¹ My manuscript was substantially complete when Benson, J. L.'s extremely interesting study Horse, Bird and Man (Amherst, 1970)Google Scholar appeared. The thesis there presented, that the Geometric artist knew and drew upon the traditions of both Mycenaean and Egyptian art, and that his visual world was as formulaic, distant, and heroic as that of the contemporary epic bard, is evidently incompatible with many of my own views. There is much of value and much to provoke thought in Benson's book, and I may have underestimated the possibility of some kind of continuity in figure-drawing traditions from Mycenaean to Geometric times. On the other hand, he perhaps goes too far in denying powers of iconographic invention to Geometric artists. But rather than indulge in detailed controversy, I have thought it best to present my argument in its original form. Iudicet lector.

² The organic interpretation fits Greek art much better than it does Roman, which is perhaps why its relevance to the former has been far less often questioned. For the history of interpretations of Roman art, see Brendel, O. J.'s fascinating essay ‘Prolegomena to a book on Roman art’, MAAR xxi (1953) 1–73Google Scholar.

³ London and New York, 1960 (Chapter iv, ‘Reflections on the Greek Revolution’).

⁴ See the symposium ’Narration in ancient art’, AJA lxi (1957) 43–91Google Scholar.

⁵ On the question of the identification of the bull-man of Mesopotamian glyptic as Enkidu, see the cautious discussion of Frankfort, H., Cylinder Seals (London, 1939) 62–7Google Scholar. If the bull-man is Enkidu, as seems possible, the fact remains that he is being portrayed in characteristic rather than specific activity; and it may be, as Frankfort suggests, that existing Enkidu-images generated the descriptions in the poem, rather than vice versa. Gilgamesh himself is not certainly identifiable, unless one allows that depicted with a bull-man may, by association, be taken as Gilgamesh. The one possibly specific incident is Frankfort's pl. xi m, but there are no inscriptions and no other certain representations in other media to clinch the identification; but cf. the statuette mentioned at the end of the next note.

⁶ The rare mythological representations in Egyptian art do not depend on any particular narrative source, and owe their appearance to their religious or symbolic importance (Kantor, in AJA lxi (1957) 44CrossRefGoogle Scholar). It is Mesopotamian glyptic which provides the richest stock of illustrations of mythology; in particular, of episodes from the Babylonian Epic of Creation whose hero is the sun-god Marduk (see Langdon, , The Babylonian Epic of Creation, Oxford, 1923Google Scholar). But this epic is hardly in the same category as the Epic of Gilgamesh, let alone the Iliad; it is rather a religious text, and its performance constituted part of the Babylonian New Year festival, the chief event of the religious year (Frankfort, op. cit. n. 5, 97). Illustrations of other myths (e.g. the myth of Zu, Frankfort 132 ff.) may likewise be seen as expressions of religious belief; though the seal apparently portraying the myth of Etana, a man who assisted an eagle and was carried up to heaven to obtain the plant of birth, is clearly a narrative illustration of a story told for its own sake. The Akkadian seal-cutters, one may conclude, definitely portrayed on their seals specific episodes from the myths of their culture; these myths concerned chiefly the gods, and hence the content of the scenes is in a sense religious. But the narrative interest is present and important. None the less, these illustrations of myth and epic become very rare after the Akkadian period (AJA lxi (1957) 60Google Scholar), and it must be very doubtful whether the occasional occurrence of similar scenes in Assyrian glyptic represents a living tradition (see Frankfort 198–9). It is impossible, in fact, to believe that Akkadian artistic practice influenced in any way the artists of the Geometric period in Greece some one and a half millennia later. This is not to claim that art did not make use of mythological material; there is, for instance, a statuette in the Louvre (to which Dr. Barnett drew my attention) showing a Gilgamesh-type figure standing on a Humbaba-type head. But this, like the Egyptian references, denotes rather the permanent state of victory than any interest in the action of the story.

⁷ For terminology and chronology I use the system of Coldstream (GGP 330). The Attic sequence, which is by far the most important runs as follows:

⁸ The question of the origin of these motifs and the reason for their use is both beyond the scope of this study and unrewarding to discuss. There are two basic viewpoints, either of which may be adopted, according to taste:

(i) that some or all of the motifs, linear and representational, are symbolic—a view which finds its extreme manifestation in Roes, Anna's Greek Geometric Art, its Symbolism and its Origin (Haarlem & London, 1933)Google Scholar. More moderate is Kraiker, W., ‘Die Anfänge der Bildkunst in der attischen Malerei des 8. Jhr. v. Chr.’, BJ clxi (1961) 108–20Google Scholar, who proposes (p. 111) that the very early horses have a chthonic symbolism, being found only on funerary vessels; this is not strictly true, since Agora P 1654 (EG I, see n. 14 below) and P 6422 (a MG II krater, Davison fig. 145) are not from graves, and it should also be remembered that the great majority of preserved pots of this epoch come from graves in any case. None the less, the possible allusion to Poseidon Hippios remains.

(ii) that the urge to draw simple figures of animals and people is common to mankind and requires no explanation (cf. pp. 38–9).

⁹ See BMQ xxiii (1961) 105Google Scholar.

¹⁰ DeVries, K., ‘Hesiodic pictures: the Greek incised fibulae’, AJA lxxiv (1970) 192Google Scholar (summary). Fittschen (213–21) dates the bulk of the material around or after 700 B.C.; while Hampe, R., Frühgriechische Sagenbilder in Böotien (Athens, 1936)Google Scholar dates the material considerably earlier.

¹¹ Protogeometric horses: amphorae, Kerameikos iv pl. 27, inv. 560, 911, 1260 (this last a sherd only); Schweitzer pl. 23; Zervos pls. 16–17. For a L.H. IIIC example, Iacovidis, S.Perati ii. 151Google Scholar fig. 24.

¹² Desborough, V. R. d'A., Protogeometric Pottery (Oxford 1952) 42Google Scholar.

¹³ NM 18045, GGP pl. 1k.

¹⁴ Agora P 1654, Hesperia ii (1933) 560Google Scholar fig. 19 no. 62.

¹⁵ Ker pl. 22, inv. 1254; Davison fig. 141; AK iv (1961)Google Scholar pl. 19.6.

¹⁶ Krater Ker pl. 20, inv. 290, grave 22 (for the dating to the second quarter of the eighth century, see AA 1963, 205Google Scholar n. 27), Schweitzer pl. 15; Krater Agora P6422, Davison fig. 145; Pyxis Louvre A 514, AHS pl. 3, GGP pl. 4 e, g, Schweitzer pl. 25, Zervos pl. 47. Kübler, (Ker 181)Google Scholar names the MG II krater from Eleusis (AE 1898, pl. 3.3) as having a ‘Pferdebild’, but I have not seen the pot and the horse is quite invisible in the photograph.

¹⁷ Ker pl. 111, inv. 2159, isolated find; Schweitzer pl. 26 (detail). I know of one other pot of this period which bears a horse: a Cretan amphora from Knossos (GGP pl. 52c) which is very strongly Atticizing, and perhaps not quite so early as Coldstream, who gives it to Cretan EG (810–790), would make it. It shows a rather rubbery-looking animal tethered to a manger, and one might perhaps be justified in inferring from it the existence of this subject in the Attic MG II repertoire. Ahlberg, (PE 211)Google Scholar sees possible Near Eastern influence in the motif of two antithetic horses with a man between them, but her cited parallels are sculptural and date from near or after 700 B.C.

