Digitized by the Internet Archive in 2013

http://archive.org/details/palaceofminoscom03evan

THE PALACE OF MINOS

AT KNOSSOS

THE

PALACE OF MINOS

A COMPARATIVE ACCOUNT OF THE SUCCESSIVE

STAGES OF THE EARLY CRETAN CIVILIZATION

AS ILLUSTRATED BY THE DISCOVERIES

AT KNOSSOS

By SIR ARTHUR EVANS

D.LITT., ETC., F.R.S., F.B.A., ROYAL GOLD MEDALLIST, R.I.B.A.

FOREIGN MEMBER OF THE R.ACAD. OFTHE LINCEI, OF THE BAVARIAN, R.DANISH, SWEDISH, AND SERBIAN ACADS.

OF THE GOTTINGEN SOC. OF SCIENCES, OF THE R.ACAD. OF SCIENCES, AMSTERDAM, OF THE GERMAN, AUSTRIAN

AND AMERICAN ARCH. INSTS. AND THE ARCH. SOC. OF ATHENS : CORRESPONDANT DE L'INSTITUT DE FRANCE

HONORARY KEEPER AND VISITOR OF THE ASHMOLEAN MUSEUM

IN THE UNIVERSITY OF OXFORD : HON. FELLOW

OF BRASENOSE COLLEGE

Volume III

THE GREAT TRANSITIONAL AGE IN THE NORTHERN AND EASTERN SECTIONS OF THE PALACE : THE MOST BRILLIANT RECORDS OF MINOAN ART AND THE

EVIDENCES OF AN ADVANCED RELIGION

WITH 367 FIGURES IN THE TEXT, PLANS, 13 COLOURED AND n SUPPLEMENTARY PLATES. (SECTION OF THE 'GRAND STAIRCASE' AND PLANS AND PERSPECTIVE DRAWING OF THE 1 DOMESTIC Q UARTER ' IN POCKET AT THE END OF THE VOLUME)

MACMILLAN AND CO., LIMITED ST. MARTIN'S STREET, LONDON

1930

COPYRIGHT

PRINTED IN GREAT BRITAIN AT THE OXFORD UNIVERSITY PRESS

PREFACE

In the Second Volume of this work, issued in two parts, a good deal of space was reserved for a survey of the neighbouring town houses and dependencies of the Palace. It was also thought desirable to take a more general survey of Knossos in its geographical relations, its position with regard to Aegean traffic, and the record of some remarkable evidences of a line of ancient road connexion across the Island to a port on the Libyan Sea. The Third Volume, now completed, is, like the First, more exclusively connected with the Palace itself. The account of the building follows on that included in the latter half of Vol. II, Part II, relating to the great Restoration of the West Wing of the building after the destructive earth- quake that took place towards the close of M. M. III.

The earlier and later history of the great Transitional Age, which embraces the whole of the Third Middle Minoan Period, is here methodically explored throughout the remaining Palace regions, beginning with the North- Western and Northern Entrance systems and working round by the Eastern slope, past the great Bastion — remarkable details of which have been now for the first time recovered — to the ' Domestic Quarter' beyond.

The artistic remains with which we are confronted in following out this circuit are in many respects pre-eminent amongst all such relics brought to light on the Palace site. The ' Miniature Frescoes ' — here illustrated by a series of Coloured Plates — that seem to stand in connexion with a small shrine of the North-West Palace section — with their vivid conversational sketches of Court ladies, are unique in Ancient Art. Fragmentary remains in the same ' Miniature ' style and some closely related small reliefs lead us on the other hand to the siege scene — here regarded as a historic record — on the silver ' rhyton ' from Mycenae, of which the original form is given for the first time.

So, too, a microscopic design of a coursing bull marvellously painted on the lower surface of a crystal plaque — illustrated by an enlarged coloured drawing — gives us a glimpse of the high perfection already attained in a Minoan Art more characteristic of the Set Cento. In view of the parallels that they present to the miniature and other small Minoan wall-paintings, it has also been thought desirable here to supply some examples of the inlaid

vi THE PALACE OF MINOS, ETC.

designs on weapons from Mycenae and elsewhere that have been well described as 'painting in metals'. Thanks to the valuable assistance of Monsieur E. Gillieron, fils, it has been possible to accompany these by a graphic record of the successive processes by which they were produced, and to add the first authentic illustration of the remarkable engrailed design of the ' Swimmers ', recently revealed by cleaning on a blade from the Vapheio Tomb.

This series of designs in the ' Miniature' style finds its counterpart in a class of intaglios on signets, the funereal association of which can often be authenticated. The subjects of these — interesting examples of which occurred on clay sealings from the Palace itself — can in most cases be shown to have, directly or indirectly, a religious character, and from this point of view, as well as from the sympathy that it shows with the particular branch of the limners' Art referred to above, it has been thought desirable to give a somewhat detailed account of the most remarkable relic of this class. This is the massive gold signet-ring, containing representations of not less than fourteen figures, found in a beehive tomb of Nestor's Pylos, and which it was my good fortune — as the result of a special journey — to rescue from the hands of its native owner.1 It is not too much to say that the successive scenes depicted on this ring supply our first authentic knowledge of the Minoan After-world — an Elysion rather than a Hades. The idea of resurgence is itself graphically conveyed by two chrysalises and corre- sponding butterflies above the Goddess's shoulders. At the same time the striking parallelism displayed by the style of the figures and of the whole composition with the miniature class of wall-paintings has suggested an actual translation of the design into colours on the model of one of these.2

The reconstitution of the structures above the Northern ' Lustral Area', by which the remains have been saved from progressive dis- integration, has made it possible to furnish an adequate record of what is one of the most elegant monuments of the site. Its sunken basin, indeed, and the sombre tone of the wall-painting, now in part restored, made its interior vault a fitting scene for rites connected with the Minoan Goddess as

1 See p. 145 seqq., and compare my fuller account The Ring of Nestor, 6-v., published by Macmillans ; and cf. J. H. S. 1925. Fresh corroboration of the association of the Minoan God- dess with the chrysalis as the emblem of resurgence to a new life after death is given here from a Vapheio ring (p. 148, Fig. 97).

2 Coloured Plate XX a, facing p. 157.

PREFACE vii

Lady of the Underworld. In the same way, the parallel work on the Northern Entrance Passage has brought with it a better knowledge of this once splendid avenue of approach, overlooked on either side by the Porticoes containing the great bull-grappling reliefs in coloured plaster. These plastic compositions, to which the noble head of the charging bull belongs, have been here brought into relation with the Elgin slabs of Knossian gypsum from Mycenae, showing a similar head and the forefeet of a stationary bull. The full illustration of dual compositions, referring respectively to the hunting of half-wild bulls and the contrasted scene in which a decoy cow is used, has been traced in detail in the reliefs on the Vapheio gold cups.

At the same time, the great work of reconstitution of the upper stories carried out in the ' Domestic Quarter ' — thanks to the new facilities for support supplied by reinforced concrete — has made possible a fuller under- standing of the structural and decorative details of the great Halls on that side. The shield frieze, a copy of which is now set up in position in the lobby opening from the second landing of the Grand Staircase, is seen to give a satisfactory key to the actual arrangement of the ' Hall of the Double Axes ' below, where restored replicas of the shields themselves have now been hung on the similar spiral bands above the dado. The evidence, now more fully interpreted, of the traces left in fallen stucco has also afforded adequate information of a wooden throne and canopy that had been fixed in the adjoining section of the Hall. The story of the contiguous ' Oueen's Megaron ' — the centre of the women's domain in the Palace — has been greatly augmented by the revelation of the earlier system, of which the fine 'mosaiko' pavement was in part uncovered beneath the floor of the later hall. The ' Mecraron ' itself, as now resuscitated, with its bath-room and other conveniences, calls up a vision of social life and amenities such as it would be hard to parallel in the Ancient World.1

The relics found in the area behind, precipitated from a windowless upper ' Treasury ', have been placed together in a collected group. Apart from the ivories, the most remarkable of these are steatite objects which as fitted together prove to be the locks of a Sphinx of Hittite type, of which a complete example from Tylissos has been here added to one already known from Hagia Triada. These objects, with a socket in the back, are shown to be descendants of similar Chaldaean stone vessels, some of which

1 See the Coloured Frontispiece to this Volume.

viii THE PALACE OF MINOS, ETC.

seem to have served as inkstands. The remote antiquity of this type has been demonstrated by the recent discovery of one in the form of a boar in a very early stratum at Ur.

Among the ivories, the figurines of leaping youths, the remains of which are here fully illustrated, may be said to excel all known works of the kind in the clan and free action that they display. The remains of a small relief of an ivory Sphinx in the Minoan style, part of a miniature painting of a pillar shrine adorned with double axes, and two bronze axes themselves of the diminutive cult type, made it strange that no figure of the divinity itself should have occurred in the deposit. On the other hand, such an ivory figure, seen by a competent archaeologist in private hands at Candia, shortly afterwards emerged on the other side of the Atlantic as the ' Boston Goddess' — divine sister of the Lady of Knossos, — holding out in this case golden snakes. The opinion, shared by our foreman and others, that this had been abstracted from the Ivory Deposit has certainly not lost in credibility from a remarkable sequel. Also emanating from private possession at Candia, but released after a further discreet interval of time, an ivory figure of a boy-God made its appearance. Having been successful in rescuing this from the midst of doubtful elements in a Parisian dealer's hands, it has been possible to ascertain the fact that it not only answers to the other in its exquisite naturalistic style and individuality of expression, but, as shown standing on tiptoe and coifed with a high tiara, corresponds within a millimetre or so in measurement. The two figures in fact form a single group of the divine Child God saluting the Mother Goddess.1

An illustration of the Minoan worship of the Mother and Child had been already supplied by the painted clay figurine of a later date found in a tomb of the Mavro Spelio Cemetery at Knossos. It is supplemented by a design on a gold signet-ring of the religious class recently found at Thisbe and published here for the first time.2 On this we see the Holy Mother seated with the Babe on her knee and approached with gifts by adorant chieftains, remote predecessors of the Magi.

It will be seen that from the point of view of Comparative Religion this evidence — like that supplied by the subjects on the ' Ring of Nestor ' — is of the highest interest.

Important as are the remains of the 'Domestic Quarter' and the

1 See the restored arrangement, p. 456, Fig. 318, below.

2 P. 471, Fig. 328

PREFACE ix

associated relics, it looks as if, in the later phase of the M. M. Ill Palace, as well as throughout the succeeding epoch, the monumental and artistic aspects of the building found their chief centre in a great East Hall, occupying the area immediately North of that quarter and approached from the Central Court by a stepped portico, facing, though on a larger scale, the ' Stepped Porch ' on the other side of the Court. An earlier East Hall, to which the beautiful ' Ladies in Blue ' fresco belonged, had been built at a lower level, somewhat below that of the pavement of the Court. A reconstruction of the plan of the later Hall * is here given, based on the well-preserved lines of the basement structures, while the existence of the central light-court is further assured by the existence of the head of the shaft by which it was drained. From the porticoes surrounding this had unques- tionably been derived the remains of a series of painted high reliefs that had been precipitated into the space to the South-East at the time of the final destruction of the building-.

These noble fragments, which must be regarded as representing the final development of the Minoan Art of plaster relief, are for the first time fully illustrated in the present Volume, and to the description of the most important specimens I am happily able to add notes kindly supplied to me by the late Sir William Richmond, R.A., whose artistic sense and special technical experience give a lasting value to his appreciations. For anatomic observations on these works I am also greatly indebted to Professor Arthur Thomson, F.R.S. The reliefs themselves are almost exclusively of an agonistic character and belong to boxing and wrestling bouts or the episodes of bull-grappling scenes. With them were also remains of a frieze of a more architectonic character, consisting of opposed pairs of Griffins, tied to columns. Fragmentary as they are, these various high reliefs represent the culminating achievement of Minoan plastic Art. The anatomical know- ledge here displayed, the natural rendering of human forms, the delicate treatment of the flesh surfaces, combined with the most powerful muscular action, raise these works — executed not later than the first half of the Sixteenth Century b. c. — to a level of artistic execution that, on the same lines, has hardly been surpassed by any later Age.

It is tantalizing, indeed, that, for the completion of the figures and composition to which these fragments belong, we should only be able, and

1 P. 49i-

x THE PALACE OF MINOS, ETC.

that occasionally, to turn to small reliefs on vases or to intaglios on gems and signets.

Another very interesting discovery made on the borders of the area occupied by the East Hall was of the same tantalizing nature. This consisted of large bronze curls belonging to the upper part of the fore- head of a gigantesque female figure l that had been wrought in a perish- able material, a clue to which may be found in the carbonized mass in which it lay. We have here in fact the evidence of a Xoanon or wooden image, such as the ' Daedalid ' works still preserved at a comparatively late date in Greece, that may have stood on the back section of this great Hall. It doubtless represented the great Minoan Goddess in one of her aspects, and its presence would mark the Hall as a temple as well as a meeting place.

