¹ The excavations were carried out under the auspices of the British School at Athens; we are most grateful to the chairmen, R. A. Tomlinson and Gerald Cadogan, and to the Managing Committee for their continued support and encouragement. We are also very much indebted to the Directors, E. B. French and R. A. Tomlinson, and the School's staff in Athens for their assistance throughout the years. Permission to excavate was given by the Greek Ministry of Culture, for which we are grateful to Dr Y. Tzedhakis. Dr N. Papadakis in Siteia very much facilitated the work and provided the supervision and collaboration of Ms E. Tzangarakis in 1994 and Ms D. Koïnakis in 1996. The necessary financial support was provided by the Institute for Aegean Prehistory, whom we continue to value as our primary sponsor. The aerial photograph in PLATE 38 was taken by J. Driessen who acknow ledges the support of the Belgian F.W.O.

The excavations were directed by L. H. Sackett and J. A. MacGillivray. The architects were J. M. Driessen and P. Jerome. S. Dandali was administrative assistant. French supervisors in 1994 were T. Cunningham. S. Hemingway, J. Mason, M. Naughton, I. Schoep, S. Thorne, assisted by A. Harris, D. and S. Jacovini. R. Krieger, K. Sendler, S. You, A. Wasse, and V. Isaakidou; in 1996 they were T. Cunningham. E. Hatzakis, A. Marcar, P. Tomkins and S. Wallace assisted by V. Kyriakidis, L. Preston, D. Romanou, E. Sackett, K. Sendler, and S. Vàzquez. Video notebooks were kept by S. Moss in 1994 and A. Lesk in 1996. The apotheke was run by H. Parton and A. Sackett, assisted in 1994 by E. Bonn and Robin O'Nan. The computer archives constructed by B. Howenstein were implemented by C. Sarther and J. Romalo in 1994 and J. Rempel in 1996. Conservators were P. Harrison, A. Brysbaert, and M. Moak in 1994 and N. Karanikolas, M. Naunton and M. Tziotziou in 1996. Pottery processing was supervised by J. A. MacGillivray, assisted in 1994 by A. Marcar and in 1996 by S. Janssen. Bone was processed by C. Walker in 1994 and S. Crowther in 1996; soil samples were processed in 1994 by J.Jones assisted by J. Stewart and T. Aretaki and in 1996 by V Isaakidou. Small finds and tools were studied by D. Evcly. Drawings are by P. Murray assisted in 1994 by R. Moak, A. Currelly, and S. Bell. C. Doherty is geological consultant. Twenty workmen and technicians were supervised by N. Daskalakis.

This report was written by J. A. MacGillivray and L. H. Sackett incorporating texts byj. M. Driessen. Special reports by J. M. Driessen, Ilse Schoep and Eleni Hatzaki are included as separate headings.

Special abbreviations:

PK I–VII = ‘Excavations at Palaikastro I’ to ‘VII’, in BSA 8 (1901–1902), 286–316Google Scholar; 9 (1902–3), 274–387; 10 (1903–4), 192–321; 11 (1904–5), 258–308; 12 (1905–6), 1–8; 60 (1965), 248–315; and 65 (1970), 203–42 respectively.

PK 1986–8, 1990–1 = ‘Excavations at Palaikastro, 1986’ to ‘1988’ in BSA 82 (1987), 135–54Google Scholar; 83 (1988), 259–82; 84 (1989), 417–45; and ‘1990’ to ‘1991’ in BSA 86 (1991), 121–47; 87 (1992), 121–52 respectively.

PK Settlement = ‘Minoan settlement at Palaikastro’, in Darcque, P. and Treuil, R. (eds), L'Habilat Égéen préhistorique (BCH supp. 19, Paris, 1990), 395–412Google Scholar.

PK Survey = ‘An archaeological survey of the Roussolakkos area at Palaikastro’, BSA 79 (1984), 129–59Google Scholar.

PKV = Bosanquet, R. C. and Dawkins, R. M., The Unpublished Objects from the Palaikastro Excavations 1902–1906, Part I, (BSA supp. paper 1 London, 1923)Google Scholar.

PKL II = ‘Unpublished objects from Palaikastro and Praisos’, BSA 40 (1939–1940), 38–59Google Scholar.

TAW = Hardy, D. A. (ed.) Thera and the Aegean World, iii (London, 1990)Google Scholar.

