^* Acknowledgements. This study was undertaken with the kind permission of the Greek Archaeological Service and was financed by a generous grant from the Knossos Donated Fund; we thank the then Ephor of the Heraklion Museum, Dr I.A. Sakellarakis and the British School at Athens for making the work possible. Driessen's trip to Crete was made possible by an N.F.W.O. Travel Grant. The study began under the auspices of Mr M.S.F. Hood while searching for Mason's Marks; we learnt much while working for Mr Hood and owe him a debt beyond the limits of this article. He also provided the photographs of our squeezes of Mason's Marks. J.-P. Oliver provided the equipment for taking the squeezes. At Knossos, the Curator and Cataloguer were of great assistance in the Stratigraphical Museum whilst others helped in the palace itself; we particularly thank Cyprian Broodbank, Jane Cocking, Vasso Fotou and Sheilagh Wall. Steve Townsend fed and watered us and, with his usual keen eye, pointed out certain features in the palace we otherwise would have missed. The Editors of the BSA have made helpful suggestions, which are incorporated in the text. The plans were inked with advice from David Smyth and Vasso Fotou.

¹ BSA 8 (1901–2) 81.

² BSA 8 (1901–2) 81–87. PM I, 225–230.

³ PM I, Figs. 171a–b.

⁴ PM I, 226, n.1.

⁵ PM I, 226, Fig. 171a.

⁶ M.S.F. Hood and W. Taylor, The Bronze Age Palace at Knossos (1978). In the process, it was discovered that Entrance Hole C is incorrectly positioned on the published plan. It should be situated about a metre to the east.

⁷ cf. PM I, Fig. 152.

⁸ op. cit.

⁹ See n.7.

¹¹ PM I, 133; PTK 166; Hood, in The Function of the Minoan Palaces (Stockholm 1987) 205.Google Scholar

¹² PM I, 226, Fig. 171a–b.

¹³ PM III, 405.

¹⁴ op. cit.

¹⁵ PM I, 227, Fig. 171b.

¹⁶ op. cit.

¹⁷ PM I, 227. Fig. 171c.

¹⁸ BSA 8(1901–2) 83, Fig. 47.

¹⁹ op. cit. 56, Fig. 29.

²⁰ op. cit. 82, Fig. 46 and 85, Fig. 48.; PM I, Fig. 171b, 172.

²¹ op. cit. 82. Fig. 46.

²² PM III, 388–389, Fig. 260.

²³ Overbeck, J.C. – McDonald, C.K. in AJA 80 (1976) 163–65CrossRefGoogle Scholar; Palmer, L.R. and Boardman, J., On the Knossos Tablets (Oxford 1963) 130–150, 52–61.Google Scholar

²⁴ For drains elsewhere on Crete see especially Shaw, J., Minoan Architecture: Materials and Techniques (henceforth MAMAT) (Rome 1971, 126–134)Google Scholar; Halbherr, F. – Stefani, E. – Banti, L., Haghia Triada nel periodo Tardo Palaziale (Rome 1980), 221–226Google Scholar; Pernier, L. – Banti, L., Il Palazzo Minoico di Festo II (Rome 1951), 186–187 and index s.v. ‘canale’Google Scholar; Pelon, O., Le Palais de Mallia V.1 (Paris 1980)Google Scholar, index s.v. ‘caniveau’ and ‘conduit’.

²⁵ Evans, (PM I, 226)Google Scholar and Shaw (MAMAT 132–133) believed that the drains had two main functions, namely drainage and ‘sanitation’; Hutchinson, R.W. (Town Planning Review 21 (1950) 213–214)CrossRefGoogle Scholar considered their primary function to have been the disposal of sewage, whereas Graham, (Palaces of Crete (Princeton 1962) 220–221)Google Scholar stressed the fact that at least five light wells were served by this elaborate system.

²⁶ PM I, Fig. 171a shows e.g. the original loop of the drain whereas Evans, in BSA 8 (1901–1902) Fig. 29Google Scholar and Hood and Taylor (op. cit.) also show the later arrangements.

²⁷ cf. infra p.256.

²⁸ PM I Fig. 152 and p.204, 327–328, 346–347; PM III 373–374, especially concentrating on the south wall of the South Light Well of the Queen's Megaron (also called the Area of the Bird Relief).

²⁹ Cf. Fig. 3, Cover Slab 11 and Section B–B. The conjectured north-south wall which could have served as a terrace wall for the buildings of the original Domestic Quarter to west, followed a line running from the easternmost pillar base of the Portico of the Hall of the Double Axes – the base itself may have been part of the original wall (PM III Fig. 217) – towards the remains of ancient walling south of the staircase leading onto the terrace.