¹⁸ Argive(?) export: krater frr. Samaria-Sebaste iii. 211Google Scholar fig. 34. 1–2. Cretan amphora: see n. 17. Argive are the pyxis from Makris grave 1, bearing two horses (?) and a man in a separate panel (GGP 124); and the kantharos Argos C₃₃, with five horses on one side and five deer on the other (CGA pls. 60, 139). Walter, H. (Samos v, Frühe griechische Gefässe, Bonn, 1968)Google Scholar publishes as ninth-century a krater-fragment (no. 25, pl. 6) bearing the head and neck of a horse at a manger. The fabric, which one cannot judge from a photograph, may of course provide an overridingargument for this dating; but the subject, virtually restricted to the Cycladic, Euboean, and Boeotian schools andto the LG period, and the confident and mannered stylization of the horse would naturally indicate an LG date. The hatching of the neck is hardly in itself sufficient reason for placing this fragment in EG, as Walter does (p. 18); note that the terracotta votive horses from the Heraion are most commonly hatched (AM lxv (1940) 101Google Scholar n. 2).

¹⁹ Ker pl. 142, inv. 642, and two without inv. nos., pls. 142, 143 (top); Schweitzer pl. 24, Zervos pls. 133, 135, 136. No. 642 is without doubt a toy since its legs are bored to take an axle. Unfortunately all three are isolated finds; but the only others of the Geometric period from this cemetery (pl. 143, inv. 1312, 1313) are from a child's grave (no. 50) containing other toys: a hen (pl. 144, inv. 1309), a donkey (pl. 144, inv. 1311), a basket (pl. 118, inv. 1307), and a miniature amphora (pl. 110, inv. 1306; cf. GGP 38 n. 1). (For a rather different interpretation of the contents of this grave, see Ker 42.) Thus of the five examples from the Kerameikos three are certainly toys; there is also a PG example, complete with wheels, from a child's grave in Amphiktyon St., Athens (ADelt xxii (1967Google Scholar) Chr 49 pl. 70, 1). The presumption is therefore strong that the remaining two from the Kerameikos are also toys; on the other hand one was found in grave α at Eleusis, a woman's grave. Cf. also the terracotta donkeys or mules carrying jars (n. 48 below).

²⁰ Agora P 19240, from grave D 16:2 (GGP 11, pl. 1g; Hesperia xviii (1949) 284, 290Google Scholar, pl. 67 no. 3).

²¹ Ker pl. 54, inv. 257, GGP pl. 4b, Zervos pl. 48. Also early are the lid Ker pl. 66 neg. 5706, minus its animal but with a vestigial knob; and the horse in Heidelberg (CVA iii, pl. 103) minus its lid and with a strange ornament on its side resembling a Dipylon shield with rectangular instead of semicircular notches. Possibly contemporary with these is the pyxis illustrated by Matz, F., Geschichte der griechischen Kunst i (Frankfurt, 1950)Google Scholar pl. 26: the horse has a vertical chain of three cross-hatched lozenges on each shoulder, and no collar, which suggests an early, experimental stage of development. On the other hand there is no vestigial knob, and the chief ornament around the edge of the lid is the dotted lozenge which first becomes common in LG I (Matz gives this pot as from the Kerameikos; Kübler, Ker 179 n. 176, denies this, but does not give the true origin or present ocation). On horse-pyxides in general see Bouzek, J., Sbornik (1959) 131–4Google Scholar.

²² e.g. Ker pl. 67, inv. 1310, grave 50.

²³ Argive: Nauplia Museum no. 4068 (or is this Attic?).

²⁴ e.g. Vienna, Kunsthistorisches Museum, inv. 3453.

²⁵ Samos, Heraion: Ohly, , AM lxv (1940) 90 ff.Google Scholar, pls. 57–9 (cattle commoner earlier, horses later).

Perachora: Payne, , Perachora i. 229Google Scholar.

Argive Heraion: Waldstein, , The Argive Heraeum iiGoogle Scholar pl. 48.

Olympia: Furtwängler, , Olympia ivGoogle Scholar pl. 17.

Sparta: Dawkins, Artemis Orthia pl. 41.

Delphi: apparently none, unless they were classed with the Mycenaean material (Fouilles de Delphes v. 15 no. 7).

²⁶ The best and most recent study is by Herrmann, H.-V. ‘Werkstätten geometrischer Bronzeplastik’, JdI lxxix (1964) 17–71Google Scholar, with full references and documentation. This has at last put the study of Geometric sculpture on a solid methodological foundation (see especially his introductory remarks, pp. 18–20); but in the absence of any firmly dated examples to serve as anchor-points it is difficult to see what real progress can be made. On the general question of a ‘long’ or ‘short’ chronology for the material from Olympia see Hampe, R.Gymnasium lxxii (1965) 76–81Google Scholar and Kunze, E.Olympiabericht viii (1967) 219 ff.Google Scholar, with a judicious evaluation of the controversy by Snodgrass, A. M. in Gnomon xl (1968) 390–1Google Scholar.

²⁷ Willemsen, F., Olympische Forschungen iii, Dreifusskessel von Olympia (Berlin, 1957)Google Scholar, gives a complete catalogue and discussion of all examples known up to 1954. He favours a ‘long’ chronology and makes the preserved series start in the eleventh century. Amandry (Gnomon xxxii (1960) 459–63Google Scholar), noting the scarcity of PG and EG pottery, and indeed of votives of any sort, before the eighth century at the two chief centres, Delphi and Olympia, is sceptical. Note, however, the two clay models from the Kerameikos (PG tomb 4) of the tenth century and the new evidence from Lefkandi—a tripod-mould fragment in a tenth-century context. Tripod-cauldrons must have been made, even if they were neither dedicated nor decorated, throughout the Submycenaean and PG periods. The question is, whether any of the extant examples go back so far.

²⁸ Olympia B 5633 (Olympiabericht viii 218–19Google Scholar, pl. 107.4), probably third quarter of the eighth century, discovered since Willemsen (op. cit. n. 27) wrote.

²⁹ The following statistics are drawn from Willemsen (op. cit. n. 37) and may be of some interest (pieces from Olympia are listed by number alone; the others may easily be tracked down through Willemsen's index):

Early, complex-section, handles

7 Bull-protomes (Br 5449, 8721, 9229, 9398, B 1868; Delphi 1946(?), 2998);

2 Birds (Br 1047, 7872);

2 Horses (Br 5471, Ithaca 6b);

In addition, 1 Horse (B 2040) and 1 Bull (Ithaca 6b) stand inside the handle, on the strap at the base; there is possibly another bull inside Delphi 1946 (BSA xxxv (1934/1935) 100Google Scholar).

Later, cast, handles (openwork, ribbed, and stepped, of near-flat section)

1 Bull-protome (Br 13504, 13518);

1 Lion (Br 11340);

1 Bird (B 1946);

12 Horses (Br 913 (tail only preserved), 2991, 8063, 9404, B 1269, 2406, Athens NM 7482, two unnumbered (Willemsen, op. cit. pls. 66, 67), Delphi 2956, Ithaca 7b, 9b);

2 Horse-and-man groups (B 4567 (Olympiabericht viii pl. 59), Delphi 4393 (inferred));

2 Doubtful (Br 9127, B 2175).

Late, hammered, handles (flat)

2 Horses (Br 9694, Athens NM 7483);

1 Horse-and-man group (B 2041).

These lists include only those instances where the connection of handle and figure is certain. I have not counted, for example, separated figures whose mounting shows them to have been fixed to a particular type of handle (in the vast majority of cases, the hammered variety). Nor have I included the Cretan material.

³⁰ See BSA xxxv (1934/1935) 95Google Scholar nos. 1–2 with pl. 11a–c. See also n. 44 infra.

³¹ The hoard: AM lv (1930) 119–40Google Scholar.

³² Tiryns tripod: Catling, H. W., Cypriot Bronzework in the Mycenaean World (Oxford, 1964) 195Google Scholar no. 10 pl. 28b. Other bull-protomes adorn a rod-tripod, Catling 194 no. 9 pl. 28c–e (cf. the clay bracket in Paris BCH xxxi (1907) 238Google Scholar and pl. 22) and a torch-holder, Catling 162 pl. 25 b, c. The Perati cemetery has yielded two gold ear-pendants of Cypriot design, stamped with bulls' heads (Iacovidis, S., Perati ii. 454Google Scholar pls. 15 a, c).

