¹ My thanks are due to the Managing Committee of the British School of Archaeology at Athens for awarding me the Macmillan Studentship for 1962 and 1963, when the substance of this article was written. I am also grateful to Mr. V. R. d'A. Desborough and to Mrs. E. French, who have discussed various points with me, and to Dr. St. Alexiou, Ephor of Crete, and his staff, who facilitated in every way my study of the material, and undertook the restoration of the crater.

Pendlebury's observation that one of the greatest necessities for Minoan archaeology is the excavation of a stratified L.M. III site is as true now as when it was written twenty-five years ago (AC 253).

The following abbreviations are additional to those normally used:

Athens = Broneer, , ‘A Mycenaean Fountain on the Athenian Acropolis’, Hesperia viii (1939) 317–429.CrossRefGoogle Scholar

Borda = Borda, , Arte cretese-micenea nel Museo Pigorini di Roma (Rome 1946).Google Scholar

Delphi = Lerat, , BCH lix. 329–75.Google Scholar

Desborough = Desborough, , The last Mycenaeans and their Successors (Oxford 1964).Google Scholar

FK = Matz, , Forschungen auf Kreta (Berlin 1951).Google Scholar

Karpki = Seiradaki, , ‘Pottery from Karphi’, BSA lv. 1–37.Google Scholar

Kastri = ‘Palaikastro VI’, pp. 278–99 above.

LDPK = Popham, , The Last Days of the Palace at Knossos (Lund 1964).Google Scholar

Mallia = Dessenne, and Deshayes, , Fouilles exécutées à Mallia, Maisons ii (Paris 1959).Google Scholar

Palestine = Bliss, Macalister, and Wunsch, Excavations in Palestine 1898–1900.

PKU = Bosanquet and Dawkins, The Unpublished Objects from the Palaikastro Excavations 1902–6, BSA Supplementary Paper 1.

Sinda = AJA lii (1948), ‘Archaeological News’ (531 and pl. lvii).

Tarsus = Excavations at Gözlü Kule, Tarsus ii (Princeton 1956).

² No. 67 in the catalogue, Fig. 2, Plate 85a, considered L.M. IIIB, and another sherd difficult to classify with certainty.

³ Subsequently restored and illustrated in Archaeological Reports for 1963–4, fig. 31.

⁴ BSA vi (1899–1900) 17 for a description of the Central Clay Area. It was apparently the filling of soil below the great stairway of which the steps had been robbed.

⁵ Pendlebury's Guide lists only three boxes.

⁶ BSA vi (1899–1900) 91; he also remarks that ‘there are not many specimens of the later class of Mycenaean ware, which under incoming Geometric influence developed a system of metope divisions and similar ideas foreign to the true Mycenaean spirit’ (89). Some sherds from the re-occupation bastion in the S. Propylaeum are L.M. IIIB; most are L.M. IIIA. Geometric pottery could have come from this general area since other Geometric fragments were found in the adjoining ‘site of Greek Temple’, Hartley, , BSA xxxi (1930–1931) 92.Google Scholar

⁷ The author has published the whole vases of the re-occupation period in LDPK, where he has outlined his reasons for believing that the desertion of the Palace site took place in L.M. IIIB.

⁸ Heraklion Museum suffered damage in the earthquake of 1926 and the material has been rehoused since the last war. Some suspicion arises from the close parallels with some of the L.M. III material from Phaistos and Ayia Triadha.

⁹ e.g., Hall, , Vrokastro 92Google Scholar, PKU 76, and Gournia 46.

¹⁰ See above on Palaikastro, , Kastri, p. 282.Google Scholar

¹¹ L.M. I, Vrokastro, fig. 46; L.M. IIIA, PM iv, fig. 965c, BSA vi (1899–1900) 74, fig. 19.

¹² Desborough 171 and MP 51.

¹³ FS 284 and 285.

¹⁴ See above on Palaikastro, , Kastri p. 284.Google Scholar On the other hand, there is no close Mycenaean parallel, though the shape could easily have developed out of the usual L.H. IIIB krater, FS 281.