Zakros = Platon, N., Zakros (New York, 1971)Google Scholar.

Metric height above sea level is indicated thus: 10.61⋀.

² The area includes ‘Block M’ of 1904, which was tested at its south and east perimeters by Bosanquet's team, was considered too eroded to excavate and used for dumping spoil. We found that while severe erosion had affected LM I A remains, an unusual opportunity was afforded to research the underlying early Neopalatial and earlier deposits and structures, and the evidence suggests that the area never had any later buildings. Meanwhile during the twentieth century the spread of the first excavators’ dump (covering up to a metre of bark on the old olive trees) has provided good protection for the northern part of this area, as far as the line of the old calderimi at Buildings 5 and 3.

³ PK 1990, 133–5; plan at pottery at figs. 11–13.

⁴ PK 1990, 133.

⁵ PK 1991, 125–9; plans at figs, 1 and 3, finds at figs. 2, 3, 6, 7.

⁶ For the wine-press and possible production of plaster here, see PK 1991, 126.

⁷ PK 1991, 128 and fig. 6.

⁸ ‘Metalworking in the Postpalatial period: a deposit of metallurgical debris from Palaikastro’, BSA 91 (1996), 213–52Google Scholar, pls. 39–40.

⁹ In an attempt to find out whether the unusual architectural features and chronology of Area 6 continued further north and west, as opposed to the normal pattern of LM IB destruction, two trenches were opened opposite (i.e. NE of) Block N. at EG 87 and EH 88. The results were inconclusive: fragmentary walls but no floor or good deposits were encountered in EG 87; a small compartment probably of LM I date but partly rebuilt and with plaster floor largely cut away by an LM III A1 pit was the principal feature of EH 88.

¹⁰ Material of this kind had fallen into the centre of the three small compartments found at ground level beneath.

¹¹ See n. 4.

¹² A similar representation of dolphins in a seascape is a ene on a gilt cushion seal, also from Palaikastro, acquired by Arthur Evans in 1894 (PM iV, fig. 441).

¹³ This factor may explain the unusual size of the well, if the roof of the underground cavity collapsed and it was necessary to cut a wider area to support the sides safely. In any event the result was a shaft large enough for four or five people to draw water at one time, for two or more excavators and one recorder to work together as far down as the bottom of the shaft, and (unusually) for a stratigraphical cross-section to be recorded on a stone by stone basis.

¹⁴ Thus it is planned to give priority to their publication, which is currently being prepared by Stuart Thorne and Eleni Hatzaki.

¹⁵ PK 1990, 126.

¹⁶ Plan at AR 41 (1994–1995), 69Google Scholar.

¹⁷ Excavation was not completed here and interpretation has been frustrated by a combination of severe ancient robbing and erosion, followed by the disturbance of the test trenches of Bosanquet's excavation with the inaccessibility of the street wall itself owing to close field boundaries with a complex legal history. See further discussion below.

¹⁸ Here an unburnt destruction deposit excavated in 1991 (Trench EP8₃) and associated with stone tumble had again suggested earthquake destruction (PK 1991, 128–9, fig. 7; cf. PK 1990, 135–7).

¹⁹ Reported on for the AIA at Atlanta (AJA 99 (1995), 312Google Scholar), and for the BSA at London (BICS 1995).

²⁰ PK 1990, 133.

²¹ PK 1331, 125, pl. 4 c.

²² J. A. MacGillivray, Knossos: Pottery Groups of the Old Palace Period (BSA Studies, forthcoming).

²³ PKU, 12, pls. vi d, vii.

²⁴ Poursat, J.-C., Artisans minoens: les maisons-ateliers du Quartur Mu (Etudes Crètoises 32, 1996), 66Google Scholar, pl. 331.

²⁵ PK II, 368, pl. 8.

²⁶ PKU, 12, pls. vi d, vii.

²⁷ MacGillivray (n. 22).

²⁸ Warren, P. and Hankey, V., Aegean Bronze Age Chronology (Bristol, 1989), 134–5Google Scholar, pl. 12 a.

²⁹ PK 1991, 126.

³⁰ Andreadaki-Vlasaki, M., ‘Sunken or “lustral basin: in Khania’, AAA 21 (1988), 56–76Google Scholar.

