The Emperor Michael III and the Battle of Bishop's Meadow (A.D. 863) Huxley, George Greek, Roman and Byzantine Studies; Winter 1975; 16, 4; ProQuest pg. 443

The Emperor Michael III and the 
Battle of Bishop's Meadow (A.D. 863) 

George Huxley 

THE FOUR BOOKS of Genesios deal with Byzantine history from 
the accession of Leo the Armenian to the death of Basil I. In 
the fourth book Genesios includes an account of the defeat by 

Petronas, uncle of the emperor Michael III, of the redoubtable cOmar 
cUbaid Allah al AqtaC of Melitene.1 According to the historian, 
cOmar (" Af'EP, C Amr) invaded the Armeniak theme and advanced 
as far as the coast at Amisos where, because he could progress no 
further, he ordered the sea to be beaten with rods (Genesios here com­
pares the behaviour of Xerxes at the Hellespont). 

The emperor, being dismayed at the number of prisoners taken by 
cOmar, appointed Petronas to command the tagma of the Schools with 
orders to attack the enemy. When news of the coming attack reached 
cOmar, his subordinates urged him to retreat to his own territory and 
to fight only if the Byzantine forces overtook them; but the emir, 
declaring that he was no coward, decided to advance towards Petro­
nas and his army. The opposed forces drew close together in the 
Abisian district on the borders of the Paphlagonian and Armeniak 
themes, with a mountain between them, at a place called Porson 
(ll6pcwv). Both sides tried to occupy the mountain, and in the ensuing 
battle COmar was killed. His troops were put to flight, and his son 
with one hundred men retreated across the Halys river, but Mach­
airas the merarch (hypostrategos) in Charsianon intercepted them, so 
that there was no survivor to bring news of the defeat to the Saracens 
of Melitene. 

Except for one detail the narrative of Genesios is internally consis­
tent (it may well come from a lost Vita ofPetronas2). 'Omar advances 
westwards from Amisos and engages Petronas on the borders of 
Paphlagonia and Armeniak-at a place, therefore, somewhere to the 
west of the lower course of the Halys river. The place was called 
Porson. After the defeat at Porson 'Ornar's son retreats eastwards or 

1 Genesius, ed. C. Lachmann (Bonn 1834) 94,1-97,8. 
2 P. Karlin-Hayter, Byzantion 41 (1971) 494. 

443 



444 MICHAEL m AND THE BATTLE OF BISHOP'S MEADOW 

southeastwards across the HaIys into Charsianon. The one inconsistent 
detail is the statement that the two armies met in the Abisian3 terri­
tory 500 miles from Amisos (a7To 'AjLLVCOV wee, jL[)\L(X <p'); no point on 
the boundary between Paphlagonia and Armeniak is so much as 500 
miles by road from Amisos, and it is clear that the numeral is corrupt 
-so p' (100) is an attractive emendation.' 

Like Genesios, the author of Theophanes Continuatus Books 1 to 4 

dedicated his work to the emperor Constantine Porphyrogenitus. 
The Continuator uses a different source and combines it with a hagio­
graphic legend. The legend related that before the battle Petronas, 
when he was at Ephesos as general of the Thrakesian theme, was 

MAP OF ANATOLIA 

assured of victory by John, a monk from Mount Latros (Latmos); not 
long after the defeat of co mar, the legend added, Petronas promptly 
followed his spiritual adviser into the next world.s The hagiographical 
accretion need not detain us here.6 The significant point is that al-

a b T4J 'A{1u;tav4J. According to Gregoire this is the correct text (By{antion 5 [1929/30] 
346), but Lachmann (p.96,6) prints ~JI TCP 'A{1vctaJlcp. A connexion with wE{1tcca is not 
certain. 

, H. Gregoire. Byzantion 8 (1933) 536. 
5 Theoplutnes Continuatus 4.25 (180.13-181,4 and 183,13-184,10) ed. I. Bekker (Bonn 1838). 
1\ In Th£Oplutnes Continuatus the hermit John of Latros meets Petronas for the first time at 

Ephesos shortly before the battle with COmar and advises the general to secure divine 
assistance by having the image of the apostle John painted on his soldiers' shields (pp.180, 
14-181,2 ed. Bekker). According to the Vita ofSt Antony the Younger. however. St Antony, 
whose earlier name had been John. had already been for some time the spiritual father of 
Petronas, when the latter was put in command of the expedition against the Arabs. 