The circuit of the building included in this Volume may be said to com- plete the general survey of the structures and of the artistic and other remains belonging to the great Transitional phase of Minoan culture that covers the whole of the M. M. Ill Period. It has also inevitably included something of the succeeding Late Minoan style in the days of its early maturity, since, indeed, much of the later decorative system of the ' Domestic Quarter' — illus- trated by the 'Shield Fresco', by the spiraliform dado bands in general, and by the traces of an extension of the Processional friezes to this region — belongs, as is fully demonstrated in these pages, to an intermediate era of partial restoration that dates from about the close of L. M. I a. To this, too, doubtless, the ' Taureador ' panels must themselves, as a whole, be ascribed.

It is clear that a good deal of the decoration of this epoch remained on the walls to the time of the final catastrophe. But we already observe its supersession in places, and notably in the Hall above that of the Double Axes and adjoining East- West Corridor, by wall-paintings in the later style that marks the very latest phase of the Palace history and corresponds with the mature L. M. II Period.

A consideration of the remains of this later class — which represent the artistic fashions of the Age that saw the final destruction of the Palace as a Palace — is reserved for the concluding Volume (IV) of this work. The purest centre of these is to be found in the ' Room of the Throne ' in the West section of the building with its antechamber and surrounding

1 See pp. 522-4, Fig 365, 3<j(>.

PREFACE xi

structures, which, as already pointed out, belong exclusively to this Period. Additional evidence is also afforded in that Palace section by the West Porch, with the remains of a bull-grappling fresco, and by the Magazines on that side in their existing state, comprising the bulk of the great store jars. There is also a mass of fine ceramic evidence from the Western borders of the building in the shape of large painted vases in the highly decorative ' Palace Style '.

In the concluding Volume of this work must also necessarily be included some account of the inscribed tablets of the Linear Class B, which equally mark the latest stage in the Palace history — though in some cases they possibly go back to the closing phase (b) of L. M. I. The most important deposits of these are also associated with the Western Palace region, though they also occur throughout the site. Of these tablets I hope to give a more complete description in the concluding part of my Scripta Minoa, but no account of the Palace in its latest stage could be adequate without considerable reference to these clay archives. Although the script itself still eludes decipherment, the general purport of many of the documents is clear owing to the illustrations that they give of the objects to which they refer, while the accompanying numeration is also intelligible. They thus supply a very extensive source of information as to the contents of the Palace Magazines and Treasury as well as of the royal Arsenal and Mews at this epoch.

It is impossible for me to give adequate acknowledgement of the varied assistance afforded me in the present Volume by many kind friends and fellow workers. I am particularly indebted to my French colleagues, Messieurs J. Charbonneaux, Fernand Chapouthier, and R. Joly, for enabling me to profit by the results of their epoch-making discoveries in the Palace of Mallia.1 For sphragistic records Dr. Doro Levi's supplementary account of the clay sealings of Hagia Triada and of Zakro has also been of special service.2 Valuable assistance in the field of Egyptian and Oriental research has, as usual, been freely given me by Dr. H. R. Hall of the British Museum, and Mr. E. J. Forsdyke has kindly looked over for me the

1 Three recent publications by these explorers in the Bulletin of the French School at Athens (1928) may here be mentioned. F. Chapouthier, Une Table a Offrandes au Palais de Mallia (cf. p. 392 seqq., below) ; R. Joly, La Salle hypostyledic Palais de Mallia ; J. Charbon- neaux, V Architecture et la Ce'ramique du Palais de Mallia.

2 Doro Levi, Le cretule di Hagia Triada e di Zakro, in the Annuario of the Italian School at Athens, 1929.

xii THE PALACE OF MINOS, ETC.

proofs of this Volume. Useful suggestions regarding the Minoan ' rodeo ' have been due to my friend Professor Baldwin Brown. On the other hand, I have suffered an irreparable loss by the untimely death of Dr. Stephanos Xanthudides, who did so much to illustrate the pre-history of his native Island, whose generous help was always forthcoming, and to whom I had been constantly indebted in the earlier Volumes of this work.

To Monsieur E. Gillieron, fils, I am again greatly beholden for many fine drawings, including those illustrative of the technical processes made use of by the Minoan artists in inlaid metal-work. To his recent labours has been also due the restoration in the Palace itself of the ' Shield Fresco ' on the Staircase Lobby and of the actual shields themselves in the great Hall below. Mr. Piet de Jong, the Architect of the British School at Athens, who has carried out, under my direction, the recent work of reconstitution in the Northern and Eastern Sections of the buildino- has executed a series of

o *

restored plans and elevations, notably of the elegant structures above the Northern Lustral Basin, of the Northern Entrance, with the porticoes above on either side, and of the Eastern Bastion, with its remarkable water- system. With the new facilities supplied by the use of ferro-concrete he has completed the work of roofing over the lower Halls and subsidiary structures of the ' Domestic Quarter' by the restoration of the upper floor, and in the case of the Grand Staircase this work of reconstitution has reached the fourth landing and adjacent lobby. For all these later undertakings the immense task already accomplished in this area by Mr. Christian Doll happily afforded a secure basis, and his meticulously accurate measurements have stood all tests.

As a result of these extensive works of conservation and resuscitation not only has a great part of the history of this part of the building been set forth in a permanent manner, but the progressive disintegration of gypsum surfaces from the effects of exposure to the heavy Cretan rainfall has been radically checked throughout a considerable area. It is with special satisfaction that I am able to record that the strength and stability ensured to the reconstituted structures by the use of this new method has enabled them to resist with complete success the fresh severe shock of earthquake that took place in February of this year.

ARTHUR EVANS.

Youlbury, Berks., near Oxford,

March 20, 1930.

CONTENTS

PAGE

§ 68. ' North-West Insula ' North of Central Court, with Ramp Passage from N.W. Portico and Lustral ARea: Basement Chambers above Early Keep i

' North- West Insula ' — its artistic revelations ; Functional importance — supplements that of Western borders of Central Court ; Area West of Central Court chief Cult Centre — retrospective view ; ' Stepped Porch ', built over earlier Cists ; Block North of Porch, of L. M. II date: Includes 'Room of the Throne'; N.E. Corner of old facade and of Early Western ' Insula'; North-Western ' Insula' formed by ' Early Keep ' ; Elements of fortification in Early Palace ; Ramp passage round Keep from N.W. Entrance ; ' Initiatory Area ' and Votaries' entrance ; ' North-West Lustral Basin' of M. M. Ilia date; Restored plan and elevation of 'Lustral Basin' and Superstructures ; Isolation of Sunken Area from other buildings ; Minoan bronze ' key ' found in doorway of later passage ; ' North-West Entrance ' partially blocked ; Basement Chambers above ' Early Keep ' ; Plinth bordering Court, and steps down ; M. M. II a paved floor over Walled Pits, belonging to large basement Chamber; Later dividing wall — M. M. II b ; Gypsum steps re-used for pavement; Central Pillar of Western Basement in M. M..\\b; Exceptional evidences of stratification in ' Room of Knobbed Pithos ' ; Large black steatite vessel with spiral reliefs ; ' Saffron- Gatherer ' Fresco ; Later floors with stone lamps and tablets of Class B ; L. M. Ill a floors above these ; Basements bordering Central Court ; Rearrangement of sup- porting pier ; Further evidence supplied by ' Room of Knobbed Pithos ' ; M. M. Ill a pottery below later pavement (M. M. Ill b) ; Extension of M. M. Ill b floor, its con- tinuous use in restored Palace ; ' Miniature Frescoes ' and Spiral Ceiling fallen on this level ; Stone lamps, and basins for ritual sprinkling.

§69. Discovery of 'Spiral Ceiling' and 'Miniature Frescoes' derived from Corner Sanctuary — Date and Comparative Materials; Embroidered Designs on Holy Robe ... 29

Small corner Sanctuary containing remains of ' Spiral Ceiling ' and Miniature Frescoes ; These fallen from Upper Chamber ; The ' Spiral Ceiling ' — Egyptian analogies ; Parallel from Tomb of Senmut ; Mature L. M. I decorative style ; The ' Miniature Frescoes ' — triple group ; Chronological materials ; Fragments found on M. M. Ill floor — law governing discoveries of fresco remains ; M. M. Ill date of frescoes ascertained ; Fragments from Thirteenth Magazine ; Characteristic specimen beneath base-blocks of later facade ; Fragment from ' Ivory Deposit ' ; True ' Minia- ture' style obsolete by L. M. I; M. M. Ill date of 'Miniature' fragments from Tylissos ; Boxers as on ' rhyton ' — unique bronze vessel ; Frescoes from ' Ramp House' at Mycenae; Fragments from 'Threshing-floor heap'; Miniature designs from embroideries on female robe ; Embossed bands ; Comparison with painted reliefs from Pseira ; Pairs of flutes ; Flutes in sacrificial scene on H. Triada sarco- phagus ; Bull's head trophy between pair of Sphinxes ; Embroidered swallows on robe of Melian fresco ; ' Miniature ' fragments of Threshing-floor heap, perhaps from robe of Goddess ; Commanding position of Shrine, at angle of Central Court and N. Entrance ; Its small dimensions ; Miniature Frescoes set over gypsum dadoes, on line of vision.

xiv CONTENTS

PAGE

§ 70. The Miniature Frescoes: i, The Temple and Grand Stands. 46

The 'Temple Fresco': its technique; The 'Grand Stand' and Spectators; Artistic shorthand in delineation of figures ; Great numbers ; The men ; The women, more carefully represented — 'Court Ladies'; Freedom in Art, without decadence; Rela- tion of figures to earlier 'Ladies in Blue' — close dependence; The Seated Ladies — Groups A-E ; Groups compared with those of the rococo Age, as depicted in the boxes of theatres, &c. ; Separate groups of women in front seats mark of Matri- archal stage of Society ; Dramatic action ; Prominence of gesture language — -that of Naples compared ; Gestures in scenes on signet-rings — primitive elements; Women segregated in front seats but mixing freely with men below ; Isolated fragments with parallel groups ; Window scenes — Cypriote, Assyrian, and Biblical parallels ; Con- trast between Knossian Ladies and Oriental ' Hierodules ' ; Presumption that scenes of the Bull-ring were depicted in lower part of panel ; The Central Columnar Shrine ; ' Superposed Pillars ' of Grand Stand — parallel examples ; Upward taper- ing posts and their analogies; Theatral significance of single pillars in agonistic scenes.

§71. The Miniature Frescoes: 2, Sacred Grove and Dance . . 66

The ' Sacred Grove and Dance ' — Centre of interest to left ; Self-absorption of the female groups ; The Dance — separate performers ; Ritual Dance on Isopata signet ; Ecstatic figures — Sacred Eye in background ; Ecstatic possession — Philistine Prince at Dor ; ' Saul among the Prophets ' ; Dancer on Vapheio gem ; Fresco of Dancing Lady in ' Queen's Megaron ' — mature L. M. I work ; Terra-cotta group from Palai- kastro — a ' ring dance ' ; Central object of the ' Grove and Dance ' religious ; Aphrodite Ariadne ; Theseus and the Delian Crane Dance ; Both sexes included in later ritual dance ; Traditional Dances of Cretan peasants — the ' Kastrinos ', &c. ; Secret Dance of the women ; Mazy course of Dances ; All ' Chain Dances ' ; 'Leaping Dance' (7^81*1-0? x°P0S) — tumbling performance, as Homeric; 'Leaping Dance ' of Cretan Apollo Delphinios ; ' Siganos ' and choral accompaniment — Matinadas ; The ' Dancing Ground of Ariadne ' at Knossos ; Its probable position and character as illustrated by Fresco ; Level site with old olive-trees beneath E. slope of Palace ; The Magic of the Spot.

§ 73. The Miniature Frescoes: 3, Fragments of Siege Scenes and Analogies supplied by the Megaron Frieze and Silver '•Rhyton' of Mycenae.- Egyptian Parallels . . . .81

Miniature Frescoes depicting Siege Scenes ; Antiquity of theme in Egypt and Crete ; Warriors hurling javelins on Knossian fragment ; Youthful Spearman ; Fragment of building, probably belonging to same subject; 'Sacral horns' — not confined to Shrines ; Female figures visible in opening beneath entablature ; Correspondence of structure with faience House fronts ; The Megaron Frieze at Mycenae — scenes of assault ; Warrior precipitated in front of wall — prototype of Kapaneus ; Analogies from Egyptian siege scenes ; House facade on crystal tablet — obliterated by work- man ; Section of outer wall on Tylissos fresco ; The Silver ' rhyton ' with Siege Scene from Mycenae Shaft Grave ; graphic character of design ; Historic import ; Restoration of ' rhyton ' in conical form ; Continuous design — land, sea, and con- ventional reticulation for shallows ; The besieged City — its gate and towers — Minoan character ; Separate fragment with ' superposed pillars ' and ' Sacral horns ' ; Non-Minoan element, however, among inhabitants ; Shields of Minoan type ; Shock- haired barbarians within The walls ; Friendly native allies, some arriving by sea ; Shipwreck and sea monster; Hostile barbarians with primitive weapons — stones, throwing sticks, and clubs ; Native ' friendlies ' better armed ; Minoan element among defenders — warriors in boat ; The relief a historic record ; Topographical data ;

CONTENTS xv

PAGE

Presumptions in favour of Anatolian Coast ; Slings typical Asianic weapons ; Clubs also traditional ; Archers paralleled by Knossian small relief; Pictorial style, akin to that of Miniature Frescoes ; Anticipations of Scenes on Shields of Achilles and Herakles ; Besieged stronghold traditional in Egypt — example from Fifth Dynasty Tomb of Anta — its dramatic character ; Middle Empire examples ; Sensational ver- sions of New Empire ; Was there a Minoan reaction on Egypt ? Cretan Miniature Art supplies link between Early Egypt and Epic tradition.

