The
besieged
still had almost complete liberty of exit, especi-
ally by the river gate on the north side.
ally by the river gate on the north side.
Cambridge Medieval History - v5 - Contest of Empire and the Papacy
Tancred
remained in command, and finally crossed into Asia Minor without
entering Constantinople. Bohemond was an hereditary enemy of the
Greek Empire, and now as at all times ready to take up arms against
Alexius if he saw any advantage in doing so.
He intended to secure a
princedom in the East, and most probably had already fixed his choice
on Antioch. Before taking the oath of allegiance he endeavoured to
obtain a promise from the Emperor to support his scheme. Alexius?
answer no doubt was that such requests were premature, and that every-
thing would depend on the issue of the Crusade. It is unlikely, in spite
of the definite statement of the Gesta Francorum, that Bohemond was
now promised territory in the neighbourhood of Antioch. At most the
Emperor may have indicated that he would afterwards consider favour-
ably such claims as the Norman chief might be able to present.
Robert of Flanders accomplished the first part of his journey through
France and Italy in the company of Robert of Normandy. He crossed
from Apulia in December 1096, and did not advance farther towards
1 The attitude of the modern historian to Anna Comnena's narrative of these
events is decisively important for the view to be taken of them. Her account is inter-
preted by some to mean that Alexius finally compelled Godfrey, by force of arms, to
take the oath of allegiance. This is no doubt the impression she conveys, but in
view of Albert's narrative the Greek account may be regarded as patriotically over-
drawn at this point. The view that the essential matter in dispute between Godfrey
and Alexius was the oath of allegiance is a possible one, but Godfrey's yielding (with-
out defeat) on such a point is more difficult to understand than Alexius' yielding on
the point of the hostages. The very improbable date that Anna gives quite incident-
ally at one point (“Thursday in boly week”= 2 April 1097) has not been satisfactorily
accounted for by those who reject it (cf. Kugler). May there not be a confusion with
Christmas week, in which, according to Albert, there was a cessation of hostilities
out of respect to the season? Christmas Day in 1096 was a Thursday.
## p. 283 (#329) ############################################
Raymond of Toulouse
283
Constantinople until the spring. He arrived later than Bohemond, and
readily took the oath of allegiance.
Raymond of Toulouse, having left home, perhaps, about the end of
October 1096, came by the north of Italy and the eastern shore of the
Adriatic Sea. Passing through Dalmatia in the winter, his army suffered
from the inclemency of the season, from scarcity of food, and from the
attacks of the inhabitants of the country, so that large numbers of the
crusaders lost their lives. At Durazzo messengers from the Emperor
brought assurances of friendship and promised supplies. Beyond this
point, however, there was frequent fighting between the crusaders and the
Greek mercenaries who watched their progress. The Provençals con-
sidered themselves the aggrieved parties, and retaliated by destroying the
suburbs of Rusa and plundering the town. At Rodosto, four days'
journey from Constantinople, Raymond received a request from the Latin
leaders already in Constantinople to hurry on, because they were pre-
paring to start and were making arrangements with the Emperor to
which it was desirable that he should be a party. When he reached
Constantinople (perhaps in the third week of April), he decisively refused
to take the now customary oath of allegiance. If the Emperor put him-
self at the head of the expedition and came with them, he would become
his follower, he said, not otherwise. News of a shameful defeat of his
army, in a conflict in which they were afterwards judged to have been in
the wrong, only increased his determination not to yield. Finally, under
pressure, he only consented to take an oath that he would do nothing
against the life and honour of the Emperor. In consequence of his
attitude he received, as the Provençal historian notes, little of the
Emperor's bounty.
Last of all came Robert of Normandy, with his powerful brother-in-
law, Stephen of Blois, and with Godfrey's brother Eustace, Count of
Boulogne. Their army included the first expedition of “Englishmen
and Britons” to join in the Crusade. Robert left home in September
and had spent the winter in the south of Italy. He embarked at
Brindisi on 5 April 1097, and reached Constantinople about the middle
of May. After spending a fortnight in the Greek capital he proceeded to
the siege of Nicaea, which had already begun.
The Emperor Alexius had good reason to be satisfied with the initial
result of his negotiations with the Latins. Formally, at least, he had
secured from the leaders of the Crusade the acknowledgment he desired.
Even Raymond of Toulouse seems finally to have admitted the Emperor's
claims in Asia Minor and Syria? An agreement so important and so
i The date rests on a doubtful calculation. Vic and Vaissete, Histoire générale
de Languedoc (1733), Vol. 11, p. 628.
2 It is not easy to say when Raymond became the friend and partisan of the
Emperor, which he shewed himself to be in Antioch. Albert, xi, 20, implies that it
was previous to the march of the crusaders through Asia Minor. The Emperor may
CD, VII.
## p. 284 (#330) ############################################
284
Treaty with Alexius
intricate must have been put in writing and signed by the contracting
parties? If it did not specify all the lands which the Emperor claimed,
it probably named at least the territories and towns in which he desired
to place Greek governors, and some also of those which might be held by
the Latins in fief. The plunder of all the captured cities may have been
assigned to the Latins, and the Emperor certainly promised military
assistance to his allies. The obligations of the Latin feudatories must
have been defined, and, it may be, also the conditions on which they
would obtain recognition as lords of the conquered territory. Of course
the adherence of the crusaders to this agreement depended entirely on
the Emperor's fulfilment of his promise to render them further assistance.
If he failed in this obligation, the Latins were inevitably released from
their pledges to him. But meantime the leaders were won partly by
the personal charm and lavish gifts of the Emperor, partly, it may be
added, by the reasonable character of his proposals, so that they judged
their treaty with him to be of value to their enterprise. It is true that
there was at the same time, especially among the rank and file, a strong
undercurrent of suspicion and hatred of the Greeks. Godfrey's troops
and Raymond's had already been engaged in serious fighting with them.
have gained Raymond's concession by a definite pledge, in the formal treaty which
regulated the relationship of the parties, to bring an army to the assistance of the
Latins. Even Raymond's oath in Constantinople may have guaranteed the Emperor's
territorial claims, although it was not an oath of allegiance. To the view that Ray-
mond suddenly changed his attitude in Antioch out of hostility to Bohemond, it may
be objected that there is no suggestion in the sources of any inconsistency in his
attitude, nor of any difference between the obligations of the individual Latin leaders
towards the Emperor. The hostility of the Provençal historian, Raymond of Agiles,
expressed in his description of the occurrences at Nicaea and elsewhere, is not a certain
indication of the attitude of Raymond of Toulouse at the time. This historian, and the
Provençals generally, never shared Raymond's partisanship for the Emperor. Ray-
mond's remaining in the camp at Nicaea in June 1097, when the other chiefs went to
visit the Emperor after the capture of the city (Hagenmeyer, Epist. et Chart. p. 140),
must be regarded as merely incidental, on the view here taken of his final reconcili-
ation with the Emperor.
1 This is also F. Chalandon's view. There is a small amount of documentary
evidence in its favour, viz. Alexias, xii, 12, where the reference to a treaty with
Bohemond may be understood of a general treaty made with the Latin chiefs, and
Hagenmeyer, Epist. et Chart. No. 12, from which it may be inferred that the date of
the agreement was the middle of May. The terms of the treaty, as these are sug-
gested above, are inferred from narratives of what took place after the captures of
Nicaea and Antioch and from the terms of the later treaty with Bohemond in 1108
(Alexias, xii, 12). The alleged grant to Bohemond of territory in the neighbour-
hood of Antioch at this time (Gesta Francorum) may be interpreted to mean that this
territory, although not assigned to any individual, was actually designated in the
treaty as a prospective Latin fief. The exclusion of the city of Antioch suggests that
it was to be placed in the hands of a Greek governor. Albert, v, 2, states that the
Latin leaders had sworn specifically to hand over Antioch and Nicaea (if captured
by them) to the Emperor (cum omnibus castellis et urbibus ad regnum eius perti-
nentibus). Raymond, ch. iv, shews that certain rights were conceded to the Latins
in these Greek towns (in Nicaea a monastery and a hospital).
## p. 285 (#331) ############################################
Siege of Nicaea
285
The Normans were really bitter and contemptuous enemies of the Greeks,
although Bohemond judged it to be expedient to acquiesce in a general
treaty, and required Tancred, much against his will, to take the common
oath of allegiance. At the same time the marked hostility of the
Western sources to the Emperor in their narratives of these events
reflects largely the anger and disappointment of a later period. The
Greeks and Latins had important interests in common, and it is likely
that the policy inaugurated by the Emperor would have held them
together until at least the foundations were laid in Syria of one Greco-
Latin state. It was Alexius' own failure to implement his promise
that finally turned the Latins into declared and irreconcilable enemies.
Before the Latins left Constantinople, their route through Asia Minor
and their plan of operations had been decided on. In the first place the
Muslim capital of Nicaea, about six days' march overland from Scutari,
was to be taken. The Emperor provided siege engines and food
supplies but only a small detachment of troops.
