Acknowledgements. I am most grateful to Mr Brian Cook, Keeper of Greek and Roman Antiquities, The British Museum, for permission to study the ivories from the tombs at Enkomi and to publish this article on the Enkomi head. I would also like to thank Mrs Helen Hughes-Brock and Dr Oliver Dickinson for reading earlier versions of this article and for making constructive suggestions. My thanks also go to Mrs Anna Grammenos, Mrs Vronwy Hankey, Dr Frank Stubbings, and Professor Lucia Vagnetti. Finally, I am indebted to Dr Doniert Evely for photographic assistance.

Abbreviations in addition to those in standard use:

Åström = Åström, L., The arts and crafts of Late Bronze Age Cyprus: Swedish Cyprus Expedition IV. 1D. (Lund 1972).Google Scholar

Caubet and Poplin = Caubet, A., Poplin, F., ‘Les objets de matière dure animale: étude du matériau, Ras Shamra-Ougarit III: Le centre de la ville [ed. Yon, M.] (Paris 1987) 273–306.Google Scholar

Excav in Cyp = Murray, A.S., Smith, A.H., Walters, H.B., Excavations in Cyprus (London 1900).Google Scholar

Helme = Borchhardt, J., Homerische Helme (Mainz 1980).Google Scholar

Kantor = Kantor, H.J., The Aegean and the Orient in the second millennium B.C. (Bloomington 1947)Google Scholar = AJA 51 (1947) 1–103.

Krzyszkowska = Krzyszkowska, O.H., ‘Ivory in the Aegean Bronze Age: elephant tusk or hippopotamus ivory?’ BSA 83 (1988) 211–34.Google Scholar

Poursat Cat = Poursat, J.-C., Catalogue des ivoires mycéniens du Musée National d'Athènes (Paris 1977).Google Scholar

Poursat IvMyc = Poursat, J.-C., Les ivoires mycéniens (Paris 1977)Google Scholar

Sakellariou = Xenaki-Sakellariou, A., Oi thalamotoi taphoi ton Mykinon (Paris 1985).Google Scholar

¹ Inventory number BM 1897.4–1. 1340 from Tomb 16.

² Excav in Cyp pl. II, 1340; p. 32. Also illustrated by Poulsen, F.JdI 26 (1911) 225 Abb. 9Google Scholar; Decamps de Mertzenfeld, C.Inventaire commenté des ivoires phéniciens et apparentés découverts dans le Proche-Orient (Paris 1954) no. 800Google Scholar; Helme Taf. 2, 4; Poursat IvMyc pl. XVI, 5; Varvarighos, A., To odontophrakton mykinaïkon kranos (Athens 1981) pl. XIIIγGoogle Scholar; Courtois, J.-C., Lagarce, J. and Lagarce, E., Enkomi et le Bronze Récent à Chypre (Nicosia 1986) pl. XXIV, 1.Google Scholar

³ Excav in Cyp 32.

⁴ Lorimer, H.L., Homer and the Monuments (London 1950) 215Google Scholar; Helme 75–76 cat. no. 17 I 1; Taf. 2, 4. Borchhardt interprets the warrior heads from Mycenae, Spata and Archanes similarly: see appendix for references.

⁵ Kantor 89; Åström 609. It has also been seen as part of a figurine: J.-C. Courtois et al., op. cit. (n. 2) 128.

⁶ See appendix for list and details. A new fragment has been found at Mitza Pùrdia di Decimoputzu in Sardinia: Ceruti, M.-L. Ferrarese, Vagnetti, L., Lo Schiavo, F. ‘Minoici, Micenei e Ciprioti in Sardegna alla luce della piü recenti scoperte’ in Balmuth, M.S. (ed.) Studies in Sardinian Archaeology III. Nuragic Sardinia and the Mycenaean World. BAR International Series 387 (1987) 7–37, esp. 12–14 fig. 2.3:1–2.Google Scholar I am most grateful to Vronwy Hankey for bringing this reference to my attention and to Lucia Vagnetti for sending me a photograph and further details. Although the fragment is very small, preserving only part of the helmet, there is no doubt that it belongs to the same type as the Enkomi and Aegean examples. See also infra nn. 7, 33 and appendix.

