^{page 4 note 1} § 20, p. 130 seqq.

^{page 4 note 2} 1·39 m. wide, 0·38 deep with a tread of 0·13.

^{page 5 note 1} See Ręport, 1902, p. 110.

^{page 6 note 1} It is shown out of its place on the top of the wall of D 1, in the view given in Fig. 2.

^{page 6 note 2} Report, 1902, p. 28, seqq.

^{page 7 note 1} It is 23 cm. high and 30 square at the base.

^{page 7 note 2} B.S.A. vol. viii, p. 299 and Pl. XVIII.

^{page 7 note 3} A summary account of this is given by DrParibeni, R., Lavori Eseguiti, &c., 1903, p. 30.Google Scholar

^{page 7 note 4} A small steatite base of the kind, obtained from a peasant at Palaikastro, has been connected with this usage by MrBosanquet, (B.S.A. viii, p. 300).Google Scholar

^{page 7 note 5} Compare the design on a painted vase from Phaestos, Pernier, L., ‘Scavi della Missione Italiana a Phaestos, 1900–1901.’ Rapporto preliminare (Monumenti Antichi, vol. xii.) Tav. VIII. 3.Google Scholar

^{page 7 note 6} Schliemann, , Mycenae, p. 242, Fig. 352.Google Scholar

^{page 8 note 1} Mycenaean Tree and Pillar Cult, p. 61, Fig. 39 and p. 62.

^{page 8 note 2} Schliemann, , Mycenae, p. 362Google Scholar, Fig. 541; Furtwängler, , Antike Gemmen, Pl. II. 42.Google Scholar

^{page 9 note 1} Knossos Report, 1902, p. 102, Fig. 59.

^{page 9 note 2} B.S.A. vi, p. 76.

^{page 9 note 3} Mycenaean Tree and Pillar Cult, p. 12 seqq.

^{page 10 note 1} There was, as usual, an inner backing of rubble masonry.

^{page 10 note 2} About 38 cm. in diameter.

^{page 10 note 3} In each case 90 cm.

^{page 11 note 1} It is only 9 cm. high; about 35–40 deep, and is edged with gypsum strips 14 cm. thick.

^{page 12 note 1} There was not in this case any socket between the horns as in case of those of the Shrine of the Double Axes. The horned object is 20 cm. wide and 19 high.

^{page 12 note 2} Its measurements are 4·37 m. from N. to S. by 2·20 from E. to W.

^{page 13 note 1} The existence of a recess in the wall at the North-West corner of Room E 1 gave rise to the supposition that there was actually an aperture into this elongated space E 1 at that point, and that it was therefore a passage. But (1) there is no evidence that there was any opening at this point, though the wall was thinner. (2) There is no trace of door-jambs, such as in that case would almost certainly have existed. (3) Room E 1 having already a doorway leading into the portico, such a passage way would have been superfluous.

^{page 13 note 2} This is Dr Mackenzie's opinion.

^{page 13 note 3} Knossos, &c., Report, 1902.

^{page 13 note 4} On this feature of the Cretan House and its architectural consequences, see Noack, F., Homerische Paläste, p. 17seqq.Google Scholar

^{page 18 note 1} See Report, 1902, pp. 106, 107, Fig. 64.

^{page 18 note 2} Report, 1902, pp. 118–120, and Figs. 70, 71.

^{page 19 note 1} See below, pp. 96–98, and Fig. 66.

^{page 20 note 1} Report, &c., 1900, pp. 25 and 59 seqq.

^{page 21 note 1} It must be borne in mind that the evidence as to the exact stage of culture existent at the time of the foundation of the Later Palace is not so clear as that which illustrates the close of its First Period. The ceramic and other relics found immediately below the remodelled floors belonging to the Second Period of this Palace sufficiently declare the character of its culture in the middle stage of its history—at the close, that is, of its First Period. We see there a more or less transitional phase which may conveniently be termed Late Minôan I. On the other hand immediately below the original floors of the first Palace we find, as under the Olive Press Room, the finest products of the ‘Middle Minôan’ Ceramic Style. It seems probable that this phase was still existent at the time of the foundation of the Later Palace. It even appears that the pictographic characters of seal impressions found in the deposits below the original floors of the Later Palace are somewhat more archaic than those of the Magazine containing the clay documents of this class. This fact seems to weigh in favour of the second hypothesis mentioned in the text, that this pictographic deposit, namely, was covered in owing to some alteration of structure that took place soon after the foundation of the Later Palace. It may be possible to describe the last phase of the culture of the Earlier Palace as ‘Middle Minôan I.’ and the first of the Later Palace as ‘Middle Minôan II.’