¹⁵ L.M. II/IIIA 1, BSA xlvii. 263, fig. 9: 1. L.M. IIIA 2, PKU, fig. 68 and PM iv, fig. 309a from the Dictaean Cave. L.M. IIIB, carinated, AM xi, pl. 3, and IIIB/C Vrokastro, fig. 49: i, shallow, AJA xlii (1938) pl. xxvii: 1. and IIIB/C ibid., pl. xxvii: 2. L.M. IIIC, MA xii. 115, fig. 45 and late examples, Karphi, fig. 18 and pl. 10b, Nécropoles du Mirabello pls. xli and xv, D 30. Sub-minoan, Vrokastro, fig. 48A. A fragmentary kylix with rounded profile of IIIB type from the Little Palace is illustrated in Plate 85d.

¹⁶ MP 29.

¹⁷ See above on Palaikastro, , Kastri, p. 284Google Scholar; Athens 391–2; Karphi 16.

¹⁸ Not illustrated here. See LDPK 13, no. 6.

¹⁹ LDPK, pl. 2b; PKU 73.

^19a Mrs. E. French kindly informs me that at Mycenae at least the reserved band does not appear until advanced IIIC. The reserved disk is introduced at Knossos early in IIB if not in late IIIA.

²⁰ There is some doubt about the provenance of these sherds, which are a selection from a box in the Strati-graphical Museum titled ‘K o8 Little Palace. Earth above Western Part; P I 8 (?)’. This is presumably the un-numbered 1908 test mentioned on p. 30 of the Guide. The label, however, in the box says, ‘NW Wagers-Area of Roman Amphitheatre’ which is T. P. 2 of 1908 mentioned on the same page of the Guide. Pendlebury from his diagnosis of its contents, on the last page of the Guide, clearly took it to be the former. The sherd on Plate 86c with the antithetic spiral has been analysed, and despite its Mycenaean appearance was found to belong to the Cretan group of clays, Catling, H.et al. BSA lviii. 115.Google Scholar The sherd is no. 9 on the list from the Stratigraphical Museum, and in Table 6, p. 112.

²¹ OpArch iii. 196.

²² See catalogue for references.

²³ PKU 97, fig. 81.

²⁴ MA xiv, pl. 1 (one of the bird alabastra from Kalyvia); there are several unpublished examples in the Stratigraphical Museum.

²⁵ See above on Palaikastro, , Kastri, p. 290Google Scholar and figs. 8 i–j and 9ƒ–g.

²⁶ AJA xii, pl. xxvii: 4 and 8; FK, pl. 48.

²⁷ e.g. CVA British Museum 1, pl. 15: 20 and Taylor, , Myrtou Pyghades 43Google Scholar, fig. 20 (209). CVA British Museum 1, pl. 5: 11, on the other hand, is a specifically Levanto-Mycenaean shape.

²⁸ Karphi, 31.

²⁹ Cf. Furumark, , OpArch iii, 215 and 228–9Google Scholar and MP 301. Considering later examples Brock, Fortetsa 9. 3, suggests the motive originated from the cuttlefish; Vrokastro fig. 49: 1 and Mallia, Maisons ii, pl. lxxia could be used to support this view. Alexiou, , KretChron. (1950) 444 and 449Google Scholar, suggests that it might symbolize the bucranium.

Furumark, , MP 458Google Scholar, speaks of the Phaistos krater, MA xii, pl. 8: 4, which is similar to our no. 62, as being ‘clearly dependent on Mycenaean prototypes’.

³⁰ See catalogue no. 62 for references.

³¹ Furumark, , MP 309Google Scholar, says of the Mycenaean version in L.H. III A2 that its immediate prototype is evidently the L.M. III A2 type. The main evidence is one stirrup-jar from Palaikastro, (BSA viii. 303, fig. 19 = FM 23d)Google Scholar but there are difficulties in this view. The date of the Palaikastro vase is by no means certain; that is to say it is uncertain whether it is earlier than any of the Mycenaean versions; there are very few examples in the extensive III A2 deposits from Knossos and those which are stylistically the earliest seem to be in imitation of Mycenaean vases.

³² Desborough 7 and Appendix B.

³³ Palace, LDPK passim; Zafer Papoura, PT; Gypsadhes, , BSA liii–liv 194–262.Google Scholar There were IIIC burials in the earlier built tomb at Isopata, PT, fig. 122, and in the Kephala, tholos BSA li. 74–80.Google Scholar Some IIIB and IIIC sherds were published in BSA xxxi, pl. xix.

