Slike stranica

Christendom,' and a Greek contemporary writer ranks it next to the capital.2


Mutasim left his palace at Samarra in April (A.D. 838), and the banners of his immense army 3 were inscribed with the name of Amorion. The Caliph was a warrior of indisputable bravery, but we know not whether it was he or his generals who designed the strategical plan of the invasion. The two most eminent generals who served in this campaign were Ashnas and Afshin. The former was a Turk, and his prominence is significant of the confidence which Mutasim reposed in his new corps of Turkish guards. Afshin had distinguished himself by suppressing rebellion in Egypt, and he had done much to terminate the war against Babek which had been so long drawn out.

The city of Ancyra was fixed upon as the first objective of the invasion. An army of the east, under the command of Afshin, advanced by way of Germanicia, and crossed the frontier by the Pass of Hadath on a day which was so fixed as to allow him time to meet the army of the west in the plains of Ancyra.

The purposes of the Caliph were not kept secret. The dispositions of the Emperor show that he was aware of the designs on Ancyra and Amorion. He left Constantinople probably in May; and from Dorylaion, the first great military station on the road to the Saracen frontier, he made provisions for the strengthening of the walls and the garrison of Amorion. The duty of defending the city naturally devolved upon Aetius, the stratêgos of the Anatolic Theme, for Amorion was his official residence. The plan of the Emperor was to attack the forces of the enemy on their northward march to Ancyra. Knowing nothing of the eastern army under Afshin, he crossed the Halys and encamped with his army not far from the river's bank in the extreme south of the Charsian district,

1 "And more valued by the Greeks than Constantinople" (Tabari, 30); cp. Masudi, 74.

2 Acta citt. 425 (cp. 1113).

According to Michael Syr. 95, Mutasim's army numbered 50,000, Afshin's 30,000. He mentions also 30,000 merchants and providers, 50,000 camels, 20,000 mules. BarHebraeus (159) says that Mutasim led 220,000 men. The Armenian version of Michael (274) mentions 30,000

negroes. Masudi (68) says that the numbers were exaggerated by some to 500,000 and reduced by others to 200,000. Tabari (30) says that no Caliph had ever made preparations for war on such a gigantic scale. These statements illustrate the value of numbers in medieval writers. We can only trust intelligent contemporaries. Here the numbers of the combatants given by Michael, i.e. Dionysios, are moderate and credible.

probably near Zoropassos, where there was a bridge. He calculated that the enemy would march from the Cilician Gates to Ancyra by the most direct road, which from Soandos to Parnassos followed the course of the river, and he hoped to attack them on the flank.1 The Caliph's western army advanced northward from Tyana in two divisions, and Ashnas, who was in front, was already near the Halys before the Emperor's proximity was suspected. The Caliph ordered a halt till the position and movements of the Romans should be discovered. But in the meantime Theophilus had been informed of the advance of the eastern army, and the news disconcerted his plans. He was now obliged to divide his forces. Taking, probably, the greater portion with him, he marched himself to oppose Afshin, and left the rest, under the command of a kinsman, to check or harass the progress of the Caliph. Afshin had already passed Sebastea (Sivas), and was in the district of Dazimon, when he was forced to give battle to the Emperor. Dazimon, the modern Tokat, commands the great eastern road from Constantinople to Sebastea, at the point where another road runs northward to Neo-Caesarea. The town lies at the foot of a hill, at one extremity of which the ruins of the ancient fortress are still to be seen.1 Situated near the southern bank of the Iris, it marks the eastern end of a fertile plain stretching to Gaziura (now Turkhal), which in the ancient and middle ages was known as Dazimonitis; the Turks call it Kaz-Ova. It was probably in this plain that the Saracens encamped. The Emperor, who may have



1 For details of the march of Mutasim and Ashnas, see Bury, Mutasim's March. Tabari's account of the campaign is fuller than any other.

2 30,000 (Michael Syr. 95, who gives no topographical indications). Afshin is evidently meant by Simeon's curious Sudeê (Sundei, vers. Slav. 97; Zovden, Add. Georg. ed. Mur. 712; Zovdéu, Leo Gr. 224).

3 Gen. 67 of (the Saracen commanders) κατὰ τὸν Δαξιμῶνα συνήχθησαν στρατοπεδευσάμενοι. Tabari's date (45) for the battle, July 22, can hardly be right. A longer time must surely have elapsed before the beginning of the siege of Amorion (Aug. 1). Moreover, Tabari refutes himself. His date

[ocr errors]

is 'Thursday, Shaban 25." But Shaban 25 July 22 fell on Monday.

4 For the plain of Dazimon, which seems to have been once part of an Imperial estate, see Anderson, Stud. Pont. i. 68; for Tokat itself and the fortress, Cumont, ib. ii. 240-243.

