Abbreviations in addition to those normally used in the Annual:

AM = Ashmolean Museum.

Boardman = On the Knossos Tablets. The Date of the Knossos Tablets (1963).

Evans, Prehistoric Tombs = ‘The Prehistoric Tombs of Knossos’, Archaeologia lix (1905), 391–562.

Furumark = The Mycenaean Pottery. Analysis and Classification (1941).

HM = Herakleion Museum.

Marinatos and Hirmer = Crete and Mycenae (1960).

Palmer = On the Knossos Tablets. The Find-Places of the Knossos Tablets (1963).

PM = Evans, , The Palace of Minos at Knossos i–iv (1921–35).Google Scholar A. Evans and J. E. A. Evans, Index Volume (1936).

¹ I am grateful to Mr. Mervyn Popham and to Mr. Sinclair Hood for reading this article and for making a number of very helpful suggestions. The Managing Committee of the School gave me permission to study the material and the Ashmolean Museum generously made available the notebooks of the Knossos excavations. The pottery from Magazine XIII has been studied and is being published by Mr. Popham in his work on the Palace destruction deposits. He has generously allowed me to make use of it here.

² Mr. Hood in 1965 first recognized the material as the remains of a gypsum vase workshop and kindly drew the writer's attention to it.

³ BSA vii (1900–1), 42 refers to a ‘perforated object of gypsum’ of careless fabric. It might be thought that this is the spindle whorl in our material (Plate 40, b, lower right), but the object Evans describes is compared with, though smaller than, the large octopus weight (ibid., fig. 12), 42 cm. high. The perforated object must have been something considerably larger than our little whorl.

PM i. 457 refers to ‘many odd slips of gypsum slabs’ from the intervals between the bottoms of the cists in the magazines. This may be or may include a reference to our material, or Evans may be referring to fragments of the gypsum slabs of the cists themselves. Knossos. Dating of the Pottery in the Stratigraphical Museum ii. 4 has the entry ‘mostly stone fragments’ under D XVII 5 (Magazine XIII, Cist 5). The material is not mentioned in the excavation notebooks.

⁴ BSA vii (1900–1) 40–41.

⁵ Evans's Sanctuary Hall, PM i. 445.Google Scholar For the tablets, Palmer, 98; Boardman, 35–36. A further point in favour of the upper floor is that it is surely impossible to imagine the workshop situated inside the Magazine just around Cists 2–6.

For the juxtaposition of ‘religious’ and ‘working’ areas there is the evidence of the Zakro Palace where a room with raw materials (Room ψ) is adjacent to the Shrine Treasury (Room Φ) (Archaeological Reports for 1963–4 28, fig. 37; Illustrated London News 7 Mar. 1964, 350) and where the Archives Room is next to the Shrine (Arch. Reps. ibid. 29, and fig. 37, Rooms Ο and Π).

Mr. Popham has suggested to me that the cist fill (see Section 2), to which the stone pieces belong, is not necessarily fall from above, but might be filling material brought from outside the Palace. This is possible but in the writer's opinion a simpler and more probable explanation of the location of the material is that it did fall from an upper floor. The fact that the material includes pieces of wall frescoes and the remains of a stone workshop would seem to support this.

A further knotty point. It is appreciated that the tablet archives and the workshop are not exactly contemporary, since the workshop material is almost certainly from the lower cists (see n. 17) which were closed just before the destruction, to which the tablets belong. Hence the workshop is earlier than the tablets. But, as shown below (Section 2), there is very little difference in date. Hence it is reasonable to discuss the archives and workshop together as indicating the nature of the upper floor—at two slightly different phases in the final Palace period (L.M. II–early IIIA 2).

⁶ Drill marks are frequently visible on the interiors of stone vases. Cf. also Section 3.

⁷ Unpublished Objects from the Palaikastro Excavations (1923), pl. XXX, E, 1; F, 1. BSA lx (1965) 305–6 and fig. 21, nos. 51–53 and pl. 79a.

⁸ PM i. 621–2, fig. 458.

⁹ Dr. Chadwick kindly examined Evans's photograph for me and gave it as his opinion that the marks are not script. Evans's transcription (fig. 459) is not very true to the original. Nor is that in Brice, , Inscriptions in the Minoan Linear Script of Class A (1961) 24Google Scholar, no. V, 13 and Pls. XXX, XXX a. Mr. Brice now agrees that the engraving is probably not script.

¹⁰ The main accounts of the construction and modifications of the cists are PM i. 448–62; iv. 630–3.

¹¹ BSA x (1903–4) 40.

¹² PM i. 442–3, 457.

¹³ PM i. 443 ff. and 527–8. For the triglyph and split rosette fragment, PM i. 443; ii. 604 and fig. 377.