³³ Catling (op. cit. n. 32) 220 f. There is even a pair of horizontal handles with a centrally-mounted inward-facing bull-protome of L.M. IIIB date from Mouliana Tomb A (AE 1904, 30 f.Google Scholar, fig. 7).

³⁴ The numerous clay imitations testify to their popularity, as does the example found (as an antique) with LG pottery on the Pnyx. See Catling (op. cit. n. 32) 194 no. 6 and 215 f.

³⁵ See the publications listed in n. 25.

³⁶ CVA Athens iGoogle Scholar, IIIHd pl. 1.9.

³⁷ Oakeshott, N. R., ‘Horned-head vase handles’, JHS lxxxvi (1966) 114–32CrossRefGoogle Scholar, believes that the double-arched handles of the PG and Geometric periods are a stylized continuation of the horned-head handles of the Warrior Vase from Mycenae. Although a second L.H. IIIC example has now come to light (Popham, M. R. & Sackett, L. H.Excavations at Lefkandi, Euboea, 1964–1966 (London, 1968) 20Google Scholar fig. 38), it is still impossible to tell whether this was normal practice in the period or a flight of individual fancy. What does seem clear is that the potters and painters of Geometric Greece did not usually (pace Oakeshott) treat these double handles zoomorphically, as the rare examples where such treatment is unmistakable show. Double handles can scarcely help suggesting a pair of horns, and one hardly needs to postulate a formalized Mycenaean tradition to explain the appearance of horned-head handles on Cypriot pottery c. 700 B.C. More remarkable, if Oakeshott is right, is the restraint of artists in the intervening period.

³⁸ GGP 69 n. 1.

³⁹ Higgins, GT 12, 16, 21Google Scholar pls. 4D, 5A, 5C; Nicholls, R. V.BSA lxv (1970) 26 ff.Google Scholar and n. 18; id. ‘Greek Votive Statuettes and Religious Continuity’, Auckland Classical Essays presented to E. M. Blaiklock (1971) 1 ff.

⁴⁰ Attic: London MsC 2532, Athens Agora P 22433, Eleusis 820 (all oenochoai with antithetical birds in a metope—for references see GGP 26); Kerameikos, skyphos from grave hS 230 (AM lxxxi (1966)Google Scholar Beil. 65). Argive: Nauplia 4253 (GGP pl. 25h); Argos C 840 (CGA pl. 77). Cretan: (i) Protogeometric B: Fortetsa no. 354; BSA lv (1960) 130Google Scholar pl. 31 (?PG); Hall, E. H., Vrokastro1 (Philadelphia, 1914)fig. 52a; on PGB see Fortetsa 143, 213 ff., GGP 235 ff.Google Scholar, BSA lxii (1967) 64 ff. (ii) MG: Fortetsa nos. 422, 596, 841.Google Scholar

⁴¹ There are no terracotta birds of the Geometric period at Delos, Perachora, or Samos. The bronzes have been comprehensively treated by Bouzek, J., ‘Die griechischgeometrischen Bronzevögel’, Eirene vi (1967) 115–39Google Scholar.

⁴² Olympia Br 1047, 7872 (a pair), B 1946.

⁴³ According to the Encyclopaedia Britannica ¹¹, s.v. BOVTDAE, ‘antelope’ may be used to describe a number of animals, some of which are deer-like in general appearance; these in fact belong to the family of Bovidae, which is distinct from that of Cervidae to which true deer belong.

⁴⁴ Fortetsa no. 45, pls. 4, 135; BSA lvi (1961)Google Scholar pl. 11. v (from Knossos); the Mouliana krater (AE 1904 pl. 3) is earlier. The goat appears to have played an important part in Cretan cult: see Dietrich, B. C. in Historia xvi (1967) 403Google Scholar.

⁴⁵ Pyxis Louvre A 514 (GGP pl. 4f)—the deer here perhaps based on the schema for horse?; oenochoe Ker 8796 from Eridanos tomb hS 55 (AM lxxxi (1966)Google Scholar Beil. 15. 2–3). The oenochoe Eleusis 820 (AE 1898 pl. 5. 2), where the decoration has spread on to the neck of the vessel instead of being restricted to the ‘bib’, is likely to be slightly later. This pot, and the banded amphora with antithetical deer in neck-panels found in the same grave, may be by the same hand. The grave was an adult cremation next to the Isis Grave.

⁴⁶ Argos C33 (see n. 18).

⁴⁷ Kerameikos iv pl. 26, inv. 641, from grave 39. This has a hollow wheel-made body and seems to be an epigone of the L.M. IIIC bull-statuettes in the same technique (Zervos pl. 137, Higgins GT pl. 7D).

⁴⁸ For the record, I note the following relative rarities from the six sites listed in n. 25: sheep (common at Delphi, rare elsewhere), several dogs, two tortoises, four pigs, two hares, a goat, and an ape. Doubtless there are others from sites less conveniently published, but they do not affect the argument. Many of these must be LG, if not later. Also worthy of separate mention are the wheel-made terracotta donkeys or mules carrying jars, probably toys, from Attica and Eretria (BSA lii (1957) 15Google Scholar pl. 3; Ker pl. 144; Higgins GT pl. 8a; and Agora nos. T 1725, MC 385, and MC 685).

⁴⁹ GGP pl. 4h.

⁵⁰ Olympia Br 11340 (Willemsen, op. cit. n. 27, pl. 54; BSA xxxv (1934/1935)Google Scholar pl. 24. 4): those who believe in Mycenaean survivals may like to identify this ‘lion’ as a dog, noting its similarity to the top animal on side A of the wooden chest from Shaft Grave V (AM lv (1930)Google Scholar Beil. 29). I cannot follow Miss Benton in seeing a lion inside a tripodleg from Olympia (BSA xxxv (1934/1935) 95–6Google Scholar no. 5, fig. 9, pl. 22.2). This, and one or two other ‘animals’ in this position, are surely the bent and corroded remains of projections such as can be clearly seen on her pls. 11c and 12c.

⁵¹ London BM 1960.11.1.44, BMQ xxiii (1961) 105–6Google Scholar, pl. 46; BSA lxiv (1969) 147Google Scholar pl. 36; Hampe, R., Frühe griechische Sagenbilder (Athens, 1936)Google Scholar pl. 7 (reversed). For the dating, the similar fibulae from a MG II grave at Anavysos are crucial (ADelt xxi (1966)Google ScholarChr 97–8).

⁵² Ker pl. 159, 160, inv. M 47, 48. The grave is transitional to MG I.

⁵³ Fogg Art Museum Acquisitions 1965, 65Google Scholar. I exclude the ‘lyre-player’ class of seals as non-Greek (JdI lxxxi (1966) 1–63Google Scholar).

⁵⁴ Agora viii 16 n. 33Google Scholar.

⁵⁵ Ker pls. 159–61, inv. M. 48; BM 1960.11.1.45 (a pair to the fibula of n. 51).

⁵⁶ BSA xliv (1949) 95Google Scholar.

⁵⁷ One-handled cup, AJA xliv (1940)Google Scholar pl. 21.6, and hydriske, ibid. pl. 22.1, both reputed to be from the same grave; the cup has two separate ships in panels, without masts and with a star as filling-ornament above, while the hydriske has at least two ships in procession round the shoulder, with sails furled on the yard (not ‘geblähten’, as Kahane describes them, ibid. 472) and a steering-oar. Also a Cretan pithos from Khaniale Tekke, given to LG by the publication (BSA lxii (1967) 73Google Scholar fig. 2) but conceivably late MG (imitating Attic?); and an Attic (?) MG oenochoe (ADelt xvii (1961/1962)Google Scholar pl. 55a; GGP 95 note).