^{page 22 note 1} Report, &c., 1902, pp. 62, 63; Myc. Tree ana Pillar Cult, p. 31, Fig. 17. See too below, p. 88, Fig. 60.

^{page 22 note 2} Report, 1901, p. 35 seqq.

^{page 27 note 1} Knossos, Report, 1902, p. 87 seqq.

^{page 27 note 2} See p. 49 seqq. and Fig. 26.

^{page 28 note 1} Report, &c., 1901, p. 44 seqq.

^{page 29 note 1} See Report, &c, 1901, p. 47.

^{page 29 note 2} See above, p. 13.

^{page 35 note 1} A preliminary notice of this is given by DrParibeni, R., Lavori eseguiti &c. nel Palazzo di Haghia Triada dal 23 Febbraio al 15 Luglio 1903; p. 30seqq.Google Scholar

^{page 36 note 1} See p. 9 and p. 149 seqq.

^{page 36 note 2} See above, pp. 31–33.

^{page 37 note 1} Myc. Tree and Pillar Cult, p. 44 Fig. 25.

^{page 37 note 2} Report, &c., 1902, p. 32.

^{page 38 note 1} Report, &c., 1901, p. 30.

^{page 38 note 2} See below § 18.

^{page 38 note 3} I observe that this conclusion, which I have already insisted on elsewhere, has been advanced independently by MrCook, in his interesting monograph on ‘Zeus, Jupiter, and the Oak’ (Classical Review, Nov. 1903, pp. 409, 410).Google Scholar Mr Cook rightly points out the religious importance of the Lily Crown as seen in the painted relief found in the South wing of the Palace.

^{page 41 note 1} For the probable meaning and composition of these cake-like objects (c, f, g, h, i, k, or Fig. 20) see below, p. 64.

^{page 41 note 2} See Hogarth, , B.S.A. vi. p. 114, Fig. 30 and Pl. XI. 2.Google Scholar A similar steatite ‘Libation Table’ was obtained by me from what appears to have been an early sanctuary at Arvi on the South coast of Crete (J.H.S. xvii. p. 357).

^{page 43 note 1} The names have been kindly supplied me by Prof. W. F. R. Weldon, F.R.S., from some specimens submitted to him. The worn state of the valves in the case of one or two examples made it difficult to attach the specific name.

^{page 47 note 1} The crystal petals are 3·4 centimetres in length and 2·3 in breadth.

^{page 47 note 2} See Report, &c., 1901, p. 78.

^{page 47 note 3} It is 27·7 centimetres in length.

^{page 51 note 1} For a fuller exposition of this ceramic influence I must refer to Dr Mackenzie's paper in the forthcoming work on the Excavations of the British School at Phylakopi.

^{page 51 note 2} See Report, &c., 1902, p. 58.

^{page 51 note 3} See my note on the Marks on the Melian pottery in the forthcoming publication on Phylakopi referred to above.

^{page 52 note 1} Halbherr, Federigo, ‘Resti etc. scoperti a Haghia Triada. Rapporto sulle ricerche del 1902,’ (Mon. Antichi xiii.), p. 21seqq.Google Scholar Further discoveries of tablets presenting the same characters were made during 1903, which by the great courtesy of Professor Halbherr I have been allowed to study.

^{page 52 note 2} A photograph of this has been kindly supplied me by Mr Bosanquet, and I have also had the opportunity of studying the original at Candia.

^{page 54 note 1} In a paper read to the British Academy on Nov. 25, 1903, on ‘the Pictographic and Linear Scripts of Crete and their Relations’ I have already called attention to the value of the evidence supplied by the inscriptions from the Temple Repository at Knossos, and to their identity in style with those of Hagia Triada (see Summary Report, Times, Nov. 26).

^{page 56 note 1} See above, p. 2.

^{page 57 note 1} Compare, too, the seal impression found in 1902. Report, &c., 1902, p. 78, Fig. 43.

^{page 57 note 2} Report, &c., 1901, p. 95, Fig. 31. A seal impression with a fuller design of the same kind was found at Hagia Triada.

