^† I am indebted to Dr. H. W. Catling for his constant encouragement. Drs. J. T. Killen and J.-P. Olivier read an earlier draft of this paper and gave much valuable criticism. I thank Miss Emma L. Faull and Mr. David Hibler for correcting the English.

¹ See Chadwick, J., The Mycenaean World (Cambridge 1976)Google Scholar (henceforth Chadwick, MW) 175–8; Schmitt-Brandt, R., SMEA 7 (1968) 69–96Google Scholar; Deger-Jalkotzy, S., E-QE-TA. Zur Rolle des Gefolgschaftswesens in der Sozialstruktur mykenischer Reiche (Vienna 1978)Google Scholar (henceforth Deger, Eqeta) 14–51; Palmer, L. R., Minos 4 (1956) (henceforth Palmer, 1956) 120–45Google Scholar; and id., Armées etfiscalité dans le monde antique (Paris 1976) 35–52.

² See, however, Hooker, J. T., in SMEA 23(1982) 209–17.Google Scholar

³ Chadwick, MW 177 believes that the Pylian eqeta were in command of a regiment of the army, whereas Deger, Eqeta 40 thinks that they controlled both troops and officers and acted as liaison officers.

⁴ For the different interpretations see especially Lindgren, M., The People of Pylos ii (Uppsala 1973)Google Scholar (henceforth Lindgren, PoP) 58, 75, 86, 90, and 149; Chadwick, MW 115; Deger, Eqeta 26; Gschnitzer, F. in Donum Indogermanicum (Festschrift A. Scherer) (Heidelberg 1971) 90–106.Google Scholar

⁵ See e.g. Schmitt-Brandt, R., SMEA 7 (1968) 80–1Google Scholar who concludes that the different designations derived either from toponyms or from the name of the equipment which the soldiers carried (or from both). See also Lindgren, P0P86, 91, 150 and Palmer, 1956 126, 145. The Korokurajo could have come from Κροκύλεια or from Κόρκυρα. The Kurewe could have a connection with the island Skyros or the name could be derived from words such as σκῦλα ‘spolia’ etc. The whole thing is still under discussion.

⁶ See Hiller, S., Ziva Antika 15 (1975) 388–412.Google Scholar

⁷ See Lencman, J. A., Die Sklaverei im mykenischen und homerischen Griechenland (Wiesbaden 1966) 200–1Google Scholar and especially Gelb, I. J., JNES 32 (1973) 91–4.Google Scholar P.O.W.s were inter alia used as mercenaries in Israel, Egypt, Mesopotamia, Anatolia, and elsewhere.

⁸ See Chadwick, J., BICS 26 (1979) 130Google Scholar and Foster, E. D., Minos 17(1981) 67–121.Google Scholar

⁹ See e.g. Evans, A. J., The Prehistoric Tombs of Knossos (London 1904)Google Scholar; Vandenabeele, F.-Olivier, J.-P., Les Idéogrammes archéologiquesdu linéaire B (Paris 1979)Google Scholarpassim; id., BCH 98(1974) 23–5.

¹⁰ The fragments mention ]o-ka[ (Xd 70), o-ka-ra[ (X 7631), ku-re[ (X 7668), u-ru-pi-ja[ (Fh 〈392〉) and ke-ki[ (F 7362 v.). Apart from ke-ki[, which could have stood for the anthroponyms Keki, Kekido, or Kekijo, all words are repeated only in the Pylian OKA set.

¹¹ The archive in the Room of the Chariot Tablets in the West Wing of the palace is regarded as a scribal school by Chadwick, J. (BICS 13 (1966) 103–4)Google Scholar. The fact that the find circumstances were remarkable, that the graphic style is not found in any other Knossian archive, that the sealings of the archive and the contents of the tablets correspond, that the information does not betray obvious ‘lies’, that there are sufficient cross-references to the other Knossian archives, that there are no real exercise tablets, all point away from Chadwick's;s interpretation. The archive is indeed awkward, but seemingly for other reasons. See KT (4) for text and comments.

