^{page 33 note 1} Ceilings: Eighteenth or Nineteenth Dynasties; d'Avennes, Prisse, Histoire de l'art égyptien IGoogle Scholar, Architecture, Plafonds, 31st plate in the list though they are not numbered, figs. 4, 6, 7, 9; Twenty-Sixth Dynasty; 34th plate of op. cit., fig. 1. As is noted further on a development of the design is quite well-known on scarabs and amulets.

^{page 33 note 2} SirWilkinson, J. Gardner, The Manners and Customs of the Ancient Egyptians (edn. Birch), iGoogle Scholar, Pl. viii, fig. 7, and p. 363 and note 4. It is a very imperfect copy and is not fully shown here, but I have myself observed the pattern on the ceiling.

^{page 34 note 1} The dates are those of Edgerton in JNES, i (1942), p. 314Google Scholar. The overlap of Amenemhat II with Senusret I is not a mistake, but is the co-regency of the two.

^{page 34 note 2} Griffith, The Inscriptions of Siût and Dêr Rîfeh, Pl. 4, first marginal note.

^{page 34 note 3} Blackman, , The Rock Tombs of Meir, iiiGoogle Scholar, Pl. ix = xxviii on a larger scale and p. 15 for the king.

^{page 34 note 4} Petrie, Antaeopolis, Pl. i, fig. 2, from which our fig. 1 is taken. For the date see Steindorff in Steckeweh, , Die Fürstengräber von Qâw, p. 8Google Scholar and the present writer's review of the two books in JEA, xxiv (1938), pp. 143 ffGoogle Scholar.

^{page 34 note 5} In the 1837 and 1847 editions of Wilkinson's, Manners and Customs, vol. iiGoogle Scholar, Pl. vii, fig. 7, Hapzefi's design is blue, not green as in the Birch edition.

^{page 34 note 6} Davies, Nina M., Ancient Egyptian Paintings, iGoogle Scholar, Pl. xxii. It shows itself a dark greenish colour to-day, but where the surface is broken away the original colour is seen to be a bright blue.

^{page 34 note 7} It is hoped to elaborate this before long in a companion article.

^{page 34 note 8} Compare the plans in Porter, and Moss, , Bibliography, iv, p. 260Google Scholar Hapzefi, and v, p. 12 Qau.

^{page 34 note 9} Hapzefi, Wilkinson, op. cit. i, Pl. viii, figs. 4, 7, 20, facing p. 363; Wahka B at Qau, Petrie, Antaeopolis, Pl. i.

^{page 34 note 10} Davies, in JEA, xxvii (1941), pp. 127 fGoogle Scholar.

^{page 35 note 1} Brunton, , Qau and Badari, iGoogle Scholar, Pls. xxxii, xxxiii; id., Mostagedda, Pls. lx, lxv, 14–17; id., Matmar, Pl. xxxiii.

^{page 35 note 2} No. 49 of the Abydos List, Stock, , Die erste Zwischenzeit Ägyptens, pp. 42, 43Google Scholar. If, as Stock seeks to establish, the Eighth Dynasty reigned at Abydos, it would be thoroughly in accord with conditions in Middle Egypt. But in reviewing his book Posener points out the difficulties of such a view, and falls back on the Manethonian tradition that the dynasty reigned at Memphis (Posener, in Bibliotheca Orientalis, viii (1951), pp. 165–172)Google Scholar.

^{page 35 note 3} Stock, loc. cit., No. 45 of the Abydos List.

^{page 35 note 4} Scharff, , Der historische Abschnitt der Lehre für König Merikarê, pp. 19–21Google Scholar.

^{page 35 note 5} id., op. cit., p. 20, l. 95 and note 61. Under the name of ‘The Wall of the Prince” it figures again in Erman, , The Literature of the Ancient Egyptians (tr. Blackman, ), p. 115Google Scholar, The Prophecy of Neferrohu, and p. 17, The Story of Sinuhe.

^{page 35 note 6} Newberry, , Beni Hasan, iGoogle Scholar, Pls. 28, 31 and p. 69, and for Absha and the donkey and for four of the women see Nina M. Davies, op. cit., Pls. x, xi.

