Hostname: page-component-cd9895bd7-gbm5v Total loading time: 0 Render date: 2024-12-25T07:36:10.608Z Has data issue: false hasContentIssue false
  • English
  • Français

Metal-Working in Homer1

Published online by Cambridge University Press:  23 December 2013

D. H. F. Gray
D. H. F. Gray
Affiliation:
St. Hugh's College, Oxford
Get access Rights & Permissions [Opens in a new window]

Extract

In discussing the transition from bronze to iron in Anatolia, Dr. Stefan Przeworski incidentally identifies Homeric conditions with the stage in the historical development of metallurgy which he calls Chalcosideric. Professor Nilsson and Miss Lorimer have argued briefly but effectively that the poems contain elements from different periods; but belief in an historical ‘Homeric Society’ dies hard and justifies a more detailed examination of all the references to metals in the poems.

Przeworski's transitional age began about 1300 B.C. in Anatolia and about a century later in Greece; in both it ended about 700 B.C. Before it began, bronze was the useful material for all industrial purposes, and the rare uses of iron were ornamental or magical. After it ended, iron was the normal industrial material, and the more malleable bronze was used for fine work or elaborate modelling. The characteristics of the intermediate period are: 1. Imitation of Late Bronze Age types in iron. 2. Simultaneous appearance of bronze and iron objects of the same purpose and type. 3. Inlay of bronze objects with iron. 4. Combination in the same weapon or tool of iron working and bronze ornamental parts. 5. Addition of iron working parts to bronze objects such as cult-wagons and utensils. 6. Use of bronze rivets on iron weapons and tools. 7. Repair of bronze objects with iron parts (Przeworski 175–6.) Most of these characteristics are so technical that they are unlikely to be reflected in poetry. Moreover, so many bronze objects were in common use at all periods, including the full Iron Age, that the most significant evidence may be taken to be the relative value of the metals, the relative frequency of bronze and iron weapons and tools, and the degree of familiarity shown with the methods of the forge as distinct from the foundry.

Type
Research Article
Information
The Journal of Hellenic Studies , Volume 74 , November 1954 , pp. 1 - 15
Copyright
Copyright © The Society for the Promotion of Hellenic Studies 1954

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

2 Die Metallindustrie Anatoliens (1939), 177. So also Forbes, R. J., Metallurgy in Antiquity (1950), 458.Google Scholar Both quote Lang, Andrew, R. Arch. VII (1906), 280Google Scholar, as their authority.

3 Nilsson, M. P., Homer and Mycenae (1933), 139–42.Google ScholarLorimer, H. L., Homer and the Monuments (1950), 111–21Google Scholar (hereafter Monuments); since this appeared, I have been able to shorten my notes considerably.

4 σίδηρος, as is clear in the description of tempering in 1 391–4, is strictly mild steel, a low carbon content being picked up from the charcoal; but the translation ‘iron’ is traditional, and it seems better to keep it than to vary the English word for the same metal or to use ‘steel’ for all the ferrous object of Homer and the Early Iron Age. So ‘bronze’ is a convenient translation of χαλκός though utensils were in fact copper, and copper ingots were probably preferred, since the tin content is reduced when bronze ingots are remeked. For weapons and tools thecommonest alloy found is tin, usually about 3·15 per cent, but lead, arsenic, and antimony are sometimes present, apparently as alloys and not merely as impurities (Przeworski, op. cit. pp. 89–110, and Hampe, and Jantzen, , JdI LII (1937)Google Scholar, Bericht, pp. 34–5). Arsenic, which is now used to give greater tensile strength to wrought metal, would also improve the hardness and toughness of hammered castings. Lead-bronze lacks hardness, but has good colour and greater malleability. Lead is mentioned in Λ 237 and ω 80 as soft and heavy. In δ 73, but not in ο 460 and σ 296, ἤλεκτρον is probably the alloy of gold and silver. All numbers refer to occurrences of words, not to objects, e.g. the same spear described four times as bronze counts as 4 and not as 1. The equals sign is used when the relevant parts of the lines are identical.

5 See Przeworski, op. cit. p. 149 for table of relative prices, based on the evidence of the Cuneiform texts given by Meissner, , Babylonien und Assyrien I, pp. 363 ff.Google Scholar

6 Wainwright, G. A., The Coming of Iron, Antiquity X, p. 21.Google Scholar See also Przeworski, op. cit. pp. 140–1 and references in the notes.