³⁰ Overbeck, J.C. and McDonald, C.K. in AJA 80 (1976) 164CrossRefGoogle Scholar and Pendlebury, J.D.S., The Archaeology of Crete (London 1939) 129–131.Google Scholar

³¹ A proper survey of the First Palace Architecture still awaits to be conducted; Branigan, K.The Foundations of Palatial Crete (London 1970) 52ff.Google Scholar is unfortunately incomplete and deals only with the published evidence.

³² For the lay-out of the palaces, see Preziosi, D., Minoan Architectural Design (New York 1984), passim.Google Scholar

³³ This is the case for the outlets of the Queen's Megaron and the later ones of the Hall of the Colonnades and the Hall of the Double Axes.

³⁴ The ashlar used in the side-drain to Shafts A and B is e.g. of much finer appearance than the ashlar normally used in the Main Drain. This observation also applies to the ashlar used in the shafts themselves. By contrast, the masonry used in the Outlet of the Queen's Megaron is of inferior quality.

³⁵ The Grand Staircase is said to belong to the original construction (PM I, 240).

³⁶ For clay pipes see PM III, 252–253 and MAMAT 198–201; similar systems, e.g. in the Court of the Stone Spout at Knossos (MAMAT, Fig. 160) may have led water from the roofs to lower areas. In Egypt, some temples and palaces were provided with sculptured spouts (Badawy, A., History of Egyptian Architecture (University of California Press, 1966, 358).Google Scholar

³⁷ According to Hood and Taylor, op. cit. the height of the south-west corner of the Central Court (the base of Evans' bust being an arbitrary 100.00 m.) is 101.78 m.; that of the north-west, 100.95 m.; the height in front of the Stepped Porch is 101.30 m. No level is given for the Central Court in front of the Domestic Quarter but their Section 4 shows a slope down from west to east.

³⁸ PM I, 225–226, 393, Fig. 286.

³⁹ We were unable to ascertain the relative architectural date of the North-South Tributary. Its construction forms an organic part of the original Main Drain. If, however, the present reconstructed east terrace wall of the East Light Well of the Hall of the Double Axes is later, the Tributary could also be later since that terrace wall uses the east side wall of the drain as a foundation.

⁴⁰ PM II, 462–463, Fig. 273.

⁴¹ PM III, 326.

⁴² PM III, 492, Fig. 341.

⁴³ Branigan, op. cit., 59, 63 remarks on the introduction of a proper drainage system: ‘The construction of such drains is a useful pointer to the further emergence of an acceptance of urban life and its communal obligations’.

Naumann, R.R., Architektur Kleinasiens (Tubingen 1971) 197–202 gives a good summary of the drainage systems in the Near East and AnatoliaGoogle Scholar; in Alaça Huyuk at least an elaborate system seems to have existed early in the 2nd Millenium (Figs. 255–266). Hult, G., Bronze Age Ashlar Masonry in the Eastern Mediterranean, SIMA 66 (Göteborg 1983) unfortunately omits all mention of the use of ashlar in drainage systems.Google Scholar

⁴⁴ It is accepted here that the Domestic Quarter is mainly a construction of MM IIIA (in Evans' terminology), with modifications in LM I and later; see Overbeck, and McDonald, in AJA 80 (1976) 164.CrossRefGoogle Scholar

⁴⁵ The re-use of architectural elements was noted at certain points and described in the detailed description above. The entire north part of the North-South Tributary was refurbished, but if this was entirely due to Evans, we could not establish. Earthquakes were often used to modify existing constructions, see e.g. Soles, J. in TUAS 4 (1979) 11.Google Scholar

⁴⁶ Shafts A and B were constructed just outside the terrace wall. They abut not only the wall itself but also each other. Evans' idea that they were originally cisterns during MM III may be correct (PM I, 228). Thus, the opening into the Room of the Plaster Couch from Shaft A (PM III, 389, Fig. 260) behind the gypsum dadoes could belong to this phase (2a). The hole was blocked, perhaps when the dadoes were put in place, the Toilette installed and the Shafts made into a single interlocking system leading to the Main Drain. It is also possible that, when the Main Drain was blocked at this point, the Shafts were made into cisterns and a hole created between Shaft A and the Room of the Plaster Couch. We tend to agree with Evans and accept the former proposal since the gypsum dadoes are to a certain extent paralleled by the installation of the gypsum lined compartment named the ‘toilette’ which we have placed in Phase 2b. A likely variation of this is that Shaft A was originally (Phase 2a) a cistern and that Shaft B was connected to the Main Drain from the beginning of Phase 2. This may account to some extent for the observed differences in construction.