^{page 35 note 7} Erman, op. cit., p. 96, The Admonitions of Ipu-wer. The Egyptians had been getting “cedars” and their “oil or resin” from Keftiu. Actually it was juniper oil that was used for embalming (Lucas, A., Ancient Egyptian Materials and Industries (3rd edn.), pp. 355, 358Google Scholar, for the resins see pp. 368 ff, and for the old-established loose use of the word “cedar” see p. 492, for the use of cedar wood see p. 489, and for juniper wood see p. 490). The sweet-scented juniperus excelsa still grows in Cilicia Tracheia (Schaffer, , Cilicia, p. 51Google Scholar (Petermanns Mitteilungen, Ergänzungsheft, Nr. 141, 1903). Kilindria, i.e. Chelindreh-Kelenderis, still exports timber and firewood besides other things (Cuinet, V., La Turquie d'Asie, ii, p. 80)Google Scholar.

^{page 35 note 8} This puts out of court Miss Kantor's claim that the design in Menkheperrasenb cannot be taken seriously, and her further claim that the ignorant and careless artist just filled up his Keftiuan kilts with a chance lot of native Egyptian motifs collected from any where in the country (AJA, li (1947), p. 43)Google Scholar. The article has been reprinted separately under the title The Aegean and the Orient in the Second Millennium B.C. as No. iv of the Series of Monographs on Archaeology and Fine Arts Sponsored by the Archaeological Institute of America and the College Art Association, 1947.

^{page 36 note 1} Fig. 2 is taken from Nina and Norman de G. Davies, The Tombs of Menkheperrasonb, Amenmosĕ, and another, Pl. v, top row, 3rd man. The pattern has often been published by students of Keftiu, for instance Evans, , The Palace of Minos, ii, p. 745Google Scholar, fig. 480 b (where it is turned upside down) which is the most often used; Wainwright, in LAAA, vi (1914)Google Scholar, Pl. xv, fig. 17, etc., etc.

^{page 36 note 2} Davies, op. cit.,, p. 15 and passim. The usually accepted belief that Menkheperrasenb was the son of Rekhmirê is not correct, p. 16.

^{page 36 note 3} Schmidt, Hubert, Heinrich Schliemann's Sammlung Trojanischer Altertümer, nos. 6003, 6133, 6134 (Beilage ii) and 6401Google Scholar, and two others are shown by Schliemann, in Ilios, p. 489Google Scholar, nos. 849, 850. Another was found by Blegen, , Caskey, , Sperling, Rawson, Troy, i, fig. 356, 37–709Google Scholar, and p. 367.

^{page 36 note 4} Schmidt, op. cit., gives a sample one on p. 236, fig. d 5 (otherwise called no. 17), from which our fig. 3 is taken.

^{page 36 note 5} Blegen and others, op. cit., pp. 207, 210.

^{page 36 note 6} Milojčić, in BSA, xliv (1949), p. 304Google Scholar; Schaeffer, , Stratigraphie comparée, p. 292Google Scholar, calls this phase Troy III and puts it slightly later, 2300–2100 B.C.

^{page 36 note 7} Schaeffer, op. cit., fig. 176, no. 12.

^{page 37 note 1} Mallowan, in Iraq, ix (1947), p. 171Google Scholar, no. 8. An important study of the pattern is made on this and the following pages.

^{page 37 note 2} id., loc. cit., A number of other spiral decorations are studied resulting in the suggestion that spirals originated in northern Persia and came westwards.

^{page 37 note 3} Murray, A. S. and others, Excavations in Cyprus (Brit. Mus.)Google Scholar, Pl. xi, 448, Tomb no. 79, and for the date Gjerstad, , Studies on Prehistoric Cyprus, pp. 283, 287, 335Google Scholar.

^{page 37 note 4} Evans, , The Palace of Minos, ii (I), p. 208Google Scholar, fig. 117 B. For the probable date, to which Dr. Stais agrees, see pp. 199, 200. Kythera had considerable connections with the east at one time and another, and actually an oriental object had been imported there at the time of our vase. It bore the name of Naram-Sin son of Ibiq-Adad who were kings of Eshnunna in Mesopotamia about then (JHS, lix (1939), pp. 137 fGoogle Scholar; Wainwright, in Antiquity, xviii (1944), pp. 59–60)Google Scholar.

^{page 37 note 5} Evans, op. cit. i, Pl. i, fig. k facing p. 231 = ii, p. 200, fig. 110 A, no. k. At this same time or slightly later the same design appears in Egypt on scarabs of the Twelfth to Thirteenth Dynasty Age, as for instance id., loc. cit., fig. 110 A, nos. c, d.