⁵⁸ Bibliography to 1956: JdI lxxiv (1959) 69Google Scholar n. 54. Further: Marwitz, H., AK iv (1961) 47Google Scholar n. 45, pl. 19.7; Davison 106–7; Himmelmann-Wildschütz, N., Gnomon xxxiv (1962) 79Google Scholar; Herrmann, H.-V., JdI lxxix (1964) 69Google Scholar n. 5; Kunze, E., Olympiabericht viii, 219Google Scholar n. 16; GGP 26, 28; Schweitzer 37. The skyphos was too often dated, in earlier discussions, without taking into consideration the other pots from the same grave-group. All are unequivocally MG; the two skyphoi 740 and 743 (AJA xliv (1940)Google Scholar pl. 21.2, 3) are very near to those from Agora Grave 12:17 (Agora viii nos. 261–2, GGP pl. 9 b, c) which Coldstream, though he classes them as LG Ia, says could well be MG II surviving in an LG Ia context. Arguments from the figure style involve a priori assumptions and are worthless until some consistent pattern of development has been established on other grounds.

⁵⁹ The whole tale of Odysseus' wanderings, which are themselves only an expanded mise en scène, occupies less space than the stretch of fighting in the Iliad which culminates in the death of Patroclus.

⁶⁰ For a full discussion, see Brunnsåker, S. in OpRom iv (1962) 187–94Google Scholar (apropos the Pithecusa ‘shipwreck’ krater). Cf. also Marwitz, H. in AuA x (1961) 13 f.Google Scholar

⁶¹ I have counted as MG II only two of the five pots listed in GGP 26. I see no reason to date either the Altheim skyphos (on which see also n. 192 below) or the Athens and Toronto stand-fragments earlier than LG. The fifth pot is the New York krater 34.11.2; because of its individuality it is impossible to date this accurately on stylistic grounds (see AHS 268) and I should prefer to fit it into the general line of development sketched in section III below and class it as the work of an individualist who took his ideas but not his style from the two great workshops of the day, those associated with the Dipylon and Hirschfeld Painters.

⁶² Fortetsa no. 1440, pls. 77, 163.

⁶³ Ibid. no. 339, pls. 24, 144.

⁶⁴ It would be difficult to down-date PGB, making it coexist with, rather than precede, Cretan EG and MG (see BSA lxii (1967) 64 ff.Google Scholar). There are pure PGB groups in tombs, uncontaminated by EG or MG. But even if one made the attempt, 775 would be as late as the end of PGB could possibly go.

⁶⁵ GGP 21. Ahlberg (PE 261) comments on the iconographic link, provided by the gesture of mourning, between Mycenaean and Geometric art. Benson (op. cit. n. 1) 94 suggests as well a possible connection with Egyptian art. I would prefer to think that the gesture remained the same in real life, and that the Geometric artist, depicting the practice of his own day and dependent on no artistic tradition, inevitably produced a schema similar to that of his Mycenaean predecessors and Egyptian contemporaries.

⁶⁶ Olympia iv pl. 15.236, 238; pl. 16.240, 241, 246.

⁶⁷ I take all the bronzes of a more complicated nature showing figures e.g. sitting, driving chariots, playing an instrument, or dancing, to be LG at the earliest. The ringdance (Olympia iv pl. 16.263; Lamb, W., Greek and Roman Bronzes (London, 1929)Google Scholar pl. 16b; F. Matz, op. cit. n. 21, 84, pl. 24b; Zervos pl. 158; for other examples see Corinth xv. 2, 42Google Scholar and BCH lxxviii (1954) 181Google Scholar fig. 41) indeed occurs in Mycenaean art, but there is no reason to suppose a connection or date any of the Geometric examples early. They are best explained as a plastic version of a subject which became very popular on LG pots (cf. Tölle, R., Frühgriechische Reigentänze (diss. Waldsassen 1964)Google Scholar.

⁶⁸ JdI lxxix (1964) 45Google Scholar figs. 28–30.

⁶⁹ GGP 29 ff. (with further refs.) and pl. 6.

⁷⁰ Hampe, R., Die Gleichnisse Homers (Munich, 1952)Google Scholar pl. 3.

⁷¹ Marwitz, H., ‘Das Bahrtuch’, AuA x (1961) 7–18Google Scholar.

⁷² Cf. Himmelmann-Wildschütz, N., Erzählung und Figur in der archaischen Kunst (Akad. Wiss. u. Lit. Mainz, Abh. der geistes- und sozialw. Kl. 1967/1962) 84 f., 87–92Google Scholar; Ahlberg, , PE 285 f.Google Scholar, believes that certain of the details show the protheseis to be individual, not typical scenes: ‘individuality in the iconographic expression’ corresponds to ‘individuality in the iconological significance’.

⁷³ Ekphorai: Athens NM 803, by the Dipylon Master (AuA x (1961)Google Scholar pl. iii, Ahlberg PE fig. 53); krater Athens NM 990, by the Hirschfeld Painter (AHS pl. 5, Davison fig. 25, Ahlberg PE fig. 54); krater fr. Bonn 16, by the same (AA 1967, 170Google Scholar figs. 1–2, Ahlberg PE fig. 55). No others are known.

⁷⁴ Numerous examples: see GGP 30 ff., 41 f.

⁷⁵ The frequency of ship-fights in art surely does not mean that this was the commonest form of encounter in this period, nor that the dead man honoured by a shipkrater was a sea-captain. Ships played an important part in early Greek warfare, largely if not entirely as a means of transport for fighting men, so it is scarcely surprising that both the Iliad and the pot-painters give considerable attention to fighting round ships. I cannot share Kirk's belief (BSA xliv (1949) 144–53Google Scholar) that the ship was popular simply because it was an intrinsically geometric and naturally attractive subject for the artist. It is noteworthy that so far no Attic figured kraters have been found outside Attica, and only one outside Athens: a fr. from Thorikos with a condensed prothesis-scene of the type found on the New York krater 34.11.2 (Thorikos iii (1965) 43Google Scholar fig. 49). This could date to MG II. There is also the pedestal of a huge krater in the Eleusis Museum.

⁷⁶ Contra: Chamoux, F.RA xxxiii (1945) 84 f.Google ScholarAhlberg, (PE 285 ff.)Google Scholar believes that the variations occurring within the ‘formulaic language’ are too many and too important to be explained away as merely stylistic differences and that the scenes have particular and specific reference to the life (and death?) of the deceased. In spite of her detailed analysis I feel that this remains a matter of opinion.

⁷⁷ IVORIES:

Mallowan, M. E. L., Nimrud and its Remains (London, 1966) iiGoogle Scholar; Barnett, R. D., The Nimrud Ivories (London, 1957)Google Scholar; Crowfoot, J. & Crowfoot, G., Samaria-Sebaste ii, The Early Ivories (London, 1928)Google Scholar; Thureau-Dangin, F. et al. , Arslan-Tash (Paris, 1931)Google Scholar.

POTTERY:

Akurgal, E., Phrygische Kunst (Ankara, 1955) 1 ff.Google Scholar, pls. 1–7 (on this style see GGP 379); Hrouda, B., Die bemalte Keramik des zweiten Jahrtausends in Nordmesopotamien und Nordsyrien (Berlin, 1957)Google Scholar; Margueron, J., ‘Trois vases du Bronze”, Syria xlv (1968) 75–96CrossRefGoogle Scholar; Schaeffer, C., Missions en Chypre (Paris, 1936)Google Scholar; Williams, M. V. Seton, ‘Painted pottery of the second millennium from Southern Turkey and North Syria”, Iraq xv (1953) 56–68CrossRefGoogle Scholar.

(The ship mentioned in the text is from Hama: Ingholt, H., Rapport préliminaire … (Copenhagen, 1940) 71Google Scholar pl. 22.2. Others of the end of the third millennium from Tello, see Hrouda, above).

METAL BOWLS:

Layard, A. H., Monuments of Nineveh (London, 1853) iiGoogle Scholar pls. 57–67; F. Poulsen, op. cit. n. 1, 6–37; Fr.Freiherr von Bissing, W., ‘Untersuchungen über die “phoinikischen” Metallschalen”, JdI xxxviii/xxxix (1923/1924) 180–241Google Scholar; Gjerstad, E., ‘Decorated metal bowls from Cyprus”, OpArch iv (1946) 1–18Google Scholar.