³¹ PK 1986, 150–1, pl. 24 c, d, e.

³² Bernini, L., ‘Ceramics of the early neo-palatial period at Palaikastro’, BSA 90 (1995), 59–61Google Scholar, fig. 7.

³³ Hood, M. S. F., ‘Back to basics with MMIIIB’, in Evely, D., Lemos, I. S., and Sherratt, S. (eds), Minotaur and Centaur: Studies Presented to Mervyn Popham (BAR S638, 1996) 10Google Scholar; C. F. Macdonald, ‘Notes on some LM I A contexts from Knossos’, ibid. 17.

³⁴ Marthari, M., ‘The destruction of the town at Akrotiri, Thera at the beginning of LC I: definition and chronology’, in MacGillivray, J. A. and Barber, R. L. N. (eds), The Prehistoric Cyclades (Edinburgh, 1984), 119–33Google Scholar.

³⁵ See Kuniholm, P.et al., ‘Anatolian tree rings and the absolute chronology of the Eastern Mediterranean 2220–718 BC’, Nature, 381 (27 June 1996), 782CrossRefGoogle Scholar.

³⁶ Bernini (n. 32), 62.

³⁷ Warren, P. M., ‘A new Minoan deposit from Knossos, c.1600 BC, and its wider relations’, BSA 86 (1991), 319–40Google Scholar.

³⁸ Discussed in detail below by Eleni Hatzaki.

³⁹ PK 1991, 137.

⁴⁰ As we interpret this, level 10 represents an early compact stony fill, predating the building. There followed a period of construction in quite grand style, represented here by the built drain, stone-lined and originally covered too. Levels 7, 8, and 9 represent silt, fill, and debris which accumulated after the events (perhaps seismic) which caused the demolition of the building and indeed after the demolition itself. There followed a powerful run-off of water cutting a course immediately parallel to but north of the now clogged drain. This is shown by the hollow later filled by levels 3–6. The fill was a build-up of alluvium, first a layer of fairly pure silt, then a sequence of wash levels containing lenses of tiny waterborne pebbles and some tephra perhaps deriving either from a first fall or from the later deposit lying immediately above (levels 5 and 6). Over this was found a layer of whitish volcanic ash, compact, horizontally layered, with larger pumice pieces at the bottom, finer ash at the top, and mixed at its upper limits with thin striations of mud again lying flat and evidently due to a sequence of rain-water puddles. Although analysis is still pending at the time of writing, it seems clear that this material should be associated with the Minoan eruption of Thera. Above this lay a thick and compact layer of orange-yellow clay (2), a wash level found close below the surface and containing nothing later than LM I A ceramic material, a factor consistent with our interpretation of this as an open area in the succeeding periods.

⁴¹ Levi, D., ‘La tomba a tholos di Kamilari a Festòs’, ASA 39–40 (1961–1962), 37Google Scholar, fig. 35; Levi, D. and Carinci, F., Festòs e la civiltà minoica ii (Rome, 1988), 107–8Google Scholar, pl. 48 c–f.

⁴² Kanta, A., The Late Minoan III Period in Crete (SIMA 58; Göteborg, 1980), 101–2Google Scholar.

⁴³ Hallager, B. P and McGeorge, P. J. P., Late Minoan III Burials at Khania (SIMA 93; Göteborg, 1992), 19Google Scholar, pl. 17 a, b, 18 a P 6339.

⁴⁴ Alexiou, S., ῾Yστερομινωϊκοὶ τάφοι λιμένος Κνωσοῦ (Κατσμπὰ) (Athens, 1967), 42, 61Google Scholar, pls. 7 a–b, 9 a–b.

⁴⁵ PKU 62, 110 and 160.

⁴⁶ Popham, M. R., The Minoan Unexplored Mansion at Knossos (Oxford, 1984), pl. 165, nos. 37–9Google Scholar.

⁴⁷ PK III, 217–22, fig. 6.

⁴⁸ The phases are clearest at the test made through the street surface in the area of Room B, and are discussed in greater detail below by Eleni Hatzaki.

⁴⁹ These are not found further east, at Room P, where the early wall line must have been set slightly further south, for the LM I A pebble and plaster floor runs directly beneath the later rebuild.

