^{page 208 note 1} George of Pisidia, Hexaemeron. The idea recurs in Theophanes (Bury, II, 244, n. 3).

^{page 209 note 1} According to the received account of the war, it was in Cilicia that Heraclius had landed and made ready his forces in his first campaign of 622 A.D. ‘No place,’ says Professor Bury (L.R.E. II. 228), ‘could offer a more secure retreat for organizing and drilling his army at leisure, and for assimilating the new recruits to the troops which he had brought with him.’ If Gerland's view (accepted by Professor Bury in his edition of Gibbon) is right, the story of the landing in Cilicia in 622 A.D. must be given up; but the suitability of that province for the preparation and equipment of an army remains unaltered.

^{page 212 note 1} έχιδνα would equally suit the metre.

^{page 212 note 2} Read, perhaps, σο⋯ προσιόν, ‘Pώμη’. The ‘day’ which the speaker longs to see is the day of destruction of Rome.

^{page 212 note 3} This use of ὼς ⋯πόταν is an extension of usage common in similes. Cf. Hom. Il. 10.5, ὼς δ’ ἂν άστράπτη πόσις ‘Hρης, … ὥς … ⋯νεστενάχιζ’ ’Aγαμέμνων. In that sentence, the verb dependent on ὥς is omitted; here, the verb of the ⋯πόταν clause is also omitted, so that ὡς⋯π⋯ταν becomes equivalent to ὡς alone. If written out in full, the sentence would run thus: ‘The peoples will taste as (men taste) when (they taste) at a pleasant feast.’ Cf. Hom. Il. 4.462, ἤριπε δ’ ὡς ὅτε πύργος έν⋯ κρατερ⋯ ὑσμινη. Od. 5. 281, εἴσατο δ’ ὡς ὅτε ῥινόν ⋯ν ἠεροειδέϊ πόντψ. The use of ὡς ε⋯ (like quasi) as an equivalent for ὡς is analogous; e.g. Il. 16. 58: τήν μ’ ἄψ ⋯κ χειρ⋯ν ἕλετο … ὡς εἴ τιν’ ⋯τίμητον μετανάστην.

^{page 213 note 1} Butler, p. 319.

^{page 214 note 1} Batler, p. 285.

^{page 215 note 1} Cf. ‘Amr's description of himself (Hishâm ibn al Kalbî, in Butler, p. 203): ‘I am the deaf adder, from whose bite none may recover, whose sting renders a man sleepless.’

^{page 215 note 2} Karabacek (Papyrus Erzherzog Rainer, Fuhrer durch die Ausst., p. 134) thus describes the equipment of the Arab horsemen who invaded Egypt: ‘Die schweren Panzerreiter waren mit einem den ganzen Korper bedeckenden, von den Schultern bis zu den Knöcheln herabreichenden Panzerhemd aus Ringelgeflecht bekleidet; über das Haupt fiel auf die Schultern herab wiederum ein eisernes Ringelgeflecht, das sogenannte Mighsar, gleichzeitig als Visir und Mücken-schutz dienend; darüber ward der konische Eisenhelm aufgesetzt.’ (How a man clothed in a shirt of chain-mail reaching to his ankles could sit on horseback, I am unable to explain; but the general description holds good independently of this detail.) An Arab warrior, then, was literally ‘wrapped’ in chain-mail; and the interwoven iron rings might very well be likened to a serpent's scales. According to Oman's Art of War in the Middle Ages, pp. 184 ff. (Butler, p. 131), ‘The ordinary equipment of the Roman cavalry soldier at this time was a steel cap, a coat of mail, gauntlets, and steel shoes’; but the mail covering the head and shoulders, which is mentioned in the case of the Arabs alone, would constitute a marked difference in appearance, and it may be this that is specially referred to in the words of the text.

^{page 217 note 1} John of Nikiou, in Butler, p. 313. (My hearty thanks are due to Dr. Charles for his kindness in lending me the unpublished MS. of his translation of John of Nikiou.)

^{page 217 note 2} Kριòς is metrically interchangeable with K⋯ρος; and both words begin with the same letter.

^{page 217 note 3} Before his appointment as Patriarch and Governor of Egypt in 631 A.D., he was bishop of Phasis in Colchis. Is it possible that Kυρήνης here means ‘from Cyrus-land,’ i.e. from the land of the river Cyrus (of Iberia), the upper course of which lies between the Colchian river Phasis to the north and an Armenian river Phasis to the south?