¹² See Chadwick, J. in Greece and the Eastern Mediterranean (Festschrift F.Schachermeyr) (Berlin-New York 1977)Google Scholar (henceforth Chadwick, GEM) 108–9. He is not certain whether the names in lines 3 and 4 are nominatives plural of ethnics or datives singular of personal names. It seems likely, however, because of the principle of parallelism, that we are dealing here with ethnics.

¹³ Chadwick, GEM 106–9.

¹⁴ Examples of Near Eastern languages in the first millennium B.C. which mention the Ionians are the Hebrew Yāwān (Ezek. 27: 13; Gen. 10: 2), the Assyrian Tawnai, the Persian Yauna, and the Egyptian jwn(n)3. The modern Turkish word for ‘Greek’ is still Yunanli.

¹⁵ They are mentioned in the Iliad together with the Boeotians and the Locrians. Herodotos (7. 94; 8. 44; 9. 26) places the prehistoric Ionians especially in Attica and the Peloponnese.

¹⁶ The third sign in line 6 is possibly ke or je. Possibilities are few, e.g. opikerijo, a place-name in the Pylian tablets (An 615–724) (cf. DMG ² 188). In Knossos there is the obscure word otake[ on fragment Xd 7547, again a tablet from the Room of the Chariot Tablets.

¹⁷ There is no reason why kurewe should not mean the sameas in Pylos, since both are accompanied by the VIR ideogram and high numbers.

¹⁸ See e.g. the OKA set in Pylos and possibly the newly discovered tablet Al 7 from Tiryns (Godart, L., Killen, J. T., and Olivier, J.-P. in AA 1983, 418Google Scholar).

¹⁹ The tablet is missing but was found in the Room of the Column Bases in the West Wing of the palace. Like the other tablets of the set, it was probably written by scribe 141.

²⁰ Deger, Eqeta 86–8. This may have been during the harvest season.

²¹ See Seymour, T. D., Life in the Homeric Age (London 1907) 589.Google Scholar

²² Breasted, J. H., A History of Egypt (New York 1905) 336, 449–50Google Scholar and Gelb, I. J., JNES 32 (1973) 91–4.Google Scholar

²³ See Vercoutter, J., L'Egypte et le monde égéen préhellénique (Cairo 1956)Google Scholarpassim, esp. 29–30, 423–30 and Faulkner, R. O., JEA 39(1953) 32–47.Google Scholar

²⁴ PM ii. 755–7, pl. XIII (LM IB).

²⁵ See KT (4) 219–23 for texts and comments and Killen, J. T. in BSA 59 (1964) 1–15Google Scholar; Killen, J. T. in Colloquium Mycenaeum (Neuchâtel 1979) (henceforth Killen, CM) 151–81.Google Scholar

²⁶ Deger, Eqeta 98–105.

²⁷ Killen, CM 153 and Deger, Eqeta 99 (references).

²⁸ Deger, Eqeta 99.

²⁹ Eqesija in Ld 571–572–575–583; kesenuwija in Ld 573–574–585 and kesenewija in Ld 649.

³⁰ DMG ² 319.

³¹ See KT (4) 18, 28, 36, 38–9 for texts and comments.

³² Deger, Eqeta 92–4 and Melena, J. L., Studies on Some Mycenaean Inscriptions from Knossos dealing with Textiles (Salamanca 1975) (henceforth Melena, Textiles) 37–42.Google Scholar

³³ This interpretation is not accepted by Melena, Textiles 42.

³⁴ Lindgren, PoP 50–1.

³⁵ Deger, Eqeta 203–4.

³⁶ Deger, Eqeta 94–8 and Melena, Textiles 33–6.