Nicaea was very
strongly fortified and was protected on the west side by the waters of
a lake. The disposition of the crusading army illustrates the separation
caused by national divisions. Bohemond's forces encamped on the north,
Godfrey and the Germans on the east of the city (6 May 1097). When
Raymond's troops arrived they occupied the south side (16 May). On
the day of Raymond's arrival a small force of Muslims attempted to
throw themselves into the city and were beaten off. Robert of Normandy
and his men joined the besiegers on 1 June; their position also was on
the south side. The siege operations, begun on 14 May, were pressed
strenuously with little result for nearly five weeks. At length the ruin
by Raymond's engineers of a large tower on the south side brightened
the prospects of the besiegers. This and the launching on the lake of
Greek vessels, brought from the sea, decided the defenders to surrender.
They opened negotiations with the Greek commander, and capitulated
to him on condition that their lives should be spared (19 June 1097).
Most likely they were allowed to remain undisturbed in their homes if
they chose to transfer their allegiance to the Emperor. In order to
prevent wanton plundering and destruction, the Latins were allowed
to enter the city as visitors only and in small parties. As previously
arranged, the spoil of the town, or its equivalent, was distributed among
the crusaders, and their leaders received in addition handsome gifts from
Alexius. No doubt the sparing of the lives of infidels became a cause
of reproach to the Emperor in the Latin camp, and perhaps the pre-
cautions taken to protect the city from plundering were resented. But
the Latins do not seem, on this occasion, to have been unfairly treated',
and some of them settled in Nicaea as the Emperor's subjects.
1 Raymond's account brings no specific charge against Alexius, although it shews
that he was disliked and hated by the Provençals.
CH. VII.
## p. 286 (#332) ############################################
286
March through Asia Minor
After the capture of Nicaea the proximate goal of the crusaders' march
was Antioch on the Orontes. It may be assumed that Alexius urged the
siege and capture of a city which had been for a century an outpost of the
Empire, and the occupation of which would be an important initial step
in the conquest of Syria. Besides, the deliverance” of Antioch had been
from the first one of the specific objects of the Crusade. The way through
Asia Minor was familiar to the Greeks and in any case easily found and
followed. It leads through Dorylaeum and Iconium and then over the
passes of the Taurus into Cilicia. But in order to rescue Armenia Minor
from the Muslim yoke and to secure for themselves friendly support in a
district near Antioch, the main body of the crusaders kept eastwards to
the anti-Taurus mountains, and then came southward to Antioch by way
of Geuksun (Coxon) and Marʻash. Cilicia, in which there was also a
friendly Armenian population, was secured by Tancred and by Godfrey's
brother Baldwin. The Latins sent letters to the Armenians of Euphra-
tesia, most probably from Nicaea, and Baldwin was joined there by an
Armenian exile who accompanied and advised him during the march
through Asia Minor. This alliance with the Armenians was afterwards
of great value during the siege of Antioch, and by it the crusaders were
enabled to make their first settlements in the East.
When the Latins left Nicaea—those who moved first started on
27 June—some cherished the hope that they might reach Jerusalem in
five weeks, if Antioch did not prove a serious obstacle in their way.
It was three months before they approached Antioch and nearly two
years before any of them reached Jerusalem. Qilij-Arslān, having
assembled his army too late to save Nicaea, attacked the Latins near
Dorylaeum on 1 July. The crusaders were in two divisions, two miles apart,
on separate roads. The first encounter was between the whole Turkish
army, which consisted exclusively of horsemen armed with bows, and the
smaller part of the Latin host, which included the Normans only, under
Bohemond, Tancred, and Robert of Normandy. An attack of the Norman
knights was repulsed by the Turks, whose advance, in turn, was checked
by the spears and bows of the Latin infantry, upon whom the knights
fell back'. The encircling Muslims now employed their usual elusive and
harassing tactics and the Normans fought a desperate battle, until they
were relieved by the arrival, in successive bands, of Godfrey and the other
leaders. The Turks having retreated on to a hill-side, the crusaders
formed themselves into line of battle and broke and scattered their
opponents by one irresistible charge. In the shock of direct encounter
the light Turkish horsemen had no chance of success. The fight before
1 The fullest account of the first stage of the battle is given by the Gesta Tan-
credi. This source alone explicitly mentions the decisive action of the infantry (deu-
sissima pedestrium hastarum silva nunc fugam impedit, nunc extiuguit). The part
played by the Latin infantry in these battles with the Muslims is frequently ignored
by the sources (cf. infra, p. 290, note 1).
## p. 287 (#333) ############################################
Alliance with the Armenians
287
Godfrey's arrival may have lasted two or three hours and the second
stage of the battle, including the pursuit, three hours more. The enemy
were pursued for several miles, and great booty was obtained from the
captured Muslim camp'.
During the march beyond Dorylaeum the Latins found the country
laid waste for a considerable distance, and suffered greatly from want of
food and water as well as from the excessive heat. They lost a large
number of their horses and baggage animals. Most probably the crusaders
now marched in one main force, where all the baggage was placed, and in
several smaller forces under independent leaders such as Tancred and
Baldwin. From Iconium eastwards the conditions seem to have improved,
and of course in Armenia Minor the friendship of the Christian popula-
tion made the way easy. The Muslims were nowhere in sufficient force to
venture another attack after their defeat at Dorylaeum. In Armenia
Minor the Turkish garrisons, which had not long been in possession, were
expelled and Armenian supremacy was restored. Several Western knights
settled in the conquered strongholds, but the only leader of importance
to remain in the district was Baldwin, Godfrey's brother, afterwards his
successor in Jerusalem. Baldwin was the founder of the first of the
crusading states in the East. After passing, as we shall see, through
Cilicia he reached the main army at Marʻash. But while it went on to
Antioch he remained to establish a Latin princedom in Euphratesia. His
first capital was Tell-bāshir (October 1097). Afterwards, when he became
ruler of Edessa (spring of 1098), he made that city his capital. His
forces in themselves were not at first large, but the friendship of the
Armenian princes secured his position. After the fall of Antioch, Godfrey
came to his assistance, and from that time he was quite able to maintain
himself. Undoubtedly, if the Latins had continued to co-operate with the
Armenians, this northern state would have been a much more effective
bulwark of their power than it ultimately proved to be.
No doubt Baldwin's settlement in Edessa was made with the consent
and approval of the Latin leaders. It was in some measure due to him,
since he had recently resigned to Tancred his claim on Cilicia. As the
crusaders arrived in the districts where the first permanent conquests
were attempted, it became perfectly clear that each leader fought not merely
for the common cause but also for a share in the territory that was being
1 Oman's representation that Godfrey and the others were six or seven miles
distant from the Normans and that Bohemond fought alone for five hours (Art of
War, 1st ed. p. 275 (2nd ed. 1, p. 277]), seems to lack foundation. The distance between
the divisions was two miles according to Raymond, who is confirmed by Albert's state-
ment (11, 35) that the intention was that the divisions should march one mile apart.
The Gesta Fruncorum makes the battle last from the third hour to the ninth, and
Fulcher says that the enemy were routed at the sixth hour. This seems to indicate
that the six miles pursuit (Albert, 11, 42) lasted about three hours. Fulcher is vague
regarding the time of the commencement of the battle, for which he names both
the second hour and the first.
CH. VII.
## p. 288 (#334) ############################################
288
Inter-national rivalries
conquered. In the smaller undertakings each national army made its own
conquests and of course claimed to retain what it thus won. The events
in Cilicia are narrated at full length by the sources, and may be taken as
the best available illustration of what has just been said. The occupation
of this province was probably part of a general scheme suggested by
Armenians who accompanied the crusaders from Nicaea, and it may have
been included in the plans of the leaders from the time they left that
city. But since Baldwin and Tancred were rivals in their operations in
that province from the first, it is not hazardous to conclude that one at
least was deputed to protect national interests against the action of the
other? . Tancred left the main army at Heraclea and made directly for
Tarsus, which he hoped to gain with the help of Armenian friends. He
had encamped beside the town and was negotiating its surrender, with
every prospect of success, when Baldwin came on the scene with much
larger forces. It is uncertain whether Baldwin had left the main army
at Heraclea, or had separated from it much earlier than Tancred and had
reached Tarsus by a different road. The result of his arrival was that the
Turkish garrison deserted the town and the inhabitants prepared to sur-
render formally to Tancred. Baldwin, in virtue of his superior strength,
required them, however, to surrender to him, and Tancred retired in anger
without fighting. At Adana he found the Turks already driven out and
an Armenian governor installed", from whom he received a welcome.