⁴⁷ Shaft C itself was found filled-in at the time of the excavation (PM III, 405). It is unlikely that this happened at the beginning of Phase 2 but rather at the end.

⁴⁸ PM I, 228 and PM II, 400, Fig. 226.

⁴⁹ Evans, in BSA 8 (1901–1902) 62–63, Figs. 17, 44–48Google Scholar; PM I, 228–230, Fig. 172; PM III, 387–389, Fig. 260.

⁵⁰ For spaces or installations elsewhere interpreted as toilets see e.g. Platon, N., Zakros. The Discovery of a Lost Palace of Ancient Crete (New York 1971) 103, 127, 195–196Google Scholar; Graham, op. cit., 108–110 and Marinatos, S., Excavations at Thera VI (Athens 1974) 26–27, Figs. 2–5, Pls. 55b and 58aGoogle Scholar; Willson-Cummer, W. and Schofield, E., Keos III. Ayia lrini: House A (Mainz 1984) 18 and Pl. 7 (Room 24).Google Scholar

⁵¹ For the elaborate toilet provisions at Amarna, see Badawy, Egyptian Architecture 94–95.

⁵² BSA 8 (1901–2) 68ff.

⁵³ PM III, Fig. 250 for an illustration of this wall.

⁵⁴ BSA 8 (1901–2) 56, Fig. 29.

⁵⁵ PM III, 405.

⁵⁶ For U-shaped channels see MAMAT 126–127. Evans, (PM I, 335)Google Scholar thought the U-shaped channel carried the ‘effluvia’ of the toilet away. Another misinterpretation (e.g. Cadogan, G., Palaces of Minoan Crete (London 1976) 80)Google Scholar, is that the U-shaped channel carried away dirty water, while the Main Drain disposed of the fresh.

⁵⁷ For Mason's Marks see MAMAT 109–111 (references), J.-P. Olivier in Pelon, Palais, 175–238 and the forthcoming publication of Knossian Mason's Marks by M.S.F. Hood. Sign l in the present catalogue was kindly pointed out to us by Mr S. Townsend. In the list, the following order is used (all dimensions in metres except where stated): dimensions of sign; depth and width of cutting; situation of block; type of stone; dimensions of block; position of sign on block.

⁵⁸ Sign m occurs on a block which belongs to Phase 2 at the earliest. The block, however, is out of place due to Phase 3 modifications. Signs k and l are on blocks belonging to Phase 2a or 2b (cf. supra n.43).

⁵⁹ For the latest discussion of Mason's Marks and their function cf. Hood, in The Function of the Minoan Palaces (Stockholm 1987) 205–212.Google Scholar

⁶⁰ MAMAT, 22 ‘It (gypsum) would not have been used for drainage channels’; ibid. 132 ‘…these (drains) were constructed of poros limestone or sandstone, never of gypsum’.

⁶¹ Pendlebury, J.D.S., Handbook to the Palace of Minos (London 1954) 44.Google Scholar

⁶² Hutchinson, in Town Planning Review 21 (1950) 209.CrossRefGoogle Scholar

⁶³ PM II, 119; PM III, 288 and Graham, Palaces. 104. For a recent discussion on the function of Lustral Basins see Marinatos, N. and Hägg, R. in OpAth 16, (1986) 60.Google Scholar

⁶⁴ For the kalderim paving in the Lair, cf. BSA 8 (1901–2) 68ff. and PM I, 335.

⁶⁵ PM I, Fig. 152.

⁶⁶ Cf. Pendlebury, Guide to the Stratigraphical Museum in the Palace at Knossos.

⁶⁷ The differences between MM I and MM IIA at Knossos are not necessarily clear unless one is dealing with substantial deposits or particularly diagnostic pieces. More information concerning this will shortly be published.

⁶⁸ BSA 8 (1901–2) 68.

⁶⁹ PM III, 405.

⁷⁰ For illustrations of the deposit cf. PM III, Figs. 268–271.

⁷¹ PM III, 405.

⁷² BSA 8 (1901–2) pls. II–III.

⁷³ BSA 8 (1901–2) 56ff.

⁷⁴ BSA 8 (1901–2) 66.

⁷⁵ BSA 8 (1901–2) 56, Fig. 29.

⁷⁶ BSA 8 (1901–2) 64, Fig. 31 left.

⁷⁷ BSA 8 (1901–2) 84.

⁷⁸ The chronology proposed here differs from that to be published by us in L'Habitat Egéen Préhistorique (forthcoming). Having presented that paper in June 1987, the evidence was reconsidered. We are now convinced that Shaft C was not blocked as a result of the LM IIIA destruction but rather during or at the end of LM I. The U-shaped Channel does not, therefore, belong to the LM IIIB period but more likely to LM II–IIIA.