^{page 37 note 6} Heurtley, in BSA, xxv (1921–1922; 1922–1923), p. 139Google Scholar, fig. 32, no. aa. Nearly five hundred years later an elaborated verson of this appeared on the ceiling of Imisibe's tomb at Thebes (for which see further, p. 43, note 2). This dates to the late Twentieth Dynasty and, therefore, to about 1100 B.C.

^{page 37 note 7} Karo, G., Die Schachtgräber von Mykenai, Text, pp. 274–9, Figs, 108–123Google Scholar.

^{page 37 note 8} id., op. cit., Pls. xxi, figs. 56, 57, and p. 52; xxxvii, fig. 234, and p. 73; lvi, figs. 649, 650, and p. 125.

^{page 38 note 1} The developed specimen in Karo's fig. 234 (our fig. 4) is specially interesting in showing what can happen to the design. Here the palmettes have been removed from within the volutes to the outside and then greatly enlarged. The whole design has then been turned sideways. Later on by some four hundred and fifty years this led to a beautiful descendant at Enkomi in Cyprus in the Late Cypriote III Period, c. 1200–1000 B.C. (see note 3, p. 40 and fig. 7). Karo's nos. 649, 650 (our fig. 5) exhibit yet another development of the design. In this case the palmettes have disappeared and each spiral is joined to its neighbour on the opposite volute by a swelling bow which arches over between the two. It is so exaggerated that it becomes the major part of the new pattern, quite over-powering the original pair of double volutes at the centre.

^{page 38 note 2} Kunze, E., Kretische Bronzereliefs, p. 121, fig. 12Google Scholar.

^{page 38 note 3} id., op. cit., p. 122, fig. 13b. The outside palmettes are represented by small triangles themselves representing the little conical cores from which the palmettes had at one time sprung. Similarly, large triangles are stood within the volutes where they represent the broad-based spikes which at Bor, for instance, fill the whole space.

^{page 38 note 4} Payne, in BSA, xxix (1927–1928)Google Scholar, Pl. xviii.

^{page 38 note 5} id., op. cit., Pl. xxiii, fig. 2. The pithos is one of the latest of the polychrome series, p. 241, no. 51.

^{page 38 note 6} Wide, in Mitt. kais. D. Arch. Inst., Athenische Abteilung, xxii (1897), p. 243, figs. 13, 13aGoogle Scholar.

^{page 39 note 1} id., op. cit., p. 242, fig. 12a = Kunze, Kretische Bronzereliefs, p. 122, fig. 13a.

^{page 39 note 2} Kunze, op. cit., Pls. 21–23 and pp. 118 f. For the date see Benton, in BSA, xxxix (1938–1939)Google Scholar, p. 62, and id. in BSA, xl (1939–1940), pp. 53, 54Google Scholar, no. 14, for some more late examples of the design, and a discussion of the shield which is shown again on Pl. 25. I have to thank Miss Benton for a number of these references and for much help with the Orientalising Period.

^{page 39 note 3} It no doubt originated far away to the east somewhere probably in the direction of northern Persia. There, in the remote past as for instance at Sialk III, patterns formed on spirals and hooks are common; the hooked diamond going back into Sialk II (Ghirshman, , Fouilles de Sialk, i, p. 49Google Scholar, Pls. Frontispiece, 3, xii, etc., xiv, etc., lxxvi B, 22, and for Period II p. 28, fig. 3, no. 6. Sialk is close to Kashan). All of this is of the ‘Ubaid Period, which is long before Jemdet Nasr (McCown, The Comparative Stratigraphy of Early Iran, Table ii at the end of the book), that is to say long before the Egyptian First Dynasty and 3000 B.C. In Sialk IV, which is of Jemdet Nasr date, there was a pin with a head of an elementary pair of outward turning spirals (Ghirshman, Pl. xxix, fig. 1, b) of which the pin of Troy II (Schliemann, , Ilios, p. 489Google Scholar, no. 848) is a glorified derivative. Similarly, the Trojan pin, no. 849, seems to represent a memory of the “tree” of spirals of Sialk III (Ghirshman, Pls. Frontispiece, xii, etc.). But at Troy the spirals turn inwards whereas at Sialk they turn outwards.

^{page 39 note 4} Prisse d'Avennes, op. cit., Pl. 31, figs. 4, 6, and Texte, p. 368.