§y$. Miniature Painting on Crystal: 'Painting' in Metal- work —

Engrailed Designs on the Blades of Weapons . . .107

Similar Subjects of Wall-paintings and small reliefs ; Other examples of painting on the back of Crystal — lens of bull's eye and pommel of dagger; Painted Crystal plaque from Room of the Throne, presenting bull-catching scene — Microscopic work ; Driven bull barred by rope ; Locks visible of leaping Cow-boy ; Perfection of Miniaturists' Art ; ' Painting in Metals ' — Minoan intarsia work ; M. M. II dagger from Lasithi ; The Mycenae daggers — Egyptian adaptations ; ' Nile pieces ' on dagger-blade — description of technique ; Hunting-leopards or Chitas ; Egyptian Caffre cats — trained to catch ducks ; Indigenous Cretan versions — pheasants for water- fowl ; Influence of Nile scenes on Minoan and Mycenaean Ceramic Art ; Also frequent on intaglios — duality of scenes on dagger-blade — other examples ; Also division into three, as on Vapheio Cups; Dividing up of designs reflects separate panels of fresco technique ; Lion-hunt on Mycenae dagger ; Successive stages in execution ; Restoration of dagger — duality of designs, again marked ; An Epic touch; Original design drawn by eyewitness of lion-hunts; Lions in Classical Greece; Comparison with African scenes of lion-hunting in modern film (Simba) ; Lion bringing down Gazelle ; Fragment of painted relief of lion from S.E. Palace Angle ; Intaglio types derived from painted reliefs ; Type of hero stabbing lion on Mycenae bead-seal; Copied by Third-Century Greek engraver — Canea find; Minoans personally acquainted with lions in every aspect ; Lion sacred to Minoan Goddess ; Vapheio dagger-blade with inlaid designs of swimmers ; Flying-fish on Vapheio blade compared with fresco ; Lilies on inlaid blade based on fresco band — M. M. Ill parallels; Masterpieces of inlaid metal-work recorded in Greek Epic — • implied knowledge of Minoan originals ; Yet the Art itself extinct long before Achaean invasions.

§ 74. Pictorial Religious Subjects on Signet-rings, reflecting Miniature Style: the ' Ring of Nestor' — Glimpse of Minoan After-World 134

Pictorial Religious Subjects on Class of Signet-rings — dependence on frescoes ; Frescoes of religious nature ; Mycenae painted tablet with descending God ; Descend- ing deities on Signet-rings ; Landscapes with rustic cult compared with Pompeian paintings ; Sacred Spring and Trees on ring ; Marine subjects ; Descending God on painted ' larnax ' ; Offertory scenes on H. Triada Sarcophagus ; ' Sarcophagal ' Art ; Origin of Minoan signet-ring from bead-seal — primarily designed for suspension ; Mourning scenes on signet from Vapheio Tomb — Dual composition of design ; Ecstatic effect of fruit of Sacred Tree ; Funereal significance of body-shield ; Parallel design on Mycenae ring ; Similar ritual refreshment of Goddess ; Separate scene of mourning at grave hung with little shield ; Association of Goddess with boy-God ; Religious scenes nearer to the Christian than to the Classical Spirit — Syrian and Anatolian affiliation • Signet-rings made for use in this World and the next — ensured protection of divinity ; The ' Ring of Nestor ' : circumstances of finding and recovery ; Arrangement of bezel in compartments and zones by trunk and branches of 'Tree of the World ' ; Comparison with Yggdrasil and ' Tree of Paradise ' ; Interpretation of design ; Butterflies and chrysalises symbolical of resurgence ; Young couple

xvi CONTENTS

PAGE

reunited in death ; Chthonic aspect of Minoan Goddess ; Her life-giving power ; Lion Guardian of Under-World; Initiatory examination by Griffin as Chief In- quisitor ; First insight into Minoan eschatology ; A true Elysion, in contrast with Hades ; Translation of design into Miniature Fresco ; Reflection of an original masterpiece in wall-painting representing the After- World.

§ 75. Upper Porticoes of N. Entrance Passage and their Painted

Reliefs of Bull-hunting Scenes : compared with Those of

Vapheio Cups 158

' Miniature Frescoes' contrasted with neighbouring deposit of painted stucco remains belonging to North Entrance Passage ; Successive stages of this ; Good ashlar masonry of Middle phase ; Fine works of reconstruction in M. M. Ilia; Proposed reconstitution of N. Entrance Passage ; Approach to Sea Gate from W. and N. ; Traces of inner Gateway ; Safeguarding measures due to considerations of police ; Lower Pillar-Hall and light-area ; Presumed Upper Columnar Hall and Corridor linking it with Central Court ; Portico overlooking W. side of Entrance Passage ; Loggia with three structural divisions ; Painted reliefs of bull-grappling scenes ; Olive-trees in background ; Scene laid in Country ; Stratum containing painted fragments ; Remains of rock-work foreground ; Bovine reliefs ; Head of gigantic bull — its noble aspect ; Foot and hoof of bull ; Part of woman's leg — ankle ring ; Part of a female thigh ; Parallels from Vapheio Cups ; Recurrence of triple division ; Did the Palace friezes supply models elsewhere ? PVagmentary reliefs in Elgin Collection ; Indications of parallel frieze in E. Gallery ; Bull's leg found in con- nexion with it ; The Vapheio reliefs ; Cup A ; Girl grappling bull ; The neck- twisting feat; Cup B— Capture through decoy Cow ; Evidences of Artistic cycle — on vases and gems; Limitations of intaglio technique; Bull on seal-impressions nosing trail — as in Cup B; Lassoing scene on gem; M. M. Ill date of painted reliefs from N. Entrance Passage ; Relatively late date of stratum with bull reliefs ; Part of reliefs seen in position by Greeks ; Olive-tree reliefs at two extremities ; Parallel survival of N.W. Portico ; Influence of remains on imagination of Hellenic settlers.

§ 76. Parallels supplied by Bull-reliefs in Elgin Collection from

the ' Atreus ' Tomb at Mycenae 192

Two contrasted groups of bull-catching scenes on Vapheio Cups and gems ; Origi- nate in Palace friezes ; Fragmentary reliefs in Elgin Collection — material proves to be Knossian gypsum ; Found in front of ' Atreus ' facade ; Supposed lion, a bull ; Bull's head resembles Knossian ; Olive-tree also introduced — from similar bull- hunting scene ; Slab with stationary bull — answers to Vapheio Cup B ; Elgin slabs from similar contrasted compositions ; Probably executed at Mycenae, by Minoan artist ; Friezes, not tympanum reliefs ; Were they from side-walls of ' Atreus ' dromos ? Possible existence of fore-hall to tomb ; Signs of restoration of dromos ; Re-used lintel-block, perhaps of entrance to fore-hall.

§ 77. The Taureador Frescoes 203

' Cow-boy ' feats in the open to be distinguished from Circus Sports ; Oriental associations of the latter — Cappadocian cylinder sealing; M. M. I a ' rhytons ' with acrobatic figures ; Covering on bulls — parallels from early Cylinders and Bull of Apis ; Elements of distinction between the two classes of subject : Theatral sports in honour of Goddess ; Miniature representations associated with shrine — the Ivory Deposit ; Fragment from ' Queen's Megaron ' ; Early example from beneath ' Kasella ' floor ; Deposit of ' Taureador Frescoes ' — probably belonging to Close of L. M. I a ; Their character and position on wall ; Elegance of female performers ; Use of ' Cestus ' round wrist ; Tumbler caught by female attendant ; Analogous

CONTENTS xvii

PAGE

bronze group from Crete — acme of Minoan metallurgic Art ; Diagrammatic sketch of acrobat's evolutions ; Conjectural form of Arena, fenced oval enclosure within rectangular; Dangerous aspect of sport — overthrown performers; ' Sacral Knots' before bulls, sign of Consecration : Comparison with sports of the Amphitheatre and Plaza ; Spanish Corridas outgrowth of sports of Roman arena ; Was there a Sacrificial sequel to Minoan sports ? Matador on Thisbe bead-seal ; Female per- formers devotees of Goddess — signs of gentle birth ; Precedence of women in Minoan Society ; Princes and Grandees participating in Spanish bull-fights com- pared ; Chevaleresque sanction in Spain just as religious in Crete ; Survival of Minoan taurokathapsia in Thessaly, in equestrian form — the Oxford relief; Parallel survival of earlier form of bull-grappling on foot ; Heroic feats of kind in Greece traced to Minoan source ; Absorbing attachment of Minoans to bull-sports : Attitude of Roman and Spanish spectators compared and contrasted ; Reaction of monu- mental remains on Greek tradition — but false idea of Captive performers.

§78. The East Postern and Bastion: runnels with Parabolic Curves — Chronological Place of That by East Steps of 'Theatral Area' 233

' Court of Stone Spout ' and Old Frontage line of N.E. 'Insula'; Stepway down East Slope ; Four lines of massive exterior walls ; East Postern — its relation to Bastion ; Recess for Warder ; Staircase down E. Bastion with descending runnel ; Fall of water controlled by parabolic curves ; Extraordinary evidence of hydraulic knowledge ; Settling basin and further course of runnel ; Presumed tank for washing linen; A second staircase; Faintly incised signs on blocks — M.M. III<£ date; Stepway with similar runnel South of 'Domestic Quarter', overlying M. M. Ill Magazines ; Further chronological materials ; Similar descending runnel by East steps of ' Theatral Area ' ; New stratigraphic results regarding this Area ; Painted Pottery with racquet and ball motives — of Senusert IPs time ; Section beneath the East steps — runnel of M. M. Ill a date.

§79. Advanced Minoan Water-system; Pipes and Conduits; Foun- tain of Fresco ; Cylinder-built Well with Minoan Signs and Melian Parallel; Sumerian Comparisons . . . 252

Advanced Minoan water-system ; Sections of terra-cotta water-pipes ; Minoan Wells ; Discovery of M. M. I a well by Villa Ariadne ; Well composed of clay cylinders — Incised signs on rims of clay drums ; Repetition of Minoan linear sign ; Cylinder- built well of Late Mycenaean date, Phylakopi ; Evidence for Greek or Greco- Roman origin of Knossian example — Glass fragments in exterior clay filling ; Was there here a surviving tradition of Minoan craftsmen ? Oriental origin of cylinder-built wells — examples at Ur in association with primitive stage of ' tholos ' tomb ; Fresh evidences of Minoan indebtedness to early Chaldaean civilization.

§80. 'East Corridor', linking East Bastion and Stepway with 'Domestic Quarter': newly discovered 'East Portico' and later 'East Stairs' 262

Fortificatory aspect of East Bastion and wall-lines — special preservation owing to recess in East slope ; ' South-East stairs ' from ' Domestic Quarter ' to ' Corridor of Labyrinth ' ; ■' North-East stairs ' from same and side access to ' Laundry ' steps ; ' East Corridor ' Main line from ' Domestic Quarter ' to stepway leading to ' East Postern ' ; Earlier Magazines of ' Corded pithoi ' underlying its entrance system ; Blocking of ' East Corridor ' — its date at first mistakenly referred to M. M. Ill ; Proves to be work of later squatters, with L. M. Ill pottery beneath it ; Division of Entrance Hall of 'East Corridor' in M. M. Wlb; Doorway opening on 'Court of in. b

xviii CONTENTS

PAGE

Stone Spout ' with massive threshold slab ; East boundary of ' Court of Stone Spout ' ; Discovery of Verandah on its North side ; ' Room of Stone Pier ' ; Lapidary's store of worked blocks of Spartan basalt ; Workshop above with unfinished Amphoras ; Bowl of Spartan basalt from ' Royal Tomb ' ; Presumed wooden steps from lower Store-room to upper Workshop ; ' Lobby of the Wooden Posts ' — block above Balustrade pier raised to original position ; Characteristic M. M. Ill timbering ; Symmetrical reversing of W. and E. balustrades ; Discovery of ' East Portico ' — massive foundation blocks of four columns ; Agreeable open-air retreat from closed areas; Later 'East Stairs'— constructed after a seismic collapse ; Great Deposit of L. M. I a pottery beneath steps — vessels of ordinary use ; Numerous Artisans within Palace ; Minoan flower-pots; Comparison with similar contemporary deposits — signs of severe dislocation c. 1520 B.C.; Evidence for dating spiral decoration in neigh- bouring halls.

