ 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 reedition 
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 oi 
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 ol 
the empire. These are the points on which an intelligent war- 
minister in the capital could attain full knowledge. The 
weakest 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 precepti- 
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 



oo] THE BYZANTINE CAVALRY 185 

ot permanent conquests ; his main end is the preservation of 
is own territory rather than the conquest of his neighbour's. 
\.fter reading the book, it is easy to see why the invaded theme to send off all his infantry 
' occupy passes in the hills, or fords on great rivers, so as to 
ock the enemy's retreat ; he would then start with his cavalry 
one to hunt down the raiders. This fact is deducible from 
eo's Tactica, but is still more explicitly stated in the excellent 
imphlet on the defence of the Asiatic border which stands 
ider the name of the Emperor Nicephorus Phocas.^ 

When infantry and cavalry acted together, as would be the 
ise against an enemy mainly composed of foot-soldiery, e.g. 
le Slavs or the Franks, or against a regular invasion of 
aracens as opposed to a mere raid, the usual tactical arrange- 
lent of the Byzantines was to place the infantry in the centre, 
ith cavalry on the wings and in reserve behind the line. The 
ifantry " band " was drawn up sixteen, eight, or occasionally 
)ur deep, with the scutati in the centre and the archers and 
ivelin-men on the flanks. If expecting to be charged by 
ivalry, or to be assailed by a heavy column of hostile foot, the 
ght troops retired to the rear of the scutati and took refuge 
ehind them, just as a thousand years later the musketeers of the 
cxteenth and seventeenth centuries used to take cover behind 
leir pikemen. The " band " was taught to fight either in single 
r in double line {bicpakayyia) : to take this latter formation the 

^ Niceph. Phoc. iii. § i. The strategos is at once, on receiving news of a raid, to 
)llect his horse and t6 ire^iKhv airav iinffvvayeLV irrl ttjv 68ov KaO' fjv bpfi-qcrovaiv oi 
oX^fjLLOL e^eXdeiv. The retreating enemy, heavy with plunder, could be intercepted 
isily in the passes by the foot-soldiery, and could be crushed between them and the 
ur suing 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, while the 
front ranks moved forward and then halted.^ 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.^ When, however, the army 
had moved out far from its camp to take the offensive, tht 
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 drungus) in second line, 01 
by forming each band into the above-mentioned h<pa.'hayyia 
with an interval of three hundred yards between its front anc 
its rear half-band. The army was never drawn out in a singk 
line without reserves ; that order of battle was discouraged b} 
all Byzantine writers on matters tactical. It was only used a.' 
a last resort when there was a desperate need to produce at al 
costs a line equal in length to the enemy's. 

Byzantine infantry were accustomed to charge in columi 
sixteen deep ; the bowmen and javelin - men having retiree 
behind the scutati, the latter received the command to close uj 
the ranks (clnvusov), and drew close together, the front rani 
locking their shields together, while the second and rear rank; 
held their shields aloft over their heads, after the manner of th( 
ancient Roman testudo. The bowmen in the rear kept up sucl 
a discharge as they best could over their comrades' heads. Oi 
getting within a few paces of the enemy, the scutati hurled thei 
spears, as did the ancient Romans their pila^ and then fell tc 
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 infantiy division (turma) when charging 
would be composed of a number of small columns, witl 
moderate intervals between them, each composed of from som< 
two hundred and fifty to four hundred men.^ The strength o 
the division might be anything between two thousand"* an( 
six thousand strong, and the number of battalions (bands) in i 

1 Tadica, vii. § 76. 2 Tactica, vii. § 73, 4. 

^ An interesting but casual notice in one of the doubtful chapters of th 
Tactica (No. xxxiv. ) says that in the Thrakesian theme the bandon was supposed t 
be three hundred and twenty strong ; in the theme of Charsiana it was three hundre 
and eighty ; in some of the Western themes as much as four hundred. 

•* Constantine Porphyrogenitus, quoted above on pp. 182, 183, mentions th 
turma of Saniana as only five bands strong. 



9oo] 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 ^ or fighting line, 
the second the defensor^es ^ or reserve line.^ 

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 
bi(pa\ayyia^ and the rear half-band, eight deep, would receive the 
order "right about face" {l'no6Tpi-^ari) 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 crtscent-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.^ 

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 

^ KOiLfpa-upes. 2 5i<f>ivaop€i. 

^ 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 lactica, vii, § 45, that 
a brigadier or divisional general is to tell off his bands into defensores and cursores, and 
to be careful that each band gets a fair share of each sort of work. 

^ Leo, Tactica, ix. § 78. 

13 



194 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.^ 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).^ Often infantry were 
altogether wanting, and the whole army was composed of 
cavalry. Both Leo's Tactica and the Ua,padpoiui.rj UoXs/xov 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) [d. 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 ; ^ 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 easterr 
frontier theme has pursued a large Saracen raiding force anc 
brought it to bay. Having left behind all weak men and horses 
all recruits, and certain necessary detachments, the general ha5 
with him two weak divisions {tiirmae), each composed of twc 
brigades (drungi) of five regiments {bandcx) each. The individua 
band has been weeded down to two hundred or two hundrec 
and fifty men, but contains only picked troopers. The total o 
the host is only about four thousand six hundred men, thougl 

^ John Zimisces in his expedition against the Russians had thirteen thousand hors> 
and fifteen thousand foot (Leo Diaconus, viii. 4). 

^ The centre was formed of TravaLd-^poi 'nnroTai, behind whom were the infantry 
the wings of cavalry also (Leo Diaconus, iv. 3). 

^ 16 rplrov iroabv, 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 thirteei 
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 hug 
one. 



9oo] THE ARRAY OF A CAVALRY FORCE 195 

the two turmae, if present with their whole effective, would 
amount 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 
hundred 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 
\turma7'cJi), the second in command of the whole force, leads the 
line : ^ 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.^ 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 ^ (g G G in plan). 

3. Behind the second line, not to its rear, but on its 
flanks,* 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 vTspjispasTai. 
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 ^Xay/o^uXaxgc. 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. 

^ xii. 77. - xii. 90. ^ xviii. § 147. 

■* xii, § 30. This point, noted in the general directions for drawing up a cavalry 
array, is not repeated in Const, 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 hsdpot 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" = icxx> men. 
Third line . . . . .2 "bands" = 500 men. 

'TirepKepdaTaL ..... I "band" = 200 men. 
IIXa7to0i^XaAre? ..... I "band" = 200 men. 

'Ev^dpoL 4 "bands" = 800 men. 

General's escort. .... ^"band" = 100 men. 
To fill 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 hsdpoi, 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 

^ xii. § 40, 



PLATE VI 






A BYZANTINE CAVALRY FORCE OF TWO 'TURMAE' 
IN LINE OF BATTLE . 

Enemy's Line of Battle 







* HB SSJ ISJy '^ 




I 



,oo] CRITICISM OF BYZANTINE TACTICS 197 

ntervals of the second line. Leo intends them to deceive the 
memy's eye, and to give an impression of continuity and 
;olidity to the array.^ If the front line is broken, they are to 
etire, leave the intervals open, and draw up in the rear of thig 
;econd line, and between the two bands of the third line. There 
hey are to serve as a rallying point for the broken troops from 
he front, who will form up on each side of them. But in 
)ractical work this retiring to the rear at the moment when the 
emnants of the shattered first line were tumbling in upon them 
vould be a very hazardous experiment. There would be a great 
hance that, instead of the fugitives rallying upon the support, 
he support would be carried away by the fugitives, and all go 
)ff the field in disorder. Only the steadiest and coolest troops 
:ould be trusted to carry out the manoeuvre. Still, as we shall 
;ee from the battles which we are about to describe as instances 
)f Byzantine cavalry tactics, the troops of the empire were quite 
:apable of rallying and returning to the charge. 



i 



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 mihtary writen 
of the East-Roman Empire are always more satisfactory thar 
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 victory. Then 
is a certain amount of truth in the sneer, for the main lines o 
Byzantine strategy during the greater part of the history of th( 
empire are somewhat one-sided. They are almost entireh 
defensive in their scope, and pay little attention to the offensive 
In this respect they do but reflect the general condition anc 
needs of those who used them. From 600 to 800, and agaii 
from 1050 to 1453, the rulers of Constantinople were making 
strenuous fight for existence, and not aiming at offensive opera 
tions beyond their own borders. Between Heraclius' Persia 
campaigns (622-28) and Nicephorus Phocas' conquest of Cilici 
(964), the East- Roman generals never were able to contemplat 
an invasion on a large scale into hostile territory. The tacticc 
offensive they might often take, but it was always with th 
object of preserving or recovering their own lands, not with the 
of annexing those of their neighbours. Summed up shortly, th 
whole military history of these centuries consists in a strugg] 
to preserve Asia Minor from the Saracen, the Balkan peninsul 
from Slav, Bulgarian, and Turk,^ and the Italian themes fror 
Lombard and Frank. Of these struggles the first was far the mo.' 
engrossing : when once the pressure was taken off the Easter 

^ i.e. Avar, Magyar, Patzinak : perhaps one ought to include the Bulgarian al 
under this name. At least the Byzantine writers often place him in that categor 
See Leo, Tactica^ xviii. 

198 



9oo] 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 : ^ 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.^ 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- 

^ The UapaopofXT] UoX^nov, 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. § 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 18 13, 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 own 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 general; 
have not always been straightforward and honourable in theii 
observance of the customs of war.^ But they do not as a rule 
proceed to glory in their ingenuity and commit it to paper as < 
precedent for the future. There is ample evidence, not only fron 
the records of chroniclers, but from the chapters of Leo's Tactica 
that the East-Romans felt no proper sense of shame for some o 
their over-ingenious stratagems in war. It is with a kind of intel 
lectual pride in his own cleverness that the Imperial autho 
advises that if negotiations with a neighbour are going on, anc 

^ Napoleon certainly committed breaches of the laws of war as odious as any c 
which the Byzantines ever were guilty. None of them ever surpassed those mastei 
pieces of treachery and lying, — the seizure of the Vienna bridges in 1805 under pretenc 
of an armistice, and the occupation of the Spanish fortresses in 1S08. 



)o] FRAUD AND FORCE 201 

is intended to break them off, the softest words should be re- 
eved to the last day but one, and then a sudden expedition be 
unched against the enemy, who has been lulled into a belief in 
le certainty of peace. He is quite ready to send bribes into the 
Dstile camp. He recommends two ancient tricks that were 
ready a thousand years old in his own day. The first is that 
.' addressing treasonable letters to officers in the enemy's camp, 
id contriving that they shall fall into the hands of the com- 
lander-in-chief, in order that he may be made suspicious of his 
^utenants. The second is that of letting intelligence ooze out 
) the effect that some important person in the hostile country is 
icretly friendly, and adding plausibility to the rumour by spar- 
ig his houses and estates when raids are going on.^ Leo is not 
Dove raising the spirits of his own soldiers before a battle by 
iventing and publishing accounts of imaginary victories in 
nother corner of the seat of war. A trick too well known in 
iter as well as in Byzantine times is that of sending parle- 
lentaires to the enemy on some trivial excuse, without any real 
bject except that of spying out the numbers and intentions of 
le hostile forces. These and similar things have been tried in 
lodern times, but they are not now reconlmended in official 
aides to the art of war published under Imperial sanction.^ It 
i only fair to say that the same chapter which contains most of 
lem {Const, xx.) is full of excellent matter, to the effect that no 
lighted treaty or armistice must be broken, no ambassador or 
arlementaire harmed, no female captive mishandled, no slaughter 
f non-combatants allowed, no cruel or ignominious terms im- 
osed on a brave enemy. A few precepts of the rather futile 
nmorality of those which we have instanced above must not be 
llowed to blind us to the real merits of the strategical system 
ito which they have been inserted. The art of war as it was 
nderstood at Constantinople in the tenth century was the only 
ystem of real merit existing in the world ; no Western nation 
ould have afforded such a training to its officers till the sixteenth, 
r we may even say the seventeenth century. If some of its 

^ A device as old as the Punic Wars ! Hannibal tried it against Fabius. 

^ The most " Byzantine" piece of writing that I can recall in a modern campaign is 
utusofif's cynical despatch to the Emperor of Russia, avowing the trick which he had 
layed off on Murat a few days before Austerlitz. " In alleging the conclusion of an 
rmistice," he wrote, " I had nothing in view but to gain time, and thereby obtain the 
leans of removing to a distance from the enemy, and so saving my corps." Many 
len might have carried out the fraud : few would have openly boasted of it. 



202 THE ART OF WAR IN THE MIDDLE AGES [90 

precepts leaned a little too much towards the side of fraud, i 
may be pleaded that at any rate its methods were more human 
than those prevailing in any other part of the world at th 
time. 

But we are at present engaged in investigating the efficac 
and not the morality of the military customs of the Byzantine 
A survey of the main lines of the strategy and tactics of the 
armies must be our next task. 

The generals of the new Rome made it their boast that the 
knew how to face and conquer the various enemies of the empi] 
in East and West, by employing against each the tactical meai 
best adapted to meet their opponents' method of warfare. Tl 
Strategicon of Maurice gives an account of the Persian, Avc 
and Lombard and the methods to be used against them : Le 
three hundred years later, substitutes for these earlier foes tl 
Frank and Saracen, the Slav and Turk. His chapter dealir 
with them {Const, xviii.) is more detailed and more interesting th? 
the corresponding passage in his predecessor's work, and deserv 
reproduction, alike as showing the diversity of the tasks set b 
fore a Byzantine general, and the practical manner in which the 
were taken in hand. They serve, indeed, as a key to the whc 
art of war as it was understood at Constantinople. 

" The Franks and Lombards," says Leo, "are bold and dari] 
to excess, though the latter are no longer all that they once wei 
they regard the smallest movement to the rear as a disgrace, a; 
they will fight whenever you offer them battle. When th< 
knights are hard put to it in a cavalry fight, they will tu 
their horses loose, dismount, and stand back to back against ve 
superior numbers rather than fly. So formidable is the char 
of the Frankish chivalry with their broadsword, lance, and shie 
that it is best to decline a pitched battle with them till you ha 
put all the chances on your own side. You should take advanta 
of their indiscipline and disorder ; whether fighting on foot or 
horseback, they charge in dense, unwieldy masses, which cam 
manoeuvre, because they have neither organisation nor di 
Tribes and families stand together, or the sworn war-bands 
chiefs, but there is nothing to compare to our own orde 
division into battalions and brigades. Hence they readily 1 
into confusion if suddenly attacked in flank and rear — a thing Ci 
to accomplish, as they are utterly careless and neglect the use 
pickets and vedettes and the proper surveying of the countrysi 



oo] TACTICS USED AGAINST THE FRANKS 203 

'hey encamp, too, confusedly and without fortifying themselves, 
D that they can be easily cut up by a night attack. Nothing 
acceeds better against them than a feigned flight, which draws 
lem into an ambush; for they follow hastily, and invariably fall 
ito the snare. But perhaps the best tactics of all are to protract 
he campaign, and lead them into hills and desolate tracts, for 
hey take no care about their commissariat, and when their stores 
un low their vigour melts away. They are impatient of hunger 
nd thirst, and after a few days of privation desert their 
tandards and steal away home as best they can. For they are 
lestitute of all respect for their commanders, — one noble thinks 
limself as good as another, — and they will deliberately disobey 
)rders when they grow discontented. Nor are their chiefs 
hove the temptation of taking bribes ; a moderate sum of 
noney will frustrate one of their expeditions. On the whole, 
herefore, it is easier and less costly to wear out a Prankish 
irmy by skirmishes, protracted operations in desolate districts, 
md the cutting off of its supplies, than to attempt to destroy 
t at a single blow." 

The chapters (xviii. 80-101) of which these directions are an 
ibstract have two points of interest. They present us with a 
picture of a Western army of the ninth or tenth century, the 
ixact period of the development of feudal cavalry, drawn by the 
:ritical hand of an enemy. They also show the characteristic 
strength and weakness of Byzantine military science. On the 
)ne hand, we see that Leo's precepts are practical and efficacious ; 
3n the other, we see that they are based upon the supposition 
:hat the Imperial troops will normally act upon the defensive, a 
imitation 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 East-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 Eastern 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 ^) 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 scatterec 
bands, which swept along the enemy's front and around hi^ 
flanks, pouring in flights of arrows, and executing partial charge.' 
if they saw a good opportunity. On a fair open field, however 
they could be ridden down by the Byzantine heavy cavalry, wh( 
are therefore recommended to close with them at once, and no 
to exchange arrows from a distance. Steady infantry also the} 
could not break, and foot-archers were their special dread, sinc( 
the bow of the infantry-soldier is larger and carries farther thai 
that of the horseman ; thus they were liable to have their horse 
shot under them, and when dismounted were almost helpless, th' 
nomad of the steppes having never been accustomed to fight oi 
foot. The general who had to contend with the Turks, therefore 
should endeavour to get to close quarters at once, and fight then 
at the earliest opportunity. But he should be careful about hi 
flanks, and cover his rear if possible by a river, marsh, or defile 
He should place his infantry in the front line, with cavalry o 
the flanks, and never let the two arms be separated. Heedles 
pursuit by the cavalry was especially to be avoided,^ for th 
Turks were prompt at rallying, and would turn and ren 
pursuers who followed in disorder. But a proper mixture ( 
energy and caution would certainly suffice to defeat a Turkis 
host, because in the actual clash of battle they were man fc 
man inferior to the Imperial Cataphracti. These chapters woul 
have been the salvation of four generations of Western Crusadei 
if their chiefs had but been able to read them. Well-nigh ever 
disaster which the Crusaders suffered came from disobeying son" 

^ Apparently also the Bulgarians (xviii. §§ 42-44), as he speaks of them as 
Scythian race very like the Turks, and again, of their " differing little or not at : 
from each other in their way of life and their methods of war." 

2 Never let the cur sores get more than three or four bowshots from the defen.^i 
is Leo's general rule. 



o] TACTICS USED AGAINST THE TURKS 205 

le of Leo's precepts — from falling into ambushes, or pursuing 
o heedlessly, or allowing the infantry and cavalry to become 
parated, or fighting in a position with no cover for rear or 
inks. The Byzantines, on the other hand, made on the whole 
very successful fight against the horse-archers who overwhelmed 
» many Western armies. It is true that one huge disaster, the 
ifeat of Manzikert, brought on by the rashness of Romanus iv., 
as perhaps the most fatal blow that the empire ever received. 
ut, with this and a few other exceptions, the East-Roman armies 
ive a good account of themselves when dealing with the Turk, 
lexius Comnenus, though not a genius, was always able to 
ifeat the Patzinaks ; his son and grandson reconquered from 
e Seljouks half Asia Minor, and, even after the Latin conquest 
' 1204, Lascaris and Vatatzes held them back. It was not the 
)rse-archers of the older Turkish tribes, but the disciplined 
nissaries of the Ottomans that were destined to give the 
up de grace to the Eastern Empire. 

The third group of nations with which Leo deals are the 
avonic tribes — Servians, Slovenes, and Croatians, who inhabited 
e north-western parts of the Balkan peninsula. The space 
ivoted to them is much less than that spent on each of the 
her categories of the enemies of the empire. Leo remarks 
tat since their conversion to Christianity in the reign of his 
ther Basil, and the treaty in 869 which had made the Dalmatian 
id Bosnian Slavs, in name at least, vassals of the empire, they 
id given no trouble. They were a nation of foot-soldiers, and 
ily formidable when they kept to the mountains, where their 
chers and javelin-men, posted in inaccessible positions, could 
inoy the invader from a distance, or their spearmen make 
idden assaults on the flank or rear of his marching columns, 
uch attacks could be frustrated by proper vigilance, while, if 
irprised in the plains when engaged in a plundering expedition, 
ley could be easily ridden down and cut to pieces by the 
nperial cavalry, since they had no idea of discipline and no 
sfensive arms save their large round shields. Leo gives no 
ascription of the Russians, though they were already beginning 
) plague the themes along the Euxine coast.^ Had he devoted 
chapter to them, we should be the richer by some interesting 
2tails of their early military customs. Sixty years later, when 

^ Their first expedition had been in 865, and there was one in Leo's own reign 
907. 



2o6 THE ART OF WAR IN THE MIDDLE AGES [qoc 

they fought John Zimisces, they had adopted the armour an( 
tactics of their Varangian chiefs, and resembled the Northmei 
rather than the Slavs of the South, fighting with shirts of mai 
long kite-shaped shields, and battle-axes, and arraying themselve 
in well-ordered columns, which could often beat off cavalry. I 
took the most strenuous efforts of the gallant Zimisces and hi 
chosen horse-guards to break into these stubborn masses, an 
the battle of Dorostolon was one of the hardest fought an 
perhaps the most creditable of all the victories of the Byzantir 
armies (971). 

The longest and most interesting paragraphs in Leo 
Eighteenth " Constitution " are reserved for the Saracens, ar 
his description of them can be amplified by details from the vei 
interesting Uspl Uapadpo/nrig UoXs/iov, a work written about 980 by 
trusted officer of Nicephorus Phocas, who desired to preser 
his late master's precepts and practice in a literary shape. Tl 
little book is practically a manual for the governors of them 
on the eastern border, giving all the methods to be employ( 
in repelling Saracen raids, and all the precautions necessary f 
the execution of retaliatory invasions of Saracen territory. It 
especially valuable because, unlike the Tactica of Leo, it giv 
lavish historical illustrations and examples, and does not confi 
itself to precept. 

To deal with the Saracen, the most formidable enemy of t 
empire, the greatest care and skill were required. " Of 
barbarous nations," says Leo, " they are the best advised and mc 
prudent in their military operations." The commander who \ 
to meet with them will need all his tactical and strategic 
ability, the troops must be well disciplined and courageous 
the " barbarous and blaspheming Saracen " is to be driven back 
rout through the " clissuras " of Taurus. 

The Arabs whom Khaled and Amru had led in the sevei 
century to the conquest of Syria and Egypt had owed th 
victory neither to the superiority of their arms nor to 
excellence of their organisation. The fanatical courage of 
fatalist had enabled them to face better-armed and betl 
disciplined troops, as it nerved the Soudanese ten years ago 
face the breechloaders of our own infantry. We, who remem ' 
the furious rush that once broke a British square, cannot won ' 
that the troops of Heraclius, armed only with pike and sw( , 
were swept away before the wild hordes of the early Calij • 



)o] TACTICS USED AGAINST THE SARACENS 207 

is greatly to the credit of the East-Roman troops and the 
3use of Heraclius that Asia Minor did not suffer the same fate 
s Persia and Spain. But when the first flush of fanaticism had 
assed by, and the Saracens had settled down in their new 
Dmes, they did not disdain to learn a lesson from the nations 
ley had defeated. Accordingly, the Byzantine army served as 
model for the forces of the Caliphs. " They have copied the 
.omans," says Leo, " in most of their military practices,^ both in 
rms and in strategy." Like the Imperial generals, they placed 
leir confidence in their mailed lancers : they were no longer 
le naked hordes of the sixth century, but wore helms, shirts of 
lain - mail, and greaves. But the Saracen and his charger 
ere alike at a disadvantage in the onset : horse for horse and 
lan for man the Byzantines were heavier, and could ride the 
)rientals down when the final shock came. 

By the tenth century the Saracens had an art of war of their 
wn. Some of their military works have survived, though 
one, it appears, date back to the times contemporary with Leo. 
'hey had advanced very considerably in poliorcetics and forti- 
cation ; they had learned how to lay out and entrench their 
amps, and how to place pickets and vedettes. But they never 
lised a large standing army, or fully learned the merits of drill 
nd organisation. The royal bodyguards were their only regular 
roops ; the rest of the army consisted of the war-bands of chiefs, 
liscellaneous bands of mercenary adventurers, or the general 
^vies of tribes and districts. 

Two things rendered the Saracens of the tenth century 
angerous foes, — their numbers and their extraordinary powers of 
Dcomotion. When an inroad into Asia Minor was on foot, the 
)Owers of fanaticism and greed united to draw together every 
nquiet spirit from Egypt to Khorassan. The wild horsemen 
>f the East poured out in myriads from the gates of Tarsus and 
Vdana to harry the rich uplands of the Anatolic, Armeniac, 
nd Cappadocian themes. " They are no regular host, but a 
nixed multitude of volunteers ; the rich man serves from pride 
)f race, the poor man from hope of plunder. They say that 
jrod, ' who scattereth the armies of those that delight in war,' is 
)leased by their expeditions, and has promised victory to them. 
Those who stay at home, both men and women, aid in arming 
heir poorer neighbours, and think that they are performing a 

^ Tactica, xviii. § 120. 



2o8 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." ^ 

Once clear of the passes of Taurus, the great horde oi 
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 day^ 
one of the richest in the world. It was only exceptionally thai 
the invaders were aiming at serious conquests and halted tc 
besiege a fortified town. The memory of the awful failures 
the two great hosts that perished before Constantinople in 6^^^ 
and 718 seems to have been deep impressed in the minds of the 
Mohammedan rulers and generals. The two last attempts a 
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 tasl 
before him that he gave up his project and retired. The armie 
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, the; 
pushed on for some days at an almost incredible speed, for thei 
baggage was all laid on camels or sumpter beasts, and their foot 
soldiery were either provided with horses of some sort or take; 
up on the cruppers of the cavalry.^ They made for the distric 
that they had marked out for plunder, and trusted to arrive i 
such haste that the natives would not have had time to gathe 
in their property and take shelter within walled towns. 

Now was the time for the Byzantine general to show hi 
mettle. If he was a competent commander, he would have ha 
regular outposts, relieved every ten or fifteen days, to watch th 
passes. The moment that these were driven in, they would tak 

^ Leo here adds, xviii. § 129: "And would that we Christians did the sam 
For if all of us, both soldiers and those who have not yet borne arms, could agree 
strengthen our hearts and go forth together, if every man armed himself, and t) 
people gave their money to equip such a host, and their prayers to help it, th< 
marching against that race which blasphemes our Lord and God, Christ, the King 
all, we should obtain victory. For the Roman armies being increased manifold, ai 
furnished liberally with all weapons of war, and abounding in military skill, ai 
having heaven as their aid, could not fail to crush the barbarous and blasphemii 
Saracen." This surely is the spirit of the Crusader, appearing two hundred yeo 
before its time. 

2 Toi>s 5^ Tre^oyj aindv (pipovaiv ■^ ^0' lttitcov Idiuv oxov/x^vovs, t) Siriadep tQiv ,Ka^a 
\apiu)P Kadrjimevovs (xviii. § 1 15). 



)oo] HOW TO DEAL WITH SARACEN RAIDS 209 

;he tidings to the chief town of the theme, and to the nearest com- 
nanders of bands and turmae. While the main body of the cavalry 
j( the theme concentrated under the strategos at a central point, 
t would be the duty of the turmarch into whose district the 
-aid had come, to collect the nearest two or three bands in haste, 
md to hang on to the skirts of the invading force at all costs. 
For even a small observing force compels the invaders to move 
:autiously, and to abstain from letting their men straggle for 
jlunder. Meanwhile, all the disposable foot - soldiery of the 
:heme would be hurried off to seize the mouths of the passes 
Dv which the enemy would probably return. These were not 
io numerous but that a competent officer might make some 
provision for obstructing them all.^ 

To ascertain the enemy's route and probable designs, the 
:ommander of the theme must spare no pains. The turmarch 
:harged 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 
sating 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' Uepl Uapa8po/xijs llo\4fiov, cap. i. § i. The chapter 
is really excellent ; it might be used on the Indian north-west frontier to-day, so 
practical is it. 



THE ART OF WAR IN THE MIDDLE AGES 



I 



Borderers) were of comparatively unfrequent occurrence, and 
was not often necessary to " set all the rest of the themes of tl 
East marching," each with its picked corps of four thousand ( 
four thousand five hundred cavalry. If needed, however, Le 
states that thirty thousand cavalry of the best quality could I 
collected in a moderate space of time. A most perfect instanc 
of such a concentration had taken place in A.D. 863 (though Le 
does not mention it ^), when a great Saracen army under Oma 
the Emir of Malatia, had been completely surrounded an 
absolutely exterminated by the skilful and simultaneous appea 
ance of no less than ten contingents, each representing a theme 
The more typical Saracen inroad, however, was on a smalk 
scale, and only included the warriors of Cilicia and Northei 
Syria, assisted by casual adventurers from the inner Mohan 
medan regions. To meet them the Byzantine commands 
would have no more than the four or five thousand horseme 
of his own theme. When he came up with them, they woul 
probably turn and offer him battle : nor was their onset to t 
despised. Though unequal, man for man, to their adversarie 
the Saracens were usually in superior numbers, and alwa} 
came on with great confidence. " They are very bold whe 
they expect to win : they keep firm in their ranks, and stan 
up gallantly against the most impetuous attacks. When the 
think that the enemy's vigour is relaxing, they all charge togeth( 
in a desperate effort." If this, however, failed, a rout generall 
followed, " for they say that all misfortunes come from God, an 
if they are once well beaten, they take it as a sign of divir 
wrath, and altogether lose heart." Their line once broken, the 
have not discipline enough to restore it, and a general sauve qi 
peut follows. Hence a Mussulman army, when routed, could t 
pursued a Voutrance^ and the old military maxim, Vince sed 1 
niniis vincas, was a caution which the Byzantine officers coul 
disregard. 

In the actual engagement with the Saracen foe, the tactic 

^ Perhaps because the reigning emperor was Michael ill., whom Basil i. (Lee 
father) had murdered. 

- Having sacked Amisus and ravaged Paphlagonia and Galatia, Omar found h 
way home blocked by the contingents of the Anatolic, Obsequian, and CappadociE 
themes ; at the same time those of the Buccellarian, Paphlagonian, Armeniac, ar 
Colonean themes encompassed him on the north ; and that of the Thracesian them' 
strengthened by European troops of the Macedonian and Thracian themes, closed ; 
on the west. The Saracens were absolutely exterminated. 

^ Nic. Phoc. xxiv. § 10. 



)oo] SARACEN METHODS OF WAR 211 

•ecommended were those of the double Hne, with flank-guards, 
eserve, and outlying detachments to turn the enemy, which we 
lave described in the section dealing with the organisation 
)f the Byzantine army. The Saracens were accustomed to 
irray themselves in one very deep line, which Leo calls a 
;olid oblong (^rsrpdyuvov xa/ s'jn/j^riKr} 'Trapdra^iv). Their cavalry were 
practically the sole force that gave trouble, the foot being a 
nere rabble of plunderers, which would never stand. Their 
)nly useful infantry were composed of Ethiopian archers, but 
hese, being wholly destitute of defensive armour, could never 
ace the Byzantine footmen. In battle the single heavy line 
)f the Orientals must under ordinary circumstances ■ give way 
Defore the successive charges of the three Byzantine lines. The 
generals of the East had already discovered the great precept 
vhich modern military science has claimed as its own, that " in 
I cavalry combat the side which holds back the last reserve 
nust win." They were equally masters of the fact that this 
ast reserve should be thrown in on the flank rather than on the 
Vont of the enemy. It was not, therefore, without reason that 
;he author of the Uapadpo,<M7i exclaims that " the commander 
,vho has five or six thousand of our heavy cavalry and the help 
)f God needs nothing more." ^ 

It would sometimes, however, happen that the Saracens were 
lot caught on their outward way, and that the forces of the 
Byzantine general only closed in on them as they were retreating.^ 
Loaded with booty, the raiders would be constrained to move 
"ar more slowly than on their advance ; their camps, too, would 
3e filled with captured herds and flocks, laden waggons, and 
:roops of prisoners. In this case Nicephorus Phocas recom- 
mended a night attack, to be delivered by infantry or dismounted 
:avalry. " Send three infantry bands, ranged a bowshot apart, 
;o charge into each flank of their camp," says the emperor, 
' assail the front a little later with your main body of foot, and 
eave 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." ^ 

^ Nic. Phoc. Preface, § 15. 
- Nic. Phoc, xvii. § 15. 

^ ei 8e (rufi^rj Xvdijvai. tt]v irapaTa^iu, 5t' eavrwv d<n'crraTot Kaidi/eirlarpocpoL yevdfievoi 
ti6vip T(p (T(j}driva(. i\a{ivov<Tiv (xviii. 116). 



2 12 THE ART OF WAR IN THE MIDDLE AGES [96. 

But success was most certain of all if the invaders could b" 
caught while retreating through the passes of Taurus. If th 
infantry of the theme had succeeded in reaching the defiles an( 
posting themselves there before the retreating enemy arrivec 
while at the same time the pursuing cavalry pressed them i 
the rear, the Saracens were lost. Wedged in the narrow roac 
with their line of march mixed with countless waggons an- 
sumpter-beasts laden with spoil, they were quite helples 
They could be shot down by the archers, and would not stan 
for a moment when they saw their horses, "the Pharii whor 
they esteem above all other things," struck by arrows from 
distance; for the Saracen, when not actually engaged in clos 
combat, would do anything to save his horse from harm.^^ 

The most noted instance of a victory of this kind was the 
won in 963 by Leo Phocas, brother of Nicephorus, over tt 
hosts of Seif-ed-dauleh ben Hamdan, Emir of Aleppo. Thoug 
he had with him only the forces of his own theme of Charsian? 
Leo captured or slew the whole of the Saracen army, recovere 
much plunder, liberated many thousands of Christian prisoner 
and bore off in triumph the standard and the silver carr 
equipment of the emir. Mohammedan historians confess tl 
greatness of the disaster, though they reduce the number 
their slain to three or four thousand.^ Seif-ed-dauleh himse 
escaped with three hundred men only, by climbing an almo 
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 Romi 
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 t 
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 firmnt 
and daring, and could be attacked with great advantage. Mu 
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 ?hto Cappado( 

1 Leo, xviii. § 135. ^ Nic. Phoc. Preface, § 15. 

^Jemaleddin, p. 134; Abulfeda, ii. 469. 



^o] SUCCESSFUL ADVANCE AGAINST THE SARACENS 213 

- Charsiana. On hearing of such a retaliatory expedition, the 
[oslems would often return home to defend their own borders.^ 
his destructive practice was very frequently adopted, and the 
ght of two enemies each ravaging the other's territory with- 
At attempting to defend his own was only too familiar to 
le inhabitants of the borderlands of Christianity and Islam, 
icursions by sea supplemented the forays by land. " When 
le Saracens of Cilicia have gone off by the passes, to harry the 
:)unty north of Taurus," says Leo, "the commander of the 
ibyrrhaeot theme should immediately go on shipboard with all 
v'ailable forces, and ravage their coast. If, on the other hand, 
le Cilicians have sailed off to attempt the shore districts of the 
nperial provinces, the clissurarchs of Taurus can lay waste the 
irritories of Tarsus and Adana without danger." 

All through the tenth century the Saracens were growing less 
ad less formidable foes, owing to the gradual dropping off of 
le outlying provinces of the empire of the Abbassides, who by 
le end of the period were masters of little more than the 
Euphrates valley, and were dominated even in their own palace 
y their Turkish guards. The Byzantine realm, on the other 
and, under the steady and careful ministers who served 
le Macedonian dynasty, was at its very strongest. For a 
undred and fifty years after the accession of Basil I., the empire 
'as always advancing eastward, and new themes were continu- 
ity being formed from the reconquered territory. There is a 
reat difference of tone between the language which Leo, writing 
bout 900, and the author of the UapudpofMTj, writing about 980, 
se concerning the Saracen enemy. To the former they are 
till the most formidable foes of the empire ; the latter opens 
is preface with the words : " To write a treatise on frontier 
perations may seem at the present day no longer very 
ecessary, at least for the East, since Christ, the one true God, 
as in our day broken and blunted the power of the sons of 
shmael, and cut short their raiding. . . . But I write neverthe- 
iss, thinking that my experience may be useful, because I was 
n eye-witness of the commencement of our successes and of 
lie application of the principles which led to them. Through 
he use of these principles I have seen small armies accomplish 

^ The author of the HapaSpofxifi speaks of this device, quoting it as a good piece 
f counsel given by Leo, and gives as example an occasion when the siege of Misthea 
as raised by means of a retaliatory raid against Adana (xx.). 



214 THE ART OF WAR IN THE MIDDLE AGES [98 

great feats. What once, when the Saracens of the border wer 
strong, seemed impossible to a whole Roman army, has been c 
late carried out by a single good general with the forces of 
single theme. By the use of these principles I have seen a force 
though too small to face the enemy in open fight, yet defea 
his purpose, and preserve our borders unravaged. The syster 
was first, as far as I know, utilised in modern times by Barda 
Caesar,^ who foiled the Saracens of the Tarsiot border not one 
but ten thousand times, and erected countless trophies over then 
Constantine Melei'nos, strategos for many years in Cappadoci? 
won magnificent successes by using these principles.^ Bu 
Nicephorus Phocas, that prince of immortal memory, accorr 
plished by their use feats that defy description and enumeratioi 
He it was who bade me write down the system, for the use c 
future generations. And this I do with the more readines 
because it can be applied not only to the eastern border, but t 
the western, as I (who have served most of my time on th 
latter) can state from my own experience." 

By the end of the tenth century the Byzantines were habiti 
ally taking the offensive against the Saracens, and, instead c 
seeing Cappadocia or Phrygia ravaged, were themselves pushin 
their incursions almost to the gates of Damascus and Bagda( 
The conquest of Cilicia by Nicephorus Phocas was but the fin 
of a series of advances which promised ultimately to restore t 
the empire the frontier that it had held in the days of Justiniai 
Antioch was conquered, the Emirs of Aleppo and Tripoli wer 
made tributary, and kept in that position for sixty years. Eve 
after the death of Basil II., the greatest soldier of the Easter 
realm, the Imperial borders continued to advance eastward 
Edessa was captured in 1032, and a new theme was establishe 
in Mesopotamia. The whole of Armenia was annexed in 104 
and Constantine IX. might have boasted that his provinces ej 
tended farther to the East than those of any of his predecessoi 
since Trajan. 

^ This, I suppose, was the unfortunate Bardas Caesar who was murdered by h 
nephew Michael iii. in 866. There had been some great victories in his da 
notably that over Omar (see p. 210), and he is said to have devoted much attentic 
to military affairs, but it is surprising to find him given such a marked place by tl 
author of the Uapadpofx.-r}. Did his exploits inspire the sections on border warfare 
Leo's Tactica'i 

^ There were several good generals of this name. I suppose this to be the 01 
who ruled Cappadocia about 960 a.d. 



37 1] THE COMING OF THE SELJOUKS 215 

But at the moment when the East - Roman boundaries 
ached 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 
le turning-point in the history of the great East - Roman 
:alm. 



CHAPTER IV 

DECLINE OF THE BYZANTINE ARMY — IO71-1204 

THOUGH the internal condition and administration of tt 
empire had been steadily deteriorating since the death ( 
Basil II. (1024), it cannot be said that its army showed any declir 
till the very day of Manzikert. Indeed, as we have already see 
the Imperial frontier continued to advance down to the momei 
of that disaster, and the first advance of the Seljouks was mt 
without wavering. For some years the Turks had no high( 
aim than to win booty by sudden inroads into Asia Minor. C 
their raiding bands some were turned back, and some cut 1 
pieces ; but their numbers were so great that the line of defenc 
could not be held everywhere, and on different occasions Caesare 
Iconium, and Chonae fell into their hands. No lodgment, hov 
ever, was made in the empire, and the fact that the decisi\ 
battle was fought so far east as Manzikert, in farther Armeni 
hard by the Lake of Van, shows that the hold of the governmei 
on its frontier provinces was not yet shaken. 

The Seljouks of Alp Arslan were in tactics just like the Turl 
whom Leo the Wise had described a century and a half befor 
They only differed from the Patzinaks and other Western trib( 
of the same blood by their enormously superior numbers. N 
such formidable invasion had befallen the empire since the da} 
of Leo the I saurian, and to meet it there sat on the Byzantir 
throne a gallant hot-headed soldier with a doubtful title an 
many secret enemies. Romanus Diogenes had been lately raise 
to the purple by his marriage with Eudocia, the widow ( 
Constantine XL, and reigned as colleague and guardian of h( 
young son Michael. He knew that he was envied and hate 
by many of his equals, who had aspired to fill the same place 
hence he was nervously anxious to justify his elevation b 
military success, as his great predecessors, Nicephorus Phocr 



07 1] BATTLE OF MANZIKERT 217 

nd John Zimisces, had done. He was in the field for almost 
he whole of the three uneasy years for which he reigned (1068- 
' I ) ; and if energy and ceaseless movement could have driven off 
he Seljouks, he must have been successful. But he was a bad 
;eneral, easily distracted from his aims, and too quick and rash 
n all his actions. 

In the spring of 107 1 Romanus collected a very large army, 
it least sixty thousand strong, and betook himself to the extreme 
eastern corner of his dominions, with the intention of meeting 
:he Turks at the very frontier, and recovering the fortresses of 
^khlat and Manzikert, which had fallen into their hands. He' 
lad retaken the latter place, and the former was being besieged 
3y a detached division of his army, when the main host of the 
Seljouks came upon the scene. It was a great horde of horse- 
irchers, 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,^ 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 
uti 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 anc 
thirsty, and he had left his camp insufficiently garrisoned, so thai 
he was anxious to return to it, lest it might be surprised in hi.' 
absence. Accordingly, he gave orders to face about and retire 
Then began the disasters of the day : the order to retreat wa.' 
not executed with the same precision in all the divisions of the 
host ; those on the flanks received it late, did not understanc 
its cause, and, when they wheeled about, did not keep theii 
dressing with the centre. Gaps began to appear between severa 

^ These are, I suppose, the eraipoL and rb apxovTCKov of which Bryennius speak 
in his account of the battle. 



oyi] BATTLE OF MANZIKERT 219 

»f the corps. The Turks, according to their custom, commenced 

o close in again when the army commenced its retreat. They 

nolested the retiring columns so much that Romanus at last 

;ave orders to face about again and beat them off. The whole 

ront line carried out this order, but the reserve under Andronicus 

lid not : out of deliberate malice, as most of the authorities allege, 

his treacherous commander refused to halt, and marched back 

apidly to the camp, observing that the day was lost, and the 

;mperor should fight out his own battle. To lose the rear line, 

md to be left without any protection against circling move- 

nents on the flanks, was fatal. The Turks began to steal round 

he wings and to molest the fighting line from behind : they 

)articularly concentrated attention on the right wing, which, 

rying to face both ways, fell into disorder in the twilight, and 

it last broke up and fled. The victors at once fell on the flank 

md rear of the centre, where the emperor made a gallant defence, 

:harged repeatedly both to flank and rear, and held his own. 

3ut the European troops in the left wing had got divided from 

he centre, and, after fighting a separate battle of their own, gave 

vay, and were driven off the field. Thus left isolated, Romanus 

encouraged his men to stand their ground, and held out till 

lark, when the Turks broke into his column and made a dread- 

ul slaughter. The emperor's own horse was killed beneath him ; 

le was wounded and taken prisoner, with many of his chief 

)flicers : the whole centre was cut to pieces, and not a man of it 

escaped. 

Thus Romanus Diogenes, like Crassusof old, paid the penalty 
'or attacking a swarm of horse-archers in a open rolling country, 
.vhere he had cover neither for his flanks nor for his rear. It is 
^nly fair to say that he would have in all probability brought 
lome his army without any overwhelming loss but for the abomin- 
able misconduct of Andronicus Ducas. When encompassed by 
'he 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 infinitum^ 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 ^gean 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 a« 
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 
107 1 -108 1. 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 Nicsea tc 
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, foi 
the would-be Caesars who were contending for the throne lefl 
the Seljouks alone. Both Michael Vll. and his foe, the usurpei 



1079] DISAPPEARANCE OF THE ARMY OF THE EAST 221 

Xicephorus Botaniates, actually bought the aid of Turkish 
luxiliaries by formally surrendering whole provinces. In 1080 
;he 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 
[lave 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 
jreat 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.^ 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 

^ '0 ^aaiXevs Mtxa^jX i8(jt}v to tt]S '"Eipas crpdrevfia dirav T)8ri iKkekonrbs, W5 virox^ipcov 
tQv TovpKwv yevdi-ievov, ^(ppdvTLcre tDs otou re crTpdrev/jia KaracrTTJaai vedXcKTOv, Kai 5?^ 
Ttvas Tuu €K TTJs 'A(Tlas BiaaTap4vTU}v Kai erl pnadi^ SovXevSvruv crvW^yuv, dcjpaKds re 
iv^dve Kai dvpeovi ididoVy etc. etc. (Nic. Bry. iv. § 4). 



222 THE ART OF WAR IN THE MIDDLE AGES [107, 

uneasy throne at Constantinople, sent against the rebel Nicephoru 
Bryennius his general Alexius Comnenus, whom he had lateh 
made " Domestic of the Scholae," i.e. commander of the Imperia 
Guard. Nearly all the European provinces had fallen away t( 
Bryennius, and as Asia had been overrun by the Turks as far a 
Nicaea and the Propontis, the ruler of Constantinople was no 
able to put into the field so large an army as the insurgents. 

The armieSjboth wholly composed of cavalry, met at Calavrytc 
hard by the river Halmyrus. Comnenus, as the weaker of the twc 
waited to be attacked, and chose a position with a comparativel; 
narrow front, apparently where a road crossed the slope of a hill 
on the left of his position were some hollows, screened from th 
eyes of those approaching from the plain by a rise in the grounc 
Comnenus drew up his main body, composed of the " Immortals 
whom Michael Ducas had organised, and a body of Frankis! 
mercenaries, across the road. He hid his left wing in the hollow; 
ordering them to keep wholly out of sight till the enemy shouL 
have passed them, and then to charge in upon Bryennius' righ 
flank. His right wing, composed of garrison troops strengthenc' 
by a considerable force of Turkish mercenaries — all horse-archer 
— was placed under the command of Catacalon ; it was i 
military terminology " refused," and ordered to devote its who) 
attention to preventing the enemy from turning the flank of th 
main body. Thus, to use the technical terms of Leo's Tactia 
Comnenus had Wihf>ui or b'rrspzspdffTai on his left wing, an 
crXay/o^oAaxej on his right. 

Bryennius, on the other hand, came on with his host divide 
into three parallel columns. The right wing, five thousan 
strong, was led by his brother John, and contained the cavalr 
of the theme of Thessaly and the veteran remnants of the ol 
army of Italy, which had long served under John Maniake 
against the Normans and Saracens. The left wing, unde 
Tarchaniotes, three thousand strong, was composed of Mace 
donian and Thracian regiments. The centre, led by the usurpt 
himself, was also formed from Macedonian and Thracian corp 
strengthened by a picked body of ap^ovng — local nobles and the 
followers. But Bryennius intended to strike his chief blow wit 
a body of Scythian (Patzinak) horse detached from his mai 
army and moving a quarter of a mile to its left, with orders 1 
turn the right of Alexius' line, — serving in fact, as Leo woul 
have said, as I'Tripxipdera/. 



079] BATTLE OF CALAVRYTA 223 

When the rebel army came level with the hollows where the 
mperialist left was concealed, the hidden troops suddenly issued 
orth and charged John Bryennius in flank, while Comnenus 
md his main body rode down upon the usurper's own central 
iivision. Both these attacks failed : John Bryennius wheeled to 
lis right in time, and beat off the attack of the troops in ambush. 
N'icephorus Bryennius defeated the squadrons of the Immortals, 
md drove them off the field, while the Prankish mercenaries 
vho formed the remainder of Comnenus' centre were wholly 
nicompassed by the rebels,^ and cut off from the possibility of 
etreat. Meanwhile, on the extreme right of the Imperialist 
irmy, the garrison troops under Catacalon had been charged 
md routed by Bryennius' flanking force of Patzinak horse. 
The victorious barbarians went off in wild pursuit of the 
ugitives, and seem to have overlooked the other corps on the 
[mperialist right, the Turkish auxiliaries, who found themselves 
eft without an enemy in sight.^ When the Patzinaks returned, 
;hey began plundering their own employer's camp, instead of form- 
ng up to aid him in an engagement as yet by no means ended. 

Alexius Comnenus had extricated himself with difficulty 
"rom the melee in the centre, and retired over the brow of the 
lill, where he at once halted and began endeavouring to rally 
lis broken troops. During the combat he had charged into the 
personal escort of the usurper, and had chanced to come upon 
:he squires who led the second charger of Bryennius, adorned 
mth 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 

^ 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. 

2 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. Bryenniue 
himself and the centre division were surrounding the Franks o 
the Imperialist army, who, when they had been cut off, hac 
dismounted, and offered to surrender. The commanders o 
these mercenaries were standing on foot before Bryennius anc 
doing homage to him just as Alexius came down the hill for hi: 
second charge. 

Though much surprised by the return of the enemy to th 
fight, Bryennius and his men came boldly forward. Alexiu 
set his Turks to skirmish, and bade them empty their arrow 
into the disordered rebels before he made any endeavour ti 
close ; he wished to fight a cautious battle, avoiding any genera 
charge. As the enemy advanced, he retired before them slowl; 
till he had reached the point far up the hill where he had left hi 
ambush. When he saw the flanks of Bryennius exposed to th 
lateral attack, he halted, faced to the front, and charged. At th 
same time the concealed troops, dashing out " like a swarm ( 
wasps," attacked the rebels on both flanks. Already muc 
disordered, and with hundreds of horses disabled by the Turkis 
arrows, the squadrons of Bryennius could not face the charg 
but broke and fled. The rebel chief himself, with a small bod 
of devoted followers, refused to give ground, fought to the las 
and was finally dragged from his charger and taken prisoner.^ 

The battle of Calavryta was fought in the time of tl 
Byzantine decadence which set in after Manzikert : there we 
many raw troops in both armies,^ and a large proportion < 
foreign auxiliaries not drilled or disciplined after the tradition 
methods of the Imperial army. Nevertheless, the incidents 
the fight show the main characteristics of the system whii 

^ Most of the details of this interesting fight came from Anna Comnena, who h: 
for a lady, a very fair grasp of things military. No doubt she accurately put down 1 
father's account of his doings, and we are really reading Alexius' versions of his fig 
Deducting the Homeric diction and the far too hairbreadth 'scapes of the narrat 
they are very favourable specimens of Byzantine military annals. 