¹⁴ BSA vii (1900–1) 40–41. Mackenzie, Day Book 1901, vol. i, 20 Mar.–2 Apr., pp. 38–57.

¹⁵ BSA x (1903–4) 40. Day Book 1904, p. 50 (quoted by Palmer, 98). Cist 2 had its upper slab removed in 1903 in that this is the year on the wooden label in the Strat. Mus. box.

¹⁶ Mackenzie, , Day Book 1923 i. 12–13Google Scholar.

¹⁷ Mr. Hood tells me that he thinks the material may have come from the upper, not the lower cists—i.e. that it was part of the Palace destruction debris. His main reason is that there is no mention of the stone pieces in the accounts of the fill of the lower cists in the Magazine (see n. 15). A consequence of this view is that the upper cists of the Magazine were not cleared out till 1904 (as indicated by the 1904 box labels of our material). It would then follow that Evans and Mackenzie did not keep separate the material from the upper and lower cists—an unlikely event in view of the careful nature of the investigations here in 1904 (cf. n. 15). Mr. Hood's view also involves rejecting the stated context of the gypsum trial piece (lower cists, see Section 1 and n. 8).

Hence it seems to the writer virtually certain that the stone pieces, labelled 1904, came from the 1904 excavation of the lower cists. The reason why they were not mentioned was that their interest and significance were not then recognized.

¹⁸ PM i. 442; ii. 600; iv. 632–3.

¹⁹ PM iv. fig. 297 a (Temple Tomb flask, L.M. IIIA 1 burial); Evans, Prehistoric Tombs, fig. 117, 76 f and 66 p (Zafer Papoura); Hall, , Decorative Art of Crete in the Bronze Age (1907) fig. 61, 2Google Scholar; Popham, , ‘Some Late Minoan III Pottery from Crete’, BSA lx (1965) pl. 86Google Scholar d (top row 5th and 6th from left). Cf. Furumark Motive 10 A g.

²⁰ Furumark Type 219 (L.H. IIB–IIIA 1). Cf. French, in ‘Late Helladic IIIA 1 Pottery from Mycenae’, BSA lix (1964) 249.Google Scholar Two more such cups come from a recently excavated L.H. IIIA 1 tomb in the Agora, Athenian, Vermeule, and Travlos, , Hesperia xxxv (1966) 76, nos. 7 and 9Google Scholar, and pls. 22 a, 23 b, 23 a. From Knossos, , Popham, , ‘The Destruction of the Palace of Knossos and its Pottery’, Antiquity xl (1966) 25, 27Google Scholar and pl. III. Mr. Popham kindly informs me that this cup shape is common in destruction deposits from the Palace.

²¹ This date for the laying of the upper slabs tells us precisely nothing about how long activities went on afterwards above them, i.e. we are not here concerned with the destruction date of the Palace or with reoccupation.

²² The fragment is closely similar in style to other rosette fragments in fresco and stone from the Palace. Of these the stone pieces from the North West Angle were part of the destruction debris (PM ii. 590–1 and fig. 368; iv. 226). The dating evidence, if there was any, for the pieces from east of the Porch, West (PM ii. 671)Google Scholar and from the West Court Well (PM iv. 222 ff. and fig. 172)Google Scholar was not stated. The South West Porch fragments were said to be in an ‘unquestionably M.M. III association’ (PM ii. 162) but no dating evidence was given and the fact that they occurred ‘near the upper borders of the adjoining remains of paving slabs’; (ibid.) suggests a Palace destruction context (pace PM iv. 225). The piece from the South Propylaeum said by Evans, (PM ii. 696)Google Scholar to have been found 70 cm. ‘beneath the M.M. III b tarazza pavement’ was shown by Boardman to have been 70 cm. below the original ground surface (Boardman, 15, n. 2) and was thus almost certainly part of the Palace destruction debris, as was the rosette fresco also from the Propylaeum, (PM ii. 704).Google Scholar It thus looks very much as if all the stone pieces and the South Propylaeum fresco were part of the destruction debris. Spiral and rosette fresco friezes decorated other parts of the Palace at this time, notably in the Domestic Quarter in the Hall of the Double Axes and the Queen's Megaron, and there was a split-rosette frieze on the Frescoes, Miniature (PM iii, pl. XVI).Google Scholar There was also the spiral and rosette relief stucco ceiling in the Room of the Cornice, Spiral (PM iii. 30 ff. and pl. XV).Google Scholar

As the fresco fragment from below the cists of Magazine XIII depicts this type of frieze it may be suggested on stylistic grounds that it is approximately contemporary with the other friezes. That the pottery with it is L.M. II–IIIA 1 confirms the suggestion.

²³ About 3,100 are known.

