[This paper was originally given to the Society of Antiquaries of London on December 8, 1959. It owes much to the advice and encouragement of Professor Albrecht Goetze, of Yale University.]

¹ For an interesting discussion of the motives for, and circumstances attending, the publication of the Anabasis, see Dürrbach, F., ‘L'Apologie de Xenophon dans l'Anabase’, REG 1893 343–86.CrossRefGoogle Scholar

² L'Euphrate et le Tigre (Paris, 1779).

³ Weidner, E. F., Archiv für Orientforschung xvi (1952) 19–20 Google Scholar; Thureau-Dangin, , ‘Numération et métrologie sumérienne’, Revue Assyriologique xviii.Google Scholar

⁴ Nineveh and Babylon (1853) 59–60.

⁵ Journal of the Royal Geographical Society ix 9.

⁶ One of the most detailed and interesting of these is the Assyrian itinerary of King Tukulti-Ninurta II (824 B.C.) who struck down the Wadi Tharthar into Babylonia, then making west to the region of the Tigris below Samarra, then via Dûrkurigalzu marched south to Sippar, then home via the Euphrates. ( Luckenbill, , Annals of Assyria i § 407–8Google Scholar; Musil, A., The Middle Euphrates (New York, 1927) 199–204.)Google Scholar

⁸ i 4.11. The location of Thapsacus (meaning ‘a ford’, from the Semitic root psḫ ‘to pass’) is disputed, but seems to have been at Samûma (Meskenê) at the great bend of the Euphrates. See Pauly-Wiss s.v. Θάψακος. Others place it at Raqqa (Nikephorion). The arguments of Farrell, W. J. (JHS lxxxi (1961) 153–5)CrossRefGoogle Scholar for placing it at Carchemish are unconvincing. The distance of Thapsacus from Babylon is given by Eratosthenes (Strabo ii 1.22, 29) as 4800 stadia or 600 miles.

⁹ Musil, op. cit., 221, followed by du Buisson, Du Mesnil, Baghouz, l'ancienne Corsôté (Leiden, 1948).Google Scholar The Dawrin canal is the Saocoras river of Ptolemy: Musil, 340.

¹⁰ i 5.4.

¹¹ Du Mesnil du Buisson, op. cit.

¹² See below, p. 17.

¹³ ‘Researches in the Vicinity of the Median Wall’, Selections from the Records of the Bombay Government xliii (1850) 263.

¹⁴ Herzfeld, , Ausgrabungen von Samarra vi (1948).Google Scholar This posthumous work, an excellent survey, unfortunately lacks maps, the plates having been destroyed in the War.

¹⁵ Musil, Alois in The Middle Euphrates, Appendix II ‘Xenophon on the Middle Euphrates’, 213–14 esp. 223 (New York, 1927).Google Scholar

¹⁶ Obermayer, , Die Landschaft Babyloniens (1929).Google Scholar

¹⁷ Marcellinus, Ammianus, Rerum Gestarum libri qui supersunt xxiv.Google Scholar For a detailed study, see Dilleman, L., ‘Ammien Marcellin et les pays de l'Euphrate et du Tigre’, Syria xxviii (1961).Google Scholar

¹⁸ i 5.5.

¹⁹ i 6.1.

²⁰ i 7.1.

²¹ i 7.14 ff.

²² i 8.1 ff.

²³ Artaxerxes.

²⁴ GSGS. 3919. 3rd ed. 1949. Quarter inch. 1–38 N.

²⁵ Op. cit., 51, and 142, 148 and 154. See Goetze, loc. cit., 64 n. 94.

²⁶ Ross, J., ‘Notes on two Journeys from Baghdad to the Ruins of al Hadhr … in 1836 and 1837’, JRGS ix 445, 1Google Scholar; also ‘Journey from Bagdad to the Ruins of Opis and the Median Wall in 1834’, JRGS xi (1841).

²⁷ Lynch, , ‘Note on a part of the river Tigris between Baghdad and Samarrah’, JRGS xi (1841) 472–3.Google Scholar

²⁸ Op. cit., 263 (see note 13).

²⁹ Op. cit., 8: Lane, , Babylonian Problems, described in 1921 Google Scholar the wall near the railway as being of unbaked brick forming two casemates 5 feet broad, with regular bastions at 60 yards' interval.

³⁰ Lane, , Babylonian Problems 42 Google Scholar, points out this feature and illustrates it in a photograph (his Plate 5).

³¹ Poidebard, , La Trace de Rome dans le désert de Syrie, pls. 38–9.Google Scholar

³² The Middle Euphrates 154. It seems that this wall or fortification was observed but never recorded, during or just after World War I; see Major Mason, loc. cit. (note 72).