^{page 39 note 5} id., op. cit., Pl. 34, fig. 1. The artist also reproduced the curious pattern something like an aeroplane within the diamond which seems to be peculiar to Qau. He also reproduced the surrounding spots which also appear at Qau, though unlike the pattern they appear at Asyut as well. Both details are lacking at Meir.

^{page 40 note 1} id., op. cit., Pl. 31, figs. 4, 6.

^{page 40 note 2} Mrs. Davies only shows the pattern in outline in the publication from which fig. 2 is taken, hence does not show the filling of the diamonds. She shows it, however, in her large coloured copy, Ancient Egyptian Paintings, i, Pl. xxii, which also shows the pitiable condition of the whole.

^{page 40 note 3} As has already been seen and will be seen again, strange things happen to the palmettes. A crude, vague, memory of them, still inside the volutes, was common at Enkomi in Cyprus, fig. 6 (A. S. Murray and others, op. cit., Pls. vi, 524, Tomb 93; viii, bottom right hand corner, Tomb 19; xi, 450, Tomb 79; xi, 192, Tomb 53 (from which our fig. 6 is taken), 367*, Tomb 67) in the Late Cypriote II Period, which dates to 1400–1200 B.C. (Gjerstad, op. cit., pp. 283, 287, 335). This design with its cross bars is no doubt also some sort of derivative from fig. 8 though all attempt at reproducing palmettes had disappeared from that.

Then, in the next Period, Late Cypriote III (1200–1000 B.C. or 1050 B.C. as Schaeffer puts it) we get a return to the original simple pattern, but with the palmettes so exaggerated that they can no longer find room inside the volutes but are pushed out beyond them. The pattern still remains upright in its original position (Schaeffer, , Enkomi-Alasia, Pl. xxxix, figs. 1, 2, pp. 170, 179)Google Scholar. The same Period produced a beautiful development (fig. 7). Here the palmettes have been so enlarged that they have been moved round to the outside of the volutes and have blossomed out into the major part of the scheme (fig. 7 is taken from Murray and others, op. cit., Pl. vii, 184, Tomb 45, and for the date see Gjerstad, op. cit., pp. 284, 335). To accommodate the great palmettes the whole design is turned sideways, and so approximates the idea which had already been expressed in the Fourth Shaft Grave at Mycenae 450 years or so earlier (see note 1 p. 38, fig. 4). In that case the palmettes had still been naturalistic, but here at Enkomi they have become stylised. The palmettes spring from a small conical core which in later days still remains long after the palmettes themselves had disappeared (see note 2 p. 43). The Enkomi development of one portion of the design until it overshadows the original basis of the whole is comparable to the other development at Mycenae which was recorded in note 1 p. 38, and fig. 5.

^{page 41 note 1} Fig. 8 is taken from Engelbach, Harageh, Pl. xxi, no. 143, and shows the appendages which, however, are generally omitted. As already mentioned on p. 38 the appendages are afterwards lost until the seventh century, when they reappear this time in Crete and as tendrils.

^{page 41 note 2} Thus, if the otherwise “gaping blanks” of the Keftiuan kilts were only filled up with “standard motives” of Egyptian art as Miss Kantor postulates (op. cit., p. 43), this would have been an extraordinarily bad copy of such patterns, and a good example of the “extraordinary carelessness” and “surprising ignorance”, etc., etc., which is too readily supposed to characterise the Egyptian artists' work. On the other hand I would suggest that in reality it is yet another version of what was evidently a “standard motive” of some Asiatic art which entered Asia Minor, Keftiu and Egypt.

^{page 42 note 1} Nina M. Davies, op. cit., Pl. xxii.

^{page 42 note 2} Ramsay, W. M., The Historical Geography of Asia Minor, p. 361Google Scholar, under the city's ancient name of Laranda, and map facing p. 330. The Christian inhabitants of Karaman still call their town Laranda, p. 336.

^{page 42 note 3} Hogarth, and Munro, , Modern and Ancient Roads in Eastern Asia Minor, p. 645Google Scholar f (Roy Geogr. Soc., Supplementary Papers, iii (1893)Google Scholar, Pt. 5.

^{page 42 note 4} Ramsay, op. cit., p. 361.

^{page 42 note 5} Beaufort, F., Karamania (1818), p. 209Google Scholar, who describes it as “a snug but very small port”.