³⁴ LDPK 9 and vase no. 8.

³⁵ The flower forms are similar, cf. especially LDPK vases nos. 9, 13, and 15. None of the octopus stirrup-jars from the Palace shows any signs of the IIIC close style, ibid., vase 10.

³⁶ The latest tomb in the Zafer Papoura cemetery seems to be T 54. It was much disturbed; vases b and c could be IIIC but vase a looks typical IIIB. They are all called IIIB by Furumark, , MP Chron 105.Google Scholar At Gypsadhes, the latest IIIB tomb is T XI; it is suggested in the report that two of its vases might be IIIB/C. Tombs VIa, VII, and IX are advanced IIIC.

³⁷ See above on Palaikastro, , Kastri, p. 281 and nn. 63 and 64.Google Scholar

³⁸ LDPK 9 where it is suggested that the closest parallels to the Palace desertion deposits are the latest vases from House E at Mallia; see especially vase 11. Desborough 170 suggests, on admittedly slight evidence, that House E may have been occupied into IIIC. There is nothing there, however, that could not be IIIB.

³⁹ MA xii, figs. 45–55 and pl. viii: 4.

⁴⁰ See above on Palaikastro, , Kastri, p. 283.Google Scholar

⁴¹ Cf. Desborough 185 for similar doubts.

⁴² See above on Palaikastro, , Kastri, p. 281.Google Scholar

⁴³ There are two gaps in the use of the Gypsadhes cemetery. The one in L.M. III A2 and the other at the end of L.M. IIIB. The first may reflect disturbance at Knossos after the destruction of the Palace and other buildings. The second, coupled with the abandonment of the Zafer Papoura cemetery, may be a sign of a similar disturbance at the end of IIIB.

⁴⁴ Desborough 191–5 has carefully stated the evidence. He is on the whole more concerned with disturbance in L.M. IIIC and the reasons for the foundation of Karphi; he considers (p. 229) that Crete was no more affected at the time of the invasions of the Mainland in IIIB than the other islands. Basic to this belief is his view that L.M. IIIB continued after the introduction of IIIC characteristics on the Mainland. But, as has been pointed out above, the IIIB abandonments at Palaikastro, Gournia, and Knossos are marked by the presence of Mainland IIIB vases; so at the time of these events at least L.M. and L.H. IIIB appear to be contemporary. The real problem is: who caused these abandonments? A threat from the sea seems the most probable in the case of Gournia, Palaikastro, and Mallia, all of which are very near the coast. The Mycenaeans are obvious candidates, repeating what one imagines happened in L.M. IB. The difficulty is that the new settlements, e.g. Kastri and Vrokastro, are predominantly Minoan in character as though the new settlers had come from some centre of fusion between Mycenaeans and Minoans elsewhere on the island. But this is guesswork.

⁴⁵ Mrs. L. French in an unpublished thesis on the development of Mycenaean terracotta figurines ascribed the figurines from Phaistos and Ayia Triadha to her groups B and C of the late ‘Psi’ type; these she dates to late IIIB into IIIC and to IIIC respectively (pp. 124–5, 127, and 129 of her manuscript in the British School Library). She interprets the distribution of these late figurines to indicate that the inhabitants of the Mainland spread over-seas after the fall of many of the Mainland cities.

⁴⁶ See Furumark's views on the origin of Philistine, pottery, OpArch iii. 264.Google Scholar

⁴⁷ Furumark loc. cit. remarks ‘It has been shown that several features typical of the Mainland style appear suddenly in the pottery of Rhodes and Crete just after the fall of Mycenae and that this Hellado-Mycenaean strain becomes very strong in the subsequent Myc. IIIC: 1c phase, resulting in a transformation of the local styles of the two islands. This process must be the result of an emigration from the Greek Mainland.’ Cf. Desborough's, similar but more tentative conclusions, Desborough 235–6.Google Scholar It has been one of the purposes of this article to suggest that some of these new and Mycenaean features may well have been introduced into the island at an earlier stage in IIIC than Furumark and Desborough, on the evidence available to them, proposed.