§ 8r. Further Reconstitution of the ' Domestic Quarter ' ; General

View of the Upper Story System 282

Erroneous popular impression of Palace as a Maze ; True derivation of name 'Labyrinth'; Maze as sphragistic motive of Egyptian origin; Regular and four- square construction of Palace ; Scientific planning of ' Domestic Quarter ' ; Com- parative isolation of Women's Chambers ; Exceptional preservation of ' Domestic Quarter ' ; Surprising discovery of Grand Staircase — difficult tunnelling ; Fallen materials, petrified by gypsum percolations, support Upper Stories ; Re-supporting and restoration of floors throughout Quarter a necessary work ; Reconstitution facilitated by use of ferro-concrete ; Flooring over of Lower Stories of ' Domestic Quarter'; Restoration of floor of ' Upper Hall of Double Axes' ; The System of Light- wells ; Light-courts necessary protection against fierce South-East and North-West winds ; Wind erosion of rocks due to S.E. blasts — slits cut in cliffs and Palace wall; Upper Story System above Great Halls; Fragment of L. M. II fresco in situ on wall; Decorative frieze associated with ' Upper Hall of Double Axes ' — contemporary with spiral friezes of this area ; West Light-well of Hall ; Special facilities for social inter- communication between Upper and Lower Hall ; Partial segregation of sexes ; Private Chamber (Thalamos) above ' Queen's Megaron ' ; Windowless Chambers for Treasury and Archives — These connected with a Shrine of the Double Axes ; Room of Stone Bench ; Fragments of Processional Fresco — belong to partial restoration towards close of L. M. I a ; Service Staircase ; Compact planning of inner region.

§82. Grand Staircase and Loggia as further restored ; 'Shield

Fresco' and its Derivatives — Military and Religious Import 299

Further reconstitution of Grand Staircase — to fifth Flight ; Impressive effect : a vision of the Past; 'Loggia of the Shield Fresco'; Fresco fragments thrown into nook below; 'Shield Fresco' replaced in replica; Comparison with Tiryns frieze — mala- chite green pigment in latter, of Egyptian derivation ; Tiryns shield frieze copied from Knossian fresco ; The great body-shields of Staircase loggias represent actual shields as hung in Hall below ; Shield decoration answers to military spirit that marks last Age of Palace ; Indications of new Dynasty — introduction of Linear Script B ; The ' Chariot Tablets ' ; Remains of smaller Shield Fresco ; Influence of Shield Frescoes on Ceramic design ; L. M. I a polychrome Goblet ; Shields and Spirals on ' Palace Style' Amphora; Shields on L. M. \b 'aryballos', Phylakopi ; Imported vessels with Shield designs at Gezer and Late Minoan sword ; Reflection of Shield friezes on seal-types: 8-shaped shield as religious emblem; ' Baetylic' function — comparison with ancilia ; Shields as decorative adjuncts ; Amuletic beads in shield form ; Minoan shield in field of signet gems as religious indication; Associated with Minotaur; Parade of Shields on Stairs suggestive both of temporal power and of divine pro- tection.

CONTENTS xix

PAGE

§83. 'Hall of the Double Axes' as reconstituted .... 318

Down Grand Stairs to lower Halls ; Control of doors — marks of their swing and bolts ; Massive wooden lintels ; Great heaps of calcined material ; Remains of car- bonized columns; Downward taper of Shafts — comparison with primitive stone pillars of Malta and Balearic Islands ; Knossian examples of such now to hand ; Impressions of flutings on clay plaster; The ' Hall of the Double Axes' — floor above reconstituted ; Gypsum elements of lower Hall thus preserved ; Wooden framework filled in with cement ; Painted spiral frieze above dado slabs ; Western Section of Hall— An 'Audience Chamber'; Central Section or 'Inner Hall'; Eastern or 'Ex- terior' Section — its Well; Symmetrically opposed doorways of Eastern Section; Back passage to 'Queen's Megaron ' ; Connexions with stepway from S.E. of Central Court; Roofed Annexe to N.E. of this Section; E. Wall and S. Corner — View of Eastern and Southern Porticoes as restored ; Discovery of descending steps by S.E. doorway — Access to Corridor of 'Labyrinth Fresco'; ' Tarazza' flooring of Light- courts ; Results of tests beneath pavements ; Early ceramic and fresco remains ; Gypsum paving of Hall M. M. Ilia; Area of Hall; High walls of Light-courts ; Traces of remains of wooden Canopy and Throne in ' Audience Chamber ' ; Inner Section of Hall ; Door-jambs serving as piers between Sections ; Elasticity of System ; Two-leaved doorways ; Window-like openings above doorways ; Red-coloured panes — ? parchment ; Inner Hall capable of isolation ; Spiral frieze compared with that of ' Shield Fresco ' ; Presumption that real shields were hung along fresco band ; Actual shields replaced, in replica, on wall ; Restored view of Inner Hall — the Chieftain at ease ; Religious side of scene ; Indications of Double Axe Cult ; ' Rhyton ' for libation ; Clay ' fire-box ' from Well ; Baetylic cult of Double Axes.

§84. The 'Queen's Megaron' as reconstituted 349

How far were sexes segregated in Palace ? Epic tradition ; Isolated Halls at Tiryns, and Mycenae ; More open system at Knossos — movable hearths ; Yet certain elements in common with Mainland plan ; Notched plume decoration of Knossian hearths on that of Mycenae ; Analogy supplied by relation of two Halls at Tiryns ; The ' Dog's-leg Corridor ' ; Double Compartments of ' Queen's Megaron ' ; Private staircase and upper ' Thalamos ' ; Pillared Stylobates of 'Queen's Megaron'; Bi- columnar Portico and Eastern Light-area ; Evidences of wind erosion ; Data for earlier history of ' Megaron ' ; Successive pavements and levels ; ' Mosaiko ' pavement below Gypsum slabbing ; ' Kalderim ' pavement below this ; ' Mosaiko ' system M. M. II b— its original extension ; Other evidences of M. M. II b date of ' Mosaiko ' pavements ; ' Nature-printed ' fresco with sponges associated with ' Mosaiko ' system ; Ceramic parallels to process ; Printed sponges a stage towards later 'Marine style'; Red gypsum border of ' Marine ' panel ; Gypsum pavement answers to later struc- tures ; Underlying stratum M. M. HI a in East section of ' Megaron ' ; In inner section L. M. I a sherds predominate; Pillared stylobate connected with M. M. Ill paving; Wooden benches along stylobates ; Double Axe socket found by Central Stylobate ; Fresco of Dancing Lady ; Painted stucco relief with papyrus pattern — recalls ceiling decoration of Orchomenos ; Parallels from Tirynthian friezes ; Fits in with L. M. 1 redecoration ; South Light-area — M. M. II walls; Incidence of light ; Eastern Light- area — successive walls; Reoccupation wall; Remains of ' Dolphin Fresco'; Over- lapping of decorative systems ; Painted stucco on Light-well wall ; Section of ' Dol- phin Fresco ' reconstituted ; Painted clay tub; Bath-room of ' Queen's Megaron' — remains of spiral frieze; Painted clay bath — L. M. II reed decoration; M. M. Ill bath compared ; Water transport by hand labour ; Back passage with spiral band ; Toilette Room and Latrine ; Plaster dais ; Great stone water-shafts in wall — possible cistern; Exceptional conveniences of 'Queen's Megaron' — ideal interior view; Holes in pavement slab for game — Parallels from Mallia and Gournia ; Miniature Fresco showing boys playing pavement game.

xx CONTENTS

PAGE

§ 85. East Treasury (later, Archives) and Derivative Relics : Bone

Fish with Alphabetiform Signs 397

Windowless chambers for Stores nnd Treasure ; ' The Lair ' ; Early ' kalderim ' pave- ment below with contemporary door-jambs ; Clay and plaster floors (M. M. Ill) of later store-room ; Relics fallen from Treasure Chamber above ; Extension of deposit under Stairs — 'The Ivory Deposit ' ; Parallel stratification under Stairs ; M. M. Ill b pedestalled Vases at top — their Egyptian pedigree; Fragments of Miniature Fresco found with 'Ivory Deposit ' relating to Bull Sports ; Upper Treasury Chamber later used for Archives — sealings and tablets of Linear Class B ; Relics from ' Treasury ' also found in 'Room of Stone Bench'; Faience plaques as from 'Temple Reposi- tories ' ; Parts of Ivory Casket and Wooden Chest ; Rock-crystal bowl ; Pendant in form of gold heart — similar amulets ; Gold fish — Scarus Cretensis ; The lion jewel ; Deposit in Drain Shaft : Bone ' fish ' with alphabetiform signs ; Simple geometrical character of such signs; Inlayers' signary ; Accompanied by varying numbers; Segments of bracelets with similar marks and numbers ; Were both classes of objects used for game ?

§86. Further Relics derived from East Treasury: Sphinxes,

mlnoan and hlttite, and leaping youths of ivory deposit 41 3

Gold-plated bronze attachments — perhaps from Lion's mane; Miniature Double Axes — bronze, gold-plated — from small shrine ; Plumed crest and part of wing of Ivory Sphinx ; Comparison with naturalistic ivory wing of small bird ; The Minoan Sphinx ; Minoan Oedipus at grips with it on Thisbe bead-seal ; Name of Sphinx Hellenic, Art form Minoan ; \Yingless Sphinx of M. M. II Signet ; Hathoric curls — of Hittite derivation distinct from Egyptian type ; Steatite locks of Sphinx's head from drain-shaft deposit ; Oriental comparisons — Hathoric elements ; H. Triada Sphinx of Hittite type — supplies key to restoration of Knossian head ; Twisted locks ; H. Triada figure compared with Chaldaean inkstands ; Minoan ink-written inscrip- tions ; Immediate source of H. Triada Sphinx Hittite ; An imported object ; Parallel example recently found at Tylissos : Minoan female figurine of steatite found with Sphinx at Tylissos — peaked cap a novel feature ; ' The Deposit of Ivories' ; Figures of acrobatic youths ; The Leaping Youth — marvellous elan ; Chryselephantine Art : Curled flying locks of bronze, gold-plated ; Part of larger arm : Figures engaged in Sports of Bull-ring — these under divine patronage ; Faience head of bull ; A miniature Minoan Corrida — perhaps exhibited in Sanctuary.

§ 87. Chryselephantine Figurinks of 'Boston Goddess' and Boy-God

connected with ivory deposit: the mother goddess and

Child 436

The acrobatic ivory figures connected with bull-ring ; Religious elements of Treasury Deposit — miniature bronze Double Axes and part of shrine on fresco; the Sphinxes; Were there also images of divinities ? Facilities here for abstraction ; ' Boston Goddess ' : parallel to Rnossian Snake Goddess — Knossos its probable source ; Locks secured as in acrobatic ivories ; Physiognomy individual and modern ; Emer- gence of ivory boy-God, in the same style and probably from same Knossian source ; Fragments of steatite cup with male heads in relief; Ivory figurine of boy-God, as found ; Belt already fitted to child's body ; Very young infants beltless — Palaikastro examples ; Rivet holes in ivory figurine for gold-plated belt and kilt ; Girdles on young girls ; Proportions of boy-God ; Comparative girth of adult male figure ; Adaptation of human form to early waist constriction ; Looser belts worn by elderly- men ; Tight waists begin with M. M. Ill ; Boy-God on tiptoes andadorant ; Delicate execution of toes; Foot of M. M. II a figurine; Waving hair and features of boy- Clod — recall ' Boston Goddess'; Correspondence in height ; Same group — Mother and adoring Child; Matriarchal stamp of Minoan Religion; Same Goddess with

CONTENTS xxi

PAGE

various attributes ; Youthful male adorants — fine bronze example ; Adorants on signets ; Armed youthful male figure grouped with Goddess ; Adult warrior God exceptional till latest Minoan epoch — figures of Resheph ; Mourning scene for youthful warrior God on Minoan signet ; Minoan Goddess, resurgent as Gaia ; Un- armed boy-God — Oriental parallels; Minoan versions of Mother and Child ; Terra- cotta idol from Knossian tomb ; Adoration scene on Thisbe signet ; Goddess with child on lap, holding cymbals ; Adorant warriors bearing gifts ; Reeds borne, as in honour of Cybele ; Metal vessels as offerings ; Adoration of Magi compared — Christian version on ring-stone ; Thammuz at Bethlehem.

Supplement to Sectio>i 87. Late Minoan and Mycenaean images of the Syrian Lightning- God Resheph : with Syrian and Hittite comparisons . . . . . .477

§88. Great 'East Hall' resuscitated 481

Main approach to lower halls of ' Domestic Quarter ' from above : its ground floor \\\q piano nobile ; Somewhat lesser height of upper stories ; the fourth landing, below level of Central Court ; Flight of Stairs thence to roof terrace ; Architectural analogies with West side of Court ; Evidences of great ' East Hall ' ; Correspondence with ground-plan of earlier 'East Hall' basements; These choked by Earthquake and subsequently filled in; M. M. Wla relics of earlier Hall — including 'Ladies in Blue ' fresco ; Earlier East Hall narrower with lateral light-area ; Break between M. M. Ill plan and M. M. II structures; Important relics found in Loom Weight area — M. M. II b ; Restored plan of later ' East Hall' ; Blocked doorway of filled-in Magazines ; Earlier Hall approached by descending steps ; New Hall and adjoining East facade in harmony with later facade West of Court ; ' Stepped Porch ' balanced by Stepped Portico of new ' East Hall ' ; Striking correspondence in levels and espacement with structures West of Court ; Agreement between Steps of ' East Hall ' and those of ' Stepped Porch ' ; Position of Columns in Portico ; Lateral openings to landing : Doorways of East Hall ; Central light-area of Hall evidenced by drain- head — course of drain to blind well ; M. M. Ill a conduit re-adapted for later Hall ; Evidence of use to L. M. II ; Square central peristyle of later East Hall ; Interior section, with side doors ; Parallelism with plan of Mainland ' Megaron ', but this Hall not isolated ; Painted plaster high reliefs precipitated from Walls ; Fragments thrown down at time of final Catastrophe; High reliefs, executed in the great Transitional Age, preserved on Walls to the last.