CYLINDER SEALS:

For an introduction see H. Frankfort op. cit. n. 5; Porada, E., Corpus of Ancient Near Eastern Seals in North American Collections. I: The Pierpont Morgan Collection (Bollingen Series xiv, 1948)Google Scholar; Buchanan, B., Catalogue of Near Eastern Seals in the Ashmolean Museum (Oxford, 1966)Google Scholar.

⁷⁸ The celebrated silver ‘Phoenician’ bowl from the Bernardini Tomb at Praeneste (MAAR iii (1919)Google Scholar no. 25, pls. 20, 21) bears scenes telling the story of a hunt and its curious sequel. This is apparently an illustration of an epic or fable. But it is of the latter half of the seventh century and could certainly have been influenced by the then well-developed Greek penchant for story-telling. This bowl has a near-double fron Kourion (Gjerstad op. cit. n. 77, pl. 8) which is unfortunately without an archaeological context.

⁷⁹ Smith, W. S., The Art and Architecture of Ancient Egypt (Harmondsworth, 1958) 217Google Scholar (with further refs.); Capart, J. and Werbrouck, M., Thebes (London, 1926) 120 ff.Google Scholar; Michalowski, K., The Art of Ancient Egypt (London, 1969)Google Scholar figs. 551–2.

⁸⁰ Aristonothos vase: Pfuhl, E., Malerei und Zeichnung (Munich, 1923) no. 65Google Scholar. The Acropolis sherds (Athens, NM) illustrated by Kirk, in BSA xliv (1949)Google Scholar pl. 40.3, 4, may be part of sea-fights; likewise the sherd published by Williams, in JHS lxxix (1959) 159–60CrossRefGoogle Scholar fig. 1, but seventh-century ship representations are rare (Williams, R. T., JHS lxxviii (1958) 129Google Scholar).

⁸¹ ‘The battle scenes cease abruptly around 750, at the end of the LG la phase.” (GGP 350.) The LG II oenochoai Agora P 4885 and Copenhagen 1628 may be taken to mark the end of the Geometric tradition (Ahlberg, , Fighting on Land and Sea in Greek Geometric Art (Stockholm, 1971)Google Scholar figs. 2, 31); then we have two Protocorinthian aryballoi, one from Lechaion (PLATE 12c), c. 690, the other from Perachora (Dunbabin, T. J., Perachora ii (Oxford, 1962)Google Scholar no. 27), c. 680.

⁸² See Davison figs. 1, 5, 6, 10b, 19, 20, for examples.

⁸³ Amandry, P. ‘Un motif “scythe” en Iran et en Grèce’, JNES xxiv (1965) 149–60Google Scholar.

⁸⁴ Munich 6402, Athens 866, both from the Dipylon Workshop, LG Ib.

⁸⁵ AM lv (1930)Google Scholar pl. II, no. 6214. See also Illustrated London News of 28 Nov. 1964, 860 fig. 8, for the same motif on one of the cylinder-seals from the cache discovered in the Mycenaean palace of Thebes; and S. Iacovidis, Perati pl. 47a for another fifteenth-century Mitannian seal, very similar to the Tiryns example, buried in a twelfth-century Greek grave.

⁸⁶ Barnett, R. D., ‘Ancient Oriental influences on archaic Greece’, The Aegean and the Near East (New York, 1956) 212–38Google Scholar, makes the case for the north-eastern route; but the Eusebian foundation-dates for Trapezus and Sinope are not above suspicion (JHS lxvi (1946) 72, 82Google Scholar), and none of the exports listed by Barnett need have travelled from Iran by way of the Black Sea. This is not to deny that there was a Black Sea route; but there is no certain evidence for its use as early as the eighth century. Herrmann, H.-V., ‘Urartu und Griechenland’, JdI lxxxi (1966) 79–141Google Scholar has demolished the long-established myth of the importance of Urartu as an art-centre (omne ignotum pro magnifico). Mrs.Birminham, J. M., ‘The overland route across Anatolia’, (AS ii (1961) 185–95Google Scholar) documents the contacts between Phrygia and the west coast of Anatolia and between Phrygia and points further east; but Herrmann's article removes the linchpin of her argument for the existence of a regular trade-route. As for the metal objects with Iranian affinities from Samos and Lindos, she herself says (192) ‘few objects of this Iranian group have so far appeared in central Anatolia’.

⁸⁷ For the ivories, see the works listed in n. 77, and JHS lxviii (1948)Google Scholar pl. II a–c. (Cf. especially Mallowan, figs. 435 and 561–2.)

⁸⁸ Ker pls. 23, 140, 141, grave 26.

⁸⁹ Op. cit. n. 1. Zervos pls. 250–1 has good pictures of Ohly's A12, A18, and A24.

⁹⁰ At the time of writing, the count of gold bands discovered since Ohly wrote is twenty:

(a) Unfigured, whether plain or decorated with linear motifs:

3 from Kavalotti St., Athens (ADelt xx (1965)Google ScholarChr. 75, pl. 44 a, b);

1 from Kerameikos, gr. hS 109 (ibid. 40);

1 from south of Olympieion, Athens (ADelt xviii (1963)Google ScholarChr 37–8, pl. 37d);

5 from Kriezi St., Athens (AAA i (1968) 23Google Scholar fig. 5, ADelt xxii (1967)Google ScholarChr 95, pl. 87a).

(b) Figured:

2 from Kerameikos graves 290, 291 (ADelt xviii (1963)Google ScholarChr 29);

1 from Eretria (AK xi (1968) 93–4Google Scholar, pl. 26.1; ADelt xxii (1967)Google ScholarChr pl. 180c);

2 from Kriezi St., Athens (see above under (a));

3 from Erysichthon St., Athens (ADelt xxii (1967)Google ScholarChr 79 f. pl. 78 c–d);

2 of unknown provenance (Early Art in Greece: Catalogue of the Exhibition held at the André Emmerich Gallery, New York, May-June 1965, 37Google Scholar no. 107; Hampe, R., Die Gleichnisse Homers (Munich, 1952) 46Google Scholar n. 35, pl. 12).

⁹¹ On the dating, cf. Higgins, R.BSA lxiv (1969) 152–3Google Scholar. Note particularly one of the bands from Kriezi St. (n. 90 (b) above) which has animals and was found with MG II/LG Ia pots.

⁹² If one agrees with Ohly's view that the matrices were of stone, they could only, at this date, have been produced by a craftsman trained in the east. Much scorn has been poured, for no good reason, on theories which postulate travelling or immigrant workmen. The Greeks themselves had a tradition of wandering smiths, the Telchines (Diodorus v. 55, Strabo xiv. 653), and in the Odyssey a smith is one of the recognized wandering professions. But even without such references one could infer that in a primitive economy like that of Geometric Greece new and specialized techniques were unlikely to become established unless a practising exponent of them brought them with him. It seems very likely that the unique series of ivory female statuettes from Dipylon Grave XIII, datable to c. 750 B.C., are to be explained by some such circumstance. The type of figure, nude female with polos, is at home in Syria, but the style is far more Greek—austere, precise, but a little ungainly—and the maeander of the polos is as near proof as one can ask for that these figures were carved on Greek soil. Ivory carving itself was, of course, a peculiarly Syrian speciality.

⁹³ Ohly (68 ff.) believes that the bands could only have had a funerary purpose, because of their extreme fragility. As J. M. Cook pointed out (Gnomon xxvi (1954) 110Google Scholar), contemporary literature, especially Hesiod, Theog. 578–84Google Scholar, is against this. It is quite possible that the bands were strengthened in some way, for example by being backed with leather or fabric.

⁹⁴ See Brown, W. L., The Etruscan Lion (Oxford, 1960) 166Google Scholar; Helly, B., ‘Des lions dans l'Olympe!’ REA lxx (1968) 271–85CrossRefGoogle Scholar.