^{page 57 note 3} Paribeni, R., ‘Lavori eseguiti dalla Missione Italiana nel Palazzo e nella necropoli di Haghia Triada, 1903.’ (Rendiconti della R. Accademia dei Lincei, vol. xii. fasc. 70, p. 17).Google Scholar The remains of an upper zone of this rhyton exhibit a hunting scene of wild bulls closely resembling that of the Vapheio Cup.

^{page 58 note 1} See Head, , Coinage of Syracuse, p. 10Google Scholar, and cf. Holm, , Geschichte Siciliens, I. p. 572Google Scholar and my ‘Contributions to Sicilian Numismatics’ Num. Chron. 1894, p. 212. The artistic tradition of similar sea-monsters goes back to the very beginnings of a later classical art: witness an amygdaloid gem of the ‘Melian’ class in my own collection found at Epidauros Limera and exhibiting a ‘pistrix’ beneath the forepart of a war galley. This gem dates from about 700 B.C.

^{page 58 note 2} For a fresh illustration, see p. 114 below, Fig. 70.

^{page 58 note 3} See above, pp. 36, 37.

^{page 59 note 1} See below, p. 86.

^{page 59 note 2} See p. 88.

^{page 59 note 3} The cross type appears five times on the side of clay disks; one with inscribed characters on its face. On the same disk appears impressions of a couched bovine animal on a base (No. 27) and a façade with masonry (No. 8).

^{page 59 note 4} See Report, &c., 1901, p. 29, Fig. 9.

^{page 59 note 5} There were eleven examples of this type in the Repository.

^{page 60 note 1} In Halbherr, Resti etc. Rapporto, 1902, p. 39, Fig. 33. This type is there reproduced from an imperfect example, the double axes, clear on some impressions of the seal since discovered, not appearing.

^{page 60 note 2} This type was represented by eight examples.

^{page 60 note 3} Myc. Tree and Pillar Cult, p. 27, Fig. 15.

^{page 60 note 4} This pard-like creature may however be intended for some kind of mastifi.

^{page 60 note 5} Length 6·78 centimetres. There had originally been a second handle plate of the same form the blade being held between them. The upper and lower faces of the plate and the side view are shown in Fig. 39.

^{page 62 note 1} Berlin Cat. No. 2; Furtwängler, Antike Geminen, Taf. II. 24.

^{page 62 note 2} On coins of Crete in genere struck under Trajan, B. M. Cat. Pl. I. 9. In a specimen in my own collection the arrow is very clear.

^{page 62 note 3} As shown below, it can only be called ‘porcelain’ in a loose popular sense.

^{page 62 note 4} So too in the excavations of the Italian Mission at Phaestos and Hagia Triada.

^{page 62 note 5} See § 16.

^{page 63 note 1} In this particular deposit the signs on the under-side 01 some of the inlays were little more than groups of notches. On the roundels from the Throne Room and faience plaques for inlaying found elsewhere regular signs occurred analogous to those of the bone and ivory inlavs from the Palace. See Report, &c., 1901, pp. 119, 120, where the marks on the faience inlays from Tell-el-yehûdiyeh, &c., are compared. Similar signs occur on the faïence plaques found by the Italians at Phaestos and Hagia Triada.

^{page 64 note 1} The notches, strokes, or dots of varying numbers which either accompany these signs or appear in place of them evidently refer to the arrangement of the inlays. Here again Egyptian analogy is conclusive—witness the slanting lines, from one to nine in number, incised on the lower surface of the gold and turquoise hawks from bracelets found in the tomb of King Zer at Abydos—to mark their order in the series (Petrie, , Royal Tombs of the Earliest Dynasties, II. 15Google Scholar).

^{page 64 note 2} Professor A. H. Church has kindly examined some specimens of this Palace faïence. He writes ‘besides silica the glaze contained lime, a little magnesia, some soda, and a larger amount of potash. The friable and rather porous ‘body’ or paste of this glazed material contains (in the state in which it was analysed),—(a) moisture and other matters—1·22 p.c.; (b) matters soluble in strong hydrochloric acid—2·22 p.c.; (c) quartzite sand with traces of mica, felspar, and clay—96—56 p.c. (= 100). (b) consists chiefly of lime and the oxides of iron, alumina, and copper. (c) consists of 97·01 p.c. of silica, 1·33 p.c. of alumina, and 0·17 p.c. of lime, with traces of lime, magnesia, copper, and alkalies. The paste when dry has received a coating of glaze, and has been fixed at a moderate heat, just sufficient to fuse the latter without softening the body.’ Professor Church considers that the glaze owes its colouring mainly to copper and that it is probably nearly related to ‘Egyptian Blue.’ The dark browns and black however are referred by him to a ferruginous origin.