Mamistra, the next town on the way, was occupied without difficulty, for
the garrison Hled almost as soon as Tancred approached the city. Mean-
time Baldwin was joined at Tarsus by a fleet of Flemings and Frisians,
which had been cruising for some years in the Mediterranean and was
commanded by a certain Winemar of Boulogne. Having left a garrison
in Tarsus, Baldwin marched on to Mamistra, where he encamped outside
the walls. Either party may have been the aggressor in the fighting which
followed. But Baldwin had now designs further east, so that peace was
quickly re-established and Tancred was left in possession. Beforethe Norman
leader left Cilicia, he had established a claim to possession which Bohemond
and he, as princes of Antioch, afterwards strenuously maintained against
the armies of the Empire. Meantime most of the population favoured
the Latins, and the small Turkish garrisons were cowed by the numbers
1 If both left the army at Heraclea they may have been declared rivals from the
first, or Tancred may possibly have tried to outstrip Baldwin and thus have provoked
the final conflict. On the other hand, if Baldwin left the main army considerably
earlier on the march from Nicaea, then Tancred must have entered Cilicia with the
purpose of securing the province for the Normans with the knowledge that he would
probably have to compete with Baldwin for its possession. Both Baldwin and Tancred
were advised by Armenian friends.
2 The Gesta Tancredi estimates Tancred's force at 100 knights (loricas) and 200
bowmen (ch. xxxIII), Baldwiu's at 500 horsemen and 2000 foot soldiers (ch. xxxvII).
Ralph of Caen. Albert makes the governor Welf; a Burgundian, who had also
separated himself from the main army.
3
## p. 289 (#335) ############################################
Siege of Antioch
289
of their opponents. Only a fortnight or three weeks were required to
subdue the principal towns of the province. Three or four weeks more
were spent in the neighbourhood of Antioch, subduing castles there.
Iskanderun (Alexandretta) was one of Tancred's acquisitions and probably
became his headquarters. It is significant that Raymond of Toulouse and
Robert of Flanders also sent on a part of their forces to make conquests
in Northern Syria before the main body of the army arrived. Each leader
was thus fighting for his own hand and anxious not to be outdone by his
rivals. The result was that before the siege of Antioch began the Latins
had gained a secure footing in Syria and Euphratesia. These preliminary
conquests, and especially the establishment of friendly relations with the
native Christian population, were the essential conditions of further success.
It was perfectly evident soon after the main army reached Antioch
(21 October 1097) that the crusaders were not able to press the siege of
such a strongly fortified city. Lack of siege engines and the moderate
number of efficient fighting men in the army may have been contributory
causes. No attempt was made to undermine the walls or to take the town
by storm. For four or five months the city could not even be said to be
strictly invested. The Latins were encamped together, with the exception
of one small party under Tancred, just on the side where they had reached
the town.
The besieged still had almost complete liberty of exit, especi-
ally by the river gate on the north side. The fighting was only a series
of skirmishes on the plains to the north of the Orontes, and on the roads
eastward to Hārim and westward to St Simeon. Although the Turkish
garrison was not more than 5000 strong, and the auxiliary troops cannot
have been numerous', the Latin army was evidently not the overwhelming
force dreamed of by poets and imaginative historians. Still the chief
cause of the weakness of the Latin army was its deficiency in supplies.
In December 1097 and in the earlier months of 1098 the number of
horses, so vital to the strength of the army, was reduced to a dangerously
low figure? The privations of the crusaders themselves would have been
intolerable but for the assistance of their Armenian and other native
Christian allies. As many as could be spared from active service were dis-
persed through the conquered towns and castles of Cilicia along the coast
and the neighbouring country. It was not until fleets from England and
other countries arrived in the spring that the strain of the situation was
Stephen of Blois, Epist. (Hagenmeyer, Epist. et Chart. p. 150): “milites Turci
plusquam v milia. . . exceptis. . . Arabibus, Turcopolitanis. . . aliisque gentibus diversis. ”
Raymond (MPL, p. 598): “duo millia optimi milites et quatuor vel quinque mil-
lia militum gregariorum atque decem millia peditum et eo amplius. " This latter
may be regarded as a genuine attempt at an estimate, but the 10,000 foot can only
represent the whole male population of the city and that at its maximum.
According to the Gesta, not more than 1000 war horses (cf. Albert, 111, 60);
according to Anselm, Epist. (Hag. 157) only 700. The Provençals at this same time
are said to have had no more than 100 horses (Raymond d'Agiles, MPL, p. 602 c).
Compare also infra, p. 290, note.
1
C. MED, H. VOL. V, CH. VII.
19
## p. 290 (#336) ############################################
290
Latın successes
relieved. On the other hand, during the winter the Muslim garrison does
not appear to have suffered much from lack of provisions. A large part
of the non-combatant population, especially Armenian and Syrian Chris-
tians, were dismissed at the beginning of the siege. In early spring the
Muslims were still able to pasture their horses in relays outside the city.
It was only from March or April that the besieged began to suffer
serious privation. Their numbers were then reduced not only by death
but by desertion. Finally, it was the treachery of a discontented soldier
which secured an entrance for the enemy (3 June 1098).
The chief events of the siege were the battles which the crusaders
fought with the relief armies of other Syrian emirs. Yaghi Bassān, the
Turkish governor of Antioch, had no reason to expect cordial assistance
from his neighbours. They did not desert him altogether, but the ease
with which they were repulsed is as much an indication of their lukewarm-
ness as of the superiority of the Latin arms. In November, raiders who
probably came from Hārim, a strong castle on the way to Aleppo, were
ambushed and severely defeated by Bohemond, Robert of Normandy, and
Robert of Flanders. These same leaders were sent out in December to
bring in supplies, and at Al-Bārah they encountered and repulsed troops
from Damascus and Hims which were on their way to relieve Antioch
(31 December 1097). In the beginning of February 1098 the Latins
learned that a Muslim army, consisting chiefly of troops from Aleppo,
was close at hand. It was decided that they should be met a few miles away
at a narrow point on the road by the full force of the Latin cavalry, 700
strong! The foot-soldiers and unmounted men were left to guard the camp.
The Muslims were attacked where they could not employ their customary
enveloping tactics, and their crowded rear increased the confusion rather
than the strength of the ranks in front. The first charge of the crusaders
was checked, but the onset of the reserve under Bohemond was irresistible.
The Latin victory (9 February 1098) was specially welcome because it
secured fresh supplies of provisions and of horses, and was followed im-
mediately by the surrender of the strong castle of Hārim.
1 This number is given by Raymond d'Agiles, Anselm, Albert, and Cafaro.
Possibly Raymond's knights remained with him in the camp. On the other hand
a small number of foot-soldiers may have accompanied the horsemen (Anselm).
As a contribution to the vexed question of the numbers of the crusaders and
their opponents, it may be noted that Raymond in describing the other engage-
ments of this paragraph estimates the Latin knights at 150 and 400 (v. l. 300)
respectively. In December the Latins, including infantry, are said to have numbered
2000 (Gesta) or 5000 men (Hist. Pereg. , Rec. Hist. Cr. ii, 187). Presumably foot-
soldiers were employed in November also, although they are not mentioned by the
sources. The almost exclusive reliance upon the knights in February was quite ex-
ceptional. The Latins probably outnumbered their opponents in November and
December, although not in February. 3000 is a reasonable estimate of the strength
of the Muslim army in February (Cafaro). The Muslim slain, whose heads were
cut off as trophies, are reckoned at 100 (Gesta) or 200 (Anselm, Albert). It is very
unlikely that as many as 400 Muslims were slain in the fight in November (Anselm).
## p. 291 (#337) ############################################
Investment completed
291
As already observed, the investment of Antioch by the crusaders was
not complete until March or even April. The city lay at this time wholly
on the south bank of the Orontes, with its northern wall' running roughly
parallel to the river. The Latin camp was on the same side of the Orontes,
round the north-east corner of the wall. In this position the crusaders
blockaded three of the city gates, which opened here on the northern and
eastern sides. They built a bridge of boats across the river to be a means
of communication with the plain on the other side, in front of the city,
and later a fort on the hill slopes beside them to protect their exposed
flank on the south. Tancred remained separate from the main army in
occupation of a monastery on the west side of the city, no doubt in order
to maintain communication with the sea and the port of St Simeon, ten
miles away. The gate in the centre of the north wall, where it approached
the river most closely, was the principal gate of the city and opened on-
to a bridge over the Orontes. By this the Muslim garrison issued out to
intercept the provision trains, which began to come more frequently in
spring from St Simeon to the Latin camp. In front of the bridge was a
low mound with a mosque and a burying-ground upon it. In order to
frustrate the sallies of the garrison, the crusaders at length determined to
seize and fortify this post. On 1 March? Bohemond and Raymond rode
with a strong escort to St Simeon in order to obtain workers and tools
for the fortification of the mound, and with the intention of escorting a
provision train on its way to the camp. A party of the garrison set an
ambush for them as they marched back (5 March). The knights seem to
have saved themselves at the expense of their companions, many of whom
lost their lives. Meantime Godfrey and the other leaders in the camp
had become aware of what was happening, and prepared to intercept the
victorious Muslims. Bohemond and his horsemen joined the main army
in time to share in this counter-attack. The garrison attempted to rein-
force their comrades, but this only increased the magnitude of their
disaster. Next day the work of fortifying the rising ground in front of
the river gate was begun. The gravestones on the hill supplied welcome
material to the builders. The graves themselves were desecrated, to the
distress and indignation of the Muslim spectators. After the fort was
his
1 More exactly, the wall which faced N. W. Similarly, what is called the western
wall faced rather S. W.