6a Regting, RE VII 973–6.

7 Specified weights were presumably in bar form, but certainly in ο 448, cf. 469, and probably in most of the other places where gold or bronze are joined to clothing, cattle, and slave-girls, the wealth in metal would be in manufactured form, as much of it was in the fifth century, Thuc. II. 13. 4–5 and VI. 46. 3.

8 φ 61–2 sounds like, and perhaps originally was, a general description of wealth, but in φ 81 it is limited to mean the iron axes.

9 Monuments, 111–15.

10 Palace of Minos II, pp. 169–70. Dr. Desch drew my attention also to a bangle of pure tin which he identified among the finds from Thermi, , Lamb, , Excavations at Thermi, pp. 165, 171–3, 215.Google Scholar It also is Early Bronze Age.

11 Monuments, 189–91.

12 The Sutton Hoo Ship-Buria, p. 25. I am told that such surface tinning would occur accidentally when tin was worked with new bronze or iron implements.

13 Wace, A. J. B., BSA XXV, p. 350Google Scholar: Schuchhardt, , Schlie mann's Excavations, pp. 145 and 302.Google ScholarDickins, Guy, The Hieron of Athene Chalkioikos, BSA XIII, pp. 139–40.Google ScholarMonuments, 429. BCH LXXVII 193, with fig. 2 and Pl. XXXII.

14 Blinkenberg, Fibules grecques et orientales, fig. 50.

15 Palace of Minos III, pp. 111–33. Karo, , Schachtgräber, pp. 313–16.Google ScholarPersson, , Royal Tombs, pp. 48–51.Google Scholar Technical report by Plenderleith, H. J. in Enkomi-Alasia I, 381–9.Google Scholar Humfry Payne has shown how improbable it is that the polychrome style on vases is derived from metal-work, when the known techniques of metal-working are quite different (Necrocorinthia, 95, cf. 19 n. 2).

16 Cf. BCH LXXI–II (1947–48), 148 f., especially 243–9 and pl. XXV. The theoretical possibility is also shown by the presence of post-Mycenaean objects in Mycenaean tombs (Ἐφ. Ἀρχ. 1937, 377–90) and by accounts of ‘Treasuries’ (e.g. Paus. ii. 16. 6, ix. 36. 5, 38. 2, 37. 5 f.).

17 Myres, J. L., JHS LIII (1933), 29.Google Scholar

18 Blümner, in Pauly-Wissowa RE VII, cols. 1575–7.Google Scholar

19 Persson, , Royal Tombs, p. 50Google Scholar, and New Tombs, pp. 89–91. Schaeffer, , Cuneiform Texts of Ras Shamra Ugarit, Pl. XXXIII and XXXV. I.Google Scholar

20 Thompson, Homer, Harvard Studies, Suppl. Vol. I, pp. 183 ff.Google Scholar

21 Cf. Bailey, K. C., The Elder Pliny's Chapters on Chemical Subjects, Part I, Nat. Hist. XXXIII. 64–5, 100 and 123.Google Scholar Dr. Desch tells me that argentum vivum and hydrargyrum are the same product, the former natural, the latter distilled from cinnabar.

22 BSA XLII, p. 88, fig. 5 and p. 89. Myres, , Handbook of the Cesnola Collection, p. 382, no. 3209.Google Scholar

23 χέω and its compounds are generally used of liquids or of solids thought of as fluid, e.g. a stream of people or grain, a shower of leaves or feathers. But the idea of ‘pouring from above’ (M 284) passes easily into the simple idea of ‘covering’ ( 114, Φ 319, π 47). There is little resemblance to water in the heaped bodies of geese and sheep (τ 539, Ε 141), still less in a wooden breastwork (ε 257). So in Ο 364 and 473 συνέχευε is used both of knocking down sand-castles and of breaking a bow. The use of περιχεύεται does not mean that the poet had in mind the pouring of liquid gold. Blümner suggests that χρυσοχόος, used in γ 425 with χαλκεύς as an alternative, means either ‘gilder’ on the analogy of περιχεύεται or ‘gold-melter,’ because the goldsmith would generally use gold scrap which had to be melted down; it does not mean that ‘gold-pourer’ was the way to describe a ‘gilder.’