^{page 64 note 3} In the case of other plaques such as those of the latest Palace Period found in the Throne Room it is of a brown colour.

^{page 64 note 4} Some of these were shown in Fig. 20 above, below the Libation Tables.

^{page 65 note 1} The lowermost matrix on the mould figured by Schliemann, (Mycenae, p. 107, No. 162)Google Scholar seems to have been made for a glass paste object of a kind representing a degeneration of the console shown below, Fig. 43.

^{page 65 note 2} Similar shells in glass paste have been found in tombs of the Lower Town at Mycenae.

^{page 65 note 3} E.g. Plychoceras gaultinus, a cretaceous species.

^{page 66 note 1} A plain example of such a console in native faïence was found in the Palace (near the Southern Terrace). Degenerations of similar consoles in glass paste are not infrequent in ‘Late’ Mycenaean deposits. For the matrix of one such found at Mycenae, see above, p. 61, note 1.

^{page 66 note 2} Schliemann, , Mycenae, I. 241, Nos. 350, 351.Google Scholar

^{page 67 note 1} In the ‘Late’ Mycenaean Period, answering to the ‘Period of Partial Occupation’ of the Palace at Knossos, Cyprus, as the excavations at Enkomi show, had become a great centre of faïence fabric.

^{page 68 note 1} The calix of the flower as seen in profile is bell-shaped, patulous at the upper margin: height 4 centimetres.

^{page 70 note 1} Over a dozen of these were found. Their lower surface is flat.

^{page 70 note 2} The nautiluses were of four sizes, 9 centimetres × 4·3, 6·6 × 3·5, and 4 × 2·5, and one intermediate between the two last. They were all modelled in the round.

^{page 71 note 1} It is also common on Minôan gems, and it is possible (as seems certainly the case on some of the bovine designs of the Eretrian dies) that these Karystian coin-types actually represent a revival of designs taken from ‘Mycenaean’ gem engravings.

^{page 72 note 1} The length of this plaque is 20·5 centimetres, the height 12·4. Besides this type, of which there were fragments belonging to several examples, there occurred parts of reliefs belonging to another similar series, the plaques of which were about ⅔rds of the dimensions of the other. A calf belonging to this smaller series is fairly preserved.

^{page 73 note 1} See Myc. Tree and Pillar Cult, pp. 78, 79, 81, and especially p. 82 (where the Ancilia are compared).

^{page 76 note 1} The l. fore-arm with the tail is restored in the figure.

^{page 78 note 1} A small fragment of the top of the tiara at the back was also wanting.

^{page 78 note 2} Sections are restored in Fig. 57.

^{page 78 note 3} The head part is restored in Fig. 57.

^{page 81 note 1} The Crocus-gatherer fresco seems to belong to the First Period of the Later Palace. Crocuses are found on the vases of the same transitional period and were also a favourite subject in Melos.

^{page 83 note 1} The upper part of the jacket is restored in Fig. 58.

^{page 83 note 2} Compare the Dove Goddess found in the shrine of the Double Axes at Knossos (Report, &c., 1902, p. 98 seqq.).

^{page 84 note 1} This shrine is perhaps contemporary (as most of the remains at Gournià) with the First Period of the Later Palace at Knossos. (In my Report, &c., 1902, p. 105 it is referred to as later.)

^{page 84 note 2} For an excellent account of these see Wide, S., Ath. Mitth, xxvi. pp. 247–257 and Pl. XII.Google Scholar

^{page 84 note 3} The head of the snake, as seen in Fig. 54, is restored, but there is no doubt whatever as to its position.

^{page 84 note 4} See Myc. Tree and Pillar Cult, pp. 52, 53.

^{page 84 note 5} See below, p. 87.