? This date is arrived at by reckoning four days back from the day of the return,
which was 5 March according to Epist
. Lucc. (in Hagenmeyer, Epist
. et Chart.
p. 166). The interval of four days is given by Raymond d’Agiles. Hagenmeyer in
Chronologie makes the day of departure 5 March and the day of return 6 March.
3 The narratives of Raymond and the Gesta here demand special scrutiny because
they indicate how, in this case, the number of the Muslim slain was computed at
1500. The basis of the calculation seems to have been the number of bodies ex-
humed from the burying ground (Raymond). But in spite of the definite assertion of
the Gesta it is extremely improbable that those slain on 5 March had already been
buried in the cemetery, and certainly the bodies exhumed included more than these.
CH. VII.
19-2
## p. 292 (#338) ############################################
292
Surrender of Antioch
completed it was occupied by Raymond's troops. Early in April Tancred's
position was strengthened, and the only other important gate, that on the
western side, was now completely blocked. The garrison was quite un-
able to dislodge the crusaders from their new position, and provisions
could no longer be brought into the beleaguered city.
In May 1098 word reached the crusading chiefs that a great army
under the command of Karboghā of Mosul, with the approval of the
Caliph of Baghdad, was on its way to the relief of Antioch. The Latin
position was now extremely perilous. Fortunately Bohemond was already
in communication with an officer who commanded one of the western
towers, and through him the Latins gained an easy entrance into the city
on the night of 3 June. Although the citadel at the southern extremity
of the town did not surrender, the crusaders were now protected by the
walls of Antioch itself against the army of Karboghā. On 5 June the
Muslim host encamped at the “ Iron Bridge,” eight miles away, and that
same day, or the day before, a party of their horsemen was seen from the
walls of Antioch and skirmished with the Latins. From 8 June to 28
June the crusaders were besieged in Antioch. Some of the nobles lost
heart at once and deserted their comrades. The ships in the harbour of
St Simeon began to set sail, crowded with fugitives. Had Karboghā's
army arrived four days sooner, it is not improbable that the crusading
movement would have been extinguished at the gates of Antioch. As it
was, the Latins endured three weeks of continuous fighting and terrible
privation.
In these circumstances the crusaders took an unprecedented step.
Neither on the march to Antioch nor during the siege had their opera-
tions been controlled by one supreme commander. The current modern
belief that Godfrey of Bouillon was the leader of the whole Crusade has
no foundation in fact. But now it was decided that one chief should take
command and the choice of the leaders fell on Bohemond". Enthusiasm
Reading between the lines it may be concluded : (a) that the Latins found on the
hill some of the recently slain Turks, probably unburied (should Raymond's in vallo
be in valle? ); (b) that some graves were opened during the digging of the ditch of
the fort, and that this led to the deliberate desecration of other graves; (c) that the
cemetery had been used as a burying-ground by the besieged throughout the siege;
(d) that 1500 may have been an estimate of the total losses of the Muslims up to this
time; (e) that if this number was based at all on the number of the bodies exhumed
it included a very liberal allowance for those known to have been slain and not
buried in the cemetery.
1 The choice of the battle-ground and the plan of battle on 9 February were
probably due to Bohemond, and he may, therefore, have exercised the chief com-
mand during that engagement also. But there was probably no formal appointment
then such as there was now. Stephen of Blois in the winter of 1097 is known to
have been appointed provisorem et gubernatorem by the leaders in council. But he
was never distinguished as a military leader; so that his office can hardly have been
that of commander-in-chief. More probably he acted as an executive officer, and
that certain decisions of the leaders were carried out.
saw
## p. 293 (#339) ############################################
Battle with Karbõghā
293
had already been stirred by supernatural visions and by the finding of
the Holy Lance (14 June), and thus encouraged the leaders had decided
to put all to the hazard of a single battle. Bohemond's part was to direct
the preparations, to marshal the army, and to exercise the chief com-
mand during the fight. His supreme authority was to remain intact for
a fortnight beyond the day of battle. It is probably not accidental that
the chosen day (28 June) was a Monday, the second octave of the finding
of the Holy Lance.
The hazardous operation of crossing the bridge into the plain north
of the Orontes, where the Muslims lay, was accomplished without dan-
gerous interference from the enemy. Karboghā's army included the troops
of the brothers Duqāq of Damascus and Ridwān of Aleppo, who were
deadly rivals, and Arab forces upon whom small reliance could be placed.
When it was known that the Latins intended to march out from the city,
there was hot debate regarding how they should be met. Those who
wished that they should be attacked as they issued from the bridge were
overruled, and some in consequence rode away almost before the fight
began. The Latins took up their position in the plain, with their front
to the east, in three divisions, stretching from the river to the hills.
Bohemond with strong forces posted himself in the rear, facing westward.
It is not clear that the Muslims had a well-arranged plan of battle.
Evidently the Syrian, Mesopotamian, and Anatolian troops operated
separately, and their chief attack was from the west and north-west,
although their main strength faced the Latins on the east. The crusaders,
therefore, were able to transfer reinforcements from the east front to the
west, and to rout the enemy in the rear before they began their decisive
movement forward'. Karboghā, who was posted on the right flank of the
Muslim army, remained strangely inactive. When he saw that the attack
from the west had failed, he drew back to his camp, set fire to his tents,
and made off in hasty flight. The number of the Muslim slain does not
seem to have been large. Yet the Latin victory was the turning-point
in the history of the First Crusade and decisive of its ultimate success.
The defenders of the citadel of Antioch now made overtures of surrender,
and the Latins took possession in the beginning of the following week.
It was determined in council that the march on Jerusalem should not be
resumed until 1 November.
The final disposal of Antioch after its capture was complicated by
jealousies and rivalries and doubtful questions of interpretation. Certainly
1 The distribution of the Latin forces is plainly given by Raymond, and the battle
is most fully described by Albert. A large force under Qilij-Arslān threatened the
Latin rear from the N. W. and was opposed by troops taken from the Latin left. A
strong attack on Bohemond was made later by the horsemen of Aleppo and Damascus,
who seem to have crossed from the south side of the river. To repel them it was neces-
sary to bring large reinforcements from the Latin centre and right wing. Oman's Art
of War does not distinguish the two forces operating against the rear of the crusaders.
CH. VII.
## p. 294 (#340) ############################################
294
Bohemond, Prince of Antioch
it had been assigned by treaty to Alexius, but only on condition that he
brought in person a sufficient army to help the crusaders. What period
might he claim for the fulfilment of this promise? In 1097 and 1098 the
naval and military forces of the Empire were chiefly engaged in subduing
Muslim towns in the west of Asia Minor'. But in June 1098 Alexius had
already marched with a considerable army half way to Antioch, following
the road traversed in the previous year by the crusaders themselves. Un-
fortunately for all concerned, he listened at Philomelium to the alarmist
stories of Stephen of Blois and the other fugitives from Antioch who met
him there. They probably told him that the crusading host had been
irretrievably defeated, and that a Turkish army was already marching
against him. He turned back to protect his recent conquests in Asia
Minor. Naturally this action was judged by most to be a surrender of
the Latin cause. At the best Alexius was now in a position hard to
retrieve. There are two accounts of the message which the crusaders
sent him in July. Albert of Aix says that the envoys were instructed
to tell the Emperor that he had been untrue to his promise, and thus
had nullified his treaty. This may have been the opinion of most of the
Latin leaders, but, as their attitude in November shewed, they were not
yet prepared completely to break off relations with the Emperor. The
Gesta Francorum says that the envoys were told to invite Alexius to
fulfil his promise and come to receive possession of Antioch. It may be
that something of this kind was said, with qualifications, setting a limit
to the delay which would be considered reasonable, and referring to the
Emperor's recent retreat. Presumably the envoys were empowered to
adhere in substance to the original treaty, provided the Emperor agreed
to carry out his engagements effectively and quickly. It is not known
what reply Alexius sent to this communication. It may be that he felt
the difficulty of his position so keenly that he sent no immediate reply.
In the spring of 1099 he promised to join the crusaders with an army
on St John's day (24 June), if they would wait for him until then.
Perhaps he was encouraged by the support of Raymond of Toulouse.
But his proposal came too late. The Crusade was nearing a successful
conclusion without the Emperor's assistance. All the leaders except
Raymond now held that the treaty had lapsed, and that the Emperor
had not fulfilled his obligation.
Bohemond, Prince of Antioch as he now became, profited most by the
Emperor's mistake. Before the capture of the city he had manoeuvred
dexterously to establish his claim to it. Under pressure of Karboghā's
approach, the leaders had reluctantly assented to his proposal that the
lordship of Antioch should fall to anyone who secured its capture or be-
trayal. Before Bohemond made this proposal he had arranged for the
betrayal of the town. Of course the rights of the Emperor were duly
reserved, but after the defeat of Karboghā's army Bohemond was practi-
1 E. g. Smyrna, Sardis, and the towns in the district of the Meander valley.
## p. 295 (#341) ############################################
March to Jerusalem
295
cally ruler of Antioch'. In November he urged that the Emperor's claim
had already lapsed.
remained in command, and finally crossed into Asia Minor without
entering Constantinople. Bohemond was an hereditary enemy of the
Greek Empire, and now as at all times ready to take up arms against
Alexius if he saw any advantage in doing so.