³⁷ In this way there would have been about 13 eqeta at Knossos and one eqeta at Tylissos and Utano. The numbers mentioned in these tablets are surprisingly similar; both are multiples of 24 (product minus 3): i.e. 24× = 216–3 = 213 (B 1055) and (24 × 10 = 240–3 = 237 (B 807). Coincidence?

³⁸ Deger, Eqeta 83–92 and Melena, Textiles 26–49 (references).

³⁹ Either Katara or Samara, if two signs are lost as in line 2, Era or Qara if only one sign is lost; the former places were situated more westerly, the latter two in Central Crete.

⁴⁰ Sijaduwe is the only possible completion of the word. The identification is supported by the occurrence of an ijereu (ἱερεúς) in Sijaduwe in Am 821 and the sheep belonging to the Potnia at Sijaduwe in the D-series. Melena, Textiles 31, is of a different opinion.

⁴¹ Also mentioned as a shepherd on Da 1108.

⁴² DMG ² 420, 564.

⁴³ Melena, Textiles 48 and Deger, Eqeta 85–6. The classical word for mercenary was sometimes ἔμμισθος (cf. Philo v. 29).

⁴⁴ Deger, Eqeta 196–8.

⁴⁵ DMG ² 218, 553 (II. 11. 779); Deger, Eqeta 104; Melena, Textiles 44–5: Kesenuwo is probably a man's name in PY Cn 286.

⁴⁶ If one accepts–as we do–that Crete was partly conquered only two generations before the palace fell, it is likely that the ‘foreign guests’ were mercenaries and not local Cretan tribes or allies. For mercenaries, see further Sandars, N. K., The Sea Peoples (London 1978) 49–50, 94Google Scholar and Parke, H. W., Greek Mercenary Soldiers (Oxford 1933) 3–6Google Scholar ‘… as long there were wealthy masters, there were also masterless men who for need or greed or adventure would sell their swords and even their lives’….

⁴⁷ Herodotus uses ξεîνοι to indicate Carian and Ionian mercenaries in Egypt (2. 163, 169). See Bétant, E.-A., Lexicon Thucydideum ii (Hildesheim 1961)Google Scholar for the use of the words ξéνοι, ξενοτροφεῖν, and ξεναγός in connection with mercenaries.

⁴⁸ Which would mean that ξéνος still had a double meaning (‘guest’ and ‘foreigner’). Cf. Gschnitzer, F. in Colloquium Mycenaeum (Neuchâtel 1979) 128–9 n. 67.Google Scholar

^‡ Acknowledgements. I am grateful to the following people for helpful comments on early drafts of this paper: Miss J. M. Cocking. Mr. J. Driessen, Miss E. L. Faull and Mr. J. A. MacGillivray. Miss P.-A. Mountjoy gave help when needed with pottery and Mr. D. Smyth with the drawings. Miss N. K. Sandars, Mr. M. R. Popham, and, particularly, Dr. H. W. Catling gave advice and detailed criticism which enabled me to produce the final draft; to them I owe a special debt. Any work on this subject must begin by acknowledging the important work carried out by Miss Sandars over twenty years; constant reference to her article (AJA 67 (1963) 117–53) is required throughout this paper.

Abbreviations not standard but used here are:

PTKA. J. Evans, Prehistoric Tombs at Knossos (1906)

Royal Tombs A. W. Persson, The Royal Tombs at Dendra near Midea (1931)

⁴⁹ Popham, M. R., BICS 23 (1976) 119–20.Google Scholar

⁵⁰ Only seven of the one hundred Zapher Papoura tombs at Knossos contained swords or dirks.

⁵¹ See Desborough, V. R. d'A., The Last Mycenaeans and their Successors (Oxford 1964) 67Google Scholar, where swords, dirks, and daggers are defined according to length. Swords: more than 0·50 m; dirks: 0·30–0·50 m; daggers: less than 0·30 m. From the supplementary catalogue it will be seen that only a handful of C and Di weapons are less than 0·40 m.