^{page 84 note 6} Compare the shell relief found by the Italian Mission at Phaestos, (Mon. Ant. vol. xii. Tav. VIII. 1Google Scholar: J.H.S. xxii. p. 92, Fig. 33, on which Mr Hogarth justly remarks, ‘A glance is enough to assure any one familiar with Egyptian art that these figures are first cousins of those Nilotic divinities whose one arm is raised in exactly the same pose while the other holds the ankh.’

^{page 85 note 1} Od. ix, 188, 189.

^{page 86 note 1} In Cer. 123, 124.

^{page 86 note 2} Τινὲς δὲ Βριτόμαρτιν τὴν αὐτὴν εἶναι τῇ ῾Εκάτῃ Schol. ad Hymn. Orph. XXXVI, v. 12. According to another account Britomartis was daughter of Hekatê, Etymol. M. s.v. Βριτόμαρτισ. Cf. Hoeck, , Creta ii. p. 175.Google Scholar Britomartis, according to a mystic tradition, was the granddaughter of Eubulos (Diod. v. 76, 3).

^{page 86 note 3} See above, p. 62 and note 2.

^{page 86 note 4} See above, p. 61, Fig. 40.

^{page 86 note 5} See above, p. 72, Fig. 49.

^{page 87 note 1} Diod. v. 77. 5.

^{page 87 note 2} Paeôn of Amathus, in Plut. Thes. c. 20. Ariadnê had there a sacred grove and grave.

^{page 88 note 1} See Myc. Tree and Pillar Cult, pp. 31, 32.

^{page 89 note 1} Stark, Gaza, Plate, Fig. 1.

^{page 89 note 2} Interpreted as an allusion to the Semitic epithet of the God: Mania, or the Lord.

^{page 89 note 3} Diod. iv. 79, 3.

^{page 89 note 4} On the obols of Eryx dating from about 450 B.C. above the dog. On the didrachms of Motya and Panormos of about the same date, beside the Goddess's head, and, again, beside the head, on the coins inscribed Ziz, belonging to one or other of the Elymian cities. The hound, which here is the sacred animal, appears from certain Minôan seal-types to have been early connected with the cult. Another frequently recurring religious emblem associated with the Aphroditê of Eryx is the Triton shell (wrongly described as a ‘Murex’) so significant in the Minôan ritual.

^{page 89 note 5} See, for instance, Mallery, Garrick, ‘Pictographs of the North American Indians’ (Annual Report of Bureau of Ethnology, 1882–1883, pp. 238, 239).Google Scholar

^{page 89 note 6} Thus the body of Hathor, as the Night Sky, is at times seen covered with crosses in place of stars (Budge, , The Gods of the Egyptians, i. 430Google Scholar).

^{page 89 note 7} On coins of the Syrian Tripolis. On the cruciform types of the star symbol, see especially Müller, L., Religiøse Symboler of Stierne-, Kors-, og Cirkel-form hos Oldtidens Kulturfolk, p. 7seqq.Google Scholar

^{page 90 note 1} A similar cruciform figure occurs as a character of the Linear Script. Compare, too, the Egyptian †, the early Dynastic form of which is an equal-limbed cross.

^{page 90 note 2} See above, p. 36.

^{page 90 note 3} Pernier, L., Scavi, etc., a Phaestos (Roma, 1902), p. 90.Google Scholar

^{page 90 note 4} At a distance, respectively, of 3·8, 3, and 2 centimetres from the edges of the three complete limbs.

^{page 91 note 1} The exposition of the objects is necessarily very incomplete. It was, for instance, impossible to set up the faïence reliefs and inlays that probably decorated the walls.

^{page 92 note 1} Described and illustrated by DrXanthudides, Stephanos A. in Ἐφ. Ἀρχ. 1900, p. 26Google Scholar, seqq., Plates 3 and 4.

^{page 93 note 1} Tsountas, , Ἐφ. Ἀρχ. 1897, Pl. 13Google Scholar, Fig. 26, p. 170.

^{page 93 note 2} The cross as a symbol or amulet was also known among the Babylonians and Assyrians. It appears on cylinders (according to Professor Sayce, of the Kassite Period), in front of seated gods (cf. Cat. De Clercq, No. 254, 255, Pl. XXV.), apparently as a sign of divinity. As an amulet on Assyrian necklaces it is seen associated as on the Palaikastro mould with a rayed (solar) and a semi-lunar emblem—in other words it once more represents a star.