³³ von Haller, , Uruk-Warka 7.Google Scholar Bericht 41–45, pl. 35.

³⁴ Andrae, , Das Wiedererstandene Assur pl. 72.Google Scholar

³⁵ Von Luschan, , Ausgrabungen von Sendschirli ii pls. 15, 16, 30.Google Scholar

³⁶ Dr Gadd points out to me that it is now claimed that battering rams were known in the Old Babylonian period, being called wašibum in the Mari letters. Kupper, , Revue Assyriologique xlii 139–45, 125.Google Scholar But the great period of their use was by the Assyrians in the early Iron Age.

³⁷ Ammianus Marcellinus xxiv 2.

³⁸ See Musil, op. cit., Appendix VI, ‘The Canals of the Middle Euphrates’. On these canals, see Streck, , Die Alte Landschaft Babyloniens (1900–1901)Google Scholar; Alois Musil, op. cit.; Lestrange, , The Lands of the Eastern Caliphate (1905)Google Scholar; and Obermayer, , Die Landschaft Babyloniens (1929).Google Scholar

³⁹ See on this, Laessoe, , ‘Reflexions on Modern and Ancient Oriental Waterworks’, Journal of Cuneiform Studies vii (1953).Google Scholar

⁴⁰ ‘Surveys of Ancient Babylon’ by Commander W. Beaumont Selby, Lieutenant W. Collingwood and Lieutenant J. B. Bewsher (London, 1885). See note 77.

⁴¹ See ‘Mesopotamian Mound Survey’, Archaeology vii (1954); Goetze, , ‘Archaeological Survey of Ancient Canals’, Sumer xi (1955) 127–8.Google Scholar

⁴² Adams, Robert M., ‘Settlements in Ancient Akkad’, Archaeology x (1957) 270–3Google Scholar; ‘Survey of Ancient Watercourses and Settlements in Central Iraq’, Sumer xiv (1958) 101–4.

⁴³ Tablets dated under Achaemenid kings have been found, dated and inscribed from the following cities: Borsippa, Babylon, Sahrinu, Sippar, Nippur, Dilbat, Hubadišu, Kutha, Uruk, Ur (information from Mr D. J. Wiseman).

⁴⁴ Dûr-kurigalzu, founded by the Kassite king, Kurigalzu I (c. 1400 B.C.) appears to occupy the site of an older Sumerian city named Esâ. Poebel, , ‘The City of Esa’, Miscellaneous Studies (Chicago, 1947).Google Scholar (I owe this reference to Dr E. Sollberger.)

⁴⁵ Poebel, op. cit., plausibly suggests that this name, used in Islamic times, and popularly said to refer to the 'Isa, the uncle of Mansur, is, in fact, an adapted recollection of Esâ.

⁴⁶ See Jacobsen, T., ‘The Waters of Ur’, Iraq xx (1960) 175.Google Scholar

⁴⁷ Unger, E., Babylon, die heilige Stadt (1931).Google Scholar

⁴⁸ Strabo ii 1.23. In xvi 1.22 he says that Mesopotamia contracts in shape, projecting to a considerable length; the shape of it somewhat resembles a boat, and the greatest part of its periphery is formed by the Euphrates.

⁴⁹ Meissner, B., ‘Pallacottas’, Mitteilungen der Vorderasiatisch-ägyplischen Gesellschaft i (1896) 177–89.Google Scholar

⁵⁰ vii 22.

⁵¹ xvi 1.9–11.

⁵² ‘Pallughtha—nicht eigentlich ein Kanalname, sondern ein Ausdruck für die Regulierung des Euphrat selbst’, Herzfeld, op. cit., 13.

⁵³ Meissner, loc. cit., 186.

⁵⁴ NH ii 5.30.

⁵⁵ i 189–91.

⁵⁶ vii 5.

⁵⁷ ch. v.

⁵⁸ Wiseman, , Chronicles of Chaldaean Kings (1956).Google Scholar

⁵⁹ According to Pallis, , The Antiquity of Iraq 10 Google Scholar, Herzfeld, op. cit., it is first mentioned in the time of Nazimaruttaš. Pinches, , JRAS 1917 137 Google Scholar, ‘An Early Mention of the Nahr Malka’, publishes a tablet from Jokha (Umma) of the time of Bur-Sin of Ur, mentioning Sura and the Canal of the King.

⁶⁰ Parthian Stations 1.

⁶¹ Ammianus Marcellinus vi 1.

⁶² NH v 21.90.