5 Afshin had been reinforced by the forces of Armenia led by Bagarat, lord (ishkhan) of Vaspurakan, the "prince of princes." This title was rendered in Greek by ἄρχων τῶν ἀρχόν TWV (Constantine, Cer. 687). Genesios has split him into two persons (67) αὐτοῦ τοῦ ἄρχ. ἀρχ. καὶ τοῦ Βεσπαρακανίτου (I am not quite sure whether Marquart follows him, op. cit. 463). Cont. Th. 127 rightly mentions only one person. Bagarat was a son of

arrived on the scene by way of Zela and Gaziura, halted near Anzên, a high hill, from whose summit the position of the enemy could be seen. This hill has not been identified; we may perhaps guess, provisionally, that it will be discovered to the south of the plain of Dazimonitis.1 The fortune of the ensuing battle at first went well for the Greeks, who defeated the enemy, on one wing at least, with great loss; but a heavy shower of rain descended, and the sudden disappearance of the Emperor, who at the head of 2000 men had ridden round to reinforce the other wing of his army, gave rise, in the overhanging gloom, to the rumour that he was slain. The Romans, in consternation, turned and fled, and, when the sun emerged from the darkness, the Emperor with his band was surrounded by the troops of Afshin. They held the enemy at bay, until the Saracen general brought up siege-catapults to bombard them with stones; then they fought their way, desperately but successfully, through the hostile ring.2

The Emperor, with his handful of followers, fled northwestward to Chiliokomon, "the plain of a thousand villages" (now Sulu-Ova), and then, returning to his camp on the Halys, found to his dismay that his kinsman had allowed, or been unable to forbid, many of the troops to disperse to their

Ashot (ob. 826), on whom the Caliph had conferred the government of Iberia. Leo V. bestowed on him the title curopalates (frequently conferred on the Iberian princes), and in A.D. 820 he besought Leo's help against a rebel. (Cp. Marquart, ib. 404.) Bagarat was also lord of Taron (the district west of Lake Van and north of Arzanene, from which it is separated by the Antitaurus. Vaspurakan is east and north-east of Lake Van).

1 Anzên recurs in a later battle in the same region; see below, p. 282, for the topographical data.

2 I have followed the account of Michael Syr. 95. Genesios (68) agrees as to the first success of the Romans, but attributes their flight to the archery of the Turks. He describes the surrounding of Theophilus, with whom were Manuel, the Persians, and the commanders of the Tagmatic troops. He also mentions the rain and explains that the Turkish archers could not shoot at Theophilus and his

companions because their bow-strings were wet; this, in turn, explains the employment of stone-hurling machines mentioned by Michael. According to Tabari (135), who professes to give the evidence of a Christian captive present at the battle, the fortune of the day was retrieved by the Saracen cavalry. It may be suspected that the discomfiture of the Romans, whether by archers or cavalry or both, occurred on that wing which the Emperor with his 2000 rode round to reinforce. Gen. 68-69 (Cont. Th. 128) relates that Theophilus was rescued by Manuel from the contemplated treachery of his Persian regiments. The story is highly suspicious (cp. Hirsch, 145), as it was also told, with little variation, of a battle in A.D. 830 (above, p. 257). But the life of Theophilus was certainly in danger, as we know from Michael. According to Masudi (68), having lost many of his officers, he owed his life to the protection of Nasr.

3 See Cumont, op. cit. 144.


various stations. Having punished the commander for his weakness, and sent orders that the soldiers who had left the camp should be beaten with stripes, he dispatched a eunuch1 to Ancyra, to provide, if there were still time, for the defence of that city. But it was too late; for the western army of the invaders was already there.2 Ancyra ought to have offered resistance to a foe. Its fortifications were probably strengthened by Nicephorus I.3 But the inhabitants, thoroughly alarmed by the tidings of the victory of Afshin, deserted the city and fled into the mountains, where they were sought out by Ashnas and easily defeated. Thus the town fell without a blow into the hands of the destroyer. The Emperor, at this crisis, did not disdain to humble himself before the Caliph. He sent an embassy, imploring peace, and offering to rebuild the fortress of Zapetra, to release all the captives who were in his hands, and to surrender those men who had committed cruel outrages in the Zapetra campaign. The overtures were rejected, with contempt and taunts, by the Caliph,5 and Theophilus betook himself to Dorylaion to await the fate of [I read πένθει, Boeckh πενθεῖ. He reads exopov Taîs in line 2, but the traces do not point to this.] Now, as no destruction of Ancyra is recorded between A.D. 805 (the restoration of Nicephorus) and A.D. 829, Michael II. cannot be meant. The storm must refer to the event of 838, and the restoration must belong to the reign of Michael III. Moreover, in the case of Michael II. (except in the first five months of his reign), Theophilus I would have been associated with him in such an inscription. The fact that Michael III. is named alone, without Theodora, points to a date after A.D. 856, and this is confirmed by ráλai. The other inscription (ten iambic trimeters), though it does not mention the disaster, is evidently of the same date, and, as Boeckh thinks, probably by the same (local) "poet."