§89. The Painted High Reliefs from East Hall; and Bronze

Locks of Hair from Gigantesque Female Image . . . 497 Agonistic and acrobatic character of High Reliefs ; Fragment from group of two wrestlers — Sir W. Richmond's technical appreciation ; Boxing bouts and hand-to- hand encounters in Minoan Art ; Upper arm — attributed to fallen pugilist ; Com- parison with small relief of H. Triada ' rhyton ' ; Forearm of boxer; Forearm of Taureador, with clenched fist — Sir W. Richmond on this ; Hand with careful render- ing of veins ; Leg seen from back, perhaps of pugilist ; Calf in profile and ankle ring ; Female breasts — probably of Taureador ; Fragments of frieze with opposed Griffins tethered to Columns ; Palm Columns ; Cornice with triple gradation beneath Griffin Frieze ; Opposed Griffins compared with signet-type ; Column here baetylic form of divinity; Oriental origin of such ' antithetic' schemes ; Frequency of opposed animal types in Late Minoan Art ; High reliefs probably on walls of back Section of Hall ; Double tiers of reliefs ; Griffin Friezes along side walls in narrower porticoes ; High reliefs evolved from lower ; Gypsum reliefs at Mycenae — by Minoan hand ; Con- current progress of works in the round ; Hand of stone figure, half natural size ; Life-sized female head in painted stucco, from Mycenae ; Influence of wood carving visible in head ; Discovery of large bronze locks of hair in carbonized deposit by N. wall of East Hall ; Belong to wooden figure of gigantic proportions ; Female sex indicated ; The ' Xoana ' of Daedalos : A gigantesque cult image of Minoan Goddess Probable place, in back Section of Hall.

XXII

CONTENTS

ARCHITECTURAL PLANS AND SECTIONS

Facing

Fig. ia. Restored Plan showing Upper Floor of Sanctuary Quarter, West of Central Court Fig. i b. Plan of Part of North Palace Section showing Area of ' Early Keep ', N.W. Portico

and Entrance Passage and Northern Entrance System Fig. 2. Revised Plan of ' Initiatory Area' and Lustral Basin . Fig. 3. Plan of Northern Lustral Basin ......

Fig. 5. Plan of Later Doorway in Corridor South of Lustral Basin . Fig. 7. Horizontal Section showing Primitive Lock: South House . Fig. 8. Isometrical Sketch of Jamb of Gypsum Doorway, South House, showing Sockets

for Bar and Bore-holes for Locking-pin .......

Fig. 9. Part of North-West ' Insula', showing Position ot Principal Finds

Fig. 12. Diagrammatic Section of East Part of Room of ' Knobbed Pithos', Sec.

Fig. 106. Restored Upper Plan and Elevations of Northern Entrance System . Facing

Fig. 108. Sections across Pillar Hall (b. b.) and Upper Passage with Porticoes (c. c.)

Fig. 1 14. Elevation of West Side of North Entrance Passage showing Earlier and Later

Levels of Roadway .......... Facing

Fig. 166. Isometric Projection of East Bastion and Postern ......

Fig. 168. Plan of East Bastion and Postern .....

Fig. 169, a, />, (a) Plan, enlarged to six inches wide, and (a) section at A of Staircase of

East Bastion .............

Fig. 169, c. Section b-b of Staircase of East Bastion ......

Fig. 172. Section under 10th Step of Eastern Flight, * Theatral Area', showing Strati

graphic Horizon of Parabolic Channel beside Stone Platform .... Fig. 173. Sections of Terra-cotta Water-pipes, below ' Corridor of Draught Board' Fig. 178. Section of 'Tholos' Tomb of Sumerian Queen and Well-like Shaft consisting of

Terra-cotta Drums to receive Libations .........

Fig. 183. Plan of Region including 'East Corridor', 'Court of Stone Spout', and 'East

Portico '............

Fig. 213. Holes for Bolts and Scorings on Thresholds marking Swing of Doors

Fig. 218. Revised Plan showing Central and Eastern Sections of 'Hall of the Double

Axes ' and Adjoining Area ......... Facing

Fig. 223. Elevation, Plan, and Section of Lime Plaster Mass showing Impress of Wooden

Throne and Canopy ........

Fig 232. Plan showing Part of the Later Akropolis (L. M. Ill) at Tiryns with Propyla and

Palace Halls

Fig. 233. Plan of Entrance to Private Staircase from 'Queen's Megaron' showing Later

Pier inserted for Support of Upper Structures .......

Fig. 244. Plan, Section, and Elevation of Central Pillared Stylobate of ' Queen's Megaron

above the Floor-level ...........

Fig. 245. Section of Central Pillared Stylobate of ' Queen's Megaron ' showing Wooden

Seats covered with Plaster as restored and Underlying System . Fig. 249. Plan of ' Queen's Megaron ' and Connected Structures Fig. 253. Plan and Elevation of Bath-room ..... Fig. 259. Elevation of Corridor Wall showing Spiral Frieze Fig. 264 a. Plan of Platform near South-west Corner of Central Court at Alallia showing

position of Cupped Table with Bench behind ....

Fig. 266. Plan of First Floor of ' Domestic Quarter' .... Fig. 338. Basement Plan beneath ' East Hall ' (revised) .... Fig. 339. Section o^f ' Stepped Porch ', West of the Central Court . Fig. 340. Conjectural Plan of Great ' East Hall' as restored in M. M. \\\b Fig. 341. Plan showing Course of Stone Conduit from Drain-head of Shaft to 'Court of

Stone Spout ' .............

Facing

PAGE

3

10 12 J3

14

19

24

161

164

171 237 239

240 241

249

253

260

270 320

328

336

352 353 367

368

375 382 388

394 400 487 490 491

493

CONTENTS

xxn

GENERAL PLANS

{In Pocket at End of Volume.)

D. Reconstructed Elevation of Grand Staircase. By Christian C. T. Doll.

E. Plan of Domestic Quarter : Ground Floor. By Christian C. T. Doll.

F. Plan of Domestic Quarter : First Floor. By Christian C. T. Doll.

G. Constructional View of ' Hall of the Double Axes ' and Structures to the West of it, with

Upper Elements removed.

TABLES AND COMPARATIVE EXAMPLES OF SIGNS

AND SCRIPT

Fig. 267. Comparative figure showing Egyptian Pedigree of M. M. Ill Pedestalled Vases Fig. 269. Alphabetiform Signs and Numbers on Bone 'Fish' .....

PAGE

402 406

LIST OF COLOURED PLATES

PLATE

XV. Ceiling Pattern in Painted Stucco Relief Facing 30

XVI. Panel of 'Miniature Fresco' restored, showing Central Pillar Shrine with

Grand Stands on either Side and Crowds looking on at Sports Facing

XVII. Miniature Frescoes of Seated Ladies on Grand Stand . . . Facing

XVIII. Miniature Fresco of the Sacred Grove and Dance . . . Facing

XIX. Bull-catching Scene painted on back of Crystal Plaque . . Facing

XX. Inlaid Designs on Mycenae Dagger-blade : Chitas and Caffie Cats hunting

Ducks on Banks of the Nile ...... Facing

XX a. Design on the 'Ring of Nestor' translated into a Painting of the ' Miniature' Class ........... Facing

XXI. Part of Fresco Panel : Leaping Girl from Scene of Bull-ring . Facing

XXII. Fountain {Jet d'Fau) from Painted Stucco Panel, House of Frescoes Facing

XXIII. Middle Band of 'Shield Fresco' from Loggia of Grand Staircase . Facing

XXIV. Restored Interior View of Inner Section of ' Hall of Double Axes ' showing

Suspended Shields. Doors and Windows to left opening on Western Section .......... Facing

XXV. Fresco Fragment showing Boys playing Game. Fresco Fragment showing Upper Part of Dancing Girl ...... Facing

XXVI. Restored View of ' Queen's Megaron ' with Doorway leading to East Section and Light-well beyond. To right, Window opening on Southern Light- area : to left, Private Staircase to Upper ' Thalamos ' . . Frontispiece

47

49

67

109

114

J57 216

254 306

346

37°

LIST OF SUPPLEMENTARY PLATES

XXXII. Interior View of Northern Lustral Basin, with Upper Part restored. Staircase Balustrade, looking towards North-West. Bands of Dark Wall-painting are shown above Gypsum Dado Slabs (here replaced in plaster) with Speckled Panels. (See p. 9.)

XXXIII. Interior View of Northern Lustral Basin looking down First Flight of Stairs to

West. (See p. 9.)

XXXIV. View of East Bastion and Quadruple Line of Walls, from the North. (See p. 234.) XXXV. a, Vertical Post (socketed in M. M. II Masonry) with Base tapering upwards, East

Corridor. (See p. 65.) b. Plaster Mass showing Impress of Fluted Columns, East Slope. (See p. 323.)

xxiv CONTENTS

PLATE

XXXVI. Steatite Locks of Sphinx of Hittite Type from Deposit in Stone Drain-shaft

(See p. 420 seqq.) XXXVII. a. Plaited Locks of Sphinx in Steatite : Mycenae. (See p. 422.) B. Ivory Figurines of Boys : Palaikastro. (Sse p. 446.) XXXVIII. Head, Forearms, and Legs of Figurines from ' Deposit of Ivories '. (See p. 428 seqq.) XXXIX. Bronze Votive Figures of Adorants. (See p. 459 seqq. and p. 461.) XL. Painted Stucco Relief of Lower Part of Man's Leg. (See p. 508.) XLI. Painted Stucco Reliefs : a, Man's Hand, and b, section (see p. 506) ; c, Upper

Part of Griffin's Hind Leg. (See p. 508 seqq.) XLII. Restored Handle of Dagger-blade from Mycenae (with Inlaid Design of Lion-hunt). (See p. 119.)

ERRATA

Page 473, line 21, for ' wreaths ' read ' reeds '.

Page 525, lines 28, 29. Parenth before the word ' including '	eses should replace and after the word	the commas ' below '.
Plan G in pocket. The Plan should be om	Well in itted.	bottom left-hand corner of this

§ 68. ' North-West Insula' North of Central Court, with Ramp Passage from N.W. Portico and Lustral Area: Basement Chambers above Early Keep.

' Arorth- West Insula '—its artistic revelations ; Functional i)nportauce — supplements that of Western borders of Central Court ; Area West of Central Court chief Cult Centre — -retrospective viezv ; ' Stepped Porch ' , built over earlier Cists ; Block North of Porch, of L. M. II date : Includes ' Room of the Throne' ; N.E. Corner of old facade and of Early Western' Insula' ; NortJi- Westeru ' Insula 'formed by ' Early Keep ' ; Elements of fortification in Early Palace ; Ramp passage round Keep from N.W. Entrance ; 'Initiatory Area' and Votaries' entrance ; ' North- West Lustral Basin' of M.M. Ill a date ; Restored plan and elevation of' Lustral Basin ' and Superstructures ; Isolation of Sunken Area from other buildings ; Minoan bronze 'key ' found in door- way of later passage ; 'North-West Entrance' partially blocked; Basement Chambers above ' Early Keep ' / Plinth bordering Court, and steps down ; M. M. II a paved floor over Walled Pits, belonging to large basement Chamber ; Later dividing wall — M.M. lib; Gypsum steps re-used for pavement ; Central Pillar of Western Basement in M. M. II b / Exceptional evidences of strati- fication in ' Room of Knobbed Pithos' ; Large black steatite vessel with spiral reliefs; ' Sa/frou-Gatherer' Fresco; Later floors with stone lamps and tablets of Class B ; L.M. Ill a. floors above these; Basements bordering Central Court ; Rearrangement of supporting pier ; Further evidence supplied by ' Room of Knobbed Pithos ' ; M. M. Ill a pottery beloiv later pavement (M. M. Ill b) ; Extension of M. M. Ill b floor, its continuous use in restored Palace ; 'Miniature Frescoes' and Spiral Ceiling fallen on this level ; Stone lamps, and basins for ritual sprinkling.

The Palace region to which we are now led, to the North of the Central The Court, may be fittingly described as the ' North-West Insula'. Its Eastern insula and Western boundaries are indeed well defined, on the one side by the ' North Entrance Passage ', on the other by a ramp passage winding up from the North- West Portico round the massive outer walling of the ' Early Keep ', which forms the nucleus of the ' insula' (see Plan, Fig. 1 b, facing p. 7). As regards its upper structures this region had suffered much dilapidation, but it proved to be of the first importance in the revelations that it has afforded of certain classes of artistic work.