24 Schol, κραίνω in Homer means ‘carry out’ a promise, threat, etc., once ‘hold sway,’ but the use here is appropriate, especially if the word is connected, by true or false etymology, with κάρα, cf. Soph. O.C. 473. So Bechtel, Lexilogus. Even if it were part of κεράννυμι, it could hardly mean more than a combination of two metals. The verb in Homer is always used of mixing drink or bathwater. It would be the exact term for ‘alloy’ and means ‘adulterate’ in (Dem. XXIV. 214). Quicksilver produces a thin layer of alloy between the base and the outer skin, but the poet could not be supposed to know this.

25 Jong, P. de and Hood, M. S. F., Late Minoan Warriorgraves, in BSA XLVII, 243–77.Google Scholar

26 Ib. 251 and Pl. 52. b. cf. Karo, , Schachtgräber, Pl. XXIV no. 35, 116 and 241.Google Scholar Some of the rivets from the Shaft Graves may have been used for the same purpose.

27 Ib. 260.

27a Monuments, 225–7, and Pl. XIII. 2.

28 BSA XLVII, 256 and Pl. 50–2.

29 Monuments, 211, 226 and pl. XIII. 1.

30 E.g. on the badly-worn sherd shown Röm. Mitt. 59. 185.

30a Compare now a similar ivory from Paphos, Old, ILN May 2, 1953, p. 710Google Scholar, fig. 8.

31 Monuments, Pl. XII. 2, and possibly 1. On the sherd from Schliemann's dump, Wace, A. J. B., Mycenae, Pl. 71.Google Scholar c. I, the chiton may originally have been fringed, as in Monuments, Pl. II. 3.

32 BSA XXXVIII, 136–41.

33 To the examples in Monuments 200–1, add AA 1927, 251.

34 Studniczka, , Ath. Mitt. XII. 21 f.Google Scholar and fig. 4.

35 Rodenwaldt, , Fries des Megatons von Mykenai, 39.Google Scholar fig. 20 and plate at end.

36 Murray, , Excavations in Cyprus 51.Google Scholar These and other bronze greaves will be discussed by Mr. Hector Catling in an article to appear shortly in Opuscula Atheniensia, to which he has kindly given me permission to refer. He has convinced me that they are Mycenaean and have parallels on the Mainland. Since, however, they still seem to be less common than the type on the Warrior Vase, I do not think that the conclusions reached on p. 8 below are affected. As evidence that bronze greaves were characteristic of the ‘Achaeans’ whom Merneptah defeated, reference is sometimes made (e.g. Schaeffer, C. F. A., Enkomi-Alasia I (1952) 342Google Scholar) to de Roudé, in Rev. Arch. XVI (1867), 44Google Scholar, repeated by H. R. Hall. De Rougé tentatively suggested ‘épies, poignards, cuirasses et cne'mides, et ustensiees divers’ as translation of a mutilated passage (= Mariette, , Karnak, Pl. 55Google Scholar, Col. 61) describing booty taken from the non-Libyan invaders. Cnémides translates the sign for razors (Budge, , Egyptian Hieroglyphic Dictionary, (1920) CXXXVIIIGoogle Scholar) and is interpreted by Breasted, , Ancient Records of Egypt, (1906) III, 251Google Scholar, as ‘knives of copper’. The cuirasses are obtained by ignoring a lacuna and taking tcharna as a ἁπ. λεγ. variant for tharin = Hebrew corslet (Budge, s.v. 851, 899, following Proceedings of the Soc. of Bib. Arch. X (1888), 472. I owe these references to Professor J. Cerny.) In any case the ‘‥and copper razors’ were captured from ‘‥, ‥, -men, Meshwesh, ‥.’ The Akaywash, Akaiwasha, Ekwesh, or ‘A-qi-w -ša, with the other non-Libyans probably appeared in the lacuna, but the unknown equipment was not characteristic of them in particular. If further discoveries show that bronze greaves were typically Mycenaean, it will increase the probability that their absence from the poems is a Geometric characteristic. The solitary of H 41 is not in its context likely to be a genuine reminiscence of the Bronze Age. It is more probable that it appeared at a time when hoplite greaves were taken for granted and that it replaced a word which was metrically objectionable; perhaps since very rarely retains its natural quantity in hiatus (Monro, , Homeric Grammar, § 380Google Scholar).