⁹⁵ Ker pl. 162; Akurgal, E., The Birth of Greek Art (London, 1968)Google Scholar pls. 39–40; Zervos pl. 232.

⁹⁶ Fortetsa no. 1569, pls. 116, 169.

⁹⁷ BSA lxii (1967)Google Scholar pl. 12.

⁹⁸ J. Boardman, ‘The Khaniale Tekke tombs II’, ibid. 57–76.

⁹⁹ Fortetsa no. 578, pls. 37, 167.

¹⁰⁰ Kunze, E., Kretische Bronzereliefs (Stuttgart, 1931)Google Scholar.

¹⁰¹ The essential correctness of Kunze's relative dating is confirmed by a fragment of an indisputably Cretan shield discovered at Delphi in a deposit containing Geometric pottery (BCH lxviii/lxix (1944/1945) 45–9Google Scholar pl. 3). Mr. J. N. Coldstream informs me that this pottery is Corinthian, not later than the third quarter of the eighth century.

¹⁰² For a discussion of the oriental models, see Hencken, H. in AJA liv (1950) 297–302Google Scholar.

¹⁰³ Dunbabin, T. J., The Greeks and their Eastern Neighbours (London, 1957) 40–1Google Scholar.

¹⁰⁴ The Cretan Collection in Oxford (Oxford, 1961) 150Google Scholar.

¹⁰⁵ Copenhagen 727, CVA ii pls. 73.5, 74.2–6, AHS pl. 8, Ohly 79, Corolla Curtius (Stuttgart, 1937)Google Scholar pls. 42, 43, GGP 72 n. 2.

¹⁰⁶ Ker pl. 77, inv. 2160, isolated find.

¹⁰⁷ Vlastos Collection, BSA xxxiv/xxxv (1934–1935) 182–3Google Scholar.

¹⁰⁸ R. Hampe, op. cit. n. 70, 32 f.

¹⁰⁹ Ker pl. 69, inv. 407, isolated find; Schefold pl. 5a; Zervos pl. 108.

¹¹⁰ Particularly relevant are its occurrences on the sculptures of Carchemish (Woolley, C. L., Carchemish i (London, 1914)Google Scholar pl. B IIb) and Tell Halaf (von Oppenheim, M., Tell Halaf iii (Berlin, 1955)Google Scholar pls. 37, 39); on the ‘Phoenician’ metalwork of the seventh century (deep bowl from the Bernardini Tomb, MAAR iii (1919)Google Scholar pls. 12–18; bowl from Regolini–Galassi Tomb, Randall-McIver, , Villanovans and Early Etruscans (London, 1924)Google Scholar pl. 38. 1; unpublished bowl, Walters Art Gallery, Baltimore (drawn to my attention by Mr. W. Culican)); and on the North Syrian ivories (NI pls. xviii, xxii–vi).

¹¹¹ NI pl. xxii, S. 2.

¹¹² Sachs, A. J., ‘The Late Assyrian royal seal type’, Iraq xv (1953) 167–70CrossRefGoogle Scholar; Millard, A. R., ‘The Late Assyrian royal seal type again’, Iraq xxvii (1965) 12–16CrossRefGoogle Scholar.

¹¹³ The details of the Kerameikos tripod scene do not, of course, fit the later version of the legend. Most obviously, Herakles has a sword. But legend is protean, and in view of the fluidity of Greek iconography before the sixth century it would be unrealistic to expect this scene to match precisely the version of the story which later became canonical.

¹¹⁴ King, L. W., Babylonian Boundary Stones and Memorial Tablets (London, 1912)Google Scholar pl. 29a (winged, two-headed, with horse-forelegs which end as scorpions); Frankfort, op. cit. n. 5, pls., 31f, 34d (both winged); Porada, op. cit. n. 77, no. 749 (winged); and others. Note that none has the human forelegs which characterize early Greek centaurs. It may be that the comparative unpopularity of the centaur in the Orient was because the horse did not arrive there before the later third millennium B.C., by which time the iconography of monsters had become fairly well fixed.

¹¹⁵ λαχνήεντας (Iliad ii. 743Google Scholar).

¹¹⁶ Buschor, E., ‘Kentauren’, AJA xxxviii (1934) 128–32CrossRefGoogle Scholar.

¹¹⁷ R. Hampe op. cit. n. 51, pl. 38; Schefold pl. 15b.

¹¹⁸ Baghdad IM 52041, Davison fig. 116, GGP 59.

¹¹⁹ Fittschen (104–11) proposes an alternative explanation to Buschor's. He objects to the equation Centaur = Monster (unspecified), on the grounds that the spectator would not know which monster, and hence which story, was being depicted. He notes that centaurs not involved in a situation (‘ohne Handlungsbezug’) become rare just when ‘Handlungsbilder’ start to come in, and argues that a change took place in their interpretation. Originally, he thinks, they had the character of a spirit of death (like the sphinx?), but when they became identified with the Centaur of myth, they lost their funerary associations, except in the West, where these survived and ultimately reappeared in the art of Rome.

¹²⁰ NM 784, AM xviii (1893) 113Google Scholar fig. 10; Lullies, R. (ed.), Neue Beiträge zur klassischen Altertumswissenschaft (Stuttgart, 1954)Google Scholar pl. 3; GGP 60 ff.

¹²¹ Centaurs: Cook, J. M., BSA xxxiv/xxxv (1934/1935) 169Google Scholar n. 1; sphinxes: Audiat, F., Mon Piot 1938, 47Google Scholar; Verdhelis, N., BCH lxxv (1951) 16Google Scholar; Dessenne, A., Le Sphinx (Paris, 1957) 201Google Scholar; Coldstream, J. N., GGP 61Google Scholar.

¹²² NI pls. xvi–xvii, S. 3, and pp. 78–9 for discussion of the scene. For an interpretation of the Greek version in Homeric terms see Webster, T. B. L., BSA 1 (1955) 48Google Scholar.

¹²³ Bossert, H., Altsyrien (Tübingen, 1951)Google Scholar fig. 309; OpArch iv (1946)Google Scholar pl. 1; Perrot, and Chipiez, , Histoire de l'art dans l'Antiquité iii (Paris, 1885) 675Google Scholar fig. 483. For the non-Greek origin of the ‘chorus’ see Barnett, , NI 79Google Scholar n. 4. The scene is also found on Cypriot pottery and Cretan bronze-work: see Lehnstädt, K., Prozessionsdarstellungen auf attischen Vasen (diss. Munich, 1970) 18–21Google Scholar; Kunze, op. cit. n. 100, 214 f., pl. 48.

¹²⁴ The eight are made up of the seven mentioned in nn. 95, 125, ia6, and 130, together with one recently discovered on the site of the new Museum at Olympia (BCH xc (1966) 819Google Scholar fig. 11).

¹²⁵ Louvre AO 4702 (Dussaud, , BSA xxxvii (1936/1937) 92–5Google Scholar; Gjerstad, , OpArch iv (1946) 3Google Scholar pl. 2).

¹²⁶ NM Athens (Olympia iv pl. 52).

¹²⁷ Other examples of this shape which have figured decoration strikingly recalling oriental prototypes are Ker pl. 130 inv. 1319, Zervos pls. 50–1; London BM 1950.11–9.1 (Davison fig. 83); and Athens NM 14475 (Kunze, op. cit. n. 100, pl. 53e detail). Cf. Ahlberg, G., ‘A late Geometric grave-scene influenced by North Syrian art’, OpArch vii (1967) 177–86Google Scholar.

¹²⁸ Pritchard, J. B., The Ancient Near East in Pictures (Princeton, 1954)Google Scholar figs. 456–9.

¹²⁹ BSA xxxvii (1936/1937) 56–72Google Scholar pls. 7–9; Gjerstad, E., SCE iv. 2, 62Google Scholar n. 1. The dating could be as much as a century earlier, in view of Mrs. Birmingham's revision of the SCE dates in AJA lxvii (1963) 15–42Google Scholar.