^{page 217 note 4} John of Nikiou (Butler, p. 303): ‘Constantine assembled great number of vessels, and sent them under Kirius and Salâkriûs [sic] to bring the Patriarch Cyrus to him.’ The names of the two commanders of this fleet, as they appear disguised in the Ethiopic translation of an Arabic translation of a Coptic or Greelf original, have a curious resemblance to the Kριòς of 1. 328; but this can hardly be anything more than a coincidence.

^{page 218 note 1} Butler, p. 307.

^{page 218 note 2} Ibid. p. 330.

^{page 218 note 3} Ibid. p. 365.

^{page 219 note 1} Supposing, as is probable, that the lost passage spoke of ‘Amr in terras of praise or approval, that might of itself be enough to account for its excision; for a Christian reviser would find cause for offence in a passage which spoke well of a victorious infidel; and it is easier to cancel objectionable verses than to alter them or compose others in their place. Doubtless others also among the many lacunae in the Jewish Sidyllina are due to a similar cause. As we know that Christians sometimes interpolated passages to express their own views, we may infer that they also-and probably even more frequently - cut out passages of which they disapproved.

^{page 219 note 2} Cyrus died four months after the signing of the armistice; and in the chronicle of John of Nikiou (Butler, p. 361) his death is attributed to ‘the remorse of having delivered Alexandria into the hands of the Muslims.

^{page 220 note 1} Butler, p. 468: ‘There were only a thousand Arab soldiers to defend the city; these were quickly overpowered and slain, very few making their escape: and Alexandria came once more under the dominion of the Caesars.’

^{page 220 note 2} The question suggests itself, whether the variant is in some way connected with ‘the Kαυχιος’ or ‘Mukaukis.’ No convincing explanation of that title or nickname of Cyrus has been put forward; it was probably derived from some unknown word in use among the Egyptian natives. A reference to Cyrus himself is imposquickly sible, as he died before the time here spoken of; but if the word is a title, there may at this time have been another Kαυχιος More probably, however, icavxiv is merely a corruption of ναν(μα)χιη.

^{page 220 note 3} Or the reading αὐτοῖς may have come from αὖτις (MSS. αι’τοῖς) in the preceding line.

^{page 221 note 1} Butler, p. 470. The Copts of the Delta villages were afterwards compensated by ‘Amr for the losses they had suffered through these raids; ibid. p. 488.

^{page 222 note 1} Butler, p. 475.

^{page 222 note 2} πόλει or πύλαις would require a verb in the sense of ‘incursum facient’ and ⋯γκύρειν could hardly bear that sense—unless, indeed, on the assumption that its meaning was modified by association with the Latin incurren, which, as a military term, may have survived in use among the Byzantines. The word πρόκονρσον, in the sense of prima acies, occurs in Chron. Pasch. ad A nn. 626.

^{page 222 note 3} Butler, p. 473.

^{page 222 note 4} Does the other reading, ⋯γκύρσουσι, point to a variant ⋯γκύψουσι, ‘they shall raise their heads again’?

^{page 223 note 1} Geffcken, though not aware of the true application of the prophecy, corrects ’Iουδα⋯ους into ‘Iουδαῖοι’, rightly remarking that ‘die Juden sollen ja zuletzt siegen.’

^{page 223 note 2} The other inhabitants of Alexandria were too much split up into hostile factions to be spoken of as acting as a united whole. Theie must have been partizans of both sides in the city.

^{page 224 note 1} The word τροπαιο⋯χος was used officially as an epithet of Roman emperors; e.g. of Justinian, in Chron. Patch.

^{page 224 note 2} E.g. Isaiah xiv. 3 ff., of the King of Babylon.

^{page 224 note 3} Bury II. 19.

^{page 224 note 4} Geffcken ad loc.

^{page 225 note 1} Alexandra, then, was after all not far astray when he said that ‘Arabs here means Jews’; though the arguments by which he supports that statement cannot be accepted. The truth is rather that the word Arabs has been wrongly inserted in a line which originally referred not to Arabs, but to Jews.

^{page 226 note 1} A treaty of peace was in fact made between Constans II. and Muawiah in 659 A.D.; another was made by Constantine IV. in 678 A.O.

^{page 227 note 1} With the exception of the few years during which it was held by the Persians under Chosroes.

^{page 228 note 1} VIII. 324 forms part of a Christian oracle, which is not likely to have been in the hands of our Jewish Sibyllist; but both may have borrowed the phrase from a common Jewish source.