37 The shields from Palaikastro may be cult objects (Benton, , BSA XXXIX, 52–64Google Scholar; XL, 52–4, 82). The votives (BSA XL, pl. 27, 17) are more utilitarian. Omphalos shields: Olympia IV. pl. 62 no. 1006–7; from Cumae, , Mon. Ant. XIII (1903) 246Google Scholar, fig. 24, and fairly commonly from Italian sites; from a sixth-century warrior grave in Macedonia, , Filow, , Die arch. Nekropolis von Trebenischte, no. 122, fig. 100. 2–4.Google Scholar Lambda shields: Fouilles de Delphes, V. 25, fig. 99: from Idalion, , Perrot, et Chipiez, , Histoire de l'art dans l'antiquité III, 869, fig. 636.Google Scholar Spike shield: ib. fig. 639. The Italian shields are discussed by Åkeström, A., Der geometrische Stil in Italien, 68, 102 f., 119 f.Google Scholar, and Pl. 28; he dates the earliest to the first quarter of the seventh century, but his chronology involves lowering Thucydides' data for the foundation of Syracuse; for criticism see Dunbabin, T. J., The Western Creeks, Appendix 1, especially 466–70.Google Scholar

38 MissLorimer, (Monuments, 233)Google Scholar concludes from the absence of metal, except for one very small tube which may be a crest holder, in Geom. graves that helmets were of perishable material. She accepts, however, bronze Geom. helmets on: figurines from Olympia and Delphi, Pl. XVII. 3: pyxis from Argive Heraion, Pl. XVII. 2: armourer, c. 700 B.C., AJA XLVIII. 1–2, fig. 1–4. This is enough for my argument, but I think that Geom. helmets with rigid stilts (Monuments, fig. 12, cf. similar helmets such as Matz, Gesch. d. gr. Kunst I Pl. 29b–31) or metallic ribbing (Matz. op. cit. Pl. 27a. I, 273, fig. 40) and probably those with heavy crests (Olympia IV, Pl. 16, no. 242–3, JdI 14. 84 fig. 42 and 85. fig. 44) or offset contours (A.M. 17. 211–15, fig. 2, 3 and 4 and Pl. X. 2. Α.ζ. 1884. Pl. 9. 1, 1885. p. 131 and 139) are bronze, and that many others may be. That would take bronze helmets well back into the eighth century. The descendants of the types are clear in the seventh, stove pipe, stilted, Corinthian and conical (BSA XL. Pl. 31. 17, Pl. 28. 31, Pl. 32. 23, Matz, op. cit. Pl. 38. a, and the archaistic ribbed helmet, ib. Pl. 69). Montelius, , Die älteren Kulturperioden II. p. 3101Google Scholar shows Assyrian helmets which might well have been thought to be non-metallic if actual examples in metal had not been found.

39 Lorimer, , Monuments 170–1Google Scholar, dates before 700 B.C., Hampe, , Die Gleichnisse Homers, 38Google Scholar, to early seventh century, and Cook, J. M., BSA XXXV. 207Google Scholar, to after 680 B.C.

40 Pfuhl, MuZ III, no. 117. Conze, , Melische Thongefasse, Pl. III.Google Scholar cf. BSA XLII, p. 88, fig. 5.

41 Hampe, , Neue Funde aus Olympia, Die Antike XV. 25.Google Scholar ‘Nur die wenigsten waren aussen ganz mit Bronze überzogen. Bei den meisten war—das hat sich jetzt herausgestellt—die hölzerne Wölbung nur mit einem Schildzeichen aus Bronzeblech beschlagen.’

42 Pfuhl, op. cit. no. 140, 218–19, 229, with comments in text. On r.f. vases it is normal.

43 CQ XLI, 109 f.

43a On this page, a dagger † indicates θώρηξ certainly of metal.

44 Since there are a few unmistakable references to the body shield, it is probable that both words go back earlier, the Greek word σάκος perhaps being the Helladic Tower-shield, and the non-Greek άσττίς the Minoan 8-shield, and What survives is a small group of phrases and details which vary the narrative without altering its course. In their duel, Aias outdoes Hector at each stage, but his actions are the same.