He intended to secure a
princedom in the East, and most probably had already fixed his choice
on Antioch. Before taking the oath of allegiance he endeavoured to
obtain a promise from the Emperor to support his scheme. Alexius?
answer no doubt was that such requests were premature, and that every-
thing would depend on the issue of the Crusade. It is unlikely, in spite
of the definite statement of the Gesta Francorum, that Bohemond was
now promised territory in the neighbourhood of Antioch. At most the
Emperor may have indicated that he would afterwards consider favour-
ably such claims as the Norman chief might be able to present.
Robert of Flanders accomplished the first part of his journey through
France and Italy in the company of Robert of Normandy. He crossed
from Apulia in December 1096, and did not advance farther towards
1 The attitude of the modern historian to Anna Comnena's narrative of these
events is decisively important for the view to be taken of them. Her account is inter-
preted by some to mean that Alexius finally compelled Godfrey, by force of arms, to
take the oath of allegiance. This is no doubt the impression she conveys, but in
view of Albert's narrative the Greek account may be regarded as patriotically over-
drawn at this point. The view that the essential matter in dispute between Godfrey
and Alexius was the oath of allegiance is a possible one, but Godfrey's yielding (with-
out defeat) on such a point is more difficult to understand than Alexius' yielding on
the point of the hostages. The very improbable date that Anna gives quite incident-
ally at one point (“Thursday in boly week”= 2 April 1097) has not been satisfactorily
accounted for by those who reject it (cf. Kugler). May there not be a confusion with
Christmas week, in which, according to Albert, there was a cessation of hostilities
out of respect to the season? Christmas Day in 1096 was a Thursday.
## p. 283 (#329) ############################################
Raymond of Toulouse
283
Constantinople until the spring. He arrived later than Bohemond, and
readily took the oath of allegiance.
Raymond of Toulouse, having left home, perhaps, about the end of
October 1096, came by the north of Italy and the eastern shore of the
Adriatic Sea. Passing through Dalmatia in the winter, his army suffered
from the inclemency of the season, from scarcity of food, and from the
attacks of the inhabitants of the country, so that large numbers of the
crusaders lost their lives. At Durazzo messengers from the Emperor
brought assurances of friendship and promised supplies. Beyond this
point, however, there was frequent fighting between the crusaders and the
Greek mercenaries who watched their progress. The Provençals con-
sidered themselves the aggrieved parties, and retaliated by destroying the
suburbs of Rusa and plundering the town. At Rodosto, four days'
journey from Constantinople, Raymond received a request from the Latin
leaders already in Constantinople to hurry on, because they were pre-
paring to start and were making arrangements with the Emperor to
which it was desirable that he should be a party. When he reached
Constantinople (perhaps in the third week of April), he decisively refused
to take the now customary oath of allegiance. If the Emperor put him-
self at the head of the expedition and came with them, he would become
his follower, he said, not otherwise. News of a shameful defeat of his
army, in a conflict in which they were afterwards judged to have been in
the wrong, only increased his determination not to yield. Finally, under
pressure, he only consented to take an oath that he would do nothing
against the life and honour of the Emperor. In consequence of his
attitude he received, as the Provençal historian notes, little of the
Emperor's bounty.
Last of all came Robert of Normandy, with his powerful brother-in-
law, Stephen of Blois, and with Godfrey's brother Eustace, Count of
Boulogne. Their army included the first expedition of “Englishmen
and Britons” to join in the Crusade. Robert left home in September
and had spent the winter in the south of Italy. He embarked at
Brindisi on 5 April 1097, and reached Constantinople about the middle
of May. After spending a fortnight in the Greek capital he proceeded to
the siege of Nicaea, which had already begun.
The Emperor Alexius had good reason to be satisfied with the initial
result of his negotiations with the Latins. Formally, at least, he had
secured from the leaders of the Crusade the acknowledgment he desired.
Even Raymond of Toulouse seems finally to have admitted the Emperor's
claims in Asia Minor and Syria? An agreement so important and so
i The date rests on a doubtful calculation. Vic and Vaissete, Histoire générale
de Languedoc (1733), Vol. 11, p. 628.
2 It is not easy to say when Raymond became the friend and partisan of the
Emperor, which he shewed himself to be in Antioch. Albert, xi, 20, implies that it
was previous to the march of the crusaders through Asia Minor. The Emperor may
CD, VII.
## p. 284 (#330) ############################################
284
Treaty with Alexius
intricate must have been put in writing and signed by the contracting
parties? If it did not specify all the lands which the Emperor claimed,
it probably named at least the territories and towns in which he desired
to place Greek governors, and some also of those which might be held by
the Latins in fief. The plunder of all the captured cities may have been
assigned to the Latins, and the Emperor certainly promised military
assistance to his allies. The obligations of the Latin feudatories must
have been defined, and, it may be, also the conditions on which they
would obtain recognition as lords of the conquered territory. Of course
the adherence of the crusaders to this agreement depended entirely on
the Emperor's fulfilment of his promise to render them further assistance.
If he failed in this obligation, the Latins were inevitably released from
their pledges to him. But meantime the leaders were won partly by
the personal charm and lavish gifts of the Emperor, partly, it may be
added, by the reasonable character of his proposals, so that they judged
their treaty with him to be of value to their enterprise. It is true that
there was at the same time, especially among the rank and file, a strong
undercurrent of suspicion and hatred of the Greeks. Godfrey's troops
and Raymond's had already been engaged in serious fighting with them.
have gained Raymond's concession by a definite pledge, in the formal treaty which
regulated the relationship of the parties, to bring an army to the assistance of the
Latins. Even Raymond's oath in Constantinople may have guaranteed the Emperor's
territorial claims, although it was not an oath of allegiance. To the view that Ray-
mond suddenly changed his attitude in Antioch out of hostility to Bohemond, it may
be objected that there is no suggestion in the sources of any inconsistency in his
attitude, nor of any difference between the obligations of the individual Latin leaders
towards the Emperor. The hostility of the Provençal historian, Raymond of Agiles,
expressed in his description of the occurrences at Nicaea and elsewhere, is not a certain
indication of the attitude of Raymond of Toulouse at the time. This historian, and the
Provençals generally, never shared Raymond's partisanship for the Emperor. Ray-
mond's remaining in the camp at Nicaea in June 1097, when the other chiefs went to
visit the Emperor after the capture of the city (Hagenmeyer, Epist. et Chart. p. 140),
must be regarded as merely incidental, on the view here taken of his final reconcili-
ation with the Emperor.
1 This is also F. Chalandon's view. There is a small amount of documentary
evidence in its favour, viz. Alexias, xii, 12, where the reference to a treaty with
Bohemond may be understood of a general treaty made with the Latin chiefs, and
Hagenmeyer, Epist. et Chart. No. 12, from which it may be inferred that the date of
the agreement was the middle of May. The terms of the treaty, as these are sug-
gested above, are inferred from narratives of what took place after the captures of
Nicaea and Antioch and from the terms of the later treaty with Bohemond in 1108
(Alexias, xii, 12). The alleged grant to Bohemond of territory in the neighbour-
hood of Antioch at this time (Gesta Francorum) may be interpreted to mean that this
territory, although not assigned to any individual, was actually designated in the
treaty as a prospective Latin fief. The exclusion of the city of Antioch suggests that
it was to be placed in the hands of a Greek governor. Albert, v, 2, states that the
Latin leaders had sworn specifically to hand over Antioch and Nicaea (if captured
by them) to the Emperor (cum omnibus castellis et urbibus ad regnum eius perti-
nentibus). Raymond, ch. iv, shews that certain rights were conceded to the Latins
in these Greek towns (in Nicaea a monastery and a hospital).
## p. 285 (#331) ############################################
Siege of Nicaea
285
The Normans were really bitter and contemptuous enemies of the Greeks,
although Bohemond judged it to be expedient to acquiesce in a general
treaty, and required Tancred, much against his will, to take the common
oath of allegiance. At the same time the marked hostility of the
Western sources to the Emperor in their narratives of these events
reflects largely the anger and disappointment of a later period. The
Greeks and Latins had important interests in common, and it is likely
that the policy inaugurated by the Emperor would have held them
together until at least the foundations were laid in Syria of one Greco-
Latin state. It was Alexius' own failure to implement his promise
that finally turned the Latins into declared and irreconcilable enemies.
Before the Latins left Constantinople, their route through Asia Minor
and their plan of operations had been decided on. In the first place the
Muslim capital of Nicaea, about six days' march overland from Scutari,
was to be taken. The Emperor provided siege engines and food
supplies but only a small detachment of troops.