⁷ The Spata head differs: see appendix for details. As the heads from Phylaki Apokoronou and Archanes are not fully published, the appearance of their undersides is not entirely certain. However, the association of ‘backingplaques’ with those from Archanes makes the provision of mortises extremely probable. The new fragment from Mitza Pùrdia has a flat underside and preserves a circular mortise in the expected position (op.cit. n. 6 p. 12). Although Poursat recognised that the Enkomi head was an attachment (IvMyc 54) he did not take this into account when discussing its style and origin (ibid. 160); below p. 114.

⁸ Poursat Cat passim.

⁹ Poursat, IvMyc 55, pl. VII.3Google Scholar; Sakellarakis, I.A., ‘Elephantosta ek ton Archanon’, Atti e memorie del 1δ congresso internazionale di Micenologia (Roma 1967) 245–61, esp. 253–7; pl. V.Google Scholar

¹⁰ Cut-out rosette inlays have been found at Megiddo: Loud, G., The Megiddo Ivories (Chicago 1939) pl. 15 nos. 74–75, 77, 84–93.Google Scholar With such simple shapes it is impossible to judge whether they are imports or local copies of a Mycenaean motif. One piece, a rosette incised on a square piece of ivory, might be interpreted as an unfinished cut-out (pl. 15 no. 96).

¹¹ Enkomi BM inventory 1897. 4–1. 820. Åström (p. 551) following Gjerstad, E. [Studies on Prehistoric Cyprus (Uppsala 1926) 256, 288Google Scholar; VIII, 2] lists this as a ‘volute-shaped pendant’. Yon, M., Caubet, A., Le Sondage L–N 13. Kition-Bamboula III (Paris 1985) 83, 87 nos. 160–1Google Scholar; figs. 41, 44.

¹² de Santerre, H. Gallet, Tréheux, J., BCH 71–72 (1947–1948) 148–254, esp. 191–8 nos. 27, 28, 34CrossRefGoogle Scholar; pls. XXVII, XXXIV, XXXV. Poursat, IvMyc 152–8, esp. 158Google Scholar; pl. XV.

¹³ Caubet and Poplin passim esp. 273–78; Krzyszkowska 211–15; eadem, Ivory and Related Materials: An Illustrated Guide (London 1990) 38–47, 84–91.

¹⁴ The disposition of the commissure varies throughout the tusk and from one tusk to another (i.e. mature or juvenile, right tusk or left). In Fig. 2 the features of the tusk are rendered conventionally, and do not, therefore, correspond precisely to the disposition of such features on the head. This would be almost impossible to achieve given the condition of the head and the variables in tusk growth.

¹⁵ According to Poursat (IvMyc 55) the head alone is represented ‘sur une base renflée qui ne peut être comprise ni comme le cou ni comme la poitrine’. I am not altogether happy with this view and, for convenience, here will retain the designation ‘neck’.

¹⁶ For Ugarit see Caubet and Poplin passim. Among the other ivories from the Enkomi tombs in the British Museum a high proportion are made from hippopotamus ivory. They will be the subject of a future article. For the French tombs see: Caubet, A. et al. ‘Enkomi (Fouilles Schaeffer 1934–66): Inventaire Complémentaire’ RDAC (1987) 23–48Google Scholar; bone/ivory 23–36.

¹⁷ Enkomi BM 1897.4–1. 1339: Excav in Cyp pl. II and infa n. 29. Kition-Bamboula: Yon and Caubet op.cit. (n. 11) 83–85 no. 151, fig. 42.

¹⁸ Krzyszkowska 228–33, esp. 232–3.

¹⁹ Ibid., 224–6, Fig. 5, Plates 29, 30.

²⁰ Korres, G., ‘To kranos tou polemistou ton Spaton AE (1966) 119–26Google Scholar; 124, fig. 3 (underside). When I last considered the use of hippopotamus ivory for warrior heads (supra n. 19) I was unaware of both this illustration and that cited in n. 21. Consequently at that time I was unable to reach any firm decision regarding the material used for the Spata and Archanes heads.

²¹ These features are clearly visible in the colour plate of the leftward facing head in Sakellarakis, I.A., Guide to the Herakleion Museum (Ekdotike Athinon: 1985) p. 56Google Scholar. There is no reason to believe that the rightward facing head is made of a different material. Colour photography is most beneficial in distinguishing ivory types and identifications can be made with considerable confidence.

²² AR (1981–2) 58 fig. 128; BCH 106 (1982) 628, 630 fig. 182. ‘Heads’ are mentioned, although only one is illustrated.