1 Doubtless Theodoros Krateros, one of the Amorian martyrs, who, as Nikitin conjectures, may have been stratêgos of the Bukellarian Theme (Acta 42 Mart. Amor. 205).

2 It had marched northward by the route west of the Halys (see above, p. 264). Michael Syr. 95 records that Mutasim found Nyssa, which lay on his road, deserted, and destroyed its walls.

3 Theoph. 481. In 806 Harun marched within sight of the city (ib. 482). It is generally said that the walls were restored by Michael II. (so Vasil'ev, 124). But the inscriptions on which this statement is based (C.I.G. iv. 8794, 8795, pp. 365-366) have, I think, been wrongly interpreted. The second (consisting of fifteen iambic trimeters) tells how Michael

Μιχαὴλ ὁ δεσπότης

μέγας βασιλεὺς ν[ικητ]ὴς στεφηφόρος has raised Ancyra from her ruins. The document begins:

πένθει φθαρεῖσα καὶ «κλι〉θεῖσα πρ[ὸς πέδω

χ]ερσὶν ὑπ ἐ[ ] μιαιφόνοις, [ἐκ] πάλαι,

νῦν [ἀνεγ]είρου τῶν κακῶν ἀνειμένη.


A poet, Husain, sang in honour of Mutasim: "Of Ancyra thou didst spare nought, and thou didst demolish the great Amorion." Ibn Khurdadhbah, 101, 74; Vasil'ev, 129, n. 2.

Yakubi, 9; Gen. 64.

6 Michael Syr. 95 relates that a report was spread in Constantinople that the Emperor was slain in the battle with Afshin, that a plot was

Amorion, for the safety of which he believed that he had done all that could be done.


The army of the Saracens advanced westwards from Ancyra in three columns, Ashnas in front, the Caliph in the centre, and Afshin behind, at distances of two parasangs. Ravaging and burning as they went, they reached Amorion in seven days. The siege began on the first of August.1 The city was strong; its high wall was fortified by forty-four bastions and surrounded by a wide moat;2 its defence had been entrusted by Theophilus to Aetius, stratêgos of the Anatolic Theme; and reinforcements had been added to its garrison, under Constantine Babutzikos, who had married a sister of the Empress Theodora and was Drungary of the Watch, and the eunuch Theodore Krateros 3 and others. But there was a weak spot in the fortification. Some time

formed to elect a new Emperor, and that Theophilus, informed of the matter by a message from his mother (? stepmother), hastened thither from Amorion and punished the conspirators. Genesios (69) mentions his being at Nicaea, and Vasil'ev suggests that this may confirm the Syriac record.

1 Tabari, 45; Acta 42 Mart. 42 (εἰσιόντος τοῦ Αὐγούστου μηνός). The city was taken on Tuesday in Ramadhan, i.e. August 13, according to Yakubi, 10. This accords with Michael Syr. 100, who says that the city was taken in 12 days, and can be reconIciled with the statement of Euodios (Acta citt. 65) that the siege lasted 13 days. For Ashnas arrived at Amorion on Thursday, August 1, the Caliph was there on Friday, August 2, and Afshin came on Saturday (Tabari, 37). Thus the duration might be described as either of 12 or of 13 days (or of 11, since active operations did not begin till August 3). Nikitin (ad Acta citt. 243), who wrongly equates the Thursday with July 31. Tabari's equation (45) of Friday with the 6th of Ramadhan is false; Thursday= Ramadhan 7 (see Mas Latrie, Trésor, p. 566). The same scholar rightly points out that a wrong deduction has been drawn by Weil and Vasil'ev from Tabari's statement (45) that Mutasim returned 55 days after the beginning of the siege. They


took this to mean that the siege lasted 55 days, and so placed the capture on September 23 or 24. But Tabari obviously means his return to Tarsus, and the 55 days include his march from Amorion, which was slow and interrupted. According to George Mon. 797, the siege lasted 15 days in August; this is nearly right.


2 Ibn Khurdadhbah.

[ocr errors]

3 The names in Simeon (Add. Georg. 805; vers. Slav. 98) and Cont. Th. 126 must be controlled by the Acta of the 42 Martyrs. The identity of the officers has been examined by Nikitin (Acta, 202-219), who has proved, in my opinion, that Constantine the Patrician is Constantine Babutzikos. In one document he is described as ἄρχων τῶν ταγμάτων (Synaxar. ecc. Const. 516), whence Nikitin infers that he was commander of one of the 'guard regiments. But Simeon's δρουγγάριος shows at once that he commanded the Arithmos (Vigla), the only one of the four Tagmata whose commander was so named. The other officers were Theophilus, a stratêgos, and Bassoes, & Spoμeús the runner. Nikitin (208 sqq.) has shown that this does not mean a courier here, but a victor in the foot-race (πεŠodрóμιov). Constantine, Cer. 358, mentions Bambaludes, ὁ τῶν Πρασίνων Spoμeús, champion of the Greeks, in the reign of Michael III.

« PrethodnaNastavi »