Here was brought to light the 'Saffron-Gatherer' fresco, the earliest figured painting of which any record has been left at Knossos. Here, too,

III. B

ARTISTIC TREASURES OF N.W. INSULA

Itsartistic derived from a corner sanctuary, lay the fallen remains of the 'Miniature' tions! frescoes, illustrating an astonishingly lively development of pictorial design in the transitional Age that heralds the ' New Era '. On the Northern borders of this area, beneath and near a later threshing-floor, great heaps of pieces of painted plaster were uncovered, some of them — like those depicting parts of an embroidered robe — of exceptional interest. On the other hand, by the bastions of the adjoining Corridor East, precipitated from the back wall of the portico above overlooking the 'North Entrance Passage', there occurred a series of fragments of painted stucco reliefs belonging to an extensive frieze representing bull-catching scenes and, included among them, the noble head of what to the Minoans was the King of Beasts. Various comparisons, as will be shown below, enable us to recover the general character of this composition. Together with a companion frieze on the opposite side of the Entrance Passage, it seems in a special way to have impressed the imagination both of contemporary and later beholders, and a reminiscence of its designs is traceable on the Vapheio Cups as well as in two sculptured plaques brought by Lord Elgin from the ' Tomb of Agamemnon '. The functional importance of this ' Insula ' was also great, owing to the fact that what is ex hypothesi a pilgrims' entrance from the ' Initiatory Area' beyond, with its ' lustral basin ', led here, as already mentioned, by a winding ramp and passage way to the N.W. Corner of the Central Court and thus to the sanctuary region on its Western side. That this sanctuary character was shared by a large part of this ' North- West Insula ' bordering the Court on the North appears from the data supplied by a series of finds made within it. This region in fact takes up and illustrates on its own lines the religious functions fulfilled in a pre-eminent degree by that West of the Court. These have received detailed consideration in the concluding Sections of the pre- ceding Volume of this work, but it may be well to recall here the salient features of the sanctuary quarter on that side to which the ramp passage primarily led.

Func- tional im- portance of ' N.W. Insula'.

Area West of Central Court : chief cul centre : retro- spective view.

Retrospective View of Sanctuary Area W. of Court.

The Palace region between the Upper Long Corridor and the Nor- thern section of the facade bordering the Central Court, as shown in the restored plan (Fig. 1 a), includes within it a group of structures that sufficiently mark it as the true cult centre of the restored building.

Its nucleus is the ' Tri-Columnar Hall' which forms the principal objec- tive of a splendid architectural suite, beginning with the ' West Porch ' and the ' Corridor of the Procession ' and directly approached by the ' South

I

ft

REMAINS OF PlEPsS - lb -

CENTRAL

TRI-COLUMNA^

HALL

COLUMN BASES R£PLACE.D OVEfL PlEJtS OF CRYPT

CENTRAL LOBBY

VESTIBULE OF UPPER. PROPYL AEU?M

ST \Q

CENTRAL COURT

-5HRINE

UPPER. VERANDAH

'5

20 METRES

1 1 1 n

I I I E

I I

I I 1

Fig. 1 a. Restored Plan showing Upper Floor of Sanctuary Quarter, West of Central Court. (Reproduced from Plan C at the end of Vol. II, Part 2.)

SANCTUARY AREA W. OF CENTRAL COURT

Retro- spect of Sanctuary Area to West.

' Stepped Porch ' built over earlier cists.

Intrusive

structural

block

N.of

Porch.

Including ' Room of Throne '.

Propylaeum ' and its stepped continuation above, through a Central Lobby. At the South-East angle of this Hall opens, as described above, a small chamber which, as its contents show, — including the finely carved ' rhytons' in the shape of lions' and lionesses' heads — served as the Treasury of a Sanctuary. Two of the columns of the Hall were in fact supported by the stone pillars of corresponding crypts below, to which — in view of the vats for the blood of victims, the double axes repeated on the pillars themselves, and the numerous analogies now available — a sacral character must certainly be assigned. These dark vaults, dedicated to the cult of the sacred weapon and its associated divinity, led in turn on the Court side to a small columnar shrine of the Minoan kind, in the North wing- of which was found a whole deposit of clay seal-impressions depicting the Minoan ' Rhea ' herself on her lion-guarded peak.

The adjoining ' Temple Repositories ' of the preceding Palace sanc- tuary, the date of which goes back within the borders of the earlier phase (a) of M. M. Ill, had been paved over by the restored basement floors. At the same time, too, the contemporary system of cists containing similar ceramic remains that ran North from the Eastern Repository had been covered over by the newly constructed ' Stepped Porch ' which gave access from the Court to the Central Staircase of this wing of the building. The steps of this Porch also form a break in the double facade that runs South along the whole border of this Section of the Central Court.1

The inner facade belonging to the earlier Palace and consisting of gypsum orthostats on a limestone plinth can still be traced beneath the steps of the Porch. Beyond this point, however, both it as well as the Northern series of M. M. Ill Cists, and indeed the whole palatial unit to which the Central Staircase belongs, are entirely broken off by an intrusive block of somewhat later date and which in fact bears every evidence of dating from the latest Palace period (L. M. II).

The principal chambers of this block are the 'Room of the Throne' and its Ante-room, but, as the decoration and contents of these connect themselves with the closing phase of the building, it has been thought better to reserve a description of them to a later Section.2 Here it need only be observed that this conglomeration of Chambers, following on to the earlier remains of the sanctuary quarter of the Palace, presents itself a strong religious character, as is clearly shown by the lustral area and small ' Con- sistory' hall round which it centres.

1 See P. of M., ii, Pt. II, p. 798 seqq., and Fig. 525, p. 803.

2 In the concluding volume of this work.

N.E. CORNER OF FACADE 5

A terminus a quo for the dating of this intrusive block of buildings was l. M. II supplied by the remains of pottery found under the second and third thres- ^^ holds (from the South) of the doorways leading down to this Ante-chamber thres- r

. holds 01

from the borders of the Central Court, which in addition to some L. M. I ante- fragments contained others of the best ' Palace Style' belonging to the early room" part of L. M. II.1 These doorways preserved the line of the outer facade and abutted on the slabs of the same limestone pavement that occurred elsewhere on this side of the Central Court. This pavement, of which the best preserved remains lay, as already noticed,2 in the North-West angle of the Court, was itself clearly contemporary with the earliest elements of the restored building. In this angle, as has been shown, it had displaced an earlier M. M. Ill pavement immediately superposed on a well-marked stratum the latest elements of which belonged to M. M. II and which itself immediately overlay the Neolithic — another proof of the levelling away of the intermediate strata on this part of the hill in order to layout the Central Court and the adjoining regions of the Minoan Palace.3 In this stratum, in a M. M. II b medium, was found the lower part of the diorite Middle Empire statuette of User.4

In the same layer, at depths varying from 50 to 70 centimetres, there Cistern came to light two small stone drains which converged on a common channel drains, running East to a larger stone-built tributary of the main ' Cloaca ' of the Central Court (see Fig. 9). The more Southern 5 of the two small conduits, running from the South-West, had been cut off by the front line of the Ante- chamber of the ' Room of the Throne '. The more Northern proceeded from a cistern bordering the North-West corner of the earlier facade. In it occurred M. M. \\b pottery including a polychrome cup. The cistern for which it provided an overflow channel had been, no doubt, filled by rain- water from the roof-terraces of the earlier facade line.

On the Northern border of the Ante-room of the ' Room of the Throne ' N.E. to which the later frontage line here belonged, the older line — blocked up to 0f old this point by these intrusive structures of the concluding Palace period — fa?ade reappears for a short distance. Its plinth and orthostats, however, almost at 'insula'. once curve Westwards forming a rounded outline which, as already observed, corresponds with what seems to have been a similar feature at the diagonally

1 19 1 3 tests 59, 60. was better preserved, was 30 cm. deep and

2 P. of M., ii, Pt. II, p. 800, Fig. 522. 12 cm. wide. The upper border of its side slabs

3 See P. of M., ii, Pt. I, p. 5. was 50 cm. beneath the surface, that of the

4 Ibid., i, p. 286 seqq. other drain 70 cm. s The more Southern of these drains, which

6 N.W. INSULA FORMED BY EARLY KEEP

opposed South-West Corner of an extensive 'insula' of the Palace in its original form.1

North- Western ' Insula ' formed by ' Early Keep '.

N.w. The Corridor that follows the Western turn of this old angle of the

formed facade — named from a stone basin found within it — separated the large

by 'Early West ' Insula' in question from another very characteristic structural island

also dating in its inception from the earliest or ' proto-palatial ' stage of the

building. This is the ' North-West Insula' already mentioned as that with

which we are in this place principally concerned. Its area, as we have seen,

was originally composed of the ' Early Keep', which dominated the Northern

approach to the Palace and at the same time the North-West Section of the

Central Court.2

This building, with its massive walls and deep dungeon-like, walled pits, recalls the donjon of a Norman Castle. However comparatively open most of the building may from the first have been, it is difficult not to recognize in this tower-like structure, guarding the Sea-Gate of the Palace, a real strong- hold for its early princes. It may be added, indeed, that the same fortifica- tory intention attaches to the adjoining Northern Entrance system and in a high degree to the mighty walls that run East of it in parallel lines, the more so as in the upper part of this area there was no such need as on the Eastern slope beyond of massive terrace walling. The blocks may well compare with those of the citadels of Tiryns or Mycenae. Elements It is true that the generally open character of this and other Cretan

fication in cities under the Pax Minoica of later days points to the neglect of such £a,rly defensive works. But we are already warned bv the massive enceinte walls

Palace. ■>

of the ' City of Refuge' on the sacred height of Juktas, going back to the very beginning of the Middle Minoan Age,3 that the need of fortification was still recognized in Crete at the epoch answering to the 'proto-palatial' stage at Knossos.

In the Cyclades,4 at Phylakopi in Melos, at Chalandriane in Syros, at St. Andreas in Siphnos, and elsewhere we have evidences of similar walled strongholds going back in some cases at least to the last Early Cycladic Period, contemporary with E. M. III. The faience mosaic, moreover, in any case not later in date than the beginning of M. M. Ill, has given us a

1 See P. of M., ii, Pt. II, p. 66 1, and a white line in the General Plan A at the end cf. Vol. i, p. 130, Fig. 96 (also Vol. ii, p. 799, of Vol. ii.

Fig. 521). 3 See P. of M., i, pp. 156, 157.

2 For the ' Farly Keep' see Vol. i, pp. 136- 4 See especially D. Mackenzie, Phylakopi, 9 and Figs. 100, 101. Itsareais indicated by pp. 255-8.

GREAT DEPOSIT OF TABLETS

PART OF FALLEN ,

Thown of consecration

Fig.

1 b. Plan of Part of North Palace Section showing Area of ' Early Keep ', N.VV. Portico and Entrance Passage and Northern Entrance System.

RAMP PASSAGE FROM PORTICO AND LUSTRAL AREA 7

glimpse of a fenced city, perhaps the port town of Knossos, with outer towers and houses on the wall, in addition to its street facades. It, more- over, during the later epoch, unified dominion and immunity from foreign attack led to the comparative disuse of such precautions in Crete itself, there is no reason to suppose that walled defences were neglected in the more out- lying districts of the Minoan World. Certain traditions, indeed, of this older system of fortification clung to both the palatial and the civic architecture of Crete down to Late Minoan times. The system of successive returns in the wall line — a survival of projecting bastions such as we see them in early Anatolian sites like Sendjirli — so characteristic of the West facade at Knossos, recurs at Phaestos, Hagia Triada, and Gournia, and again in the early Palace of Mallia.1

Ramp Passage from N.W. Portico and Lustral Area.

It is clear that the ' Early Keep' must have blocked the direct access Ramp from the North-West Portico to the Central Court. The course of what froSmage was probably an open ramp may be traced, however, stepping up past its N-w- N.W. angle and thus abutting on a passage-way leading directly to the 'Corridor of the Stone Basin' referred to above and so to the adjoining angle of the Central Court.2

The access to this passage-way from below was through the double gateway opening on the ' North-West Portico '. From the inner vestibule, bordering this entrance on its Southern side, there opens West an elongated space that must always have been uncovered, of the same width as the entrance of about two and a half metres, representing the first section of the original ramp and somewhat overlapping the Northern substructures of the ' Keep '.3 The ramp, after running a little over four metres West, turned at right angles South, ascending still till it reached a level answering to the original level of the Central Court. Up to this point we must suppose it to have been open, but the passage now passed under cover, debouching by a doorway into the ' Corridor of the Stone Basin ' above mentioned and thus gaining access to the North-West corner of the Central Court.

1 See P. of M., ii, Pt. I, p. 270. sections of the ramp, the pavement had not

2 See Plan A at end of Vol. ii, and cf. Vol. i, been preserved, though there are some traces pp. 422, 423. of the points where it stepped up. The inter-

3 A fine limestone corner block of the North space that it occupied was to a large extent wall of the passage-way marked with the ' spray ' excavated to a considerable depth in 1913 in sign was brought out in 1928. Of the lower tracing the substructures of the Keep.

' INITIATORY AREA ' AND N. LUSTRAL BASIN

Palatial Function of ' Initiatory Area ' and the N. Lustral Basin.

It has been already pointed out that this passage-way fulfilled a special function in bringing this part of the building — itself, it appears, con- ' Votaries' tainino- more than one shrine — into direct connexion with the entrance

entrance. °

Initia- tory Area

Fig. 2. Revised Plan of 'Initiatory Area' and Lustral Basin.