Nicaea was very
strongly fortified and was protected on the west side by the waters of
a lake. The disposition of the crusading army illustrates the separation
caused by national divisions. Bohemond's forces encamped on the north,
Godfrey and the Germans on the east of the city (6 May 1097). When
Raymond's troops arrived they occupied the south side (16 May). On
the day of Raymond's arrival a small force of Muslims attempted to
throw themselves into the city and were beaten off. Robert of Normandy
and his men joined the besiegers on 1 June; their position also was on
the south side. The siege operations, begun on 14 May, were pressed
strenuously with little result for nearly five weeks. At length the ruin
by Raymond's engineers of a large tower on the south side brightened
the prospects of the besiegers. This and the launching on the lake of
Greek vessels, brought from the sea, decided the defenders to surrender.
They opened negotiations with the Greek commander, and capitulated
to him on condition that their lives should be spared (19 June 1097).
Most likely they were allowed to remain undisturbed in their homes if
they chose to transfer their allegiance to the Emperor. In order to
prevent wanton plundering and destruction, the Latins were allowed
to enter the city as visitors only and in small parties. As previously
arranged, the spoil of the town, or its equivalent, was distributed among
the crusaders, and their leaders received in addition handsome gifts from
Alexius. No doubt the sparing of the lives of infidels became a cause
of reproach to the Emperor in the Latin camp, and perhaps the pre-
cautions taken to protect the city from plundering were resented. But
the Latins do not seem, on this occasion, to have been unfairly treated',
and some of them settled in Nicaea as the Emperor's subjects.
1 Raymond's account brings no specific charge against Alexius, although it shews
that he was disliked and hated by the Provençals.
CH. VII.
## p. 286 (#332) ############################################
286
March through Asia Minor
After the capture of Nicaea the proximate goal of the crusaders' march
was Antioch on the Orontes. It may be assumed that Alexius urged the
siege and capture of a city which had been for a century an outpost of the
Empire, and the occupation of which would be an important initial step
in the conquest of Syria. Besides, the deliverance” of Antioch had been
from the first one of the specific objects of the Crusade. The way through
Asia Minor was familiar to the Greeks and in any case easily found and
followed. It leads through Dorylaeum and Iconium and then over the
passes of the Taurus into Cilicia. But in order to rescue Armenia Minor
from the Muslim yoke and to secure for themselves friendly support in a
district near Antioch, the main body of the crusaders kept eastwards to
the anti-Taurus mountains, and then came southward to Antioch by way
of Geuksun (Coxon) and Marʻash. Cilicia, in which there was also a
friendly Armenian population, was secured by Tancred and by Godfrey's
brother Baldwin. The Latins sent letters to the Armenians of Euphra-
tesia, most probably from Nicaea, and Baldwin was joined there by an
Armenian exile who accompanied and advised him during the march
through Asia Minor. This alliance with the Armenians was afterwards
of great value during the siege of Antioch, and by it the crusaders were
enabled to make their first settlements in the East.
When the Latins left Nicaea—those who moved first started on
27 June—some cherished the hope that they might reach Jerusalem in
five weeks, if Antioch did not prove a serious obstacle in their way.
It was three months before they approached Antioch and nearly two
years before any of them reached Jerusalem. Qilij-Arslān, having
assembled his army too late to save Nicaea, attacked the Latins near
Dorylaeum on 1 July. The crusaders were in two divisions, two miles apart,
on separate roads. The first encounter was between the whole Turkish
army, which consisted exclusively of horsemen armed with bows, and the
smaller part of the Latin host, which included the Normans only, under
Bohemond, Tancred, and Robert of Normandy. An attack of the Norman
knights was repulsed by the Turks, whose advance, in turn, was checked
by the spears and bows of the Latin infantry, upon whom the knights
fell back'. The encircling Muslims now employed their usual elusive and
harassing tactics and the Normans fought a desperate battle, until they
were relieved by the arrival, in successive bands, of Godfrey and the other
leaders. The Turks having retreated on to a hill-side, the crusaders
formed themselves into line of battle and broke and scattered their
opponents by one irresistible charge. In the shock of direct encounter
the light Turkish horsemen had no chance of success. The fight before
1 The fullest account of the first stage of the battle is given by the Gesta Tan-
credi. This source alone explicitly mentions the decisive action of the infantry (deu-
sissima pedestrium hastarum silva nunc fugam impedit, nunc extiuguit). The part
played by the Latin infantry in these battles with the Muslims is frequently ignored
by the sources (cf. infra, p. 290, note 1).
## p. 287 (#333) ############################################
Alliance with the Armenians
287
Godfrey's arrival may have lasted two or three hours and the second
stage of the battle, including the pursuit, three hours more. The enemy
were pursued for several miles, and great booty was obtained from the
captured Muslim camp'.
During the march beyond Dorylaeum the Latins found the country
laid waste for a considerable distance, and suffered greatly from want of
food and water as well as from the excessive heat. They lost a large
number of their horses and baggage animals. Most probably the crusaders
now marched in one main force, where all the baggage was placed, and in
several smaller forces under independent leaders such as Tancred and
Baldwin. From Iconium eastwards the conditions seem to have improved,
and of course in Armenia Minor the friendship of the Christian popula-
tion made the way easy. The Muslims were nowhere in sufficient force to
venture another attack after their defeat at Dorylaeum. In Armenia
Minor the Turkish garrisons, which had not long been in possession, were
expelled and Armenian supremacy was restored. Several Western knights
settled in the conquered strongholds, but the only leader of importance
to remain in the district was Baldwin, Godfrey's brother, afterwards his
successor in Jerusalem. Baldwin was the founder of the first of the
crusading states in the East. After passing, as we shall see, through
Cilicia he reached the main army at Marʻash. But while it went on to
Antioch he remained to establish a Latin princedom in Euphratesia. His
first capital was Tell-bāshir (October 1097). Afterwards, when he became
ruler of Edessa (spring of 1098), he made that city his capital. His
forces in themselves were not at first large, but the friendship of the
Armenian princes secured his position. After the fall of Antioch, Godfrey
came to his assistance, and from that time he was quite able to maintain
himself. Undoubtedly, if the Latins had continued to co-operate with the
Armenians, this northern state would have been a much more effective
bulwark of their power than it ultimately proved to be.
No doubt Baldwin's settlement in Edessa was made with the consent
and approval of the Latin leaders. It was in some measure due to him,
since he had recently resigned to Tancred his claim on Cilicia. As the
crusaders arrived in the districts where the first permanent conquests
were attempted, it became perfectly clear that each leader fought not merely
for the common cause but also for a share in the territory that was being
1 Oman's representation that Godfrey and the others were six or seven miles
distant from the Normans and that Bohemond fought alone for five hours (Art of
War, 1st ed. p. 275 (2nd ed. 1, p. 277]), seems to lack foundation. The distance between
the divisions was two miles according to Raymond, who is confirmed by Albert's state-
ment (11, 35) that the intention was that the divisions should march one mile apart.
The Gesta Fruncorum makes the battle last from the third hour to the ninth, and
Fulcher says that the enemy were routed at the sixth hour. This seems to indicate
that the six miles pursuit (Albert, 11, 42) lasted about three hours. Fulcher is vague
regarding the time of the commencement of the battle, for which he names both
the second hour and the first.
CH. VII.
## p. 288 (#334) ############################################
288
Inter-national rivalries
conquered. In the smaller undertakings each national army made its own
conquests and of course claimed to retain what it thus won. The events
in Cilicia are narrated at full length by the sources, and may be taken as
the best available illustration of what has just been said. The occupation
of this province was probably part of a general scheme suggested by
Armenians who accompanied the crusaders from Nicaea, and it may have
been included in the plans of the leaders from the time they left that
city. But since Baldwin and Tancred were rivals in their operations in
that province from the first, it is not hazardous to conclude that one at
least was deputed to protect national interests against the action of the
other? . Tancred left the main army at Heraclea and made directly for
Tarsus, which he hoped to gain with the help of Armenian friends. He
had encamped beside the town and was negotiating its surrender, with
every prospect of success, when Baldwin came on the scene with much
larger forces. It is uncertain whether Baldwin had left the main army
at Heraclea, or had separated from it much earlier than Tancred and had
reached Tarsus by a different road. The result of his arrival was that the
Turkish garrison deserted the town and the inhabitants prepared to sur-
render formally to Tancred. Baldwin, in virtue of his superior strength,
required them, however, to surrender to him, and Tancred retired in anger
without fighting. At Adana he found the Turks already driven out and
an Armenian governor installed", from whom he received a welcome.
Mamistra, the next town on the way, was occupied without difficulty, for
the garrison Hled almost as soon as Tancred approached the city. Mean-
time Baldwin was joined at Tarsus by a fleet of Flemings and Frisians,
which had been cruising for some years in the Mediterranean and was
commanded by a certain Winemar of Boulogne. Having left a garrison
in Tarsus, Baldwin marched on to Mamistra, where he encamped outside
the walls. Either party may have been the aggressor in the fighting which
followed. But Baldwin had now designs further east, so that peace was
quickly re-established and Tancred was left in possession. Beforethe Norman
leader left Cilicia, he had established a claim to possession which Bohemond
and he, as princes of Antioch, afterwards strenuously maintained against
the armies of the Empire. Meantime most of the population favoured
the Latins, and the small Turkish garrisons were cowed by the numbers
1 If both left the army at Heraclea they may have been declared rivals from the
first, or Tancred may possibly have tried to outstrip Baldwin and thus have provoked
the final conflict. On the other hand, if Baldwin left the main army considerably
earlier on the march from Nicaea, then Tancred must have entered Cilicia with the
purpose of securing the province for the Normans with the knowledge that he would
probably have to compete with Baldwin for its possession. Both Baldwin and Tancred
were advised by Armenian friends.