GEORGE HUXLEY 445 

though the Continuator follows a different source, his account of the 
battle is topographically consistent with the narrative of Genesios. 
Petronas found COmar to be encamped in a naturally defensible place 
called Poson (IIocwv). Here a river called Lalakaon flowed from north 
to south, and there was a meadow called in rustic dialect Gyris or 
Gyrin (riJpLV). The East Roman commander determined to surround 
the enemy, and accordingly the generals of Armeniak, Boukellarioi, 
Koloneia and Paphlagonia were stationed to the north; those of 
Anatolikoi, Opsikion and Kappadokia were drawn up on the south, 
together with the commanders of the kleisourai Seleukeia and Char­
sianon, while Petronas himself with the four imperial tagmata and the 
generals of Thrake and Makedonia took up positions to the west. The 
men of the Thrakesian theme were with their general Petronas. 

When COmar saw that he was surrounded, he asked a prisoner the 
names of the locality, the meadow and the river. The prisoner an­
swered IITcfJcwv, Aa'\aKawv and riJpw. cOmar, whom the inventor of 
the conversation supposed to be fluent in Greek, immediately inter­
preted the names to be ill-omened-they signified 'downfall' (7TTWCLC), 
'ruin of his army' ('\aoiJ KaKwcLc) and 'encirclement' (yvpLc8fjvaL). Of 
these names Porson or Poson is common to Genesios and the Continu­
ator, and Lalakaon as the place of the battle is found in the Logo­
thete,7 but neither Lalakaon nor Poson can be certainly identified.8 

Perronas, who was commander of the Thrakesian theme, was ordered by the emperor to 
fight a defensive campaign only, not to confront the enemy in a pitched battle. St Antony 
came to Petronas at Ephesos and assured him that COmar (¥Ap.{3poc) would not see Syria 
again if Petronas marched against him in the name of the Lord and sang the fourth canon 
of St Nicolas (Vita 14-15, ed. F. Halkin, AnalBo11 42 [1944] 187-225, at 218-20). In another 
hagiographical version St George of Mytilene foretells to Petronas the victory to be won by 
him twenty years later-over the Assyrian Ammor (COmar, CAmr): AnalBoll 18 (1899) 252. 
As for St Antony, he had formerly been governor €K 1TPOCW1TOV of the theme Kibyrrhaiotai; 
it is possible therefore that the advice given by him to Petronas at Ephesos, if he ever gave 
it, was military as well as spiritual; for St Antony's governorship see F. Halkin, art. cit. 188. 
In Skylitzes Petronas visits John at Larros (ed. J. Thurn [Berlin/New York 1973] p.100,2). 
As in Theophanes Continuatus, Petronas in the Vita of St Antony dies close in time to his 
spiritual father Antony (17-18, pp.220-21 ed. Halkin). The detailed order-of-battle may 
well have been supplied to the Continuator from the imperial archives by the Porphyro­
genitus. 

? Georgius Monachus (Cont.) 825,3 (with Theoph.Cont., op.cit. [supra n.5]). 
8 Gregoire draws attention to a southward-flowing river Halmyros; this ran east of 

Gangra and so close to the border between Armeniak and Paphlagonia. It is mentioned in 
the Vita of St Hypatios of Gangra (BYZantion 8 [1933] 538-39); but Gregoire's proposal to 
link 'AAP.Vp6c with AaAaKawv through aAvK6c is not helpful. Oddly, Skylitzes exchanges 
the names of the river and meadow (p.100 ed. Thurn). 



446 MICHAEL III AND THE BATTLE OF BISHOP'S MEADOW 

tOmar fought desperately within the trap set by Petronas, but he was 
killed, and his son, who escaped, was caught by the kleisourarches of 
Charsianon. 

Thus the topography of the battle in Theophanes Continuatus is more 
detailed than in Genesios, but it does not contradict Genesios' narra­
tive. The fact of troops from Charsianon having been stationed on the 
south is consistent with the battle having taken place northwards of 
Charsianon, as in Genesios. The Continuator does not state, as Gene­
sios does, that COmar advanced from Amisos to meet Petronas, and 
according to the Logothete, Petronas the commander-in-chief (to­
gether with Nasar of the Boukellarians) trapped COmar at Lalakaon 
on his retreat-Aox~cav'TEC T~V Tfjc tmOcTPocPijC aVTOV cvvaVTWCLV aVTcp 
€lc TOV AaAaKawva;9 but there is nothing in the Logothete to suggest 
that the battle of Lalakaon took place far away from Amisos: Amisos 
is not mentioned by him here. (The immediately preceding mention 
of COmar at Sinope refers to an earlier campaign). A notable variant 
in the Logothete is that a KOp:Y]C overtakes the fleeing emir.1o In 
Genesios and Theophanes Continuatus it is cOmar's son who escapes 
from the trap and is caught later. 