45 At end of line, 24, so with Argives and Trojans 2 each, Epeians and Boeotians 1 each, 694, 287. N 439–40 may be a solitary survival of the same tradition; it sounds more like an adaptation of to the pattern of Η 371–2 = 397–8. may also belong here, and which is put with it for consistency in Table C but seems in the poems to mean ‘with bronze helmet’; it is used 8 times out of 9 of Hector, cf. κορυθαίολος and Ζ 469. Il. 31, Od. 5, έταῑροι Od. 5. κνημῑδες appear in arming scenes with θώρηξ, sword, helmet, and one (Γ 330–8, Τ 369–91) or two (Λ 17–46, Π 131–44) spears, but not in those with shield, helmet and single spear (E 736–47, 0 125–7, 479–82). This in fact agrees with their introduction in Myc. III, though it is unexpected in passages so easily compounded of a series of formulae.

46 Poulsen, , Der Orient u. frühgr. Kunst, 170Google Scholar, notes that snakes were not a Mycenaean decorative motif. The coloured bands of the long scale corslet (Monuments, fig. 17) are an attractive parallel, but conflation of two Bronze Age types, though possible, is less probable.

47 See p. 6 above. It is more probable that the poet mentioned the corslet when it was relevant and left it out when it was not (Φ 179–83) than that a rhapsode introduced it in Ψ because the blow was aimed low in Θ (Monuments, 204). The deduction of a corslet from wounds in the belly is doubtful. The advice given in bayonet drill to ‘aim below the belt’ does not imply a modern corslet.

48 It has 13 different epithets, used only 18 times in all. θώρηκα (11) always comes before the caesura and θώρηξ (6) at the end of the line. Other cases have no preferences. (Δ 133 = Y 415) may be a misunderstood formula.

49 BSA XLII, 114.

50 A late figurine, de Ridder, Les Bronzes antiques du Louvre I, Pl. 14, no. 124, shows a simple use of round, concave plates. Geometric bronze helmets are certain, but none has been found. Starr, , Nuzi I, 476–80Google Scholar and II Pl. 126 D, J, shows fifteenth-century plates ‘which could only be used when sewn on to a stout fabric or leather base’; it is known from the texts that 109–242 were needed for one garment but only 4 were found. Valeton suggested this meaning of γύαλα in Mnemosyne 47 (1919), 187 f. If a bronze corslet of any kind was original, it is easier to explain the inorganic lines in which it seems intrusive.

51 For the gold ring on Hector's spear see Monuments, 260; three of the four spears from the L.M. II warrior graves had bronze rings, BSA XLVII, 267, no. II. 4, 271, no. III. 14, and 275 no. V. 7. ἐπισφύρια are unparalleled; the gold ornaments from Schliemann's Shaft Graves, as is shown by finds from the new Grave Circle, were not worn on the legs (ILN 6 March 1954, 365, fig. 19). μΐτραι and are too obscure to be used as evidence.

52 See p. 8 above.

54 The classification of χαλκός etc., depends on the context. Often there is reference to a specific weapon or tool. For cutting throats or wax, mutilating an enemy and peeling bark, a knife or dagger seems most suitable; in Γ 271 = 252 Agamemnon cuts off hair with the μάχαιρα which he carries beside his sword. Perhaps Ψ 412, where horses are to be butchered, should go with them, but it has been grouped as ‘unspecified’, a class which consists almost entirely of the weapons in such phrases as and therefore in fact mainly refers to spears. T 222 is a good example of ambiguity, since χαλκός is equally appropriate to the weapon in battle or to the sickle with which it is compared.

55 Op. cit. pp. 138–44 and 149–52. Cf. especially Starr, , Nuzi I pp. 194Google Scholar, 470 and 475 and Pl. 125 KK, Ugaritica I, chap. 3, and Carter, , The Tomb of Tut-ankh-amen II, Pl. LXXXVIIGoogle Scholar B cf. LXXVII B, LXXXII A and III, Pl. XXVII.