2 The Gesta Tancredi estimates Tancred's force at 100 knights (loricas) and 200
bowmen (ch. xxxIII), Baldwiu's at 500 horsemen and 2000 foot soldiers (ch. xxxvII).
Ralph of Caen. Albert makes the governor Welf; a Burgundian, who had also
separated himself from the main army.
3
## p. 289 (#335) ############################################
Siege of Antioch
289
of their opponents. Only a fortnight or three weeks were required to
subdue the principal towns of the province. Three or four weeks more
were spent in the neighbourhood of Antioch, subduing castles there.
Iskanderun (Alexandretta) was one of Tancred's acquisitions and probably
became his headquarters. It is significant that Raymond of Toulouse and
Robert of Flanders also sent on a part of their forces to make conquests
in Northern Syria before the main body of the army arrived. Each leader
was thus fighting for his own hand and anxious not to be outdone by his
rivals. The result was that before the siege of Antioch began the Latins
had gained a secure footing in Syria and Euphratesia. These preliminary
conquests, and especially the establishment of friendly relations with the
native Christian population, were the essential conditions of further success.
It was perfectly evident soon after the main army reached Antioch
(21 October 1097) that the crusaders were not able to press the siege of
such a strongly fortified city. Lack of siege engines and the moderate
number of efficient fighting men in the army may have been contributory
causes. No attempt was made to undermine the walls or to take the town
by storm. For four or five months the city could not even be said to be
strictly invested. The Latins were encamped together, with the exception
of one small party under Tancred, just on the side where they had reached
the town.
The besieged still had almost complete liberty of exit, especi-
ally by the river gate on the north side. The fighting was only a series
of skirmishes on the plains to the north of the Orontes, and on the roads
eastward to Hārim and westward to St Simeon. Although the Turkish
garrison was not more than 5000 strong, and the auxiliary troops cannot
have been numerous', the Latin army was evidently not the overwhelming
force dreamed of by poets and imaginative historians. Still the chief
cause of the weakness of the Latin army was its deficiency in supplies.
In December 1097 and in the earlier months of 1098 the number of
horses, so vital to the strength of the army, was reduced to a dangerously
low figure? The privations of the crusaders themselves would have been
intolerable but for the assistance of their Armenian and other native
Christian allies. As many as could be spared from active service were dis-
persed through the conquered towns and castles of Cilicia along the coast
and the neighbouring country. It was not until fleets from England and
other countries arrived in the spring that the strain of the situation was
Stephen of Blois, Epist. (Hagenmeyer, Epist. et Chart. p. 150): “milites Turci
plusquam v milia. . . exceptis. . . Arabibus, Turcopolitanis. . . aliisque gentibus diversis. ”
Raymond (MPL, p. 598): “duo millia optimi milites et quatuor vel quinque mil-
lia militum gregariorum atque decem millia peditum et eo amplius. " This latter
may be regarded as a genuine attempt at an estimate, but the 10,000 foot can only
represent the whole male population of the city and that at its maximum.
According to the Gesta, not more than 1000 war horses (cf. Albert, 111, 60);
according to Anselm, Epist. (Hag. 157) only 700. The Provençals at this same time
are said to have had no more than 100 horses (Raymond d'Agiles, MPL, p. 602 c).
Compare also infra, p. 290, note.
1
C. MED, H. VOL. V, CH. VII.
19
## p. 290 (#336) ############################################
290
Latın successes
relieved. On the other hand, during the winter the Muslim garrison does
not appear to have suffered much from lack of provisions. A large part
of the non-combatant population, especially Armenian and Syrian Chris-
tians, were dismissed at the beginning of the siege. In early spring the
Muslims were still able to pasture their horses in relays outside the city.
It was only from March or April that the besieged began to suffer
serious privation. Their numbers were then reduced not only by death
but by desertion. Finally, it was the treachery of a discontented soldier
which secured an entrance for the enemy (3 June 1098).
The chief events of the siege were the battles which the crusaders
fought with the relief armies of other Syrian emirs. Yaghi Bassān, the
Turkish governor of Antioch, had no reason to expect cordial assistance
from his neighbours. They did not desert him altogether, but the ease
with which they were repulsed is as much an indication of their lukewarm-
ness as of the superiority of the Latin arms. In November, raiders who
probably came from Hārim, a strong castle on the way to Aleppo, were
ambushed and severely defeated by Bohemond, Robert of Normandy, and
Robert of Flanders. These same leaders were sent out in December to
bring in supplies, and at Al-Bārah they encountered and repulsed troops
from Damascus and Hims which were on their way to relieve Antioch
(31 December 1097). In the beginning of February 1098 the Latins
learned that a Muslim army, consisting chiefly of troops from Aleppo,
was close at hand. It was decided that they should be met a few miles away
at a narrow point on the road by the full force of the Latin cavalry, 700
strong! The foot-soldiers and unmounted men were left to guard the camp.
The Muslims were attacked where they could not employ their customary
enveloping tactics, and their crowded rear increased the confusion rather
than the strength of the ranks in front. The first charge of the crusaders
was checked, but the onset of the reserve under Bohemond was irresistible.
The Latin victory (9 February 1098) was specially welcome because it
secured fresh supplies of provisions and of horses, and was followed im-
mediately by the surrender of the strong castle of Hārim.
1 This number is given by Raymond d'Agiles, Anselm, Albert, and Cafaro.
Possibly Raymond's knights remained with him in the camp. On the other hand
a small number of foot-soldiers may have accompanied the horsemen (Anselm).
As a contribution to the vexed question of the numbers of the crusaders and
their opponents, it may be noted that Raymond in describing the other engage-
ments of this paragraph estimates the Latin knights at 150 and 400 (v. l. 300)
respectively. In December the Latins, including infantry, are said to have numbered
2000 (Gesta) or 5000 men (Hist. Pereg. , Rec. Hist. Cr. ii, 187). Presumably foot-
soldiers were employed in November also, although they are not mentioned by the
sources. The almost exclusive reliance upon the knights in February was quite ex-
ceptional. The Latins probably outnumbered their opponents in November and
December, although not in February. 3000 is a reasonable estimate of the strength
of the Muslim army in February (Cafaro). The Muslim slain, whose heads were
cut off as trophies, are reckoned at 100 (Gesta) or 200 (Anselm, Albert). It is very
unlikely that as many as 400 Muslims were slain in the fight in November (Anselm).
## p. 291 (#337) ############################################
Investment completed
291
As already observed, the investment of Antioch by the crusaders was
not complete until March or even April. The city lay at this time wholly
on the south bank of the Orontes, with its northern wall' running roughly
parallel to the river. The Latin camp was on the same side of the Orontes,
round the north-east corner of the wall. In this position the crusaders
blockaded three of the city gates, which opened here on the northern and
eastern sides. They built a bridge of boats across the river to be a means
of communication with the plain on the other side, in front of the city,
and later a fort on the hill slopes beside them to protect their exposed
flank on the south. Tancred remained separate from the main army in
occupation of a monastery on the west side of the city, no doubt in order
to maintain communication with the sea and the port of St Simeon, ten
miles away. The gate in the centre of the north wall, where it approached
the river most closely, was the principal gate of the city and opened on-
to a bridge over the Orontes. By this the Muslim garrison issued out to
intercept the provision trains, which began to come more frequently in
spring from St Simeon to the Latin camp. In front of the bridge was a
low mound with a mosque and a burying-ground upon it. In order to
frustrate the sallies of the garrison, the crusaders at length determined to
seize and fortify this post. On 1 March? Bohemond and Raymond rode
with a strong escort to St Simeon in order to obtain workers and tools
for the fortification of the mound, and with the intention of escorting a
provision train on its way to the camp. A party of the garrison set an
ambush for them as they marched back (5 March). The knights seem to
have saved themselves at the expense of their companions, many of whom
lost their lives. Meantime Godfrey and the other leaders in the camp
had become aware of what was happening, and prepared to intercept the
victorious Muslims. Bohemond and his horsemen joined the main army
in time to share in this counter-attack. The garrison attempted to rein-
force their comrades, but this only increased the magnitude of their
disaster. Next day the work of fortifying the rising ground in front of
the river gate was begun. The gravestones on the hill supplied welcome
material to the builders. The graves themselves were desecrated, to the
distress and indignation of the Muslim spectators. After the fort was
his
1 More exactly, the wall which faced N. W. Similarly, what is called the western
wall faced rather S. W.