2 Alexius complained that the majority of the Immortals were recruits x^f's re 
irp(brjv ^icpovs Tjix/iiuoi. Kai dopara. 



ii5o] 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. Frankish 
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 : ^ 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 Asi^ 
Minor from the Seljouks, subdued the Franks of Antioch, anc 
routed the Magyars beyond the Danube. But they never buil 
up a real national army. When the strong hand of Manuel wa: 
removed, and the wretched A^ngeli sat upon the Imperial throm 
(i 185— 1204), the military machinery of the empire went t( 
wrack and ruin. The weak and thriftless emperors Isaac II 
and Alexius in. were neither able to find money to pay thei 
troops nor to maintain their discipline. A state which relies fo 
its defence on foreign mercenaries is ruined when it allows then 
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, th< 
defenders struck for higher pay and refused to charge. Th( 
city fell, and the old Byzantine military organisation passee 
away. 

^ There seems to have been some revival of local native forces during th 
existence of the empire of Nicsea (1204-61). We hear of militia in Bithynia unde 
Lascaris and Vatatzes, and their disbandment by Michael Palaeologus is said t 
have been one of the causes of the successful advance of the Ottoman Turk 
(Pachy meres, i. 129). 



BOOK V 

THE CRUSADES 
IO97-I29I. 



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 effect a lodgment south 
of the Mediterranean.^ In Spain, too, the balance had turned 
definitely in favour of the Christians ; Toledo had fallen in 
1085, and with 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. 

^ The landmarks in the history of the struggle of the ItaHans 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 lOil. 



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-important 
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 1000 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 
the 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 sudden 
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 ir 
1 291.1 

With the causes of the Crusades we are not concerned ; nor 
are their religious, social, or commercial aspects our province. 
It is with their military side alone that we have to deal — a 

^ In a way we might say that the last effects of the Crusades were not over ti 
the Turks evicted the Venetians from Cyprus (1571), Crete (1669), and the Moret 
(1715). 



096] STATE OF THE LEVANT IN 1096 231 

ubject sufficiently vast and varied to fill many volumes if 
ve had space to descend into detail. 

Stated broadly, the problem which was started in 1096, and 
asted till 1291, was whether feudal Europe, with the military 
;ustoms and organisation whose development we have been 
racing, would prove strong enough to make a permanent 
odgment in the East, or perchance to make good the whole of 
he ancient losses which Christendom had suffered at the hands 
)f the Saracen and Turk from the days of Heraclius to those of 
rlomanus Diogenes. 

The state of the Moslem powers of the Levant in 1096 was 
)n the whole favourable for the assailants who were about to 
hrow themselves upon Syria and Asia Minor. It had seemed 
n the early days of the Turkish invasion, and soon after the 
atal day of Manzikert, that a single great empire might establish 
tself in Western Asia under the house of Alp Arslan. But no 
5uch result had followed the conquests of the Seljouks. At the 
noment when the first Crusaders crossed the Bosphorus, the 
Sultanate of Roum had separated itself from the main body of 
:he 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 ; 
1 few years earlier they would have foutid their task far harder, 
md the heroic courage which habitually saved them from the 
:onsequences 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 
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 [10915 

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 efificiency 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. 



1 



CHAPTER II 

THE GRAND STRATEGY OF THE CRUSADES 



^ OOKED at from the most general point of view, the 
I V Crusades, as a whole, may be said to have had two main 
ejects. The first was to relieve the pressure of the Turks on 
onstantinople, which had been so dangerous ever since the day 
:' Manzikert. The second was to conquer the Holy Land and 
istore its shrines to the custody of Christendom. Both of these 
urposes were to a certain extent accomplished : the Turkish 
ontier in Asia Minor was thrust back many scores of miles, and 
early two centuries elapsed before the Seljouk Sultans were 
ble to recover their lost ground. Jerusalem was stormed, and 
)r ninety years remained in the hands of the Franks. But 
lese ends were achieved in the most wasteful manner, by the 
lost blundering methods, and at the maximum cost of life and 
laterial. 

One of the main causes of the disasters of all the crusading 
rmies was a complete lack of geographical knowledge. A 
ursor}^ glance at the itineraries of the various expeditions 
hows that the majority of them were chosen on the most 
inhappy principles, and were bound to lead those who adopted 
hem into grave peril, if not to utter destruction. We must not 
)lame the men of the eleventh and twelfth centuries overmuch 
or their errors : to a great extent they were inevitable in face 
)f their utter want of geographical information concerning the 
:ountries of the Levant. Any misdirection was possible in days 
vhen the whole available stock of information in the West con- 
sisted of garbled fragments of the ancient Roman geographers, 
einforced by a certain amount of oral information gathered 
rom 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 



234 THE ART OF WAR IN THE MIDDLE AGES [109 

the control of the seas beyond Sicily, and the navies of thi 
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 o 
the sea. The road to the Bosphorus across Hungary anc 
Servia had only become practicable after the conversion of the 
Magyars to Christianity (1000-61).^ It had not yet been adoptee 
as a channel for commerce or a route for pilgrimages. Beyonc 
Constantinople there was only such information to be obtainec 
as the Greeks could give. This information was not always 
honestly purveyed : the Byzantine emperors had purposes o. 
their own to serve, and often sent the pilgrim hosts on itinerario* 
which suited themselves rather than those which were be$1 
adapted for the purposes which the Franks had in view. Wc 
need not believe the constant complaints of the Western 
chroniclers that the Comneni deliberately guided the pilgrims tf 
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 v/hat he 
could. The recovery of Lydia and Mysia was tindoubtedly due 
to the first Crusade, and that of Northern Phrygia and Galatia 
to the Crusade of i loi. 

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 iioi, 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 dAgiles 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. 



;oo] THE CRUSADERS' IGNORANCE OF GEOGRAPHY 235 

ult if, after recovering Ancyra for the empire, they were starved 
id harassed in the trackless lands beyond the Halys, so 
lat only a few thousands of them finally struggled back to 
inope. It must also be remembered that the Byzantines them- 
;lves, though they had all the old Roman road-books, and 
aborate data for the distances in their own lost " themes " in 
sia Minor, were not able to give accurate information concern- 
[g the present condition of the land. The Turks had wrought 
) much damage in the last twenty years, burning towns, filling 
3 cisterns, and extirpating the population of whole districts, 
lat the old information concerning the interior had no longer 
s full value. Routes easy and practicable before 1070 were 
roken and desolate by 1097. The many perils which the 
omneni suffered in their own campaigns in inner Asia Minor 
-e sufficient proof that their information as to the land was no 
mger reliable. 

It would be unfair, therefore, to attribute to wilful misdirection 
a the part of the Greeks the whole of the misadventures of the 
rusaders in Asia Minor. The larger part of their troubles were 
f their own creation, and came from carelessness, presumption, 
nprovidence, and selfishness. Even when put upon the right 
md, they were apt to go astray from blind conceit or want of 
iscipline. This comes out most clearly from the fact that 
lany crusading expeditions miscarried in Hungary or the 
lavonic lands just to the south of the Danube, before they ever 
cached Constantinople. For an elaborate example of a wrong- 
eaded choice of route, nothing can be more striking than that 
^hich Raymond of Toulouse and the Provencals selected in 
096. In all South-Eastern Europe there is no district more 
estitute of roads and more inhospitable than the Illyrian coast- 
ne. But Raymond chose to march from Istria to Durazzo 
hrough the stony valleys and pathless hills of Dalmatia, 
VFontenegro, and Northern Albania, among the wild Croats and 
' ^orlachians. It is surprising that he was able to bring half his 
ollowing to Durazzo : he must have failed altogether had not his 
xpedition been by far the best equipped and the most carefully 
)rovisioned of all those which set out for the first Crusade. 

For the pilgrimage to Syria there were two great alternatives 
>pen — the land voyage by Constantinople and the sea voyage 
lirect to the Levant. The latter was in every way preferable 
vhen once the sea routes had been surveyed. But at the time 



236 THE ART OF WAR IN THE MIDDLE AGES [lo 

of the first Crusade it was practically unknown : only t 
adventurous sailors of Venice, Pisa, and Genoa attempted : 
the French, Burgundians, Provencals, Germans, and Lombar 
all preferred the longer road by Constantinople. Even in lat 
times the landsmen's horror of the water drove a majority oft 
Crusaders to shun the voyage by sea : all the greater chiefs 
the second Crusade, and Frederic Barbarossa among the leade 
of the third, persisted in taking the land route. The first gre 
expeditions made by sea by any save the Italian powers we 
those of Philip Augustus and Richard of England in 1190. B 
from that time onward the advantages of the direct voyage 
Palestine seem to have been recognised, and all the lat 
Crusaders preferred it. It was obviously better to arrive fre: 
and unwearied at Acre or Tyre, rather than to run the thousar 
risks from Hungarian, Greek, and Turk which threatened c 
who marched by land. 

(A) The Land Routes through Asia Minor. 

Since, however, the majority of the early Crusaders we 
unaware of the superiority of the sea route, and chose to mal 
Constantinople their basis for the march on Jerusalem, we mu 
begin by pointing out the strategical aspects of their unde 
taking. In 1097 almost the whole of Asia Minor was in tl 
hands of the Seljouks : the Emperor Alexius held little moj 
than Chalcedon, Nicomedia, the Mysian coast-region, and a fe 
isolated towns on the Black Sea, like Sinope and Trebizon- 
The Turks were established on the Sea of Marmora : they ha 
chosen Nicaea, only twenty-five miles from its shore, as the 
capital. All the inland plateau of Asia Minor was in the 
hands, and all the coast-line also, save the few Byzantine sec 
ports and a patch or two in Cilicia, where Armenian mountair 
chiefs maintained a precarious independence. 

If Alexius Comnenus had been able to direct the crusadin 
army at his own good pleasure, he would have used it to cleg 
Bithynia, Lydia, and Phrygia of the Seljouks. If the Frank 
on the other hand, had been entirely their own masters, the 
would have marched straight across Asia Minor to the Cilicia 
gates, and made Antioch their first halting-place. But sine 
neither party could disregard the wishes of the other, a kind c 
compromise was concluded : the Crusaders took Nicaea fc 
Alexius, and then went on their way. The reduction of th 



,97j THE ROUTES OF ASIA MINOR 237 

urkish capital was of inestimable advantage to the emperor : 
onstantinople could breathe freely when the Seljouks were dis- 
dged from the stronghold almost in sight of its walls which 
ley had been holding for the last fifteen years. With this 
lexius had to be content for the present Murmuring bitterly 
lat they had been restrained from plundering and occupying the 
ty, the Crusaders moved forward into Phrygia. The route 
:ross Asia Minor which they adopted was, except in some small 
stalls, the right one. Their successors in later years would have 
jen wise if they had always adhered to it. 

The great peninsula consists of a high central plateau sur- 
mnded by a number of small coast-plains. For those who wish 
) march from west to east there is no good road either along 
le Euxine shore or the shore of the Sea of Cyprus. On the 
3rth the mountains of Paphlagonia and Pontus, on the south 
lose of Lycia and Isauria, come down to the water's edge at 
lany points, and cut the practicable route in so many places,, 
lat it is for all intents and purposes impassable for an army, 
[o traveller in his senses would attempt to use the coast-roads, 
'he inland roads, one of which he must choose, are practically 
iree in number. Two of them suit those who start from Nicaea,. 
le third those whose base is Sardis, Miletus, or Ephesus. This 
ist was not available for the Crusaders of 1097; they had no 
'ish to make the long detour along the ^gean, through Mysia 
nd Lydia, which would have brought them to Sardis or any of 
le other suitable starting-points for the march to Philadelphia- 
'hilomelium-Iconium-Tarsus. There remained for their choice 
le two other routes, one of which passes north, one south, of the 
reat Salt Lake of Tatta (the Tuz Gol of the Turks) and the little- 
nown region of the Axylon ^ which lies around it. The southern 
DUte is that which they chose : it runs by Dorylaeum, Philo- 
lelium, Iconium, and Heraclea-Cybistra to the Cilician gates.^ 
The northern and the longer way leads to the same pass by 

^ Mr. Hogarth informs me that the Axylon does not deserve its well-known 
2putation for barrenness and desolation. 

2 Why Godfrey of Bouillon and the larger half of the crusading host diverged from 
le obvious route by Heraclea, the Cilician gates, and Tarsus, and only sent Baldwin 
nd Tancred upon it, it is hard to discover. But they undoubtedly took the extra- 
rdinary and circuitous road by Nigdeh, Csesarea-Mazaca, Coxon (Cucusus-Goeksun), 
nd Marash, and suffered severely from privations in the Anti-Taurus while crossing 
tie Doloman Dagh, between Coxon and Marash. Probably they were attracted by the 
riendly Armenian population of Eastern Cappadocia. 



238 THE ART OF WAR IN THE MIDDLE AGES [k 

Tatiaeum, Ancyra, Caesarea-Mazaca, and Tyana. Both w< 
good Roman roads, and had been kept in order by the Byzantii 
down to the disastrous year 1071. Now, however, the land '. 
desolate : bridges were broken, cisterns empty, and for ma 
stages the whole population had been slain or driven off by t 
Seljouks. There were no insuperable natural obstacles on eitl 
road : the two perils to the Crusaders were starvation and 1 
chance of being wearied out and brought to a stand from e 
haustion by the incessant attacks of the Turkish horse-arche 
More fortunate than any of their successors, the hosts of Godfi 
and Bohemund opened their march by inflicting a crushing def 
on the enemy, who was so utterly awed that he held off, and c 
not venture to harass the marching columns for many wee 
They moved by Philomelium, Antioch-in-Pisidia, and Iconiu 
with no let or hindrance. It was not till they reached Herack 
Cybistra that they again met the Turks in arms, and then they< 
feated them with ease. Though unmolested by the Seljouks, t 
Franks suffered dreadfully from want of stores and forage. TJ 
was unavoidable in a desolate land, for the Western armies of t\ 
age had no proper conception of commissariat arrangemen 
they depended mainly on the districts they passed through ; a 
if the countryside was barren, they were bound to suffer. T 
trouble was made far worse by the long and useless train 
non-combatants of both sexes which the crusading host dragg 
behind it. If they had endured many privations in Christi 
regions like Hungary and Bulgaria, it was obvious that t 
passage through Asia Minor was bound to be accompanied 
terrible loss of life. Nevertheless, the greater part of the he 
struggled through, some to Marash, others to Tarsus, where th 
could rest and recruit themselves for a space among the friend 
Armenian population of Cilicia. 

On the whole, therefore, the passage of the first Crusade 
through Asia Minor may be described as fairly successful wh 
their difficulties are taken into consideration. Far otherwise w 
it with their successors of iioi. The miscellaneous bands und 
Sweyn the Norseman, Archbishop Anselm of Milan, William 
Poictiers, Stephen of Blois, and Eudes of Burgundy, all fared f 
worse. Some were wholly destroyed, others were turned ba( 
with the loss of nine-tenths of their numbers ; of the remainder 
few stragglers only succeeded in pushing their way to Tarsus ar 
Antioch. The causes of their disasters are sufficiently obviou. 



.loi] THE CRUSADE OF iioi 239 

hey showed even less discipline than their predecessors, and 
hey had formed a wholly erroneous conception of the easiness of 
heir task from the comparative immunity enjoyed by Godfrey 
i;nd Bohemund's army during its passage. They were so puffed 
rp with the idea of their own invincibility that they declared 
heir intention of " crossing the mountains of Paphlagonia and 
brcing their way into Khorassan, in order to besiege and take 
3agdad." ^ It was in pursuit of this mad design that the majority 
)f their host started off on the route Ancyra-Gangra- Amasia, 
vhich, if they had been able to pursue it to the end, could only 
lave stranded them in the mountains of Armenia. After a 
errible march among the highlands of Pontus,^ where the foot- 
oldiery died by thousands of weariness and starvation, and the 
avalry were almost entirely dismounted, the Lombards and Pro- 
encals were brought to a standstill by the army of Mohammed 
bn Danishmend, Emir of Cappadocia, whose light troops hovered 
.round them day after day, cutting off their stragglers and for- 
-ging parties. When the Turks thought the Crusaders sufficiently 
xhausted to fall an easy prey, they offered them battle at a 
)lace named Maresh (or Marsivan), somewhere in the neighbour- 
lood of Amasia. The combat was indecisive, but on the follow- 
iig night Raymond of Toulouse, the man of greatest note in the 
lost, fled away by stealth and deserted his comrades. Others 
lasted to follow his example, and, in the disorderly retreat which 
hen set in, Danishmend cut the whole army to pieces, with the 
xception of a few thousands who succeeded in distancing tbear 
^ursuers and finding shelter in the Greek fortress of Sinope. 

Meanwhile, the smaller division of this band of Crusaders, 
/ho had refused to take the unwise route along the northern 
dge of the plateau of Asia Minor, had been reinforced by 
Villiam Count of Nevers and a large band of French pilgrims. 
They then marched fifteen thousand strong^ by the long but not 
(rational line of Ancyra-Iconium-Heraclea. All the way 
rom Iconium to Heraclea they were encompassed by the hordes 
•f Danishmend and Kilidj-Arslan, fresh from their victory over 
he Lombards at Maresh. Harassed incessantly, day and night, 

^ Albert of Aix, viii. p. 7. Cf. the identical statement in Anna Comnena, 
00k xi. § 8. 

'■^ We get from Anna only the fact that they had crossed the Halys ; the Frankish 
iironiclers thought they were still in " Flagania," i.e. Paphlagonia. 

^ Albert of Aix, viii. p. 29. 



240 THE ART OF WAR IN THE MIDDLE AGES [no 

by the enemy, and suffering horribly from thirst, they wer 
reduced to the most pitiable condition when they reached Heracle 
and had the passes of the Taurus in sight. Then the Turk.^ 
fearing that their prey was about to escape them, closed in an( 
offered battle. In a long straggling fight between the city am 
the foot of the Taurus the Christian army was gradually brokei 
up and shot down in detail. Seven hundred knights, who at las 
abandoned their unhappy foot-soldiery ^ and took to the hill.^ 
got off in safety over one of the minor passes of the Taurus, an( 
reached Germanicopolis in Cilicia, where they took shelter witl 
the Byzantine garrison. William of Nevers himself finally reache( 
the same spot with only six companions. The rest of the fifteei 
thousand Franks had been slain ; the Parthian tactics of tb 
Turks had not been frustrated by any such happy chance as tha 
which saved Bohemund and Robert of Normandy at Dorylaeum. 
A very similar fate befell a large body of Aquitaniai 
Crusaders, led by their duke, William of Poictiers, who had startec 
shortly after the departure of the Count of Nevers from Constanti 
nople. This host, a much larger one than either of those whicl 
preceded it, followed the same route as Godfrey and Bohemunc 
had taken four years before. They had little trouble from tht 
Turks till they reached Iconium, and were successful in taking 
and pillaging the towns of Philomelium and Salabria.^ But a 
Iconium their provisions gave out, and they learned of th( 
destruction of the army of the Count of Nevers. Nevertheless 
they resolved to press forward, and soon found themselves bese 
by Kilidj-Arslan and Danishmend. Their immunity from attacl 
hitherto had only been secured by the fact that the division o 
Nevers was eight days ahead of them, and had attracted all the at 
tention of the Seljouks. The fifty-five miles between Iconium anc 
Heraclea proved as fatal to the Aquitanians as it had been tc 
their predecessors. The want of water was their ruin,* and wher 
they approached the river near Heraclea they broke their ordei 
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 

^ Albert of Aix, viii. 30. 

2 See the account of this battle on pp. 271-274. 

^ 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 Nicaea-Iconium. 

^ Robert the Monk, book iii., tells us how Godfrey of Bouillon avoided this danger 
by taking water with him. 



iioi] 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 fighting, 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 i lOi, 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 
irmies had been in fair fighting trim, it is doubtful whether they 
A^ould 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 
:avalry in combination, and wished to do all the work with 
their mounted men alone. Hence they were bound to fail : only 
1 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 1 148-49, 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 

foiindly 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 Dorylaeum, Philadelphia, and 
Laodicea. They had also recovered the whole southern coast oi 
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 Nicaea. It was just possible to march frorr 
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 : - 
" From the Bosphorus [or the Arm of St. George, as it was ther 
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'.' 
residence, and five days after that it enters Cilicia, a Christiar 
land. A strong army, fortified by the faith and confident in iti 
numbers, might despise its obstacles ; but in winter the snowj 
which cover the mountains are very terrible." This is the ok 
route of the first Crusaders by Dorylaeum, Iconium, Heraclea, anc 
the Cilician gates. " Secondly, there is the road most to the 
right, which is better in some ways, as supplies are to be hac 
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 a: 
bad as the Turks and the snows on the first route." By this roac 
Odo means the long, circuitous passage by Pergamus to Ephesus 
and thence along the Carian, Lycian, Pamphylian, and Isauriai 
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 bu 
poorer and less safe than the second." The middle route o 
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 winte 
of 1 148-49. The Emperor Conrad and the Germans took the 
" left-hand road," i.e. the short and dangerous line through the 
^ Odo of Deuil, book v. 



1 148] DEFEAT OF THE EMPEROR CONRAD 243 

midst of the Turkish territory, which passes by the gates of 
Iconium. 

The fates of the two expeditions were not wholly 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 Nicaea, 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 iioi a stage or two 
farther to the east. The Germans, in spite of all the warnings 
of previous Crusades, had no provision of crossbowmen ^ 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.^ 
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 

^ This is especially remarked upon by Odo of Deuil, book v. p. 343. 
2 Odo of Deuil, book v, p. 347. 



244 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. Neat 
Antioch-on-Maeander 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 ir 
good order, and a vigorous charge of the French knights swept 
the Seljouks away ; they gave no trouble for some days, so thai 
the army arrived safely at Laodicea, the border town of tht 
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, bac 
been harried bare by the Seljouks. There was hardly ar 
inhabited village on the route, which turned out to be no les.' 
than fifteen days in length, though the French had calculatec 
on taking a much shorter time to traverse it. But the governo 
of Laodicea refused to sell any provisions to the Crusaders— 
from treachery, according to the French chroniclers, but mon 
probably because he dared not exhaust his stores when th( 
Turks were known to be in the immediate neighbourhood. 

It was accordingly with a very insufficient stock of food tha 
the French marched past Laodicea and started on their way b^ 
the pass between the Baba Dagh and the Khonas Dagh whicl 
leads up into the highlands. On the second day after leavinj 
Laodicea their disasters began. The army was marching with ; 
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 hundret 
feet above the sea level), the van, under Geoffrey de Rancogne an( 
Amadeus Count of Maurienne, the king's uncle, was ordered t< 
seize and hold the exits of the defile till the whole army ha( 
passed. But, preferring to spend the day comfortably in the plaii 



1 149] LOUIS VII. IN ASIA MINOR 245 

of 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,i j^g 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 

^ Odo of Deuil, book vi. p. 363. 



246 THE ART OF WAR IN THE MIDDLE AGES [n^ 

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 Oestrus, 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 som.e 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 iioi) 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 



iiQi] FREDERIC BARBAROSSA'S CRUSADE 247 

Deat the Turkish hosts in every encounter. The great emperor 
lOok in the first half of his march (March-April 1190) a route 
lot very unlike that which had been followed by Louis VII., 
keeping well inside the Byzantine border in Mysia and Lydia. 
He passed by Philadelphia and Tripolis into the valley of the 
IVIaeander, and reached Laodicea. But from this point he did 
not turn south like the French king, but set his face due east, 
md moved by the great Roman road which passed by Apamea 
ind the Pisidian Antioch to Iconium. This was the main artery 
3f the communications of the central plateau, and it is curious to 
find that no other crusading army had tried it. The Turks 
:losed round Frederic and attacked him at the sources of the 
Maeander, 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 
^May 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." ^ 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 

^ See the Epistola de Morte Fy-ederici, p. 346. 



248 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 theii 
coats of mail effectively kept out the Turkish arrows. But theii 
horses, not yet armed in steel like those of later times, suffered 
terribly. By the 13th 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^ 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," ^ 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. 

^ Not Malek Shah. See Boha-ed-din, p. 272. ^ £p^ ^g j^gj^te Frederici, 350. 



iQi] THE CAUSES OF DISASTER 249 

Our first deduction must be couched in the form of a testi- 
nonial to the very efficacious nature of the Seljouk methods of 
varfare. The Turks had deliberately established a broad belt 
)f wasted and uninhabited territory between themselves and the 
Byzantine border. Moreover, when a Christian army passed 
hrough their dominions, they did not hesitate to destroy their 
)wn crops and sacrifice their villages. The cattle were driven 
nto 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 
vas the most obvious. 

A thoroughly disciplined regular army, with an organised 
vaggon-train, could no doubt have triumphed over this system 
jy bearing its own food with it. But the Franks were a mixed 
nultitude, with little or no organisation, always clogged in their 
Drogress by the hordes of non-combatants, largely paupers, whom 
:hey dragged with them. Against such foes the Turkish system 
A^as most efficacious. We may, indeed, express our wonder that 
Grodfrey and Frederic Barbarossa struggled through in spite of 
ill opposition. That the Crusaders of iioi and 1148 failed is 
less a matter of surprise. 

The second among the main causes of the disasters of the 
:rusading 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 
Ciiician 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 Vll. 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 [109 

must place their own want of providence, discipline, and sel: 
control. Even the best-behaved of their armies were, by the cor 
fession of their own chroniclers, terribly addicted to riot an 
plunder. Their interminable quarrels with the Greeks mostl 
arose from their own fault. That there existed a very considerabl 
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 wonde 
at the resentment they provoked. Their want of discipline was a 
well marked as their proneness to plunder : deliberate disobedienc 
on the part of officers was as common as carelessness an 
recklessness on the part of the rank and file. This was alway 
the case in feudal armies : in the East the fault was seen eve 
more clearly than elsewhere. Most notable of all is the eviden 
inability of the Franks to learn from the unhappy experience 
of their predecessors. The thousands of veterans who driftet 
back from the East did not succeed in teaching their successor 
to observe the precautions appropriate to Turkish warfare. Fift; 
years after the first Crusade, Conrad III. and Louis VII. com 
mitted exactly the same mistakes as the contemporaries c 
Godfrey and Bohemund. They marched without caution ; the; 
did not properly combine infantry and cavalry ; they had no 
provided themselves with the necessary proportion of men arme< 
with missile weapons such as the bow and arbalest ; their stocl 
of food was always running short. It seemed that the art c 
learning by experience hardly existed in the military circles c 
the West. The description of the faults of the Frank as a soldie 
which Maurice wrote in 580, and Leo the Wise repeated in 90c 
might still be utilised almost word for word in describing th' 
Crusaders of 11 50. 

(B) Tpie Strategy of the Conquest of Syria. 

The primary impulse of the men of the first Crusade wa 
religious, not strategic. Their end was to recover Jerusalem, no 
to establish a sound military base for the ultimate conquest o 
the whole of Syria. There were those among the Frankisl 
leaders who saw that it was dangerous to march from Antioch t* 
Jerusalem, leaving hostile towns to right and left, and sacrificing 
the connection with their only base ; but they were overruled b} 
the majority, whose ruling desire was to get possession of th( 
Holy Places. We must not, therefore, criticise the campaign o 
1099 as if it had been carried out on logical military lines. 



)99] THE GEOGRAPHY OF SYRIA 251 

It was only when Jerusalem had fallen, and the Crusaders had 
^termined to establish a permanent feudal state in Palestine, 
lat strategical considerations came to the front 

When Godfrey was crowned, the new kingdom consisted of 
Dthing more than the towns of Jerusalem and Jaffa. Whether 
ohemund, isolated at Antioch, and Baldwin in his distant county 
:' Edessa, would ever truly become the vassals of their theo- 
itical suzerain was most uncertain. The future of the Franks 
I Syria was not settled for many years : indeed it was not till 
Dout 1 125 that any general conclusions as to the new states 
Duld be formulated. 

Before passing on to consider the military history of the 
Dnquest, it is necessary to understand the general strategical 
5pect of Syria. It may be divided into four narrow zones 
inning from south to north, one behind the other. The first of 
lese — the shore — consists of a series of coast-plains of very 
arying size and width ; they are cut off from each other by 
mountains running down to the water's edge, like Carmel, the 
purs of Lebanon, and the " Black Mountains " by Antioch. 
lost of these level coast-tracts are narrow, but the southmost of 
lem, the celebrated plain of Sharon, is larger than the rest, and 
verages fifteen miles in breadth. Occasionally, too, the coast- 
lain runs inland up a river valley, as in the plain of Esdraelon 
ast north of Carmel, and in the valley of the Orontes near 
intioch. In the central districts of the Syrian shore, however, 
bout Tripoli and Beyrout, it is exceptionally narrow and much 
•roken up. 

The second zone of territory comprises the mountainous 
ipland overhanging the coast-plain. This region consists of the 
purs of three main chains — the mountains of the Ansariyeh (the 
]asius of the ancients) in the north, Lebanon in the centre, and 
he mountains of Itphraim and Judaea in the south. The two 
orm.er are lofty ranges rising at some points to eleven thousand 
eet above the sea level ; the last has a broader and less well- 
lefined crest, and seldom rises to a greater height than three 
:housand feet. The spurs and shoulders of all these chains con- 
:ain many fertile and populous tracts. 

The third zone consists of the deep-sunk valleys of three 
jreat rivers — the Orontes, Leontes (Litany), 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 [SI 

Dagh), the second by a much narrower defile just north of Tyj 
But the Jordan, whose course is mostly below the level of tl 
Mediterranean, falls into the Dead Sea, a sheet of water wath i 
exit. The Orontes and Leontes have broad and fertile vallej 
while that of the Jordan is a narrow, precipitous, and marsl 
defile, only to be crossed at a limited number of points. Tl 
deep depression through Central Syria formed by these thr 
streams and by the Dead Sea is continued yet farther south I 
the gorge of the Wady-el-Arabah, which runs down to the easte 
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 othe 
mountainous. It runs into the Great Desert, and is its€ 
barren in many parts. But it contains many fruitful and we' 
watered districts, such as those around the great cities 
Aleppo and Damascus. 

Syria as a whole is eminently defensible : the sea and dese 
cover it on three sides — the west, east, and south ; on the nor 
the Amanus and the Euphrates give an excellent and we] 
marked frontier. But the Crusaders never got possession of tl 
whole country : they only held the coast, the greater part of tl 
mountain, and certain regions of the central valley. The largi 
half of the latter and the whole of the eastern plateau remainc 
unconquered. It was for this reason that the kingdom < 
Jerusalem was always in a precarious position. A chain • 
Mohammedan states always shut it out from expanding to tl^ 
eastward and reaching its natural boundary. 

The cause of this anomaly is not hard to find. The crusac 
ing states were never really strong enough to complete the coi 
quest of Syria : they would not even have succeeded in subduir 
the whole of the coast if they had been forced to rely on the 
own resources and could have counted on no external aid. Bi 
the great Italian republics were deeply interested in the conque 
of the Syrian shore. It was of high importance to the 
commerce that the whole of the ports of the Levant should 1: 
in Christian hands. Hence they co-operated with the greate 
zeal in the sieges of the coast-cities : they and not the kings ( 
Jerusalem were really the conquerors of the whole coast-plaii 
The Venetians were the real captors of Sidon (iiio)^ and Tyi 
(1124). The Pisans gave assistance to the Prince of Antioch < 

^ Largely aided by King Sigurd of Norway on this occasion. 



125] WEAKNESS OF THE CRUSADING STATES 253 

.aodicea (1103) and to Count Bertram at Tripoli (1109); they 
v^ere also present at the siege of Beyrout (mo). The Genoese 
/ere still more energetic : to them were due the falls of Caesarea 
iioi), Tortosa (1102), Acre (1104), Giblet (1109), Beyrout 
1 1 10). Casual aid was often given to the kings of Jerusalem by 
ither crusading fleets, such as those of the Englishmen Harding 
.nd Godric, and the Norse king, Sigurd the Jerusalem-farer 
1 109-10). But it was mainly by the aid of the Italians that 
he Syrian coast became Christian. 

Inland, the aid of these all-powerful allies was not available, 
[heir interests did not bid them equip armies to conquer 
Damascus or Aleppo. Hence it was with their own weak feudal 
evies alone, aided by occasional hosts of Western pilgrims, that 
he kings of Jerusalem and princes of Antioch carried on their 
vars with the emirs of the inland. The military resources of 
he Prankish states were more than modest : the largest army 
hat they ever put into the field was one of thirteen hundred 
cnights and fifteen thousand foot,^ a number only obtained by 
:ollecting every available man and leaving the towns and castles 
dmost ungarrisoned. Larger numbers were of course assembled 
when a crusading host from the West was present ; but the help 
)f the pilgrims was transient : they always returned home after 
I short sojourn in the Holy Land. As a rule, the domestic 
brces of the Syrian Franks seldom took the field more than six 
Dr seven thousand strong. Often, when the fate of the kingdom 
vvas 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 i loi 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 11 18, 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.^ 

^ To withstand Saladin's invasion of 1 183. 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 

Want of numbers, then, was the real cause of the failure 
the Franks to conquer inner Syria. That they ever succeed* 
in establishing themselves firmly on the coast, and in holdii 
many districts of the mountain zone, must be attributed to t 
divisions of the Moslems. As long as the interior lands we 
divided between three or four independent emirs, the Crusade 
not only held their own, but actually advanced their frontie 
Down to the rise of Zengi, the first prince who began to uni 
the emirates, the Franks were slowly but surely occupying t' 
cities of the Infidel.^ Nothing, indeed, could have been mo 
opportune than the fact that, in the early years of the twelf 
century, Damascus, Aleppo, Kayfa, Mosul, Mardin, were in t] 
hands of different families, all bitterly jealous of each other, ai 
sometimes even ready to ally themselves with the Christian 
thereby they might do their neighbours an ill turn.^ This fact 
was which enabled a few hundred Frankish knights to ric 
roughshod over Syria for some twenty years, till in 1 1 27 Zen 
took up the governorship of Mosul. The interesting picture 
the state of the land in this year given by the Moslem chronicl 
who wrote the history of the Atabegs ^ is well worth quoting. 

"At the moment when Zengi appeared, the power of tl 
Franks extended from Mardin and Scheikstan in Mesopotam 
as far as El-Arish on the frontier of Egypt, and of all tl 
provinces of Syria only Aleppo, Emesa, Hamah, and Damasc 
were still unconquered. Their bands raided as far as Amida 

^ The dates of the changes of dynasty in the emirates are all-important i 
\mderstanding the history of the Crusades. They are as follows : — 

Aleppo. Held by the house of Tutush-ibn- Alp- Arslan, 1094-1117. 

Held by Il-Ghazi of Mardin and his nephew Soliman, 1117-112 
Held by Balak-ibn-Bahram, 1123-II25. 
By Il-Borsoki and his son Massoud, 1125-1128. 
Surrendered to Zengi, 11 28. 
Damascus. Held by Dukak the Seljouk, 1095-1103. 

Held by Toktagin and his house, 1103-1154. 
Surrendered to Nur-ed-din, son of Zengi, 1 154. 
Mosul. Held by Kerboga, 1096-1102; by Jekermish, 1102-1107 ; 
Javaly, 1107-1108 ; by Maudud and his nephew Massoud, no; 
1113; by Il-Borsoki, 1113-1127. Taken over by Zengi, 112; 
- The strange battle of Tel-basher in 1108 is worth notice. Tancred of Antioi 
and Joscelin, Lord of Tel-basher, had quarrelled. So had Ridwan of Aleppo ai 
Javaly of Mosul. Each allied himself with a stranger against his own co-religioni^ 
and in the fight Frank fought with Frank, and Turk with Turk. Tancred and Ridwi 
were victorious. Albert of Aix and William of Tyre both allude to the story. 

^ The Turkish deputies or generals of the great Seljouk Sultan, who ruled 
practically independent princes in Syria and Mesopotamia. 



ri27] WIDEST EXTENSION OF THE FRANKS 255 

;he province of Diarbekir, and in that of El-Jezireh [Upper 
Mesopotamia] as far as Nisibis and Ras-Ain. The Mussul- 
mans of Rakkah and Haran [Carrhae] were exposed to their 
oppression and the victims of their barbarous violence. All 
:he 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 
ihe 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." ^ 
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, Frankish 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 Faneas (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 1 1 16 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 (11 17). This stronghold communicated with Palestine by 
means of two other castles, Montreal (Schobek) near Petra in 

^ Quoted in Michaud's Bihliotheque des Croisades, vol. iv. p. 61. 



256 THE ART OF WAR IN THE MIDDLE AGES [Ml 

the centre of the Edomite desert, and Kerak in the land ( 
Moab. The fief of Montreal-Kerak or of " the land beyon 
Jordan " was one of the four great baronies of the Latin kingdon 
It formed such a dangerous outpost, and its position was s 
forbidding, lost as it was in the desert, that we are surprise 
to find that the Franks held it from 11 16 till 11 87, the year ( 
the fall of Jerusalem.^ As long as it survived, it made th 
-communication between Damascus and Egypt very precarious 
Moslem caravans had to pay blackmail to its lords, or suffe 
untold danger of starvation and misdirection in passing b 
stealth between the three fortresses in the wilderness. Militar 
communication between the Fatimites and the rulers of Damasci 
was equally hard ; armies marching through the sands and rocls 
of Idumea were always exposed to sudden attacks from thes 
garrisons. They were such thorns in the side of Islam tha 
repeated attempts were made to capture them, all of whic 
failed — even when Saladin himself took the matter in banc 
They only fell with the fall of the Latin kingdom, and Kera 
actually held out longer than Jerusalem. 

North of Kerak the frontier of the Franks was guarded by 
-chain of castles watching the defiles which lead down to th 
fords of the Jordan. The line was composed of Paneas, Beai 
fort, Chateau-Neuf, Safed, Castellet, and Beauvoir. South of th 
last-named, where the valley of the Jordan is most deep an 
rugged, there seems to have been a gap left, the natural defence 
being apparently too formidable to require strengthening. 

Stretching along the coast from Beyrout northward lay th 
county of Tripoli, the weakest of the four crusading states. It 
rulers never succeeded in pushing inland through the passe 
of Lebanon or getting a lodgment in Coele-Syria. They onl; 
possessed the series of narrow coast -plains round the stron: 
<:ities of Markab, Tortosa, Tripoli, and Giblet, together with th 
•spurs of the mountains above and between them. The grea 
-chain of Lebanon, however, gave a strong frontier for defence 
In commanding positions, watching the few practicable passe 
through the range, were the inland castles of Montferrand, Krat 
and Akkar. Weak for offence, but strong for resistance, th 
county of Tripoli preserved its mountain boundary far into th 
thirteenth century. 

^ Kerak fell in ii88 only, but Elath had been recovered by the Moslems in 117c 
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 always 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 Frankish 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 disaj^^pearance 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 

17 



i 



258 THE ART OF WAR IN THE MIDDLE AGES [112: 

weak, and so exposed that there was no hope of permanentl} 
retaining it. Edessa would have fallen long before if it had no 
been repeatedly saved by the intervention of its neighbours t( 
the south. Tancred and King Baldwin T. led armies fron 
Antioch and Jerusalem to save it : without their aid it mus 
have succumbed in mo, or perhaps even in 1104. It wouk 
undoubtedly have been better for the general defence o 
Syria if the first conquerors of the land had seated themselve 
at Turbessel rather than at Edessa, and contented themselve 
with holding only the districts west of Euphrates : the^ 
might then have made the great river their boundary, an( 
served as efficient guardians of the marches of North-Easteri 
Syria. 

The extension of the Prankish dominion ceased immediately 
on the appearance of Zengi. The only important conques 
made after the year 11 27 was that of Ascalon, taken from th' 
Fatimite Sultan of Egypt by Baldwin III. in 1153. Before th< 
end of the long reign of the great Atabeg, the balance had begui 
to turn definitely in favour of the Moslems. The great mark c 
the change was the destruction of the northernmost crusadinj 
state, the county of Edessa, by Zengi's hand, in 1144. Th 
union of Mesopotamia and Northern Syria under Zengi's rul 
completely checked the expansion of the Prankish dominioi 
inland. There remained the three surviving Christian states— 
the kingdom of Jerusalem, the principality of Antioch, th 
county of Tripoli, forming a long straggling strip of territor 
along the coast, much cut up by mountains, and nowhere muc 
more than fifty miles broad. They had no good land communi 
cations with each other, and depended for their union solely o 
the maritime predominance of the Italian republics. 

One chance only of triumph remained to the Franks — th 
possibility of the arrival of a new crusading host from the Wes 
sufficient to enable them once more to take the offensive. I 
was obvious that the strength of the Latin states of Syri 
unassisted would not even suffice to preserve themselves. Fc 
one moment in 11 49 it appeared as if this chance might com 
into realisation. Deeply stirred by the news of the fall c 
Edessa, the nations of the West sent out the great hosts c 
Conrad III. and Louis Vll. on the second Crusade. Only th 
broken wrecks of these expeditions ever reached Palestine, bu 
even these were numerous enough to encourage the King c 



1 149] 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 sw^ore 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 
:o be part of Islam.^ It is certain that the King of Jerusalem 
tressed the leaguer slackly, and at last departed homeward, to 
:he great disgust of the emperor and the other pilgrim princes 
rom the West. Thus ended the one serious attempt of the 
Franks to establish themselves in inner Syria and carry their 
Vontier up to the desert. 

The fact that Zengi's dominions were divided up among his 
ons (Nur-ed-din taking Syria and Seyf-ed-din Mesopotamia), — 
o that for a time the unity of command was lost, and the Franks 
)btained a respite, — did not lead to any permanent change in the 
ortunes of the crusading states. The King of Jerusalem turned 

^ See Ibn-Alathir on p. 96, vol. iv. of the Bibl. des Croisades. Cf. also William 
f Tyre, book xvii. pp. 14, 15, who says that the Count of Flanders was to be made 
'rince of Damascus by the Westerns, which the Syrian Franks would not endure. 



26o 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 nc 
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 thar 
once King Amaury seemed on the point of succeeding. B) 
adroitly taking part in the war between the Egyptian viziei 
Shawir and Shirkuh, the general of the Syrian prince Nur-ed 
din, he obtained a free entry into Egypt, and occupied man\ 
towns as the ally of Shawir. For a short time a Frankist 
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 ii 
Palestine, exposed in his absence to the raids of Nur-ed-din': 
governors in Damascus and Coele-Syria. It was bad strateg) 
to strike at the Nile while Jerusalem and Antioch still had ai 
enemy encamped only a few score miles from their gates. I 
was the consciousness of the danger of his own realm tha 
always kept Arriaury anxious and preoccupied during hi 
Egyptian campaign. He had always, so to speak, to " keep on' 
eye behind him " : a demonstration on Jerusalem by Nur-ed 
din might bring him back from Cairo at any moment. This i 
the true reason why he lost the fruits of successful campaigns, b; 
allowing himself to be bought off by great sums of money 
Hence it came that he levied great fines from Egypt, and fo 
several years received a regular tribute from Shawir, but neve 
made a firm lodgment in the land. At last, the most unhapp; 
contingency for the Franks came to pass. Shirkuh murderc' 
Shawir, and seized Egypt for his master Nur-ed-din (1169 
Syria and Egypt were at last united in the hands of a singl 
prince, for the Fatimite caliph did not long survive his vizie 
meeting, like him, a bloody end at the hands of Nur-ed-din 
lieutenants (1171).^ Amaury made one last invasion of Egyf 
after the fall of his ally Shawir, leaguing himself with th 
Byzantine emperor, Manuel Comnenus. But the Greek flet 
and the Frankish army lay long before Damietta, and failed t 

^ So at least say the Frankish historians. Saladin's biographers either pass ovt 
the event without details, or say that El-Adid died a natural death. See tf 
Mohammedan authorities quoted in the Bibliothcqvc des Croisades, iv. 147. 



ii87] THE RISE OF SALADIN 261 

take it. Presently came the news that Nur-ed-din was 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 Frankish 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 
for 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 Frankish 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 Frankish 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 1 571, and then only did the maritime domination of 
the Franks in Eastern waters come to an end. 

(C) The Attacks on Egypt. 

Before dismissing the subject of the grand strategy of 
the Crusades, we have still to deal with two^ 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 1163-69. 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 Ayubite 

^ The expedition of St. Louis to Tunis has no bearing on the general history of 
the Crusades, and was inspired by a rehgious, not a military object — it being supposed 
that the riiler of Tunis might be converted to Christianity ! 



264 THE ART OF WAR IN THE MIDDLE AGES 



n 



house in 12 19: 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 
1 2 19 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 oi 
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 Brienne and St. Louis neglected 
this obvious and easy line, and chose instead to land at Damietta. 



22o] FAILURE OF JOHN DE BRIENNE 265 

The road from that place to Cairo leads through the very midst 
f the Delta, over countless canals and four considerable branches 
f the Nile. Across it lie a dozen strong positions for the defend- 
ig army. It is not too much to say that the invasion of Egypt 
y this line is bound to fail, if the masters of the country show 
rdinary vigour and intelligence. The fates of the two Prankish 
xpeditions are a sufficient commentary on the wisdom of their 
jaders. John de Brienne only took Damietta after a siege of 
ight months ; his troops were already much exhausted when he 
dvanced into the Delta ; they were brought to a stand by the 
ne of the Ashmoun Canal, behind which lay the army of the 
>ultan El-Kamil. They made several unsuccessful attempts to 
Teak through, and were already despairing of success when 
hey learned that the land between them and their base at 
)amietta had been inundated ; the Nile was rising, and the 
Egyptians had cut the dikes. They hastily retreated towards 
)amietta; but the waters were out everywhere, the Sultan 
Dllowed hard behind them, and, to save themselves from starva- 
ion or drowning, the Crusaders had to come to terms. El- 
Camil granted them a free departure, on condition that they 
hould evacuate Damietta (August 1221). 

Far worse was the fate of St. Louis when he tried the same 
oute in 1249. Considering how John de Brienne had fared, we 
an only marvel that he ventured to choose the same road. He 
tarted with somewhat better fortune than his predecessor, for 
)amietta fell into his hands after a very slight engagement with 
he Moslems. But he then wasted no less than six months in 
■aiting for stores and reinforcements ; all this time was employed 
y the Sultan in increasing his army and in preparing obstacles 
3r the march of the French. When, in November 1249^ 
Cing Louis did at last begin his advance, he was promptly 
hecked by the same bar which had ruined John de Brienne, 
he impassable Ashmoun Canal, defended by the Egyptian 
rmy. Time after time the bridges and causeways which he 
trove to construct were swept away by the military machines 
f the enemy. At last Louis got across by night with his cavalry 
t a deep ford practicable only for horsemen ; the infantry could 
;0t follow. The Egyptians were for a moment surprised, but 
he king's brother, Robert of Artois, threw away all chance of a 
ictory by charging rashly into the streets of Mansourah with 
he van long ere the king and the main body had come upon 



266 THE ART OF WAR IN THE MIDDLE AGES [i2f 

the field. He and the whole of his division were cut to piece 
and when Louis arrived he only succeeded in forcing his way 1 
the neighbourhood of Mansourah at the cost of half his knight 
At last, however, he worked his way to the bank opposite h 
own camp, and his infantry^ were able to finish the causeway ; 
which they had long been labouring, and so to join him. TI 
French thus obtained a lodgment beyond the Ashmoun, bi 
the success had cost them so dear that they could advance r 
farther. They lingered near Mansourah for some months, unab 
to move forward and unwilling to turn back, till at last famir 
and pestilence broke out, and compelled them to abandon tl: 
invasion. But the Egyptians had broken the road between thai 
and Damietta, and as they straggled northward they were c\ 
to pieces in detail in a long running fight extending over sever; 
days. At last the king was surrounded and taken prisoner, ar 
soon after the few surviving wrecks of the army laid down the 
arms. They could not even make terms for themselves, as Jot 
de Brienne had done in 1221, and the greater part of the captiv< 
were put to death in cold blood by the Egyptians.^ 

As a comment on King Louis' strategy we need only poii 
out that, even if he had successfully forced the passage of tl 
Ashmoun when he first reached it, he would yet have had 1 
pass three broad branches of the Nile and numerous canals, all su 
ceptible of easy defence, before reaching Cairo ! Nothing but tl 
entire want of geographical knowledge in those mapless da} 
can explain the madness of the Crusaders in twice selectir 
the utterly impracticable route Damietta-Mansourah-Benhc 
Cairo, when it was open to them to use the easy and obvioi 
road by Salahieh and Belbeis. Apparently they were attracte 
by the port and fortress of Damietta, which seemed to offer a 
excellent base and storehouse, while there was no town at a 
in the tract east of the ancient Pelusium, the proper startim 
place for the descent. There was nothing else to account ft 
the preference : one landing-place was as open as another to a 
armament in full command of the sea, and the coast east < 
Pelusium, though shallow inshore, does not present any re 
obstacle to the approach of vessels of such light draught as wei 
those of the thirteenth century. A careful examination of tl 
Government Survey maps of the Delta seems to show that east ( 
Pelusium and its marshes there is a sandy shore, with sufificiei 

^ For a more detailed account of Mansourah, see pp. 340-347. 



5o] FAILURE OF ST. LOUIS IN EGYPT 267 

; ;pth of water for light vessels to get close in. The region is 
s remote from the military centres of Egypt that no local 
i sistance need have been feared. 

,' We may fairly say, therefore, that the two great invasions of 

i ^ypt in the thirteenth century failed mainly because they were 

> idertaken with insufficient geographical knowledge, and con- 

j icted along an impossible route. That they would have had 

\ fair chance of success if they had been more wisely directed, 

j best shown by the fact that the Moslem historians one and all 

i sure us that their compatriots had completely lost heart after 

I e first successes of the Christians. In 1220 El-Kamil actually 

j fered to surrender Jerusalem, Tiberias, Giblet, Ascalon, 

( azareth, and Laodicea, if the Crusaders would but restore 

i amietta and return home. In 1249 Damietta was evacuated 

most without the striking of a blow, and the army which 

ustered behind the Ashmoun was in great disorder and deep 

;pression. If forced to fight not covered by a broad water- 

•urse, but in the open country about Salahieh or Belbeis, it 

>uld certainly not have held its ground. 