⁶³ For the identification, see Honigmann, and Maricq, , Res Gestae Divi Saporis 110 ff.Google Scholar, from the newly discovered trilingual of Naksh-i-Rustem. Henning, W. B., ‘ Βεσήχανα πόλις: ad BSOAS xiv 512 N. 6’, Bulletin of School of Oriental Studies 1953 xv/2.Google Scholar

⁶⁴ The identification of the site of Agranis is highly obscure, but it should be near Al-anbar. Dilleman (loc. cit.) has complicated matters by twisting a passage in Pliny ‘sunt etiamnum in Mesopotamia oppida: Hippareni, Chaldaeorum doctrina et hoc sicut Babylon—juxta fluvium qui cadit in Narragam, unde civitati nomen’ (NH vi 30.123) to bring Agranis into connection with Sippar. In about 1090 B.C. Tiglathpileser I marched against Babylon and captured the cities of Dur-kurigalzu (= Aqar Qûf), Sippar-of-Shamash (= Abu Habbah), Sippar-of-Anunitum and Babylon, and returned via Opis. The site of ‘Sippar-of-Shamash’ (Abu Habbah) is well known, but that of the other Sippar, that of Anunitum, is not. It clearly lay between Abu Habbah and Babylon, but is not mentioned again. It is, however, known that Sippar-of-Anunitum was a next-door neighbour of the city of Akkad (the site of which is likewise unknown), being separate from it only by a canal called the nâr Agâde or ‘river of Akkad’ (Ebeling, Reallexikon der Assyriologie s.v. ‘Akkad’). But the reference in the Bible, 2 Kings xix. 34, to the name of Sepharvaim, a city which the Assyrians claim to have destroyed, being a dual form, has suggested to some that both Sippars existed into the seventh century B.C. In fact, both Akkad and Sippar-of-Anunitum survived certainly into the sixth century B.C., for Nabonidus rededicated an identically-named temple in each (Ebeling, loc. cit.). Sippar-of-Shamash lasted certainly into Achaemenid times. While, therefore, it would seem that Sippar-of-Anunitum formed a twin city with Akkad, it is not clear that Sippar-of-Anunitum had any intimate geographical connection with Sippar-of-Shamash. Dilleman, however, claims that Sippar(-of-Shamash) formed a double city (‘ville jumelle’) with a non-existent city, Agané (which is evidently a misreading of the name Agadẽ = Akkad), and sees in a hypothetical *nar-Agane the interpretation of both Agranis and ‘Narraga’ of Pliny. But Andrae and Jordan examined the terrain around Sippar (-of-Shamash) in 1927 in detail and could find no trace of a second twin city (‘Abu-Habbah-Sippar’, Iraq i (1934)). It is, however, perfectly possible that Pliny, in mentioning in his almost certainly garbled passage the river Narraga, near Sippar, was referring to the Nar-Agadẽ. The passage should probably be amended: ‘Hippareni, Chaldaeorum doctrina et hoc sicut Babylon juxta fluvium Narragam qui cadit in <Euphratem> unde civitati nomen’, the last statement being perhaps a clumsy attempt to derive Hippareni from Euphrates; it may even reflect some confused knowledge of the fact that the Euphrates was once called ‘the river of Sippar’.

It is usually assumed that Hippareni refers to Sippar (Ptolemy's Sippara), but even that requires proof, since the change of ‘s’ to ‘h’ is strange. In 1921 Andrae and Jordan, a short distance to the east of Sippar, examined another massive ruined site named Tell-ed-Deir, surrounded by a wall, dated at least to the ist dynasty of Babylon.

To the east side, the still-visible defences of Tell-ed-Deir are formed by a dried-up stream bed. Whether another ‘twin’ city lay on the far side of the bed cannot be stated.

⁶⁵ See Dilleman, loc. cit., for a detailed evaluation of Ammianus' and Zosimus' testimony.

⁶⁶ Adler, Elkan, The Itinerary of Benjamin of Tudela 53 (34).Google Scholar ‘Thence (from Rahbah) it is a two days’ journey to Karkisiya. [Circesium, Deir-ez-Zor.] Thence it is two days to El-anbar, which is Pumbedita in Nehardea.’ Benjamin evidently travelled by boat or raft. Nehardea is used apparently to describe a district.

⁶⁷ Obermayer, op. cit., 70 ff.

⁶⁸ Ray, John, A Collection of Curious Travels and Voyages, 2 vols. (1693).Google Scholar

⁶⁹ See note 6. The Patti-Bêl was also called the river Pittia, Waterman, op. cit., 883.