? This date is arrived at by reckoning four days back from the day of the return,
which was 5 March according to Epist
. Lucc. (in Hagenmeyer, Epist
. et Chart.
p. 166). The interval of four days is given by Raymond d’Agiles. Hagenmeyer in
Chronologie makes the day of departure 5 March and the day of return 6 March.
3 The narratives of Raymond and the Gesta here demand special scrutiny because
they indicate how, in this case, the number of the Muslim slain was computed at
1500. The basis of the calculation seems to have been the number of bodies ex-
humed from the burying ground (Raymond). But in spite of the definite assertion of
the Gesta it is extremely improbable that those slain on 5 March had already been
buried in the cemetery, and certainly the bodies exhumed included more than these.
CH. VII.
19-2
## p. 292 (#338) ############################################
292
Surrender of Antioch
completed it was occupied by Raymond's troops. Early in April Tancred's
position was strengthened, and the only other important gate, that on the
western side, was now completely blocked. The garrison was quite un-
able to dislodge the crusaders from their new position, and provisions
could no longer be brought into the beleaguered city.
In May 1098 word reached the crusading chiefs that a great army
under the command of Karboghā of Mosul, with the approval of the
Caliph of Baghdad, was on its way to the relief of Antioch. The Latin
position was now extremely perilous. Fortunately Bohemond was already
in communication with an officer who commanded one of the western
towers, and through him the Latins gained an easy entrance into the city
on the night of 3 June. Although the citadel at the southern extremity
of the town did not surrender, the crusaders were now protected by the
walls of Antioch itself against the army of Karboghā. On 5 June the
Muslim host encamped at the “ Iron Bridge,” eight miles away, and that
same day, or the day before, a party of their horsemen was seen from the
walls of Antioch and skirmished with the Latins. From 8 June to 28
June the crusaders were besieged in Antioch. Some of the nobles lost
heart at once and deserted their comrades. The ships in the harbour of
St Simeon began to set sail, crowded with fugitives. Had Karboghā's
army arrived four days sooner, it is not improbable that the crusading
movement would have been extinguished at the gates of Antioch. As it
was, the Latins endured three weeks of continuous fighting and terrible
privation.
In these circumstances the crusaders took an unprecedented step.
Neither on the march to Antioch nor during the siege had their opera-
tions been controlled by one supreme commander. The current modern
belief that Godfrey of Bouillon was the leader of the whole Crusade has
no foundation in fact. But now it was decided that one chief should take
command and the choice of the leaders fell on Bohemond". Enthusiasm
Reading between the lines it may be concluded : (a) that the Latins found on the
hill some of the recently slain Turks, probably unburied (should Raymond's in vallo
be in valle? ); (b) that some graves were opened during the digging of the ditch of
the fort, and that this led to the deliberate desecration of other graves; (c) that the
cemetery had been used as a burying-ground by the besieged throughout the siege;
(d) that 1500 may have been an estimate of the total losses of the Muslims up to this
time; (e) that if this number was based at all on the number of the bodies exhumed
it included a very liberal allowance for those known to have been slain and not
buried in the cemetery.
1 The choice of the battle-ground and the plan of battle on 9 February were
probably due to Bohemond, and he may, therefore, have exercised the chief com-
mand during that engagement also. But there was probably no formal appointment
then such as there was now. Stephen of Blois in the winter of 1097 is known to
have been appointed provisorem et gubernatorem by the leaders in council. But he
was never distinguished as a military leader; so that his office can hardly have been
that of commander-in-chief. More probably he acted as an executive officer, and
that certain decisions of the leaders were carried out.
saw
## p. 293 (#339) ############################################
Battle with Karbõghā
293
had already been stirred by supernatural visions and by the finding of
the Holy Lance (14 June), and thus encouraged the leaders had decided
to put all to the hazard of a single battle. Bohemond's part was to direct
the preparations, to marshal the army, and to exercise the chief com-
mand during the fight. His supreme authority was to remain intact for
a fortnight beyond the day of battle. It is probably not accidental that
the chosen day (28 June) was a Monday, the second octave of the finding
of the Holy Lance.
The hazardous operation of crossing the bridge into the plain north
of the Orontes, where the Muslims lay, was accomplished without dan-
gerous interference from the enemy. Karboghā's army included the troops
of the brothers Duqāq of Damascus and Ridwān of Aleppo, who were
deadly rivals, and Arab forces upon whom small reliance could be placed.
When it was known that the Latins intended to march out from the city,
there was hot debate regarding how they should be met. Those who
wished that they should be attacked as they issued from the bridge were
overruled, and some in consequence rode away almost before the fight
began. The Latins took up their position in the plain, with their front
to the east, in three divisions, stretching from the river to the hills.
Bohemond with strong forces posted himself in the rear, facing westward.
It is not clear that the Muslims had a well-arranged plan of battle.
Evidently the Syrian, Mesopotamian, and Anatolian troops operated
separately, and their chief attack was from the west and north-west,
although their main strength faced the Latins on the east. The crusaders,
therefore, were able to transfer reinforcements from the east front to the
west, and to rout the enemy in the rear before they began their decisive
movement forward'. Karboghā, who was posted on the right flank of the
Muslim army, remained strangely inactive. When he saw that the attack
from the west had failed, he drew back to his camp, set fire to his tents,
and made off in hasty flight. The number of the Muslim slain does not
seem to have been large. Yet the Latin victory was the turning-point
in the history of the First Crusade and decisive of its ultimate success.
The defenders of the citadel of Antioch now made overtures of surrender,
and the Latins took possession in the beginning of the following week.
It was determined in council that the march on Jerusalem should not be
resumed until 1 November.
The final disposal of Antioch after its capture was complicated by
jealousies and rivalries and doubtful questions of interpretation. Certainly
1 The distribution of the Latin forces is plainly given by Raymond, and the battle
is most fully described by Albert. A large force under Qilij-Arslān threatened the
Latin rear from the N. W. and was opposed by troops taken from the Latin left. A
strong attack on Bohemond was made later by the horsemen of Aleppo and Damascus,
who seem to have crossed from the south side of the river. To repel them it was neces-
sary to bring large reinforcements from the Latin centre and right wing. Oman's Art
of War does not distinguish the two forces operating against the rear of the crusaders.
CH. VII.
## p. 294 (#340) ############################################
294
Bohemond, Prince of Antioch
it had been assigned by treaty to Alexius, but only on condition that he
brought in person a sufficient army to help the crusaders. What period
might he claim for the fulfilment of this promise? In 1097 and 1098 the
naval and military forces of the Empire were chiefly engaged in subduing
Muslim towns in the west of Asia Minor'. But in June 1098 Alexius had
already marched with a considerable army half way to Antioch, following
the road traversed in the previous year by the crusaders themselves. Un-
fortunately for all concerned, he listened at Philomelium to the alarmist
stories of Stephen of Blois and the other fugitives from Antioch who met
him there. They probably told him that the crusading host had been
irretrievably defeated, and that a Turkish army was already marching
against him. He turned back to protect his recent conquests in Asia
Minor. Naturally this action was judged by most to be a surrender of
the Latin cause. At the best Alexius was now in a position hard to
retrieve. There are two accounts of the message which the crusaders
sent him in July. Albert of Aix says that the envoys were instructed
to tell the Emperor that he had been untrue to his promise, and thus
had nullified his treaty. This may have been the opinion of most of the
Latin leaders, but, as their attitude in November shewed, they were not
yet prepared completely to break off relations with the Emperor. The
Gesta Francorum says that the envoys were told to invite Alexius to
fulfil his promise and come to receive possession of Antioch. It may be
that something of this kind was said, with qualifications, setting a limit
to the delay which would be considered reasonable, and referring to the
Emperor's recent retreat. Presumably the envoys were empowered to
adhere in substance to the original treaty, provided the Emperor agreed
to carry out his engagements effectively and quickly. It is not known
what reply Alexius sent to this communication. It may be that he felt
the difficulty of his position so keenly that he sent no immediate reply.
In the spring of 1099 he promised to join the crusaders with an army
on St John's day (24 June), if they would wait for him until then.
Perhaps he was encouraged by the support of Raymond of Toulouse.
But his proposal came too late. The Crusade was nearing a successful
conclusion without the Emperor's assistance. All the leaders except
Raymond now held that the treaty had lapsed, and that the Emperor
had not fulfilled his obligation.
Bohemond, Prince of Antioch as he now became, profited most by the
Emperor's mistake. Before the capture of the city he had manoeuvred
dexterously to establish his claim to it. Under pressure of Karboghā's
approach, the leaders had reluctantly assented to his proposal that the
lordship of Antioch should fall to anyone who secured its capture or be-
trayal. Before Bohemond made this proposal he had arranged for the
betrayal of the town. Of course the rights of the Emperor were duly
reserved, but after the defeat of Karboghā's army Bohemond was practi-
1 E. g. Smyrna, Sardis, and the towns in the district of the Meander valley.
## p. 295 (#341) ############################################
March to Jerusalem
295
cally ruler of Antioch'. In November he urged that the Emperor's claim
had already lapsed.