TheN. Lustral Basin : M. M. Ilia date.

system that bordered the N.W. Portico — itself essentially sacral in character. This entrance system indeed — forming as a whole the ' North- West Bailey ' — seems to have been specially designed to suit the religious require- ments of pilgrims or votaries entering the Palace-Sanctuary from that side. As shown in the revised Plan, Fig. 2,1 it centres in the remarkable early example of a Lustral Basin, free-standing in a separate Court, described as the ' Initiatory Area ', and approached by a kind of Entrance Court, recalling the Fore-hall of the ' Room of the Throne ', which contained a similar sunken basin on a smaller scale. Within the ' Initiatory Area', in a con- temporary stratum that covered its North- West corner, there came to light a series of remains of vessels such as decorative stone ewers and inlaid

1 The original plan (/'. of A/., i, p. 405 with observations made in the course of the seqq., and p. 406, Fig. 291) has been slightly work executed in 1928 and 1929. modified by Mr. Piet de Jong, in accordance

RESTORATION OF N. LUSTRAL BASIN 9

bowls of types now known to belong to the initial phase, a, of M. M. Ill and exactly answering to similar remains found in the basin itself. In the same deposit, which seems to have belonged to some kind of Sacristy or Treasury in that part of the Area, was found the inscribed lid of an alabastron bearing the name and titles of the Hyksos King Khyan.1

It is clear that both the stone vessels found in this deposit and those derived from the sunken basin had a sacral character and certain thick- walled vessels of greyish clay found on or near its floor, as in other -similar basins, had served for the oil or unguents used for the anointing. 1 he evidence of the use of ' holy-water sprinklers ',2 resembling the aspergilla of the Roman Pontifices and still in use in the early Christian Church, makes it probable that some sprinkling of this kind was also included among the lustral rites performed in these sunken basins. For this they were quite adapted, though they could not well have been used for holding water.

In this connexion the recurrence of solid stone basins of various materials, resembling fonts, to be described later on in this Section 3 may be thought to have a special significance.

Restoration of Northern ' Lustral Basin '.

In order permanently to preserve the fine gypsum material of the Restored

riivTi t i t-» • a t i • i i ri plan and

interior ot the Northern Lustral Basin4 1 decided, as a part ot the cam- elevation paign of conservation, to replace its roof and the supporting columns above p^S' its balustrade. This work, begun in 1928 and concluded in 1929, was en- trusted to Mr. Pietde Jong (see Figs. 3 and 4 and Suppl. PI. XXXII). From fallen remains of the painted stucco it has even been possible to restore its general effect on the walls, including the somewhat sombre blue fields above and, below, black panels speckled with white, in imitation of stone- work.5 This sombre colouring was in keeping with the chthonic cult within. This little structure — now a jewel of the site — was bordered by open spaces on its Western and Northern sides, on which it was probably lighted, as indicated in the restored plan, by small upper windows. The researches of 1929 showed that the N.E. corner of the building abutted on a small portico, the substructure of the central column of which was brought out. This faced the Initiatory Area and was entered by the N. doorway of the Entrance Court (see Plan, Fig. 2). A stone bench probably occupied the recess behind.

1 Ibid., i, p. 419 seqq., and Figs. 304 a, b. of the basin and staircase at the time of its

2 See ibid., ii, Pt. II, pp. 792-5. excavation see P. of M., i, p. 407, Fig. 292.

3 See below, pp. 25, 26 and Fig. 13. 6 Compare the M. M. JII a fresco, ibid., i,

4 For Mr. Theodore Fyfe's admirable sketch p. 446, Fig. 321.

IO

NORTHERN LUSTRAL BASIN

In its construction this Northern Lustral Basin most nearly recalls that of the Little Palace,1 though it was over twice the depth of the other and was approached by sixteen descending steps instead of six. Here, too, in this

Isolation of basin from other struc- tures.

Fig. 3. Plan of Northern Lustral Basin.

case on two sides, it was flanked by a narrow passage receiving its light from between the columns of the parapet.

Careful supplementary excavations carried out in 1928 all round the building entirely corroborated the conclusion, embodied in the plan, Fig. 3, that the basin was isolated on all sides. To the North and West it was cut off from any other structure and on the other two sides it was bounded by Corridors. A trench round its outer walls showed that, unlike ordinary

P. of M., i, p. 405 seqq.

Fig. 4. Restored View of Northern Lustral Basin showing Upper Structures : looking West.

By Piet de Jong.

12

PRIMITIVE KEY FOUND IN DOORWAY

basements, they had no connexion with any other walls, the stones being simply built up against cuttings in the virgin soil. This direct contact with Mother Earth and the descent into her bosom beneath the

sombre vault above fits in with the su gores- tion already made that these sunken purifica- tory basins connected themselves with the Goddess in her chthonic character, as Lady of the Underworld and wreathed with its em- blematic snakes. In this earthquake - stricken region it is hardly neces- sary again to recall the special need of divine protection from the infernal Powers.1 A flagcllo terraemotus libera nos may well have formed part of a Minoan litany, long before the days when the prayer was taken over by the Christian Church.2

Fig. 5.

Plan of Later Doorway in Corridor South of Lustral Basin.

Minoan ' key ' found in doorway of later passage.

Fig. 6. Bronze Locking-pin or Primitive Key from Doorway South of Lustral Area.

Dis aliter placitum. It looks, indeed, as if an earthquake shock about the close of M. M. Ill a — of which there seems to be other evidence on the Palace site 3 — played a leading part in the ruin of the Lustral Basin itself. The destruction over this area was indeed so severe that in the succeeding restoration no attempt was made to rebuild this structure. Its basin, choked with charcoal and rubble, was found overlaid by later walls traversing it from South to North. The principal of these, which ran out at a slightly higher level from the massive line of walling that bordered this area on the South side, was shown by the exploration of 1928 to have a doorway consisting of massive gypsum jambs with a threshold of the same material, Fig. 5, above which an interesting find occurred. In the rubble debris near the threshold was a kind of pointed bronze instrument, Fig. 6, in which we may

1 See P. of M., ii, Pt. I, p. 322.

3 Ibid., p. 320 and note 2.

Ibid.

MINOAN LOCK IN SOUTH HOUSE

venture to recognize a primitive Minoan key r of the kind used to lock the cellar door of the South House.2 Mr. Christian Doll's sketches of the door jamb and lock are here reproduced, which curiously anticipate the present discovery (Figs. 7, 8).

An interesting feature is to be observed with regard to the locks here illustrated. As will be seen from the horizontal section, Fig. 7, the locking-pin that secured the wood- en bar of the door could be in- serted on either side of it. The bar itself could only be with- drawn on the inner side by some one in the basement room to which it gave access, but this could not be done unless the locking-pin was first removed from the other side, at the foot of the little staircase leading from the ' Megaron ' and entrance sys- tem of the house. This arrange- ment is very characteristic of the carefully devised control system of the Minoan dwellings large and small. The basement room to which the door led gave beyond into a smaller inner cellar or store-room where a hoard of bronze tools was found, the door of which was barred on the inner side while it was itself, apparently, accessible from an interior room above by means of a trap-door and ladder.

The doorway of the later structure bordering the Lustral basin fitted on South to a wall-line underlying a line of massive blocks that had apparently been taken from the earlier Palace ruins to patch up its main boundary-wall on this North-West section at the time of the general restora- tion of the building towards the close of M. M. III.3 The doorway itself

Minoan lock in South House.

CHRISTIAN C T DOLL AUGUST I9;0

Fig. 7. Horizontal Section showing Primitive Lock : South House. (See, too, Fig. 8, p. 14.)

Later

structures over N. | Lustral Basin.

1 From the position in which this object was found it seems less probable that it was a stilus, as was supposed to have been a pointed bronze instrument from Palaikastro in the Fitzwilliam Museum.

2 Ibid., pp. 382-4 and Figs. 217, 218.

3 To determine the date when it was placed in its present position, I decided, during the work of 1928, to temporarily turn over one cf

the limestone blocks of this construction. It was of somewhat wedge-shaped form with a good face, showing two double-axe signs of Early Palace type (M.M. I— II), and weighed as nearly as possible 2\ tons. In an un- touched element beneath it, among forty sherds, two were intrusive Neolithic, one was a part of a M. M. I a cup, and the rest M. M. Ill, including one characteristic piece

i4 M.M. III« SIGNS ON BLOCKS BY N.W. ENTRANCE

N.W. emrance partially blocked.

Incised signs in M.M. III.

t

2cit-IfoU 055XV4-5X

whole thickness

"**%&

•0*8

< 20 > "

tt

40

I

I

and the connected wall built over the choked-up basin must itself have be- longed to an intermediate date following on the local catastrophe, which, as we have seen, took place when M. M. Ill a was already stratified.1

But, although other structures were erected after the filling in of the basin, they had no re- lation to the original plan, and the special functional character of this area as a place set apart for the purification of those about to enter the building through the adjoining ' North-West Portico ' seems to have been given up.

The interior area of the building was entered from this Por- tico by a double door- way, the inner walling of pIG 8

Socketfyr Locking bar

„, -J05x 02x

035 cCeefi

^^Sorxng for Lot'

u> yoitiii o

ISOMETRICAL SKETCH OF JAMB OF GYPSUM DoOR-

which consisted of some- way, South House, showing Sockets for Bar and Bore- what elongated lime- holes for Locking-pin By Christian Doll. (See, too,

15 g _ details in Jig. 7, p. 13.)

stone blocks incised al- ternately with the ' double-axe ' and the ' spray ' sign in the same some- what superficial manner as those of the neighbouring bastions of the Northern Entrance passage. This style of incision — which contrasts with the bolder character of the signs on the interior, for instance, of the existing West facade of the Palace — wherever its date can be approxi- mately fixed — seems to mark the considerable restoration that was carried out some time after a great catastrophe that took place towards the close of M.M. II; in other words it may be described as M.M. \\\ a. The evidence that has been already adduced shows that the construction of the 'Northern Lustra! Basin' itself and its connected system, including the whole ' Initiatory Area ', owed its laying out to the same epoch, at the expense, it would appear, of some earlier arrangement.

With the filling in of the Lustral Basin and the consequent diversion of

of its earlier phase (a), and several charac- ' As a matter of fact sherds of M. M. Ill ^

teristic M.M. lll/> fragments. OfLateMinoan type including parts of typical low cups were there was no trace. found beneath the threshold.

N.W. ENTRANCE NARROWED IN M. M. Ill b 15

this purificatory station to other purposes the N.W. entrance of the building Partial

.1 r 1 m ^i • blocking

lost its importance as a special route for votaries and pilgrims on their way to of N w the central sanctuary. Accordingly, in the restored arrangement, following ^a^e on the great seismic catastrophe towards the close of M. M. Ill b, one wing in b. of its double portal was blocked with materials taken from the ruins.1 It looks, moreover, as if the width of the initial section of the ramp passage was reduced to the same extent.

The subsequent formation of a lustral area in connexion with the ' Room of the Throne ' would have supplied the means of religious lustration for those entering the North-West corner of the Central Court during the concluding period of the Palace, but the evidence of a similar arrangement for the intermediate period ranging from M. M. Ill a to L. M. II has not been preserved. It must be remembered, however, that the laying out of the Throne Room system, which dates from L. M. II, had entirely obliterated the earlier plan of the area that it occupied.

The existence of the small cistern, dating from M. M. II, on the borders of this area may, however, connect itself with lustral functions, and it is worth noting that inside the adjoining facade on the North side of the Court two alabaster ' fonts ' were found, similar to that illustrated in Fig. 13, below.

The Basement Chambers above the 'Early Keep'.

Unfortunately, the upper structures situated behind the Northern part Area

of the facade of the Central Court, in its earlier and later shape, that lay n.w\

West of the Northern Entrance Passage, have suffered an exceptional f,ect!on °J

o ' . r border of

amount not only of ruin and denudation, but of disturbance through make- Central shift arrangements due to later squatters. These structures were superposed on the much more mighty walls of the ' Early Keep ' 2 that occupied this area, and which must be included among the ' proto-palatial ' elements of the site, the sherds derived from its inner interstices dating; no later in fact than E. M. III. The cells or walled pits of this, which seem to have been largely Basement coated with red-faced plaster, went down somewhat over seven metres (about aDove 25 feet) into the Neolithic, their major axis usually running E. to W., and '?ar1^. though some of the thick cross-walls of this ' proto-palatial ' structure served new to support later lines, the general tendency of the later builders was to neglect tjon< the early lines. Their foundations were sunk deep down, in some cases as much as their full depth, into the earlier shafts and at right angles to their

1 In the course of supplementary researches tain number of sherds most of them belonging

undertaken in 1928 I raised some of the upper to the advanced stage of M. M. ZII& No

blocks belonging to this miscellaneous block- L. M. fragments occurred, ing and found in the interstices below a cer- 2 See Vol. i, p. 136 seqq. and Fig. 100, 101.

16 NORTHERN BORDER OF CENTRAL COURT

general axis — from North to South, that is, instead of from East to West.

In exploring the lower part of the cells parts of these foundation piers had

to be removed and the later wall-lines resupported by means of arches.

