^{page 112 note 1} For further references to the topography and history of the Baliḫ valley the reader should consult the monumental work by R. Dussaud, Topographie historique de la Syrie antique et médiévale. See op. cit., p. 481 for a discussion on the source of the Balih, the location of which appears to be a matter of some doubt. I believe, however, that the headwaters of the Baliḫ proper derive from the perennial spring called ‘Ain al ‘Arus, just south of T. Abyad. It may be that the upper reaches of the Baliḫ are also fed by a spring in the neighbourhood of Harran, but this can hardly be regarded as more than a subsidiary source, since for the greater part of the year the amount of water feeding the Baliḫ from Turkey itself is a mere trickle. References to the Roman limes on the Baliḫ will be found in the comprehensive account written by Father A. Poidebard, entitled La Trace de Rome dans le désert de Syrie. The geographical aspect of the valley is also discussed by L. W. Lyde in The Continent of Asia, but in so far as the pre-Christian period is concerned his historical deductions should be treated with reserve; cf. also B.A.S.O.R. No. 78 for W. F. Albright's location of Ibla at the mouth of the Baliḫ. This however does not agree with the location of Ibla and Yarmuti proposed in Ur Excavations Texts I, p. 79 Google Scholar.

^{page 113 note 1} Cotton in antiquity. See the Nonesuch Press Herodotus, Book 3, section 106, note 3, referring inter alia to one of Sennacherib's inscriptions which is said to record the introduction of the plant to Nineveh. For references by the Arab geographers to the cultivation of cotton and the olive in the Ḫabur valley during medieval times see Oppenheim, M. Von, Vom Mittelmeer zum persischen Golf, Band 11, p. 4 Google Scholar.

^{page 113 note 2} For die discovery of Prunus sensu lato see du Mesnil du Buisson in Berytus, II, 1935, p. 125 Google Scholar, article entitled Souran et Tell Masin.

^{page 115 note 1} Cuneiform records from the Palace of Mari. See Dossin, G., Les Archives épistolaires du Palais de Mari (Syria, XIX, pt. 2)CrossRefGoogle Scholar.

^{page 113 note 2} Barbarians in Ḫanigalbat. See Smith, Sidney, Early History of Assyria, p. 249 Google Scholar.

^{page 113 note 3} T. Zaidan. See Albright, W. F. in Man, 1926, p.25 Google ScholarPubMed.

^{page 113 note 4} T. Halaf ware in the district of Van. See article by Reilly, E. B. entitled Test Excavations at Tilkitepe in Türk Tarih, IV. 1940 Google Scholar. A translucent green obsidian which occurs on the Balih is probably Vannic, and a speclded tortoiseshell variety discovered at Brak is also known to occur at Van.

^{page 113 note 1} The following are the references to examples of T. Halaf ware, showing the wideness of its dispersion. Mersin in Cilicia: Garstang, J. in A.A.A. XXV, pl. XXVIGoogle Scholar; Ras Shamra: Schaeffer, C. F. A. in Ugaritica, 1939, Figs. 4, 5Google Scholar; Carchemish: SirWoolley, Leonard in Iraq, 1, Pt. 2, pls, XVIII-XXGoogle Scholar; Til-Barsib: F. Thureau-Dangin and M. Dunand, Til-Barsib, pl. XXXVI; T. Halaf: M. Von Oppenheim, Der Tell Halaf; Chagar Bazar: M. E. L. Mallowan in Iraq, III Pt. 1, Figs. 21-7; Jebel Sinjar: Seton Lloyd in Iraqt v, Pt. 2; Hassuna: Seton Lloyd and Fuad Safar in J.N.E.S. IV, 10 1945, No. 4Google Scholar; Nineveh: Thompson, R. Campbell and Mallowan, M. E. L. in A.A.A. XX Google Scholar; Arpachiyah: M. E. L. Mallowan and J. Cruikshank Rose in Prehistoric Assyria, The Excavations at Tall Arpachivah, 1933; Tepe Gawra: E. A. Speiser in Asia, Sept. 1938, Closng the Gap at Tepe Gawra, pp. 536 ff., also A.f.O. XII, Heft 3, 1938, p. 169. ‘Amuq Plain: R. J. Braidwood in O.I.P. XLVIII.

^{page 113 note 2} The overlap between the Samarran and T. Halaf ware is clearly illustrated from the site of Samarra itself by E. Herzfeld, Die Ausgrabungen von Samarra, pl. XLIV, especially the sherds numbered 295, 296. See also McCown, D. E., The Comparative Stratigraphy of Early Iran, 35–8Google Scholar, for evidence of Samarran cognates in Persia.