¹³⁰ One from the Acropolis of Athens (JHS xiii (1892/1893) 248Google Scholar fig. 19); one from Delphi (Poulsen fig. 11, Perdrizet, , Fouilles de Delphes vGoogle Scholar pls. 18–20); and two from Olympia, (i) Kunze, op. cit. n. 100, Beil. 4c; Dunbabin, op. cit. n. 103, pls. 6, 7.1; (ii) Olympia B 1145, Herrmann, H.-V., Olympische Forschungen vi. Die Kessel der orientalisierenden Zeit (Berlin, 1966) 178Google Scholar n. 6, pl. 76.

¹³¹ Note that the sphinx does not occur at all in Homer, either in shape or in name; and Hesiod mentions the creature but does not describe it (Theog. 326 f.). Walter, H. (AuA x (1961) 63)Google Scholar remarks that it is ‘das äussere Kleid, das die Griechen übernehmen und mit Vorstellungen ihres eigenes Dämonenglaubens füllen’.

¹³² E. Akurgal, op. cit. n. 95, 171 pl. 48; Schefold pl. 4a; Dunbabin, op. cit. n. 103, pi. 5.1; Richter, G. M. A., The Sculpture and Sculptors of the Greeks (New Haven, 1965) 345Google Scholar fig. 13; JdI lxxix (1964) 40Google Scholar fig. 21 and, for further refs., 42 n. 115.

¹³³ Olympiabericht iv (1944) 114Google Scholar.

¹³⁴ JdI lxxix (1964) 43Google Scholar n. 119.

¹³⁵ Ibid. 42–3, figs. 26–7.

¹³⁶ AM lv (1930)Google Scholar Beil. 38. 2.

¹³⁷ Desborough, V. R., Nicholls, R. V., and Popham, M. R., ‘A Euboean centaur’, BSA lxv (1970) 21–30Google Scholar, pls. 8–10.

¹³⁸ Higgins, , GT 20Google Scholar pl. 6 A–B; BdA xxxv (1950) 320Google Scholar.

¹³⁹ AJA lxiv (1960) 11–12Google Scholar nos. 43–6, pl. 4; add to the examples there cited AM lxxviii (1963) 149–53Google Scholar Beil. 54. Note that Lefkandi is probably one of the few known settlements which displays continuity from L.H. into Geometric times.

¹⁴⁰ In Kunze's catalogue (see n. 100) nos. 1, 18, and 20 carry the typically North Syrian helmeted sphinx; nos. 2, 4, 5, 7, 8, 10, and 21, the common female sphinx. Other Cretan metalwork portraying the sphinx: Kavousi plaques (Kunze pl. 56e, A. Dessenne, op. cit. n. 121, 199 f.); shield from Delphi (see n. 101).

¹⁴¹ GGP 60.

¹⁴² For the pitcher, see JdI xiv (1899) 207 fig. 75Google Scholar, GGP 66 no. 4. For the grave see AM xviii (1893) 111–15Google Scholar: it was apparent even to Brückner and Pernice that the grave did not belong to the latest stage of the Dipylon cemetery.

¹⁴³ GGP 331.

¹⁴⁴ For a list of Geometric and Early Protoattic centaurs on pots, see R. Tölle, op. cit. n. 67, 96. No. 303 in that list (Florence, Museo Archeologico, unpublished) carries in a neck panel a single centaur with a cross-hatched torso and reserved eye, and is certainly LG IIb, as are all its fellows. On the gold bands centaurs first appear on Ohly's A19 (Group III), then on A20 and A20a (Group IV), which were made on the same moulds and of which the latter was found inside the ‘Stathatou’ amphora (GGP 59 no. 15) and is therefore definitely dated to LG lib. A19 may be a little earlier. Lists of centaurs may also be found in Fittschen (93 f., 111 f.).

¹⁴⁵ Winged centaurs only on the hydria in the Vlastos Collection, Athens, GGP 64 n. 5 (detail, not the centaurs, BSA xlii (1947) pl. 22a)Google Scholar. The earliest sphinxes on Greek pots are on a fragmentary ArgiveLG II (late) krater from Tiryns (Nauplia Museum no. 4268, GGP 143), where they are helmeted and sickle-winged; and on an LG IIb/EPA transitional standed bowl (Ker pl. 126, inv. 302), where they are bare-headed and straight-winged. Verdhelis, N. (BCH lxxv (1951) 1–37)CrossRefGoogle Scholar dates the appearance of the sphinx in Greek art earlier than I do; he places the Ephesus ivories very early, and calls an early seventh-century Argive-influenced krater from Syracuse Corinthian Geometric, For sphinxes on Cretan metalwork, see n. 140; the dating of these remains in dispute, but the examples must certainly span the latter half of the eighth century.

¹⁴⁶ See n. 114 for oriental centaurs. It is possible, of course, that one of the very rare seal stones carrying a centaur reached Greece; but since none of the oriental examples has human forelegs, as the Greek centaur does ab initio, and wings are usual, not, as in Greece, exceptional, the direct copying of an oriental image seems to be excluded.

¹⁴⁷ Olympia B 1665: Willemsen, op. cit. n. 27, pl. 46; Zervos pls. 220, 223.

¹⁴⁸ Akurgal, E., Art of the Hittites (London, 1962) pl. 104Google Scholar; B. Buchanan, op. cit. n. 77, nos. 1028–9; Jaarb. ex Oriente Lux xviii. 280–1 pl. 7.B3Google Scholar; A. Parrot, Assur figs. 84, 85; Bossert, H., Altanatolien (Berlin, 1942) figs. 768, 775Google Scholar; Woolley, L., Carchemish iii (London, 1952) pl. B33Google Scholar; Porada, E. in The Aegean and the Near East (NewYork, 1956) 201 fig. 12Google Scholar ( = JdI lxxxi (1966) 26 no. 44 fig. 34)Google Scholar.

¹⁴⁹ RDAC 1964, 116 pl. 8.2Google Scholar. For the dating, see n. 129. On the transmission of motifs from the Syrian culture-area to Cypriot pottery, see Yon, M., ‘Sur une representation figurée chypriote’, BCH xciv (1970) 311–17CrossRefGoogle Scholar.

¹⁵⁰ Kunze, op. cit. n. 100, no. 74; Dunbabin, op. cit. n. 103, pl. 10. 1; Zervos pl. 233.

151 BCH xlv (1921) 384–5 fig. 45 (no. 154)Google Scholar.

¹⁵² Möbius, H., ‘Griechisch-orientalische Bleimedaillons aus Ionien’, AA 1941 cols. 1–36Google Scholar. Another specimen is known from Delos.

¹⁵³ For what follows, see also Johansen, K. F., The Iliad in Early Greek Art (Copenhagen 1967) 17–25Google Scholar.

¹⁵⁴ Many of these motifs can also be matched from Minoan-Mycenaean art, and it may be that the poetic tradition preserved a reminiscence of works such as the Siege Rhyton and the Harvesters' Vase. But the main argument, that for Homer and his audience the essential subject-matter of art was the everyday world, remains unaffected even if this was the case.

¹⁵⁵ Matters were very different for the composer of the Hesiodic Scutum, and for Virgil.

¹⁵⁶ Op. cit. n. 103, appendix B, Attic nos. 1–9, ProtoCorinthian no. 1, Argive nos. 1-2. Add the krater Louvre A 517 (seen. 164).

¹⁵⁷ Copenhagen 7029 (CVA ii pl. 73. 3, GGP 59 no. 17) and Athens NM 784 (see above, p. 46 f.).

¹⁵⁸ So Fittschen 105f.

¹⁵⁹ BM 1899.2–19.1 (GGP 55 no. 4, Davison fig. 98, Schefold pl. 5c, Zervos pl. 123).