^{page 97 note 1} Cf. Mackenzie, , ‘The Pottery of Knossos,’ J.H.S., xxiii., p. 166.Google Scholar

^{page 97 note 2} Petrie, , ‘Egyptian Bases of Greek History,’ J.H.S., xi., Pl. XIV., Fig. 6.Google Scholar

^{page 97 note 3} Tsountas, , Κνκλαδικά II., Pl. 9, 4.Google Scholar

^{page 98 note 1} Diameter at top 1·8 centimetre.

^{page 100 note 1} The system consists of blocks of limestone alternating high and low. The best preserved of the higher blocks is that against the bastion, the other being much weathered and worn away. The thickness of the construction is only 36–40 centimetres. The first ‘pier’ is 60 centimetres in length: then follows a lower block 67 centimetres long; then two higher blocks 70 and 72 centimetres in length respectively with an interval of 45 between them which seems to have been originally filled. Beyond this, apparently there was another lower interval followed by a similar longer ‘pier.’ The higher blocks were at most 37 centimetres high and the lower 12 centimetres.

^{page 101 note 1} Excepting a small strip cut out of the first step by the keying in of the lower steps of the Eastern flight.

^{page 101 note 2} They are somewhat finely cut, but of the usual calibre. The axe is 20 centimetres in width. In the neighbouring Palace area were also found broken blocks belonging to earlier constructions with a Double Axe of archaic form, a deep-cut ‘star’ sign 26 centimetres in diameter and a ‘zigzag’ sign also deeply cut, 26 centimetres by 3.

^{page 101 note 3} See Report, &c., 1902, p. 111 seqq. and Figs. 67, 68. Another similar runnel was found this year by the remains of a staircase in the South-East quarter, below the ‘Court of the Sanctuary.’

^{page 101 note 4} Dr Mackenzie observes that ‘the highest point of the parabolic curve comes at the transition from one step to another.’

^{page 102 note 1} This is not adequately brought out in the plan, Mon. Ant. xii. Tav. II.

^{page 104 note 1} The Eastern section of the part preserved of this supporting wall is 6 metres in length; there is then a gap of about 10 metres, after which from a point under the fourth step it is continued for another 13 metres.

^{page 104 note 2} The restored parts are indicated by dotted lines in Mr Fyfe's plan (Fig. 68).

^{page 106 note 1} Traces of an earlier wall line were found (as shown in the plan) a little West of this, which was cut through when the paved path was made.

^{page 107 note 1} The lower of these is ·74 centimetres deep, the other ·85; the tread is ·12 centimetres.

^{page 109 note 1} Remains of the original ‘Middle Minôan’ wall have been lately found behind a wall belonging to the later Period of the Phaestian Palace, now restored by the Italian Mission (Pernier, L., Lavori eseguiti nel Palazzo di Phaeslos: Marzo-Luglio, 1903Google Scholar.)

^{page 109 note 2} Pernier, Luigi, ‘Scavi della Missione Italiana a Phaestos’ (Mon. Ant. xii., 1902, pp. 33, 34).Google Scholar

^{page 109 note 3} As noticed above, p. 21, note, the later Palace at Knossos itself probably goes back to the close of the Middle Minôan Period. But many fragments of polychrome pottery found above the level of the West court at Phaestos and in the neighbouring house (the so-called ‘Altar’)—such as some of those imitating metal work and others with plain geometrical designs—correspond with wares actually found below the early floor levels of the later Palace at Knossos.

^{page 109 note 4} One feature of the Phaestian arrangement not observable in the Theatral Area at Knossos is the continuation of the causeway in a line of steps ascending the long stone seats in the manner of a diazoma.

^{page 110 note 1} Recent investigations point to the fact that the wall above these may have enclosed a rounded bit of staircase.

^{page 110 note 2} See above, p. 57.

^{page 110 note 3} See Report, &c., 1902.

^{page 110 note 4} Ἐφ. Ἀρχ., 1889, Pl. X. 12.

^{page 110 note 5} Halbherr, , Resti, &c., scoperti ad Haghia Triada (Mon. Ant., vol. xiii., Roma, 1903), p. 39, Fig. 33.Google Scholar The heads of the double axes are visible on a more recently discovered impression from the same seal.

^{page 110 note 6} See above, p. 87.