So far, we are entitled to infer that Porphyrogenitan historiog­
raphy, in spite of variations of detail and emphasis, was unanimous 
in placing the battle between Petronas and COmar west of the Halys, 
close to the boundary between the Paphlagonian and Armeniak 
themes. The inference was, however, rejected by J. B. Bury, who in a 
discussion of the Arabo-Byzantine campaigns of A.D. 838, argued that 
the battle of 863 was fought at a place known to Arab writers as Mardj­
al-Uskuf ('Bishop's Meadow').l1 This he proposed to identify with a 
place (Nazianzos) or a district (extending from Nazianzos to the 
Halys, perhaps as far as Nyssa) in Cappadocia. We can agree with Bury 
that Mardj-al-Uskuf lay somewhere in Cappadocia, south of the 
HaIys, since the narrative of the campaign of 838 given by Tabar! 
shows that to have been so. According to Tabar!, Al MuCta~im com­
manded A~inas to enter Byzantine territory by way of the Cilician 

9 825,1-4 Bonn. Nasar is also mentioned with Petronas at the battle of Lalakaon by Leo 
Grammaticus p.238,21 (Bonn 1842). 

10 Georgius Monachus (Cont.) 824,12 Bonn. For the tendency of Arab expeditions into 
Anatolia to make for the coast, see H. Ahrweiler, "L'Asie Mineure et les invasions arabes," 
RHist 227 (1962) 8-10 (repr. in Etudes sur les structures administratives et sociales de Byzance 
[Variorum, London, 1971] IX). 

11 "Mutasim's March through Cappadocia in A.D. 838," ]HS 29 (1909) 120-29. 



GEORGE HUXLEY 447 

Gates; Asinas set out on 19 June; the Calif's own advance guard fol­
lowed, and MuCta~im himself set out on 21 June. When Asinas had 
reached Bishop's Meadow, a letter came to him from MuCta~im, who 
was then at al Maramlr, warning him that the emperor (Theophilos) 
lay in wait ahead of him and was intending to surprise his force at 
the crossing of the 'Lamis'. Asinas was therefore ordered to wait at 
Bishop's Meadow.12 Since the invaders had been advancing north­
wards through Cappadocia, the name of a river 'Lamis' (Lamos), far 
away on the Arabo-Byzantine frontier in western Cilicia, cannot be 
correct here; 'Lamis' is a mistake for 'Halys',13 and it is clear that 
Asinas was at most two days' march ahead of MuCta~im, within easy 
reach of a messenger from al MaramIr. Al MatamIr was a district 
where there were subterranean strongholds, to the north of 
Tyana in southern Cappadocia. The narrative in TabarI shows that 
Bishop's Meadow cannot be far from al Maramir, and it is impossible 
that Tabari thought of Mardj-al-Uskuf as lying far away from Cappa­
docia, on the borders of the Paphlagonian and Armeniak themes. 
Bury therefore accepted the distance 500 miles in Genesios, rejected 
his reference to the border between the Paphlagonian and Armeniak 
themes, and maintained that Petronas fought COmar of Melitene in 
Cappadocia, somewhere to the south of Nyssa. To account for the 
large number of miles from Amisos, 500 according to the unamended 
text of Genesios, Bury supposed COmar to have followed a devious 
route southwards from the Black Sea coast.14 

H. Gregoire, while agreeing with Bury that Bishop's Meadow was 
shown by Tabari and also by the geographer Ibn Khurdadbih to lie in 
Cappadocia, nevertheless insisted that Genesios was correct-the 
battle between Petronas and COmar was fought on the frontier of the 
Armeniak and Paphlagonian themes. Gregoire also believed that 
there was a memory of the battle in the story of an encirclement of 
Arabs by Byzantines in Digenes Akrites ;15 this fight was said in the 

12 Tabarl (French transi. pp.295-96) in Vasiliev, op.cit. (infra n.l3). 
13 Bury, art. cit. (supra n.ll) 122-23. M. Canard, in A. A. Vasiliev, Byzance et Ies Arabes I, 

edd. H. Gregoire and M. Canard (Brussels 1935, repro 1959) 412, thinks that 'Uimis' conceals 
the name of the river Halys, since MuCta~im would not have needed to cross the Halys 
on the way from al Ma~amir to his objective Ankyra; but the original orders given to 
Asinas and later countermanded by the letter sent from MuCta~im at al Matamir to him at 
Bishop's Meadow may well have required him to cross the Halys, not to advance directly 
to Ankyra. 