¹⁶⁰ ‘… die vorgeschlagene Deutung (viz. as the Rape of Helen) … wäre nur denn sicher, wenn sich annehmen liesse, dass der Frauenraub damals nur noch im Liede vorkam’ (F. Matz, op. cit. n. 31, 65). A more plausible identification is that of Robert, revived by Coldstream in BICS xv (1968) 91Google Scholar, which makes the man Theseus and the woman Ariadne. This explanation depends on identifying the object the woman holds so prominently in her right hand as the Crown of Light, the attribute of Ariadne in later Greek art. But there is nothing to show that it is not a wreath or garland (as held, for example, by so many of the lead Potnia Theron figures from the shrine of Artemis Orthia at Sparta), which would be very appropriate at a farewell party.

¹⁶¹ Copenhagen, Ny Carlsberg Glyptothek (on loan): Brommer, F., Herakles (Münster-Köln, 1953) pl. 18Google Scholar, Schefold pl. 5b.

¹⁶² Ithaca no. 163, BSA xliii (1948) 42 fig. 29Google Scholar.

¹⁶³ Munich 8696: R. Hampe, op. cit. n. 90, pl. 7–11; GGP 76 no. 3 (LG IIa), Schefold pl. 8.

¹⁶⁴ Krater Louvre A 519, GGP 31 no. 17, Davison fig. 11, Hampe, op. cit. n. 51, pl. 34; krater New York 14.130.15, JHS lxxxvi (1966) 1–4Google Scholar, pls. i–iii; oenochoe Athens Agora P 4885, GGP 55 no. 9, Davison fig. 97, Schefold pl. 7a; Argive (?) sherd, Athens NM, Hampe pl. 34. To these four instances should be added the krater Louvre A 517, GGP 30 no. 4, pl. 7a.

¹⁶⁵ BSA xxxv (1934/1935) 206Google Scholar.

¹⁶⁶ Ahlberg, , PE 240–52Google Scholar.

¹⁶⁷ GGP 351.

¹⁶⁸ BSA xlii (1947) pl. 18a, Schefold pl. 7bGoogle Scholar.

¹⁶⁹ e.g. CGA pl. 63.

¹⁷⁰ See also von Bothmer, D., Amazons in Greek Art (Oxford, 1957) 1 no. 1Google Scholar, with bibliography.

¹⁷¹ Agora P 10201: AK ii (1959) 35–7Google Scholar, Agora viii no. 311. For objections, see Cook, J. M. in Gnomon xxxiv (1962) 823Google Scholar (identification) and Coldstream, J. N. in JHS lxxxiv (1964) 217 (date)Google Scholar.

¹⁷² CA 2509: GGP 38 n. 4, Zervos pl. 73.

¹⁷³ Johansen, K. F., ‘Aias und Hektor’, Historisk-filosofiske Meddelelser xxxix no. 4 (1961) 1ffGoogle Scholar.

¹⁷⁴ Himmelmann-Wildschütz, N., Marburger Winckel mannprogramm 1961, 1–5Google Scholar, followed in essentials by Ahlberg, op. cit. n. 81, 21–5.

¹⁷⁵ Olympia B 1730: Willemsen, op. cit. n. 27, pl. 63, Schefold pl. 6b, Akurgal, op. cit. n. 95, pl. 59, Zervos pl. 221–2.

¹⁷⁶ On dating and interpretation, see Kunze, E.,‘Archaische Schildbänder’, Olympische Forschungen ii (Berlin, 1950 115Google Scholar; Willemsen, F., JdI lxx (1955) 94Google Scholar, and op. cit. n. 27, 104, 109; Amandry, P., Rev. Phil. lxxxii (1956) 275Google Scholar and Gnomon xxxii (1960) 461Google Scholar.

¹⁷⁷ Ischia: Expedition viii/4 (1966) 6 fGoogle Scholar. Samos: AM lxi (1941) 416, pl. 11Google Scholar.

¹⁷⁸ Webster, T. B. L., ‘Homer and Attic Geometric vases’, BSA 1 (1955) 38–50Google Scholar.

¹⁷⁹ Snodgrass, A. M., Early Greek Armour and Weapons (Edinburgh, 1964) 58–60Google Scholar. Webster's thesis is accepted by Benson (op. cit. n. 1, 105), since it conforms with his general ideas about the Mycenaean inspiration of Geometric art. For the case against, see also Fittschen 32 f. and Ahlberg (op. cit. n. 81) 59–66.

¹⁸⁰ e.g. Exekias' Vatican amphora showing Ajax and Achilles at dice (AHS pl. 62).

¹⁸¹ For a full discussion see Snodgrass, op. cit. n. 179, 37–51.

¹⁸² ‘Boeotian’ shields are depicted on the Apadana relief at Persepolis, with a central ornament of very similar type to one found at the Samian Heraion, probably of eighth-or seventh-century date. See AM lxxxiii (1968) pl. 124Google Scholar.

¹⁸³ Op. cit. n. 153, 20.

¹⁸⁴ Lorimer, H. L., Homer and the Monuments (London, 1950) 152Google Scholar.

¹⁸⁵ i. 21. I.

¹⁸⁶ AJA lxxv (1971) 98Google Scholar.

¹⁸⁷ Catling, op. cit. n. 32, 142–6.

¹⁸⁸ Assyrian reliefs of the reigns of Sennacherib (704–81) and Ashurbanipal (668–627), from Nineveh, show three different types of shield in use simultaneously in the Assyrian army: a rectangular ‘tower’ shield with a rounded top; a convex round shield (of hide ?) with a distinct boss and rim (of metal?); and a flat round shield made of some plaited material (osier?). For illustrations, see Strommenger, E. and Hirmer, M., The Art of Mesopotamia (London, 1964) figs. 223, 237, 240Google Scholar. It may be no more than a coincidence that this is just the period when the Greeks were by experiment evolving the hoplite armour which later became standard (Snodgrass, op. cit. n. 179, 189 f.).

¹⁸⁹ All dates and statistics given here are from Snodgrass (op. cit. n. 179) 39-41.

¹⁹⁰ For a fuller statement of this defence, see Courbin, P. in Vernant, J.-P. (ed.), Problèmes de la guerre en Grece ancienne (Paris–The Hague, 1968) 79–82Google Scholar.

¹⁹¹ Sophocles, , Ant. 332–3Google Scholar.

¹⁹² e.g. Altheim skyphos (Robinson Studies ii, pl. 10), round and Dipylon shields; Kerameikos frr. (AA 1963, 643, nos. 1, 2, 3, 4Google Scholar); amphora from the Kynosarges cemetery (BSA xii (1905/1906) 82 fig. 2a)Google Scholar, oblong and Dipylon shields; amphora Markopoulo (Ergon 1960, 35 f. fig. 48 pl. 11–12Google Scholar), round and oblong shields; fr. in Vienna (AZ 1885, 139–40Google Scholar), round and Dipylon shields; and those listed in BSA 1 (1955) 42 n. 33Google Scholar.

¹⁹³ Snodgrass (op. cit. n. 179) 61.

¹⁹⁴ BSA 1 (1955) 42Google Scholar.

¹⁹⁵ See p. 54.

¹⁹⁶ e.g. the amphora Louvre CA 1823 (GGP pl. 12c).

¹⁹⁷ Cf. the terracotta model in the British Museum, Lorimer (op. cit. n. 184) pl. vii.

¹⁹⁸ Snodgrass (op. cit. n. 179) 60 and pl. 15. The only inference, pace Snodgrass, which can safely be made from the way the three ‘Dipylo-Boeotian’ shields on this lekythos are being handled by their owners is that the artist was not au fait with the finer points of shield-drill. Surely the two leading warriors' shields should have telamones visible? And how does one hold a convex shield at arm's length by a central hand-grip and still manage to carry a spear in the same hand, as these two are doing?

¹⁹⁹ Perachora ii. 417 pl. 177 A 35Google Scholar.

²⁰⁰ Zervos, pl. 175-7; Lamb, op. cit. n. 67, 43 pl. 17B.

²⁰¹ Ahlberg, , PE 285–91Google Scholar.