Acknowledgements. The work reported here took place in May 1983 and was given at all times the fullest co-operation and assistance of the local Ephor, Dr. K. Davaris, the Epimelete at Sitea, Dr. N. Papadakis, and the Phylax at Palaikastro, Mr. M. Neranzakis. We are most grateful to them and to the Director of the British School at Athens, Dr. H. W. Catling, and the Managing Committee for their financial and other support. We are also most grateful to Professor and Mrs. J. W. Myers for their generous and prompt help and for permission to publish their excellent balloon photograph at Plate 8. L. H. Sackett also wishes to acknowledge his support from the Dillon Grant of Groton School, Mass.

J. A. MacGillivray and L. H. Sackett are responsible for the archaeological survey and interpretation, D. Smyth for the detailed surface survey of walls at fig. 2, as well as for the other maps and plans at Figs, I, 3–5, assisted in the field by J. Driessen, who also contributed part II. D. G. Lyness undertook the magnetometer survey, reported with B. A. Hobbs in part III, and was assisted by B. Howenstein, M. Burnham, and J. Cocking, who also took part in both the architectural and sherd survey work. A. Peatfield examined the Kouremenos Area. M. MacGillivray was responsible for the care and feeding of the field team, for which we are extremely grateful.

Abbreviations other than those in standard use

PK I–VII. ‘Excavations at Palaikastro I’ to ‘VII’. I in BSA 8 (1901–2) 286–316; II in BSA 9 (1902–3) 274–387; III in BSA 10 (1903–4) 192–321; IV in BSA 11 (1904–5) 258–308; Vin BSA 12 (1905–6) 1–8; VI in BSA 60 (1965) 248–315; VII in BSA 65 (1970) 203–242.

PKU. The Unpublished Objects from the Palaikastro Excavations 1902–6 (BSA Supplementary Paper 1), 1923.

PKU II. ‘Unpublished Objects from Palaikastro and Praisos’, BSA 40 (1939–40) 38–59.

¹ Called To Kephalaki or Tou Kona to Kephalaki after a former owner.

² Some 10,000 sq. m of the central part of the town were opened up, but excavation tended to be delimited by the need to complete well-preserved areas to which a commitment had already been made. There were occasional exceptions to this—for instance, on the Promontory, where Bosanquet identified and partially excavated a building of great importance, cf. PK I 306, but was unable to fulfil his stated intention of returning to complete the work. The problem of erosion—and the consequent question of the value of committing resources to substantial excavation in a particular area—was there from the beginning: the areas north and east of Block Χ were abandoned as too eroded, PK IV 285; Blocks Κ, Λ and Σ, Υ were in much eroded areas, PK III 202, 215; and again in 1962–3 several small tests within the town area were unproductive, e.g. AA, CC, EE, and Trenches 1, 5, and 6 at DD, PA VII 210–11. The truth is that on this kind of sloping site, lined with terraces, some areas are likely to be severely eroded, others, possibly even nearby, to preserve a surprising depth.

³ Kanta, A., The Late Minoan IIIB Period in Crete, SIMA 58 (1980) 189–192.Google Scholar

⁴ Intensive sherding produced very few post-Minoan sherds from this or the surrounding area, PA VII 241 fig. 25 nos. 1, 3, and 4.

⁵ Other cemeteries were found at Kephalaki, Angathia, the cliff edge to the south of the Promontory, and Petsophas lower slopes and cave.

⁶ The sequence of EM/MM pottery awaits full publication.

⁷ Smee, M. H., BSA 61 (1966) 157.Google Scholar

⁸ Davaras, K., BSA 75 (1980) 115.Google Scholar

⁹ Id., ADelt 27 (1972) 652.

¹⁰ Damage was also done by deep ploughing to the north-west sector of Block B, not Block M as formerly thought.

¹¹ A classic case occurs at Kouremenos, where a rectangular, roofed animal shelter is built of excellent cut Minoan blocks including schist, sandstone, and limestone. Nearby can be seen the plan of a Minoan house in negative form, i.e. robbing trenches in rectangular form around an area of surviving floor with purple-schist paving.

¹² There are two obvious areas with ‘nil’ readings protruding into the occupied zone. The first, a separate rectangular field south-west of Block Γ, has a thick cover of sedge grass because a stream deposits silt in the area. The second is a large hollow set between two ridges at the south of the town; this area may also be partly covered with recent silt, but our examination of the surface suggested that much of it was never occupied.

¹³ Warren, P. M., AE 1979, 86, 108.Google Scholar

¹⁴ Platon, N., Zakros (1971) 9.Google Scholar

¹⁵ e.g. Hogarth's Houses, The Unexplored Mansion and the town area being cleared behind the Stratigraphical Museum, cf. Hood, S. and Smyth, D., Archaeological Survey of the Knossos Area, BSA Suppl. 14 (1981) 8–10 and fig. 2.Google Scholar

¹⁶ Cadogan, G., AR 24 (1977–1978) 74 figs, 3c, 11.Google Scholar For a complete discussion of Minoan fortifications see Alexiou, S. in Κρητολογια 8 (1979) 41–56.Google Scholar

¹⁷ Part of this north-south route, where it approaches the coast, was reused more recently as a water channel. This complicates the identification of the Minoan construction here, but the channel appears to follow an ancient line.

¹⁸ See n. 2.

¹⁹ The rocks and the beach below are covered with fallen Minoan building material from these structures.

²⁰ It is possible that there is in fact no gap, but that buildings run all along the north slope of the Promontory and around its south end to form a link with the houses at the East Beach, discussed here. See also n. 21 below.