It was the same utter want of geographical knowledge which 
id ruined the Provencal Crusaders of iioi, and the French 
)st of Louis VIL in 1248, that brought to such disastrous ends 
e two formidable expeditions which endeavoured to subdue 

gypt. 



CHAPTER III 

THE TACTICS OF THE CRUSADERS 

Section I. — The Early Battles and their Tactics : Dorylcsum 
Antioch, Ascalon, Rafiileh. 

THE Western countries which contributed the largest prop 
tion of warriors to the first Crusade were precisely th( 
in which cavalry were at the time most predominant — Frar 
and Aquitaine, Lotharingia, Western Germany, and Italy b( 
Lombard and Norman. In each of the contingents wh; 
marched out in 1096 to join the great host which mustered 
Constantinople, the horsemen were considered the main comb 
ant force. If foot-soldiery followed by tens of thousands, it v 
not because their lords considered them an important part 
the line of battle, but because the same religious enthusiasm h 
descended upon the poor as upon the rich, and all were equa 
bent on seeking the path to the Holy Sepulchre. It was evide 
too, that infantry would be required for sieges, the service of 1 
camp, and the more onerous and less attractive labours of w 
So little, however, were they esteemed, that in the first gene 
engagement in which the grand army of the first Crusade engag 
— the battle of Dorylaeum — the foot-soldiery were left behind 
the tents, and the horsemen alone drew up in the line of batt 
Nor did the infantry even prove competent to keep the cai 
safe — they did not prevent the flanking parties of the Tui 
from entering it and massacring hundreds of the non-combatai 
committed to their care. 

The Crusaders then were accustomed only to one developmt 
of tactics — the shock-tactics of heavily-armed cavalry. Tfc 
regarded infantry as fit — at best — to open a battle with a d 

^ See p. 271. 



D97] THE CRUSADERS AND THE TURKS 269 

large of missiles, before the serious fighting began, or to serve 
^ a camp-guard. 

Ranged to oppose them, however, they found enemies of whom 
le most formidable were the Turks, a race long accustomed to 
efeat by their Parthian tactics the most powerful and the best 
isciplined heavy cavalry of the day — that of the East- Roman 
;mpire. The other Moslem powers who still employed the 
Ider methods of Saracen war, such as the Egyptian Fatimites, 
ere far less dangerous to the Crusaders. They — like their 
rcdecessors described by Leo and Nicephorus Phocas — still 
epended on the impact of their mailed horsemen, who were 
idividually inferior to the Byzantine trooper, and still more so to 
le Prankish knight. But the Turkish horse-archers were the 
)e who were destined to prove the main danger to the Crusaders, 
s they had long been to the emperors of Constantinople. It was 
tiey who were to teach not only the first invaders of the East, 
ut every army that followed them, many a bitter lesson. 

We have already recapitulated in an earlier chapter the 
anons 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 
»ody of infantry into the field ; ^ (2) to maintain an elaborate 
creen of vedettes and pickets round the army, so as to guard 
.gainst surprises; 2 (3) to avoid fighting in broken ground 
^here the enemy's dispositions could not be descried ; ^ (4) to 
:eep large reserves and flank-guards ;^ (5) to fight with the rear 
and if possible the wings also) covered by natural obstacles, 
uch as rivers, marshes, or cliffs, so as to foil the usual Turkish 
levice of circular attacks on the wings or the camp-guard ; ^ 
6) always to fortify the camp ; (7) never to pursue rashly 
md allow the infantry and cavalry to get separated after a 
irst success.^ With the necessity of all these precautions well 
anderstood, the Byzantines had yet suffered many disasters 
it the hands of the Turks. How was it to be expected that 
:he Crusaders would fare, to whom some of these precau- 
:ions would have seemed impossible, some ignominious, all 
infamiliar ? As a matter of fact, they knew nothing of them, 
>ince they utterly despised the Greeks and their methods of 
.varfare, disdained to learn anything from them, and took 

1 Leo, Tactica, xviii. 63. 2 j^^^ 58. 3 j^^^ 5^^ 

^ Ibid. Ti, ^Ibid.'Jl. ^Ibid.'j/^. 



2 70 THE ART OF WAR IN THE MIDDLE AGES 



I 



nothing but guides and money from the emperor.^ In 
first campaign they were as successful in violating every T 
of these rules as if they had committed them to memory 
the express purpose of not carrying them out. 

The hordes under Peter the Hermit and Walter the Pennih 
which first crossed the Bosphorus can hardly be called an arn 
Even in the eyes of their own countrymen they scarcely count 
as a military force, since they comprised but a very few mount 
men. When they were destroyed by the Seljouks near Nics 
they are said to have numbered only five hundred horse 
twenty-five thousand foot : ^ they had lost many thousands on t 
way by the hands of the Greeks and Bulgarians, but it is certc 
that in these earlier disasters the infantry had suffered infinite 
more than the cavalry, so that the original force must ha 
shown a still larger preponderance of men on foot. 

The great army which started from Constantinople in M 
1097 was a very different host. According to Western ideas, 
was a most formidable instrument of war. Many rich cour 
and dukes and their well-equipped retainers served in its rani 
Its numbers are given as high as a hundred thousand horse ai 
six hundred thousand foot — figures impossible in themselves, b 
showing a proportion between the two arms which was infinite 
more suited to the practice of the day than that which had pr 
vailed in the unfortunate horde of Walter the Penniless. 

Yet this great host came very near to suffering a comple 
disaster in its first serious conflict with the Turks. After layii 
siege to Nicaea and repelling with success the attempts of Sulti 
Kilidj-Arslan-ibn-Soliman to relieve it, they forced the place 
surrender. On June 27 they started forth to march into tl 
interior of Asia Minor, following the great Roman road whic 
leads by Dorylaeum, Philomelium, and Iconium to Tarsus. Tl 
countryside was wholly desolate : " Romania, a land once ric 
and excellent in all the fruits of the earth, had been so cruel] 
ravaged by the Turks, that there were only small patches < 
cultivation to be seen at long intervals." ^ Food for man an 
horse was difficult to procure, and it was perhaps to cover 

^ Save, indeed, Raymond of Toulouse, who borrowed some " Turcopoles," /. 
cavalry taught to act as horse-archers after the Turkish fashion, for his secor 
expedition. But he got no use out of them, except to escort his flight (Albert 
Aix, viii. p. 19). 

2 William of Tyre, book i. 

^ Fulcher of Chartres, chap. v. 



097] BATTLE OF DORYL.^UM 271 

reater space for foraging, and not out of mere carelessness,^ that 
he army split into two columns, marching parallel to each other 
t a distance of some seven miles. The right-hand corps was 
omposed of the followers of Godfrey of Bouillon, Raymond of 
^oulouse, Hugh Count of Vermandois, and most of the French 
nd Lotharingian contingents. The left column included 
)ohemund and Tancred with the Sicilian Normans, Robert of 
landers, Robert of Normandy, and Stephen of Blois. They 
eem to have been fairly equal in size and composition. 

Battle of Dorylcuitm, July i, 1097. 

After debouching from the Bithynian mountains, the Crusaders 
ound themselves in a broad upland plain, watered by the 
rhymbres, a tributary of the Sangarius. It was a rolling 
ountry, destitute of strong positions, and very well suited to the 
jeculiar tactics of the Turks. Flj^ing parties of their light horse 
oon began to hover around the advancing columns, but the 
rusading leaders did not take the obvious precaution of draw- 
'(\g together, or at least arranging to keep in close touch. On 
ighting the enemy they merely contracted their straggling line 
>f march and kept vedettes out to prevent a surprise. On June 
;o they camped some miles on the north side of the Thymbres, 
.nd not very far from the ancient and ruined town of Dorylaeum. 
)n the 1st of July the left division, with which we are most con- 
erned, moved forward to resume its route, and had marched for 
ibout an hour when its scouts reported the approach of the 
Turks in huge numbers. Bohemund, to whom the other chiefs 
lad committed the general charge of the host, ordered the tents 
o be pitched and the baggage unladed by the side of a reedy 
narsh ^ which gave a certain amount of cover, and deployed his 
nen in front of it. The infantry were left to guard the impedi- 
nenta,^ the cavalry alone drew up in line of battle. The camp 
vas not fully pitched, nor the squadrons completely ranged in 
)rder, when swarms of Turks suddenly appeared from all 
iirections, pressing in on the flanks and rear of the army as well 

^ Fulcher (chap, v) says that the parting was accidental, owing to the divergence 
)f one column at a cross-road, and the failure to get into touch again. Albert of Aix 
.ays that it was deliberate, and ordered for the reason stated above. William of Tyre 
'ays that it was uncertain whether it was accidental or not. 

2 "Juxta quoddam arudinetum " (Fulcher, v.). 

^ Gesta Francorum, 6 : " Pedites prudenter et citius extendunt tentoria, milites eunt 
/iriliter obviam iis [Turcis]." 



272 THE ART OF WAR IN THE MIDDLE AGES [ic 

as upon its front. The Sultan had gathered all his availal 
forces, and, though too late to relieve Nicaea, trusted to aven 
himself on its conquerors by a battle in the open field. The mc 
distant Seljouk hordes of Asia Minor had now had time to jc 
him, and his host was enormous — the Crusaders estimate it 
from a hundred and fifty thousand to three hundred and fii 
thousand strong. What struck the Franks with the great( 
surprise was that every man was mounted : the whole horde \\ 
composed of horse-archers, and not a foot-soldier was visible. 
In a few minutes the Crusaders found themselves envelope 
The Turks pressed in from all quarters at once ; some appear 
in the rear and cut to pieces many belated parties who had r 
reached the camp at the moment when the fight began ; ^ oth< 
threatened the flanks ; the majority advanced against t 
Prankish line of battle. But they were not drawn up in a 
regular array or order : in loose swarms they kept riding alo 
the crusading line and discharging their flights of arrows ir 
the masses of heavy-armed cavalry. There was no main bo 
which the Franks could charge, and Bohemund, lest his m 
should fall into disorder, refrained from ordering a gene 
advance, hoping that the enemy might ere long close with hi 
But they showed no intention of doing so, and fresh horc 
were continually pressing up, emptying their quivers, and th 
sweeping off to the flanks. At last the Crusaders grew restlt 
and angry : many bands from various parts of the line bro 
out and dashed to the front. But they could not reach t 
Turks, who rode off at their approach, overwhelming them w: 
showers of arrows and slaying their horses by scores — t 
mail-clad men suffered much less than might have been expectt 
But when they turned to make their way back to the line, t 
enemy closed around them, cut off" the stragglers, and destroy 
many of the parties wholesale. Seeing the little profit that t 
sallies brought them, the Crusaders soon desisted from attempti 
to drive off" the enemy, and contented themselves with closi 
their ranks and standing firm. But this passive policy or 
made them a more helpless prey to the Turks, whose arro 
fell so thickly among the crowd that the line began to grow loc 
and disordered. This unequal combat, in which the Fran 
suffered heavy loss and the Turks little or none, went on i 
several hours. At last the host grew more and more unsteac 

^ Raymond d'Agiles, i. 



1097] DORYLvEUM: 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." ^ 

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.- 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 

^ Fulcher, i. 5. 2 Albert of Aix. 

18 



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 tc 
rally, though the Franks in their hasty deployment and hurried 
advance must have come in upon them in considerable disorder.^ 
Struck by a simultaneous impulse of panic, the whole Turkish 
host swept off the field in wild rout : only the Sultan's bodyguard 

^ The ground over which the right cohimn reached the field was mountainous 
(Baldric of Dol; Guibert of Nogent. See Delpech, ii. 153). I conclude, therefore, thai 
they cannot have marched in line : they had started off in haste, and no doubt the reai 
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 regulai 
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 joic 
Bohemund and the left column in their final attack. 



I097] 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 

I 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. 

I 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- 

1 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, 
wrho had been massacred among the tents.^ 

Doryla^um 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 ^ 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 Bo^hemund's messengers, sent off when the Turks 

^ Fulcher, i. 5. Raymond d'Agiles, 239. 

2 Figures taken from William of Tyre — a late authority, though a very capable one. 
^ Fulcher gives five or six hours as the duration of the engagement, and also 
retealrks 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 theii 
cavalry, either by setting those armed with missile weapons tc 
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 itJ 
charges. Probably in their untrained state the foot-soldien 
would have been unable to discharge either function very 
effectively — we have seen that they were not even able to defenc 
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 hole 
a position. This course was almost as dangerous as the om 
which they avoided — that of making a general charge witi 
unprotected flanks into the midst of the great circle of Turkisl 
horsemen. To wait and receive the enemy's shafts withou 
being able to reply to them could only retard disaster, and no 
avert it. As a matter of fact, after five hours of endurance th< 
Franks had recoiled to their tents in a disorderly mass, am 
were about to break up and suffer massacre when their comrade: 
came to their aid. 

Undeserved as the victory had been, it yet gave the Crusader: 
a free passage through Asia Minor. They were not agaii 
obliged to fight a pitched battle till they had arrived at Antiocb 
By the time that the siege of that place had been formed, th< 
condition of the army had greatly changed. The privation 
which it had been forced to endure on its long march had fear 
fully thinned its ranks. The infantry had fallen by the way ii 
tens of thousands : the cavalry had lost the greater part of it. 
horses. For the Western chargers could not stand the heat 
and the forage provided for them was both insufficient ii 
quantity and different in form from that to which they wen 



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,^ the Franks built a tower 
to command the western gate,^ 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 
i the sea at the port of St. Simeon, where a Genoese flotilla had 
I 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 
I engagements during the siege were fought in places where the 
I 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 

^ The siege began October 21st. The new works were not begun till February. 
' The gate of St. George. 



278 THE ART OF WAR IN THE MIDDLE AGES [109^ 

at least to harass the besiegers. Hearing of its approach, thf 
crusading chiefs determined to make a bold stroke at it befort 
it closed in upon them. The Turkish force had met at th^ 
town of Harenc (Harim), sixteen miles east of Antioch. Theii 
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 routt 
had one dangerous defile upon it. About seven miles east o 
Antioch, there is a place where the great lake of Begras at it: 
southern end approaches within a mile of the Orontes : ^ th< 
road passes through this narrow neck of land. This was th< 
point at which the Crusaders resolved to intercept the relievin| 
army : the neighbourhood of their camp was now well known U 
them, and Bohemund had noted this spot not only as giving i 
narrow front where superior numbers would not avail, but als( 
as affording opportunities for a surprise, for the approach wa: 
hilly, and there were many dips in the ground where a consider 
able force could lie hid. 

The Franks could only put into the field seven hundrec 
well-mounted men: their horses had fallen into such bad con 
dition that only that number could be found fit to face a shor 
night march and a battle to follow. With this small bam 
Bohemund, to whom the command had been given for the day 
marched out under cover of the darkness, and, " passing ove 
seven valleys and seven ridges," ^ took post close to the narrov 
neck between the lake and the river. At dawn the Turks wer 
seen advancing, with a swarm of horse-archers thrown out ii 
front to cover their main body. When the whole were in tfe 
defile, the Crusaders, having formed a line of five small squadrons 
with a sixth in reserve, galloped in upon this vanguard. Th 
Turks yielded after a smart skirmish, and fell back in disorde 
on their main body. If there had been room and time for th 
Infidels to deploy,^ the Crusaders must have been crushed, bu 

^ The distance was apparently much shorter in 1098 than now ; probably th 
marshy southern end of the lake is drying up and receding. 

- Raymond d'Agiles, p. 253. 

^ Raymond d'Agiles and William of Tyre agree on this. The latter says : " Con^ 
primentibus eos locorum angustiis, hinc lacu inde fluvio licentiam evagandi inhibent 
ad consuetas discurrendi artes et sagittandi habilitatem discurrere non dabatur. 
William of Tyre, however, does not seem fully to have grasped the topograph 
when he speaks of the Turks as having "crossed the river during the night at th 
upper bridge." There is no river between Harenc and the battle-spot. The onl 
stream between the Bridge-Gate and Harenc is the Iferin (Labotas), the river whic 



1 



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, 1098.^ 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. 

^ For a description of the walls of Antioch, their topography, and ihe Crusaders' 
entry, see chap. vii. of Book vi. 



28o THE ART OF WAR IN THE MIDDLE AGES [1098 

plain of Antioch, and forced the Crusaders to keep within theii 
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 oi 
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 theii 
place of vantage. 

Battle of Antioch^ June 28, 1098. 

The position of the Crusaders, therefore, was hardly alterec 
for the better. Though they had taken Antioch, they were 
themselves practically besieged by Kerboga. After waiting foi 
more than three weeks, during which things went from bad tc 
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,^ they could no longer suffice to cope with the vast 
army of Kerboga. There were many hundreds of knights oi 
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.'^ The infusion ol 
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 

^ 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. 

^ Albert of Aix, iv. 



098] ANTIOCH: SORTIE OF THE CRUSADERS 281 

nemies like the Turks, who placed their whole confidence in 
leir arrows, troops armed with long-range weapons would be 
specially valuable. We have already had occasion to remark 
lore than once that the foot-archer is the most efficient check 
n the horse-archer, because he can carry a larger weapon with 
longer range. Probably Western archery, save in some few 
istricts, was not very efficient, yet it would still be of much 
vail against the Turk. Of course, however, it was not by the 
rrow that the crusading chiefs intended to win. The infantry 
-ere to be mere auxiliaries in the fight, and the charge of the 
lailed horsemen was to deal the decisive blow. The battle 
rder was to consist of lines of infantry with small bodies of 
avalry in the rear of each, the former to open the fight, the 
Ltter to end it. 

On Monday, June 2S, the army was drawn up in the streets 
f Antioch, corps by corps, with the van lying just inside the 
iridge-Gate, and ready to sally out when the signal should be 
iven. It is most difficult to make out the exact disposition of 
le various divisions ; various chroniclers give almost every 
umber between four and thirteen for them. Of the two really 
ood authorities, Raymond d'Agiles and the Gesta Francorum, 
le first gives eight, the second six.^ But Raymond adds the 
arious statement that "the princes had arranged eight 
3rps, but when we had got outside the city, with every man 
ble to bear arms put into the ranks, we found there were 
ve more corps, so that we fought with thirteen instead of the 
riginal eight." ^ Comparing the elaborate list of names in each 
ivision which two or three of the chroniclers give, we find that 
lere is little or no dispute about the first four and the last two 
r the corps, but that in the middle of the line we have a difficulty 
1 reckoning the bodies formed by the Burgundian, South-French, 
nd Provencal contingents. In these parts of the army, which 
ere led by Godfrey of Bouillon and Adhemar, Bishop of Puy, 
)me reckon only two large masses, some four, some as many as 
, iven smaller ones. The general result of our investigation 
I iems to be that though the original intention had been to com- 
ose the centre of two corps of Lorrainers and Burgundians, and 

^ Fulcher of Chartres gives four, the Gesta six (as also many chroniclers who 
I tpy the Gesta), Anselm of Ribeaumont and Orderic Vitalis seven, Raymond eight, 
I r thirteen), Gilo nine, Albert of Aix and William of Tyre twelve, 
f ' Raymond d'Agiles, p. 287. 



282 THE ART OF WAR IN THE MIDDLE AGES [lo 

the left of two corps of Aquitanians and Provengals, yet on gettii 
into the plain these two grand divisions were re-formed respective 
into three and four brigades. If we can trust Raymond d'Agil( 
it was an inspiration of the moment, caused by the fact that t 
numbers of these contingents had been underrated in the coun 
of war which drew up the order of battle.^ 

Summing up our authorities, we may conclude that t 
probable order was as follows: — (i) North- French, under Hug 
brother of the King of France ; (2) Flemings, under their Cou 
Robert; (3) Normans, under Robert, son of William the Conquer 
These three divisions formed the right wing, and headed t 
column. The next to issue from the gate were the contingei 
(4) (5) (6), three corps of Lorrainers, Burgundians, and Mi 
French, under the general command of Duke Godfrey — the otl 
two bodies in this division seem to have been under Reginc 
Count of Toul, and Hugh Count of St. Pol. The whole w 
destined to form the centre in the fight. Next were (7) (8) (9) (] 
four corps of Provencals, Aquitanians, and West-French, unc 
the general command of Bishop Adhemar, the three other lead< 
in this wing (the left) being Raimbaud Count of Orange, Isoc 
Count of Die, and Count Conan the Breton.^ Finally, (: 
Tancred and Gaston de Beam, with Apulians and Gascons ; a 
(12) Bohemund, with the main body of the Normans of Italy a 
Sicily. The last-named corps was to form a reserve divisi 
behind the others, and to guard the rear when all should he 
defiled over the bridge and into the plain.^ 

The only useful notice which we have concerning the numl 
of men in each division is Albert of Aix's statement that Di 
Godfrey's own corps consisted of no more than two thousa 
men, horse and foot all told. Albert grossly exaggerates ' 
weakness of the Franks in all his account. But Godfrey's co 
may have been smaller than the rest — we are told at least ti 

^ The original design, according to Raymond, was to make four grand divis 
— (i) North-French, Plemings, and Normans ; (2) Lorrainers and Burgundians; 
Aquitanians and Proven5als ; (4) Sicilian and Apulian Normans (Raymond, p. 2 
Each grand division was composed of "duo orpines duplices," i.e. two corp 
two lines, one of foot and one of horsemen. So there were to be eight c< 
in all. 

2 Raymond of Toulouse should have shared the command of this wing ^ 
the bishop, but was left behind in Antioch to observe the citadel with two hunc 
knights. He was too sick to ride that day. 

^ All this array is given wit/i resei'vaiions ; there may be, and probably are, ff 
in it. But the divergences of the chroniclers only allow us to give probabilities. 



PLATE VIT 



Siege AND Battle 

of 

A\NiTrHQ)€:Hl. 

•^Oct I097-Junel098. 

lone Foot 




,098] ANTIOCH: ADVANCE OF THE CRUSADERS 283 

Boheriiund's corps was much larger,^ 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,^ — 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 outfliank and surround 
them in the usual Turkish fashion.^ " The farther they come out 
the more they will be in our power " are said to have been his 
words.^ 

^ 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. 

^ See chap. i. of Book vir. 

^ Albert of Aix, not an eye-witness, and William of Tyre, writing a centurj' 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. 

•* Gesta Francorum, xxix. 3. 



284 THE ART OF WAR IN THE MIDDLE AGES [io< 

It was only when corps after corps came pouring over tl 
bridge, and it became evident that the Christians were far mo 
numerous than he had supposed, and might when drawn up f 
the whole breadth of the plain, and prevent any turning mov 
ment, that Kerboga roused himself and put his army in motio 
Apparently, the divisions of Hugh, the two Roberts, and Godfn 
were already in line, and that of Bishop Adhemar was beginnit 
to take ground to their left, when the Emir endeavoured to thrc 
his right wing across the level ground at the foot of the hill whe 
the remaining Christian corps were intending to draw up. Fiftec 
thousand horse, filing along the foot of the hills, succeeded 
getting round the flank of the Crusaders and placing themselv 
perpendicularly to the still incomplete left wing. These a 
said to have been the Turks of Kilidj-Arslan of Roum, ar 
Ridwan of Aleppo.^ The corps of Bishop Adhemar and tl 
three which followed it had the greatest difficulty in fightii 
their way into line with the centre and right.^ But tht 
succeeded in doing so, and thereby cut the army of Kerboga 
two, the detached corps under the Sultan of Roum becomii 
completely separated from the main force.^ Hence the batt 
consisted of two independent fights — one between the ma 
Christian army and the Turkish centre and left, the oth 
between the detached right wing of the Infidels and the Christie 
reserve under Bohemund. For the latter prince, seeing tl 
fatal consequences which might ensue if Kilidj-Arslan attack( 
Godfrey and Adhemar in the rear, hurried forward and deploye( 
his corps facing westward, with their backs to the main bod 
His position must have been parallel with the divisions 
Adhemar and Godfrey, i.e. behind the left centre of the ma 
army. Godfrey, according to some of our sources, hastily se 
the corps of Reginald of Toul to assist in keeping off the attac 
from the rear. 

In the main battle the Crusaders won a complete victo: 

^ But this we have only from two secondary chroniclers, William of Tyre and 
authority, Albert of Aix. 

^ Raymond d' Agiles, p. 286 : ' * We had to strive hard in the space at the foot 
the hills, as the enemy was trying to envelop us, and had their largest corps in frc 
of us." 

' '* Denique divisi sunt Turci : una pars ivit contra mare ; alii steterunt contra no; 

* Ralph of Caen compares the Christian army so arrayed to the snake of the fal 
which had a head at each end, or to a monster with two faces, and specially mentic 
that Bohemund "turned his back to his friends, and his face to his enemie 
(pp. 169, 170). 



t 



398] ANTIOCH : VICTORY OF THE CRUSADERS 285 

ith astonishing ease. Kerboga was a bad general, and his 
^lleagues, the Emirs of Damascus and Aleppo, were mistrustful 
f him and of each other. Moslem historians tell us that at the 
loment of action a great body of Turcoman auxiliaries, with 
hom Ridwan of Aleppo had been tampering, treacherously took 
) flight and threw the whole line into confusion. It is certain, 
t any rate, that when the Christian armies advanced in steady 
ne, with archers in front and knights behind, the Turks retired 
cm their first station towards their camp. There they again 
lade a front, but there was no further chance of putting their 
sual tactics into play, since the Franks filled the whole plain from 
le river to the hills, and could not be outflanked. Their first 
itreat had some semblance of order, but when pressed again the 
afidels broke up more and more, and finally fled at full speed, the 
Dwardly Kerboga at their head. They made off by the road 
etween the Orontes and the lake of Antioch, abandoning their 
imp and the masses of unfortunate camp-followers to the sword 
f the Franks. " No man of rank fell," says Kemal-ed-din, " but 
lere was a horrid slaughter of our foot auxiliaries, grooms, and 
irvants." ^ 

The combat in the rear had been much more serious. The 
urks of Roum and Aleppo fell with fury upon Bohemund's 
Drps, where the infantry threw themselves into a dense circle 
nd did their best to hold firm. They were in great danger,, 
xposed to the Turkish arrows and attacked at intervals by 
arties who abandoned their usual tactics and charged in with 
le sword. The corps of Reginald of Toul when it came up was 
Iso assailed with great vigour, and suffered heavy loss : accord- 
ig to some authorities, nearly the whole of its infantry was cut 
) pieces. But presently the Turks saw their own main army 
ying, and knew that the battle was lost. Apparently, too, the 
ictorious Crusaders detached more troops to help Bohemund. 
'iring the grass to cover their retreat,^ 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 
tterly failed because of the cowardly conduct of Kerboga and 
lie main army. 

^ See the quotations from Kemal-ed-din, Abulfeda, and Abulfarag in Michaud's. 
Hblioth^que des Croisades, iv. 9. 

^ We need not believe the unlikely story about the smoke signals concerted 
etween Kerboga and his lieutenants. 



286 THE ART OF WAR IN THE MIDDLE AGES [lo^ 

The battle of Antioch filled both Frank and Turk wi 
wonder. The Christians marvelled at their own victory : wi 
inferior numbers and men debilitated by famine and the heat 
the Eastern sun, they had swept the Infidels before therh in 
single desperate charge. They attributed their success who] 
to supernatural causes : the Holy Lance borne before Bish( 
Adhemar, they said, had turned the enemy to flight, and tj 
hosts of heaven, headed by St. George and St. Demetrius, h; 
been seen co-operating in the victory, " clothed in white, ridii 
on white horses, and bearing white banners before them." TJ 
Moslems attributed the victory of the few over the many, tl 
famished over the well-nurtured, to the inscrutable will of Heave 
desiring to chastise the emirs for their sins. 

To those in search of more earthly explanations the meanii 
of the fight is obvious enough. The Turks had fought on 
more in a space too confined for their usual tactics : the rig 
wing of the Franks rested on the river, and could not be turne 
Their left wing, the point in real danger, broke through the hord 
sent to surround it and got in touch with the hills. When bo 
flanks were protected, they had only to execute a straightforwa 
charge, and the Turks must choose between the hand-to-har 
combat, which they always disliked, and flight. They chose tl 
latter alternative, and the day was won. If the rear had not bet 
guarded by Bohemund and Reginald of Toul, a disaster mig 
well have occurred ; but while the attack on the rear was held 
check, the main Turkish army could do nothing. 

The lessons of Dorylaeum and Antioch should have remaint 
fixed deep in the minds of both Christian and Moslem, but v 
shall see that only the keenest minds on each side suspected tJ 
meaning. Both parties for the next hundred years frequent 
repeated their original blunders — the Turks that of fighting 
cramped ground, the Franks that of failing to combine hor 
and foot in due proportions. 

Battle of As caloUy August 14, 1099. 

There was no general engagement of importance beside tl 
battle of Antioch during the conquest of Syria. The rest of tl 
history of the year 1098-99 consists of a series of sieges, wil 
which we shall have to deal when treating of the siegecraft of tl 
early Crusaders. It was not till August 1099 that another batt 
in the open field was fought, and this time the enemy was not tl^ 



I 



BATTLE OF ASCALON 287 

irk, but the Fatimite ruler of Egypt, El-Mustali Abul-Kasim 

timed. The Egyptians had been in possession of Palestine at the 

iment of the arrival of the Crusaders, and it was from them that 

-^alem had been wrested. Shortly after it had fallen (August 

.j], El-Mustali sent his general, El-Afdal, with a large army 

drive off the Crusaders and recover the Holy City. The 
rces of El-Afdal were unlike those with which the Crusaders 

cl hitherto had to contend. They resembled the old Saracen 

mies with which the Byzantines had so often fought: there 

many thousand infantry, all black Soudanese, armed with 

- and iron maces (or flails) ; while the cavalry consisted partly 

]\Ioorish and Bedouin light horse, partly of mailed troopers of 
e Caliph's regular army. All of these were spearmen, and not 
chers like the Turks. Having long been at war with the 
jrkish princes of Syria, El-Mustali had no help to expect from 
em. But there seem to have been a few mercenaries of 
irkish blood in his ranks. The whole army is estimated at 
usual vague figure of three hundred thousand by the crus- 
j writers. It may possibly have reached in reality some fifty 
ousand or sixty thousand in all.^ 

The Franks marched out from Jerusalem on August 13, 
ith five thousand horse and fifteen thousand foot.^ The knights, 

will be observed, were all remounted since their victorious 
arch through Syria, having found Arab horses for themselves 

replace their lost chargers. Hence the proportion of cavalry 

infantry is far larger than it had been at Antioch. When 
ey arrived in the neighbourhood of the enemy, they feared to 
; surprised and surrounded on the march, and formed the army 

nine small corps, each composed of foot and horse. These 
>rps marched three abreast, so that whether attacked in front, 
ar, or flank there would always be three divisions to face the 
.ock, three to sustain them, and three more in reserve.^ So far, 
)wever, were they from suffering from any such danger, that 
ey themselves surprised and captured the flocks and herds of 
l-Afdal's army, which were grazing, under the guard of three 
mdred men, in a valley some miles north of Ascalon. 

i ^ The Moslem Ibn-Giouzi says no more than twenty thousand. This is probably 
\ understatement. Perhaps it only includes the Caliph's regular troops. 

* So say the Princes in their letter to the Pope. The usually trustworthy Raymond 
^es the number as twelve hundred knights and nine thousand foot only. 

^ Raymond of Agiles, p. 388. 



288 THE ART OF WAR IN THE MIDDLE AGES [la 

The fugitives soon brought the news to the Egyptian vizit 
who prepared to fight next morning. He took up his positi( 
on the shore north of the town of Ascalon, with his left wii 
resting on the sea and his right on the hills, which here run son 
two miles inland from the water. In his rear was the town wi 
its orchards and plantations, and the camp pitched outside t 
Jerusalem gate. He placed his Soudanese archers in the fro 
line, his regular cavalry behind them. On the right a corps 
Bedouins were to endeavour to encircle the enemy's flank : i 
the left the sea rendered any such attempt impossible. 

On August 14 the Crusaders came in sight, marching dov 
the sandy plain between the water and the hills, which gradual 
broadens as it approaches Ascalon. When they came into t 
neighbourhood of the Egyptians, they proceeded to deploy in 
line from the order of march in nine columns which they h 
hitherto kept. Apparently the front three columns, unc 
Robert of Normandy, halted, while the second line, unc 
Raymond of Toulouse, took ground to their right next the S( 
and the rear line, under Godfrey of Bouillon, filed off to the I 
and took post towards the hills.^ The whole nine corps th 
came up into a single line, and no reserve was left behind : 
each corps the infantry were formed in front, the cavalry 
the rear. 

When the two armies were within bowshot, the Soudanc 
opened fire on the Crusaders, " falling on one knee to sho 
according to the custom." ^ At the same time the whole Sarac 
army struck up a horrible din of trumpets and nakers to dai 
the Christians, and the Bedouin squadrons rode out to the rig 
to encircle the left flank of the enemy. The opening of t 
fight by the Infidels is described by one good authority 
resembling " a stag lowering his head and extending his hoi 
so as to encircle the aggressor with them ; " ^ but there can ha 
been no attempt to do this on the western flank, where the s 
was too close to allow of any such manoeuvre. 

The turning movement was easily stopped by Duke Godfr 
who charged with his knights and easily rode down the lig] 

^ This deployment seems certain from the words of the Ces^a Francomm, xxx 
which say that Raymond fought on the right, Godfrey on the left, and all the otl 
between them : it names Robert of Normandy, Tancred, and Robert of Flanders 
among those who commanded in the centre, but says that ' ' alii omnes " were there a 

2 Albert of Aix, vi. ^ Fulcher of Chartres, xix. 



PLATE VIII. 



ASCALON, 

Au ^ .14.1009. 

Crusaders ci J^M 
Egyptians ^ c±] 




Q 9 



6 ^* 

^^ o 






<?--^<?:<?.<? p'- p 



Q 
Q 



'^o^'^Qo^Q-' ' ' '^ ^ 



q n .'^ Q Q "^ 



Robert 
Fulcoy 



Battle qtHab , 

Aii g .19,1119. 

Christians ^ p?^ 
7i/rA.s 6 6 6 6 



Infai-itry 



The King 



AritiocKeiae 
Buroiia 



P p P o P ^ 

O XD D o 

^ ^ D O 
^ O o 

Pons of 
Tripoli 



I 



1099] BATTLE OF ASCALON Hi 9§$ 

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.^ 
[n every quarter the Egyptian foot were rolled back on to their 
lorse, and the whole army fell at once into complete confusion. 
Ihey seem to have made a very poor resistance, and the 
Drusaders penetrated everywhere through their line. Robert 
)f Normandy slew the vizier's banner-bearer, Tancred charged 
■ight into the Moslem camp, Raymond of Toulouse hurled the 
lostile left wing into the sea. Some of the Egyptians got into 
:he town, others fled away to the south, some even swam out to 
;heir fleet, which lay moored off the shore. But thousands were 
ilain on the field, many more crushed to death as they tried to 
brce their way in at the crowded gates, and a considerable 
lumber were drowned. For some hours after the fight ended, the 
3rusaders were hunting down fugitives who had concealed them- 
lelves in the orchards or even climbed up trees to hide in their 
ops.^ They captured the hostile camp with vast spoils, and 
larrowly escaped slaying or taking the Emir El-Afdal. The 
dttory was a far more crushing one than either Antioch or 
Dorylaeum, for the enemy had not so good an opportunity of 
jetting ofl', and suffered much more severely. His wretched 
nfantry were completely cut to pieces. jij; ;.^j 

Obviously the Egyptians were an enemy to be treated far 
nore unceremoniously than the Turks. They tried to face 
he heavy cavalry of the Crusaders with less efficient horsemen, 
irmed only with the spear, and their infantry were in no wise 
uperlor to that of the Franks. Hence in an open field they 
v^ere sure to be beaten, even though their numbers were largely 
uperlor, as undisciplined Asiatic armies have usually been when 
hey meet Europeans under competent leaders. The Crusaders 
ame to hold the Egyptians in such contempt that they neglected 
he most common precautions against them, and would attack 
hem if they were but one to ten, and even in most unfavourable 
;i!Ound. This rashness was chastised a few years later at the 
•a^ttle of Ramleh, where King Baldwin suffered heavily at the 
^nds of the despised foe. 

^.^ .^ Albert of Aix, who was not an eye-witness, gives an unintelligible account of the 
^t : I follow the Ges^a, checked by Fulcher and Ra}Tnond. 

• ^This is mentioned by the Arab chronicler Ibn-Ghiouzi as well as by several of 
le Christian writers, <?.^. Albert of Aix. 

19 



296 THE ART OF WAR IN THE MIDDLE AGES [no 

Battles of Ramleh^ September 7, i loi, and May 1 102. 
It had taken the Vizier El-Afdal two years to recover fron 
the shock which the defeat of Ascalon had imprinted on hi 
mind. But in the autumn of iioi he sent out a new army t( 
invade Palestine : Ascalon, still in Egyptian hands, served a 
a base for the operation of the host. Baldwin I. had nov 
succeeded his brother Godfrey, who had only worn the crowi 
of Jerusalem for a year. His little kingdom consisted c 
nothing more than his capital and the three seaports of Jaffc 
Arsouf, and Caesarea : the last two he had only just subdued b 
the aid of a Genoese fleet in the summer of iioi. Baldwin a 
his life through was a rash and reckless leader, one of th 
typical Prankish generals on whom the Byzantine authors poi: 
so much scorn. The Egyptian troops were not so strong a 
they had been at Ascalon, but still very numerous : Fulcher c 
Chartres estimates them at eleven thousand horse and twent} 
one thousand foot;^ the Moslem chroniclers state that the 
were led by the Emir Saad - ed - Dowleh. Baldwin, howeve 
resolved to march against them with the scanty force that h 
could collect in Jerusalem at a few days' notice. He would nc 
wait for outlying parties of his own followers, much less sit sti 
for weeks while reinforcements should be summoned froi 
Antioch or Edessa. The Egyptians having moved out froi 
Ascalon, Baldwin left Jerusalem and marched down to Jaffa c 
September 5. The Egyptians did not come to meet him ther 
but pushed in between the king and his capital, marching 1 
Ramleh — a point equidistant from Jerusalem and Jaffa. Thith 
Baldwin followed them with two hundred knights, sixty sergeani 
hastily mounted on borrowed horses to swell the number of h 
cavalry, and nine hundred infantry. He divided this little arn 
into six corps, each containing both horse and foot, and march( 
recklessly into the midst of the Egyptian host, who had bee 
warned of his approach, and had formed up with a front far ot 
flanking the Crusaders on both wings. Baldwin and his litt 
band plunged in among them " like fowlers into the midst of 
covey of birds." Of the exact order of the Franks we have i 
further particulars than the vague statement of Fulcher, th 
they were "arrayed according to the rules of military ar 
Even the simple critics of the twelfth century, however, 

^ Fulcher, chap. xxvi. 



iioi] FIRST BATTLE OF RAMLEH 291 

ready to grant that Baldwin's attack was made with a rash 
disregard of possible dangers.^ 

It seems that when the lines were a thousand paces apart ^ 
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,^ 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 
:ame 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 
knights w^hom 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 all belonged to the two squadrons which had perished at 
ihe opening of the battle.^ A much larger number of the 

^ " Minus caute," says Ekkehard in his /lie ro so /f mi/a. - Ekkehard. 

^ Ekkehard says that one squadron only was cut to pieces by a fiank attack ;: 
Fulcher (a better witness) that two were destroyed. Albert of Aix, exaggerating 
earfully, makes four perish, and says that the king won the battle with his own forty 
lifrigbts alone. 

* Fulcher, p. 125. 



J9« THE ART OF WAR IN THE MIDDLE x\GES [iioi 

infantry had also been cut down. The loss of the Egyptians i< 
put as high as fifteen thousand men — an impossible figure ; the} 
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. 
The whole fight had only occupied a short hour.^ 

After having been within an ace of destruction in Septembei 
IIOI, it is astonishing to find that Baldwin repeated his reckless 
tactics in an aggravated form on the very same battlefield, onl} 
eight months after.^ In May 1 102 the Egyptians once more brokt 
forth from Ascalon and marched on Ramleh, where they pitchec 
their camp. Baldwin set out against them with his militar) 
household, without waiting for any reinforcements from the out 
lying towns of his little kingdom. He picked up at Jaffa a banc 
of pilgrim knights, survivors of the unfortunate Crusade of iioi 
who were just embarking to return to France : they were led b^ 
Stephen of Blois and Stephen Count Palatine of Burgundy. Thi 
gave him no more than two hundred lances ; nevertheless, h< 
marched straight on Ramleh, believing (it is said) that the enem; 
w^ere only a raiding party some eight hundred or a thousan( 
strong : ^ as a matter of fact they were a whole army, . about a 
large as that which had been beaten on the same spot in th 
previous year.^ " It was pride and presumption that led th 
king,'' says Fulcher, " not to w^ait for more troops, not to mov 
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 marc 
until he saw before him, and far closer than he wished, such 
huge multitude of the enemy." With no more than his tw , 
hundred knights Baldwin rode straight at the hostile centn 
hoping to repeat his exploit of the previous year. But the odd 
were too great, and this time he had no infantry with him t 
protect his rear and take off some of the pressure. The Frank j 
were engulfed in the hostile mass, and slain off almost to a maj j 
Baldwin and a few more cut their way out of the melee, but tlr 

^ See the chronicle quoted on p. 1 7 of vol. iv. of the Bibliothcque dcs Croisades. 

- Fulcher. 

^ In all that follows I have taken Fulcher as guide, not Albert of Aix, who vari 
hopelessly from Fulcher's tale, and was not, like the former, on the spot. 

^ Fulcher, chap, xxvii. p. 135. 

° Ibid. : ' ' Twenty thousand horse and ten thousand foot, the latter all Soudanes< 
— numbers grossly exaggerated, we need not doubt. 



1 102] FINAL DEFEAT OF THE EGYPTIANS 293 

Counts of Blols, 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 
^^ii 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. 

oi oiii li .xili: 

■ ■ - ' — di obnn: 
:>»o odi ih 



CHAPTER IV 

THE TACTICS OF THE CV^USA.'D'E.VsS—COntmucd 

Section II. — Tactics of the Later Battles : Victories at Hob, 
Hazarth, Marj-es-Safar, Arsoiif, Jaffa. 

A S our task is not to write a history of the kingdom o: 
jfjL Jerusalem and its wars, but to indicate the main militarv 
tendencies of the crusading age, w^e must not attempt to give ir 
detail each of the numberless fights of Frank and Moslem, bui 
only to comment on such of them as show features of import- 
ance. Speaking in general, we may say that the same point.' 
of interest which we have observed at Dorylaeum, Antioch 
Harenc, Ascalon, Ramleh, and Jaffa, are to be found repeatec 
in all the fights of the twelfth century. 

Against the Turk the Crusaders were generally successful i 
they took care (i) to combine their cavalry with a solid body o 
infantry armed with missile weapons ; (2) to fight on grounc 
where the Infidel could not employ his usual Parthian tactics o 
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, tin 
Crusader was far more certain of success ; he had only to ust 
the common military precautions, and he might fairly count 01 
victoiy. The battles of the Franks with these less formidabli 
foes sometimes remind us of the early English battles in India 
where the few striking boldly at the many were so oftei 
victorious in spite of ever)'- disadvantage. The one all-importan 
canon which had to be observed was that there must b< 
infantry on the field to serve as a support and rallying point fo 
the cavalry. If the foot-soldiery seldom won the battle, the; 
always made the winning of it by the knights possible. 

If, on the other hand, the Frank chose to advance recklessb 

S34 



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 
:lass-pnde to lead him astray, and left his infantry behind, he 
.vas liable to suffer terrible disasters. 

We have selected as examples of typical victories of the 
Crusaders the battles of Hab (i 1 19), Hazarth (1125), Marj-es- 
Safar (i 126), Arsouf (i 191), Jaffa (i 192). As instances of defeat 
wrought about by neglect of first principles, we may take the 
nghts of Carrhae (1104), Tiberias (1187), Acre (1189), and 
i\Iansourah (1250). 

Battle of Hab, August 14, 1 1 19. 

On the 27th of June 11 19, Roger, Prince of Antioch, had 
[alien with many of his knights in the disastrous fight of Cerep. 
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 own 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.^ 

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 

^ All this comes from Gautier the Chancellor. William of Tyre, Fulcher, and 
the rest are vague, and speak at second-hand. 



296. THE ART OF WAR IN THE MIDDLE AGES [11 19 

Antioch ; another party of Antiochenes seems to have beer 
detailed as a rearguard, if our chief source, Gautier the 
Chancellor, can be trusted.^ The squadrons of Pons anc 
Robert were placed not parallel to, but somewhat to the reai 
of, the front line, in order that they might defeat attempts tc 
turn its flanks, while the king could strengthen it if the mair 
pressure of the enemy was thrown upon its centre. Whethei 
by chance or design, this order bears a striking resemblance tc 
that which the Byzantine Leo the Wise advocates for ust 
against the Moslem. A comparison of the plan on p. 195, wit! 
the sketch of the battle of Hab on p. 290, makes this clear in c 
moment. The only difference is that Baldwin had infantry 
perhaps two thousand or three thousand strong, behind his firsi 
line of horse, while Leo is describing the order of a division o 
cavalry unassisted by any foot-soldiery. The nine squadrons 
each about eighty strong, were three in the first rank, three ir 
the second, one on each flank, and one behind. Il-Ghazi anc 
Toktagin seem to have hoped that they might be able to surprise 
the Franks at daybreak, but when the sun rose Baldwin'.' 
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 throvv^n 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 or 
the right, while Toktagin, with the men of Damascus anc 
Emesa, held the left. The Turks, were well aware that the 
greatest danger to themselves lay in the combination of infantrj 
and cavalry by the Christians. Il-Ghazi had therefore resolvec 
to do his best to overwhelm the front line of the enemy, anc 
prepared a desperate assault on Baldwin's centre, where all the 
foot-soldiery were collected. They and the three squadrons o 
knights in front of them were very fiercely assailed ; ^ the 

^ This is the only way ©f construing "acie comitis Tripolitani a dexteris posita 
aciebus baromim a sinistris et post : jussu regis quibusque sno loco positis " (Gautier 
p. 460). If the Antiochenes had all been on the left of Baldwin in one body, >vi 
should have had acie, not acichtts. Bongars prints the colon before et, but evidenti; 
it should be before y^jx?/, making no good sense if introduced after sinistris. 

2 That the knights were in front of the infantry and not behind, is shown b^ 
Gautier's wording: "Turci, ambitiosi manum pedestrem prosternere, qua gravio. 
refrenabantur, cum banc praccedentibtis aciebus, et acies hac protegi videbant, v 
maxima . . . arcubus brachiis immissis, strictis ensibus, nostros percutiunt," etc 
The \sox^ praecedcntibus is conclusive (p. 461). 



119] BATTLE OF HAB 297 

orsemen were driven back on the foot, and the latter attacked, 
ot with the usual arrow-shower of the Turks, but by vigorous 
larges home with lance and sword. The Prankish footmen, 
hen the knights were driven off, proved unable to bear up 
gainst the Mesopotamians. Armed with missile weapons to 
ithstand the Turkish bow, they were less fitted for close 
Dmbat. They fought well, but began to fall into disorder, and 
)st heavily. 

Meanwhile, the fortune of battle on the wings had been 
/enly divided ; on the right Il-Ghazi's men had assailed and 
eaten back the Count of Tripoli, whose whole corps was 
nally driven in and thrown on to the flank of Baldwin's own 
ivision in the second line. On the left, however, Robert Fulcoy 
nd the Antiochenes had charged the men of Damascus with 
ich vigour that they had completely scattered them, and 
riven them off in confusion. Robert might have won the day 
y promptly charging the hostile centre from the flank. But 
o such idea entered into his head ; his main desire was to 
ilieve his ov/n castle of Zerdana, whose fall had not yet reached 
is ears.^ Accordingly he pursued the Damascenes for a space, 
nd then rode straight for Zerdana without making any further 
ttempt to join in the battle. He and his corps were absent 
cm the field during the remainder of the engagement. 

Il-Ghazi's men on the other flank made no such mistake, but 
osed in on Baldwin's second line. The fight now became 
2ry confused ; the van and right wing of the Franks were 
riven in on their centre in a disorderly mass, and it remained 
) be seen whether the king would be able to save the day with 
is reserve. Time after time he charged out with his knights 
nd drove off each swarm of Turks as it pressed in to complete 
lie victory. Whether the attack threatened front or flank or 
iar, he and his chivalry were always at the point of danger, 
igain and again the cry of " Holy Cross ! " and the impact of the 
eavy squadron of men-at-arms drove back the Infidels from 
leir prey.'^ Towards evening Il-Ghazi gave up the struggle 
nd rode off, leaving Baldwin in possession of the field. .<;(>>[ 

^ So Keraal-ed-din, who seems very well informed. Gautier the Chancellor 
lagines that the news had already reached the Christians, which is improbable. 
obert 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, 
ic exclamando Sanctae Crucis protectionem et auxilium, velocissime irruit, 
irfidos prostravit et in dispersionem impulit," etc. (Gautier, p. 461). 



298 THE ART OF WAR IN THE MH:)DLE AGES [in 

As he retired, the lord of Mardin came into coliision wit 
the corps of Robert Fulcoy, returning tardily to join the kir 
after they had discovered that Zerdana was already in the hanc 
of the Infidels. The Antiochene knights, marching in disordi 
and without proper precautions, were easily dispersed by tl 
Turks, and Robert himself, falling from his horse in the fligl- 
was made prisoner. He was put to death by Toktagin in co 
blood some days after the battle : it is impossible to s^ 
that his fate was undeserved, as his selfish abandonment 
his comrades at the outset of the battle merited the heavie 
punishment. 

Baldwin, unaware of this disaster, held the field till nig 
and then retired to Hab. He returned next morning to bu 
his own dead and strip those of the enemy. As the Turks h', 
entirely disappeared, he with justification regarded himself 
victor. The battle had in truth been indecisive ; but as t' 
enemy made no further advance against Antioch, the end f 
which it had been fought was achieved. The losses had bc< 
very heavy : Baldwin counted a hundred knights and sevi 
hundred footmen among the slain, and many more were d 
persed and did not rejoin for several days. The Turks had Ic 
from two thousand to three thousand horse. 