56 Mercer, , The Tell el-Amarna Tablets I no. 22. pp. 83 and 85Google Scholar, cf. no. 25.

57 Luckenbill, , A. J. Sem. Lang. XXXVII, p. 206.Google Scholar

58 Id., Ancient Records of Assyria and Babylon I, no. 222, cf. no. 236 and 247.

59 Contenau, and Ghirshman, , Fouilles de Tepe Ghiyan, p. 44, Pl. 8 ff.Google Scholar

60 Ghirshman, , Fouilles de Sialk II, pp. 9 and 45–9.Google Scholar

61 OIP XIX, figs. 360–2; XXIX, figs. 494–502; XXX, fig. 111.

62 For sites in Palestine see Wright, G. E., Iron: the date of its introduction into common use in Palestine, AJA XLIII, pp. 458 ff.Google Scholar

63 L.C. III at Idalion, , SCE II, 597 ff.Google Scholar C.G. generally SCE I–III passim. A few objects not assigned to one particular period are omitted. In the discussion of types in IV. 2, p. 212, the difficulty of fixing the chronological sequence of metal objects is emphasised.

64 BSA XXXVIII, 112–22.

65 Hall, , Vrokastro. 103–6, 138–9, 143.Google Scholar

66 Swedish Messenia Expedition, 102–4, 367–73: Monuments, 112.

67 From the grave inventories in Kerameikos I and IV.

68 Payne, , Perachora, 69–75 and 167–90.Google Scholar

69 Weber, H., Olympische Forschungen I, 146–71.Google Scholar

70 See Mireaux, E., Les poèmes homériques I (1948), 298.Google Scholar

71 op. cit. 147.

72 The Coming of Iron, in JEOL IX, 207–14.

73 op. cit. (n. 62 above) 460.

74 Alexander, and Street, , Metals in the Service of Man, 131.Google Scholar

75 Casson, , The Technique of Early Greek Sculpture, 227–29.Google Scholar ‘In the Foundry Vase no less than six hammers are shown. All alike have long, slender handles and light heads which are not pointed…. The long, slender hafts of the hammers show plainly enough that the hammers were used lightly. Hard hitting would soon break so slender a shaft. The hafts on the hammers on the Acropolis fragment are even more slender, and obviously intended for only the very lightest tapping. For the whole process of bronze-working admits only of light work with the hammer or rasp or gouge. Heavy blows would soon crack cast bronze or bring out latent weaknesses or flaws. Steady and continuous tapping would produce good results.’

76 Hdt. I. 67–8.

77 Przeworski, op. cit. 157.

78 See Monuments, 121, for reasons for identifying Temese with Tempsa in Bruttium rather than Tamassos in Cyprus.

79 Forbes, R. J., Metallurgy in Antiquity, 190Google Scholar: ‘In Asia Minor there are no less than 26 important deposits, seven of which are located in Pontus in the district south of Trabzon.’ The galena deposit of Karasar is exceptionally rich in silver content (ib. 180).

80 Strabo 549–552.

81 Przeworski, op. cit. 180: ‘Für die chalkosiderische Stufe ist die Armut an Edelmetallsachen höchst bezeichnend.’

82 Demargne, , La Crète dédalique, 199–216.Google Scholar

83 Monuments, 273–4.

84 See Lorimer, H. L., Gold and Ivory in Greek Mythology, in Greek Poetry and Life, 14–33.Google Scholar

85 In P 376 is unnecessarily translated ‘by the pitiless weight of their armour.’ (Π 739) is a nice adaptation.

86 Many are conveniently expansible, e.g. or and are strict. Generally the use is freer, e.g. άτειρής (of bronze 7: of fighters and their strength, once compared to an axe, 3: of voice 3. The entry in L. and S. (1925) needs correction). It comes in three patterns and one individual line:

But Ο 697 Rare epithets are

87 Not a Bronze Age parade weapon. The weapon, not the material, is the oddity from which he takes his title.

88 A bronze cult-wagon from an eighth-seventh-century context at Toprah Kaleh has iron axles. Przeworski, op. cit. Pl. XII. Knowledge of such bronzes could have reached the Greeks through Al Mina, but any influence on Hera's chariot is improbable.

89 Przeworski, op. cit. 142, and Wainwright, , The Coming of Iron, Antiquity X, 5–13.Google Scholar

90 Monuments, 190–1, 509–15.

91 Monuments, 119–20.

92 BSA XLII, 93, fig. 7, and 100, fig. 9d.

93 At Smyrna the Protogeometric pottery corresponds with a stage when iron was in general use in Attica, (JHS LXXII, 104).Google Scholar I was allowed access to the Register of finds at the end of the excavations, and I found that, except for bronze arrowheads which continued side by side with iron until c. 600 B.C., no bronze tool or weapon had come from the Greek levels. This is of course subject to correction when the material is published.