⁵² There is iconographic evidence from both the fifteenth and twelfth centuries to show that the sword/dagger was also used in hunting. A seal from Shaft Grave III at Mycenae depicts a man fighting a lion with a sword (CMS I, 20 no. 9). The Griffin Slayer (Enkomi) and the Lion Slayer (Kouklia), both carved on Late Cypriot ivory mirror-handles are similar scenes (Buchholz, H.-G. and Karageorghis, V., Prehistoric Greece and Cyprus (1973) 480–1.Google Scholar). However, spears and bows are much more common in such scenes.

⁵³ e.g. Zapher Papoura Tomb 66 (PTK 71) and Dendra, Chamber Tomb 2 (Royal Tombs 91–107).

⁵⁴ AJA 67(1963) 128. Examples could be the Di sword from Zapher Papoura (ZP) Tomb 36 and the huge Ci swords from the Dendra Tholos and Ialysos New Tomb (NT) 4.

⁵⁵ Lorimer, H. L., Homer and the Monuments (London 1950) 272, 274.Google Scholar

⁵⁶ CMS i, 22 no. II; 23 no. 12; 27 no. 16.

⁵⁷ AJA 67 (1963) 117–53. Other, more recent articles include: Kilian, K., ‘Nordgrenze des ägäischen Kulturbereiches in mykenischer und nachmykenischer Zeit’, Jahresbericht des Instituts für Vorgeschichte der Universität Frankfurt a. M. (Munich 1976) 112–19Google Scholar; Matthaus, H., ‘Two Mycenaean Bronzes’, BSA 74 (1979) 163–73Google Scholar; Foltiny, St., ‘Schwert, Dolch and Messer’ in Archaeologica Homerica vol. 1, chapter E, part 2 (1980) 231–74.Google Scholar For the purposes of this article, constant reference is required to Sandars' paper; full references are not given here to swords which appear in her catalogue.

⁵⁸ Sandars's format has been followed here. The numbers against each sword in the Catalogue refer to those in my Supplementary Catalogue, which lists all examples of types C, D, E, and G. These numbers are often used in the text thus: (Ci 10) i.e. sword no. 10 of type Ci from Dendra Tholos. The numbers on the distribution maps (figs. 2–4 and 6) correspond to the Supplementary Catalogue.

⁵⁹ By no means all are well dated, but none have to be later than LM/LH IIIA₂.

⁶⁰ e.g. dagger from Poros Wall founder's hoard, Mycenae (Dii5).

⁶¹ For distribution of type A and B see Kilian (1976) fig. 4.

⁶² Several swords were found with the last burial on the floor of the tholos, and Sandars mentions that one may have been of type A (AJA 65 (1961) 26).

⁶³ Spears are more common than swords in tombs. In addition, there is iconographic evidence from Knossos for the use of the spear by groups of men, presumably soldiers (Knossos Fresco Atlas pl. iv, figs. 1–2).

⁶⁴ AJA 67 (1963) 128.

⁶⁵ One-piece spearheads have been found in the following areas: Knossos, Archanes, Argolid, Attica, Boeotia, Thessaly, Elis, Kos, Rhodes. For details of provenance and variations in form see Avila, R. A. J., ‘Bronzene Lanzen und Pfeilspitzen der griechischen Spätbronzezeit’, Prähistorische Bronzefunde (PBF) v. I.Google Scholar; Höckmann, O., ‘Lanze und Speer im spätminoischen und mykenischen Griechenland’, Jahrbuch des Römisch Germanischen Zentralmuseums Mainz 27 (1980) 13–158.Google Scholar

⁶⁶ Type E is found with types C and/or D in the following tombs: Ayios Ioannis (A. J.) Tomb 2 (Di), Mavrospelio Tomb XVIII (Di), Sellopoulo Tomb 4 (Di), Argive Heraion Tomb 25 (Di), Pylos Chamber Tomb E (Di),Tanagra Tomb 16 (Cii), Ialysos NT₄ (Ci and Dii).