The incidents of this battle, in which the fortune of the d 
was for a long time so equally divided, remind us of those 
MontThery, and Gautier's account of the flight of each si 
may w^ell stand beside the well-known passage in Commin 
" Our fugitives," he wTites, *' fled to Hab, to Antioch, and ev 
as far as Tripoli, reporting that the king and the whole arr 
had been exterminated. On the other hand, those of the Tur 
who had been driven off the field (by our left wing) pour 
into Aleppo, swearing that Il-Ghazi and Toktagin and all i 
Turkomans had been slain to the last man."^ If Baldw 
could claim that he had held the field at sunset, Il-Ghazi coi 
display as trophies one of the royal banners of the La' 
kingdom, torn from the king's squire who bore it, a 
Robert Fulcoy and many other noble prisoners. That, af 
massacring thirty of them, he then returned to Mardin to ra 

^ In face of Gautier's explicit statements, it is impossible to believe Kemal-cd-d 
allegation that at nightfall the Turks pursued the Christians to the gates of Hab. 
any of them did follow, it must have been at a safe distance, and as scouts rat 
than pursuers. 






:25] BATTLE OF HAZARTFI 299 

ore troops instead of pursuing his campaign, is a sufficient 
oof that the claim of victory which he made was a very 
npty one. But it seems to have deceived his chronicler, 
emal-ed-din, from whose pages we should never gather 
lat Baldwin also could declare himself the conqueror in 
[G Strife. The events of the succeeding months plainly 
lowed who was the real victor. Il-Ghazi returned home ; 
aldwin kept the field, and retook in the autumn Zerdana 
id most of the other castles and cities which the Infidels 
id captured after the death of Prince Roger. 

This battle of Hab or Danit has many points of interest. It 
lows us the Crusaders adopting for the first time a much more 
)mplex order of battle than the simple line of infantry sup- 
3rted by cavalry which they had displayed at Antioch, Ascalon, 
id Ramleh. Baldwin, instructed by his many battles with 
le Turk while he was but Count of Edessa, had employed 
; king the fruits of his experience. The Turks, too, have learned 
uch : they no longer trust entirely to the bow, but charge 
3me vigorously with sword and lance. They have come to 
^e that the Prankish foot-soldiery with their missile weapons 
*e even more dangerous to them than the knights, and devote 
lost of their energy to clearing away the infantry, not en- 
savouring to shoot them down, — an attempt in which Turks 
ildom succeeded, owing to the inferiority of the horseman's 
3\v to the arbalest, — but to ride over them with the lance, 
hat they succeeded on this occasion was apparently due to 
aldwin's mistake in drawing up his three squadrons of knights 
I front of and not behind the infantry of the centre. 

For a further development of the tactics of both sides, we 
lUst advance a few years, to the battles of Marj-es-Safar and 
[azarth. 



Battle of H azarth^ Jmic 1 1 , 1 1 2 5 . 

At Hazarth, which was fought on June 11, 1125, Baldwin 
iejTis to have returned to the simple order of battle of the 
ays of Antioch and Ascalon. He drew up his army in 
lirteen small corps, each consisting of infantry and cavalry, 
s there were eleven hundred knights and two thousand foot, 
le squadrons must each have been about eighty strong (much 
le same as at Hab), and the infantry divisions have mustered 
)mevvhat over a hundred and fifty. These thirteen bodies were 



300 THE ART OF WAR IN THE MIDDLE AGES [ii: 

divided into a centre and two wings : the right was compose 
of the troops of Antioch, the left of those of Edessa ar 
TripoH ; the centre,^ the strongest of the three divisions, w. 
formed of the king's own vassals from Palestine. Presumab 
the wings contained each four and the centre five corps, b 
neither Fulcher nor William of Tyre, our two authoritit 
definitely state the fact. 

Il-Borsoki, the opponent of Baldwin, arrayed his fifte< 
thousand horse in twenty-one corps, and pressed forward 
attack the Frankish infantry — we have no mention of t 
attempting any encircling movement after the usual Turki 
fashion. The interchange of missiles had gone on for son 
short time, and close fighting had begun, when Baldw^in ga 
orders for a general charge of the cavalry.^ The Infidels stO( 
firm for a moment, but, when the knights burst in among thei 
lost heart, broke, and fled. Two thousand of them fell, wh 
the Christians only lost twenty-four. The proper combinati< 
of infantry and cavalry had secured an almost bloodless victoi 

Battle of Mar j-es-S afar, January 25, 11 26. 

In the following year the Turks for the first time put foe 
soldiery in the field. They had evidently realised at last th 
the combination of the two arms was more effective than th< 
own horse-archery. In January 11 26 King Baldwin had cross 
the Jordan and advanced toward Damascus, harrying the lai 
far and wide, in revenge for a similar raid which Toktagin h 
directed against Palestine in the preceding autumn. Agair 
him came forth the Atabeg chief and his son, bringing wi 
them not only their riders, but " chosen youths trained to spri 
up armed behind the horsemen, who, when the enemy dn 
near, descended and fought on foot : for so they hoped 
disorder the Franks by attacking them with infantry on o 
side and cavalry on another."^ The Jehad had been preach 

^ I do not think we are justified in concluding from Fulcher's (chap. Ixii. ) call 
Baldwin's own corps " densior et posterior" that he was in a second line. Proba 
only *' last and largest " is meant. William of Tyre evidently read it so wl 
he wrote "iVz medio dominum regem," and woi pone or post. Fulcher says t 
"Baldwin charged, bidding the rest follow, for they dared not commence the f 
before he gave the word."- 'If he' was^^in- a second line, this would have bi 
impossible. '> ]{;'■ 'cfxi .: >od : 

- "The bows had been bent and.-i|^e di[»wn sword was being used at cl 



quarters," says Fulcher (chap. Ixii.). 



Fulcher, last words of chap. Ixxi; 



:.vAr 



r2 6] BATTLE OF TvlARJ-ES-SAFAR 301 

I Damascus and its subject towns, and many thousands of un- 

ained citizens went out on foot to fight for Islam. 

The armies met at Marj-es-Safar, not far from Damascus, 

• 1 the 25th of January, the day of the conversion of St. Paul. 

; aldwin drew up his men in twelve corps, each containing 

: 3th infantry and cavalry, "that the two arms might give each 

;her the proper support." ^ The Damascenes were not in any 

j iry great numerical superiority, save in the number of their 

\ regular foot-soldiery ; the Christian clironiclers confess that 

1 le two armies were not very unequal, and do not ascribe 

I le usual vast preponderance to the enemy. But whether it 

as that they were fighting close to their capital to protect 

leir own homes and families, or whether it was the unwonted 

ssistance of infantry which helped them, it is certain that they 

lade a much fiercer stand than usual. It was one of the 

;iffest, though not the most bloody, fights in which the Franks 

ad engaged for many years.^ Fulcher allows that for a space 

le battle seemed going against Baldwin ; the arrow-shower 

as too bitter, and " no part of body or limb seemed safe against 

le shafts, so thickly did they fly." The host recoiled for a long 

Dace, and it was only by a desperate rally in the afternoon that 

saved itself and resumed its advance. "But our king bore 

imself w^ell that day, as did all his knights and vassals, and 

Jmighty God was with them."^ At dusk the Turks fled, and 

le day was won. Two thousand Damascene horse and an 

inumerable number of the Infidel foot had fallen ; of the 

'.hristians twenty-four knights and eighty infantry only were 

.ain. William of Tyre, in his rather unsatisfactory narrative 

f this battle, says that the Christian foot, fired by the example 

f the king and his knights, charged the enemy at the supreme 

loment along with the cavalry, and that they did most damage 

3 the Turks by shooting their horses, so that the dismounted 

I . 

V 1 " Ordinatae sunt tarn militum quam peditum acies duodecim, ut ab alterutra 
i )rroboretur caterva, si necessitas admoneret" (Fulcher, cap. Ixx.). This can only mean 
I lat foot might help horse and horse foot, not that each of the twelve corps might 
'; elp the other. It is hardly necessary to point out that altertiter can only be used 
r f two, not of many ; bat I have seen several accounts of the battle by modern 
i alhors where this simple rule of Latinity is neglected. 

■^ William of Tyre is of course wholly in error when he calls it the most 
^ angerous and doubtful fight since the foundation of the realm (xiii. § 18). At 
lab, only seven years before, the Christian losses were eight times as great and the 
asult far more uncertain, 

^ Fulcher, cap. Ixx. 



302 THE ART OF WAR IN THE MIDDLE AGES [u 

Infidels fell easy victims to the pursuer.^ But it is not easy 
make out whether the infantry, as he conceived the fight, we 
behind or in front of the knights. For, on the one hand, 
makes the foot-soldiers " pick up and carry back to the bagga 
their wounded comrades, and set on their feet again those w 
had been overthrown ; " w^hile, on the other, they are said to she 
the Turkish horses, so that the riders "fall into the han 
of their companions who follow behind." The first statcmc 
seems to indicate that the knights had already charged o\ 
the ground which the Infantry were crossing; the second th 
they were following behind them. But William is not alwa 
happy in following his authorities for battles that took pla 
before his own day, and his picture here is decidedly confust 
In all probability the action began with the infantry in the fi 
line, and the cavalry in support. When it grew hot, the cava! 
must have charged out to the front, and in the final advar 
the foot-soldiery must have been following in the wake of t 
knights to complete the victor}^ rather than preceding the 
It is a pity that we have not any detailed account of t 
battle from Moslem sources ; if it existed, we might clear 
its difficulties, as we can those of the fight at Hazarth, by t 
comparison of the two hostile chroniclers. 

There are many Christian successes worth recording 
the years between Marj-es-Safar and the fall of Jerusalem 
1 187. But as they are not of any special tactical importan 
presenting merely the same features that we have alrea 
noted, they may be passed over wn'lhout any detailed nar 
tion. The defeats of this period are more interesting than t 
victories : notes on several of them will be found in t 
succeeding chapter, where we treat of the causes of the ma 
failures of the Franks. 

The battle which must next arrest our attention is the 1; 
of the great triumphs of the Christians, and the most notable, 
it was won over the finest general whom the Infidels ever own' 
the great Saladin himself, commanding the most powerful a 
most formidable — if not the largest — host which the Mosle: 
ever put into the field. The Christians, too, were in far lar^ 
force than ever before in any battle of the Holy Land. It 

^ " Equis hostium sauciandis omnem dabant operam, eorunique sesFores i 
sequentibus sociis parabant ad victimam " (W. T. xiii. §, i8). This, I presu: 
means shooting rather than stalbing the horses. 



I 



i9i] RICHARD I. AT ACRE 303 

)rtunate that we have excellent accounts of the fight from both 
des, and that its topography can be easily ascertained. Every 
ctail of it is well worth study. 

Battle of Arsoiif, September 7, 1 191. 

i. After a siege of nearly two years, Acre had been recovered 
ythe Franks on July 12, 1191. The garrison had laid down 
s arms and surrendered to the kings of France and England, 
fter having protracted its defence to the last possible moment, 
aladin had done his best to succour the place, and delivered 
erpetual assaults on the camp of the besiegers, but all to no 
urpose. Seeing that there was no hope of relief, and that Acre 
lUst fall by assault in a few days, the Emirs Karakush and 
lashtoub opened the gates, after promising that they would 
iduce the Sultan to pay two hundred thousand bezants as 
insom for the garrison, and also to restore the True Cross and 
fteen hundred Christian prisoners, the survivors of the disaster 
f Tiberias, who were in chains at Damascus and elsewhere. 

For some weeks after the fall of the great fortress, the 
'hristians remained encamped in and around Acre, while 
>aladin still observed them from his camp on the mountain to 
he east. The delay was caused partly by the exhaustion of 
lie victors, partly by the necessity for repairing the shattered 
^alls of the city, partly by the protracted negotiations concern- 
ig the ransom of the garrison. Meanwhile, Philip of France 
Dok his way homeward amidst the curses of the whole arm\*, 
wearing that on his return he would be a quiet and peaceful 
eighbour to the dominions of the King of England. " How 
lithfuUy he kept that oath is sufficiently notorious to all men, for 
he moment that he got back he stirred up the land, and set Nor- 
aandy in an uproar." ^ He left the bulk of his army in the camp 
nder the Duke of Burgundy and Henry Count of Champagne. 

The attempts to come to an agreement with Saladin failed 
lopelessly. Into the ugly story of the massacre of the Turkish 
.arrison, when their ransom was not forthcoming, we need not 
:nter. On Tuesday, August 20, Richard and the Duke of 
burgundy beheaded the two thousand six hundred unfortunate 
aptives, and all chance of peace was gone. Two days after, the 
rusading army set out upon its march. 

^ Iliiierarium Regis Ricardi, iii. § 22 : ** Quam vero fideliter hoc steterit jura- 
tiento satis innotuit universis," etc. 



304 THE ART OF WAR IN THE MIDDLE AGES [ii 

Richard had as his objective Jerusalem, whose recovery \v 
the main end of the Crusade. But to move directly from Ac 
on the Holy City is impossible. The mountains of Ephrai 
interpose a barrier too difficult to be attempted when ; 
alternative route is possible. For a march on Jerusalem t' 
best base is Jaffa, and to that place Richard resolved to trar 
fer himself and his army. He accordingly arranged that t 
host should march along the great Roman road beside the s 
by Haifa, Athlit, Caesarea, and Arsouf, while the fleet shou 
advance parallel with it, and communicate with it at eve 
point where it is possible to get vessels close to the shore. Tl 
co-operation was all-important, for the army was lamentat 
deficient in means of transport, and depended on the ships 1 
its food. So few were the beasts of burden, that a great pe 
of the impedimenta had to be borne on the backs of t 
infantry, who loaded themselves with tents, flour-bags, ai 
miscellaneous necessaries of all kinds. Nearly half of the 
were employed in porter's work, and thereby taken out oft 
ranks when the host began to move forward. No food was 
be found on the way, for Saladin had already ravaged the sho 
and dismantled Haifa, Caesarea, and Arsouf. 

It was obvious that the Crusaders would be harassed 1 
Saladin the moment that they started on their march. T 
temptation to assail a host strung out in one thin column alo] 
many miles of road would certainly draw the Turks down frc 
their strongholds in the hills. Richard had therefore to provi 
an order of march which should be convertible at a momen 
notice into an order of battle. His front, rear, and left flai 
were all equally liable to assault. Only the right would alwa 
be covered by the proximity of the sea. 

In view of this danger the king made the best dispositr 
possible. Next the sea moved the beasts of burden and t 
infantry employed to carry loads. Inland from them were t 
cavalry, distributed into compact bands and spaced out at eqi 
intervals all along the line of march. Inland again from t 
cavalry were the main body of infantry, marching in a co 
tinuous column, and so covering the whole eastern flank of t 
army. Though the contingents were placed so close that 
gaps were left between them, they were for purposes of organic 
tion divided into twelve bodies, to each of which there w 
attached one of the cavalry corps, which marched level with 



i9i] RICHARD MARSHALS HIS ARMY 305 

i Thus there were twelve divisions of foot and twelve of horse ; 
' hese smaller units were united into five main corps, of which 
he exact composition is not easy to ascertain. The Templars 
i nd the Hospitallers, who knew the country well, and had in 
I heir ranks many '' Turcopoles," i.e. horse-bowmen armed like 
i he Turks and specially fit to cope with them, took the van and 
he rear, the two points of greatest danger, on alternate days. 
Vith the centre division of the army moved the royal standard 
'f England fixed on a covered waggon drawn by four horses, 
ike the carrocJiio which the Milanese had used at Legnano a 
2VV }-ears before. The order of the various corps was, as v/e 
;ather, somewhat varied on different days. On one occasion 
lichard and his own military household took the van, but 
:sually he reserved for himself no fixed station, but rode 
•ackward and forward along the line of march with his house- 
■old knights, carefully supervising the movement of the whole 
.nd lending aid wherever it was required. The heat was great, 
September being not yet come, and the king was determined 
lot to harass the army by long stages. Accordingly he moved 
ery slowly, using only the early morning for the march, and 
eldom covering more than eight or ten miles in the day. 
iloreover, he habitually halted on each alternate day, and gave 
lis men a full twenty-four hours (or even more) of rest. Thus 
he host took as much as nineteen days to cover the distance 
f eighty miles betv/een Acre and Jaffa. It is well worth 
•hile to give Richard's itinerary, in order to show the care 
^hich he took of his troops. 

Thursday ^ August 22. — From the neighbourhood of Acre to the river Bekis [2 miles]. 

Friday^ Atr^nst 23. — The army crosses the Belus [2 miles]. 

Saturday, Aztpist 24. — Rest in camp and preparations for march. 

Sunday, August 25. — To Haifa [ii miles]. 

Monday, August 26. — Rest at Haifa. 

Tuesday, August 27. — From Haifa to Athlit, round the shoulder of Mount 

Carmel [12 miles]. 
Wednesday, August 28, — Rest in camp. 

Thursday, August 29. — Rest in camp. The fleet arriv-es and lands stores. 
Friday, August 30. — From Athlit to EI-Melat [Merla] [13 miles]. 
Saturday, August 31. — From El-Melat to Caesarea [3 miles]. The fleet lands 

stores and reinforcements. 
Sunday, Septeviber i. — From Caesarea to the "Dead River" [Nahr Akhdar] 

[3 miles]. 
Monday, Septeinber 2. — Rest in camp. 
Tuesday, September 1. — From the Dead River to the "Salt Kiver'' [Nahr Isken- 

deruneh] [7 miles]. 



3o6 THE ART OF WAR IN THE MIDDLE AGES [119 

Wednesday, September 4. — Rest in camp. 

Thursday, September 5. — From the Salt River through the Forest of Arsouf t 

Rochetaille [Nahr Falaik] [10 miles]. 
Friday, Septeviber 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 lO. — Nahr-el-Aujeh to Jaffa [5 miles]. The fleet lands fres 

stores. 

Throughout the march the army was incessantly worried b 
the attacks of the Turks, especially on the 25th and 30th c 
August and the ist and 3rd of September. The respite on th 
26-yS-gih was due to the fact, that while Richard had hugge 
the coast from Haifa and gone round the shoulder of Mour 
Carmel, Saladin had struck across country, passed the hil.' 
farther east, and come down on to the neighbourhood c 
Caesarea, before the Crusaders, moving slowly and on a long( 
road, had drawn near the place. From August 30 to Sej 
tember 7, on the other hand, he was always within a fe 
miles of them, waiting for his opportunity to dash down froi 
the hills if they exposed themselves. The author of tl 
Itinerarium gives an interesting description of the Turkish tactii 
during these days: — • 

" The Infidels, not weighed down with heavy armour lit 
our knights, but always able to outstrip them in pace, were 
constant trouble. When charged they are wont to fly, ar 
their horses are more nimble than any others in the world ; 01 
may liken them to swallows for swiftness. When they see th 
you have ceased to pursue them, they no longer fly but retu 
upon you ; they are like tiresome flies which you can flap awi . 
for a moment, but which^come back the instant you have stopp< 
hitting at them: as long as you beat about they keep off: t] 
moment you cease, they are on you again. So the Turk, whi 
you wheel about after driving him off, follows you home witho 
a second's delay, but will fly again if you turn on him. Wh 
the king rode at them, they always retreated, but they hu: 
about our rear, and sometimes did us mischief, not un frequent ! 
disabling some of our men " {Itift. iv. § 8).^ 

1 Note on the Battle of Arsouf. 

In my account of this fight I have followed the Itinerarium, Boha-ed-din, : 
KinfT Richard's letter to the Abbot of Clairvaux in Hoveden. All these* t? 
accounts fit into each other admirably. On the other hand, the narrative of Bene 



ii9t] 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 
^ :he Itmerarhun. He is describing the events of Saturday, 
August 31. 

" The enemy moved in order of battle : their infantry 
narched between us and their cavalry, keeping as level and 
irm as a wall. Each foot-soldier had a thick cassock of felt, 
md under it a mail-shirt so strong that our arrows made no im- 
Dression on them. They, meanwhile, shot at us with crossbows, 
,vhich struck down horse and man among the Moslems. I noted 
imong them men who had from one to ten shafts sticking in 
:heir backs, yet trudged on at their ordinary' pace and did not fall 
DUt of their ranks. The infantry were divided into two halves : 
Dne marched so as to cover the cavalry, the other moved along 

)f Peterborough is absolutely irreconcilable with them. He makes much of the 

ighting turn on the crossing of a river by the Christian army, and puts the engage- 

nent on the i6th instead of the 7th of September. It is satisfactory to know that 

i lis story is rendered wholly impossible by the topography of the place. For a mile 

y lorth of the Nahr-el-Falaik the road is bordered by the impassable swamp of the 

i iirket-el-Hamadan. North of this again it runs over flat sand dotted with salt-water 

•ends, and with the forest running down into it. This will not do for the battlefield 

' s described by the Itinerarhun and Boha-ed-din. On the other hand, the country 

outh of the Nahr-el-Falaik suits the narrative excellently. See my map, carefully 

educed from the i-inch-to-the-mile Ordnance Survey of Palestine. 



5o8 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 betweer 
the two halves of the infantry, and only came out when i 
wished to charge. It was formed in three main corps : in th( 
van was Guy,^ formerly King of Jerusalem, with all the Syriai 
Franks who adhered to him ; in the second were the Englisl 
and French ; in the rear the sons of the Lady of Tiberias 
and other troops. In the centre of their army there was visibl 
a waggon carrying a tower as high as one of our minarets, o; 
which was planted the king's banner. The Franks continued t 
advance in this order, fighting vigorously all the time : th 
Moslems sent in volleys of arrows from all sides, endeavourin 
to irritate the knights and to worry them into leaving the: 
rampart of infantry. But it was all in vain : they kept the; 
temper admirably and went on their way without hurryin 
themselves in the least, while their fleet sailed along the coaf 
parallel with them till they arrived at their camping-place ft 
the night. They never marched a long stage, because they ha 
to spare the foot-soldiery, of whom the half not activel 
engaged was carrying the baggage and tents, so great was the 
want of beasts of burden. It was impossible not to admire tl 
patience which these people showed : they bore crushing fatigu 
though they had no proper military administration, and we: 
getting no personal advantage. And so they finally pitchc 
their camp on the farther side of the river of Caesarea." ^ 

From the 29th August to the 6th September, Saladin hz 
been perpetually seeking an opportunity for delivering a serio' 
attack. But the caution and discipline which Richard had ir 
posed upon his army foiled all the hopes of the Infidel. It b 
came evident that, if the Christians were to be stopped befo 
•they reached Jaffa, a desperate attempt must be made 
i)reak in upon them, in spite of their orderly march and fir , 
array. Saladin resolved, therefore, to try the ordeal of battle I 
the ground between the Nahr-el-Falaik (the river of Rochetail' 
.and Arsouf There was every opportunity for hiding his he 

^ This account of the distribution of the Christians does not tally with 
Jtiuerariumy and is probably wrong. Boha-ed-din calls Guy " Geoffrey" by a ci 
•error. 

- Barons of the party among the Syrian Franks who opposed King Gu) 
wished to recognise Conrad. 

•^ Boha-ed-din, p. 252, in the Chroniqueurs Orientaux. 



ii9i] 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 5 th and 7th of September) ran 
between this forest and the sea. He was not less conscious 
than Salad in 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.^ 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 v/hole of Salad in'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.^ 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 

^ Itinerarittm, iv. § 16. ^ Ibid. iv. § i7. 



5IO THE ART OF WAR IN THE MIDDLE AGES [1191 



were made up from the French, the barons of Syria, and tht 
miscellaneous small contingents from other lands. Lastly, tht 
Hospitallers brought up the rear. Probably the French con 
tingents were divided into four " battles," under (i) Jamc; 
d'Avesnes, (2) the Count of Dreux and his brother the Bishoj 
of Beauvais, (3) William des Barres and William de Garlande 
(4) Drogo Count of Merle. Henry Count of Champagne wa 
charged with the duty of keeping out on the left flank to watcl 
for the breaking forth of the Turks from the woods. The Dukt 
of Burgundy, the commander of the French host, rode b; 
Richard's side up and down the line, keeping order and read; 
to give aid wherever it was wanted. The whole twelve corp 
were divided into five divisions, but it is not stated how the; 
were thus distributed. Some of the five must have include< 
three, some only two, of the. brigaded bodies of horse and foot. 

Saladin allowed the whole Christian host to emerge fror 
the camp and proceed some little way along the road before h 
launched his army upon them. While threatening the whol 
of the long line of march, he had resolved to throw the mai 
weight of his attack upon the rearguard. Evidently he hope 
to produce a gap, by allowing the van and centre to proceec 
while delaying the rear by incessant assaults. If the Hospita 
lers and the divisions next them could be so harassed that the 
were forced to halt or even to charge, while the van still wer 
on its way, it was obvious that a break in the continuous wa 
of infantry would occur. Into this opening Saladin would hav 
thrown his reserves, and then have trusted to fighting the battj 
out with an enemy split into at least two fractions and probabl 
more. He had, as we shall see, wholly underrated the prudenc 
and generalship of King Richard, and was preparing for himse 
a bloody repulse. 

The Crusaders were well upon their way when the Moslen 
suddenly burst out from the woods. In front were swarms < 
skirmishers both horse and foot — black Soudanese archers, wi] 
Bedouins, and the terrible Turkish horse-bowmen. Behind we: 
visible deep squadrons of supports — the Sultan's mailed Mam< 
lukes and the contingents of all the princes and emirs of Egyf 
Syria, and Mesopotamia. The whole space, two miles broa 
between the road and the forest, was suddenly filled with the. 
imposing masses. " All over the face of the land you could sc 
the well-ordered bands of the Turks, myriads of parti-coloui 



PLATE IX. 




Battle of 

ARSOUF, 

Sept. 7. 1191. 



B ■!■ Crusaders 
r~7^ Baggage Train. 
'c^ Turkish Horse 
6 6 Skirmishei's. 

AA. Crusaders' Infantry 

and Baggage 
B.Templers 
C Angovins 
D Poitevjus 

E. English *Nonnnns 

with the Standard 

F. Hospitallers. 
CO Infantry. // 
ti.H Turkisli Skirmishers 
I.I Saladin's Main Army 



;i9i] ARSOUF: SALADIN ATTACKS RICHARD 311 

mnners, marshalled in troops and squadrons ; of mailed men 
ilone there appeared to be more than twenty thousand. With 
inswerving course, swifter than eagles, they swept down upon 
)ur line of march. The air was turned black by the dust that 
heir hoofs cast up. Before the face of each emir went his 
nusicians, making a horrid din with horns, trumpets, drums, 
;ymbals, and all manner of brazen instruments, while the troops 
)ehind pressed on with howls and cries of war. For the Infidels 
hink that the louder the noise, the bolder grows the spirit of the 
varrior. So did the cursed Turks beset us before, behind, and 
)!! the flank, and they pressed in so close that for two miles 
u-Qund there was not a spot of the bare earth visible ; all was 
:Qvered by the thick array of the enemy." ^ 

• I While some of the Turks rode in between the head of the 
irmy and its goal at Arsouf, and others followed the rearguard 
ibng the road, the majority closed in upon the left flank and 
DHed their bows against the wall of infantry and the clumps of 
' lorsemen slowly pacing behind it. The pressure seems to have 
; )een hardest upon the rear, where the right wing of the Turks 
I lelivered a most desperate attack upon the squadron of the 
I Sospitallers and the infantry corps which covered them. The 
\ French divisions opposite the Turkish centre were less hardly 
i tressed; the English, Poitevins, and Templars in the van, though 
:onstantly engaged, were never seriously incommoded. 
'1: In spite of the fury of the attack, the Crusaders for some 
;ime pursued their way without the least wavering or hesitation. 
The crossbowmen gave the Turks back bolt for bolt, and 
vrought more harm than they suffered, since their missiles were 
leavier and possessed more penetrating power than those of the 
memy. The cavalry in the centre of the column rode slowly 
)n, though their horses soon began to suffer from the incessant 
•ain of arrows. Many knights had to dismount from mortally 
vounded chargers, and to march lance in hand among the foot. 
Dthers picked up crossbows, stepped into the front rank of the 
nfantry, and revenged themselves by shooting down the Turkish 
iorses.2 

The slow march southward went on for some time; the 
nfantry held firm as a wall, and no opportunity was given for 
he enemy to break in. Saladin, seeing that he was making no 
progress, flung himself among the skirmishers, followed only by 

^ Itinerarium^ iv. § 18. - Ibid, 



312 THE ART OF WAR IN THE MIDDLE AGES [119 

two pages leading spare horses, and continued to urge his mei 
on and to press them closer in upon the Prankish foot. Th' 
stress soon became very severe in the rear division of Kin$ 
Richard's host, which was exposed to a double fire from flan) 
and rear. Some of the crossbowmen began to waver, but th 
majority held firm, forced though they were to walk backward 
with their faces to the pursuing encmy,^ for, when they turne( 
for a moment to move on, the Turks rushed in so fiercely tha 
there was grave danger that the corps of the Hospitallers migh 
be broken up. " They had laid their bows aside, and were no^ 
thundering upon the rearguard with their scimitars and mace 
like smiths upon anvils." 

The Grand Master of the Hospitallers repeatedly ser 
forward to the king, asking leave to charge. The horses wer 
being shot down one by one, he complained, and the knight 
could no longer endure this passive kind of battle, in which the 
were struck themselves, but not allowed to strike back. Richar 
returned the reply that the rear was on no account to brea 
their order : he had settled that there should be a general charg 
of the whole line when he bade six trumpets blow ; before th 
signal no one must move. His design was evidently to get th 
whole Turkish army committed to close combat before he roc 
out upon it. At present the rear alone was seriously engagec 
the van and centre were only being harassed from a distanc 
Moreover, there would be great advantage in waiting till ll 
van had reached Arsouf, whose gardens and houses would gi\ 
good cover for its flank when the moment for the decish 
charge came. 

In obedience to these orders, the Hospitallers endured f 
some time longer, but they were growing restive and angry ; 
horse after horse fell, and man after man was disabled b 
arrows in the parts of his body which the armour did not ful 
protect. Presently the whole rear division lurched forward 
disorder and joined the French corps which was marchu 
immediately in front of it. At last, just when the head of tl 
army had reached the outskirts of Arsouf, the patience of tl 
rear was wholly exhausted. Ere the king had bade the s 
trumpets sound, but (as it would seem) only just before tl 
moment that he would have chosen, the Hospitallers bui 
forth. The ringleaders in this piece of indiscipline were two 

^ Itineiariiivi, iv. § 10, p. 264. 



iQi] ARSOUP^ THE CRUSADERS CHARGE 313 

heir leaders, their marshal and a notable knight named 
Uldwin de Carron, who suddenly wheeled their horses, raised 
he war-cry of St. George, and dashed out through the infantry 
pon the Infidels. Those immediately about them followed ; 
hen the French divisions ranged next them took up the 
novement. It spread all down the line, and Richard himself, 
eeing the die cast, was constrained 'to allow the cavalry of the 
an and centre to follow up the attack. To the Saracens it 
)ore the appearance of a preconcerted movement. " On a 
udden," says Boha-ed-din, " we saw the cavalry of the enemy, 
v'ho were now drawn together in three main masses, brandish 
heir lances, raise their war-cry, and dash out at us. The 
nfantry suddenly opened up gaps in their line to let them pass 
lirough." 1 Thus the attack of the Crusaders was delivered in 
•chelon, the left {i.e. the rear) leading, the centre starting a 
noment after, and the right {i.e. the van) a little later than the 
:cntre. 

The Turks did not endure for a moment tlie onset of the 
Ireaded knights of the West. The sudden change of the 
nusading army from a passive defence to a vigorous offensive 
:ame so unexpectedly upon them, that they broke and fled with 
iisgraceful promptness. Nothing can be more frank than 
i3oha-ed-din's account of the behaviour of his master's host.^ 
'' On our side," he says, " the rout was complete. I was myself 
n the centre: that corps having fled in confusion, I thought to 
ake refuge with the left wing, which was the nearest to me ; 
xit when I reached it, I found it also in full retreat, and making 
jff no less quickly than the centre. Then I rode to the right 
vving, but this had been routed even mere thoroughly than the 
eft. I turned accordingly to the spot where the Sultan's body- 
:^uard should have served as a rallying-point for the rest. The 
oanners 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 PVench 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 

^ Boha-ed-din, p. 258, in Chroniqueurs Orieutaux. 2 j^:^^ p 259. 



314 THE ART OF WAR IN THE MIDDLE AGES [119 

they had been exposed since daybreak. Before the Moslerr 
could scatter and disperse to the rear, they had been mow 
down by thousands. In the centre and the southern end of th 
battle the Turks had an easier flight, since their pursuers wei 
not so close. Here the contact and the slaughter must ha\ 
been much less. We know from the author of the Itinerariu, 
that the English and Norman knights who formed the fourt 
division, counting from the van, never reached the flying enem; 
though they follow^ed in echelon the movement of the rear an 
centre corps.^ The same was probably the case with the oth( 
three corps of the van, for King Richard, in his letter to tl 
Abbot of Clairvaux, states that only four. of his twelve divisior 
were seriously engaged, and that these four alone really d< 
feated the whole host of Saladin.^ 

Having pursued the Turks more than a mile, the Crusade 
halted and began to re-form — there w^as no rash pursuit like th; 
which had so often ruined the Franks in earlier fields. Thoi 
of the Infidels who still kept their heads, ceased to fly wIk 
they were no longer pursued, and turned to cut off the scatterc 
knights, who had pushed far to the front, and were now ridir 
back to fall into line with their comrades. Of these some fe 
were cut off and slain — among them James d'Avesnes, 
notable knight, who had commanded one of the rear divisioi 
of the line of march. Among those of the Turks who rallic 
most quickly and came back first to the fight was Taki-ed-di 
Saladin's nephew^ with the seven hundred horsemen wl 
followed, his yellow banner. 

When the Christian line was once more in order, Richai 
led it on to a second charge ; the Turks broke again and mac 
no stand. Yet when the king cautiously halted his men, aft 
sweeping the enemy backward for another mile, there was st 
a considerable body which turned back and once more showf 
fight. A third and final charge sent them flying into the foref 
which was now ciose at their backs. Here they dispersed in t 
directions, and made no further attempt to resist. Richar 
however, would not pursue them among the thickets, and Ic 
back his horsemen at leisure to Arsouf, where the infantry he 
now pitched their camp. 

That evening many of the foot-soldiery and camp-followe 
went out to the field of battle, where they stripped the de^ 

* Itinerariiim, p. 272. ^ Letter printed in Iloveden, Rolls Series, iii. 131. 



igi] ARSOUF: RICHARD VICTORIOUS 315 

id found much valuable plunder, since the Turks, like the 
lamelukes in later days, were wont to carry their money sewed 
p in their waist-belts or under their clothing. They reported 
lat they had counted thirty-two emirs among the slain, and 
lore than seven thousand of the rank and file> Boha-ed-din 
ames as the most prominent of the Moslems who had fallen 
lousec, the prince of the Kurds, and two emirs named Kaimaz- 
, - Adeli and Ligoush.^ Among the Christians, James of 
.vesnes was the only man of distinction who was slain : their 
)tal loss was under seven hundred men. 

So ended this important and interesting fight, the most 
jmplete and typical of all the victories of the Franks over 
leir enemies. The old morals of the earlier encragfements are 
nee more repeated in it. With a judicious combination of 
orse and foot, and a proper exercise of caution, the Crusader 
light be certain of victory. But we note that Richard, though 
ew to the wars of the East, shows far more self-restraint, 
isdom, and generalship than any of his predecessors. He 
3uld have driven off Saladin at any time during the day, but 
is object was not merely to chase away the Turks for a 
loment, but to inflict on them a blow which should disable 
lem for a long period. This could only be done by luring 
lem to close combat ; hence came the passive tactics of the 
rst half of the day. The victory would have been still more 
ffective, as the author of the Itinerarium remarks,^ if the charge 
ad been delivered a little later. But the precipitate action of 
le marshal of the Hospitallers caused it to be made a moment 
arlier than the king had intended. Nevertheless, the results of 
le fight were very well marked. Saladin reassembled his army, 
ut he never dared close in upon his enemy again : he resumed 
is old policy of demonstrations and skirmishes. As Boha-ed- 
in remarks, the spirit of the Moslem army was completely 
roken. Recognising that he could not hold the open country 
gainst the Franks, the Sultan at once dismantled all the 
Drtresses of Southern Palestine -^ Ascalon, Gaza, Blanche- 
larde, Lydda, Ramleh, and the rest. He dared not leave 
arrisons in them, for he was fully aware that his men would 

'^ liiiieyariumy -p. 2'j $, p[^-j;,>i fj^r v-,/;JBoha-ed-din, p. 260. ,;^ 

^ ^ iv. 19: "Quodsi [mandatum regis] fuisset observatum, universi illi Turdi 
lissent intercept! et confusi : praedictorum vero militum nimia properatio cedebat 
1 detrimentum universi negotii " (p. 258). 



3r6 THE ART OF WAR IN THE MIDDLE AGES [ii( 

not hold firm : the fate of the defenders of Acre and the resu 
of the fight of Arsouf were always before their eyes, and th( 
would not have maintained themselves for long. How wc 
founded was this fear, became sufficiently evident from the oi 
exception which Saladin made to his rule. He left a force 
Darum, the last fortress of Palestine on the way to Egyj 
Richard made a dash against it with the knights of his ov 
household alone, a force inferior to the garrison in numb( 
Yet so half-hearted had the Moslems grown, that the kii 
stormed the place in four days. The Turks surrendered t] 
citadel on the bare promise of life, though, if they had shown 
tithe of the courage of the garrison of Acre, they would certain 
have been able to hold out for weeks, if not for months.^ 

Arsouf therefore gave the Franks the whole coast-land 
Southern Palestine. After repairing the walls of Jaffa, 
serve them as a basis for the attack on Jerusalem, they we 
free to resume the offensive. But the jealousies and divisio 
in the host ruined the campaign which had begun so brilliant 
and, though there were several gallant feats of arms perform 
during the stay of Richard in Palestine, the Holy City w 
never recovered, and the war ended in a treaty which c 
no more than confirm the Syrian Franks in the possession 
the coast-region which the English king had reconquered i 
them. 

One fight, little more than a skirmish in itself, dcser\ 
mention as illustrating Richard's methods of war. This w 
the engagement of August 5, 1192. While the king h 
returned to Acre with his army, Saladin had descended to t 
coast and endeavoured to retake the newly-fortified town 
Jaffa. The garrison had been driven into the castle, and w 
on the point of surrendering, when Richard hastily returned 
sea with eight vessels only and saved them (August i). T 
Turks were driven off for the moment, but, learning that th 
enemies were very few in number, came down at daybreak 
the 5 th of August to surprise the Christian camp. Rich^ 
had with him only fifty-five knights and two thousand infant 
the latter largely Genoese and Pisan crossbowmen drawn fn 
the ships which had brought him. Warned in time that sev 
thousand horse, all Mamelukes and Kurds, were swooping do 
upon the sleeping camp, he promptly proceeded to get his nn 

^ Tiitterarium, p. 356. 



192] COMBAT OF JAFFA 317 

1 order. He composed his front line of infantry armed with 
pears, who knelt down with one knee fixed in the sand, and 
ith the points of their weapons levelled at the height of a 
orsc's breast. Behind stood the crossbowmen, one in each 
Ucrval between two spearmen : it was this soldier's duty to 
ischarge as fast as possible the arbalests handed to him by 
nother, who stood behind him, bending and loading each as 
: was handed back. Thus there was no intermission in the 
ischarge. The Turks swept down, band rapidly following band, 
gainst the front of the Christian line, but never dared to close. 
Lach squadron swerved and passed away without daring to 
iish on the spears ; they did little harm with their arrows, but 
uffcred far more from the constant rain of arbalest bolts which 
cat upon them. When they were all in disorder, Richard boldly 
barged out upon them, though no more than fifteen of his 
nights were horsed. He cut right into their midst, and then 
icwed his way back again, saving by his personal valour the 
'^.arl of Leicester and Ralph of Maulcon, who had been sur- 
oimded and were nearly made prisoners. The fight lingered 
m for some hours after the surprise had failed, but when the 
:ing brought up some small reserves from the fleet (he left only 
ive men on each galley) the enemy fled, leaving seven hundred 
nen and fifteen hundred horses dead upon the field. Of the 
Crusaders only two men had fallen, so secure had their order 
)f battle kept them l^ 

^ All this from the excellent account in Itinerarium, vi. g§ 21-24. 



i Offi 1i, b^llo'/ol ?. 

di 'jidi? 
CHAPTER V 

THE GREAT DEFEATS OF THE CRUSADERS — CARRHAE, 
HARENC, TIBERIAS, ACRE, MANSOURAH 

HAVING now given fair typical instances of the metho 
by which the Franks won success in the interminat 
campaigns which followed the establishment of the Latin Stat 
in Syria, it remains that we should show in the same fashi( 
the manner and causes of their defeat. With those which we 
the inevitable consequences of strategical blunders we have de.' 
in our chapter on Strategy. It is with tactical errors that \ 
are now concerned. As illustration we have chosen four battl( 
Carrhae (1104) will show the result of careless pursuit and t 
neglect of the proper precautions required in Turkish warfai 
Tiberias (11 87) displays a complicated series of blunders — t 
neglect of commissariat arrangements, the choice of unsuitah 
ground, the imperfect reconnoitring of the enemy, and (mc 
important of all) the fatal results of dividing the infantry ai 
cavalry. The battle in front of Acre (i 190) proves that a victo 
practically won might be turned into a defeat by the want of 
guiding hand and neglect of the most rudimentary disciplir 
Mansourah (1250) points out that a fault originating in be 
strategy may logically lead to bad tactics, and illustrates as w( 
the normal want of discipline in all Western hosts. 

The battle of Carrhae may be taken as an example of tl 
manner in which even the most practised veterans of the fir 
Crusade could fail when they neglected obvious precautio- 
and fought on unfavourable ground. In the spring of iic 
Bohemund, now for the last six years Prince of Antioch, ar 
Baldwin of Bourg, Count of Edessa, resolved to make a bo 
push into Mesopotamia. The Turks had lately threatenc 
Edessa ; in retaliation the princes formed a project for seizii 
and garrisoning the strong town of Harran (Carrhae), the fronti 

318 



104] BATTLE OF CARRHAE 319 

lost of the Moslems. It was close enough to Edessa to be a 
roublesome neighbour, — only twenty-five miles separated the 
wo places, — while at the same time it was a favourable point 
serve as a base for further progress eastward. Baldwin 
:alled in to his aid his cousin Joscelin, to whom he had granted 
. great lordship west of the Euphrates, round the town of 
furbesel. Bohemund brought with him his kinsman Tancred, 
he hero of so many exploits in the first Crusade. The oppor- 
unity seemed fair, for by systematic ravagings Baldwin had 
iiined the countryside round Carrhae, and knew that the place 
vas straitened for provisions. Moreover, the two Turkish 
)rinces who ruled in Mesopotamia, the Atabeg Sokman ibn- 
Jrtuk of Kayfa, and Jekermisch the successor of Kerboga in 
he emirate of Mosul, were engaged in bitter strife with each 
)ther. 

At the head of what passed for a considerable army among 
he Syrian Franks, the allied princes marched on Carrhae and 
brmed the siege. The place, as Baldwin had known, was ill 
stored, and ere long the famished citizens began to treat for a 
iurrender. But while the terms were being disputed, a relieving 
irmy came in sight : Sokman and Jekermisch had come to 
erms in face of the common danger, and had combined their 
brces to save Carrhae. The former brought to the field seven 
housand Turkish horse-archers ; the latter, three thousand Kurds, 
bedouins, and Turks.- They had resolved to threaten an attack 
)n the Christian camp, and to throw a convoy into the city 
vhile the besiegers' attention was distracted. Their success 
vas far greater than they could have hoped : when the Franks 
)aw them, they formed in three "battles," each composed of 
lorse and foot, and marched out to attack them. Bohemund 
leld the right, Tancred the centre, Baldwin and Joscelin the 
eft, in the Christian host. When the Franks advanced, the 
Turkish princes applied the ordinary stratagems of their race : 
:hey retreated into the broad plain eastward of the city, 
harassing the advancing enemy with their arrows. Old 
ioldiers Hke Bohemund and Baldwin should have known better 
iiow 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 [no 

infantry were tired to death, and, when the afternoon was h 
spent, Bohemund at last gav^e the signal to halt, and orderc 
his host to encamp for the night, not dreaming that the encm 
was likely to suddenly take the offensive. This was the momer 
for which the Turks had been waiting. When they saw tt 
Franks falling out of line, dism^ounting, and taking off the 
arms, they suddenly came charging in with loud shouts an 
dashed among their enemies, using the sword as well as tl 
arrow. Baldwin's division was caught wholly unprepared, an 
ridden down before it had time to re-form ; both he and h 
cousin Joscelin were taken prisoners, and with them Bencdic 
Archbishop of Edessa. The camp and all its stores fell in1 
the hands of the Turks. Tancred, more cautious than tl 
Count of Edessa, had not allowed his men to disperse, and wj 
able to rally them and form up on a hill a mile or two behir 
the camp ; here Bohemund joined him with the main body > 
the Christian right wing, which had been disordered, but ni 
wholly destroyed.^ The two princes waited to be attacked, b' 
the Turks only demonstrated against them ; they had no inte; 
tion of closing, and were well satisfied with their partial victor 
and eager to share the plunder they had taken. When nig! 
fell, the Franks found themselves in evil plight : they had lo 
not only their camp, but all their provisions ; horses and mc 
alike were famished and exhausted after the long day's m.an 
in the sandy plain. Nevertheless, the princes resolved to rcne 
the combat next morning, and bade the starving army prepa 
for a second battle. But the Franks were demoralised : und 
cover of the darkness their foot-soldiery melted away towan 
the fords of the Chobar, drove off the guard which had be( 
placed there to stop desertion, and made off towards Edess 
When the flight of the greater part of the infantry was observe 
many knights stole away after them, and Bohemund ar 
Tancred ultimately found themselves deserted by all save tl 
men of their own military household. It was impossible 
await the dawn and the Turkish advance, so the princes followc 
their panic-stricken host towards the ford. It was fortuna 
that the enemy kept a bad watch, or the whole Christian arn 
might have been destroyed in detail. But the Turks we 

^ So Ralph of Caen ; the Arab Ibn-Ghiouzl says' that Tancred was at soi 
distance from Baldwin, on the other side of a hill, and that the Count of Edcs?a v 
routed before his ally could come up to help him. 



104] BATTLE OF CARRHAE 321 

! pending the night in a hot dispute ; Sokman's men had been 
lundering the Prankish camp while Jekermisch's troops had 
een observing Tancred's rallied division. On their return at 
usk, the Mosulite horsemen demanded their share of the prey, 
nd Jekermisch seized the person of Baldwin, the chief of the 
aptives, who had been placed in Sokman's tent. The Turks o 
layfa drew their swords to resent this insult to their master, 
nd a general combat would have followed had not Sokman 
ucceeded in appeasing his men, and at the same time bought 
ff Jekermisch by a promise to divide the spoil fairly.^ 

Meanwhile, the Christians got a long start, and were all over 
le river and straggling back towards Edessa before the day 
awned. They were, of course, pursued the moment that their 
eparture was ascertained, and many stragglers were cut off; 
le main body, however, reached the city in safety.^ But the 
low had been a heavy one : more than half the army was 
lissing,^ and the Christians were thrown upon the defensive 
Dr some years. It is astonishing that the Turks did not 
lake more profit from their victory, but, after besieging Edessa 
1 vain for fifteen days, they dispersed and returned to their 
omes. 

It is strange to find that the Crusaders were routed on the 
ame field where the younger Crassus and his fifteen hundred 
rallic horsemen were cut to pieces by the Parthian archers be- 
Dre the eyes of his father the Triumvir nearly eleven centuries 
■efore. That cavalry from the far West armed with the lance 
hould strive again on that sandy plain with the Turanian horse- 
owmen, and should succumb again, was one of the most curious 
oincidences of history. The march of the Triumvir and his 
igions among the evasive Parthian s suggests somewhat the 
dvance of Baldwin and Bohemund, but the Roman was worse 

1 Ibn-Alathir says (see Michaud, Bibliotheque des Croisades, iv. 19) that Sokman 
Kclaimed, "Islam will have no joy from this victory if we quarrel after it. I will 
ither lose my spoil than let the Christians taunt us with folly." 

^ See in Ralph of Caen, 281, 282, the story of the flight, especially the comic tale 
f Archbishop Bernard, who, " when no one was pressing, thought he had behind him 
osts of Turks with bended bows and drawn swords," and cut off his palfrey's tail to 
ee the faster. 

^ Ibn-Alathir no doubt exaggerates when he says that twelve thousand Franks 
■ere slain or taken, and that Tancred got away with six knights only. But the 
nportance of the disaster is vouched for by William of Tyre's statement that "in 
o battle of the East down to our own day were so many strong and valiant men slain, 
^d never did a Christian army fly so shamefully" (x. 110). 



322 THE ART OF WAR IN THE MIDDLE AGES [ii8 

off than the Franks. He was fighting, as it were, blindfok 
against a foe whose tactics were wholly unknown to him : whil 
the veterans of Dorylaeum and Antioch were experienced i 
Turkish wiles, and ought never to have been caught unpreparec 
Their failure to observe common precautions was all the mor 
inexcusable, and if their host got off more cheaply than th 
unfortunate followers of the two Crassi, it was by good luck an 
not by their deserts.^ 

Battle of Tiberias^ July 4, 1 187. 

Disastrous as was the battle of Carrhae, it cannot compai 
either in its scale or in consequences with the great fight eight 
years later which gave Jerusalem to the Infidel. The battle ( 
1 104 did not even destroy the single principality of Edessf 
that of 1 187 was the great turning-point in the whole history < 
the Crusades, since it entirely deprived the Crusaders of the 
hold on inner Syria, and left them for the future masters < 
nothing more than a narrow strip of coast-land. 