^{page 42 note 6} Leake, W. M., Journal of a Tour in Asia Minor (1824), pp. 103, 118Google Scholar. He landed at Tzerina or Cerina (Turkish Ghirne), a small port on the north coast.

^{page 42 note 7} Ramsay, op. cit., p. 362.

^{page 42 note 8} Apollodorus, iii, xiv, 3. He also says that Sandokos had come to Cilicia from Syria and married the daughter of the king of Hyria. According to Stephen of Byzantium (s.v) Hyria was the ancient name of Seleucia at the mouth of the Calycadnus. It is evidently at least as old as the eighth century B.C., for Sargon, having won a battle on the seashore, established a fortress at a place he calls Ḫarrua at the far western end of Que, i.e. Cilicia Pedias (Forrer, , Die Provinzeinteilung des Assyrischen Reiches, p. 71Google Scholar; Naster, , L'Asie mineure et l'Assyrie, etc., p. 37Google Scholar (Bibliothèque du Muséon, Louvain, 1938, vol. viii))Google Scholar. It has often been thought that the KRNTRYS which is named at Karatepe in the eighth century B.C. would have been Kelenderis in Cilicia Tracheia. But this need not have been so, for there proves to have been a place Kylindros in the valley of the Pyramus itself and, therefore, much nearer Karatepe than is Kelenderis (Levy, Isidore in Bull, de la classe des lettres et des sciences morales et politiques (Ac.roy.de Belgique), 5e serie (1949), xxxv, p. 472)Google Scholar. These inscriptions have already become the subject of a vast literature, a bibliography of which is published by Bossert, and others, Die Ausgrabungen auf dem Karatepe (1950), pp. 76–83Google Scholar.

^{page 42 note 9} The Encyclopaedia of Islam, s.v., p. 744.

^{page 43 note 1} Garstang, , The Land of the Hittites, Pl. lvi facing p. 186Google Scholar, bottom row but one, or id., The Hittite Empire, Pl. xxxiii facing p. 162.

There is another case of a pattern lasting on in this neighbourhood almost indefinitely. This time it is one from Ivriz, a place about 50 miles from Bor. One of the patterns on the dress of the king there is a swastika of an unique shape, the four arms springing as they do, from a central bar instead of from a common centre (in the bottom row of the tunic). The identical pattern has been found again on a carpet bought at Konia, the modern capital of the district and about 95 miles away. The carpet is undoubtedly of local make for besides having been bought at Konia its knotting, colour scheme and severe geometrical design all indicate an Asia Minor origin (Sarre, F. in The Burlington Magazine, xiv (Oct., 1908, to March, 1909), pp. 143–7 and plate)Google Scholar. The carpet is no older than the eighteenth century, if as old, and the king at Ivriz whose tunic shows the pattern was reigning about 740 (Bossert, , Altanatolien, p. 70Google Scholar and note to no. 796). He is Urballa again, the same man as at Bor (Delaporte, in RHA, iv (1936–1938), pp. 138 fGoogle Scholar, and Pl. i shows the best photograph of the sculpture). The Ivriz monument, and probably also that of Bor before it was damaged, is of interest in another direction, for, the king's costume shows resemblances to that of a certain Keftiuan. The robe at Bor also has swastikas arranged in squares as at Ivriz. They are like those at Ivriz not only in the many turns they make, apparently five, but also in their position on the robe, which in each case is along the bottom edge. But so far as can be seen from the published photographs, the ones at Bor are of the normal construction and therefore to this extent unlike the others. The two sculptures may easily be compared in Garstang, The Land of the Hittites, Pls. lvi, lvii or in id., The Hittite Empire, Pls. xxxiii, xxxiv.

^{page 43 note 2} As a matter of fact the disappearance of the palmette leaving only the little conical core and the removal of that to the outside of the volutes had already taken place at the very time of the splendid Enkomi palmette. This is dated to the Late Cypriote III Period, c. 1200–1000 B.C. (see p. 40 note 3) yet on the ceiling of Imisîbe at Thebes in Egypt, fig. 10, the little conical core outside the volutes is already shown, though without filling the space between them as it should (Prisse d'Avennes, op. cit., 31st Plate, fig. 5, and Texte, p. 368, where the owner is called Aïchesi. Our fig. 10 is taken from there). Imisībe lived towards the end of the Twentieth Dynasty and, therefore, about 1100 B.G. (Porter, and Moss, . Bibliography i, p. 94, No. 65)Google Scholar. For further remarks on Imisîbe's pattern see p. 37 note 6, and for Enkomi see p. 40 note 3. His pattern also shows the bands tying together the two pairs of volutes, which are so regularly seen in the late examples.