⁷⁰ Meissner, loc. cit.; according to Pallis, op. cit., 10, the ME-Enlil was the name of the section from Pallukat to Sippar. This seems a little difficult. Jacobsen claims the ME-Enlil ‘left the Euphrates’ right bank at Kish’, ‘The Waters of Ur’, Iraq xxii (1960) p. 176 n. 1, p. 177. He bases this statement ostensibly on Kraus’ article in Zeitschrift für Assyriologie li 57. But neither Kraus nor the interesting Sumerian text he published say anything of the kind.

⁷¹ Colonel Lane, Babylonian Problems, calls it Tell Aqar Kanisah, and adds, ‘it is 32 metres high, 182 ft. above sea level’. He, however, gives its distance from Babylon as 57½ miles. Lieut. Bewsher, J. B., ‘On part of Mesopotamia contained between Sheriat el-Beytha on the Tigris and Tel Ibrahim’, JRGS xxxvii (1867)Google Scholar quoting Chesney, gives it as 51½ miles in an air-line from Babel.

⁷² Mason, Major Kenneth, ‘Notes on the Canal system of ancient sites of Babylon in the time of Xenophon’, JRGS lvi (1920) 468 ff.Google Scholar He concludes that the battlefield cannot have been more than 35, or less than 28 miles from al-Anbar, the site of Cyrus' review.

⁷³ Op. cit.

⁷⁴ Obermayer, op. cit., 244–78.

⁷⁵ Ibid., 254.

⁷⁶ Ibid. 73 n. 1; 248–9.

⁷⁷ Selby, and Bewsher, , Survey of Mesopotamia; Sheriat el Beythra to Tel Ibrahim (1862–1865).Google Scholar

⁷⁸ i 7.14.

⁷⁹ i 8.4.

⁸⁰ iii 1.1; cf. ii 2.3; 4, 5.

⁸¹ ii 4.5.

⁸² ii 3.14.

⁸³ Al-ehmedi: Musil, op. cit., 149.

⁸⁴ ii 4.9, 12.

⁸⁵ ii 4.13.

⁸⁶ ii 4.25.

⁸⁷ Wadi Brisa inscriptions: Weissbach, , Die Inschriften Nebukadnezars II in Wadi Brisa (1906).Google Scholar A similar text on a cylinder from Dêr is published by Levy, S., ‘Two Cylinders of Nebuchadnezzar II’, Sumer iii (1947).Google Scholar

⁸⁸ i 185–6.

⁸⁹ ii 1.

⁹⁰ L. Waterman formerly claimed that two cuneiform inscriptions bearing names of kings of Opis were found in the excavations of Seleucia. The texts (on two basalt slabs) are published, in translations only, by Waterman, , ‘Preliminary Report on the Excava tions at Tel Umar’, Iraq 1931 6 Google Scholar, as ‘Urur, king of Sumer, king of Akshak’ and ‘Undalulu, king of Akšak, six years’. See also BASOR 32 (1956) 18; Archiv für Orientforschung v 121; vi 35. [But Professor Waterman now kindly informs me that these inscriptions were really too worn to be deciphered, and withdraws these readings.]

⁹¹ xvi 1.9.

⁹² i 189.

⁹³ Nat. Hist. ii 30, 31.

⁹⁴ Waterman, , Royal Correspondence of the Assyrian Empire 89.Google Scholar In the reign of Sharkališharri, twenty-third century B.C., the king of Elam advanced as far as Akshak; this suggests it was on the east bank. Cameron, , History of Early Iran 37–8.Google Scholar

⁹⁵ Luckenbill, op. cit.

⁹⁶ ii 1.

⁹⁷ Lieut. Bewsher, J. B., ‘On part of Mesopotamia contained between Sheriat el Beytha on the Tigris and Tel Ibrahim’, JRGS xxxvii (1867).Google Scholar

⁹⁸ in Pauly-Wissowa, RE.

⁹⁹ Edzard, , Die Zweite Zwischenzeit Babyloniens (1952).Google Scholar

¹⁰⁰ U. 16885. This letter will be published in Gadd, and Kramer, , Ur Excavations, Texts, vol. viii.Google Scholar

¹⁰¹ See above.

¹⁰² I.e., bẽru, double-hour's march.

¹⁰³ A town called Aiabu is mentioned as on the Euphrates in a letter from Mari, , Syria xix 121 ff.Google Scholar It is probably the same as lâbi', mentioned in a text of the seventh century B.C., apparently in the neighbourhood of Ramadi—see Musil, op. cit., 212–13.

¹⁰⁴ This letter is published in part only by Jacobsen, , ‘The Reign of Ibbi Suen’, JCS vii 39–40.Google Scholar I owe a complete translation to the kindness of Professor S. N. Kramer. On this period see Jacobsen, loc. cit., and Edzard, op. cit., ch. 5.