In 1 187 Saladin, after having cut short the borders of tl 
Christians in many quarters, resolved to risk an attack on tl 
centre of their strength, by a direct invasion of the kingdom ( 
Jerusalem. He first despatched a considerable force to execui 
a raid into its northern parts : it was put in charge of Modhaffe 
ed-din. Prince of Edessa and Haran, who crossed the Jorda 
harried the hill-country of Galilee, and cut to pieces at the blood 
encounter of Saffaria (May i) the knights of the Temple an 
the Hospital, who had come forth against him with more ze 
than discretion, before any succours could reach them. His sa 
return emboldened the Sultan to ride forth in person. 

In June he gathered all his disposable forces from Egyp 
Syria, and Mesopotamia at Ashtera in the Hauran. There we 
ten thousand mailed Mamelukes of his regular army, beside tl 
innumerable contingents of his provinces : the total may ha^ 
amounted to some sixty or seventy thousand men. On Jui 
26 he led them down to the vicinity of the Jordan, ar 
encamped at Sennabra, close to the bridge of El-Kantara, whic \ 
crosses the river a mile south of the point where it issues fro 



^ We find that there were men in Latin Syria learned enough to observe t 
coincidence. William of Tyre remarks that ' ' this was that same Carrhae wht 
Crassus the Dictator (!) had his celebrated mouthful of the Parthian gold for wbi 
he had been so greedy" (W. T. Ijook x.). 



fi87] SALADIN INVADES GALILEE 323 

he Sea of Galilee. Three days later he passed the stream and 
idvanced into Christian territory. His first aim was to capture 
he town of Tiberias, the capital of the principality of Galilee. 
Posting his main army on the hills east of that place, he sent a 
:orps to lay siege to it. The town yielded with unexpected ease, 
)ut the garrison and their mistress, the Countess of Tripoli, with- 
Irew into the castle, a strong fort overhanging the water, which 
vas capable of holding out for many weeks. 

Meanwhile, the Christians were assembling in great strength. 
Vlodhaffer-ed-din's raid had seriously disturbed them, and, when 
hey heard that Saladin was concentrating his army in the 
:Iauran, they had resolved to draw together in full force. King 
juy summoned in all his barons and knights ; the military Orders 
)ut all their available men into the field, thinned though their 
anks had been by the disaster at Saffaria. The towns sent 
:ontingents even larger than they were bound to furnish. The 
Tount of Tripoli, who had onl}' lately reconciled himself to his 
uzerain, did his best to atone for past disloyalty by bringing 
he full levy of his county to the muster. The True Cross was 
etched out from the Church of the Holy Sepulchre and sent to 
he front, in charge of the Bishop of Lydda. The castles and 
ities of Southern Palestine were left with garrisons of danger- 
)usly small numbers. By this concentration, the Franks were 
.ble to assemble twelve hundred knights, many hundred 
furcopoles or mounted bowmen equipped in the Turkish fashion, 
ind eighteen thousand foot,^ the largest force that they had ever 
)ut into the field save that which had been mustered for the 
ibortive campaign of 1184.^ Their meeting-place was the village 
)f Saffaria — the spot where the disaster to the Templars had 
)ccurred seven weeks before. It lies in a well-watered upland 
alley, three miles north of Nazareth and seventeen east of Acre. 
^rom thence to Tiberias is sixteen miles, by a road passing 
Lcross one of the most desolate and thinly-peopled districts in 
he Galilean hills.^ The time was the hottest month of the 
ummer, and Saladin's raiders had burned the villages and 
lestroyed the wells all around. They had even defiled the 
Zhurch of the Transfiguration on the summit of Mount Tabor. 

^ So Ralph of Coggeshall, the best authority for the campaign, p. 218. 
2 On that occasion they had raised what WilHani of Tyre calls the largest host ever 
een in the kingdom (xxii. p. 448), 

' There are only two small villages, Toron ajid Lubieh, on the road. 



324 THE ART OF WAR IN THE MIDDLE AGES [ii8 

There was long talk and hot disputing at Saffaria as t 
whether the army should march to the relief of the castle c 
Tiberias. The Franks had mustered in such full force that the 
could never hope to raise a larger army. Saladin had place 
himself in a position where defeat would mean ruin, since h 
had the broad Sea of Galilee at his back, and his retreat eithe 
north or south would be through difficult and dangerous defile; 
On the other hand, it was hazardous to risk the whole resource 
of the kingdom in a single fight. If the army at Saffaria ws 
beaten, there were no reserves left on which it could fall back. 

The Count of Tripoli, the most experienced warrior in th 
Christian host, took the side of caution. He pointed out that 
they did not march against Saladin, the Sultan would be force 
to march against them, since he could not long abide in th 
desolate country round Tiberias. His only other alternativ 
would be to return to Damascus, a course which he certainl 
would not consent to take when his pride had risen so hig 
and when his army was so strong. It would suit the policy ( 
the Christians to be attacked at Saffaria, where they had a goo 
position, plenty of food, and an ample supply of water. Saladii 
on the other hand, would arrive with an army tired out by 
fatiguing march and discouraged by the distance from its base 
for the Turks must fight, knowing that they had no shelte 
nearer than Damascus, and with the lake and the Jordan i 
their backs. Raymond added that he, if anyone, should fei 
interested in the preservation of Tiberias, since his own wil 
and children were being beleaguered in the citadel ; nevertheles 
he advised that a waiting policy should be adopted, and tl 
responsibility of the initiative thrown on the enemy. If tl . 
Christian army marched over the mountains, it would have 1 
fight when worn out by thirst and heat ; it was far better th< 
the Infidels should have these disadvantages on the day < 
battle.i 

Unfortunately the advice of Raymond was ill receive 
His enemies whispered that he was the king's enemy, and thi 
his cowardly counsel was that of a deliberate traitor. TI 
majority of the barons voted that it would be shameful t 
abandon the garrison of Tiberias. The king assented, and c 
Thursday, July 3, the army marched out from Saffaria light! 

^ Ralph of Coggeshall, p. 222, here agrees wonderfully well with the Mosle 
chronicler, Ibn-Alathir. 



FiSy] THE FRANKS MARCH ON TIBERIAS 325 

equipped, and leaving all its impedimenta behind in the camp. 
The order of march is not very clearly stated ; but we know 
hat 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 
old off to the defence of the True Cross, was in the middle. 
rlow many divisions the whole army contained we are not 
old, nor is it explicitly stated that each consisted of horse and 
bot combined, though this must almost certainly have been 
he case. 

The Franks had marched about nine or ten miles, when 
hey began to be surrounded by swarms of Turkish skirmishers. 
Saladin did not display his main force, but enveloped their army 
vith a cloud of horse-bowmen, whose orders were to make the 
narch slow and painful. By the time that the host drew near 
he deserted village of Marescalcia,^ it was terribly weary and 
larassed. Only some six miles now separated it from the 
own of Tiberias and the lake.^ The van, which had pushed 
lown into the lower ground and was still advancing, was within 
hree miles of the water. But between the weary Crusaders 
md their goal lay the hills of Tiberias, a range rising to about 
)ne thousand feet above sea level : the northern point, Kurn- 
^^attin, is eleven hundred and ninety-one feet high. Behind the 
:rest of these hills the ground falls suddenly towards the deep- 
unk hollow of the Sea of Galilee. Tiberias itself is no less than 
ix hundred and fifty-three feet below the level of the Mediter- 
anean. All along the range the Turks were arrayed, and it 
vas necessary for the army to cut its way through them by 
)ne of the two passes which cross at its lowest points — the 
lepressions called the Wady-el-Muallakah and the Wady-el- 
iammam. 

Tired as the army was, there was an absolute necessity that 
t should push on, for there was no water available for three 
niles around, and men and horses were already perishing of 
hirst. The Count of Tripoli sent back to King Guy, begging 
lim to hasten the advance at all costs, as the day was drawing 
)n, and the lake must be reached ere nightfall if the army was 

^ Probably the modern Lubieh. 

- The distance is grossly understated in Coggeshall, who says that there were 
■nly three miles between Marescalcia and the lake, and that the van under the 
'cunt 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 
Wady-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."^ 

That night the Franks camped, huddled together arounc 
the royal standard on the hillside. There was little food anc 
hardly a drop of water in the host : even sleep was impossible 
for the Turks came close in under cover of the darkness, anc 
kept up a constant shower of arrows into the camp. They alsc 
fired the dry grass to windward of the Crusaders, so thai 
stifling clouds of smoke were drifting over it all night. *' Goc 
fed the Christians with the bread of tears, and gave them t( 
drink without stint of the cup of repentance, till the dawn o 
tribulation came again." ^ The Saracens were not much more 
easy in their minds than their enemies : with the lake at thei; 
back and the formidable Christian army still intact, they hac 
many qualms of spirit when the fight was renewed on th( 
morning of Friday, June 4.^ 

King Guy had once more ranged his army in order, with the 
same divisions as he had drawn up on the previous day — th( 
Count of Tripoli in front, the military Orders in the rear 
Swerving from his original route, he now ordered the march tf 

^ Ralph of Coggeshall, p. 223. * /l>u/. p. 224. 

^ Boha-ed-din, p. 94. 



PLATE X. 




OF 



LI LEE 

650 ft below Sea Level.) 



BATTLE OF ACRE. 
Oct. 4^." 1 1 89. 



Christians: 



A. I&i^ Guy B. Conrad of Montferrat. 

C. Lewis of Thurin^ia. D Master of the Templars . 




[iSy] TIBERIAS: DESTRUCTION OF THE INFANTRY 327 

)e directed towards the Wady-el-Hammam and the village of 
iiattin, aiming at the nearest water, and no longer taking the 
hortest way to Tiberias. Saladin had now brought up his 
vhole host, which encircled the Christians on all sides, though 
he thickest mass lay across the road to the lake. The 
Jrusaders moved forward for some distance, and were about to 
oin in close combat, when the king detected great unsteadiness 
n his infantry. They had been told off to the various corps of 
:avalry, and were directed to form line in front of them, " that 
he two arms might give each other the proper support, the 
^-nights protected by the arrows of the foot, and the foot by the 
ances of the knights." ^ At the moment of close combat, 
lowever, the greater part of the infantry, after wavering for a 
noment, shrank together into one great mass, and, swerving off 
:he road to the right, climbed a hill (probably Kurn-Hattin) 
\vhich lay to that flank, and formed in a dense clump on its 
summit, deserting the horsemen on the road below.^ The king 
sent messenger after messenger to them, imploring them to 
:ome down and play their part in the battle. The only answer 
\vhich 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 
tenis. So the battles broke up, and all huddled together in a 
confused mass around the True Cross." ^ 

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 

^ Ralph of Coggeshall, p. 224. 

^ " Conglobati sunt in unutn cuneum, et veloci cursu cacumen excelsi montes, 
relinquentes exercitum, malo suo ascenderunt" {ibid. p. 225). 
^ Ibid. p. 225. 



328 THE ART OF WAR IN THE MIDDLE AGES [iH 

their way through the Moslems and seek refuge in flight. " Th 
battle is hopelessly lost ; let every man save himself if he can," 
cried Raymond, and, forming his corps in a close body, h 
charged the Turks immediately in front of him, aiming n 
longer at the lake, but at the hills to the north-west. Hi 
desperate assault burst right through the circle of horse-archer: 
and he, with his comrades, Balian of Nablous and Reginald c 
Sidon, and the whole of their retainers, got safely away to th 
north. The Moslem chroniclers say that Saladin's nephev 
Taki-ed-din, who commanded in this part of the field, made n 
serious effort to check or pursue them, because he judged tha 
it would be more profitable to let them go, — for their departur 
enfeebled the Christian army by a third, and left the remainde 
a more certain prey to Saladin. It is permissible to suspec 
that the plea was an afterthought, and that the Turks were i 
truth cowed by the sudden charge of these desperate men. 

Meanwhile, all had gone to ruin in the rear. While on 
swarm of Moslem horse beset the confused mass of knight 
huddled together around the king's banner and the True Cros 
the rest turned to assault the infantry. The wretched fugitive 
on the hill were too exhausted to offer any real resistance. Th 
first charge of the enemy split up their ill-compacted ranks 
some were ridden down, some were cast by the impact ovc 
the cliff at the back of the hill, and met their death in the fal 
The majority threw down lance and arbalest and held out the 
hands to the conquerors. The Turks slew many, and accepte 
the rest as captives. 

The fate of the king and his knights was no less disma 
They held out for a long time, though neither victory nc 
retreat was any longer possible. Encompassed on all sides b 
the dense swarm of Turks, they could only stand to be she 
down. At last, though their horses were reduced to the la 
pitch of fatigue, and though they themselves had drunk the 
last drops of water on the previous night, the whole or part ( 
the host resolved to make one more push for liberty. The 
might perhaps cut their way through to safety, as the Coui 
of Tripoli had done a few hours before. A Mohammeda 
chronicler^ has preserved a good account of this last charg 

^ "Qui potest transire transeat, quoniam non est nobis praelium." A perfe 
mediaeval rendering of " Sauve qui peut." (Ralph of Coggeshall, p. 225.) 
2 Ibn-Alathir. 



1187] TIBERIAS: THE FRANKS SURRENDER 329 

vhich he drew from the memory of an eye-witness, Saladin's 
;on, 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 
resetting the Crusaders. 

" When the king of the Franks and his knights," said Malek- 
-Afdal, " found themselves pressed together on a hillock on 
he side of the hill of Kurn-Hattin, I was with my father. I 
aw the Franks make a gallant charge at those of the Moslems 
vho were nearest them, and drive them back close to the spot 
vhere we stood. I looked at my father and saw that he was 
leeply moved ; he changed colour, grasped his beard in his 
land, and moved forward crying, * Let us prove the devil a 
iar ! ' ^ At these words our men precipitated themselves upon 
he Franks, and drove them back up the hillside. I began 
nyself to be overjoyed, and to cry, * They fly ! they fly ! ' But 
he enemy presently came back to the charge, and for a second 
ime cut their way to the foot of the hill ; when they were 
.gain driven back, I began to cry afresh, ' They fly ! they fly ! ' 
Then my father looked at me and said, ' Hold your tongue, and 
lo not say that they are really routed till you see the king's 
ent fall.' Shortly after we saw the tent come down ; then my 
ather dismounted, prostrated himself to the earth in thanks to 
Tod, and wept tears of joy." 

When the second attempt to pierce the Moslem circle had 
liled, and all hope was gone, we are told that in their despair 
[le Franks dropped from their exhausted horses, cast down 
lieir lances, and threw themselves sullenly upon the ground. 
'he Turks ran in upon them and took them captives without 
nother blow being struck. To their great surprise, they found 
hat very few of the knights were seriously hurt ; their mail- 
hirts had protected them so well from the arrow-shower that 
;w were badly wounded and hardly any slain. Thirst and 
xhaustion had brought them down, rather than the shafts or 
-imitars of the conquerors. On the other hand, there was 
ardly a horse that was not sorely hurt, and not one that could 
;ave carried his rider out of the battle. The poor beasts were 
'tterly worn out by two days' deprivation of water and forage. 

In the corps which thus surrendered with the king were all 

^ Meaning, I suppose, that as God had promised victory to the True Believers, 
iy thought of defeat must b^ an inspiration from Satan. 



330 THE ART OF WAR IN THE MIDDLE AGES [ir8 

the great barons of Palestine save those who had got off ii 
company with the Count of Tripoh*. They included the king' 
brother Amaury, Constable of Jerusalem, the Marquis o 
Montferrat/ Joscelin, titular Count of Edessa, Reginald o 
Chatillon, lord of Kerak and Montreal^ Humphrey of Toror 
Hugh of Tiberias,^ Hugh of Giblet, the Bishop of Lydda, th 
Master of the Hospitallers, and many scores of knights o 
wealth and name. Few persons of any note had fallen — th 
Bishop of Acre, who had borne the Holy Cross throughout th 
battle, is the only magnate reckoned among the dead. 

That evening Saladin held a review of the prisoners. H 
kindly entreated King Guy and most of the barons, but h 
called out and slew with his own hand Reginald of Chatilloi 
who had earned his hate by breakir^g a truce and by plunderin 
some pilgrims to Mecca who had passed by his castle f 
Montreal. He also bade his bodyguard slay off-hand all th 
knights of the Temple and Hospital who had fallen into h 
hands. Not content with this, he proclaimed throughout h 
host that any private soldier who had captured any member ( 
the military Orders must give him up. For each knight s 
surrendered he paid the captor fifty dinars, and then sent tl 
prisoner to join his cpmrades in death. More than two hundrc 
Templars and Hospitallers were thus slain in cold bloo 
Saladin looked upon them as the professed and profession 
enemies of his faith, and never gave them quarter. When v 
remember that he had committed such atrocities, we need n 
blame too bitterly misdeeds on the other side such as Coeur ( 
Lion's massacre of the garrison of Acre. 

Few victories have brought in their train mere importa 
results than that of Tiberias : within a few months the whole 
the kingdom of Jerusalem save a few coast-fortresses was in t 
hands of Saladin. The realm had been drained dry of men 
supply the army which perished on the hillside of Hattin, ai 
its towns and castles fell helplessly before the Moslem for she 
lack of defenders. Places that had braved the assaults oft 
Infidel for eighty years opened their gates at the first summoi 

^ Boniface, father of the more celebrated Conrad of Montferrat, who figure.' 
the third Crusade. 

2 I suppose that the "son of the Lady of Tiberias," named by Boha-ed-dir 
this Hugh, eldest son of the lady, who had by now married as second husb; , 
Raymond of Tripoli. 



i88] THE SIEGE OF ACRE BEGUN 331 

)ecause there were none but clerks and women left within them, 
erusalem itself surrendered after a siege of only twelve days. 
\ few remote castles like Kerak and Montreal had been left 
>etter garrisoned, because they lay in the extreme limit of the 
:ingdom, and some of these held out till 1 188. Montreal, endur- 
ng the extremities of famine, did not surrender till May 1 1 89. 
]ut in the main body of the realm. Tyre, whither the sad 
urvivors of Tiberias had retired, was the only stronghold of 
:rst-rate importance which remained in Christian hands. 

Such were the consequences of the overhaste of King Guy, 
.nd of his determination to cut his way to the relief of Tiberias 
v'ithout having taken account of the character of the country- 
ide in which he was to fight. We may safely say that if he had 
aken more care about supplies, and especially about his provision 
f water, and had carefully planned out his itinerary, he might 
lave reached his goal. The Saracens were in a very uncomfort- 
ble position, with the lake at their backs and no place of refuge 
lear ; one more such push as the Count of Tripoli had advised 
n the evening of the first day would probably have led to 
heir withdrawal. But a much more easy alternative would 
ave been to have encamped in some well-watered spot, such as 
^affaria, and awaited the retreat of Saladin. The Sultan must 
ave soon retired for want of provender (and especially of 
odder) in the wasted country about Tiberias, and he could not 
ave dared to disperse his army for foraging purposes in the 
ice of the Christian host, while it remained intact and con- 
entrated in front of him. The whole battle, therefore, was 
nnecessary, and the details of Guy's bad generalship are 
omparatively small blunders when compared with the enormous 
litial mistake of fighting at all. 

Battle of Acre, October 4, 1 189. 

When, only two years after the fatal day of Tiberias, we once 

^':^re find the Christians capable of contending on equal terms 

!i Saladin, it is of course due to the arrival of reinforcements 

;n the West. The exhausted remnant of the Syrian Franks 

ould have done nothing. When King Guy was freed from 

aptivity in 1188, and set himself to gather forces for the 

ecovery of some foothold in his lost realm, it took him a year 

3 collect seven hundred knights and nine thousand foot, and 

hese were not for the most part his own vassals (though Tripoli 



332 THE ART OF WAR IN THE MIDDLE AGES [ii8< 

and Antioch lent him some succour), but early arrivals fron 
among the men of the West who had taken the Cross when th 
news of the capture of Jerusalem reached Europe. Guy wa 
not even in possession of Tyre, the one important city of hi 
realm which still remained in Christian hands. His rival an< 
brother-in-law, Conrad of Montferrat, shut its gates and refuse^ 
to admit him. 

It was, therefore, an act of no small daring when, on Augus 
28, 1 1 89, Guy and his little army boldly challenged the powe 
of Saladin by marching on Acre and encamping before i1 
walls. The siege began as a blockade and nothing more, fc 
the Turks were able to pass in and out of the place at wil 
But gradually the crusading contingents began to drop in on 
by one from the West, and, less than a month after the sieg 
began, nearly forty thousand men were assembled around Acr 
On September 14 they engaged in a bloody and indecisi\ 
fight with a relieving army which Saladin in person had led t 
the succour of the garrison ; the Sultan succeeded in throwin 
a large convoy into the city, but failed in his design of drivir 
off the besiegers. This encouraged the Crusaders, who; 
numbers were still growing every day, to attempt a counte 
stroke. They first completed the investment of Acre by extent 
ing their pickets from sea to sea across the neck of land ( 
which the city stands. Then, after having shut off the garrisc 
from the army without, they resolved to offer battle in the op( 
by marching upon the Sultan's camp. 

The crusading host lay in a semicircle round Acre, wi 
the king's pavilion pitched on " Mount Turon " (Tel-el-Fokha 
a low hill ninety feet high, which lies about fourteen hundn 
yards from the walls. The Turkish army formed a much larg 
semicircle, separated from the Franks by an interval of abo 
two miles. Its central rallying-point was the hill of Ayadie 
rising two hundred and fifty feet above the plain : here Salad 
himself lay. His subordinates stretched out to right and le 
watching the whole of the plain from the river Belus (Nahr-e 
Namein) on the south to the sea on the north. That the armi 
engaged were really very large, and that the chroniclers for on 
cannot be very far mistaken in the numbers that they give, 
best shown by the fact that the length of the Prankish lir 
must have been more than two miles, and the front covered 
the Sultan's host no less than three miles. 



189] ACRE : THE FRANKS ADVANCE 333 

Descending from Mount Turon into the plain of Arab-el- 
ihawarneh, which stretches away to the foot of the hill of 
Vyadieh, the Crusaders formed themselves in four corps. The 
rst (counting from the right) was commanded by King Guy, 
nd consisted of the Hospitallers, the king's own following, and 
he French Crusaders under the Count of Dreux and the Bishop 
f Beauvais. In the second corps were the Archbishop of 
vavenna and Conrad of Montferrat, with the greater part of the 
talian Crusaders and such of the barons of Palestine as adhered 
3 Conrad ^ in his feud with King Guy. In the third was Lewis, 
.andgrave of Thuringia, with the greater part of the German 
ontingents and the Pisans under their archbishop. In the 
)urth marched the Templars, under their Master, Gerard of 
Lideford, the Counts of Bar and Brienne with the Crusaders 
om Champagne ^ and the smaller part of the Germans. 

Geoffrey of Lusignan, the king's brother, and James of 
u'esnes remained behind in the camp with a reserve.^ They 
ad to watch the city, whose investment had to be relaxed when 
le army took the field. Apparently the space from Mount 
uron northward to the sea was no longer observed, nearly a 
lile being left open ; only the eastern face of the wall was 
Dvered by the camp, the northern face was free. 

In each of the four marching divisions of the Christian host 
le proper disposition of horse and foot was carried out. The 
owmen and arbalesters formed a long continuous first line : 
ehind them marched the knights in close order. The whole 
est fronted north-east, and set its face towards the Sultan's tent, 
lainly visible on the hill of Ayadieh. The line looked very 
)rmidable and strong : the chroniclers give its numbers at 
)ur thousand horse and a hundred thousand foot — figures 
om which some deductions may be made. 

On seeing the Christians moving forth from their camp, 
ialadin had promptly drawn up his host in front of them. The 
rmy reached from the sea to the Belus, with a semicircular 
ont of more than three miles : the centre was somewhat refused, 
le wings somewhat thrown forward. The array of the various 

^ Conrad had been temporarily reconciled to King Guy, and had lately come to 
;lp 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, 
enry of Champagne himself came later to Acre, but the Counts of Bar and Brienne, 
>th Champenois, were already in the field. 

^ Probably Syrian Franks and Netherlanders. 



i 



334 THE ART QF WAR IN THE MIDDLE AGES [iic 

contingents is elaborately set forth by Boha-ed-din : to the sout 
next the river, were the garrison of Egypt, the old troops 
Shirkuh ; next to them were the followers of Modhaffer-ed-di 
lord of Edessa and Haran ; beyond these the contingent 
Sinjar in Mesopotamia ; next were the whole of the tribes 
Kurdistan, under their great emir, Mashtoub. These fo 
corps formed the left wing. The centre consisted of the Sultar 
bodyguard and the Mesopotamian troops from Diarbekr, Mosi 
and Hisn-Kayfa. The Sultan himself, his son Malek-el-Afd 
and his nephew Malek-ed-Dafer, were here in commar 
The right wing, which lay towards the sea, was composed of t 
Syrian contingents, headed by Saladin's nephew Taki-ed-d 
Prince of Emesa. 

When the Christians began to advance into the plain, th 
soon found that the intervals between the four corps in their li 
of battle were growing greater. This was necessarily the Cc 
when they marched out from a comparatively narrow positi 
into a wide plain whose whole breadth was held by the enen 
When they began to extend their front to make it equal to tl 
of the Turks, each step farther forward brought about a wic 
separation between the centre and the wings. This was 
disastrous fact for the Franks, whose main chance of victory ] 
in their being able to keep a well closed-up line. In the act" 
fighting, as we shall see, this was so far from being the case tl 
three separate engagements were fought by the left wing, i 
right wing, and the two centre divisions. 

The first contact occurred in the northern part of the fit 
where the Master of the Templars faced the Syrian continge 
of Taki-ed-din. After a few minutes the Infidels began 
give ground : Boha-ed-din assures us that the movement \ 
voluntary, and that the Frince of Emesa was desirous of draw 
away the Christian left wing from the main body by his retr( 
Whether this was so or not, it is at any rate certain that Salac 
seeing his right wing retiring, sent to its aid heavy reinforceme 
from his centre. These succours enabled the Syrians to reti 
the offensive, and the Templars had to re-form their line on a 
lying toward the sea (probably the rising ground now known 
Kisr-el-Hammar). Here the battle stood still for some ti 
without marked success on one side or the other. 

Meanwhile, Saladin's despatch of troops from his cer 
towards his right had been observed by the Franks, and 



1189] ACRE: THE FRANKS ROUTED 335 

wo central divisions of the Christian host, led by Conrad of 
Montferrat and Lewis of Thuringia, delivered a fierce assault 
)n the Sultan's main body. They marched at a moderate pace 
vith the infantry in front shooting hard, till they came in 
:ontact with the Mesopotamian troops from Diarbekr and Mosul. 
iVhen the lines closed, the knights passed through intervals 
)pened out for them by the foot-soldiery, and crashed into the 
Turkish ranks. The Infidels could not stand the shock : their 
ine was broken, and they fled in wild confusion toward their 
amp on the hill of Ayadieh. Saladin could not rally them, and 
riany of the fugitives were so panic-stricken that they rode 
without drawing rein as far as Tiberias, or even Damascus, 
'ollowing the routed Turks, the two divisions of the Prankish 
entre stormed up the hill and plunged into the camp. It would 
lave been hard to keep them in order among the tents and 
ther obstacles which broke their line ; but, as a matter of fact, 
10 one made any attempt to restrain them. Horse and foot 
cattered themselves through the encampment and turned, some 
slaughter and some to plunder. The Sultan's own pavilion 
ras sacked and cast down, three of his body servants being slain 
herein. Some of the Franks turned to cutting down the camp- 
oUowers, others burst into the sutlers' quarter and plundered 
he market. No one made any attempt to prevent the routed 
Turks from rallying, or to take in flank the still intact wings of 
Paladin's army. 

Meanwhile, King Guy and the right wing of the Franks 
eem not to have come to a decisive engagement with the 
vurds and Mamelukes of Saladin's left. Neither Western nor 
Eastern writers give any clear account of the movements in this 
lart of the field. It seems likely, however, from a passage in 
br-Alathir, that the Moslems were somewhat outflanking the 
Christians, since the latter had partly followed the advance of 
heir centre. Lest the enemy might use the opportunity and 
et between him and the camp, the king may probably have 
leld back. 

By the most untiring personal exertions Saladin at last 
ucceeded in gathering together a great part of his routed centre 
omewhere at the western foot of the Ayadieh hill. His officers 
)esought him to lead them to storm their lost camp, but he 
efused, and bade them wait till the Franks should leave it, and 
hen to charge them when their backs were turned to the 



336 THE ART OF WAR IN THE MIDDLE AGES [ii8f 

Moslems. Before long the Germans and Lombards began t( 
evacuate the hill, some burdened with plunder, others wishing t( 
re-form on the open ground and then to go to the help of th( 
king or the Templars. The retreat was executed in great dis 
order, and not without panic : many thought that some disaste 
had happened in the rear to account for the fact that thei 
comrades were tramping down hill. The author of the Itmerariuf. 
tells us that in one part of the field a knot of Germans, runnim 
to catch an Arab horse which had broken loose, were suppose- 
by the rest to be flying, and caused a senseless rush to the rear 

When the Christians were trooping in disorderly masses bac 
to the plain, Saladin suddenly let loose his rallied Mesopotamia 
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 wei 
hurried back across the plain by the shock of the Turkis 
horsemen. In utter rout some fled toward King Guy's corp 
some straight to the camp. Saladin followed, slaying the hinc 
most and easily driving all before him. The crusading rigl 
wing seems to have made some attempt to rescue the fugitive 
and Guy himself is said to have saved the life of his old enem 
Conrad of Montferrat, by hewing out a passage for him when 1 
had been surrounded by the pursuers.^ But the king and tl 
Hospitallers could not restore the battle, and were themselv' 
thrust back towards the camp by the rushing mass of pursue 
and pursued. Apparently the Turkish left wing tried to puf 
itself between the Franks and their place of refuge,^ and, thou^ 
it failed to cut ofl" their main body, its movements must ha' 
hastened the retreat. The flight only ceased when James 
Avesnes and Geoffrey de Lusignan led the reserve out of t: 
camp and covered the flight of the disorderly crowd of horse ai 
foot to their tents. Saladin halted below Mount Turon, ai 
would not allow any attempt to be made to storm it : he dreads 
the strength of the Franks when acting on the defensive. 

Meanwhile, a separate battle had been fought on the hillsi 
to the north by Taki-ed-din and the Master of the Templa 
We have already mentioned that, after the first shock, the fig 
had come to a standstill in this quarter, owing to the reinforc 
TTients which Saladin had sent to his nephew. A second acct 
■sion of forces to the Moslems settled the fate of the comb 

^ Itinerarium^ ?• 7I» cap. xxx. * Ibn-Alathir. 



iSg] ACRE: THE CAUSES OF DEFEAT 337 

Seeing the Christians engaged in the battle and paying no heed 
o the town, the garrison of Acre sallied out five thousand strong, 
rom the northern gate, that most remote from Mount Turon.^ 
Then, taking a circuitous route, they came out upon the rear of 
he Prankish left, and fell upon the Templars and the Champenois 
/hile the latter were hotly engaged with Taki-ed-din. The 
ntervention of this new corps broke the spirit of the Crusaders. 
They gave up all for lost, and merely strove to cut their way 
•ack to their camp. Being beset in front and rear, it was only 
portion of them who succeeded. Eighteen knights of the 
emple fell, and their Grand Master, Gerard, was captured, and 
eheaded by Saladin's orders. Andrew of Brienne, the brother of 
he Champenois count, was also slain, and forty knights more. 
io great was the slaughter in this part of the field that the numbers 
f the fallen in the Christian left wing far exceeded those lost 
y the right and the centre.^ Thus ended in defeat a battle 
^hich might under proper guidance have led to the complete 
iscomfiture of the relieving host. The Franks had risked much 
y engaging in the vast plain of El-Ghawarneh, where their 
orps were certain to get separated the one from the other. 
[evertheless, the misbehaviour of the Sultan's centre put the 
ictory into their hands. If, instead of falling on the camp, and 
lere wasting a precious hour, Conrad and the Landgrave had 
arned to take the Turkish wing-divisions in the flank, the 
nfidels could not possibly have escaped a dreadful disaster, 
aki-ed-din's corps might have been hurled into the sea, and the 
lords and Egyptians thrust into the marshes of the Belus, if 
ther of them had delayed a moment too long before taking to 
ight. But when the battle was really won, the leaders and the 
:d were equally incapable of using their advantage. The men 
irned to pillage, and we have no proof that any of their 
fficers thought of calling them off or conducting them to 
aother part of the field. Hence the Sultan, with his usual 
bility, was able to rally his men, and snatch a victory out of 
le jaws of defeat. 

^ Itinerarium, p. 70, and letter of Theobald and Peter Leo in Ralph de Diceto, 
^ Boha-ed-din (p. 145) took great pains to make out the sum of the Christian 
sses. He considered the number of seven thousand, that which was generally 
xepted in the Sultan's camp, as exaggerated. But having questioned the officer who 
id been charged to make away with the Christian corpses on the northern part of 
e field, he was told that four thousand one hundred had been carted off. He 
erefore estimated the losses of the right and centre at less than three thousand. 
22 



338 THE ART OF WAR IN THE MIDDLE AGES [ii 

Our only wonder is that he did not utilise his success for 
further assault on the Franks. But he had a wholesome drej 
of the enemy when acting on the defensive, and (as we are tol 
his own army was in the greatest disorder. Not only tl 
Crusaders, but the Turkish camp-followers had turned to pilla 
ing the tents on the hill of Ayadieh, and for the whole day aft 
the fight, as we read, the troopers were occupied in seeking the 
lost goods and extracting them from the plunderers. When 
few hours were past, the Christians, whose losses had been f 
less than might have been expected, — only the left wing h; 
really suffered much slaughter, — were safe in their camp, and mc 
angry than afraid. When the Sultan held back, they were so i 
from being cowed that their next move was to run a line 
circumvallation from sea to sea, and actually seal up the garris^ 
of Acre within its walls. 

As to losses, we have no good account of those suffered 1 
the Moslems. The contemporary letter of Theobald and Pet 
Leo to the Pope estimates them at fifteen hundred horsemen 
not improbable figure. Boha-ed-din names as slain the Kurdi 
Emir Modjelli and a few more chiefs, together with about 
hundred and fifty persons of no importance. Considering t 
rout of the centre, these numbers are wholly improbable, a 
cannot be accepted. On the other hand, the Christian soun 
give the loss of the Crusaders at fifteen hundred only,i nami 
Andrew of Brienne and Gerard the Grand Master as the or 
notable men among the slain. These figures are equally incre 
ible, especially in face of Boha-ed-din's statement as to t 
counting of the corpses.^ On the whole, we may perhaps gu( 
that each side made a better estimate of its enemy's losses th 
its own, and put them at fifteen hundred Turks to seven thousa 
Franks. 

Battle of Mansotirah, Febnimy 8, 1250. 

In our chapter on the Strategy of the Crusades we he 
already had occasion to mention the battle of Mansourah as 1 
ill-fought end of an ill-planned advance into Egypt, We poin1 
out the madness of a march across the canals and waterways 
the Delta, and showed how the campaign was certain to end 
a check, owing to the numerous and strong defensive positit 
which were in the hands of the Egyptian army. 

* Itineran'iim, p. 72. 2 gee p. 337. 



249] ST. LOUIS IN EGYPT 339 

St. Louis started on his adventure under much more favour- 
.ble circumstances than his predecessor King John of Jerusalem 
lad met .thirty years before. The Crusaders of 1219 had only 
ecured themselves a basis of operations by the capture of 
)amietta after besieging the place for a year. Their strength 
;as exhausted before they even started on their march up-country. 
ly an extraordinary chance St. Louis in 1249 took the town 
without striking a blow. All Egypt was in disorder owing to the 
aortal sickness of Sultan Malek-Saleh/ and there was no single 
trong hand at the helm. When the troops who had been told 
]{ to oppose the landing of the French were beaten back, and 
etired towards the interior, the corps which had been selected 
D garrison Damietta evacuated the place in a panic and fled 
fter the rest.^ It was to no purpose that the Sultan roused 
imself from his sick-bed to order fifty of their officers to be 
anged : the strong city had passed into the hands of the 
'rusaders, and gave them a secure starting-point and place of 
rms : it was full of stores and in perfect order, since there had 
een no occasion to batter its walls with siege engines (June 
,1249). 

Having begun so well, it was incumbent on the French king 
iP^utilise his first success and push forward while the enemy 
/fcre still panic-stricken. It is therefore with nothing less than 
stonishment that we hear that King Louis waited nearly six 
lonths at Damietta before he began his march on Cairo. The 
ircumstances explain, but do not excuse, this halt : a large part 
f'the armament had been blown into the Syrian ports by a 
'OAtrary wind, and it was thought necessary to await its 
ppearance : the summer was at its height, and the Nile flood 
^as rising over Lower Egypt, so that the face of the land was 
-ell-nigh covered with water. These would have been good 
masons for delaying the attack on Damietta till the approach of 
le cold weather and the sinking of the flood : it was obviously 
le worst possible month for an advance when the heat was at 
;s greatest and the country most water-logged. Undoubtedly 
une was a bad season for the invasion, but, having once begun, 
le French were bound to go on : their delay merely enabled 

^ The Sultan was dying of a malignant ulcer in his thigh, which contemporary 
imour ascribed to his having lain on a poisoned mat spread for him by one of his 
aves. 

^ Makrizi in the Biblioth^ue des Croisades, iv. 42. 



340 THE ART OF WAR IN THE MIDDLE AGES [12 

the Sultan to organise his resistance with a clear knowledge 
the route which his enemies must take. There had been a fee 
ful panic at Cairo when the news of the fall of Damietta arrivt 
but the long quiescence of the Franks enabled the Egyptians 
recover their self-possession and bethink them of the best mea 
of defence.^ 

It was not till October that the last contingents of the Fren 
army straggled in from Syria : they had brought with them 
number of the barons of the Holy Land, who placed themseh 
under the Count of Jaffa.^ There was some discussion when t 
whole host was mustered as to whether it should not be trai 
ferred to Alexandria,^ and attack Egypt from that side. Tl 
plan was supported by Peter of Brittany and many other baro 
and had its advantages, for the march into Egypt from Alexand 
presents far less difficulties than that from Damietta. But 
must have begun with a second disembarkment and a toilsoi 
siege. When the king's brother, Robert Count of Arte 
explained that those who wish to kill the snake strike at 
head,* and voted for an immediate advance on Cairo along t 
Damietta branch of the Nile, he carried the king and the coun 
with him, and the hopeless march began. 

On the 20th of November^ the army commenced its mar 
moving slowly forward past Fareskour, Scharemsah, and Fa 
moun, while the flotilla advanced parallel with it on the N 
A few miles after Faramoun was passed, the advance came 
a standstill (19th December), when four weeks had been occup: 
in advancing fifty miles. The check was caused by the, 
that the king found in front of him the first formidable 
course which cuts the way from Damietta to Cairo. Al 
town of Mansourah the Damietta branch of the Nile di 
itself into two parts: the one flows down to Damietta, the otl 
turns east and falls into the swamps of Lake Menzaleh. It v 
in front of the latter that the Christian army found itself stopp( 
this second waterway, which the natives call the Ashmc 
Canal, lay across its path. Behind it the whole levy of Eg\ 
was massed ; the Sultan had taken post there when Damie 

^ Jemal-el-din in the Bib. des Croisades, iv. 451, 452. 

^ John of Ibelin. He had himself been with the king at the first landing (J 
ville, p. 215). 

^ By sea, I presume : not even the French barons can have dreamed of marcl 
over three branches of the Nile and the whole breadth of the Delta. 

^ Joinville, p. 219. ^ William of Nangis, p. 374. 



cup: 

i 



249] THE FRENCH REACH MANSOURAH 341 

ell, knowing that it was the first strong defensive position which 
he French must attack. Just as the critical moment was 
ipproaching, his old malady carried him off in the last week of 
N^ovember, and he had been dead some time when St. Louis 
cached Mansourah. His widow and his ministers, however, kept 
lis death secret, and orders were still issued in his name. The 
eal charge of the defence of Egypt fell to the Emir Fakr-ed- 
iin, the commander of the army, on whom it was agreed to 
:onfer dictatorial powers. Meanwhile, swift messengers were 
jcnt to seek Malek-Saleh's son and heir, Turan Shah, who was 
ar away at Hisn-Kayfa in Mesopotamia. Till he should arrive 
:he Sultan's death was concealed from his subjects. 

The French army now found itself at the point of a narrow 
;ongue of land, an " island " as Joinville calls it, between the 
Tiain branch of the Damietta Nile and the Ashmoun Canal. 
[t was necessary to force the passage of one or the other of 
ihese waterways ; and, both because it was smaller and because 
t covered the direct road to Cairo, the king chose the Ashmoun 
IS his objective. 

Opposite him lay the tents of the Egyptian army, stretching 
■or two or three miles along the farther bank. In their midst 
'ose 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, 
-he Sultan El-Kamil had built a new city to commemorate his 
/ictory. The strategical exigencies of the roads of the Delta 
lad placed St. Louis and Malek-Saleh in exactly the same 
position as was occupied by their predecessors during the fifth 
C^rusade. 

The Egyptian army was now composed of better stuff than 
lad been the case in 1220. It was Malek-Saleh who had first 
organised the celebrated corps of the Circassian Mamelukes 
vvhich was to dominate Egypt for the next six centuries. The 
mercenary troops of his predecessors had been mainly Kurds 
md 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 [i 



I 



formed a great mass of troops, both horse and foot, but wer 
little military value. The whole brunt of the war fell upon 
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 dam up the channel and cross on foot. Accord- 
ingly he set his foot-soldiery to cast earth into the Ashmoun or 
a broad front ; the causeway advanced a few yards, but soor 
the discharge of missiles from the opposite bank became s( 
deadly that the work was stopped ; the king saw that the earth 
bearers must be protected, and therefore built along th( 
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 woodei 
towers called " belfreys " placed at the water's edge. To batte 
down these protections the Egyptians soon set their military 
machines to work, and sixteen perrieres and mangonels hurlec 
large stones or barrels of combustible matter at the covered 
ways and wooden towers. The French replied by setting U] 
against them eighteen similar engines, and the two parties sho 
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 bed of the canal fo 
more than half its breadth. But on the other side the Egyptian 
began cutting away the bank, and, the force of the curren 
aiding them, they succeeded in keeping the Ashmoun oper 
" In one day they undid what it had taken us three weeks t 
accomplish," says Joinville, "for all our work in stopping th 
channel was useless when they enlarged it on the other side." ^ 
i-^t' Meanwhile, Fakr-ed-din threw a detachment across the cam 
lo\Ver down its course, and sent them to fall on the rear of th 
French camp : they were, however, beaten off with some loss b 
the king's brothers, the Counts of Anjou and Poictiers (Decembe 
25, 1249). This was but a diversion : the real centre of th 
fighting was the causeway ; here the matter finally went ill wit 
the French. By hurling barrels of Greek fire at the belfreys an 
" cats," the Infidels finally succeeded in setting them in flame 
Nothing could be done till they were rebuilt with ship-timb( 
which the king bought for the purpose. But only a few da} 

^ Joinville, p. 221. 



PLATE XI. 



Scbennvtic 




To Illustrate the Crusades 



Jila.ni; of the Joiner Ccuuils 
and/WcUevcourses chre omitted.. GizeK 



j\. Place of S*^Louis DyKe & Engines. 
B. Place of the Egyptian Engines. 




Neighbourhood 

OP 

Mansoukah, 

124-9-50 




I 



1^6] 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 ^ who told him that 
four miles to the east of Mansourah therq 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,^ 

^ 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 WAR IN THE MIDDLE AGES [125c 

had brought to the Crusade. They had with them all the 
king's mounted crossbowmen. In the second battle among th( 
Champenois was John of Joinville, who has left us our besi 
account of the campaign ; unfortunately he has omitted to give 
us the complete list of those who marched with him. Charles o 
Anjou was probably commander of the corps.^ The king anc 
his household knights, with his brother Alphonso of Poictier 
and Henry Count of Flanders, rode in the third division. Loui 
had issued strict orders that no knight should straggle from hi: 
corps, and that the three battles should keep close together 
the van was not to advance till all three had passed the ford.^ 

The Egyptians kept a careless watch along the canal, anc 
though the ford was only four miles from their camp, at th( 
village of Sahnar, the French reached it unobserved. The vai 
division crossed, not without some difficulty, for the bottom wa 
muddy and the opposite bank scarped and slippery : a fev 
knights lost their footing and were drowned. When they wer< 
already over, a body of three hundred Arab horse appeared, bu 
promptly took to flight when the Count of Artois charged them 
they saw that the passage was lost,^ and rode off to warn thei 
comrades. 

Flushed with this trifling success, Robert of Artois forgo 
his brother's orders, and began to move off in pursuit. Th 
Master of the Temple rode up to him and besought him to stof 
but the hot-headed count would not listen to his remonstrance' 
and spurred off towards the Egyptian camp. Thinking that h 
would be shamed if he abandoned his place in the van, th 
Master unwillingly followed, and after him all the other con 
tingents of the van battle.* 

Count Robert rode so hard and so recklessly that he cam 
hurtling into the eastern end of the Egyptian camp almost a 
soon as the flying Bedouin whom he was chasing. He fouu' 
the Infidels in a state of disarray and unpreparedness, whic 
reflects little credit on their commander. The horses were nc 

^ So I gather from the fact that he rescued Joinville before the king and the thir 
corps had reached the field (Joinville, p. 226). 

- Rothelin MS., p. 602. 

^ Joinville, p. 224. They appeared when Joinville himself was crossing, i.e. aft, 
the van had passed. 

* Joinville tells a curious tale of a deaf knight who was pulling the count's brid 
and shouting " Forward and at them ! " at the top of his voice all the time that tl 
Master was pleading for delay. 



z5o] MANSOURAH : COUNT ROBERT'S CHARGE 345 

iddled nor the men armed. The French rode through the 
imp, slashing right and left and driving all before them, till 
ley came to the place where were the perrieres and mangonels 
hich commanded the unfinished causeway. They wrought 
reat slaughter, and killed the Emir Fakr-ed-din himself, fresh 
cm his bath and without his coat-of-mail, as he rode up and 
Dwn trying to rally his men. Hitherto Robert's haste had not 
Dne any irreparable harm : if he had halted and taken post 
Tiong the machines to guard the spot till the infantry should 
)mplete the work, he would almost have justified his reckless 
large. For if he had waited till the second and third battles 
id crossed the narrow ford, the enemy would have had ample 
me of warning, and would not have been surprised in their 
imp. 

But the fiery count was now to take the fatal step which 
lined the whole enterprise. Seeing the Egyptians in hopeless 
sarray, he imagined that he had gained the day with his own 
vision alone, and thought of nothing but pursuit and slaughter, 
iter a very short breathing space, he ordered a second advance 
•wards the town of Mansourah, into which many of the fugitives 
ere pouring. The Master of the Temple again besought him 
' pause and await his brother's coming, and William of Salis- 
ary added his remonstrances to those of William of Sonnac. 
ount Robert replied with inexcusable discourtesy, telling the 
emplar that the military Orders loved to protract the war for 
eir own ends, and did not really wish Christendom to triumph, 
St their own occupation should be gone.^ Then, turning to the 
arl of Salisbury, he flung in his face the old taunt about 
Englishmen with tails " and the curse of cowardice that rested 
1 them. " I shall go this day where you will not dare to keep 
vel with the tail of my horse," replied Salisbury, and, replacing 
s helmet and lowering his lance, he rushed forward with the 
St to meet his fate.^ 

^ Artois' language to the Templar, as reported by Matthew Paris (v. 149), deserves 
-ord as showing the suspicion which the Crusaders entertained of the military Orders. 
) antiqua Templi proditio ! Hoc est quod diu praecinimus augurio, quod terra tota 
ientalis jamdiu fuisset adquisita nisi Templi et Hospitalis fraudibus nos seculares 
pediremur. Timent autem Templarii et eorum complices quod si terra juribus subdatur 
iristianis, ipsorum expirabit (qui amplis reditibus saginantur), dominatio. Hinc est 
od fideles ad negotium crucis accinctos variis inficiunt potionibus, et Saracenis con 
:derati proditionibus interficiunt." 

^ Matthew Paris makes a bad error in placing this altercation after instead of 
fore the irruption into the town of Mansourah. 



346 THE ART OF WAR IN THE MIDDLE AGES [125 

The Egyptians were still so discouraged that Artois and h 
followers were able to penetrate within the walls of Mansoura 
and to ride through the town, cutting down the fugitives ; son 
of the knights even emerged at its western gate, and almo 
reached the Sultan's suburban palace. But they were scattere 
in the streets and separated one from another, so that tl 
impetus of their charge and the advantage of combined acti( 
were lost. The Egyptians fled into the houses and flung dar 
and tiles upon the knights as they galloped up and down tl 
narrow lanes. Presently the troops from the camps west of tl 
town, who had not shared in the panic of the rest of the Mosle 
army, began to pour into Mansourah. They found the Fren 
scattered in small bands, some intent on plunder and some < 
slaughter, but all unprepared to receive a fresh attack. Hen 
the new-comers won an easy success over the Christians : ma? 
were slain in the streets, others hunted out of the town a 
cut down in the open. The only route which the fugitix 
could take lay through the eastern camp of the Egyptia 
where the Mamelukes were now rallying and getting into bat 
order. Hence it is not surprising to find that nearly the wh( 
of Artois' corps was annihilated. He himself was slain in t 
town, and his surcoat with the royal French lilies was exhibit 
to the Moslems as a proof that the King of the Franks h 
fallen. With Robert there died William Longsword, the Mas 
of the Temple, the lord of Coucy, and many barons mc 
Joinville tells us that three hundred knights perished, besi< 
the sergeants and horse-arbalesters who accompanied the 
The Temple alone lost two hundred and eighty horsemen 
various ranks. The Moslems say that fifteen hundred Frei 
were cut off in all,^ and the figure is very probably corn 
Only a few scattered bands escaped, among whom were 
Duke of Brittany and the Count of Soissons. 