^{page 111 note 1} The close connexion of the great Knossian Goddess with Ariadnê, as to which I had been independently impressed, has been rightly insisted on by Noack, F., Homerische Paläste, p. 86seqq.Google Scholar

^{page 111 note 2} Il. xviii. 591 seqq.

^{page 111 note 3} West of Ida the &#x03B1πЄντοζα'ληѕ; prevails and in Sphakia the σου'στα;

^{page 112 note 1} Plut. Theseus, xxi. (on the authority of Dicaearchus) Θησεὺς ἀναθεὶς τὸ ᾿Αφροδίσιον, ὅ παρὰ τῆς ᾿Αριἀδνης ἔλαβεν, ὲχόρευσε μετὰ τῶν ἠϊθέων χοπείαν, ἤν ἔτι νῦν ἐπιτελεῖν Δηλίους λέγιυσι, μίμημα τῶν ἐν τῷ Λαβυρίνθμͺπεριόδων καὶ διεξόδων ἔν τινι ῥυθμῷ παραλλάξεις καὶ ἀνελίξεις ἐχοντι γιγνομένην. The Κερατών altar about which the dance took place has been aptly brought into relation with the ‘sacral horns’ of the Minôan altars by Noack, F. (Homerische Paläste, p. 87).Google Scholar

^{page 113 note 1} It has been traced uninterruptedly in this direction over 40 metres.

^{page 114 note 1} Compare the examples given, Report, &c., 1901, p. 53, Fig. 15, and by Mackenzie, D., ‘The Pottery of Knossos’ (J.H.S., xxiii. 1903, p. 204).Google Scholar Dr Mackenzie rightly insists on the fact that the Double Axe is foreign to the ordinary decorative repertory of the Minôan vase painters, and that its introduction must be due to a special religious motive.

^{page 114 note 2} Compare the gold figures from Mycenae, , Schliemann, , Mycenae, p. 218Google Scholar, Nos. 329, 330, and the lentoid gem from the Heraeum at Argos, Schliemann, , Mycenae, p. 362, No. 541Google Scholar; Furtwängler, , Ant. Gemmen, Pl. II. 42.Google Scholar The design also occurs on a vase from Old Salamis (see Myc. Tree and Pillar Cult, p. 9 seqq.).

^{page 116 note 1} The drawing is by Mr Halvor Bagge. The use of white in the decoration of the original may be noted as a survival of the earlier style.

^{page 120 note 1} Now in the Ashmolean Museum at Oxford. The great apparent chronological gap between this and the Middle Minôan types in question has yet to be explained.

^{page 121 note 1} It is 22 centimetres in height and its body the same in diameter.

^{page 123 note 1} An example of this form of decoration will be seen in the vase from a Knossian house excavated by MrHogarth, , B.S.A. vi. (1899–1900), p. 76, Fig. 21.Google Scholar

^{page 123 note 2} The height of the basin, apart from the handle, is 8 centimetres.

^{page 124 note 1} An example of this, apparently of ‘Middle Minôan’ date, was also found at Phaestos (Pernier, L., Scavi, &c., a Phaestos, 1900–1901, p. 83, Fig. 22Google Scholar).

^{page 125 note 3} A smaller bronze bowl with a border showing a similar design was found in one of the tombs near Phaestos.

^{page 125 note 1} The width of the basin (without the handle) 33 centimetres.

^{page 126 note 1} The width of the basin, without the handle, is 32 centimetres.

^{page 128 note 1} Good examples of these are in the Ashmolean Museum. For a bone pendant of a similar kind from a room near the Men's Megaron at Mycenae see Tsuntas, , Ἐφ. Ἀρχ. Pl. XIII. 15.Google Scholar

^{page 128 note 2} This vase is illustrated by MrHall, H. R. in his article ‘Keftiu and the Peoples of the Sea,’ B.S.A. 1902–1903, p. 173, Fig. 7.Google Scholar

^{page 130 note 1} See above, p. 56, Fig. 35 and p. 57.

^{page 130 note 2} According to the analogy of the remains found in the Palace the upper part here would be a gypsum slab and the lower a wooden beam.

^{page 132 note 1} See especially p. 4.

^{page 133 note 1} The wall to the left was of rougher construction, originally concealed by plaster.

^{page 133 note 2} There are two narrow courses at bottom and top (the topmost 20 centimetres high) and seven between, ranging from 57 centimetres in height (the course next to bottom) to 40.