^{page 43 note 3} At this time the design is well known away. to the east. It decorates the crossbar of Ashur-nasir-pal's seat, 885–860 B.C. (Budge, Wallis, Assyrian Sculptures in the British Museum: Reign of Ashur-nasir-pal, 885–860 B.C., Pl. xxxi)Google Scholar. In this case, as in the others quoted here, the design is set upright as it should be, but the original internal palmettes or bars have been transformed into a binding holding the two pairs of volutes together as they were in the beautiful example from Enkomi in Cyprus of three hundred years earlier. The splendid palmettes of Enkomi have however disappeared and are now represented only by the little conical core on the outside of the volutes from which they originally sprang.

An actual seat from Assyria bears the same ornament in the same position (von Luschan, , Ausgrabungen in Sendschirli, iv, p. 347, fig. 256)Google Scholar as do the lower cross-bars of Sennacherib's great throne of state, some hundred and seventy-five years later, Sennacherib having reigned from 705–681 B.C. (Kleinmann, Assyrian Sculptures, Pl. lxix = A Guide to the Babylonian and Assyrian Antiquities: British Museum (1922), Pl. xix and p. 49Google Scholar where, however, the details are not clear). In between the two extremes of Cilicia on the west and Assyria on the east the Assyrian version of the design is seen on the cross-bar of another seat of the same sort as Ashur-nasir-pal's. This is the one used by Barrakub at Sendschirli (Sinjirli, Zincirli) about 730 B.C. (von Luschan, op. cit., Pl. lx). It is conveniently reproduced by Bossert, , Altanatolien, fig. 952 and references on p. 76)Google Scholar. His footstool has it also in the same position.

The double pair of volutes were shown attached to a sort of tree on a Late Mycenaean vase from Livadia in Boeotia (Kunze, E., Kretische Bronzereliefs, p. 121, fig. 12)Google Scholar. The idea of doing this sort of thing had, therefore, grown up by the period 1400–1200 B.C. Can there be some connection between the earlier pair of volutes and the much later double pair of horns so often bound in the same position to the sacred tree in Assyria, in the reign of Ashur-nasir-pal for instance (Wallis Budge, op. cit., Pls. xi, xlii bottom, xliii, xliv, xlv).

^{page 44 note 1} Strabo xii, i § 4 (tr. Loeb). He may well have been a pirate, for he was presumably living before Pompey suppressed these people in 67 B.C. Cicero was travelling in Asia Minor about 79 B.C., and twenty-five years later he speaks of the mutual hospitality and closest intimacy which he had with a certain Antipater of Derbe, when recommending him and his sons to Philippus, Q., proconsul of Asia, ad Fam., xiii, 73Google Scholar.

^{page 44 note 2} See p. 42, note 8.

^{page 44 note 3} Baedeker, , Konstantinopel und Kleinasien (1905), p. 166Google Scholar; Encyclop. Brit, (11th edn.), s.v. Frederick I, p. 46.

^{page 45 note 1} See further p. 46, note 6 for some information apropos of Greater Cappadocia and its bearing on the Septuagint rendering of the Hebrew Caphtor by Cappadocia.

^{page 45 note 2} Strabo, xiv, v § 6.

^{page 45 note 3} E.I., s.v. Karaman-Oghlu.

^{page 45 note 4} Strabo also tells us (xii, vi § 2; xiv, iii § 3) that Isaura was a robber stronghold. It is in the mountains close to the upper course of the Calycadnus River.

^{page 45 note 5} Sterrett, , The Wolfe Expedition to Asia Minor, p. 51Google Scholar (Papers of the American School of Classical Studies at Athens, iii).

^{page 45 note 6} id., op. cit., p. 96.

^{page 45 note 7} Hogarth, op. cit., p. 654.

^{page 45 note 8} Bent, in Proc. R. Geogr. Soc., 1890 (New Series xii), p. 445Google Scholar. Frazer, J. G., The Golden Bough, iv (Adonis, Attis, Osiris), 2nd edn., pp. 116ffGoogle Scholar, gives an account of Cilicia Tracheia and a bibliography.

^{page 46 note 1} Wainwright, in JEA, xvii (1931), pp. 31–38Google Scholar; id. in AJA, lvi (1952), pp. 200–203.