Meanwhile, during the hour which Artois had wasted by 
mad charge, the remainder of the French cavalry had b 
gradually crossing the Ashmoun. Joinville, who was in 
front of the middle corps, seems to have followed Count Rol '■ 
at a distance, before the king was well over the ford. At 
rate, he saw, when he reached the Egyptian camp, that son: 
the enemy were already rallying, having retired from the tt 
into the open fields where they were drawing up in line of bai • 

^ Joinville, p. 224. ^ Makrizi. 



250] MANSOURAH: THE MAIN BATTLE 347 

;^he seneschal charged the nearest squadron, but was soon swept 
.ack to the edge of the canal by the advance of the mass of the 
nfidels, whom he estimated at about six thousand horse. He 
nd his followers only saved their lives by retiring into a ruined 
iouse, where they maintained themselves, fighting on foot in the 
Gorway, till Charles of Anjou and the main body of the second 
orps came up and delivered them by driving off their assailants. 

Soon after. King Louis himself and the rear division came 
pen the scene of battle. They were at once assailed by the 
rlamelukes, who were now rallied and in good order. A fierce 
truggle began in the outskirts of the camp, and was maintained 
JT many hours. The Mamelukes poured a constant rain of 
rrows into the ranks of the French, and Louis was compelled 
charge them again and again before he could resume his 
dvance towards the all-important spot where the half-finished 
lam lay. It was absolutely necessary to reach it, in order that 
he infantry might have their chance of joining the horse. But 
>eing continually attacked on their left flank, the French could 
lot advance as they wished, but were always having to face 
outhward to beat off the Mamelukes. Seeing their enemy 
.rowing weary, and noting that hundreds of the knights were 
lismounted owing to the loss of horses under the rain of arrows, 
he Mamelukes at last threw their bows over their backs and 
harged down with mace and scimitar upon the king. Louis 
/as hard pressed, and some of his followers lost heart and 
>lunged into the Ashmoun to swim back to the Christian camp, 
^ut he persisted in his original plan of advancing to the cause- 
v-ay, and at last came level with it. 

Then the French infantry, throwing earth, planks, fascines, 
>roken military engines, and all manner of miscellaneous rubbish 
nto the unbridged half of the canal, succeeded in making a rough 
»ut sufficient bridge over the gap. The arbalesters and pikemen 
•egan to pour across the crazy structure by thousands. Humbert 
f Beaujeu, the constable, at once drew up the first crossbowmen 
hat arrived so as to cover the harassed cavalry. They opened 
destructive fire upon the Mamelukes, and the battle took a 
lew turn. 

The moment that the Egyptian leaders — Bibars, who twenty 
ears later became Sultan, is chiefly named among them — saw 
hat the French infantry were entering on the scene, they ordered 
heir horsemen to draw back. Retiring out of bowshot, they still 



h 



348 THE ART OF WAR IN THE MIDDLE AGES [125 

maintained a threatening- attitude. The king might now hav 
advanced, but his knights were so thoroughly tired out an 
harassed that he refrained from doing so. He contented hin: 
self with ordering the infantry to construct a large circular tete 
du-pont covering a considerable space of ground on the farthe 
bank of the canal. The work was easily and rapidly finishe 
by using as materials the woodwork of the captured Egyptia 
machines. 

Thus King Louis had acquired a solid lodgment on tl: 
southern side of the watercourse which had so long held hii 
in check. But he had failed to defeat the Egyptian army, whic 
still watched him at the distance of no more than a few hundre 
yards, and was rather encouraged than abashed by the resul 
of the day's fighting. The losses of the French had been .< 
much greater than those of their adversaries that the Moslen 
regarded themselves as the successful party. Louis had lost, ; 
far as can be calculated, nearly half his cavalry and a still great 
proportion of his horses. The real meaning of the battle w 
sufficiently shown by the fact that three days later ^ tl 
Egyptians assumed the offensive, and vigorously attacked t 
tete-du-pont, while the French stood entirely upon the defensi^ 
and even after beating off the assault made no further attem 
to advance. The invaders had lost their impetus and th( 
desire to push on : not long after we find them thinking 
retreat. The battle, though it had ended in the crossing of t 
Ashmoun Canal, had so exhausted the Crusaders that th' 
despaired of the result of the campaign. We cannot call 
anything but a check and a disaster.^ 

Such were the main features of the fight of Mansoural 



'% 



^ The battle had been fought on Shrove-Tuesday, and the Moslem attack on 
French lines followed on the first Friday in Lent. 

^ Joinville's interesting personal adventures after the king had come upon 
field are well worth reading, but evidently had no important influence on the forti 
of the day. He had been employed to ride on to Mansourah to look for the Count 
Artois, who was said to be yet alive, but got involved in a long skirmishing encour 
with a body of Egyptians on and about a little bridge which crossed a brook runn 
into the Ashmoun from the south. He succeeded in detaining opposite him a bod} 
the enemy who would otherwise have gone to aid in the attack on the king, 
their arrival would not have turned the event of the day — indeed, these were Egyp: 
rabble, not Mamelukes, as many of them were on foot, and they pelted Joinville ; 
his men with clods and shot at them with fire-arrows instead of charging in. His 
227-228 are of great interest, but we could wish that they contained more details al 
the king's main fight with the Mamelukes. 



25o] MANSOURAH : THE CAUSES OF FAILURE 349 

ist of the great pitched battles of the Crusades. It displays, 
ven more clearly than the other engagements with which we 
ave dealt, the absolute interdependence of cavalry and infantry 
1 the Christian hosts when dealing with the formidable horse- 
rchers of the East. For want of men armed with missile 
weapons (all the mounted crossbowmen had been slain along 
'ith Robert of Artois) the king and his chivalry were on the 
ery verge of destruction. They were saved the moment that 
leir infantry succeeded in getting across the canal and joining 
lem. Without that succour they would probably have been 
estroyed to the last man, for they had been cut off from their 
2treat to the ford, and the watercourse at their back proved 
npassable to such fugitives as attempted to cross it. 

It is curious to note that the Mohammedan writers grasped 
luch more clearly than the Christian the fact that the tardy 
rrival of the French infantry turned the engagement into a 
rawn battle, and that their earlier appearance would have made 
a decisive victory for St. Louis. Joinville^ and William of 
langis 2 mention the coming up of the crossbowmen indeed, but 
eep all their interest and admiration for the king's feats of 
ersonal valour. It is left for Jemal-ed-din and Makrizi to 
bserve that " if the first division of the Christian cavalry had 
eld out " {i.e. if Artois had remained by the engines instead of 
lunging into Mansourah), "and if the whole of the Christian 
ifantry had been engaged, Islam would have been ruined,"^ and 
lat " if the French infantry could have joined their cavalry, the 
efeat of the Egyptians and the loss of the town of Mansourah 
ould have been inevitable." * Blinded by chivalrous enthusiasm 
id class-pride, the French chroniclers omit to draw the moral 
hich to the Moslem writers was obvious. 

The separation of horse and foot while St. Louis was making 
s turning movement was unfortunate, but absolutely necessary. 
/e cannot blame the king for it, as he had no other alternative 
^fore him. All the more must the gravest censure fall on 

^ " It happened that towards evening the king's constable, Humbert de Beaujeu, 
ought us the foot-arbalesters, who drew up in front, while we dismounted, Incon- 
lently the Saracens went off and left us in peace," says Joinville — a very inadequate 
count of the crisis of the day, when whole pages have been devoted to individual 
ploits. 

^ "Nostri usque ad horam nonam graves sustinuerunt impetus. Tandem balistariorum 
bsidio multis Saracenorum vulneratis . . . nostri campum obtinuerunt " (p. 374). 

^ 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 direc 
disobedience to his brother's orders. If he had only halte( 
among the Egyptian engines opposite the French camp, an( 
held his ground there till the infantry could complete the cause 
way, and till his brother could arrive with the main body c 
the horse, the day would have gone well for Christendom. Th 
king did his best to detain him, sending ten knights to bid hir 
halt and wait,^ but Robert, in deliberate defiance of his chie 
chose to make the second mad charge, which lost the day ani 
ended his own rash career. Even the leader of a feudal arm 
could not have rationally expected to see his plans wrecked b 
such a piece of wanton and wicked indiscipline. 

1 Joinville, p. 224. 



I 



BOOK VI 

WESTERN EUROPE 

FROM THE BATTLE OF HASTINGS TO THE 

RISE OF THE LONGBOW 



Ih 



CHAPTER I 

INTRODUCTORY 

[N studying the Crusades we have seen the miHtary art of the 
nations of Western Europe at its best and its worst- 
owhere are more reckless displays of blind courage, or more 
upid neglect of the elementary rules of strategy and tactics 
) be found, than in the great expeditions to the Levant. On 
le other hand, we have also had to observe among the more 
ipable leaders of the crusading armies a far higher degree of 
telligent generalship than was usual among their contem- 
jraries in the West. If the Crusades of iioi and 1147 are 
3cidedly more distressing to the critic than the average wars of 
rance,England,or Germany, there are also battles and campaigns 
-such as that of Arsouf — which show very favourably beside 
lose of the lands nearer home. Many of the Crusaders seem to 
ive been at their best when facing the new problems of the 
ast. Richard Coeur de Lion at Acre, Arsouf, and Jaffa rises 
r above his ordinary level : we find ourselves wondering how 
le very capable general of 1190-91 can on his return waste so 
uch energy and ability to no purpose in the wretched peddling 
rench wars of 1 194-99. We may add that the great Frederic T. 
Germany never shows to such good effect in his home cam- 
ligns as in the conduct of his expedition through Asia Minor, 
any of the lesser figures of the Crusades, including the good 
odfrey of Bouillon himself, are obscure and undistinguished in 
e wars of their native lands, and only show the stuff that is in 
em when they have crossed the high seas. 
The worst military errors of the Christians in the East came, 
we have seen, from their gross ignorance of the conditions of 
arfare in Syria or Asia Minor, and of the tactics of the enemies 
ith whom they had to deal. At home leaders and led alike 
2re safe from such dangers, since they knew the military 
23 



354 THE ART OF WAR IN THE MIDDLE AGES [hoc 

character and usages of their neighbours, and had some rougl 
idea of the geography, climate, and productions of their neigh 
hours' territory. But if this knowledge preserved them fron 
certain dangers, it seems, on the other hand, that in the familia 
border wars of the West the best qualities of a commander wer 
often not developed. It is new and unforeseen dangers an( 
difficulties that test most adequately the stuff that is in a man. 

When we turn from the history of the Crusades to conside 
the contemporary history of the Art of War in Western Europ( 
the first thing that strikes us is the comparatively small influenc 
which the great campaigns in the Levant seem to have had upo 
the development of strategy and tactics at home. Tens c 
thousands of barons, knights, and sergeants came back s 
veterans from the East, and one would expect to see th 
lessons which they had learned in fighting the Turk and Syria 
perpetually applied to the wars of their native countries. Y( 
it is by no means easy to point out obvious instances of sue 
application of new principles of war, save in the provinces ( 
fortification and of arms and armour. In strategy and tacti< 
it is difficult to detect from a broad survey much direct influenc 
flowing from the Crusades. 

W^e may take as the clearest example of this the enti 
neglect by the Western nations of the most important tactic 
lesson of the Crusades. We have shown by a score of exampl 
that the one great principle which settled the fate of wars wi 
the Turk was that generals who properly combined infantry ai 
cavalry in their line of battle were successful, and that genera ■ 
who tried to dispense with the support of foot-soldiery alwa i 
failed disastrously. The fact that the combination of the tv j 
arms is better than simple reliance on one had been shown 
Hastings long ere the Crusades began, but the lesson was ev' 
more clearly visible in the details of such fights as Antioch 
Ascalon as compared with the disasters of iioi or the narrc 
escape from destruction at Dorylaeum. 

We should expect, therefore, to find that the return home 
the warriors of the first Crusade would be followed by t 
development of a rational use of infantry and cavalry in ck 
alliance and interdependence. But we find little of the kin 
over the greater part of Western and Central Europe t 
cavalry arm still maintains its exclusive predominance, a 
infantry is still despised and distrusted. In Italy, it is true, t 



2oo] THE PREPONDERANCE OF CAVALRY 355 

'orkings of the experience of the Crusades are to be recognised 
1 the sudden growth of the popularity of the crossbow, and 
robably also in the increased importance of the civic infantry, 
lut in the only other parts of Europe where foot- soldiery show 
D any effect — England and the Netherlands — we are dealing 
ith an old Teutonic survival, not with any new development. 

In many of the twelfth-century battles of Western Europe, 
hen by some rare exception we do find combatants on foot 
ntrusted with a principal part in the fight, we discover on 
oser inquiry that they are not ordinary foot - soldiery, but 
nights who have dismounted in order to carry out some 
esperate duty. We are, in short, merely witnessing a recurrence 
) that ancient habit of the Teutonic races which Leo the Wise 
ad described two hundred years before.^ Such instances are 
) be found on the part of the English and the Normans at 
enchebrai^ (1106), and again at the first battle of Lincoln^ 
146), where both King Stephen and the rebel earls dis- 
lounted the pick of their knights to form a solid reserve. The 
ime is the case in the English army at Bremule (1119), and at 
le battle of the Standard * (i 138), where the Yorkshire knights 
ft their horses and joined the yeomanry of the fy rd in order to 
iffen the mass when it was about to be assailed by the wild 
ish of the Scots. The Emperor Conrad's German chivalry 
shaved in a similar way at the chief combat during the siege 
■ Damascus in 1 148. 

Such expedients, however, are exceptional. On the other 

md, we not unfrequently find battles in which neither side 

'ought any foot-soldiery to the field, such as Thielt (1128), 

agliacozzo (1268), and the Marchfeld (1278). Cases where one 

de had no infantry whatever in the battle line are still more 

jmerous. Such are Bremide (11 19), Legnano (1176), Muret 

274). 

When true infantry are engaged on both sides, it is rare 

find them actually settling the fate of the day. Generally 

icy are only used as a very subsidiary force, employed merely 

r skirmishing and not for the decisive charge. The main 

cceptions to this rule are to be found, as we shall have to show 

ter on, in Italy and the Netherlands. But if the infantry 

most battles had no great part in the winning of the day, 

ey were often the chief sufferers in a defeat. As a rule, those 

1 See p. 202. 2 See p. 379. 3 See p. 392. * See p. 386. 




356 THE ART OF WAR IN THE MIDDLE AGES [125 

of the beaten army were fearfully mishandled by the knights c 
the victorious side. When the day was won, the infantry of th 
vanquished party were nearly always cut to pieces in the mo5 
ruthless manner, while their countrymen of the knightly classe 
were not slaughtered, but reserved for ransom. 

The mailed horseman, then, maintains his place as the chi( 
factor in battle down to the end of the thirteenth century, an 
the main features of the two hundred years from Hasting 
onward are the feudal knight and the feudal castle. We sha 
have to note that while tactics and strategy make comparative] 
small and slow progress in these two centuries, the art of fort 
fication grows very rapidly. Between the simple castle of tl 
time of William I. and the splendid and complicated fortress* 
of the end of the thirteenth century there is an enormous ga 
The methods of attack made no corresponding advance, and I 
1300 the defensive had obtained an almost complete mastei 
over the offensive, so that famine was the only certain weap( 
in siegecraft. It is not till the introduction of cannon and gu 
powder in the fourteenth century that the tables begin to 1^ 
turned. 

In chapter iii. of Book III. we dealt with the origin ai 
evolution of the feudal knight and the feudal castle. We ha 
now to treat of their further developments. 



CHAPTER II 

THE ARMIES OF THE TWELFTH AND THIRTEENTH 
CENTURIES 

Sectioji A. — England. 

[ ^ TE have first to concern ourselves with the knighthood of 
VV Western Europe and its tactics. Fortress- building 
nd siegecraft, though equally important in their influence on 
he general history of the period, must take the second place, 
^n English writer is inevitably forced to illustrate the period 
lainly from English military history, but we shall conscien- 
iously endeavour to point out all the details in which continental 
■ractice differed from that in use in our own island. 

The Norman Conquest brought about a complete change in 
tie military organisation of England : under William the 
Jastard the system of raising the armed force of the realm, the 
ictics that it employed, and the weapons that it used, were all 
like transformed. For the next two hundred years the 
I'orman castle and the Norman horseman were to be the main 
matures in the military history of England. 

The kings continued to call out the fyrd on occasion, but 
ley never treated it as the chief part of their host : it was 
ideed mainly employed when the feudal levy of the realm was, 
3r some reason or another, not to be wholly trusted. William 
lufus summoned the fyrd once for real active service, and once 
s a mere means of getting money. It was employed in the 
rst year of his reign for the sieges of the castles of the barons 
'ho had rebelled against him under the pretence of supporting 
is brother Robert. Infantry were always required for siege- 
•ork — the knights would have resented the hewing and digging, 
nd a large force of pioneers was needed. The second occasion 
n which we hear of the mustering of the old national host was 

357 



358 THE ART OF WAR IN THE MIDDLE AGES [iiJ 

when Ralph Flambard taught the king to turn a dishone. 
penny by a new device. Rufus called the shire-levies t 
Hastings, nominally for a campaign in Normandy (1094). The 
came to the number of twenty thousand, each bearing the te 
shillings which the shire was bound to provide for hir 
William took the money from them and then told them th; 
they might disperse, as they were not needed.^ Henry I. alj 
used the fyrd early in his reign, in circumstances much lit 
those which had forced his brother to employ it. Robert < 
Belesme and his fellows were in rebellion, and had manned an 
stored their castles. Large forces were needed for siegewor 
and Henry called upon the English, who came gladly, and, .- 
Orderic tells us, greeted him, after the surrender of tl 
enemy's great stronghold at Bridgenorth, with the joyful cr 
" Rejoice now, King Henry, and say that you are truly lord ' 
England, since you have put down Robert of Belesme ar 
driven him out of the bounds of your kingdom." ^ Still later tl 
shire-levies were raised by Stephen for the battle of tl 
Standard,^ and by Henry II. to put down the great feudal risii 
of II 74. The Assize of Arms of 1181 shows us how miscc 
laneous and heterogeneous was their armament : even wh( 
providing for the improvement and reorganisation of the fore 
the king does not dream of enforcing uniformity, and the poor 
classes are allowed to come to the muster armed with nothii 
better than swords, knives, and darts. There is evidently a wi 
to assimilate the wealthier men to the armament of i. 
mercenary Brabangon pikemen whom Henry was employing 
large numbers at the time, as the sheriffs are directed to see th 
persons owning sixteen marks of chattels are to bear mail-shi 
steel cap, shield, and spear. 

But alike for foreign expeditions and domestic wars, t 
Norman and Angevin kings relied mainly on the masses 
mailed horsemen provided by their feudal vassals. Still armt 
like their fathers at Hastings, with the long mail-shirt, the peak 
helmet with its nasal, and the kite-shaped Danish shield, t 
Norman knights were the flower of the chivalry of Euroj 
whether they served in their own land, in the conquered rea^ 
of England, in the new kingdom which they had built up 
Apulia and Sicily, or in the Crusades of the far East. 

^ Florence of Worcester, sud antto 1094. " Ord. Vit. xi. 3. 

^ Richard of Hexham, c. 321. 



iioo] 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 ^ that " knight-service " and " castle-ward " were 
deas not altogether unfamiliar before the Conquest, and that 
::he 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 
imong his tenants, seems often to have dealt loosely and 
iberally with the old system, frequently letting off his vassals 
A'ith less men than their acreage should have called for. 
' Beneficial hidation," the counting by favour of four or five 
lundred acres as if they were but a mere hundred and 
:wenty, was as prevalent in military arrangements as in 
Taxation. It was specially frequent when Church lands were 
Deing dealt with ; e.g: we know that the Abbey of Ramsey had 
seventy hides, and should therefore have provided fourteen 
<nights, 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 
Tiajority were obliged to supply their proper contingent. 

It has been clearly shown of late, by an eminent inquirer 
nto early English antiquities, that the hidage of the townships 
.vas very roughly assessed, and that the compilers of Domesday 
Book incline towards round numbers.^ Five-hide, ten-hide, 
Dr twenty-hide townships are so common that there was little 
difficulty in apportioning the military service due from the 
:enants-in-chief who owned them. Hence there was not so 
Tiuch difficulty from fractions as might have been expected. If 
2states had been assessed with absolute accuracy in acres and 
/ards, nearly every landholder would have been responsible for 
acentric fractions of a knight, over and above the units which 
lis manors gave when their extent was divided by the normal 
ive hides. But estates were not accurately measured and 

^ See pp. Ill, 112. 

' See Professor Maitland's Domesday Book, eic.,passifn. 




I 



360 THE ART OF WAR IN THE MIDDLE AGES [M 

assessed, and so the knights of " the old enfeoffment," ^ 
William's arrangement was entitled, are generally foun( 
round numbers : the fractions which occur are for the most 
quite simple ones. 

The landholder, knowing his servitium debittini accordinj 
the assessment of the vetus feoffamentum of the Conqueror, he 
to provide the due amount of knights. This he could do in tv 
ways : he might distribute the bulk of his estate in lots rough 
averaging five hides to sub-tenants, who would discharge tl 
knight-service for him, or he might keep about him a househo 
of domestic knights, like the housecarles of old, and mainta 
them without giving them land. Some landholders prefern 
the former plan, but some adhered, at least for a time, to tl 
latter. But generally an intermediate arrangement prevaile< 
the tenant-in-chief gave out most of his soil to knights who 
he enfeoffed on five-hide patches, but kept the balance 
dominio as his private demesne, contributing to the king for t 
ground so retained the personal service of himself, his sons, ai 
his immediate domestic retainers. 

An interesting series of documents, just a century later th; 
the Conquest, survives, and can be used to show what t 
barons had been doing with their land during the three genei 
tions which had elapsed since the first assessment. These a 
the Cartae Baronujn of 1166,^ a series of answers given by t 
tenants-in-chief to Henry II. in response to certain inquiri 
which he made from them. The king demanded a stateii^i 
as to the number of knights whom each tenant-in-chief oMJ^I 
as sub-tenants, how many were under the "old enfeoffmen 
of William I. and how many of more recent establishment,, 
also whether the lord provided his due contingent whollj 
means of sub-tenants, or was accustomed to contribute 
personal service of himself and his household for land held 
demesne. It is interesting to find that the answers show th 
the majority of the baronage had given away the larger sha 
of their estates, but still kept a certain amount in demesne 1 

^ I think that there is no doubt that Mr. Round in his Feudal England 1 
proved that we may be reasonably certain that the vetus feoffamentum really r 
back to the Conqueror, and was a formal distribution. The other view, that it \ 
irregularly and gradually established under Rufus and Henry i., seems less probal 
On the other hand, Mr, Round's "Constabularies of Knights" are not convincing. 

^ The Cartae Baronum are printed in extenso in Hearne's Liber Niger Scacca^ 
They are unfortunately incomplete, and do not cover nearly the whole of England. 



i 



i66] THE ''CARTAE BARONUM" 361 

^'hich their own personal service was due. The smaller men, 
esponsible only for the service of one or two knights, had 
iot usually enfeoffed sub-tenants, but served themselves. At 
irst a few great landholders, mostly abbeys, had refrained as 
ar as possible from cutting up their estates into sub-tenancies, 
n account of the financial advantages of keeping land in 
emesne. But this plan had the corresponding disadvantage 
f compelling the abbot to keep up a household of idle knights, 
;ho drank and roistered about the abbey precincts, and made 
hemselves an intolerable nuisance.^ Thus the house was usually 
riven, even if unwilling, to give the knights their fiefs in order 
get them away from headquarters. Where, as in the case 
f Ramsey, the abbey was very lightly assessed for knight- 
ervice, the proportion of its land which it would have to 
istribute to fulfil its servitmm debituni would not necessarily 
e a large one. But though economy dictated the enfeoffing 
f as i^w^ knights as possible, nepotism, the curse of the 
lediseval monastery, often drove abbots to give land to their 
vvn needy kinsmen, so that not un frequently it was found that 
house had created far more sub-tenants than it required. In 
ach cases the "due service" was sometimes obtained by 
laking the body of enfeoffed knights undertake to send as 
lany of themselves to the host as was necessary ; - a private 
rrangement settled who was to go on each individual expedi- 
on. 

In the twelfth century the hard-and-fast rule that five hides 
ught to make a knight's fee came gradually to be disregarded, 
n some cases a liberal lord gave his sub-tenant a good deal more 
lan the normal holding ; in other cases knights were enfeoffed 
n a good deal less — occasionally on patches no larger than two 
ides. Thus we can find a tenant describing his holding as 
pauperrimum," and grumbling at its counting as a fee at all. 
ut such cases, in spite of their numbers, were theoretically 
bnormal, and the notion which connected five hides with the 
night survived down to the time of Henry il. In the Cartae 

^ Liber Eliensis, 275. 

'At Ramsey "Homines faciunt quattuor milites in communi in servitium 
)mini regis, ita quod tota terra abbatiae communicata est cum iis per hidas ad 
aedictum servitium faciendum ; " i.e.y though only four knights are required (a very 
lall contingent from seventy hides), the abbey has not designated four particular 
itches to discharge its knight-service, but all the tenants, as well as the abbey 
:mesne land, club together to *' make " four knights for the host. 



362 THE ART OF WAR IN THE MIDDLE AGES [in 

Baronum we get a good example of this : Roger de Berkeh 
owed two knights and a half on the " old enfeoffment " : givii 
more details than his fellows generally supply, he explains 
follows : ^ — 

" The first knight is thus made up — 



Michael holds one hide ^ 

William Fitz-Baldwin, two hides I. — fi h* 1 

Helyas de Boivill, one and a half hides / "~ * 

Hugh de Planta, half a hide 



) 



I 



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 (f hide) \ = two and a 

Roger de Albamara, one virgate {\ hide) / half hides, 

Simon de Coverley, one virgate ,, 

The Prior of Stanley, one virgate ,, / ^Hj 

and here you have half a knight. 

" For making up another knight, Walter de Holecom] 
Gerard, and Reginald de Albamara hold between them t 
hides, but deny their full obligation and say that they do i 
service only for one virgate each. From them you can ma 
up a knight, and so you have two and a half knights enfeoffec 

Roger's argument in the third paragraph is hard to folio 
either the figures in the text have got corrupted, or he thir 
his disputed claim to ten hides will be compounded for haMl 
value, and that Walter, Gerard, and Reginald will do^BI 
knight's service between them. However this may be, t 
first two paragraphs of his answer amply show that he conceiv 
five hides to be the proper and normal allowance of la 
which should provide a knight. He concludes his " Cart 
with a list of his demesne land, which shows that (unlike m 
of his fellows) he had let to sub-tenants only the smaller p 
of his ancestral estates. 

As a rule, no one except a very great baron with plenty 
house-room in his castle cared to have many domestic knig 
dwelling with him throughout the year. Most of the holders 
middle-sized estates had carved the greater portion of th 
into knights* fees, and only kept in demesne as much as tJ 
themselves and their sons could do service for. 

^ Hearne, Liber Niger Scaccarii, p. 165. 



ri66] THE DANGERS OF SUBINFEUDATION 363 

There was always a great deal of trouble in keeping the 
mb-tenants up to their work. In times of civil strife, a tenant- 
n-chief might rebel, or might remain loyal. If he rebelled, 
;ome of his vassals would try to save themselves from confisca- 
ion at the king's hands by refusing to join in the rising. 
Such indeed was the bounden duty of the English sub-tenant, ever 
iince the Conqueror at the great moot of Salisbury had impressed 
jpon the English knighthood the fact that their allegiance was 
Drimarily due to the Crown, and not to their immediate lords. 
3n the other hand, when the tenant-in-chief adhered to the 
cing, it was not unusual for some of his knights to slip into the 
ebel camp : if the rising succeeded, they would have every 
:hance of shaking off their lord and freeing themselves for 
he future from the service that they owed him. In Stephen's 
eign, when anarchy prevailed for well-nigh a score of years, 
he relations of countless lords and vassals had been confused : 
lisputed claims to overlordship were found on every side. 
\lany of the answers of the barons of 11 66 show that they 
vere not quite certain as to all their own rights and possessions. 
They qualify their statements with clauses to the effect that 
hey have replied to the best of their knowledge and belief, or 
lote (like Roger of Berkeley quoted above) that some of their 
ub-tenants deny their obligations. The clerical tenants are 
pecially bitter against spoilers who have robbed them of 
lomage, or compelled them to enfeoff knights contrary to their 
/ill. We are surprised to find such a respectable person as the 
Teat Chancellor Roger of Salisbury reported as an oppressor 
•y the Abbey of Abbotsbury in Dorset.^ 

The importance of King Henry's inquest of 1166 was 
wofold. It not only gave him the information that he 
equired as to the proper maintenance of the debitum 
ervitium due under the "old enfeoffment" of the Conqueror, 
'Ut showed him how many more knights had been planted out 
ince that assessment. Having possession of this valuable 
iformation, he was able to demand for the future, when raising 
ids and scutages from his tenants-in-chief, payment not 
:ierely for the theoretical number of knights whom they owed, 
lut for the real number which they actually possessed. This 

^ Liber Niger ^ p. 76 : "Cum Rogerus episcopus habuit custodiam abbatiae, duas 
idas apud Atrum, ad maritandam quandam neptem suam dedit Nicolao de Meriet, 
Dntradicente conventu." 



364 THE ART OF WAR IN THE MIDDLE AGES [116 

gave a welcome relief to the Treasury, as in many instances th 
"old enfeoffment" had been — as we have already mentioned - 
very lax and liberal, and did not adequately represent the r< 
sources of the land. 

The Cartae Baronum are unfortunately incomplete: if the 
had all been preserved, we should have been able to say bot 
what was the number of knights due from the whole of Englan 
under the " old enfeoffment " of the eleventh century, an 
what was the number of knights' fees actually existing in 116 
A careful and ingenious calculation has been worked out b 
supplementing the Cartae from other sources, which makes 
clear that the full feudal force of England was well over foi 
thousand five hundred knights, but little, if at all, over fi^ 
thousand. Of these the Church fiefs supplied about eight hundre 
the lay tenants-in-chief between four thousand and four thousar 
two hundred.^ These modest figures contrast most strange 
with the vague numbers given by contemporary chroniclci 
who were so far from appreciating the actual size and resourc 
of the land that they often state that England could supp 
thirty thousand or even sixty thousand ^ knights for the king 
service. The whole fyrd of foot-soldiery added to the knigb 
hood would probably not have reached the latter figure. 

We must be careful, when dealing with the knight of' 
eleventh and twelfth centuries, to clear away from our minds 
chivalrous connotation of the same word in the fourteen! 
fifteenth century. The knight of William the Conqueror's 
was not necessarily nobly born, nor had he gone through^ 
elaborate ceremonial of admission to the knightly order 
prevailed three centuries later. He was simply a soldier 
fought on horseback, and who received from the king, or 
one of the king's tenants-in-chief, a patch of land on condition th 
he should do mounted service in return for it. The origir 
knights of the "old enfeoffment" were a mixed multitude 
many races drawn from many different stations in life : sor 
were the kinsmen of great Norman barons, others were milita 

^ Mr. Round's calculations on this point in his Feudal England, pp. 289-293, 
most valuable and convincing. The result is certainly surprising, and shows -a 
clearly the extraordinary want of appreciation of large figures in the thirteer 
century chroniclers, and even in Government officials who ought to have knc 
better. 

^ Swereford in the Liber Ruletis says thirty-two thousand ; Ordericus Vitali 
responsible for the still more monstrous sixty thousand. 



i66] THE EARLY KNIGHTS 365 

dventurers who had drifted in from all parts of the Continent, 
nto this heterogeneous body were incorporated the remains 
f the old English thegnhood, all the lucky survivors who had 
leen permitted to " buy back their land " from the king by 
aying him a fine and doing him homage on feudal conditions 
fter his coronation. English-speaking men applied to this 
evvly-formed and miscellaneous class of military tenants and 
ub-tenants the word " cniht," which had been used before the 
'onquest for the military dependants of the great landholders.^ 
t was really equivalent to the clieits, serviens, or famulus of the 
Continent, and has the same original meaning .of subordination 
nd subservience. But names chance on different histories in 
ifferent countries ; and while " knight " became in England the 
quivalent of miles, the name servietis came across the Channel 
3me generations later, in the form of " sergeant," to express a 
lass of men distinctly below the knightly rank. It is curious 
1 note that in Germany knecht^ starting with much the same 
leaning as " knight " in the eleventh century, gradually came 
3 denote persons of a more and more inferior status, sinking 
3 mean combatants who were not of noble blood,^ and finally 
enoting 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 
ith a fewhundred acres of land would probably have been called by 
chronicler of the time of Henry I. a " miles," by a chronicler of 
ic time of John or Henry ill. a " sergeant," ^ and by a chronicler 

^ For a picture of pre-Conquest "knights" in England, see the interesting 
-scription of the rights and duties of the "radknights" of Bishop Oswald of 
'orcester, which Professor Maitland has worked out in his Domesday Book and 
eyond," pp. 305-3 1 1. 

* The word Edelknecht was invented to denote the non-knightly combatants of 
:)od birth ( = English esquire), and then knecht without the prefix came to distinctly 
iply want of birth. 

^ That "sergeant" originally means not a professional soldier, nor a knight's 
tendant, but a landed military dependant who is not a knight, is well shown by 
le letter of Geoffrey Ferland, Sheriff of Leicester and Rutland in 1216, giving 
the names of all the knights and sergeants domiciled in his district who have 
ihered to Louis of France " (Rymer, 144). Another good example is John's writ 
• 12 1 3, to call out the full feudal levy : " Rex vicecomiti de X. salutera, etc. 
ummone comites barones milites et omnes liberos homines et servietties, de quocumque 
:neant,ut sint apud Doveram cum armis et equis," etc. (Rymer, I. 1 10). The "armis 
equis" clause shows that we are dealing with mounted men, and the *'de quo- 
imque teneant " that we are dealing with sub-tenants and not merely small tencnts* 
i-chief. 



366- THE ART OF WAR IN THE MIDDLE AGES [115c 

of the time of Edward III. a " squire " {armiger 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 fine 
large bodies of men commanded by mere esquires in the wars o 
Edward III. The reigns in which the change first made itsel 
felt were those of Henry III. and Edward I., whose repeatec 
attempts to make holders of knightly fees take up the knighth 
title by the writs of " distraint " are well known.^ But thi 
attempt did not succeed, and ere long we find the king conceding 
that even the parliamentary knights of the shire may be person 
who have not actually received knighthood, because that ii 
many counties there cannot be found sufficient competen 
persons who have taken up the required status. 

Before proceeding to investigate the character of the battle 
of the twelfth and early thirteenth centuries, we must tak 
note of one marked feature of the early Plantagenet reigns- 
the prominent place taken in their military affairs by mercenar 
troops. From the time of Stephen onward, we perpetually fin 
the feudal levies of the realm supplemented by great bodies c 
professional soldiers, nearly all foreigners. There had for 
long time existed a large floating body of adventurers i 
Western Europe : from them William the Conqueror had draw 
no small proportion of the host that fought at Hastings. Th 
original Norman conquerors of Apulia had belonged to th 
class no less than the Varangian Guards of the Eastern emperor 
During the early Norman reigns we not unfrequently fin 
mention of stipendiarii milites in England,^ but it is not ti 
the time of Stephen that we begin to find them appearing 
great force and forming a prominent feature in the 
Stephen, deserted by the greater part of the baronage, supi 
the place of the missing contingents by bringing over greT 
bodies of Flemings and Brabangons, under leaders such ; 
William of Ypres and Alan of Dinan. Henry II. and Richard 
kept up the system : without the aid of a permanent arn 
they could not have maintained their long wars over sea. F 
sieges in Normandy and Aquitaine the service of the Engli 
feudal levy would have been almost useless to them. 1 
forty days would have ended almost before it could arrive 

1 Especially in xix. Henry in. and in vi. Edward I. 

2 See Florence of Worcester, stib attno 1085. 



t tl . 

1 



173] THE ORIGIN GF SCUT AGE 367 

le distant seat of war. Moreover, a feudal host was untrained, 
ndisciplined, disorderly, and sometimes disloyal. The mer- 
^naries, on the other hand, were trained professional soldiers, 
ho served with fidelity as long as they were regularly paid, 
nd had no wish to cut the war short by an intempestive 
^turn to their homes. Hence for foreign service Henry 
id Richard preferred the steady squadrons of mercenaries 
ho kept the field all the year round, to the short and uncertain 
d of the knighthood of England. To repel a Scottish foray 
■ to carry out an expedition into Wales, on the other hand, 
le servitium debituin of the English tenant-in-chief was still 
<acted. Such campaigns were short, and cost less if carried 
at by the levies of the border shires. Henry II., therefore, very 
^Idom brought over his mercenary bands to England : the 
ily occasion when they appeared in force on this side of 
le Channel was to aid in suppressing the feudal rebellion of 
173-74- Ii^ this campaign they met their likes in battle, for 
le rebel Earl of Leicester had enlisted a great body of Flemish 
nitiei'Sy and was fighting at their head when he was taken 
risoner at Fornham. 

When the king did not wish to call out the feudal levy of 
ngland, he was accustomed to exact from all the exempted 
eights a scutage. By this arrangement the holder of a fief 
)mpounded for his personal service by paying a fixed sum for 
-ery shield {scutunt) that he should have brought to the host. 
he usual sum raised was 26s. 8d. — two marks — which seems 
) represent forty days' service at 8d. a day, the normal pay of 
knight in the twelfth century. The individuals from whom 
le servitium' debitum was due seem to have been allowed the 
loice of attending in person or paying the scutage.^ If the 
impaign was near at hand, the majority would appear in arms ; 
it was distant, only a few — mainly the larger tenants — would 
>llow the host. 

^ The whole body of feudal tenants do not seem to have been so prone to accept 
e alternative of composition as might be inferred from the chroniclers. For example, 

Mr. Round has shown, Robert de Monte tells us that in 1159 King Henry took 
ith him "capitales barones suos cum paucis, solidarios vero milites innumeros" ; but the 
utage figures show that the sum received was ^{^17 14, i.e. the money representing 1280 
lights, not more than a third of the number liable to serve from the lay fiefs, so that 
)t only the great barons must have followed the king, but some two-thirds of the smaller 
en also {Feudal England, p. 280). The reason advanced for the king's preference 

a scutage is obviously not the right one. In reality he wanted the money to pay 
ercenaries. . 



368 THE ART OF WAR IN THE MIDDLE AGES [121 

Scutage appears as a recognised institution under Henry i. 
but it was his greater grandson who made it normal an 
customary. By the end of his reign the bulk of the ruri 
knights had grown into the habit of compounding instead c 
going on wearisome expeditions to Poitou or Aquitaine, ove 
the stormy seas so hateful to the mediaeval mind. The pa} 
ment of scutage became the rule, and the hiring of mercenar 
horsemen with the proceeds of this imposition gave th 
king a more permanent and trustworthy army than he coul 
otherwise have kept together. It was mainly at the head ( 
these 'professional soldiers that Henry II. and Richard Coei 
de Lion fought out their weary and uninteresting Frenc 
campaigns. 

John, because he was more hated by his subjects than h 
father and brother had been, was still more prone than the 
to employ mercenary troops. No small part of his unpopularil 
in England came from the fact that after he had been driven 01 
of Normandy in 1204 he brought back with him the horde 
foreign adventurers who had followed his unlucky standai 
on the Continent. They were, as] might have been expecte 
very undesirable guests : the barons resented the favour whi( 
the king showed to the leaders — unscrupulous ruffians, for tl 
most part, like Fawkes de Brdaute. The common peep 
suffered from the plundering propensities which the mercenari 
had picked up on the Continent. To the hatred they won fro 
rich and poor alike, the adventurers owe their dishonourat 
mention in the Great Charter. The king is forced to promi 
to dismiss all the " alienigenos milites et balistarios et servient 
stipendiarios " who " venerunt cum armis et equis ad nocumentu 
regni." ^ A special clause names several of the leaders who \^J 
condemned to banishment — Gerard of Athies, Philip of jH 
Mark, Englehard de Cigognes, Guy de Cancelles, and other'' 
As everyone knows, John slipped easily out of the obligati( 
— the mercenaries were not expelled, and formed the be 
part of the army with which the king fought his unfortuna 
campaign of 1215. The troopers of Fawkes de Breaute, ai 
also his crossbowmen, are specially mentioned as having do 
good service, early in the reign of Henry ill., at the secoi 
battle of Lincoln. It is not till the reign of Edward i. th 

^ See the proofs in Mr. Round's Knight-service, pp. 268, 269. 
* Magna Carta, clause 51. ^ Ibid. 50. 



loo] CONTINENTAL ARMIES 369 

)reign mercenaries cease to form a prominent part in the armies 
f the Plan tagenets. 

Section B. — The Conlment. 

During the twelfth and thirteenth centuries the armies of 
le English kings differed less from those of the sovereigns of 
le Continent than at any other period in history. The Norman 
ifluence had assimilated the military forces of our island to 
lose of the rest of Western Europe. Tlie chief points of 
ifference worth noticing are, firstly, that in England there was 
iver such a clear line of division between the various classes 
" feudal tenants as elsewhere ; and, secondly, that shire-levies 
foot-soldiery, the lineal descendants of the fyrd, though 
xupying a very secondary place in war, are yet much more im- 
)rtant than the infantry of most continental districts. Only in 
e Netherlands and to a certain extent in Italy do foot-soldiery 
)me prominently to the front. In other regions the mercenary 
ossbowmen are the only dismounted men who receive much 
ention, till we come to the attempt of Philip Augustus to 
rn the levies of the French communes to account. 

The normal army of an emperor or a French king was com- 
)sed of the same elements as those with which our Norman or 
ngevin monarchs took the field — a mass of mounted feudal 
nants and sub-tenants, often supplemented by a certain propor- 
)n of mercenary horsemen and crossbowmen. Occasionally we 
d civic militia in the field — it develops in Italy and the Low 
' )untries long before it is found elsewhere. Very rare is the 
; pearance for any practical purpose of the foot-levies of the 

< untryside, which the feudal lords could as a last extremity 
<ag out to battle. 

In the eleventh century the important part of a continental 

< ny consists of all the warriors holding fiefs, either directly 
i m the Crown or as sub-tenants, on condition of doing service 

< horseback. The chroniclers often speak of the whole mass 
c them as " viilites" whether they be small men or great, but a 
c eful inquiry into the character of the body shows that it is 
r t homogeneous. When we find phrases like " miles pi'imi 
chnis" or ''miles gjrgarius" we see that within the body 
c milites there are class distinctions. The highest rank is 
c nposed of free vassals of noble blood holding considerable 
» s: this is the only class which retains the knightly style in 

24 



370 THE ART OF WAR IN THE MIDDLE AGES [ii. 

subsequent ages, but the name miles in 1080 (abroad as 
England ^) is far more vague, and covers far more persons th^ 
it does in 1 180 or 1280. 

Below these milites p7'imi ordinis are a number of oth 
horsemen, some of noble but more of non-noble blood. Son 
are the king's personal retainers, serving him as minor officials 
guardsmen : a twelfth-century German chronicler would probab 
call them " miiiisterialesl' an English or a French chronic) 
" servientes regis" Much more numerous are the persor 
retainers of the barons, bishops, and abbots, whether enfeoff 
or not enfeoffed on land. These " men " of the king or 
the tenants- in-chief are sometimes styled milites gregai 
milites ignobiles, milites plebei, or milites mediocris nobilita, 
They are also found with names which differentiate them m( 
clearly from the knights of higher rank, and point to their si 
servient and dependent condition — e.g. satellites^ servien. 
clientes, famuli. As a rule, they served on lighter horses, a 
wore less complete armour than the knightly vassals. Down 
the thirteenth century they much exceeded in numbers 
nobler and more heavily-armed horsemen.^ 

When in the later twelfth century the title miles becor 
strictly confined to the upper ranks of the military class, j-^;^ 
(sergeant) is the most usual term for the horsemen of fl 
status. In France it grew to be the only recognised name 
them. In Germany it was not so common, i-^r/<^;// (the Ge| 
form of the word) being used indifferently along with 
appellations, such as scutifer^ armiger, strator. These t) 
and thirteenth-century servteittes or scutiferi are not to be confi i 
with the squires of the fourteenth and fifteenth centuries ' 1 
were the personal attendants of a knight. In the earlier age X 
knight had no mounted follower. His armour-bearer ace i 
panied him on foot, and was not necessarily a combatant a1 j 
The " sergeants " were often formed into separate corps, a ] 
from the knights, and used for the purposes for which 1 i 
cavalry are required ; or, again, they were placed in the 
important parts of the battle-array. Not unfrequently we 
sergeants placed in the front line to open the combat, while 
kniehthood is held in reserve to deal the decisive blow. 



1 

;we " 



^ See p. 440. 

2 e.g. we shall see that at Legnano the emperor's host comprised five hi 
knights to fifteen hundred sergeants. See p. 442, 



J 



[2oo] KNIGHT AND SERGEANT 371 

jhall find Philip Augustus employing this arrangement in his 
ight wing at Bouvines (1214)} Frederic il. did the same at 
rortenuova in 1237. But it was by no means the regular rule 
o separate the lighter and the heavier horsemen. It was more 
:ommon to compose each of the divisions of an army of 
ergeants, " stiffened " by the admixture of a certain proportion 
•f knights, as did, e.g:, the elder Montfort at Muret (i2i3).2 

A further complication is- introduced into the nomenclature 

f the military class when, in the twelfth century, the word 

'tiles has its meaning still further changed by the spread of 

he new idea of chivalry. When the notion is introduced that a 

night must be solemnly invested with the arms and insignia of 

le knightly rank by his feudal superior or some other personage 

f importance, and must not call himself miles till he has 

; een so honoured, there necessarily comes into existence a class 

, f holders of knightly fiefs who have not yet received the 

» nightly name. A young baron with very large estates may 

[ irve for some time before earning the title. On the other 

* and, a warrior of approved courage, whether of noble or non- 

Dble blood, may receive knighthood from king or duke for 

)me notable feat of arms. Thus a baron not yet knighted 

as often followed to war by vassals who had attained the rank 

' which he was still aspiring. 

Hence, in the later twelfth or in the thirteenth century, when 
I examine the composition of that part of the personnel of a 
udal host which does not consist of knights, we find quite a 
':g^ variety of classes represented in it. We may notice — (i) 
)ung holders of knightly fiefs who have not yet received the 
lightly title ; (2) men of knightly blood, holding small fiefs, 
10, on account of poverty (or some such other reason) do not 
i :end to take up the honour ; (3) younger sons of barons and 
1 ights, who have no land and therefore cannot afford to aspire to 
i ighthood (this was a class out of which the mercenary cavalry 
^ re very largely'recruited) ; (4) various degrees of persons of 
1 n-knightly blood enfeoffed on land by their lords. The first 
t -ee sections are men of the knightly class, but not knights : 
t ) last is the one to which the title of sergeant properly belongs. 
i cross-division is made by the fact that a wealthy sergeant 
r y sometimes succeed in providing himself with a heavy war- 
trse 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 [125 

classes 2 and 3 may be serving in incomplete armour and o 
inferior chargers. 

In the later thirteenth century we find the three latter classe 
tending to melt together, and to be considered as all equall 
forming part of the military aristocracy, so that most of th 
sergeants ultimately became " noble." Though not knights, the 
form the lower ranks of the knightly caste. It is easy to undei 
stand that when the knightly title became restricted to a com 
paratively few individuals of the knightly houses, and when th 
poorer members of them were continually serving along wit 
the richer sergeants, the latter should ascend a step on th 
social ladder. It was more natural that the sergeants shoul 
advance to a better status, than that the brothers and young( 
sons of the holders of knightly fiefs should descend to a low( 
one. So by the fourteenth century the French noblesse ar 
the German Add have extended their ranks so as to incluc 
classes which two hundred years earlier would not have been co 
sidered to belong to the nobly-born. The term sergeant pass 
out of use as meaning a feudal horseman of the lower rank,^ ai 
armies are reckoned not as containing inilites and servient 
but by the number of " helms " or " barded horses " that th 
muster. No one now stops to inquire whether the warrior w 
wears the full panoply and rides a heavy charger has or has r 
received the knightly spurs and girdle. He is an equa 
efficient member of the host, whether he bears the knigh 
name or not. The general body of the feudal horsemen w 
have not won their spurs are now called squires {eaiyei's^knea 
armigeri)^ or men-at-arms. 

It is, of course, impossible for an army to dispense altoget' 
with light cavalry ; they are needed for purposes of foraging ; 
reconnoitring. In this capacity the place once held by 
servientes is occupied in the fourteenth century mainly 
mercenaries, but partly also by the incompletely armed serva 
of the knights and squires, who brought with them to the I: 
a certain number of mounted attendants {valets arm^s, Dieri 
There were, however, to be found light horse who were neit 
mercenaries nor mere dependants of the men-at-arms. S 
Iroops certainly existed in England ; we recognise them in 

^ Remaining in use, however, as we shall see later on, for certain individual? 
tlie king's personal retinue of " sergeants-at-arms," employed by him for va 
small official duties. It also survives in occasional use for foot-soldiery. 



[250] CONTINENTAL MERCENARIES 373 

)auncenars and hobilars of the Calais muster-roll of Edward 
II. (1347).^ On the Continent, too, they appear as panzer ati 
)r remier in Germany, as hmibergeons in France. 

No account of the armies of the twelfth and thirteenth 
;enturies would be complete without mention of the mercenary 
;avalry. We have already seen that in England they occupy a 
^ery prominent place in military history, and the same is the 
ase on the Continent. From the days of the Norman adven- 
urers who ousted their unfortunate employers from Apulia and 
3enevento, the mercenary is always intermittently in evidence. 
Robert Guiscard and William the Conqueror were able to 
ecruit them by the thousand, and in most continental wars we 
ind them serving side by side with the emperor's or king's 
iege vassals. Their bands would include a much smaller pro- 
)ortion of knights and a much larger proportion of combatants 
>f lesser status than did the normal feudal host. The knights 
v-ho left their fiefs to follow the career of adventure were 
laturally not so numerous as the smaller men. The bulk of a 
mercenary band would be composed of the landless younger 
ons of sub-tenants, mixed with adventurers of lower birth who 
.ad taken to the profession of arms from love of fighting or 
rem the wish to escape from villeinage. Whatever the origin 
f these mercenary horsemen, all who were not knights were 
ommonly known as " sergeants," the escaped villein no less than 
is better-born companion. At first it was more common to 
uy the service of mercenaries by the gift of land, but by the 
ivelfth century there was enough money in circulation to enable 
ings and emperors to retain the hired horsemen in service by 
le regular payment of a daily, monthly, or yearly salary, 
'his was in every way better for the employer : the enfeoffed 
lercenary was generally a bad and turbulent subject (we need 
nly recall to the English reader such instances as Fawkes de 
Ireaute), while the adventurer hired for a fixed term could be 
uly discharged when he was no longer needed. 