⁶⁷ Eii I from the Unexplored Mansion.

⁶⁸ Metal vessels have been found with types C, D, E, and G in the following tombs: ZP 14 (bronze); ZP 36 (bronze); ‘Silver Cup’ Tomb, Sellopoulo Tomb 4 (bronze); AJ. Tomb 2 (gold cup); Fourni, Archanes (silver cup); Phaistos, ‘Tombe dei Nobili’, Tomb 8 (bronze); Dendra Tholos (gold, silver and bronze); Dendra Chamber Tomb 2 (bronze); Dendra Chamber Tomb 12 (silver and Bronze); Mycenae Tomb 78 (2 silver cups); Ialysos OT 4 (bronze).

⁶⁹ BSA 69 (1974) 146.

⁷⁰ Dendra Chamber Tomb 2 (Royal Tombs pl. XXXIV).

⁷¹ Spyropoulos, Th., AAA 7 (1974) 322.Google Scholar

⁷² Iakovides, Sp., Perati (Athens 1970) 360 fig. 158.Google Scholar

⁷³ BSA 69 (1974) 253.

⁷⁴ See Benton, S., PPS 18 (1952) 297Google Scholar; Catling, , PPS 22 (1956) 110Google Scholar; Sandars, , AJA 67 (1963) 135–7CrossRefGoogle Scholar; Spyropoulos, Th., Ὓστερω Μυκηναϊκόι Ἑλλαδικοί Θησαυροί (1972) 15–16 figs. 14 15. pls. 7γ, 8α–β.Google Scholar

⁷⁵ AJA 67 (1963) 126.

⁷⁶ BSA 69 (1974) 243.

⁷⁷ Mylonas, G., Ὁ Τάφικος Κύκλος Β τῶν Μυκηνῶν (Athens 1973) 419.Google Scholar

⁷⁸ BSA 69 (1974) 252.

⁷⁹ ZP 36 (Ci and Di); ZP 42 (Di); ZP 44 (Ci); Sellopoulo Tomb 4 (2 Di); NHS Tomb II (Di); Phaistos, ‘Tombe dei Nobili’, Tomb 8 (Ci); Dendra Tholos (Ci and Di); Dendra Tomb 12 (Ci); Mycenae Tomb 91 (Di); Mycenae 1890 hoard, Athens 2536 (Gi); Apollo Maleatas (Di). It is highly likely that some other undecorated swords came from this workshop e.g. Di21 from Mycenae Tomb 91.

⁸⁰ PBV V, t, 871 pl. 34, 40 pl. 6. There is one sword (unpublished and therefore not included in n. 79) which is decorated with exactly the same combination of motifs as the Archanes spearhead. It comes from Crete.

⁸¹ Chadwick, , Killen, , and Olivier, , The Knossos Tablets (Cambridge 1971) 267–9.Google Scholar

⁸² PM IV 854 fig. 837.

⁸³ AJA 67 (1963) 121.

⁸⁴ Some of the ‘northern’ finds of swords are probably genuine imports (Galatin, Romania (Cii12); Mati, Albania (Di40)). Such imports may have been the source of inspiration for the a typical C and D swords outside Greece.

⁸⁵ BSA 47 (1952) 245.

⁸⁶ BSA 69 (1974) 253, 255.

⁸⁷ BSA 69 (1974) 256.

⁸⁸ BSA 28 (1926–7) 243–96. BSA 53–4 (1958–9) 194–262.

⁸⁹ BSA 28 (1926–7) figs. 17–18, Tomb VII, Chambers A and B.

⁹⁰ Hood, and Smyth, , Knossos Survey (1981) 12Google Scholar.

⁹¹ PTK II; Dickinson, , Origins of Mycenaean Civilization (1977) 51.Google Scholar Also see Dickinson, , BSA 78 (1983) 56Google Scholar for precise definition of shaft grave and parallels.