^{page 133 note 3} The edges of these are bevelled off and this interval between the stone piers and the wood-work of the window frame must have been filled with plaster.

^{page 134 note 1} Dr Mackenzie to whom this observation is due observes: ‘the sides of the H. Triada window are so well preserved that it is quite clear that the window was as wide as the stair and the stair antae taken together. The window in that case had to light not only the stair hut a corridor on either side.’

^{page 134 note 2} The dimensions of these steps are the same as the others, except that the tread was slightly higher to make nine steps (including the threshold) correspond with ten in the companion flight to the West.

^{page 135 note 1} Of these steps, Nos. 1, 2, 3 are of gypsum, 4, 5, 6, 7, 8 of limestone, 9 of gypsum and 10 of limestone.

^{page 135 note 2} This is Dr Mackenzie's opinion. The actual interval in the wall here is broader (2·53 metres). The sill seems to have been about 1·30 above the landing floor.

^{page 135 note 3} Compare especially the landing-blocks of the fourth flight of the Quadruple Staircase (Reports, 1902, p. 103; 1903, pp. 32, 33).

^{page 136 note 1} Dr Mackenzie makes the just observation that these ledges are only cut in landing-blocks where there is such a hollow interval below. Where the steps rested on a solid foundation there was no such necessity. Thus in the present case no ledge has been cut in the opposite or Western side of the same landing-block, since the steps there rested on the solid ground.

^{page 138 note 1} Report, &c., 1902: Fyfe, , ‘Painted Plaster Decoration of Knossos,’ Journ. R.I.B.A., 1902, p. 128, Fig. 69.Google Scholar (The figure has been placed there in a reversed position.)

^{page 140 note 1} An indication of this style however appears in a pithos of Magazine 9 which shows rosettes painted on slightly convex disks enclosed by raised circles.

^{page 141 note 1} Its dimension are 2 metres East to West, by 1·70 North to South.

^{page 142 note 1} Along the South wall of the room runs a strip of gypsum paving 25 centimetres wide, raised slightly above the level of the rest of the pavement.

^{page 143 note 1} These slabs do not go down to the level of the pavement hut leave an interval of 12 centimetres. Both North and South walls are of rubble construction.

^{page 143 note 2} The North Wall was found to rest on a broader foundation wall which projects 20 centimetres beyond the line of its South face.

^{page 143 note 3} Its height is 71 centimetres and breadth 47 centimetres.

^{page 144 note 1} These are shown restored. They project 20 centimetres from the wall. The length of each balustrade is 1·70 metre; the breadth 0·70, and the height 0·80.

^{page 144 note 2} 34 centimetres square.

^{page 144 note 3} They are 62 centimetres broad, 32 deep, and have a tread of 12 centimetres.

^{page 144 note 4} The lamp is 52 centimetres high, and its receptacle above is 22 centimetres in diameter.

^{page 146 note 1} This interval itself is 1·30 metre wide.

^{page 146 note 2} It was found necessary to replace the upper parts of these slabs which were in a much decomposed condition. On the South side of the niche only the lower slab was preserved.

^{page 147 note 1} The same evidence tends to show that the ‘tank’ of the Throne Room, and the Northern and South-East bath with their gypsum parapets and facings were lit in a similar manner.

^{page 147 note 2} Not including the depth of the niche.

^{page 148 note 1} See especially Lange, K., Haus und Halle, Studien zur Geschichte des antiken Wohnhauses und der Basilika (Leipzig, 1885), p. 60seqq.Google Scholar

^{page 149 note 1} See above, p. 9.

^{page 149 note 2} The plan is given in Fig. 6, p. 15. See p. 14, No. 1. It is there called ‘North door of North-East house.’

^{page 149 note 3} That to the West is hollowed out of a single block 01 gypsum, that to the East has its West side formed of a separate piece. The West ‘Vat’ is 44 centimetres North to South by 37 East to West and 12 centimetres deep. The East ‘Vat’ is 46 centimetres North to South by 35 East to West and 16 deep. The latter receptacle had been plugged by a small block.

^{page 149 note 4} There is a similar dowel-hole in the North border and another in the South-West corner of the room.

^{page 151 note 1} The steps are 30 centimetres deep and have a tread of 18 centimetres.

^{page 151 note 1} The lower of these is of gypsum, the upper of limestone.