The mercenary bands were increasing in importance all 
irough the period with which we are now dealing. Only local 
ars could be conducted by the regular feudal levy ; all long 
ad distant campaigns and all large schemes of conquest 
squired the co-operation of hired soldiery. Kings with a wide 
id scattered empire, like Henry II. of England, were necessarily 

^ See p. 366. Brady, vol. iii., Appendix. 



|t4 THE ART OF WAR IN THE MIDDLE AGES [130^ 

driven to employ them. Adventurers in search of a realm, lik( 
Charles of Anjou in the succeeding century, naturally reliec 
upon them. Long-continued wars hardened them into compac 
masses, till by the end of the thirteenth century we find th( 
condottiere system coming into existence — noted mercenary 
chiefs have collected huge bodies of men numbered by th< 
thousand, and hawk their services about from court to court 
The first ^ of these hosts of free-companions which comes intc 
prominence is the " great company " of Roger de Flor, formec 
from the discharged mercenary bands of the King of Aragor 
turned loose when Peter ended his long struggle for Sicily wit! 
Charles n. of Anjou. Roger's horde was strong enough t< 
shake the whole Levant, to bring the Byzantine Empero 
Andronicus to his knees (1308), and to carve out for itself a ne\ 
home in the duchy of Athens. 

Turning to the continental foot-soldiery, we find that w 
need not in the twelfth century concern ourselves greatly wit] 
France or Germany ; the Netherlands and Italy are the two dis 
tricts which demand our attention. Closely akin to the Englisl 
the inhabitants of Flanders, Brabant, and the neighbouring region 
had, like their kinsmen on this side of the water, taken late t 
horsemanship. Unlike England, the Netherlands had neve 
been conquered and divided up by any invader, and it seem 
likely that their steady infantry descends directly and withoi 
a break from the times of the Carolingians. The growth of a 
indigenous feudal cavalry in the duchies and counties of th 
Low Countries did not entirely extinguish the foot-soldiery, a 
was the case in most other regions. As early as 1 100 we hav 
notices of Netherlandish infantry armed with the pike whic 
enjoyed a reputation far above that of the foot-levies of oth( 
countries.2 In the earliest cases they are called geldons — th 
same word, it will be remembered, which Wace uses for th 
English axemen at Hastings.^ We may guess that the maile 

1 We can perhaps hardly count Stephen's Flemish captain William of Ypres 
Richard Coeur de Lion's follower Mercadier as real condottieri, as it does not see 
that they hawked about already formed bands for service, but rather that they gather 
and kept together new corps at the king's expense. 

2 In 1 106 the Annals of Hildesheim, 3. 1 10, mention that the Duke of Brabant se 
to aid the Archbishop of Cologne ** quoddam genus hominum qui vocantur Gehiuv 
viri bellatores et strenui, et nimis docti ad praelia." 

2 Wace, 12927 : 

" Geldons Engleiz haches portoient 
E gisarmes ki bien trancheoient." 



1 2 14] 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 Brabangons, 
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 ^ 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 F'lemish 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 

^ " Homines de Braibanto, pedites quidem, sed in scientia et virtute bellandi 
equitibus non 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 w^inning 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 Netherlanders 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 Crusade* 
of the Levant.^ Seafarers like the Venetians, Genoese, anc 
Pisans naturally developed into foot-soldiery. It is as cross- 
bowmen that they appear at every siege and battle in Syric 
during the twelfth century. Of all the peoples of Europe, non( 
had such skill in the use of the arbalest : after winning a hig^ 
reputation as marksmen in the battles of the East, we find thes( 
Italian foot-soldiers, and especially the Genoese, passing north o 
the Alps as mercenaries, and fighting in the French service a' 
Courtray, Sluys, and Cregy. 

While the inhabitants of the seafaring towns were mainl) 
skilled in the use of the crossbow, the civic militia of the inlant 
cities was chiefly composed of pikemen. The army of ar 
important municipality like Milan or Verona consisted of i 
mass of infantry, backed by a certain proportion of horse. Foi 
the Lombard states owned a not inconsiderable amount o 
cavalry, provided partly by the nobles of the countryside, whc 

1 See pp. 252, 253. 



1200] ITALIAN FOOT-SOLDIERY 377 

liad 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,^ 
where the steadiness of the Milanese foot saved the day, by 
allowing the routed Lombard horse time to rally and resume 
the charge. 

^ See p. 442. 



CHAPTER III 

ENGLISH BATTLES AND THEIR TACTICS, II00-I200 

Tejzckebrai {iio6) — Bremiile (i 1 19) — Northallerton (i 1 38) 
L incoln ( 1 1 4 1 ) — Battles in h^eland ( 1 1 69-7 1 ). 



% 



1 



IT has been often observed that the period of the completes 
supremacy of cavalry in the West, the twelfth century, wa 
not a period of great battles. There are more important fights i 
England in the open field during the sixteen years of the War 
of the Roses, or the six years of the Great Rebellion, than i 
the whole century between iioo and 1200. The same is th 
case on the Continent, though in not quite such a notice 
able degree. The main reason of this was, that the develof 
ment of fortification during the century was so enormous, that : 
was more profitable for the weaker side to take the defensiv 
behind strong walls than to fight in the open. Hence th 
century is pre-eminently one of sieges rather than of pitche 
battles. Henry l.'s victories of Tenchebrai and Bremule wer 
very small affairs, in which only a few hundred knights too 
part. The long civil wars of Stephen and Matilda abound wit 
sieges, but only supply the two battles of Northallerton an 
Lincoln. All the long French wars of Henry II. do not give v 
a single first-rate engagement in the open ; the skirmish c 
Fornham and the surprise of Alnwick arc the only fights in h 
reign that we need notice. The same is the case with the Ion 
bickering of Richard L and Philip Augustus along the Norma 
and Poitevin borders. It is hardly too much to say that betwee 
Lincoln (1141) and Bouvines (12 14) no English troops wei 
present at an engagement of first-rate importance in Wester 
Europe, If it had not been for the distant crusading battle ( 
Arsouf (1191), we might have said that there was no really gre; 
battle in the whole period in which they were engaged. 



378 



ic6] BATTLE OF TENCHEBRAI 379 

For the most part, these unimportant conflicts of the 
vvelfth century were both simple and short. Another 
lotable point about them was, that they were accompanied 
)y very little effusion of blood, save when some luckless 
nfantry had been dragged into the field by one side or the 
)ther : in that case there was often cruel butchery in the 
xirsuit ; otherwise the knights gave each other quarter, and 
he main loss of the defeated side consisted of prisoners and 
lot of slain. 

Battle of Tenchebraiy September 28, 1 106. 

Henry I. of England had invaded the lands of his brother 
R-obert and overrun most of the duchy of Normandy. He was 
beleaguering Tenchebrai, a castle of the Count of Mortain, when 
:he duke resolved to make a desperate attempt to raise the 
>iege. Gathering all the forces that he could muster, he 
;narched 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.^ 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 

^ "Rex namque et dux et acies caeterae pedites erant ut constantius pugnarent" 
(Henry of Huntingdon, 235). 



38o THE ART OF WAR IN THE MIDDLE AGES [uo6 

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,^ while the centre and right of 
the king's army held their ground. But immediately after, 
Helie 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,^ 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.^ Robert of Belesm^, 
who fled too early for his own good fame,* was the only man'^Hl 
note in the duke's army who got away. 

The whole fight had not occupied an hour, and not a sin^ 
knight on Henry's side had been slain. We have to turn:' 
Italian chronicles of the fifteenth century to find such a blo^ 
less fight followed by such great results — for the victory 
Tenchebrai gave King Henry the whole duchy of Norman( 
He had used horse and foot combined, against isolated infantry 
and had been properly rewarded for his adherence to his father'.' 
example at Hastings.^ It is curious to see that it was the 

^ ** Consul Willelmus aciem Anglorum de loco in locum turbans promovit " {ibid.) 

2 William of Jumieges, p. 573. 

' Anglo-Saxon Chronicle, sub anno 1 106. The king shortly released him, thougV 
he condemned the others to perpetual bonds. 

* Orderic Vitalis, 701. 

° Matthew Paris (writing a hundred and fifty years after the fight) thinks tha 
Henry's ** English and Normans on foot" are a different body from the three corp; 
under Ralph of Bayeux, Robert of Melient, and William de Warenne. This is ar 
error, produced by misunderstanding Orderic's ** Primam aciem, etc. . . . Rex auten 
Normannos et Anglos pedites secum detinuit, Cenomannos et Britones longius in campc 
posuit." The three corps are the pedites. 



igk 
idv 



1 1 19] LOUIS VII. INVADES NORMANDY 381 

orother who had stayed at home, and not the brother who had 
Deen to the far East, that had best reahsed the military meaning 
d{ the experience of the first Crusade. 

Battle of Bremiile {Brenville) August 20, 1 1 19. 

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-Andelle, intending to cover the duchy from the destructive 
inroads of the French. On the 20th of August, the smoke rising 
from the burning barns of the monks of Bucheron ^ 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 Henr}^ 

^ The name of the place on which the abbey of Noyon-sur-Andelle was built. 



382 THE ART OF WAR IN THE MIDDLE AGES [11 19 

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.^ 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 ^ 
and Richard, which was strongest of the three.^ 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.^ 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.^ 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.^ 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. 

2 "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). 

* Orderic says that ' ' Ricardus filius regis et c milites equis insidentes ad belluqa 
parati erant : reliqui vero pedites cum rege in campo dimicabant " (p. 722). 

^ There were no more than the five hundred knights present on Henry's side. 
The "grand pleinte de sergeants" whom the Grands Croniqties de France introduce 
are an invention, as can be seen by carefully comparing them with Suger. 

^ 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 etfifl 
extraordinarie insiliat, ponit : milites armatos pedites deponit." 



;ii9] BATTLE OF BREMOLE 383 

)n which our authorities are hopelessly at issue, is that Orderic 
;ays that the horse were commanded by the king's sons, while 
Henry of Huntingdon says that they were led by the " proceres 
Normannorum," ie. the Counts of Eu and Warrenne, and that the 
-oyal bastards led the infantry reserve. We cannot hope to 
reconcile them on this point. The French can hardly be said 
;o have had any battle-array at all:^ they rode up in disorder 
in three troops, of which the first was headed by the Norman 
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.^ 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.^ 

^ Rex autem, nullum praelii 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. 

^ Suger speaks of the Vexin knights as being in the first charge : " Priores qui 
manum applicuerunt Velcassinenses primam Normannorum aciem ... 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." ^ Of the 
leaders of the two 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 J 
few years after Bremule and would have escaped notice but foi i 
its tactical interest,^ suffice to show that the combination oJ I 
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 1 124 Waleran Count of Mellent was in 
rebellion against King Henry, and had drawn his kinsmen 
Amaury Count of Evreux, and Hugh of Neuchatel, intc 
his plot. But the royal forces were too much for him ; most 
of his castles fell, and he and his knights became wanderen 
on the face of the land. He had been raiding near Bouri 

. .^Orderic, p. 722. 

2 M. Delpech must have the credit of bringing it into notice. 



II24] COMBAT OF BOURG TH^ROULDE ■'' 3^5 

Th^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- 
lain William of Tankerville, and Ralph of Bayeux.^ 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.^ Probably they were mercenaries, who had 
been furnished with a mount in order that they might be able 
to move rapidly along Avith the knights when pursuit was 
aeeded. There were forty of them in the party; these men 
Ralph and William placed on the left of their force, but thrown 
forward en potence, so that they would take in flank any body of 
i:ien which charged up the road.* They w^ere posted on the 
eft, in order that they might shoot at the unshielded right sides 
3f the rebels. Probably they dismounted in order that they 
night use their bows to better effect. Waleran of Mellent 
night have turned back and escaped by the way that he had 
:ome. But, as his adversaries had calculated, the desperate: 
:ount had no such intention. He harangued his companions and* 
:)ade them ride down the pack of "mercenaries and rustics"^ who 
lared to block the way. He himself, with forty knights of his 
neinie, headed the charge ; the rest, under the Count of Evreux,^ 

^ His pleasing habit was to cut ofF one foot of the peasants who fell into his. 
lands (Orderic, p. 740). 

^ Orderic and William of Jumieges speak as if Ralph had been in command, but- 
lenry of Huntingdon and William of INIalmesbury mention Tankerville only. 

^William, p. 576: *'Denique catervis more pugnantium, necnon et equitibus, 
agittariis (quorum inibi exercitus regis maximam multitudinem habebat) in dextra- 
•arte hostium praemissis . . . clamor utrinque attollitur." W^illiam is a tiresome and 
onfused author, but can hardly have gone wrong on a point like this. 

* "In prima enim fronte quadraginta architenentes caballos occiderunt, et ante- 
uam ferire possunt sunt dejecti " (Orderic, p. 740). 

^ " Gregarios et pagenses milites." 
25 



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 Hugh 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 III. was tc 
employ at Creqy 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 tht 
fruits of the day, argues a high degree of soldierly skill on the 
part of the victorious commanders. Horsed archers are rarel} 
found in Western Europe in the twelfth century : they were nc 
doubt the predecessors of the mounted crossbowmen of the 
time of John and Henry III. Such troops were called intc 
existence by the need of having men armed with missiles,- wh( 
could keep up with the cavalry in their rapid marches agains 
raiders. Foot-bowmen could not have intercepted Waleran' 
raid : but if provided with mounts of some sort, they migh 
reach the field ; they w^ould then leave their horses an< 
join the knights, who had also sent their chargers to th 
rear. 

Battle of Northallerton^ August 22, 1 138. 

The celebrated " Battle of the Standard " differs in characte 
from the other fights which we have been investigating, in th? 
the enemy was not the mailed and mounted chivalry of Franci 
but the wild hordes of Celtic tribesmen from beyond the Tweec 
We might have expected that the commanders of the Yorkshii 
levies would have endeavoured to turn their superiority i 
horse to good effect against the disorderly masses of Highlandei 
and Galwegians ; but as a matter of fact they dismounted ever 
rider, as Robert of Normandy had done at Tenchebrai, an 
the sole cavalry charge of the day was delivered by the sma 



1 138] 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 with Eadgar 
Atheling, or of the Norman settlers who had drifted in somewhat 
later. The Scots harried Northumberland and Durham with 
^reat ferocity, slaying the priest at the altar, and the babe at its 
mother's breast. Hence the Yorkshiremen looked upon the 
war as a crusade against savages, and marched out under the 
Danners of their saints, St. Peter of York, St. John of Beverley, 
ind St. Wilfrid of Ripon, all of which, together with that of St. 
Cuthbert of Durham, were placed on a chariot and borne in 
'.he midst of their host. The large majority of the English 
:onsisted 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 
^ho had brought their contingents from shires south of the 
Humber, such as Derby and Nottingham.^ The chief person 
Dresent was the young William of Albemarle, but Walter I'Espec, 
Sheriff of Yorkshire, seems to have shared the command with 
lim. They drew up their whole force in one deep line along a 
lillside on Cowton Moor near Northallerton, with the chariot 
rearing the standards in the rear of their centre. The knights 
ill dismounted and served on foot with the shire - levies, 
ipparently forming a mailed front line behind which the 
lalf-armed country-folk arrayed themselves. There were a 
:onsiderable number of archers among the Yorkshiremen, who 
ire said to have been " mixed " with the spearmen. Presumably 
hey stood in the mass and shot over their friends* heads, down 
he slope, for there is no statement that they ^took position 
either on the flank or in front of the main body. Some of 
he elder knights formed a sacred band in reserve around the 
standard: among them stood the commanders of the host, 
\lbemarle and L'Espec.^ 

* See John of Hexham, p. 262, for the men from Derby and Notts. 

* Richard of Hexham, p. 322 ; Aelred of Rivaulx, p. 343. 



388 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 Gallov/ay 
armed with nothing more than a dart, a target, and broad- 
sword. Seeing the solid mass of the English awaiting him or 
foot, David resolved to assail them with their own tactics, anc 
ordered his knights to dismount and form the head of the 
attacking column, while his archers were to advance along witl 
them. The rest of the host was to follow, and to try to break ii 
when the knights made a gap in the English front.^ 

But David had forgotten to reckon with the pride anc 
headlong courage of his Celtic subjects : they refused to let th( 
Lowland knights strike the first blow. The leaders of tb 
Galloway Picts claimed that they had an ancient right to taki 
the front place, and the Highlanders refused to give precedence 
to the Norman and English strangers.^ When the kin; 
persisted in his design, Malise Earl of Strathern, one of th 
chiefs from beyond the Forth, angrily exclaimed, " Why trus 
so much, my king, to the goodwill of these Frenchmen ? Non 
of them, for all his mail, will go so far to the front as I, wh 1 
fight unarmoured in to-day's battle." At this the Norman, Ala 
Percy, cried, " That is a big word, and for your life you coul 
not make it good." The earl turned on him in wrath, and s 
hot an altercation burst out between the Highlanders and th 
Southern knights, that the king in despair withdrew his fin 
order of battle, and granted the Galloway men the forema 
place. 

In the second scheme the Scots were drawn out in foi 
masses : as far as we can follow Aelred of Rivaulx's descriptic 
of the array, the Galwegians were in the centre of the froi 
line, somewhat in advance. The two wings were formed, tl 
right by the king's son, Henry, with the greater part of tl 
knights of the Lowlands and the levies of Strathclyde ar 
Teviotdale, the left by the English of Lothian combined wi- 
the West Highland clans of Lorn, Argyle, and the Hebridt 

1 " Placuit ut quotquot aderat militum armatorum et sagittarii cunctum praeiw 
exercitum, quatenus armati armatos impeterent, milites congrederentur mililib' 
sagittae sagittis obviarent" (Aelred, p. 342). 

a Ibid, 342. 



[138] NORTHALLERTON: THE GALWEGIANS CHARGE 389 

^Cing David was in reserve, with the men of Moray and the 
Eastern Highlands: he also kept about him as a bodyguard a 
ew of his modest contingent of mailed knights.^ 

When the Scots drew near the hillside where the English 
vere arrayed, Robert Bruce, a Yorkshire baron, who held also 
he lordship of Annandale in Scotland, rode down to the 
lostile army and tried to induce the king to consent to terms 
)f peace. But the young knights about David's person taunted 
lobert as a traitor, so that he had to withdraw, solemnly 
Usavowing his feudal allegiance for Annandale ere he went. 

A moment later the Galloway men dashed at the English 
:entre, raising a terrible shout of " Albanach, Albanach ! " Their 
vild rush made the fyrd waver for a moment, but the knights 
allied and sustained the common folks, and restored the line 
without a moment's delay .^ The Galwegians soon came back 

the charge: they shivered their light darts on the serried 
ine of shields which the Yorkshire men opposed to them, and 
hen laid on with their claymores. But they could not break 
n a second time, and in the intervals between their charges 
he archery galled them sorely. Yet they furiously returned, 

many of them looking like hedgehogs with the shafts still 
sticking in their bodies,"^ to make one last bid for victory. 

At this moment Prince Henry and his corps moved in 
:ipon the English left wing. He and his few scores of knights 
ed the charge on horseback, the mass of Strathclyde men 
•allowing on foot. The charge was fairly delivered, and the 
. gallant prince with his horsemen hewed their way right through 
he line of the Yorkshire men till they came out at the back of 
he mass, scattering disorder all around them. Henry then 
aw the horses of the enemy, held by the grooms of the English 
:nighthood, a short way to the rear. He rode on to seize 
hem, thinking that the infantry of his corps would penetrate into 
he entry that he had made, and reckoning the battle as won. 

^ Richard of Hexham, whose account of the Scottish array is incomplete, only 
lys that the Galwegians were in front, the king and a bodyguard of English knights 

1 the mid-battle, while the clans were around him, " cetera barbaries circumfusa 
rat " (p. 322). 

* " Galwegensium cuneus tanto impetu irruit in australes, ut primes lancearios 
:ationem deserere compelleret, sed vi militum iterum repulsi in hostes animum et 
igorem resumunt " ( Aelred, p. 345). 

' Videres ut ericium spinis, sic Galwegensem sagittis undique circumseptum 
ihilominus vibrare gladium, et caeca amentia proruentem nunc hostem caedere, nunc 
nanem aerem cassis ictibus verberare" (Aelred, p. 345). 



390 THE ART OF WAR IN THE MIDDLE AGES [1138 

Herein he was sadly mistaken : he wasted but a few minutes 
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 whc 
strove to thrust themselves into it. Meanwhile, in the centre 
the fire and fury of the Galwegians was used up : leaving theii 
chiefs Donald and Ulgerich dead on the field, they dispersed 
and fled. On the Scottish left wing the men of Lothian anc 
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 Davie 
now ordered his reserve of Highlanders to advance, and spranj 
off his horse to lead it forward. But, seeing the disasters ii 
the front line, the fickle Celts began to melt off to right ant 
left, and David soon found himself alone with his small body 
guard of English and Norman knights. It was hopeless t< 
proceed, so he bade his standard - bearer turn back, and with 
drew to a neighbouring eminence, where there presentl; 
assembled round him the wrecks of his host. The mass lookec - 
so formidable that the Yorkshiremen dared not attack it, bu 
waited till it began to retreat. Then they followed at 
distance, slaying stragglers and taking many knights prisoners. 

Prince Henry, having (as we have seen) worked his way t | 
the very rear of the English line, was left in a position c 
desperate danger when the Scottish host broke and retiree 
He saved himself by a ready stratagem : he wheeled and face 
to the north, then, bidding the few knights around him thro^ 
off their badges^ and mingle with the advancing line of th 
enemy, he pushed on unobserved along with the English, an 
gradually passed through them. When safely in advance < 
their foremost ranks, he moved off at a moderate pace, so as nc 
to awaken suspicion, and finally got clear away, rejoining h; 
father by a circuitous route on the third day. The Sc61 
suffered very heavily in the fight, though the ten thousand c 
eleven thousand dead of which the chroniclers speak are onl 
one more instance of the usual mediaeval inability to deal wit 
high figures. It is more credible that of two hundred knighl 

^ " Projectis itaque signis quibus a caeteris dividimur, ipsis nos hostibus inferariSfW 
quasi insequentes cum iis." What were the signal Probably not coats-of-ann ' 
\vhich were only just coming into use, but some common token which the Scots sit 
all wearing to distinguish them from the English. ^'^ 



II38] 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,^ 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 

^ Aelred, p. 346. He gave it to a peasant by the way, saying, "Accipe quod 
mihi est oneri, tuae consulat necessitate " 



392 THE ART OF WAR IN THE MIDDLE AGES [1141 

and treachery,^ " so that they came home not like comra( 
but like very bitter enemies." 

First Battle of Li7tcoln, February 2, 1 141. 

When we turn to the battle of Lincoln, we find ourselves 
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 1 140-41 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 ^ 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,^ 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 dangerou, 
the corps which Stephen had set to observe them was eas^ 
brushed away. 

Hearing of their approach, the king drew up his army in fr 
of the walls of Lincoln. In the absence of any precise indica- 



tne 

J 



•^ " Rex, recollectis suis qui sparsim de pugna, non ut consortes, sed potius sicut 
hostes inimicissimi fugerant, obsidionem apud Carham 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 the town "circa natale domini" (December 25), and was still 
before the castle on February i, when the enemy appeared. 

^ 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. 



i4i] LINCOLN: KING STEPHEN'S ARRAY 393 

ion of the battle spot, we have to put the following facts 
ogether in order to identify it (i) The earls forded the 
''ossdike somewhere west of Lincoln. (2) They fought with it 
t their backs, so that defeat meant disaster ; i.e. they faced 
orth or north-west. (3) The routed cavalry of Stephen's host 
scaped into the open country, not into the town ; i.e. they were 
rawn up so as to give a free flight to the north. (4) The infantry 
ed into the town, which was therefore quite close. Probably 
he battlefield lay due west of the city, and the Royalists 
pparently faced south or south - west. Stephen used the 
ictics which his uncle Henry I. had employed at Bremdle: 
lie greater part of his knights were ordered to dismount 
nd fight on foot around the royal standard ; with them 
'ere incorporated some infantry of the shire-levy, mainly 
omposed of the citizens of Lincoln.^ In front of this mass 
f dismounted men were drawn up two small " battles " of 
orsemen ; that on the left was headed by William of Albe- 
larle, whom the king had made an earl for his services 
t Northallerton, and by William of Ypres, a mercenary captain, 
'hat on the right was under a multitude of chiefs — Hugh Bigot 
^arl of Norfolk, William Earl of Warrenne, Simon of Senlis Earl 
f Northampton, Waleran Earl of Mellent,^ and the mercenary 
dan of Dinan, whom the king had created Earl of Richmond, 
lut these great names represented no great following ; several 
f them were pseudo-coinites, men whom the king had made 
arls in title, though their power and estates did not justify 
le promotion \^ it was said that they had no more connection 
ith the counties whose names they bore than that of receiving 
le third penny of the shire-fines. The rest had come to Lincoln 
ithout their full servitiiim debitum of knights, "as if to a 
olloquy, and not to a battle.""* The two squadrons between them 
rily mustered a very few hundred knights. 

The rebel earls likewise drew up their host in three main 
Drps. One was headed by Ralph of Chester, the second division 
y Robert of Gloucester, the third was composed of the numerous 

^ We get this fact from the speech of Earl Ralph in Henry of Huntingdon, 
ecapitulating the king's forces, he says: "Gives Lincolnienses, qui stant suae urbi 
oximi, in impetus gravedine ad domos suas transfugere videbitis " (p. 269). 

^ The vanquished rebel of the skirmish of Bourg Theroulde (see p. 384). 

^ "Paucosenimmilitessecum;f<r/2^/_^f<r/w«comitesadduxerant"(Gervase, p. 1354). 

* " Persuaserunt Seniores regi congregare exercitum, sese enim inermes ad regis 
Hoquium occurrisse, non ad praelii precinctum profitentes." 



394' THE ART OF WAR IN THE MIDDLE AGES [i 



I 



Midland knights and barons whose estates Stephen had decla 
confiscated for rebellion ; the chroniclers call this corps the 
" array of the disinherited." Robert had also brought with hin: 
from the marches a body of Welsh light infantry under two brother.' 
named Meredith and Cadwallader ; these wild levies, " courageou: 
rather than formidable," ^ as the chronicler calls them, were throwr 
out on the flank of the front line. Ralph of Chester and hi; 
knights dismounted and formed the reserve, incorporating witl 
themselves (just as Stephen had done) the remaining infantry 
of their host.^ In the front line the " disinherited " faced Bigot 
Mellent, Alan, and the other earls, while Robert and the Welsl 
were opposite Albemarle and William of Ypres. The numbe 
of horsemen on the two sides was about equal ; ^ the king ha( 
the advantage in foot-soldiery. 

The first clash came when the cavalry divisions of the fron 
line charged. On the one wing the " disinherited " completel; 
broke and scattered the five earls, whose whole squadron was h 
a moment either slain, captured, or in wild flight* On the othe 
flank William of Albemarle and William of Ypres came int 
collision with Gloucester's knights and the Welsh light infantrj 
The Royalists rode down the Welsh and drove them to tak 
shelter with the Earl of Chester and the barons' .reserve. But whe: 
they were assailed at once by Gloucester's horse and Chester 
mailed foot, they gave way, and the two Williams fled in rou 
as prompt and complete as that of the earls in the other win^ 
None of the beaten Royalist horse made any attempt to rally 
looking back on the field, William of Ypres observed that " th 
battle was lost, and that they must help the king some othe 
day," and continued his flight. 

Then the whole army of the rebel earls concentrated the 
efforts on the king's column of infantry ; apparently Chester an 
his dismounted knights charged it in front, while the " disir 
herited " and Gloucester beset it on the flanks and rear. Th 
Royalists made a gallant resistance, but at last the mass ws 
broken up ; those who could sought refuge within the gates ( 

^ "Audacia magis quam armis instruct!" (Huntingdon, 268). 

2 '* Animosam legionem Cestrensium peditum " (Orderic, 769). 

^ See Baldwin's speech in p. 272 of Henry of Huntingdon : " Nobis numerus 
equitibus non inferior, in peditibus confertior." This is more probable than Orderic 
'* hostes nimia multitudine peditum et Wallensium praevaluerunt " (769). 

* ** In ictu oculi dispersi sunt, et divisio eorum in tria devenit : alii namque occ 
sunt : alii capti ; alii aufugerunt" (Henry of Huntingdon, 273). 



PLATE XII. 



Tenchebrai 

Sept. 28. //06. 



Normans Cb ^ English . 



Vllliam Duke Robert 
gfMottaiM Robert ofBelesm 



cb cb cb 

Ralph Robert WUiam 
if B^WJJtofJWlenl. deVftname. 



Bremule 

Aug. 20.11 19 

King Louis ^^^ 
G.de Serranzl^J 
WXrispin W^ 



Robert & r&\ r-^ 
Richard ^^^ ^'^ 



Fbench ^ | ) 

En^bah cf] CfH *^"^^ Henry 



Northallerton 

AUG.22.//38. 

KlniPpaxid 
^ ^]^^ Lothian 



Scots ^ cS 
English (±1 



English 



is^Batde of 
JL IWCOLN 




Royalists [^ Cb 
Rebels 1^ ^ 



2"^ Battle of 
LlWCOLW 

MAY 19.1217, 




Pooll 

Place of Fawkcs de 
Breaute'».Sortie 

Die Blocked Gate. 

Place of lastfi^ht. 
and death of tlie 
Count of Perche 

Castle Bostem. . 

The loyalist Army. 



i4i] LINCOLN: THE KING CAPTURED 39.5. 

^incoln, where the foe promptly pursued them and cut them up 
n the streets. But Stephen and his truest followers stood firm 
)y the standard, and held out long after the rest of the fighting 
vas over. The king fought till his sword was broken, and then 
ised a Danish two-handed axe which a citizen of Lincoln 
lipped into his hand.^ His terrible strokes long held the rebels 
t bay, but at last a final rush swept down his faithful band, and 
le himself was thrown to the ground by William de Caimes, a 
)owerful knight, who caught him by the helmet and dragged 
lim over. With him were captured Bernard Baliol, Roger de 
viowbray, William Fossart, William Peverel, William de Clerfait, 
Baldwin Fitz-Gilbert, Richard Fitz-Urse, and many other gallant 
mights and barons.^ 

The first battle of Lincoln is a perfectly normal and typical 
hirteenth-century engagement. Each side used the same tactics 
)f a front line of horse and a reserve of dismounted knights : 
he Welsh light infantry on the rebel flank are the only unusual 
"eature, and they had no influence whatever on the event of the 
lay. Probably they were South Welsh archers, intended to gall 
he flank of the Royalist horse by a cross-fire, like the bowmen, 
it Bourg Theroulde in 11 24. Putting them aside, we see that 
he battle was lost because Stephen's cavalry were so dis- 
:omfited that they could not rally behind the reserve and 
•eturn to the fight. When they had left the field, the king's 
ate was sealed : like his uncle Robert at Tenchebrai, he found 
;hat infantry unsupported must fail before horse and foot 
:ombined. 

Of the reign of Henry II. even more than of the rest of the 
:welfth century is the statement true that the age was one of 
jieges rather than of battles. All through his reign the king was 
ighting hard, yet he was never present at an engagement of 
irst or even second-rate importance in the open field. Only twice 
.vas he even on the edge of a great battle — once at the raising 
3f the leaguer of Rouen in 1174, and once when, in 1187, he lay 
3y Ch^teauroux with a great host, while Philip of France on the 
3ther side of the Indre was drawing out his army day after day, 
md offering to fight if the Anglo-Normans should endeavour to 

^ John of Hexham, p. 269. 

* For the list see John of Hexham, p. 269. He is by far the most full in 
snumeration. 



39^ THE ART OF WAR IN THE MIDDLE AGES [1173 

pass the river. Both kings were prudent, and would not risl< 
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 enemie; 
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 a 
Alnwick was a curious instance of a great invasion stopped h] 
the chance encounter of a few hundred knights. The King o 
Scots had invaded Northumberland with an army not less thai 
that which his grandfather led to the Battle of the Standard 
He lay before Alnwick with part of his force, while the rest wer< 
raiding far and wide in the valleys of the Tyne and Tees. Mean 
while, Robert d'EstoutevilIe,the Sheriff of Yorkshire,had musterei 
the shire-levies of the great county, and the loyal barons of th 
north had gathered to his aid. They resolved to march toward 
Alnwick, but cautiously, since they knew that the Scots out 
numbered them fourfold. In the long march from Newcastl 
to Alnwick the knights outrode the weary infantry. On th 
morning of June 13, 11 74, they found themselves close to th 
beleaguered castle, but a heavy fog lay over the face of the lane 
and it seemed reckless for four hundred knights to try to pic 
their way between the besiegers' camps in the darkness. The 
attempted the dangerous feat, and were rewarded by an unex 
pected prize. When they had ridden some miles, the fog clearec 
and Alnwick was seen close at hand ; but closer still was a sma 
body of mailed men riding at leisure round the castle. It wa 
King William and a party of his knights : the rest were ou 
raiding or scattered in distant camps. The king at firj 
thought the English were some of his own host, and cantere 
unsuspiciously toward them. Only when he was too close t 
flee did he recognise the hostile banners : seeing his danger, h 
cried, " Now shall we see who is a true knight," ^ and, levelling hi 
lance, rode at the Yorkshiremen. This foolish feat of chivalrou 
daring had the natural result : his horse was slain, and h 
himself and all his companions were captured. His host brok 
up and retired in confusion into Scotland the moment that th 
disastrous news got abroad. Thus a great invasion was foile 

^ " Modo apparebit quis miles esse invenit" (William of Newbury, 185). 



173] BATTLE OF FORNHAM 397 

)y a trifling skirmish, in which less than five hundred knights 
00k part. 

Of the fight of Fornham (October 17, 1173), the other blow 
vhich crushed King Henry's enemies, we could wish that we had 
)etter details. The rebel Earl of Leicester was marching across 
Suffolk from Framlingham towards his own county with eighty 
cnights and three thousand Flemish mercenaries, horse and foot, 
vhom he had imported to strengthen his rebellion. To inter- 
:ept him, the Constable Humphrey de Bohun and the Earls of 
\rundel and Cornwall marched to Bury St. Edmunds with a few 
oyal knights and three hundred of King Henry's stipendiary 
lorsemen. The shire-levy of Suffolk and Cambridge joined 
hem in great force, for the Flemings had made themselves 
lated by their cruel ravages in Norfolk. They were reported to 
lave sung to each other, 

*'Hop, hop, Willeken, hop ! England is mine and lliine," 

md the fyrd came out readily against them, though many were 
irmed with nothing better than flails and pitchforks.^ The 
lost of the Constable outnumbered the rebels fourfold, but, as 
R.alph de Diceto remarks, if only properly armed men counted, 
:he earl had far the more formidable following.^ De Bohun, 
bllowing, caught him as he was passing a marsh near Fornham, 
md, falling upon him suddenly, discomfited the rebels in a few 
Tioments. Apparently the whole fight was a surprise, for the 
Flemings seem to have found themselves in a helpless plight, 
md Leicester and his knights fled early.^ The infuriated 
oeasantry gave no quarter, and thrust the foreigners into bog 
md ditch till more were drowned than slain with stroke of 
sword.* Only a very few survived to share the captivity of the 
:arl and his high-spirited countess, who had gone through the 
:ampaign 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 

^ 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 comite quam ex 
adverse." 

^ " In ictu cculi victus est comes et captus " (Hoveden, 307). 

* Jordan Fantosme, p. 294, line 109 1. 



398 THE ART OF WAR IN THE MIDDLE AGES [ii6 

allowed themselves to be attacked by the Royalists, made n 
attempt to take the offensive, and hardly stood for a momen 
If the ground had been firm and open, they must surely have ha 
the better of the fyrd. 



The English in Ireland, 1169-75. 

We have, as it chances, a far better knowledge of anothc 
set of Anglo-Norman fights than of those of the great rebellio 
of 1 173-74. The Expiignatio Hiberniae and the invaluable Son 
of Dermot and the Earl^ enable us to form a very clear notio 
of the tactics and strategy by which a few hundred knights c 
the Marches of Wales subdued within the space of five yeai 
the better half of Ireland. Of all the many conquests of tb 
Normans in East and West, this was perhaps the most astonisl 
ing, for the resources of the invaders were weaker even tha 
those of the conquerors of Naples and Sicily, and the Iris 
dwelt in one of the most difficult and inaccessible regions < 
Europe. 

Ireland in 1169 was one vast expanse of wood, bog, an 
mountain, in which the tracts of open land were few and fc 
between. Between every tribal settlement lay difficult pass( 
over marshes or between woods and rocks. The natives, 
fickle and ill compacted, were not wanting in wild courage, an 
had in their long wars with the Danes evolved a system < 
defensive warfare which was well adapted to the character ( 
their country. On every trackway which led from district 1 
district there were well - known positions which the tribesme 
were wont to fortify with considerable skill. In the bogs the 
dug trenches across the road and erected stockades on the farth< 
side, so that the passage was almost impassable for horseme 
In the forest tracts they " plashed the woods," i.e. cut down tl ; 
underwood and wattled it together in abattis across and alonj | 
side of the roads, so that those who tried to force their wa 
through found themselves beset on flank and front by unsee 
enemies, who could only be reached by hewing down tl 
screen of thick boughs. The Song of Detmot and the Ea 
is full of descriptions of barriers of these two kinds : tl 
account of the pass of Achadh-Ur (Freshford in Kilkenny 
may serve as an example. This was a passage between tl 

^ I have of course used Mr. Orpen's excellent edition of 1892. 



169] THE ENGLISH IN IRELAND 399 

iver Nuenna and steep wooded hills. Mac-Donnchadh, king 
f Ossory — 

"Un fosse fist jeter aitant 
Haut e large roist e grant. 
Puis par afin ficher 
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 
nd large, and to strengthen it at the back with stakes and in 
rent with hurdles, in order to dispute the passage of King 
)ermot the stout-hearted" (lines 1013-19). 

Whenever the English marched out, the Irish "plashed the 
'oods and dug across the roads" (line 1 595), and it was hard to get 
rom place to place " on the hard field and by the open ground." 
>uch tactics were most distressing to jnvaders accustomed to win 
y the ponderous charge of mailed cavalry across the unenclosed 
.elds and hillsides of England or Normandy. Yet, as we shall 
ee, they succeeded in triumphing over these difficulties, and 
•:rmly estabHshed themselves in the conquered land. 

The weak point of the Irish was their want of defensive 
.rmour and their inability to stand firm in the open. If once 
-he enemy could close with them, and catch them far from the 
helter of stockade and trench, they were easy to deal with, for 
hey dreaded above all things the impact of the mailed horse- 
nan, and had never learned to stand fast, shoulder to shoulder, 
.nd beat off the charge of cavalry. Neither they themselves 
lor their old enemies the Danes were accustomed to fight on 
lorseback, and they were utterly cowed by the Norman knight 
.nd his reckless onset. Their arms, indeed, were very unsuited 
o resist cavalry : only the Scandinavian settlers of the coast- 
owns and a few of the chiefs of the inland wore mail ; the rest 
:ame 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."^ 

Nor were the offensive arms of the Celts very suitable for 
epelling cavalry ; they carried two darts, a short spear, and 

^ Poem of Gilla Bhrighde M'Conmidhe, quoted by Mr. Orpen in Dermot at: d the 
^arl, p. 268. 



400 THE ART OF WAR IN THE MIDDLE AGES [116 

large-headed axes wielded by one hand, but had no long pike 
nor any skill in archery.^ They hurled darts and stones 2 
close quarters from behind their stockades and fosses, but coul 
not keep off their enemy by the distant rain of arrows. In shor 
they were formidable while skirmishing in woods and bogs, bi 
easily to be routed in the open. 

The Anglo-Norman leaders soon learned to adapt the: 
tactics to those of the enemy. They had to avoid, as far a 
possible, fights in woods or bogs, and to lure the enemy into th 
clear ground. If this was impossible, and if the Irish stood fin 
behind their defences, the only courses open were either to essa 
surprises and night attacks — the Celts habitually kept a ver 
poor watch — or to gall the defenders with arrows from a di: 
tance. Fortunately for themselves, the knights of the Wels 
March had close to their hand the very associates most suited t 
aid them in such difficulties. The men of South Wales were tl" 
most skilled of all the inhabitants of Britain in archery, and dre 
the longest and the strongest bows. It was by their aid that tl 
invaders were accustomed to triumph over the Irish horde 
None of the barons who won Ireland ever marched forth withoi 
a large provision of bowmen, and after a time they habitual 
mounted them, in order that they might be able to keep up wil 
the knights in every chance of war, and might not be left behir 
in rapid advances or pursuits. Giraldus Cambrensis in h 
Expugnatio devotes the best part of a chapter to explaining tl 
advantage which the Welsh archers gave to the invaders, ar 
urges the leaders of his own day to enlarge the proportion 
Welsh in their bands,^ on account of their lightness and swii 
ness, which enabled them to follow the Irish into heavy or mou 
tainous ground, where the mailed men could pursue only slow 
or not at all. A few descriptions of battles will show how tl 
Anglo-Normans contrived to deal with their adversaries. 

Battle on the Dining 1 169. 

Dermot of Leinster, with his allies, Robert Fitz-Stephen ai 
Maurice de Prendergast, had executed a successful raid in 
the lands of his enemy MacDonnchadh, King of Ossory. Thi 
had with them three hundred knights and archers of Wales, ai 

^ Topographia Hiherniae of Giraldus Camb. p. 151. 

-See his Expugnatio^ book ii. chapter xxxviii. : "Qualiter gens Hibern 
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, 
md a move onward to seize the well-manned pass seemed 
equally hopeless. 

Maurice de Prendergast at once proposed a retreat from the 
/alley and the woods up to the high open ground from which 
:he army had descended in order to attempt the pass. If the 
nen of Ossory should follow them, as was likely, it would be 
Dossible to turn upon them where neither trees nor marsh pre- 
lected them from the charge of the Norman horse. His advice 
.vas promptly carried out ; the Anglo-Normans retired up the 
lillside with every sign of hurry and dismay. When they began 
:o approach the end of the wood, they dropped forty archers 
jnder a certain Robert Smiche (Smithe ?) by the wayside, with 
orders to hide in a thicket till the Irish should have passed 
:)y, 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 
Dssory " came out in haste from their impregnable position, and 
bllowed them across the valley and up the hill. They passed 
:he ambush without noticing it, and swept out into the open 
ground. When they had left the wood some way behind, they 
vere surprised to see the Normans turn and form line of battle. 
Before the meaning of the movement was realised, the knights 
:harged in among them, the archers and sergeants following 
:lose behind. The Ossory men were six or seven to one, — their 
lumbers are given at from seventeen hundred to two thousand ^ 

^ In line 659 the author of Dermot and the ^ar/ calls them "mil e set scent," 
)ut in7i8 "par aime erent ii millers." Neither figure seems too high, considering 
he usual exaggeration of the mediaeval poet, 
26 



Ih 



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 o 
archers fell upon the fugitives with much effect we are not told 
but the cowardly men of Hy-Kinselagh emerged from thewood.^ 
where they had been skulking, and hunted the fugitives for somt 
distance. They brought back two hundred and twenty heads- 
no quarter was given in Irish war — and laid them at Kinj 
Dermot's feet. To the horror of his auxiliaries, the brutal kin^ 
was seen ^ 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 ai 
old Norman device, whose most famous example was seen a 
Hastings. Without its use the army of Dermot and Fitz-Stephei 
must have been crushed in the valley between the marsh an< 
the wood, where no cavalry charge would have been possible. 

The next two engagements which we must notice were bot 
fought close to the walls of Dublin,^ which had fallen into th 
hands of the English in the autumn of 1 170, its Danish lore 
Haskulf Thorgilson, having been expelled and driven to see 
refuge in the Western Isles. Richard de Clare, the famou 
" Strongbow," was now at the head of the invaders, and had lai 
claim to the whole kingdom of Leinster, since the death of hi 
father-in-law, King Dermot, in May 1171. It was only a fori 
night after his accession that a Viking fleet cast anchor in Dubli 
Bay. Haskulf had sought aid from the Scandinavian settlei 
in Man, Orkney, and the Hebrides, and had gathered a fleet ( 
sixty sail to restore him to his lost possessions. His auxiliarie 
were led by an adventurer named John " the Madman " or "JJB 
Furious,"^ a famous "Baresark," who had won much glorj^B 
the wars of the North. The Norsemen landed, ten thousan 
strong, or even more, according to the estimate of their enemie 
which must be wholly futile : Orkney and Man could not ha\ 
supplied half that number of warriors. They formed up on tl 

^ Giraldus, Expvgnatio, \. 4. The author of Dermot and the Earl does not g\ 
this discreditable trait of his hero's conduct. 

2 It is strange to find that Giraldus and the author of Dermot differ as to t 
.order of the two sieges : Giraldus puts the Danish siege in May and the Irish sie 
in June, while the poet makes the Danish siege so late as September, three mont 
;9.fter Roderick's. 

'Joannes "Insanus" or "Vehemens" or " Le Wode" in Giraldus (p. 26^ 
The Song of Dermot calls him Jean le Deve (from desver, to go mad). 



ii7i] BATILE OF DUBLIN 403 

shore and marched toward the city in a solid column, all clad 
n mail-shirts and bearing their Danish axes on their shoulders. 
This was a host very different from the hordes of naked Irish 
.vith whom the invaders had hitherto had to cope, and far more 
'ormidable. 

Battle of Dublin, May 1171. 

Miles Cogan was in command of Dublin in the absence of 
lis master. Earl Richard. He had with him about three 
lundred mounted men,^ besides archers and sergeants on foot, 
probably fifteen hundred men in all, if the infantry bore to the 
;avalry the proportion that was usual in the bands with which 
;he Anglo-Normans overran Ireland. Miles came out at first 
nto the open, with his archers and spearmen in front and his 
•cnights in second line. But he was unable to break into the 
V^iking ranks, and was forced back against the eastern gate of 
Dublin (St. Mary's Gate or Dame's Gate). Foreseeing that this 
■night occur, he had previously detached his brother, Richard 
3ogan, with thirty knights, to issue from the town by its western 
jate (Newgate), fetch a compass around the walls, and fall on 
:he rear of the enemy. The main body of the English was 
rarely holding its own about the east gate when a shout from 
;he back of the Viking host told them that the diversion had 
3egun. Richard and his knights had made a desperate charge 
nto the rear ranks of the Norsemen. " When John the Wode 
jcented the noise of those behind and the shouting, he departed 
rom the city, he wished to succour his friends who were left 
)ehind ; John and his meinie, ten thousand strong or nine (I 
mow not which), departed from the city to succour their com- 
panions in the rear."^ 

The diversion, trifling as it was, had checked the Norse attack, 
md in the confused movement towards the rear the solid column 
lad been broken up, and gaps showed in it. Miles and the main 
)ody of the English, horse and foot, threw themselves upon the 
nass. The knights succeeded in penetrating into the heart of 
he column, and wrought so much damage among the Vikings 
hat they began to retire in disorder towards their ships. John 
he Wode refused to fly, and fought with astonishing strength 
md courage ; he struck one knight such a fearful blow with his 
wo-handed axe that he hewed off his thigh in spite of hauberk 

^ Songof Dermot, line 2384. " Song of Dermot^ lines 2375-80. 



;4tSH 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 • 
at the same time, and the Vikings ere long fled in complete 
rout. Some Irish levies of uncertain loyalty under one Domnah 
Macgille Moholmog- 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 rushec 
down and aided in the slaughter of the fugitives. Two thousanc 
were slain and five hundred drowned on the beach before th( 
survivors succeeded in thrusting their galleys out to sea anc 
getting into the offing.^ 

Surprise of Castle Knocks July 1 171. 

Only a month after the Vikings had been beaten, anothe 
army appeared under the walls of Dublin. This time it wa 
Roderic O'Connor, the high-kjng of all Ireland, with sixt,^ 
thousand men levied from all the clans of the island. The^ 
encamped around Dublin in four separate bodies — the high-kinj 
and his men of Connaught at Castle Knock ; Macdunlevy anc 
the clans of Ulster at Clontarf — the site of Brian Boroihme' 
old victory ; O'Brien of Thomond at Kilmainham ; and Murtoug' 
M'Murrough with the men of Leinster at Dalkey. Earl Richar 
had by this time returned to his capital and taken over th 
command from Miles Cogan, but he was in despair at the ovei 
whelming strength of the array which O'Connor had brougl: 
out against him, and did not dare to stir from the walls. Aftc 
a siege of six weeks, famine began to threaten the garrisoi 
"The measure of wheat was sold for a silver mark, and th 
measure of barley for half a mark."* Nor was there any hop 

* He was beheaded after the fight. He had been reserved for ransor 
but so angered his captors by his haughty answers that they slew him (Girald 
p. 265). 

2 The Song of Derinot tells us that Miles Cogan, knowing Domnahl's ficklene^ 
had bade him stand afar off and strike in against the losers. " If these men be d: 
comfited, then you shall aid us with your force to overthrow them. But if we 
recreant, then you shall aid these men to cut us to pieces and slay us." To tl 
the Irishman readily consented (lines 2300-2310). 

2 The 5<?;/^ ^ Z^^r;;/^/ says that two thousand Norsemen escaped, two thousai 
were slain, and five hundred drowned. This would give a total of four thousand fi 
hundred for their army — a far more probable figure than the nine thousand or t 
thousand given above, or the impossible twenty thousand which is also attributed 
the Vikings. 