⁹² Hood and Smyth (1981) 26.

⁹³ It is difficult to separate the features of a burial which may be indicators of military, as opposed to social, status and vice versa; both may depend on economic status and be inseparable archaeologically, indeed this would not be at all surprising. The quantity, quality, and type of weapons should be the primary evidence for military standing, and the accompanying goods for social and/or economic status. In practice, the finest arrays of weapons tend to be found in graves which are also rich in non-military grave gifts, e.g. ZP 36; ZP 14; Sellopoulo Tomb 4 Burial I; ‘Tombe dei Nobili’, Tomb 8; Dendra Tholos; Dendra Tomb 12; Mycenae Tomb 91. If Evans is right when he suggests that the main burial in ZP 14 had been placed in a wooden sarcophagus, which was subsequently removed along with most of the more valuable objects such as jewellery (PTK 38), it is possible that this tomb was the richest in the cemetery; the large elaborate doorway and chamber, and the exceptionally long dromos (for Knossos) would support this hypothesis. Indeed, the robbed Isopata Royal Tomb may have been one of the few Knossian sepulchres to rival it.

⁹⁴ e.g. ZP 42, ZP 43, ZP 55, ZP 98, Mavro Spelio Tomb XVIII.

⁹⁵ e.g. ZP 7 (PTK 25–7), ZP 66 (PTK 71–2), ZP 99 (PTK 87–90), Sellopoulo Tomb 3 (BSA 69 (1974) 195–257), Tholos, Archanes (PZ 45 (1970) 133).Google Scholar A major problem is that some of the rich graves are likely to be those of women since one would not expect them to have been buried with weapons.

⁹⁶ Zervos, , L'Art de la Crète (1956) pls. 536–7.Google Scholar

⁹⁷ Savignioni, L., Scavi e scoperte nella necropoli di Phaestos (1905) 37.Google ScholarRoyal Tombs 13–14 figs. 12, 16, pl. xviii 2.

⁹⁸ ZP 7 (PTK fig. 22), ZP 66 (PTK 76 fig. 85).

⁹⁹ PZ 45 (1970) 135–219.

¹⁰⁰ The Knossian warrior burials can be divided into three groups although the boundaries between them are, to a certain extent, arbitrary. Group I: ZP 14(?), ZP 36, Sellopoulo Tomb 4 Burial I, Group 2: Sellopoulo Tomb 4 Burial 2, Ayios Ioannis Tomb 2, ‘Silver Cup’ Tomb, Group 3: ZP 42, 43, 44, 55, 98, Ayios Ioannis Tomb 1, NHS Tomb II, NHS Tomb V, Mavro Spelio Tomb XVIII.

¹⁰¹ Mycenae, Argive Heraion, Dendra-Midea.

¹⁰² The evidence for a Mycenaean presence in Rhodes from LHIIB is more conclusive than that from Crete. The important point in the argument is the beginning of a cemetery of chamber tombs at Ialysos; much of the pottery of LH IIB–IIIA is likely to be Argive. See Mee, C., Rhodes in the Bronze Age (1982) 81.Google Scholar

¹⁰³ See Demakopoulo, K. and Konsola, D., Archaeological Museum of Thebes. Guide. (Athens 1981)Google Scholar for excellent description of finds and full bibliography.

¹⁰⁴ See Catalogue of Ships, Iliad 381 refers to the great wealth of Orchomenos. Also Oxford Classical Dictionary (Oxford 1970) under various sites.

¹⁰⁵ The tholos is discussed by Pelon, O., Tholoi, tumuli et cercles funéraires (Paris 1976) 233–7 pl. cx.Google Scholar

¹⁰⁶ In this respect, it is unfortunate that the date of the stirrup-jars with painted Linear B signs has not been solved. For the latest discussion see BSA 75 (1980) 95–7.

¹⁰⁷ PTK 20.

¹⁰⁸ AJA 67 (1963) 133.