¹ That Hatti did at times exercise a certain control over Cilicia is proved by monuments (e.g., Muvatalli on the rock-carving at Sirkeli; Garstang, and Güterbock, in LAAA xxiv, Nos. 1–2, p. 64 ff.Google Scholar), but even so the country seems to have been always a separate and generally a more or less independent kingdom.

² In 1935 Ras Shamra had not produced any Minoan sherds such as have been found since then.

³ I. Kings ix. 18.

⁴ The flat and marshy plain of Alexandretta is of recent formation, and in antiquity the sea washed the foothills, making it more dangerous than it is to-day for a sailing-vessel that might drag its anchor.

⁵ So Gudea, in the twenty-fourth century B.C., brought the timbers for his temple of Nin-gir-su from ‘the Amanus, the cedar mountain.’ Thureau-Dangin, F., SAKI p. 68Google Scholar (16.5.28).

⁶ As, most recently, by Dussaud, R. in Topographie historique de la Syrie, pp. 418–421Google Scholar.

⁷ See below, p. 30.

⁸ For a brief report of a reconnaissance at Basit see Schaeffer, C. in Syria, xvi. ii. (1935), pp. 173sq.Google Scholar

⁹ M. Prost was killed in a motor accident in the summer of 1936. His death was a serious blow to archaeology in North Syria, for to keen scholarship and untiring energy he added a detailed knowledge of his province such as no other man possessed.

^9a See for the coins of our 1936 season, Numismatic Chronicle, Fifth Series, vol. xvii.

¹⁰ To be published by Mr. E. A. Lane in the Antiquaries' Journal.

¹¹ In the deep trench Cypriote sherds were found below the Attic as well as above; they were not really in situ, but may imply the existence of early buildings below the acropolis. The trench was filled with boulders and rubble from the acropolis wall, lying above the 4th-century pottery, which proves that the ruin of the town dates to after that period.

¹² See on this my report on the excavations at Atchana in the Antiquaries' Journal, xviii, p. 1Google Scholar.

¹³ Normal, that is, to Moslem houses; the Christian and (usually) the Alouite house of to-day has a tiled roof.

¹⁴ The analogy cannot be pressed very far. It looks as if there had been a complete remodelling of the town between Levels V and IV, at least in the N.E. part of it, and only on the N.W. side do the lines of the streets more or less agree.

¹⁵ E.g., the short right-angled turn in Sq. E. 8 and the deflection of the road in Sq. D. 7, after which the line is brought back to the true. (Level III, Plan.)

¹⁶ The impost-stone against the middle of the N.E. wall of room has an exact parallel in the impoststone against the middle of the N.E. wall of room I in House E of Level III.

¹⁷ They belong to Levels II and III; none of Level IV were found.

¹⁸ As I have explained above, Level IV extended to the south beyond Level V, so that the houses here were built not on the tell formed by the ancient settlement, but at its foot; they were at about the same height above the river as the houses of Level VI.

^18a Compare also Ugarit (Ras Shamra) with its adjacent harbour.

¹⁹ Classifications des céramiques antiques: Cypriote Pottery, pp. 42–43.

²⁰ I published pottery of this sort in Syria, ii (1921), pp. 177et seq.Google Scholar

²¹ One supposes that in this case the contents were for sale but were decanted and sold in smaller vessels.

²² Or fit together only to form a larger but none the less isolated fragment; this happens where a large fragment has been broken again by being trampled on after burial.

²³ Good early black-figured ware from Athens has been found far inland in northern Syria, e.g., at Devi Huyuk on the caravan-route from the Orontes valley to the Euphrates ford at Carchemish viâ Aleppo; see LAAA vii. (1914), Pl. XXVIIGoogle Scholar.

²⁴ Probably a good deal earlier; we have a moderate amount of black-figured ware, going back to c. 520 B.C., and no Corinthian or other Greek fabrics of anything like so late a date, and if the trade with those centres had gone on after 520, evidence of it ought to have been forthcoming. On negative evidence it is perhaps rash to say that the Attic monopoly dates back to 520, although it is certain that by then Athens had secured the vast bulk of the trade.

^24a See Robinson, E. S., Numismatic Chronicle, Fifth Series, vol. xviiGoogle Scholar.

²⁵ The first time we found them thus I supposed that we had to do with the ruins of a kiln, for the floor and sides of the clay enclosure were burnt red and many of the lekythi blackened by heat; subsequently we were able to see that this was the result of the oil which the lekythi had contained catching fire.

²⁶ Liquid mercury is found also in considerable quantities in the soil in one part of the site of Seleucia, and the explanation of it is presumably the same, the trade having simply shifted there when the new harbour was built. I am indebted to Lord Rayleigh for the note re Almaden.

²⁷ For Ποταμοὺς Καρῶν Dussaud, , Topographie historique de la Syrie, p. 419Google Scholar, suggests with good reason Ποταμοὺς Ὑδάτων, which was the old name of Seleucia according to Strabo xvi. 2, 8. The name Seleucia was first introduced in 301–300 B.C., but there had clearly been an earlier settlement on the site. According to the Gurob Papyrus (Holleaux, in BCH 1906, pp. 330–348Google Scholar), Ptolemy III Euergetes in 247 B.C. again captured the fortified position of Posidium as a base for his attack on Seleucia; this Φρόυριον I take to be the hill-town at Sabouni, which may well have been maintained after the abandonment of the harbour.

²⁸ Herod. III. 91, Rawlinson's translation.

²⁹ The Stadiasmos does put Charadrus between them, but see the next note. Ptolemy V. iv. 2, enumerating the harbours of N. Syria, gives ‘Seleucia, Posidium, Heraclea, Laodicea.’

³⁰ The site of Heraclea is unknown; it is described as being 100 stades south of Posidium and as the next harbour to the north of Laodicea, from which it was distant 20 stades; the figures in the Stadiasmos are here obviously corrupt. According to Pliny, , H.N. v. 17Google Scholar, the port of Charadrus lay between Posidium and Heraclea, in which case an identification of Heraclea with Basit could scarcely hold good.

³¹ Stad. m. m. 143; see Dussaud, , Topographie historique de la Syrie, p. 421Google Scholar.

³² Priscian, , Perieg. v. 856Google Scholar.

³³ Malalas, , Chronographia, viii. 257 (Ed. Dindorf, )Google Scholar.

³⁴ Evans, Palace of Minos, IV, p. 781Google Scholar, citing Dussaud, in Syria, x. (1929), pp. 301–3Google Scholar.

³⁵ In modern Arab parlance 'Am(u)ki; the name of the plain is spelt either Amk (which spelling I have used) or Amuk, as in the French maps. The name is ancient, appearing in Assyrian texts. According to Malalas the Plain was named after the lady, but the reverse is equally probable.

³⁶ Cypriote Bronze Age pottery is common at Atchana.