⁵⁰ These Cape Sidero blocks are not preserved in a continuous line along the south facade. Alongside Room A they were robbed out, as is made clear by the presence of stone chips and the erratic position of the stones which later filled the robbing trench. The line continues past Rooms B–D but in less regular shape and partly in different materials, though the robbing is not so easily determined. The sideropetra line does continue further west (fronting the southwest sector of Building 6 at EJ/EN 82/84), but there is set c. 0.30 m higher.

⁵¹ The slab is broken and fragmentary, but measures 0.56 (E–W) × 0.67 (N–S) × 0.27 m, and resembles the LM III A threshold slab of Building 1.

⁵² The two lines merge further north and are not yet distinguished, nor is the relationship with the exterior wall of Room K clear. The curving wall shown derives from the 1904 excavation plan, and could be a later modification of an original squared corner. Its line lies outside the permitted excavation area.

⁵³ In Room Q the edge of the plaster floor runs along the wall line and suggests a superstructure of mud-brick here.

⁵⁴ This is consistent with the general downhill slope of the terrain from SW to NE. There is a drop of nearly 2 m from Main Street, where bedrock is found at 11.16 m, to Building VII, where it is at c. 9.50 m.

⁵⁵ See n.9.

⁵⁶ The damage was on one occasion so severe that the plane of the outer face of the wall came to rest inside (east of ) the original plane of its inner face—apparently shifted bodily (as an unbroken unit) by earthquake. The erratic, sinuous, line of what remains at the surface is shown on the air photo in PLATES 38–39.

⁵⁷ Continued deterioration after exposure in 1994, as a result of two bouts of immersion in seasonal winter flooding, made it clear that drastic conservation was necessary. This was carried out by the Knossian workmen under Pamela Jerome's supervision, during the summer of 1996 in a fully reversible manner by covering the slabs with sand and placing concrete replicas of slabs and plastered beams above. An attempt was made to reproduce the original colour variation (red and yellow). The restoration is seen in the views at PLATE 40.

⁵⁸ PK IV, 278, fig. 9.

⁵⁹ Ideally excavation should proceed further east to reveal the building's façade on the street fronting on Block B here, but this is disputed land, with a complex legal history since the days of the earlier excavations.

⁶⁰ PM ii. 672.

⁶¹ The construction is of small stones (H. 5–10 cm; W. c.30 cm), suitable for a mud-brick superstructure. One line runs north–south, parallel to the west wall of the court, and shows a semi-circular gap which could have held a column base of diameter c.25 cm.

⁶² For an account of the steps, and their latest use (MM III A), see discussion of Room L below.

⁶³ The line appears as a one-course reconstruction of the footings on the air photo at PLATE 39.

⁶⁴ Excavation at the associated space adjacent to the West Façade has begun to show a sequence of MM III to LM I A deposits of some quality here, including fresco fragments.

⁶⁵ For a possible blocked door, once entering from the west, see under Room D₂ below.

⁶⁶ Shown on the 1904 plan of the site, as an entrance to Block M (PKU, pl. 1).

⁶⁷ e.g. at EQ 93 (on plan at PK 1987, 260 and fig. 1).

⁶⁸ It is finely cut on all sides but left rough beneath. It measures 45 × 27 × 16 cm and has a Γ-shape with projection 4 cm deep and 10 cm wide. For parallel shape cf. PLATE 45 a.

⁶⁹ PK I, pl. xx; PK VII, 207.

⁷⁰ Although no access steps have yet been revealed for this period, it remains possible that there had been an earlier stepped entrance, perhaps at the gap visible in the lower courses of the south wall, as there was in Room L, described below. Later structural alterations at the south side of the room are shown by a short stretch of wall immediately to the west of the steps, directly above a possible first period entrance. For further discussion of the painted plaster, see under Room L below.

⁷¹ See discussion above, relating to possible murex dye working in MM III A. cf. also n. 26.

⁷² Cf. Durkin, M. K. and Lister, C. J.. ‘The Rods of Digenis’, BSA 78 (1983), 69Google Scholar.

⁷³ Cf. n. 28.

⁷⁴ This would place it inconveniently close to the floor level at this period, but it was evidently part of an earlier framework.

⁷⁵ These are slightly smaller than those used for the floor of the room (e.g. 23 × 50 cm as against 50 × 50 or 90 × 55 cm).