^{page 113 note 3} Speiser, E. A. in Asia, 09 1938, p. 542 Google Scholar, refers to the syncretistic period represented by Gawra XII-XIX. There now appears to be some doubt whether Samarra ware is in fact represented at Tepe Gawra, see Perkins, Ann L. and Braidwood, R. J. in J.N.E.S. III, no. 1, p. 69 Google Scholar, but it is none the less becoming increasingly obvious that the painted fabrics of T. Halaf, Samarra, and Al-‘Ubaid have to some extent a common background of design. In the same article Edna Tulane has made a useful and significant repertoire of the Samarran painted pottery style. It should, however, be emphasized that the site from which any specific ware derives its name does not necessarily provide the full repertoire which the ware can display. I would also emphasize the importance of shape, which is often even more significant than the quality of the fabric, firing, and Discuit, for indicating the particular school to which a specific ware belongs. In this connexion it should be noted that in the Al-‘Ubaid period there are certain bowl shapes common to prehistoric Assyrian sites and absent from South Mesopotamian sites. None the less it is obvious that there are many characteristics common to the numerous varieties of the ware throughout the Tigris-Euphrates valley at this period. The term Al-‘Ubaid is therefore a useful designation for a specific fabric which is found in abundance as far north as the region of Nineveh and as far south as Bundar Bushire on the Persian Gulf.

^{page 118 note 1} O.I.P. LIII for the oval plan at Khafajah. For a similar form of construction at Al-‘Ubaid see P. Delougaz in Iraq, V, Pt. I, A Short Investigation of the Temple at Al ‘Ubaid. I would like to take this opportunity of correcting a slip for which the author of this article was not responsible: it occurs in footnote i on p. 6. I did not conduct any sounding at Al-‘Ubaid. See also Seton Lloyd and Fuad Safar in J.N.E.S. II no. 2, for an account of T. Uqair, and Delougaz and Lloyd, Pre-Sargonid Temples in the Dtyala Region.

^{page 118 note 2} The inner town defences of Carchemish of the second millennium B.C. were constructed on a roughly oval plan, but Woolley considers that this was dictated by conditions peculiar to the site at that time, and does not agree with Koldewey's suggestion that the plan was analogous with that of Zinjirli; cf. Carchemish, Pt. II, The Town Defences by C. L. Woolley, pp. 43 and ff., and pl. 3. For Zinjirli see also J. Gakstang, The Hittite Empire, chapter ix, and von Luschan and others in Mitt, aus den Orient. Sammlungen, XI, XII, XIII.

^{page 118 note 3} Phraaspa (modern Takht- i- Suleiman), see Illustrated London Newsy Feb. 26th, 1938, and article by Pope, Arthur Upham in Bulletin of the American Institute for Iranian Art and Archaeologyt V, no. 2, 12 1937 Google Scholar.

^{page 119 note 1} Oval layout of Parthian Ctesiphon and Hatra. See Oscar Reuther in A Survey of Persian Art, edited by Arthur Upham Pope, 1. 575.

^{page 119 note 2} There is an interesting discussion of a number of eariy circular and oval city plans in Creswell, K. A. C., Early Muslim Architecture, II, chapter i, pp. 18 ff.Google Scholar For the oval plan of Assyrian Arslan-Tash see F. Thureau-Dangin, A. Barrois, G. Dossin, M. Dunand, Arslan-Tash, site plan at end of the text.

^{page 119 note 3} See A. Poidebard, op. cit., pp. 149, 150 for T, Mu‘ezzar, T. Bati, and T. Baindar: the last-named mound is referred to by him as T. Beydar. Poidebard describes these as Mitannian types of fortifications, a plausibie conjecture, but lacking archaeological proof. Reference should also be made to the oval-shaped walled towns at T. Khoshi and T. Hadhail in the Jebel Sinjar; cf. Seton Lloyd in Iraq, v, Pt. 2, 126 and VII, Pt. 1, 19.

^{page 120 note 1} See Mallowan, M. E. L. in Mélanges syriens offerts à Monsieur René Dussaud, tome II, pp. 887–94Google Scholar.

^{page 120 note 2} Smith, Sidney, Alalakh and Chronology, pp. 38 ff.Google ScholarPubMed

^{page 122 note 1} See K. A. C. Creswell, op. cit. II. 49.

^{page 126 note 1} Sherds of a ware which may well be ascribed to the neolithic period and are probably contemporary with some of the earliest pot fabrics of the kind in Western Asia were found on virgin soil at Chagar Bazar. Compare particularly the black and grey burnished wares, some of them incised with a ‘rocker’, from Chagar Bazar, in Iraq, III, Pt. 1, pl. III, no 12, with similar ware from the neolithic levels at Mersin illustrated in A.A.A. XXVI, pl. XLII, no. 3. The evidence, scanty as it is, suggests to me that the earliest black and grey wares of SE. Anatolia were diffused over the north Syrian steppe at a very early period.

^{page 128 note 1} See also pl. XXVIII.

^{page 132 note 1} I have followed Dr. Sidney Smith's interpretation of the evidence from Alalakh. For the reference see p. 120, n. 2 above.

^{page 135 note 1} See Iraq III, p. 55 Google Scholar, G 6 and Fig. 7, nos. 27, 29 for the type of faience duck-beads. The Chayar Bazar grave can now be securely assigned to the Sargonid period on account of the burnished pottery found with the beads. The Sardonici ducks differ from those of the Jamdat Nasr period which are usually engraved under the base and are represented with head resting on the back.