* Song ofDermot^ lines l82S-3a 



i7i] STRONGBOW'S SORTIE FROM DUBLHTOli' .^ 

)f bringing in provisions by water, for Guthred, King of Man, 
vas lying in the bay with a Viking fleet — the relics, no doubt, of 
he armament of John the Wode. 

Richard endeavoured, therefore, to make peace with King 
:^oderic, offering to hold Leinster as his vassal and do fealty to 
lim. But O'Connor replied that he might hold the three towns 
)f the Ostmen, Dublin, Waterford, and Wexford, but not a foot 
nore. These terms appeared so hard to Earl Richard that he 
esolved to hazard a sortie, in spite of the desperate odds against 
lim. On the very afternoon of the abortive negotiations he 
narshalled the forces which could be spared from garrisoning 
he ramparts, and marched out against the camp at Castle 
vnock (five miles from Dublin) in three small columns. Each 
vas composed of forty knights, sixty mounted archers,^ and a 
lundred sergeants on foot.^ Miles led the first, Raymond Le 
jros the second, and the earl himself the third. They hurried 
it full speed from the west gate and reached the camp of the 
nen of Connaught before the alarm was given. The Irish were 
:aught entirely unprepared ; they were lounging about their 
•abins and huts, and the king himself was in his bath. They 
lad surrounded their encampment with a stockade, but no one 
vas in arms to guard it. The invaders broke in easily at three 
)oints, and rode through the lanes between the huts, hacking 
md hewing at every band that strove to concentrate against 
hem. In a few minutes the fight was over, for the Irish broke 
ip and ran off with disgraceful alacrity, the king, all naked 
rom his bath, leading the flight. Fifteen hundred were slain, 
vhile the English only lost one single sergeant. On hearing 
)f Roderic's defeat, the Irish in the other three camps dispersed 
.nd went homeward, and the siege was raised (July 1171). 

Thus ended a fight which bears a strong similarity to 
.nother sortie made by an English garrison from Dublin, five 
lundred years later. Colonel Michael Jones in 1649 was be- 
eaguered like Earl Richard by a vastly superior host dispersed 
n several distant camps. Like the earl, he hazarded a sortie 
.gainst one of the hostile corps, and was successful in surprising 
.nd dispersing it. And when Ormond's men had been routed 

^ That the archers were mounted seems to follow from the correction of * * satellites 
questres" for "arcarii" in the later texts of Giraldus, i. xxiv. 

' Giraldus makes the first two columns led by only twenty and thirty knights 
espectively, and says that Raymond rode before instead of after Miles Cogan, 



4o6 THE ART OF WAR IN THE MIDDLE AGES [1180 

at Bagotsrath, the other Irish divisions dispersed and retired 
without fighting.^ The rebels of 1649 were as divided in their 
counsels and as chary of giving each other prompt aid as the 
levies of 117 1. 

The three battles which we have thus set forth give us the thre( 
main tactical devices by which the Normans won their victories— 
the feigned retreat, the flank attack by horsemen, and the sudder 
surprise. After three years of fighting, the Irish were so cowec 
that they habitually retired to wood or bog when the invader 
advanced, and never fought save in night surprises or behind im 
pregnable stockades and ditches. These defensive tactics hande( 
over the open country to the conquerors, who forthwith secure* 
it by erecting castles everywhere, structures against which th 
Irish could seldom prevail — indeed, a castle, when once completec 
never fell save by treachery. On the other hand, the Angle 
Normans were almost equally incapable of mastering the wood 
and bogs in which their enemies took refuge. Hence cam 
that unhappy division of the island, destined to last for foi 
centuries and more, in which the natives held out in the 
fastnesses, while the invaders dominated the open land — eac 
levying unending war on the other, yet neither able to get the ac 
vantage. The land could make no progress, and in the sixteent 
century the natives were as barbarous as in the eleventh, whi 
the invaders had almost sunk to their level, instead of advancir 
in civilisation parallel to the English and the other nations < 
Western Europe. The wars of Elizabeth's day in Irelar 
exhibit the " mere Irish " absolutely unchanged from the 
ancestors of the twelfth century : their primitive tactics, the 
arms, their plashed woods and wattled stockades are absolute 
the same as those of the days of Strongbow. Except that son 
of their chiefs had learned to ride ^ to battle, we see no change 

^ Ormond was caught in bed — just as Roderic O'Connor was caught in his Ix 
by the sortie party. 

2 And that as early as the fourteenth century, as is shown by the description 
the Irish by the captive squire in Froissart, p^. p. 429. 

• hrtfv ,?<^ 



CHAPTER IV 

ENGLISH BATTLES OF THE THIRTEENTH CENTURY 

Second Battle of L incoln (121 7) — Taillebourg (1242) 
— Leives (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, 12 17) 
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 12 17 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 



4o8 THE ART OF WAR IN THE MIDDLE AGES [121; 

at-Gowts, beyond the Witham. The besiegers lay within the 
walls, and pressed the siege by battering the south and eas 
sides of the castle with perrieres. They had shaken part of th( 
curtain, and hoped to see the battlements crumble within a fev 
days. 

The Royalist army mustered at Newark under Willian 
Marshall, the Earl of Pembroke: he had with him the Earl 
of Chester, Salisbury, and Derby, and the greater part of thi 
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, witl 
three hundred and seventeen crossbowmen and a considerabl 
number of foot-sergeants.^ They marched from Newark north 
westward when they heard of the straits to which the castl 
was reduced, and slept on the night of the i8th at Torksey an( 
the neighbouring village of Stow, some nine miles fron 
Lincoln. From thence they ascended the high ground alonj 
which the Roman road (Ermine Street) runs, and move< 
cautiously toward the north front of the city. This route gav 
them a chance of communicating with the castle, unless th 
enemy should choose to fight at a considerable distance from th 
walls. The host was marshalled in four - " battles," — the first le( 
by the Earl of Chester, the second by the Marshal, the thin 
by the Earl of Salisbury, the fourth by Peter des Roches, Bisho] 
of Winchester, the most unpopular but the most able of the lat 
king's foreign favourites. The crossbowmen under Fawkes d 
Breaut^ moved a mile in front of the knights. The baggag 
with a guard of infantry followed, the same distance in the rea 
of the four corps of cavalry.^ 

Second Battle of Lincoln^ June 19, 1 2 1 7. 

The besiegers of Lincoln received timely warning of the ap 

proach of the relieving army, and sent out Saher de Quincey, Ear 

of Winchester, and Robert Fitzwalter to reconnoitre the advancin: 

columns. They soon returned with the report that the Royalist 

^ So the Song of William *the Marshal, 16264-8. Matthew Paris (p. i 8) says foi 
hundred knights and two hundred and fifty crossbowmen, as also " multi servientt 
qui vices militum possent pro necessitate implere." 

2 Matthew Paris (p. 19) says seven "battles," but the Song of William th 
Marshal is so clear and full that it would be dangerous to refuse to follow it and'^t 
choose the later authority, r-r'n > <-.» iivi/j '« ,ijji>f4 ^ii. 

3 Matthew Paris, p. I^g ^rfj qJ TjnibBOl (ejjbilH 

TO* 



I? J 7] THE ROYALISTS APPROACH LINCOLN 409 

•.eemed somewhat weaker than themselves, and that it would be 
idvisable to attack them in the open, far from the city, in order 
:hat they might not be able to communicate with the garrison 
)f the castle. The estimate was not far wrong, as the besieging 
irmy counted six hundred and eleven knights and a thousand 
bot-sergeants,^ a force decidedly superior to the Marshal's host. 
3ut the Count of Perche, who commanded the French contingent 
n the rebel army,^ insisted on going forth in person to take a 
econd view of the enemy, before committing himself to a battle. 
Vlistaking the distant baggage-guard and its column of sumpter- 
)easts and waggons for an integral part of the Royalist army, he 
:ame back with a firm belief that he was largely outnumbered, 
md insisted on keeping his men within the walls of the city, and 
aking the defensive.^ This line of tactics seemed to promise 
ibsolute security, since it appeared impossible that the very 
nodest host of the Earl Marshal would be able to do serious 
larm to the rebels, when the latter were covered by the strong 
brtifications of Lincoln. The storming of a city or castle by 
nain force and without a long preparatory leaguer was an almost 
mknown thing in thirteenth - century warfare. Accordingly 
he barons continued their operations against the castle, and set 
heir machines to play upon its walls with redoubled energy. 
The only precaution which they took against the relieving army 
vas to tell off detachments to guard the four gates by which 
he Marshal might attempt to enter the city, — the north gate 
vhich lay immediately opposite him, the east gate and Potter's 
jate on the right flank, and the Ncwland gate on the extreme 
eft between the castle and the river Witham. It cannot have 
:scaped the notice of the commanders of the baronial army 
hat their tactics allowed free communication between the castle 
.nd the Royalists, and that it was possible for the Marshal to 
:nter the castle and sally forth into the town by the great gate 
n its eastern curtain. But this exit was well guarded by the 
letachment told off to operate against the castle, and such a 

* William the Marshal, 16336-9. 

^ The chroniclers only preserve the names of three of the French chiefs in the 
ost, though the French contingent seems to have been strong. These are the 
bant of Perche, the Marshal Walter of Nismes (Matt. Paris, p. 20), and the Chatelain 
f Arras {Song of William the Marshal, 16607). 

^ Matthew Paris, p. 19. He says that the barons had left many standards with 
le baggage-guard, and that their appearance misled the count into taking it for a 
eserve corps in the rear of the Royalist line of battle. 



4r6 THE ART OF WAR IN THE MIDDLE AGES [i2t: 

sally on a narrow front appeared to present no very grea 
danger. Any transference of troops from the relieving arm} 
into the castle must take place under the very eyes of th< 
defenders, and could be easily provided against by a corre 
sponding shifting of their own forces. 

When William the Marshal and his host approached Lincolr 
they were somewhat surprised to find that the enemy woul 
not come out to meet them in the open. Drawing up at 
cautious distance from the city, they proceeded to communicat 
with the castle. John Marshall, the earl's nephew, swept roun^ 
the north-west corner of the place with a small party, an 
entered the castle by its postern gate. He learned that th 
garrison were reduced to great straits, and bore back th 
message to his uncle. On leaving the postern he was pursue 
by a party of rebel knights who issued from the Newland gat 
to chase him, but outrode them and reached the main army i 
safety.^ 

The Marshal then resolved to send into the castle Bisho 
Peter, who was renowned for his good military eye, that h 
might decide whether the proper course of action would be t 
throw troops into the castle and sally forth from it, or to attac 
the gates and the city. The bishop made a rapid survey ( 
the place, and fixed his main attention on the point where t\ 
castle joined the north-west front of the town wall. Here thei 
lay, quite unguarded, and close under the castle, so as to t 
swept by its fire, an old blocked-up gate, on which tl 
barons had set no guard.^ Pie bade a party of the garrisc 
steal out and tear down the stones which closed the gate, i 
as to make an opening in this unguarded front. Meanwhile, 1 

^ Song of William the Marshal, 16438-40. 
^ *' Une vielle porte choisi 

Qui ert de grand antiqnite 

Et qui les murs de la cite 

Joigniet avec eels del chastel, 

Mes el fut anciennement 

Close de piere e de ciment. 

Quand. li evesques ont veiie, 

La fist abbatre et trebuchier 
E que Tost veist et seiist 
Que seiire entree i eiist " (16509-17). 
This gate must have been that generally known as Westgate ; it must have be 
rendered comparatively useless when the castle-building destroyed the north w 
houses of the town, and was temporarily blocked up. 



21 



7] DE BRfiAUTE'S SORTIE 411 



apidly returned to the Marshal, and advised him to throw part 
)f his men into the castle and make a sally from it, but to direct 
lis 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 do 
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 
>vere slain, and the whole body thrown into confusion. Fawkes 
:hen sallied out with his troops and made a vigorous attack on 
:he besiegers, but they were too many for him, and he was 
jeaten back into the castle with loss.^ 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 
?vlarshal 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.^ 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.^ 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.^ Having thus 

^ Matthew Paris, p. 21. 

^ That they were among houses seems to follow from line 1 6600 of the Song of 
William the Marshal^ where the earl before charging says — 
" Attendez niei a cest ostal 
Tant que j'ai mon helme pris." 
^ The assailants (line 16657) 

*' Entra sis filz en la cite 
Par 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. 
* Line 16633. 



4Ii^ THE ART OF WAR IN THE MIDDLE AGES [121.7 

won an entry into the place, the earl pushed his men through 
the breach into the streets as fast as he was able. They coulc 
not advance with ease, for the barons had rallied and massec 
their forces against the assailants, who were obliged to advance 
on a narrow front down the tortuous lanes of the town, anc 
could not deploy. A fierce jousting took place in all the north 
western streets of Lincoln, and it was only by very vigorous 
fighting 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 nea 
the castle was occupied. The enemy fell back, part along th< 
high ground towards the cathedral and the north-east quarte 
of the place, part down the broad street leading to the bridge a 
Wigford and the south gate. In the open space before tb 
minster the Count of Perche rallied the best knights of th 
baronial army, and made head for some time against th" 
Marshal and the main column of the Royalists. At last his mei 
gave way, and he himself was surrounded ; he was offeree 
quarter, but " would not yield to any traitor Englishmen," ^ am 
was slain by a thrust which pierced the eyehole of his helnr 
After his fall the rebels lost heart and rapidly gave ground 
some flying by the east gate, others southward towards th- 
river and the bridge. At both exits there was soon a cro\v( 
massed in hopeless confusion, the passages being too narrow tt 
allow so many fugitives to pass out at once. The south gat' 
had a swing door, which closed automatically after each passer 
by pushed it open ;2 the east gate is said to have been jamme( 
on a frantic cow which got mixed with the horsemen.^ HenC' 
the pursuers were able to make prisoners of an enormou 
proportion of the rebel knights and barons. About four hundrec 
in all out of the six hundred and eleven who had engaged in th 
battle were captured. They included three earls, Saher de Quince; 
of Winchester, Henry de Bohun of Hereford, and Gilbert d' 
Gand of Lincoln. Among the other captives were several of th 
twenty-five signatories of the Great Charter. The slaughter, 01 

^ "Juramento horribili affirmavit quod se Anglico alicui nequaquam reddert 
qui propri regis proditores fuerunt" (Matt, Paris, pp. 21, 22). 
,*:«).<■« Matt. Paris, p. 22. 

'jdt « This was the east gate ; the poem of William the Marshal describes it as "th 
one that leads" " dreit a I'Hospital," i.e. St. Giles' Hospital, founded by Remigim 
outside the east gale (line 16943). 



I 

1217] 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 ; 
3f the vanquished, only the Count of Perche and one other 
knight are recorded as slain, though many of the foot-soldiery 
DU 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 

irery high idea of the tactical accomplishments of either side. 

; The arrangements made by the rebels were ill conceived and 

:arelessly carried out. Their neglect to watch the blocked 

7ate is most extraordinary, and, even when it was forced, they 

\ night have had a good chance of victory if they had barricaded 

I ;he streets and fought on foot, instead of endeavouring to 

j :xpel the Royalists by cavalry charges. vd 

\ To the victors the only praise that we can give is that they 

\ ^new how to utilise a false attack in order to distract attention 

Vom the real one. Bishop Peter must apparently take more 

:redit for the plan adopted than the Marshal ; the poem written 

n praise of the latter ascribes the idea to the Churchman, and 

)nly the execution of it to the earl — a piece of evidence 

:onclusive as to the attribution of the design, for William's 

[encomiast would certainly have claimed the glory for his 
lero had he been able to do so. The details of the fisrhtingf 
\ ifter the breach was once forced show nothing but hard blows ; 
^ .ve have no evidence that the crossbowmen were used in the 
street fighting, as they well might have been, or that the enemy 
.vere evicted by flanking movements by side streets. AH 
ipparently was done by vigorous jousting down the main 
:horoughfares and in the open space by the minster. 

Nearly fifty years elapsed before Englishmen fought another 
Dattle on English soil, and we shall see, when we pass on to 
nvestigate Lewes and Evesham, that the art of war had moved 
:>n considerably in the interval. But there is no material for us 
;o use in filling up the gap save the insignificant battle of 
laillebourg, where the imbecile Henry III. allowed himself to 
DC defeated by Louis IX., a general whose strategy we have 
earned not to admire in studying the campaign of Mansourah.^ 
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 

'- See p. 339. 



414 THE ART OF WAR IN THE MIDDLE AGES [124 

crossbovvmen, and the general levy of the towns of Guicnne, la; 
on one bank of the Charente near Taillebourg. The army wa 
almost wholly composed of his continental vassals ; only eight; 
English knights were present. Louis, with a much larger force 
appeared on the other side : the river was broad and swift, an' 
there appeared to be no means of crossing save the bridge, wher 
Henry set a strong guard. Relying on the safety of his positioi 
he kept no proper watch on the enemy. St. Louis determine 
to risk an attempt to force the passage, and prepared for tha 
purpose a number of large boats. He then vigorously attacke 
the bridge, and at the same time threw across a body c 
crossbowmen, dismounted knights, and sergeants by means of hi 
vessels. The guards of the bridge, fearing to be attacked behin 
by the newcomers, gave ground, and so allowed the main body ( 
the French to evict them from the passage they were sent to guan 
When King Henry saw the bridge lost, he did not make an 
attempt to fall on the small part of the French army which ha 
crossed, but drew off and sent his brother Richard to ask for 
truce. It was granted, and under cover of it he withdrew at nigh 
fall with shameful haste, abandoning his camp and baggage. 

A capable commander would have had his army in orde 
would not have been caught off his guard, and would have falle 
on the French van when it had passed the bridge, and ove 
whelmed it before the main body could come to its aid. Sue 
were the tactics employed in a similar case by Wallace at tl 
battle of Stirling Bridge.^ But Henry was the most helple. 
and imbecile of leaders, and threw away his chances in the mo 
faint-hearted manner. At the moment that he sent to ask for 
truce, the number of French who were over the river did n^ 
amount to a tithe of his own army, yet he parleyed instead 
charging.2 If Louis had not listened to his demand, he wou 
probably have given the signal for flight at once, and wou 
have got off in even worse plight than was actually the case. 

Lewes and Evesham show a distinct advance in the art 
war, which we may fairly set down to the influence of Simc 
de Montfort, who, though not a general of the first class, had 

1 See p. 563. 

2 Joinville says that there were " not one hundred part as many" French tree 
over the bridge as Henry mustered. Matthew Paris conceals the facts of the d 
graceful skirmish in a way not creditable to his veracity, when we consider what 
capable writer he was and how fully he tells the rest of the campaign. 



264] KING HENRY TAKES NORTHAMPTON 415 

east a quick eye and a wide experience. He had been brought 
ip on the traditions of Muret and the rest of his father's victories, 
ie had himself seen several campaigns both on the Continent 
md in the East. Though not an innovator, he was a capable 
exponent of the best methods of his own generation. But it is 
mly as a tactician that he shines : strategy is nowhere apparent 
n his campaigning, and in 1265 he was hopelessly outgeneralled 
)y the young Prince Edward. We shall see that he relied, like 
ill his predecessors, on the force of cavalry ; the infantry count for 
lothing in his battles. He triumphed, when opposed by the 
ncapable Henry IIL, because he possessed decision, rapidity of 
novement, and a cool head. But it was only in the fight of 
^ewes that his abilities shone out : in the preceding campaign 
le does not show to much more advantage than his incompetent 
)pponent. 

Far otherwise is it with the victor in the campaign of Evesham, 
lere we shall see Edward showing a real mastery of strategy as 
)pposed to mere tactics. When we study his operations in 1265, 
ve shall be quite prepared to find him, thirty years later, presid- 
ng at the inauguration of a new epoch in war at the bloody 
ield of Falkirk. But in his youth he was still, as regards 
actics, employing the old methods which he had learned from 
viontfort as his teacher. 

Battle of Lewes, May 14, 1264. 

Down to the day of battle the operations which led up to the 
ight of Lewes show all the characteristic incoherence and in- 
onsequence of a mediaeval campaign, and do no credit to either 
)f the parties concerned. King Henry had raised a considerable 
a-my in the Midlands, while the baronial party had made itself 
trong in London, but had also seized and garrisoned the im- 
)ortant towns of Northampton, Leicester, and Nottingham. 
The king resolved to subdue the three midland centres of revolt 
)efore undertaking any further operations. Northampton fell 
vith unexpected ease, owing to the treachery of the monks of 
3t Andrew's Priory, who admitted the royal troops through a 
massage into their garden. This was a severe blow to the 
)arons, for some of their chief leaders were made prisoners, 
ncluding Simon the Younger, the second son of the great Earl 
Simon, his kinsman Peter de Montfort, and fifteen barons and 
:)annerets more (April 5, 1264). 



4i6 THE ART OF WAR IN THE MIDDLE AGES [126. 

A few days later (April 11) Leicester was sacked, anc 
Nottingham, the spirit of whose defenders was shaken by th( 
disaster at Northampton, surrendered at the king's summon 
(April 13). Having thus cleared the eastern Midlands 
enemies, Henry should at once have marched on London witl 
his victorious army. The fall of the capital would have settlec 
the fate of the war, and, in spite of all the efforts of D 
Montfort, the spirits of his followers were sinking low. Simoi 
himself had started to relieve Northampton, and had reached S' 
Albans when the news of disaster reached him. He immediatel 
fell back and prepared to defend the city. Finding, howeve: 
that the king showed no signs of striking at London, and ha^ 
marched northward, the earl resolved to make a rapid stroke a 
Rochester, the one Royalist stronghold in the neighbourhood c 
the capital. He stormed the bridge, penetrated into the towi 
and drove the garrison within the walls of the castle (April 18 
He captured its outworks, but the massive strength of Gundulf 
Norman keep was too much for such siege appliances as the eai 
could employ. The garrison, under John de Warrenne, the Ea 
of Surrey, held their own without difficulty. 

Meanwhile, the king had received news of the siege, an 
left the Midlands. He should undoubtedly have risked all oth( 
objects, and thrown himself upon London. The mere news c 
his having turned southward was enough to draw Simon an 
his host back from Rochester to defend the capital (April 26 
The earl merely^ left a few hundred men stockaded in front c 
the gate of the keep to hold the garrison in check — a thing easil 
done, because the narrowness of the exits of a Norman cast' 
rendered sallies very difficult. 

But, instead of striking at London, King Henry merely ser 
forward his son, Prince Edward, with a small cavalry force, to se 
if the city was in a state of defence,^ and then committed th 
extraordinary error of coasting round it by a vast circular mard 
Returning down the Watling Street, he struck off it by St. Alban 
passed the Thames at Kingston, hastily rushed across Surre 
by way of Croydon, and arrived at Rochester on April 2 
The blockading force was easily driven off, and the few prisone: 
made were cruelly mutilated. 

This huge flank march had no merit but its swiftnes 
Prince Edward and the mounted part of the royal arm 

^ See Annals of Dunstable. 



1264] HENRY III. MARCHES ON LEWES 417 

marched from Nottingham to Rochester — a hundred and fifty 
miles — in five days,^ 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.- 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 VVarrenne. 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 
\say through the wood-tracks — for the men of Kent and Sussex 
:ut 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, 
ind did not halt till they had reached Fletching, a village and 

^ Wykes. 1264, § 4. 2 Knighton. ^ Wykes. 1264, § 5. 

27 



4i8 THE ART OF WAR IN THE MIDDLE AGES [126. 

clearing nine miles north of Lewes (May 6th-ioth). Fron 
thence they addressed proposals for peace to the king, datec 
with prudent vagueness "in bosco ju5cta Lewes." They mus 
have known well enough that Henry would refuse them, afte 
his late successes at Northampton and Tunbridge, and on re 
ceiving his angry reply prepared for instant action. Althoug] 
he had the smaller force, Simon was resolved to take th 
initiative, trusting to his own skill, the greater enthusiasm c 
his supporters, and the king's well-tried incapacity in war. 

The town and castle of Lewes lie at a point where the lin 
of the South Downs is cut through by the river Ouse. To th 
east of the place the steep sides of Mount Caburn rise directl 
above the water, hardly leaving room for the suburb of Clifi 
along the river-bank. To the west of the Ouse there lies 
mile and a half of gently-undulating ground before the ascer 
of the Downs begins. In this comparatively level spot lies th 
town of Lewes, flanked to the north by De Warrenne's cast) 
on its lofty mound, to the south by the great Cluniac Priory ( 
St. Pancras, including within its precinct-wall some twenty acre 
of ground. The Ouse in the thirteenth century was still 
tidal river as far north as Lewes, and at high water the sout 
wall of the priory and the southern houses of the town looke 
out on a stretch of mingled pools and mud-banks which forme 
an impassable obstacle. 

North and east, therefore, Lewes is protected by the rive 
and on the south by this tidal marsh, but to the west it ha 
no protection but the castle and the town wall. If an enem 
approached from that side, the king's army would have eith< 
to stand a siege, or to retire behind the Ouse, or to come 01 
and fight at the foot of the hills. 

On this side the main range of the Downs descends rath( 
gently towards the river, not with a uniform slope, but in thn 
spurs separated by slight valleys. The road from Fletching 1 
Lewes passes over the easternmost of these spurs by the hamL 
of Offham, and by this path would have been the shorte 
approach from the barons' camp. But Simon had wisely r 
solved not to come down a road cramped between the hil 
and the river. Marching at early dawn on May 14, 1 
turned off the road north of the Downs, and ascended them 
a hollow slope called the Combe, four miles from Lewes.^ Th 

^ Blaaw and Prothero seem undoubtedly right on this p(»nt ©f tiepography. 



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 

■ pfevious day a small party on a high point of the Dowms 

' to keep watch. No measures had been taken to relieve the 

' .vatchers on the 13th, and, being tired and hungry, they 

, slipped back into Lewes to rest themselves, leaving a single 

1 nan on guard. This individual lay down under a gofse-bush, 

' md was caught sound asleep by the first of De Montfort's men 

; A^ho climbed the slope. Thus the earl was able to put his whole 

I *orce in array On the ridge of the Downs before the Royalists 

lad the least idea that he was within two miles of them. Simon 

md spent the previous day and night in distributing his men into 

[ :orps, and assigning the position of each on the march and m 

j )attle-line — a task which, as the chroniclers tell us, no other 

I nan in his raw army was competent to discharge.^ Now he 

• lad full leisure to see that his exact intentions were carried 

)ut, and to settle the smallest details of the marshalling. 

i Owing to the disasters at Northampton and Nottingham, 

I he barons' army was much smaller than might have been 

[ aised by the full levy of the party, for many of their most 

\ mportant leaders were prisoners in the king's hands.^ The 

Estimate of forty thousand men given by several chroniclers 

t IS Simon's force is one of the hopeless and habitual exaggera- 

i ions of the mediaeval scribe. But, small though the army was, 

, t was divided not into the usual three battles, but into four. 

I There is no doubt that the fourth, which was led by the earl 

> u'mself, was a reserve corps placed behind the others, but none 

^ »f the chroniclers expressly state this fact. It can be inferred, 

i without any danger of doubt, from the circumstance that the 

hree first-named battles of Simon's army each engaged with 

•ne of the three bodies which formed the king's left, right, and 

eritre, and that the earl's division came later into the fight 

han the other three. 

As arrayed on the Downs before descending to battle, the 

* Rishanger, p. 31. 

^ Including Simon de Montfort the Younger, Peter de Montfort and his sons 
eter and William, Adam of Newmarch, one of the greatest of the barons of the 
Velsh border, Baldwin Wake, William de Furnival, all captured at Northampton, 
VilHam Eardolf, captured at Nottingham, and the young Earl of Derby who had 
een taken in his own castle of Tutbnrv. 



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 oi 
the numerous infantry of the Londoners, and of a body o: 
knights commanded by Nicholas de Segrave, Henry dc 
Hastings, John Giffard, and Hervey of Borham. The earl'j 
reserve corps lay behind the centre ; the horsemen in it con 
sisted of his own personal retainers, the foot were probabl) 
Londoners, as they were commanded by Thomas of Pevelsdon 
an alderman of the city, who had always been one of Simon' 
most sturdy adherents. 

Deployed in this order, and probably with the knights o 
each division in front and the infantry behind, Simon's force 
halted just as the bell-tower of Lewes Priory came in sight, t< 
engage for a moment in prayer, after a short address from thei 
leaders. Scattered over the slope of the Downs were sma! 
parties of the grooms of the Royalists, grazing their lord; 
horses, for forage had failed in Lewes. They caught sight c 
the baronial host as it came down the hill, and fled back to th 
town to rouse their masters. Simon's host followed close s 
their heels, leaving on the upper ridge of the hill such sma 
impedimenta as they had brought with them, the chief of whic 
was the earl's chariot,^ to which he had bound his great banne 
after the manner of the Milanese at Legnano or the Yorkshin 
men at our own Battle of the Standard. Inside the carriag 
were three (or four) citizens of London whom Simon ha 
arrested for opposing him, and was determined to keep in sa 
custody. The banner and baggage were left in charge of 
guard of infantry under William le Blound, one of the signatorl 
of the agreement for arbitration which had ended so unhappi 
at Amiens.^ 

^ Simon had broken his leg in the previous year, and was forced to use t 
carriage for many months. 

2 Of the twenty-four laymen who signed for the barons' party in 1263, 1 
following were at Lewes : — Earl Simon, Ralph Basset, William le Blound, Humph: 
de Bohun, John de Burgh, Hugh Despenser, John Fitz-John, Henry de Hastin 
Henry de Montfort, WiUiam de Montchensy, Nicholas de Segrave, Robert de K 



J 



PLATE XIII. 




Earl Simons Army. , King Henrys Army. 

B Horse ^BFoot ^^Horse f~^Foot 

A.Covand Henry D.Earl Simon. F. Kine Henry 
" de Mont fort -. ° 

B. Gloucester. E.TheChanot O. Richard of 
^ „ Cornwall 

C Segrave and the Londoners n. Pnnce Edward 




The EVESHAM 
CAMPAIGN 

1265. 



I 



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.^ Perhaps Henry might 
hcLve 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. 

2 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 rnore 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 tlie Londoners 
on the baronial left had a shorter space to cover before contact 
took place : some of the chroniclers observe that they were s^ 
much in advance that the Royahsts 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 Giffard were 
broken by the furious charge of the prince. Giffard was takeil 
prisoner ; Hastings turned his rein too soon for his own good 
repute ; ^ 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 w^iole 
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 tfee 
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 w^ere 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 formidinem anteponens, e proelic 
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," ^ 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.^ 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 pjobable 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.^ Others slipped 

^ Chron. de Mailros, 1264, § i. 

^ 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 
ei:posed to certain death Le Blound, one of his most trusted followers, and the whole 
affair was (no doubt) a mere chance. 

^ 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 Ic Despenser, 



424 THE ART OF WAR IN THE MIDDLE AGES 



41 



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 two 
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 Wilton 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. 
''X il.t 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 



264] LEWES : A CAVALRY BATTLE 425 

rmy, another mentions crossbowmen, but of archery there is 
word, though the Assize of Arms of 1252 had named the bow 
5 the yeoman's special weapon. In the whole campaign we 
1 nly once hear of the use of that arm — when the king on 
is march to Lewes was molested in the woods by Simon's 
/elsh bowmen, and drove them off with some loss. It is 
Dvious that the supremacy of cavalry was still well-nigh un- 
lecked, and that the proper use of infantry armed with missile 
capons was not yet understood. 

The main interest of the fight is tactical : Simon won because 
3 chose his ground well, because he surprised his enemy and 
Tced him to fight in disorder before he could get his host com- 
etely arrayed, and still more, because he kept his victorious 
oops in hand, and employed his reserve at the proper moment 
id in the proper place. Henry lost, partly because he was sur- 
•ised,and forced to fight in an unfavourable position, but far more 
icause the victorious part of his army threw away its advantage, 
id was absent from the field during the critical hour that 
ttled its fortune. Rash adventure and hot-headed eagerness 
pursuit cost the Royalists the day. But neither discipline 
•)r self-restraint were likely to be prominent in any army over 
lich the imbecile Henry Plantagenet bore rule. 

Battle of Evesham^ August 4, 1265. 

We have already had occasion to remark that while at Lewes 
e tactics are all-important, in the campaign of Evesham we 
•ve to deal primarily with strategy : the actual battle was 
mparatively insignificant. 

In May 1265 all England seemed at De Montfort's disposal: 
ere were only a few small storm-clouds on the edge of the 
rizon. Certain barons of the Welsh March, headed by 
)ger Mortimer, were in arms beyond the Severn ; a small 
rty of Royalists had been holding for many months the 
>lated castle of Pevensey. The Earl of Gloucester was 
own to be discontented, but it was not supposed that he 
mid lightly betray the cause for which he had fought so well 

Lewes. 

To hunt down the insurgents in the March, Earl Simon left 
ducester in the middle of May, accompanied by several of 
5 firmest adherents — his eldest son Henry, Despenser the 
Sliciar, John Fitz-John, Ralph Basset, and Humphrey de 



I 



4:z6 THE ART OF WAR IN THE MIDDLE AGES [1265 

Bohun. He took with him King Henry and Prince Edward 
who, though nominally free, were never allowed to stir far froir 
his side, except under safe custody. At Hereford on May 2<: 
the prince escaped from his guardians by a swift horse anc 
an easy stratagem. He fled to Mortimer at Wigmore Castle 
and soon met Gloucester at Ludlow. There De Clare die 
homage to him, and concluded a formal alliance with him 
They at once raised their banners, and summoned all faithfu 
subjects of the king to join them. Shropshire and Cheshir 
rose at once at their call. 

Simon, still lying at Hereford, had now the choice w4iethe 
he would strike at once at the earl and the prince, whether 1) 
would continue his campaign against the Marchers, or whethe 
he would promptly fall back into England by Worcester c 
Gloucester, and take up a central position. He chose th 
second alternative, underrating, it would seem, the importanc 
of the earl's rebellion. But as a matter of precaution he ser 
a detachment of three hundred men-at-arms under Robert fi 
Ros to hold Gloucester, and so to provide him with a safe bridg 
over the Severn and good communication with London. H 
also bade the sheriffs of the western counties raise their levk 
against the insurgents, and made the king set his seal t 
documents outlawing both the prince and De Clare. 

Montfort stayed at Hereford till June 10, thus giving tin 
for his enemies to draw together in dangerous strength. The 
seized Bridgenorth and Worcester, broke their bridges, destro}'* 
or removed all the boats on the Severn, and spoiled the neig] 
bouring fords by dredging them deeper. Then, on June 13-^1 
they passed down the river-bank to Gloucester and laid siej 
to it. The town fell, but De Ros held out gallantly in tJ 
castle for fifteen days, in spite of the fact that he had b« 
caught almost destitute of provisions. 

Simon had not given his enemies credit for any suchstratc^ 
as they had displayed. He had moved out from Hereford < 
June 10, to confer with Llewellyn Prince of Wales, and 
enlist his services against the enemy — a task which he kvi^ 
would not be hard, on account of the Welshman's an< 
quarrels with the Royalist barons of the Marches. They m( 
conference at Pyperton, where Llewellyn, in -return for 
restoration of many lands and castles which had been tak 
from his predecessors, promised his aid. He undertook to s^ 



1265] EVESHAM: PRELIMINARY MOVEMENTS 427 

five thousand spearmen to join the earl, and to start himself at 
the same time on a raid into the Mortimer and De Clare estates. 
The treaty was concluded on June 19 ; on the 22nd the king 
solemnly signed it at Hereford, to which place Simon had led 
him back. They then marched southward to Monmouth, 
probably intending by this move to place themselves between 
the prince's army and the great De Clare estates in Gwent and 
Glamorgan ; at the same time, they were in a good position for 
moving to relieve Gloucester, the all-important avenue for 
communication with the Midlands and London. But the fates 
were against Earl Simon : he stormed the great castle of 
Monmouth, — one of De Clare's chief strongholds, — but when he 
prepared to move eastward, a large division of the Royalist army, 
detached to cover the siege of Gloucester against any attempt 
at a relief, showed itself on the other side of the Wye. It was 
headed by John Giffard, a baron who had fought for Simon at 
Lewes, but had now deserted his cause on account of a private 
quarrel. Giffard fortified himself in a good position com- 
manding Monmouth bridge, and defied the earl to come over 
and attack him. Simon saw that Giffard was unassailable, and 
that he must find some other way of continuing his movement 
eastward. The best course seemed to be an attempt to cross the 
Bristol Channel; accordingly, he sent a message to the citizens 
Df Bristol, who were his good friends, though their castle had 
3een for some time held by a Royalist garrison, bidding them to 
jend ships over to Newport, at the mouth of the Usk, and thence 
erry him and his host over the Channel. Another message was 
»ent at the same time (June 28) to the earl's son Simon, who 
vas occupied far to the east, in the siege of the castle of Pevensey, 
■Q warn him that the rebellion was spreading so rapidly that he 
nust at once raise the leaguer, collect his friends, and march 
igainst Prince Edward. 

Meanwhile, De Montfort left Monmouth and marched on 
Jsk, a strong De Clare castle, which he successfully stormed 
uid took, as it had been left with an inadequate garrison. He 
lext seized Newport and ^Abergavenny, and (being now joined 
)y Llewellyn's promised succours) spread his troops abroad, and 
iercely harried the Earl of Gloucester's lands in the neighbour- 
lood. Probably he designed by this move to draw De Clare into 
South Wales, and so to secure an undisturbed march for his own 
einforcements from the east. His intention was to abscond by 



I 



428 THE ART OF WAR IN THE MIDDLE AGES [1265 

sea, by means of the Bristol ships, when the prince and De Clare 
should come upon him. 

Gloucester Castle had fallen on June 29, and the Royalists, 
having now all the bridges over the Severn in their hands, 
marched to join the corps of observation under Gift'ard which had 
already been watching Simon. Prince Edward and De Clare 
retook Usk only three days after it had yielded to the enemy, 
and then marched to seek Simon at Newport. Before leaving 
Gloucester, they had heard of the fleet of transports which was 
being fitted out at Bristol, and sent against it three galleys which 
they had found at Gloucester, filled with a chosen band of men- 
at-arms. These vessels came upon the Bristol ships just as they 
had reached Newport harbour, and were being laden with De 
Montfort's baggage. They dashed into the river-mouth, and 
took or sunk eleven of them — practically the whole flotilla. At 
the same time, the Royalist army fell upon Simon's troops near 
Newport, and routed them by dint of very superior numbers. 
Their advance was only stopped when the bridge and town were 
fired in their faces by the retreating enemy, who took refuge 
behind the Usk (about July 8). 

Simon was thus deprived of his chance of crossing the ; 
Bristol Channel, and thrown back into Wales ; his prospect oi 
reaching England and rejoining his partisans seemed more 
remote than ever. The only course that remained open to him 
was to strike northward again, keeping the Usk between him anc 
the enemy, and regain Hereford by a toilsome march. In the wile 
and thinly-peopled country between Abergavenny, Crickhowell 
Brecon, and Hay, his army suft'*ered dreadful privations, the 
English troops complaining that they could not live on a Welsl 
diet of mutton and milk, and were lost without their daily ratior 
of bread. Simon reached Hereford somewhere about July 20, 
with a half-starved and dispirited army, and was obliged to pause 
for some days to allow his men to recover their strength. The onlj 
cheering feature in the situation was that news reached him fron 
the east that his son and his friends were marching at last to his aid 

But meanwhile Prince Edward and Gloucester, after pursuing 
De Montfort in vain up the Usk, and capturing Brecon,^ ha( 

^ The exact chronology of De Montfort's movements in July is (most unfortv 
nately) not to be made out. But the dates given cannot be far wrong. 

^ Battle Chronicle. Prince Edwnrd captured Brecon, while Gloucester retoo 
Monmouth. 



1265] EVESHAM: THE BARONS AT KENILWORTH 429 

hastened back to Worcester, and prepared once more to hold the 
passages of the Severn. The last ten days of July were spent 
by Earl Simon in two unavailing attempts to force his way over 
the river. He was foiled, and got little profit by his single 
success — the capture of the Royalist garrison at Leominster. But 
the old chief w^as not yet disheartened, in spite of the unexpected 
skill and strategy which his enemies had displayed. He knew 
that his son and the army of succour were now closing in on the 
prince's rear, and encouraged his men by promising that they 
would catch the enemy in a trap between their two divisions. 
Having at last procured some large boats, he secretly brought 
them down to the water's edge, and determined to make a third 
attempt to cross, at a spot opposite Kempsey,^ which he thought 
might be the more carelessly guarded, because it was so very 
close to the prince's main camp at Worcester. 

Meanwhile, Simon de Montfort the Younger had wasted 
much time by marching to his father's aid by a most extra- 
ordinary and circuitous route. He moved from Pevensey to 
London, from London to Winchester (July 14), from Winchester 
to Oxford, and from Oxford to Kenilworth, where he arrived on 
July 31. Speed should have been his main object, but he had 
preferred instead to gather as large an army as possible by 
calling in all his father's partisans. Hence he came on the field 
far too late, but with an imposing force, quite capable of facing 
the Royalists. With him were most of the leaders of the baronial 
party — the young Earl of Oxford, William of Montchensy, 
Richard de Grey, Baldwin Wake, Adam of Newmarch, Walter 
Colville, Hugh Neville, and some fifteen other bannerets. They 
reached Earl Simon's castle of Kenilworth on July 31, and 
encamped below its walls, for the castle enclosure was not nearly 
spacious enough to hold such a large force. 

All the combatants were now gathered in a space of thirty 
miles, and the campaign came to a sudden end with a short sharp 
shock. Prince Edward and Gloucester took the offensive; it 
was all-important to them that the two Simons should not meet, 
but should be dealt with separately. The old earl was still 
behind the Severn : his exact whereabouts was not known, but 
it was obvious that he could not cross the river and join his son 
in less than two days : he had also the less formidable of the 
two forces with which the Royalists had to contend. The prince, 

^ It is only four miles south of Worcester, the enemy's base of operations. 



430 THE ART OF WAR IN THE MIDDLE AGES [1265 

therefore, resolved to leave the earl Unwatched for a moment, 
while he dealt a rapid, vigorous stroke at Kenilworth. He learned 
from a traitor in the baronial camp that Simon the Younger 
was keeping a careless watch, trusting to the thirty miles which 
separated him from Wofcester. Accordingly he determined to 
copy the tactics of Earl Simon at Lewes, and to make his stroke 
in the early morning, so as to get a chance of surprising the 
enemy in his camp. 

Starting on the evening of August i, the prince made a 
forced march throughout the night, and reached Kenilworth in 
the early dawn. As he neared the place he heard the sound 
of a moving multitude, and imagined that young Simon had 
got wind of his approach, and was ready to meet him. Btit, 
pushing on, he found nothing but a train of waggons, bearing 
food and forage to the enemy. They were seized in an instant, 
and not a single man got away to warn the careless barons. 
A few minutes later the Royalists rushed into the streets of the 
sleeping town, cutting down the half-roused enemy as they 
poured out of tents and houses, and sweeping right up to the 
walls of the castle without a check. Well-nigh the whole of 
the barons fell into their hands, without giving or receiving a 
stroke. The young De ]\Iontfort escaped into the castle half- 
naked, but Oxford, IMontchensy, and all the rest were captured 
in their beds. The baronial army was practically annihilated ; 
only those who had slept in the castle escaped. Edward 
tarried no longer than he could help in the place ; the moment 
that the prisoners and the booty w^ere secured, he hurried back 
to Worcester, to look after Earl Simon. 

While the wearied Royalists were pouring back towards 
Worcester, a busy scene was in progress at Kempsey. The 
earl had launched his boats, and was throwing load after load 
of his men across the river, rejoicing greatly that no interrtiptior 
came from the direction of Worcester. By the evening all w^re 
across, and Simon, on learning that his son was at Kenilworth 
prepared to start on his way thither next morning. He darec 
not march past Worcester, and therefore chose the southefr 
road by Pershore and Evesham. On x'\ugU3t 3 he started, aric 
covered the fifteen miles from Kempsey to Evesham. Mean 
while, the prince had returned to Worcester and learned that hi5 
enemy had crossed the long-guarded river in his abseiVce. Bui 
Simon was not too far advanced to make it impossible to heac 



1265] EVESHAM: MOxNTFORT PASSES THE SEVERN 431 

him off and intercept his path eastward. Though his men 
must have been even more fatigued than the earl's travel-worn 
host, the prince struck out from Worcester once more, and 
marched eastward on the evening of August 3.^ 

There are two roads from Evesham to Kenilworth — one by 
Alcester, the other by Stratford-on-Avon. It was Edward's 
object to throw himself across both these paths. His exact route 
is not specified by any chronicler, but we know that, having 
marched all night and an hour or 'two after dawn, he lay across 
the Evesham-Stratford road with his own " battle." He had 
divided his army into three corps, giving the second to De Clare, 
and the third to Mortimer and the Marcher barons. It appears 
that each body marched by a different road, with orders to 
:onverge on Evesham. The prince approached from the 
aorth, Gloucester from the north-west, on Edward's right, 
Mortimer from the west, and in the rear of the town. The 
routes of the three corps were probably therefore, (i) VVorcester- 
Flyford-Dunington-Norton ; (2) Worcester- Wyre-Craycombe ; 
3) Worcester-Pershore-Hampton.^ 

:: The town of Evesham, where Montfort's little army was 
•esting on the morning of August 4, lies at the southern end 
Off a deep loop of the Avon. The roads from Alcester, 
Worcester, and Stratford join at the base of the loop, and, after 
jniting, descend into the place by the gentle slope called Green 
^ill. At the southern end of the town lies the abbey, where 
, Simon and the king were lodged, overlooking the bridge and 
:he suburb of Bengeworth. Beyond the bridge the other roads 

' ^ He is said to have suspected that there were traitors in his ranks, and therefore 

L o have marched to Claines, three miles north of Worcester, as if about to move on 
■ ^ridgenorth, and then to have suddenly swerved east, and hurried off to get between 
I >iiWon and Kenilworth. 

^ I cannot agree with Professor Prothero's view (in his Simon de Montfort) that 

ulward marched with his whole army by Alcester, crossed the Avon at Cleeve Prior, 

nd recrossed it at Offenham, sending Mortimer by the south bank of the river to 

3engeworth. The double crossing seems unnecessary, and has no justification but 

Changer's statement that Edward crossed a river unnamed, " juxta Clive," no second 

1 rossing being spoken of. That a whole army, twenty thousand strong, should pass at 

; )ffenham in full daylight, without being seen by anyone from Evesham — less than 

: wo miles away — is a sheer impossibility. We know that Edward came in sight of 

' 5imon on the Norton road, and was descried at some distance. We also know that 

Viortimer approached from the west {i.e. from Pershore), by Hemingford's statement 

hat the earl's look-out saw "vexilla Rogeri de Mortimer ab occidente et a tergo." 

therefore agree with Mr. New of Evesham, whose view Professor Prothero 

efuses to accept. 



432 THE ART OF WAR IN THE MIDDLE AGES [126 

diverge in the directions of Pershore, Tewkesbury, and Broadway 
Evesham is a good position to defend against an attack fron 
the south, being well covered by the river, and approachabl 
only by a single bridge. But if attacked from the north it i 
far less defensible, as the advancing enemy has the advantag 
of the slope, and the defenders must fight with a single narrow 
bridge at their backs. But if assailed at once on north an 
south by superior forces it is a fatal trap, for no escape i 
possible, owing to the loop of the river, which encircles it o 
three sides. 

Simon's men took their morning meal and heard mass ; bu 
just as they were mounting to commence their march, new 
came in that a large force was approaching by the Duningtor 
Norton road.^ The earl hoped that this was the army of h 
son Simon, marching in from Kenilworth, for he was still whol) 
ignorant of the disaster that had befallen his friends on tV 
2nd. He was at first encouraged in this delusion, for Princ 
Edward had ordered that the banners taken at Kenilwort 
the White Lion of Montfort, the silver star of De Vere, ar 
the three escutcheons of Montchensy, should be borne in h 
van to disarm suspicion. But to gain certainty Earl Simc 
rode to the crest of Green Hill,^ according to one account, < 
sent a keen-sighted attendant up the abbey tower,^ accordir 
to another. Very shortly the royal banner was seen wavir 
over the main body, and the earl recognised his mistake, ar 
saw that he must either fight or fly. Shortly afterwards tl 
red chevrons of De Clare were descried pressing on at tl 
head of a new column, which was only just coming into sig] 
to the prince's right. Only a few minutes later the blue ar 
white banner of Mortimer was descried on the Pershore roa 
coming from the west, and in the rear of the baronial hoi 
" Now may God have mercy on our souls, for our bodies are 
the power of our enemies," cried the earl when the full horr 
of the situation dawned upon him. There w^as still a chan- 
for well-mounted horsemen to escape over Evesham bridge ar 
dash eastward ; but the army was evidently doomed, unless 

^ The Chronicle of Mailros says that Prince Edward was sighted whflJP I 
much as two leagues away by the earl's scouts. If this is correct, the whole stc 
of his having crossed the Avon at Cleeve and Offenham fails. 

2 Hemingford calls it Mount Elyn. 

' Chronicle of Mailros and Hemingford. 



1265] EVESHAM : EARL SIMON SLAIN 433 

could cut its way through Edward's host. Henry de Montfort 
hastily bade his father fly, and swore that he would hold the 
enemy at bay long enough to get him a good start. But the 
old earl laughed the proposal to scorn. He had brought them 
there, he said, and must take the consequent responsibility. 
He had never fled from battle before, and would not begin in 
bis old age. He besought Despenser, Basset, and the other 
barons about him to save themselves, but no one would flinch 
"rom him, and all made ready for battle. There was still some 
:wenty or thirty minutes to spare before Mortimer would be able 
:o close in on their rear. Simon employed the time in forming 
lis host in a deep column, the knights at its head, the foot 
Dehind, and steadily marched up the Green Hill, making directly 
or the centre of Prince Edward's division. The front came on 
;teadily enough, but the Welsh infantry in the rear began to 
■nelt away before a blow had been struck, slipping off into the 
ields and gardens on each side of the road, and then plunging 
nto the Avon and swimming over as best they might, so as to 
ilude Mortimer's approaching corps. 

The earl himself, meanwhile, dashed into the middle of the 
Drince's corps with such a desperate shock that the Royalists 
vavere