14 art.cit. (supra n.ll) 128. 
15 E. Trapp, Digenes Akrites (Wien 1971) G675, E495, Z920. 



448 MICHAEL III AND THE BATTLE OF BISHOP'S MEADOW 

epic to have taken place at Malakopia-Melegob (near Nazianzos in 
Cappadocia), and Gregoire argued, most implausibly, that the story 
had been transferred from the borders of Paphlagonia to Cappadocia 
owing to a confusion of a mountain called Mala near Gangra with 
Malakopia in Cappadocia.16 It is time to leave these frail hypotheses 
and to look again at the sources. 

YaCqiibi, who died at the end of the ninth century and so was con­
temporary with the campaigns of 863, describes them briefly. He 
states that MustaCin sent Gacfar al Khayyar in the year A.H. 249 (863/4) 
to make the summer expedition. With Gacfar was cOmar, governor of 
Melitene. When they entered Greek territory, COmar asked Gacfar's 
permission to advance further. COmar was at the head of eight thou­
sand men. The enemy surrounded him, and he and his force were 
defeated.17 

Next comes Tabar!, who died in 923. Concerning A.H. 249 he writes 
that the events of the year included the summer expedition of 
Gacfar ibn Dinar. He occupied a strong point and also al Maramir. 
COmar asked his permission to march against a certain country of the 
Greeks. Permission was granted, and COmar departed, taking with 
him many men of Melitene.18 The reference to al Maramlr, and 
YaCqiibl's statement that Gacfar had COmar with him, together show 
that the two armies had advanced through the Cilician Gates before 
COmar had marched with his own force further north from the 
neighbourhood at Maramir, north of Tyana. Bury reasonably sug­
gested that Gacfar may have been Governor of Tarsos.19 

Tabar! continues: «He [COmar] met the emperor, who was accom­
panied by a large number of Byzantines, at a place called al Arz [or 
<_rz'20] in Mardj-al-Usquf. He engaged him with his troops in a fierce 

16 Gregoire, op.cit. (supra n.4) 534-39; idem, in Vasiliev, op.cit. (supra n.13) 252. See also Gre­
goire, By{antion 5 (1929/30) 339. It is most unlikely that the mention of an encirclement at 
Melegob in Digenes Akrites recalls the encirclement of cOmar. In the epic it is Digenes' 
father Mousour-Ioannes who escapes from encirclement at Malakopia-Melegob, but 
COmar was trapped, and the epic recalls him in Ambron, the maternal grandfather of 
Mousour-Ioannes: see GRBS 15 (1974) 319 and 322-23. 

17 YaCqiibi (French trans!.) in Vasiliev, op.cit. (supra n.13) 277. 
18 Tabari in Vasiliev, op.cit. (supra n.13) 325. In view of the division of forces it is note­

worthy that the Vita of St Antony the Younger (para. 14, ed. Halkin [supra n.6J) refers to 
two expeditionary forces of the godless Ishmaelites (Mo .poccaTa TWV &.8/wv 'IcJLaTJAtTwv) in 
connexion with the story of Petronas and the saint. 

19 A History of the Eastern Roman Empire (London 1912) 283 n.1. 
20 For the text see Bury, art.cit. (supra n.ll) 128. 



GEORGE HUXLEY 449 

battle, and many died on both sides." Thus COmar fought the emper­
or Michael III at Bishop's Meadow in Cappadocia after the emir had 
parted from Gacfar and advanced northwards from al Ma{amir. The 
battle was a hard one, but there is no sign that COmar was killed at 
this stage in the campaign. On the contrary, the emperor was not 
able to check his advance. 

It is habitual in Arab chroniclers to leap abruptly from one stage in 
the sequence of events to the next, as Bury himself recognized in dis­
cussing the invasion of 838.21 Having mentioned the heavy losses of 
Michael III and COmar in the battle of Bishop'S Meadow, Tabari 
jumps to the next engagement: "Then the Byzantines to the number 
of 50,000 encircled him and killed COmar and one thousand amongst 
the Mohammedans. "22 The thousand were survivors of the eight 
thousand who had been under his command, according to YaCqiibi, 
when the emir parted from Gacfar. Thus TabarI describes two battles, 
not one as Bury and Gregoire supposed. The first was between COmar 
and Michael III in Cappadocia; in the second, COmar was surrounded 
with the remnant of his force. From the Greek sources we learn only 
about the second. 