^{page 46 note 2} de G. Davies, N. in Bull. Metrop. Mus. of Art, Nov. 1929Google Scholar, The Egyptian Expedition, 1928–1929, pp. 38, 39Google Scholar.

^{page 46 note 3} Garstang, in LAAA., x (1923), pp. 21–26 and mapsGoogle Scholar; id., Index of Hittite Names (Brit. Sch. of Arch. in Jerusalem, Supplementary Papers, i, 1923), pp. 7, 8; Götze, , Kleinasien, pp. 53, 94, 168Google Scholar and map at the end; id., Kizzuwatna and the Problem of Hittite Geography, map at the end of the book.

^{page 46 note 4} Garstang, in Belleten, v (1941), p. 36Google Scholar; id. in AJA, xlvii (1943), p. 40 and map.

^{page 46 note 5} Garstang, , Index, p. 8Google Scholar; id. in AJA, xlvii (1943), p. 39, and for the text see Goetze, , Kizzuwatna, p. 22Google Scholar.

^{page 46 note 6} It will be remembered that the Septuagint equates Caphtor with Cappadocia. The time of the translators was that of Greater Cappadocia which included Dana-Tyana (Xenophon, , Anabasis, i, ii)Google Scholar and this was still one of the prefectures of that kingdom in the time of Archelaus and his predecessors (Strabo, xii, i, §4), presumably therefore from 350 B.C. onwards. Strabo also says that in his time (about the turn of our era) Cappadocia was bounded by Lycaonia and Cilicia Tracheia (xii, i, § 1). There was evidently good reason for the equation, Caphtor = Cappadocia.

^{page 47 note 1} Goetze, op. cit., p. 22. This was in the reign of Tudhaliyas III (p. 24), and therefore about 1400 B.C., and so in the time of Amenhotep III (Bossert, , Altanatolien, the Conspectus of dates inserted after p. 48)Google Scholar.

^{page 47 note 2} Knudtzon, J. A., Die el-Amarna-Tafeln, Letters nos. 31, 32. No. 31 is translated on pp. 271 ffGoogle Scholar, and no. 32 by Hrozny, in JA, ccxviii (1931), pp. 307–320Google Scholar.

^{page 47 note 3} Götze, , Kleinasien, pp. 119, 168 fGoogle Scholar, map at the end; Kizzuwatna, map at the end. Except Ḫapalla they are all somewhere off the map to the north, north-west and west, it being apparently agreed that the Šeḫa River is the Maeander on the north of Caria. Hitherto Ḫapalla has been grouped with Mira and Kuwaliya but now Garstang (Belleten, v (1941)Google Scholar, and p. 34, 36 and map; id. in AJA, xlvii (1943), p. 39 and map) finds reason for putting it on the coast of Pamphylia where Arzawa used to be put and for moving Arzawa westwards.

^{page 47 note 4} Ships sailed north from Egypt to Keftiu, and Caphtor is described as an ai, a word meaning “coast, region, border” and finally “islands” (Brown, Driver and Briggs, , A Hebrew and English Lexicon, pp. 15, 16)Google Scholar. The Kaptara trade, so well known at Mari on the middle Euphrates, would also have come by sea as far as the Syrian coast, and not by the difficult, and no doubt dangerous mountain country of the interior. Moreover, Furumark points out that Cretan influence is non-existent at Mari, hence the Kaptara trade could not have come from Crete. (Opuscula Archaeologica, vi (1950) pp. 216, 243.Google Scholar)

^{page 48 note 1} While of itself the Taurus forms no very serious barrier, it does divide two very different countries the one from the other. It is an obvious and natural division as has been recognised throughout the ages. Strabo (xiv, v § 1) speaks of Cilicia, both Tracheia and Pedias, as being “outside the Taurus”. On the other hand the Turks speak of the Cilician coast as Itshili “the Interior Country” as lying within or behind the mountains, Leake, W. M., Journal of a Tour in Asia Minor (1824), p. 99Google Scholar.

^{page 48 note 2} Garstang, in JNES, iii (1944), pp. 27, 36Google Scholar. It may be the country otherwise called Walwaras which Hattusil could not hold.

^{page 48 note 3} It was also the time that the Egyptians began to hear of other new peoples, such as the Danuna, Lukki, and Shirdani who like Arzawa first appear in the Tell el Amarna Letters.