⁷⁶ The slab which originally formed the threshold and top step had slipped down inside the room and was found on top of three lower slabs. A test at this point in the entrance revealed a packing fill similar to that noted in the test at the SW corner of the colonnade in court H. Below this the original line of the east wall of Room L continued.

⁷⁷ An area of light grey ashy material, c.1.0 × 1.0 m and 0.10 deep covered the centre of the room. Roofing domatochoma or volcanic ash? Analysis pending.

⁷⁸ The column base is of green serpentine (original D c.32 cm and H.6.9 cm; actual fragment 16.1 × 14.6 × 6.9 cm). The top has a border 2.9 cm wide which is raised 2 mm above the centre (to prevent the wooden column from slipping). The flat base shows the remains of a drill-hole (D.2 cm), probably for a wooden peg more than 2 cm long which held the column base in position on a wooden surface. (It cannot be excluded, however, that this fragment is part of a pyramidal stand, similar to an example from the Psychro Cave.)

⁷⁹ Refs. at n. 55.

⁸⁰ PK 128 and fig. 6.

⁸¹ Analysis by Chris Doherty of the Archaeological Research Laboratory, Oxford.

⁸² These tests were at the north terrace wall, in the southwest corner of the central court (H) and the deep drain at R1, and reveal a building of good quality with fine plastered floors. Little evidence for destruction, perhaps just abandonment.

⁸³ For the protection of livestock and water sources after the Santorini eruption cf. Driessen, J. and MacDonald, C. F., The Troubled Island: The Santorini Eruption and the LM I B Destruction of Minoan Crete (Aegaeum 17, 1997)Google Scholar.

⁸⁴ PK I, 314, fig. 27.

⁸⁵ PK II, 280, fig. 2.

⁸⁶ D'Agata, A. L., ‘Late Minoan Crete and horns of consecration: a symbol in action’, in Laffineur, R. and Crowley, J. L. (eds), ΕΙΚΩΝ: Aegean Bronze Age Iconography (Aegaeum 8; Liège, 1992), 247–55Google Scholar.

⁸⁷ For a detailed discussion of the 1902 excavation of Block Delta cf. PK II, 292–4.

⁸⁸ See PK II, pl. 6.

⁸⁹ This was done by visual comparison with previously identified samples proven to be tephra. The identification was confirmed by Professor F. McCoy, who was present during the excavation.

⁹⁰ Warren has demonstrated that the foundation of a road system at Knossos at MM I B/MM II A provides evidence for a central authority, cf. Warren, P.M., ‘The Minoan Roads of Knossos’. in Evely, D., Hughes-Brock, H., and Momigliano, N. (eds), Knossos: A Labyrinth of History (Oxford, 1994), 189–210Google Scholar.

⁹¹ The tablet and painted inscription were published in GORILA vols. 1 and 4; for Zb 19 and Zb 21, see BSA 82 (1987), 153Google Scholar; 86 (1991), 146–7.

⁹² We follow E. Hallager's new classification of Linear A W-documents (cf. E. Hallager, ‘Nomenclature of administrative Linear A documents’, in S. Dager-Jalkotzy, S. Hiller, and O. Panagl (eds), Proceedings of the Tenth International Conference on Mycenaean Studies, Salzburg 1995 (in press). Since the object here does not show any seal impressions on the edge as roundels do and is pierced, whereas roundels never are, it deserves to be classified separately. Some other awkward documents, previously classified as roundels, have now also been brought into separate categories (e.g. Ph Wc 47 as Zb 47, Ph Wc 42 as Wy 42).

⁹³ Many Linear A documents were intentionally cut, hence there is no reason to speculate about the possibility of the disc ever having been formed in a way similar to that hidden behind the so-called 'simili-joins’, records first entirely inscribed and then intentionally divided up in separate documents for different purposes (see Driessen, J., ‘Observations on simih-joins in the Room of the Chariot Tablets at Knossos’, in Killen, J. T., Melena, J. L., and Olivier, J-P. (eds), Studies in Mycenaean and Classical Greek Presented to John Chadwick (= Minos 20–2; Salamanca, 1987), 151–62)Google Scholar.

⁹⁴ For a detailed discussion of wine logograms in Aegean scripts, see Palmer, R., Wine in the Mycenaean Palace Economy (Aegaeum, 10; Liège, 1994), 2Google Scholar.