²¹ These probably link with building fragments shown here in plan, fig. 2 (top right.).

²² We owe this suggestion to Alan Peatfield.

²³ My thanks go to Miss Emma Faull for correcting my English and Miss Jane Cocking for her assistance in planning the quarries.

²⁴ See Chalikiopoulos, L., Sitia, die Osthalbinsel Kreta's. Eine geographische Studie (Berlin 1903) 69.Google Scholar

²⁵ Verbal information from A. Zervandonakis, and from M. Neranzakis, the Phylax at Palaikastro, who informed me that the gypsum was quarried in the area twenty years ago. This would be the most convenient source for the gypsum at Roussolakkos and perhaps for all three blocks recorded from the palace at Kato Zakros, cf. Shaw, J. W., Minoan Architecture: Materials and Techniques, Annuario 49 (1973) 23.Google Scholar

²⁶ Warren, , BSA 59(1964) 99, note on gypsum.Google Scholar

²⁷ PK I 311, 315; Shaw, op. cit., 19–20.

²⁸ Broken poros door-jamb bases, apparently never in situ, were noticed, especially in the south-east areas of the site. The excavated houses contain door-jamb bases in Blocks Β, Δ, Γ, Χ, Ν, and at Kouremenos. Their common use contrasts with the scarcity of similar bases in the houses at Zakros and elsewhere in east Crete. The façade of House Ν and parts of Block Β also use roughly dressed poros limestone.

²⁹ At A in fig. I, at the sea there is a rectangular cutting (about 5 × 2 m.), about 1 m deep with a narrow rectangular trench in the centre. The fact that there is a modern well in one of the corners, together with the absence of the usual quarrying channels, suggests that this quarry is modern.

³⁰ Soles, J. S., ‘A Bronze Age Quarry in Eastern Crete’, JFA 10 (1983) 45 figs. 15–16.Google Scholar This report was published after our examination of the quarries at Ta Skaria.

³¹ Minoan quarries are known at Mochlos {about 80 m long, but very narrow), at Pelikata near Zakros (30 × 50 m), at Mallia (2·5 km long, but the amount of extracted blocks cannot have been very large from such a low ridge), at Knossos, near Herakleion, Phaistos, Nirou Khani, and Archanes, cf. Shaw, op. cit. 30–42.

³² In δ 20–36, PK II 293; Block X, PK IV 283; Γ 1–12 and 13–22, PK II 290–1; Block Π, PK IV 286; parts of Block B, PK I 310, 315 and PK II 287).

³³ Shaw, op. cit. 30–4; Soles, op. cit. 33–43.

³⁴ Three Hellenistic sherds and one Byzantine sherd were also found.

³⁵ MM III walls using ashlar were discovered in Block B and elsewhere, PK II 287. A very fine ashlar façade was excavated in area DD, PK VII 211.

³⁶ For this work the mason probably wet the surfaces to be quarried and made channels using a long pick, since long vertical grooves can be discerned in most quarries and on several blocks in the Minoan town.

³⁷ The range of block measurements in each quarry, given in centimetres, is as follows: the blocks in Quarry B measure 90–110 long, 50–70 wide, and 17–45 high; those from Quarry E are 80–100 long, 40–70 wide, and 20–30 high; those from Quarry F are 65–130 long, 40–80 wide, and 35–50 high; those from Quarry G are 100–30 long, 60–80 wide, and 50–80 high; those from Quarry J are 135–60 long, 55–90 wide, and 30–80 high; those from Quarry K are 85–125 long, 55–85 wide, and 15–25 high. The table gives approximate averages.

³⁸ Thirty-five blocks were measured; sixteen from Block Δ, seven from Blocks Β and Γ, four from House Ν, and one from the main street.

³⁹ Some obvious Minoan blocks, reused as sheep-troughs, are still visible at Ta Skaria; they measure 110 × 45 × 30 cm, 95 × 45 × 30 cm, 85 × 50 × 40 cm, and 75 × 35 × 40 cm. They may have been extracted from Quarry F.

⁴⁰ About 10,000 sq. m of the Roussolakkos site have been excavated, but the occupied area was at least three times larger. Shaw, op. cit. 32 n. 3, notices that the width of the blocks of the façades of the central court of the palace at Zakros range from 58 to 70 cm, too large to come from the Pelikata quarry. These blocks may come from the more distant quarries at Ta Skaria.

⁴¹ PM II 664 n. I. The ‘window sign’ occurs at Knossos between Magazines 2 and 3 and on two blocks behind the Throne Room. My thanks to Mr. Steve Townsend for the photograph of the Knossos signs in plate 13f. The incision at Ta Skaria is about 5 mm deep; the lines being about 15 mm wide.

⁴² Quarry E seems to have been one of the latest opened since only a small area is quarried and many blocks are left unfinished on the site. The bedrock face on which the sign occurs would be very suitable for quarrying. Could the sign be taken as an indication that the face on which it was carved is an area proposed for future work?

⁴³ The authors would like to acknowledge the financial support of their field-work by the British School at Athens. Field observations were carried out by D. G. Lyness, with the assistance of J. Cocking, B. Howenstein, and M. Burnham. Many of the programs used in the reduction of the data were written by Mr. G. Dawes of the Department of Geophysics, Edinburgh University.

⁴⁴ The spelling Kouremenos occurs on the most recent large-scale map. It is defended at the local ephorate as the correct form, signifying ‘shorn’, while the former spelling Kouramenos is characterized as incorrect.