²⁴ BSA vii (1900–1), 90–92. PM iii. 269; iv. 896–900.

²⁵ Popham, , Antiquity 1966Google Scholar, loc. cit.

²⁶ First pointed out by Evans, , BSA vii (1900–1901) 92.Google Scholar Spirals on the amphoras, PM iv, figs. 875–6; on the alabastrons, PM iv, fig. 910. This particular form of spiral decoration is purely Mycenaean, occurring on the Mainland from L.H. IIA onwards. Cf. the ivory pyxis from the Routsi/Pylos tholos, Marinatos and Hirmer, pl. 223 left, and the carvings on the Treasuries of Atreus and of Minyas, op. cit. pl. 149 top right and below, and pl. 161. It is also found at Knossos and Mycenae on some of the triglyph and rosette friezes (PM ii, fig. 368; BSA xxv (1921–3) 236, fig. 47b) and is important evidence for the Mycenaean character of the final years of the Palace.

²⁷ The use of the compass here was pointed out by Casson, , The Technique of Early Greek Sculpture (1933) 33.Google Scholar Casson also observed its use on a Throne Room alabastron in the Ashmolean Museum (op. cit. 215, n. 1). See Marinatos and Hirmer, pl. 161, where some compass holes in the centres of rosettes and spirals on the ceiling of the Treasury of Minyas can be seen.

²⁸ PM iii. 269.

²⁹ Kenna, , Cretan Seals (1960) 159.Google Scholar

³⁰ A complete rhyton from the Zakro Palace (HM 2712, Πρακτ. 1963, pl. 147 β); two bowl fragments from Knossos and an unpublished chalice base from Karnari near Mt. Juktas, and the lower part of a rhyton from the Rhyton Well at Mycenae (Nauplion Mus. 8352, BSA xxiv (1919–21) 201 and pl. xi). The famous bowl from the Royal, Isopata Tomb, PM i, fig. 56Google Scholar and PM iii, fig. 182, is not lapis Lacedaemonius but is an Egyptian early Dynastic vase made of a typically Egyptian porphyritic rock (cf. Egyptian Museum, Cairo, Nos. 4844, 88288, 6056, of just this rock).

³¹ BSA vii (1900–1) 20. PM iv. 594 ff. For the date, Boardman, 12–13, 13, n. 1, 19–20 (L.M. IIIA 1); Popham, , The Last Days of the Palace at Knossos, Studies in Mediterranean Archaeology v (1964) 6Google Scholar and 14–15, nos. 12–15 (L.M. IIIB vases).

³² PM ii. 296–9.

³³ Chapouthier, and Demargne, , Mallia, Troisième Rapport (Études Crétoises vi) (1942) 24, 37 (date), 54Google Scholar and pl. LII. Chapouthier, , Demargne, and Dessenne, , Mallia, Quatrième Rapport (Ét. Crét. xii) (1962) 8Google Scholar and pl. XLIII (HM 2250). The interior of this vase had begun to be cut out in a large circular core, D. 15·7 cm., presumably with some kind of drill and lathe. The cutting had penetrated only a short way when work only the vase stopped.

³⁴ They make up 1·17 per cent. of the corpus of Minoan stone vases. Gypsum is hydrous calcium sulphate. Alabaster is a name applied to gypsum and to calcite (calcium carbonate).

³⁵ The Vat Room Deposit (BSA ix (1902–1903) 94–98; PM i. 165–71) is mainly M.M. IA with earlier material. But there are difficulties in Evans's M.M. IA terminal date. Some of the pottery is described as fully developed polychrome eggshell ware (PM i. 168), which should not be earlier than M.M. IB. The rock crystal bore-core (BSA ix. 98; PM i, fig. 119 b) would be much more at home in M.M. III–L.M. I and later, as would the piece of white-spotted obsidian (PM i, fig. 120). The two gypsum lids are the only examples of the use of this stone for vases before L.M. II–IIIA 1 (though gypsum is common at Knossos and two such lids could have been made in M.M. I). One wonders whether the deposit was contaminated.

³⁶ The vase fragments, called alabaster, are mentioned by Platon, Kretika Khronika 1959, 385. I am indebted to Dr. Alexiou for permission to study them. The tomb had no pottery to give a date but the period of the Katsamba cemetery is L.M. II–IIIA 1. This date is confirmed by the L.M. IIIA patterns on two of the gypsum vases from the tomb.

³⁷ A vase described as of white stone, perhaps gypsum (not seen by the writer), comes from an L.M. IIIB tomb at Giophyrakia, westof Herakleion (Platon, op. cit. 367–8; cf. BCH lxxxiv (1960) 840). If the vase is of gypsum the total of gypsum vases is thirty-seven.

³⁸ Professor Platon kindly allowed me to study these vases.