We can now reconstruct the order of events in 863 in Asia Minor. 
Gacfar and cOmar, having joined forces at Tarsos, advanced through 
the Cilician Gates. They campaigned together in al Ma{amIr and then 
parted company. COmar advanced to Mardj-al-Uskuf, where he was 
intercepted by the emperor. Michael III, in a delaying action, des­
troyed much of COmar's force but suffered heavy losses and was not 
able to stop his progress. COmar reached the coast at Amisos and then 
turned westwards, against the advice of his officers, to engage Petro­
nas on the borders of Armeniak and Paphlagonia. The emir was 
surrounded by the imperial forces and killed. His son fled homewards 
across the Halys23 but was overtaken and killed. Thus there is no dis­
agreement about the topography of cOmar's campaign between the 
Greek and Arabic sources. There was a battle in Bishop's Meadow in 
Cappadocia, and there was another battle, at Lalakaon,24 on the 

21 art.cit. (supra n.ll) 121. 
22 Tabari, Zoe.cit. (supra n.18). 
23 Skylitzes (p.IOO,55 ed. J. Thurn) says that he was making for Melitene-a reasonable 

inference. 
:I, An officer having the family name Lalakon was CTparrrt0C of Armeniak in the time of 

the emperor Leo VI (Const.Porph. De Admin.Imp. 45, p.206, 47 M.-J.); the name may have a 
local connexion, as Bury suggests, art. cit. (supra n.ll) 125 n.20. W. M. Ramsay, The Historical 



450 MICHAEL ill AND THE BATTLE OF BISHOP'S MEADOW 

border of Armeniak and Paphlagonia. Since such a large part of the 
combined thematic and tagmatic armies had been placed under the 
command of Petronas, it is not surprising that the emperor had a 
hard fight in the first of the battles. 

The damnatio memoriae of the emperor Michael III perpetuated by 
Genesios and the Continuator on behalf of Constantine Porphyro­
genitus required that the credit for the success of 863 be given to 
Petronas alone, or to Petronas and John of Latros. But the emperor 
Michael's share of glory was not entirely forgotten in Byzantium: in a 
work of the Porphyrogenitus himself, De Ceremoniis, there are pre­
served Akta in honour of the victory over the emir, and in the hymn 
the triumphant emperor is given his due guerdon of praise, together 
with the two Augustae (his mother Theodora and his wife Eudokia) :26 

I "e ~ ~ <'~I 
• . • XaLPOLC ava"" TOV cTpaTov cov 1] avopLa' 

xatpOLc ava~, DL' 00 <> DELva (sc. "AfLEp) KaTE11'Tw()1]' 

xatpoLc av~, <> DE Lva (sc. "AfLEp) Ka()aLplTa' 
<..a I .J.. \ 'e ' ~ .J..I , , " 1 ~ o O'EOC CE 'f'Vl\aSEL EV T'[J 11'Op'f'vpCf. EtC TLfL1]v KaL aVEYEpcLV TWV 

'p I , ~ I , 1 , ~ .J..I 
wfLaLWv CVV TaLC TLfLLaLC aVYOVCTaLC EV T'[J 11'0p'f'VpCf.· 

, I <;01 ~\ ~<~ 

ELcaKOVCEL 0 ~EOC TOV I\aov VfLWV. 

THE QUEEN'S UNIVERSITY OF BELFAST 

June, 1975 

Geography of Asia Minor (London 1890) 77 and map opposite p.I96, proposed to identify the 
river Lalakaon with a southward-flowing tributary of the Halys between Gangra and 
Andrapa. This would suit quite well the distance p' miles from Amisos in the emended 
text of Genesios (p.96,6 Lachmann). 

25 De eer. 1, pp. 332-33 (Bonn 1829). See also Bury, op.cit. (supra n.19) 284, and eundem, 
EHR 22 (1907) 434, on the W AKTa £171. p.£ytCT/zvcp ap.~tpfi. £.. 17oMp.cp TJTT7J8mt Ka1. dvatp~8mt. 

Notably absent from Greek and Arab accounts of the fighting in 863 is the ally of cOmar, 
the Paulician leader Karbeas of Tephrike, but he may already have been on his death­
bed at the time: cf P. Lemerle, Travaux et Memoires 5 (1973) 95-96. Neither Genesios nor 
Tlreophanes Continuatus mentions that the victory of Petronas in 863 was followed by a 
campaign to the frontiers of Mesopotamia; this was opposed by the governor of Armenia 
and Azerbaijan, cAli-ibn-Ya~ya, who fell in battle against the imperial forces in the 
month of rama~an of the same year (18 October-16 November)-see Vasiliev, op.cit. 
(supra n.13) 256-as we learn from Tabari (Vasiliev, op.cit. 325).