¹⁰⁵ Jacobsen, loc. cit.

¹⁰⁶ Letter from Kish, quoted by Edzard, op. cit., 47 and n. 208. The AP.KAL and ME.Enlil are again linked in a text of Halium (of Kiš?) a contemporary of Shumu-abum of Babylon, who dammed them. Edzard, 113.

¹⁰⁷ Letter from Puzur-Numushda, governor of Kazallu, to Ibi-Sin, apparently in his twentieth year, quoting Ishbi-Irra's proclamation, and describing the latter's forcible annexation of several cities, Nippur, Subir, Hamazi, Girkal, his pardoning of Eshnunna, Kish and Bad-zi-abba, which have defected to him, and his seizure of the ‘banks of Tigris, Euphrates, NUN.ME [AP.KAL] and ME.Enlil canals’; Falkenstein, , Zeitschrift für Assyr. xlix 60.Google Scholar

¹⁰⁸ Edzard, 33.

¹⁰⁹ ‘In this year in Ur they sold gold and silver and other precious objects in the temples to pay Isin’; Edzard, 47.

¹¹⁰ In his thirty-fifth year. See Meissner, , Babylonien und Assyrien (1920–1925).Google Scholar For the possible site of Rapiqu, see Goetze, , ‘An Old Babylonian Itinerary’, JCS vii, map on p. 72.Google Scholar

¹¹¹ ii 2.11–12.

¹¹² Polybius v 51.6. Zeuxis warns Antiochus that if he marches from Liba along the right bank, he would, after six marches, come to the ‘Royal Ditch’ and would have to return if it were held by Molon and he were unable to force a crossing.

¹¹³ ii 4.9–12.

¹¹⁴ Item Apamea, cui nomen Antiochus matris suae imposuit; Tigri circumfunditur haec, dividitur Archôo ( Pliny, , NH vi 21.132 Google Scholar). In his campaign against Bit-Yakin, Sennacherib brought Phoenician and Greek shipwrights and sailors to Nineveh and built galleys there, which he sailed down the Tigris to Opis, then dragged them on sledges (?) to the Arahtu to be refloated. There were two Apameas, Upper and Lower. The Upper is identified with Yakut's Zur-Famia (thirteenth century A.D.) and lies near Numaniya; Obermayer, op. cit., 86.

¹¹⁵ ii 4.13.

¹¹⁶ Op. cit.

¹¹⁷ Loc. cit., Inter has gentes (the Medes and Adiabenis) atque Mesenen Sittacene est, eadem Arbelitis et Palaestine dicta. Oppidum ejus Sittace Graecorum, ab ortu et Sabdata, ab occasu autem Antiochia inter duo flumina Tigrim et Tornadotum.

¹¹⁸ xi 13.6 cf. xv 3.12; xvii 17.

¹¹⁹ Lestrange, op. cit., 37.

¹²⁰ Collingwood, 's map, From Hillah to the Ruins of Niffer (1861/1862).Google Scholar

¹²¹ The Arahtu-Archôus?

¹²² ii 4.25.

¹²³ ii 4.27.

¹²⁴ Lestrange, op. cit., 92.

¹²⁵ For references see Obermayer, op. cit., 142. Here the Roman army crossed the Tigris after Julian's death (Ammianus xxv 6.8). ‘Da Dura nicht mehr in die eigentliche, reich kultivierte baby lonische Zone fällt indem schon unterhalb Dura der reine alluvialboden Babyloniens seine Nordgrenze gefunden hat, so dürfte in alten Zeiten, ebenso wie Gegenwärtig, die Umgebung von Dura als Weideplatz für Schafherden gedient haben’ (Obermayer, loc. cit.).

¹²⁶ ii 4.28.

¹²⁷ For Takritâin, or Birtu, see Musil, op. cit., 363.

¹²⁸ ii 5.1.

¹²⁹ iii 4.7.

¹³⁰ Layard, , Nineveh and Babylon (1853) 123–6Google Scholar for the ziggurat. For the circuit of the walls, see Jones, Felix's map, Nimrud and Selamiyeh (1852).Google Scholar

¹³¹ ii 4.10.

¹³² Thompson, R. C. and Hutchinson, R. W., A Century of Excavation at Nineveh (1929) 138.Google Scholar

¹³³ Thompson and Hutchinson, op. cit. Herzfeld, Encyclopaedia of Islam, s.v. ‘Mosul’, doubts the identification.

¹³⁴ Lestrange, op. cit., 87.

¹³⁵ Plutarch, , Antonius 45.Google Scholar