²³ See appendix for complete list of identifications. With ingenuity a carver might manage to extract a pair of heads and their ‘backing-plaques’ from a mature lower canine. The plaques would need to be made in pieces and joined (see Krzyszkowska 224–5, Fig. 5a–b). But this solution is not economical in material or in labour. Nevertheless I can no longer believe that the use of hippopotamus tusk for warrior heads reflects some kind of temporary shortfall in elephant ivory: it is difficult to argue that all examples were made at the same time, much less the same place. See appendix for contexts and dating.

²⁴ It is virtually impossible to compensate for the loss of the underside when positioning the head for photography (or drawing). This accounts for the differences in the views on Fig. 1, for instance, between the positions of the eye and nose (Fig. 1a–b) or the width of the helmet (Fig. 1 a, c).

²⁵ It seems likely that some of the discolouration and surface damage was caused by post-excavation treatment and not by deposition. Unfortunately no details of conservation exist for this or indeed for many other items from Enkomi. Hot-wax was, however, a common consolidant and its use can mimic the effects of high temperatures and burning in antiquity. See Krzyszkowska Guide (op. cit 13) 8, 36–37.

²⁶ Demargne, P., La Crète dédalique (Paris 1947) 196.Google Scholar

²⁷ Kantor 89.

²⁸ Kantor, H.JNES 15 (1956) 171 n. 53.Google Scholar

²⁹ Åström 608–9. For the mirror handle: ibid. 612 and Kantor 97 ‘an uncontestable product of a mainland workshop’. Also infra n. 42.

³⁰ Courtois et al., op.cit. (n. 2) 128.

³¹ Catling, H.W., Cypriot Bronzework in the Mycenaean World (Oxford 1964) 161 n. 1.Google Scholar Borchhardt (Helme 75) apparently also prefers to regard the head as the survival of an old motif.

³² Poursat IvMyc 158; ibid. 54 for identification as an attachment.

³³ As on the new example from Mitza Pùrdia where, according to Ferrarese Ceruti, the tusks are not disposed in alternating rows ‘ma sono rigorosomente orientate dalla stessa parte’ (op.cit. n. 6 p. 14). However, there is actually more variation in the Aegean examples than Ferrarese Ceruti realises. On the Sardinian example the height of the two preserved registers (ca. 1.5 cm) demands a rather cursory treatment. Moreover the double relief ridges also serve to limit space, as on the Spata and Enkomi heads.

³⁴ Poursat IvMyc 55–57.

³⁵ This is Poursat's interpretation: ibid 56.

³⁶ Ibid 57. The eye on the Enkomi head looks rather low-set: this is chiefly due to distortion (supra n. 24).

³⁷ Ibid. Traces of the nostril are preserved on the Enkomi head and help to establish the size and shape of the nose (Fig. 1b).

³⁸ Sandars, N.K., The Sea Peoples (London 1985) figs. 18, 131Google Scholar; Excav in Cyp. pl. I fig. 19 (BM 1897. 4–1. 996). Note that many illustrations, including Sandars', show the gaming board with feet. These were additions which have now been removed.

³⁹ Excav in Cyp pl. II nos. 872, 883. Sandars sees the kilt and body armour on the handle as similar to those worn by the Sea Peoples (op.cit. n. 38 fig. 97). The ‘pyxis’ fragment (no. 883) will be discussed in a future article in this journal.

⁴⁰ Poursat, IvMyc 157 pl. XIV.Google Scholar He has effectively dismissed the interpretation put forth by Kantor and others that this plaque is a Mycenaean import. Rather, the facial features and proportions of the body, justly compared to those of the bronze horned god from Enkomi, point to a Cypriot origin.

⁴¹ The Aegean manufacture of duck pyxides (an eastern type) from hippopotamus lower canines (the correct material) is equally implausible. The bodies from Zapher Papoura and Mycenae (considered ‘boats’ by Sakellarakis, : AE [1971] 188–233Google Scholar), a small neck from Mycenae (Tiryns apotheke, Nauplia 1090), and the fine head from Asine (Frodin, O. & Persson, A.Asine [1938] 388Google Scholar fig. 254) are all surely imports. See Krzyszkowska 233–4.

⁴² pace Kantor (supra n. 29) on the mirror handle. She also sees the Saqqara lid (in wood) as another LH III import to the east (ibid.) See also Poursat, IvMyc 141–51 esp. 147–8.Google Scholar

⁴³ Even allowing for the vagaries of deposition and retrieval, I would have expected other elements from the original composition (e.g. lilies, rosettes, tricurved arches and, above all, figure-of-eight shields) to have been found, had that composition remained intact up to the time of burial.