Plinth The Northern boundary of the Central Court very nearly corresponds

in^ Court m its Western section with the outer Southern line of the 'Early Keep'.

and steps Unlike the Western border of the Court, it shows no siens of a double

down. - # _ . .

facade, but there are remains of a single boundary line starting from a point answering to the abutment of the earlier Western facade line of the Court, where the ' Corridor of the Stone Basin ' opens, and running originally East as far as the upper angle of the Northern Entrance Passage, (See Plan, Fig. 9.) From the N.W. corner of the paving of the Central Court in its later stage a plinth resembling that of the Western border of the Court runs East for a distance of 4*75 metres. At this point it shows an opening that stood in relation to a flight of three descending steps of gypsum a good deal worn x (see Fig. 9). These, as they were found, seem to have led down to a plaster floor about 43 centimetres below the limestone pavement of the Court. This floor, which is of great importance owing to the relics found on its surface, marks the basement levels on this side as they existed during the period that succeeded the great Restoration of the Palace towards the end of M. M. Ill b and continued to its close. Basement Subsequent investigations brought out the remains of three more steps,

deeper! Y or s'x m a^>2 demonstrating that the flight had been originally designed for a more deep-lying floor-level. The internal wall construction, however, shows that this flight had been first so centred as to extend a little West of the remaining borders of the uppermost steps on that side in the position in which they were found. The original width, so far as could be judged, was about 92 centimetres, answering to the width of an ordinary doorway.

Beyond this point the plinth of the later facade of the Central Court had been much destroyed, but another small flight of descending steps,3 that came to light about two metres East of the first, indicated that there had here been a similar opening of the same width.

1 Dr. Mackenzie notes that the second and 3 These were not like the others of wedge- third step had been slightly cut away at the shaped form, and originally designed as steps, Western extremities in order to get the founda- but mere elongated gypsum blocks, re-used, tion block of the plinth into position. The They were evidently of later construction and plinth itself in its existing shape was an in- only led down to the M. M. Ill b floor-level — tegral part of the restored Palace, dating from ex hypothesi belonging to the restored Palace, the post-seismic M. M. Ill b epoch. On this floor-level the bulk of the remains of

2 In the Plan, p. 19, Fig. 9. the first four the 'Spiral Ceiling ' and 'Miniature Frescoes' steps only are shown. were found (see below, p. 30 seqq.).

LARGE BASEMENT CHAMBER 17

It thus appears that the facade of the Central Court on this side as on the West flanked a series of basement chambers, the floors of which lay well below the level of its pavement (see Plans, Figs. 1 b, 9). In their original shape they lay even more below the level of the Court as it existed in the earlier part of the Middle Minoan Age.

The stratigraphy of this whole area bordering the North- West section of the Central Court was in many respects more complicated and difficult to decipher than that of any other quarter of the Palace site. It was only indeed with the accumulated experiences of the successive Minoan phases represented in the building, acquired after long years of investigations, that the problem could be successfully attacked.

Supplementary explorations undertaken by me in 1928 brought to light m.M. some gypsum slabbing immediately overlying the wall-tops of the Early flyover Keep and belonging, therefore, at latest to the very beginning of the Second walled Middle Minoan Period. This floor lay 1-14 metres below the level of the later paving of the Central Court. An interesting fact in connexion with it, brought out in the course of the same investigations, is that a paving slab of this system runs under the South end of the wall that in the later plan separated the Central basement chambers from that to the East.1 More than this, a further gypsum slab belonging to the same level was found in the S.W. corner of the Eastern basement space and clearly continuous with the other. It thus appears that, according to the original arrangement, Large after the filling in of the deep walled pits of the Keep, this part of the area maent~ formed a single chamber. As a matter of fact the original arrangement of chamber- the descending steps ignored the dividing wall.

The separation into two rooms by a dividing wall was the work of the later succeeding phase of restoration, dating from an advanced stage of M. M. II wau mg and marked here by a second gypsum pavement about 25 cm. above that ^,M- already described as immediately overlaying the wall-tops of the ' Early Keep '. The evidences of this extend to both sides of the wall, and in both cases pottery of the fully developed M. M. II class was found in the inter- vening deposit — part of a characteristic cup of that epoch being embedded in the interior of the wall itself.

An interesting feature in the remains of the gypsum paving on the West

1 West of this wall the floor-level to which filling material of its deep-walled cells. The

this gypsum slabbing belongs could only be remains of this floor underlay the paving of

traced by means of the continuation of its re-used steps at a depth of 56 cm. beneath its

underlying layer of clay and plaster, which was surface. The gypsum pier block belonging

immediately superposed on the top of the to the same later system rested on this early

great walls of the Early Keep and of the floor.

HI. C

18 RE-USED GYPSUM STEPS AND CENTRAL PILLAR

Gypsum steps re- used for pave- ment.

Central pillar of Western basement in M. M. 116.

Excep- tional evidences of strati- fication in ' Room of Knobbed Pithos'.

side of the wall of division threw a retrospective light on the earlier architec- tural arrangement of the Palace in this region. The paving proved to have been largely composed, as their acute triangular section shows, of thin gypsum steps with a very broad tread of 65 cm.1 and a rise, in the state in which they are preserved, of 12 cm. Of their original width there is no certain evidence, but they had apparently belonged to some very gradually descending Corridor on this side, somewhat suggestive of the contemporary ' Stepped Portico ' of the South-West corner of the building. They were of two widths laterally, of 1-50 and 1-16 metres, which if we may assume, as was often the case, an alternating arrangement, points to a total width for the step-way of 2-66 metres. It is possible that they may have belonged to some covered section of an early phase of the ramp passage described above.

The slabbing in the Western of the two spaces stands in relation to a central gypsum block2 (see Fig. 9), which shows, indeed, traces of the ledge on which the edge of the pavement here rested. This block clearly belonged to a central pillar and marks a permanent feature of this basement chamber though, for better security, this supporting pier was in the succeeding age linked to the South wall. We may deduce from this, as shown in the Plan (at the end of Volume II), that there was here an upper chamber with a single column.

A valuable commentary on what may be regarded as the M. M. II b arrangement of this basement region was supplied by a discovery made during the very elaborate supplementary investigations of 1928. The small cell-like space bordering this central basement chamber on the West, and known from the principal object that it contained as the ' Room of the Knobbed Pithos ', to which access was later gained by an opening in the East wall, was found to have had an earlier opening on its South side (see Fig. 9), corresponding with a floor-level of the same date as the M. M. II b pave- ment of the Central Chamber, though differently constructed. The opening itself gave access to a small Magazine with a plaster floor at the same level, the latest sherds under the intact Eastern section of which were of the fully developed M. M. II class. But a contemporary slab fitting on to this found in the opening itself and protruding over the inner cell in an irregular fashion had considerable significance. It clearly belonged to a small patch

1 The breadth of the slabs was 75 cm., but at least 10 cm. of this would have been covered over.

2 This block, as is shown by two dowel- holes on its \V. side, had been taken from the

earlier Palace (M. M. II a). The cutting in its E. side which served to receive the slabbing of the later floor (M. M. 11/') may therefore have been originally made in an earlier connexion.

FINDS IN NORTH-WEST 'INSULA'

19

of polygonal pavement of the familiar ' mosaiko ' class, formed, as in this instance, of iron-stone (d/ivy8a\6\i6os) and typical of a considerable restoration of the building that took place towards the close of M. M. II. The chrono-

PIET DE JOMG

Metres

Fig. 9. Part of North-West ' Insula ', showing Position of Principal Finds.

logical data here corresponded with previous evidence, and this iron-stone slabbing is seen to mark the level where the ' Knobbed pithos', dating from the earliest phase of that Period, had been truncated.

This series of levels brought to light in this small chamber affords one

c 2

20

STEATITE JAR WITH SPIRAL RELIEFS

Stratifica- tion in area N. of 'Pillar Base- ment '.

of the best pieces of stratification to be found on the whole of the Palace site. An ideal sketch of this is given in Fig. 12, p. 24, below.

The latest level here reached of post-seismic M.M. 1116 date ex- tended over this whole area and, from the character of the relics found on

Fig. 10. Black Steatite Jar from Early Floor of Inner Basement.

it, must have remained in use to the very end of the Palace and the closing days of L. M. II. The same continuity is observable in many of the ' West Magazines', where late 'Medallion pithoi' stand side by side with L. M. I and L. M. II jars.

In the space, again, immmediately North of the room with the central pillar — which lies at a somewhat lower level owing to the slope of the ground and seems indeed to have been approached by two descending steps — slabs of gypsum also appear of the same fine alabaster-like quality, such as is characteristic of the earlier palatial stages. Here, too, a careful exploration

DISCOVERY OF 'SAFFRON-GATHERER' FRESCO 21

of the material underlying the slabs 1 produced similar chronological results. In the inter-space, in this case only about 5 cm. wide, between the slabs and the earliest Palace floor-level of rough cement-covered blocks — immediately overlying the wall-tops of the Keep — the latest sherds were M. M. 1 1 and

the deposit underlying a con- nected patch of ' tarazza ' pave- ment repeated the same story. This conclusion as to the date of the paved floor-level of this area has a special interest in relation to two discoveries here made, both apparently be- lonp'incr to this level.

One of these finds was a large black steatite vessel 48*5 cm. in height (Fig. 10) with a spiral decoration in bold relief. This spiral, as will be seen from its representation in Figf. 10, differed in two sections of circumference of the vase. On one side it is a simple coil. On the other side it presents a characteristic inner return, such as is frequent throughout the earlier phases of the Middle Minoan Age. It is seen already on the M. M. \a ivory bead-seal from Knossos, Fig. 11, probably of M. M. I a date.2 It recurs on polychrome pottery of egg-shell type (M. M. II a),3 and is found on jars of the earliest M. M. Ill fabric («).4 The jar itself has a somewhat archaic aspect.

The other discovery, made near the S.E. corner of the room, consisted of the fragments of the fresco panel representing the naked figure of a boy gathering saffron flowers in a rocky landscape, described and illustrated in the first Volume of this work.5 The unique bluish tint of the figure points to an Age when the Egyptian colour convention of deep red for males, and

Fig. 11. a, l>, c, Ivory Bead-seal from Knossos, M.M.I a (f).

Large

black

steatite

vessel

with

spiral

reliefs.

' Saffron- Gatherer' fresco.

1 Supplementary researches of 1928. appears on b.

■ Found N.W. of the Palace site : drawn at 3 E. g. P. of M., ii, Pt. I, PI. IX c, 1-2.

the time but now untraceable. The 'tree' 4 Seager, Pachyanunos, PI. VIII (xi, a).

motive of E. M. Ill and M. M. Is seals 5 P. of M., i, pp. 265, 266, and PL IV.

22 LATER FLOORS OF BASEMENTS

white for women, generally prevalent from the Third Middle Minoan Period onwards, had not yet been adopted. At the same time the colour scheme of the rocks, with their black and scarlet striations, corresponds with that of the great days of Cretan polychromy on vases, which does not extend beyond M. M. II. The latest investigations as to the floor-level on to which it had fallen corroborate the view that this remarkable fresco — doubtless part of a frieze above a high dado — which in its early characteristics stands apart from all known Minoan wall-paintings, is of exceptionally early date. Nor, indeed, does the fact that it was found in apparent connexion with a M. M. II b floor by any means exclude the possibility that it was executed at a somewhat earlier date. For it must always be borne in mind that, while small objects, such as vases resting on a floor, may be generally referred to the latest epoch when it was in use, frescoes on walls often survived many interior changes. Later The stone lamps found in this space (see below, F'ig. 14, a, b, c) must

w?t°h stone ^e taken in connexion with a later floor-level — 50 cm. above the earliest lamps and M. M. II laver — on which some inscribed tablets of Class B also came to Class B. light. In the space immediately West the pavement of this, consisting of clay with a plaster face, was better preserved, and on it lay a mass of much decayed tablets of the same kind. Twenty centimetres above this, again, L. M. was a clay floor, resting on which were numerous vessels belonging to the floors earlier phase of the ' Reoccupation ' period (L. M. Ill a), including 'linked' above. or double pots, a ' cheese strainer ', and stirrup vases, showing symmetrically decorative but degenerate octopuses, which at the time gave their name to the room. It was observable that these late remains did not extend to the contiguous spaces, a phenomenon frequently repeated and which illus- trates the partial and sporadic character of the reoccupation. Base- For the decorative remains of the Palace as restored after the seismic

bordeHn" catastrophe towards the close of M. M. Ill the basement rooms above

Central described, bordering the Central Court and South of the area that contained Court. ' & (

the ' Saffron-Gatherer , were specially productive. Some account has already

been given of the earlier stratification of these, and of the subsequent rearrangement by which a central ' pillar-room ' was formed, the Eastern- most space being separated off by a dividing wall. There is considerable evidence to show that at the time of the great Restoration new plaster floors were laid clown in both these basement rooms the surface of which was about half a metre above the earliest Middle Minoan floor, itself immediately overlying the wall-tops of the Keep.

The mark of this clay and plaster floor is clearly seen on three sides of

STRATIFICATION OF 'ROOM OF KNOBBED PITHOS' 23

the limestone block that was now set above the original gypsum base of the Re- central pillar already described,1 and this was at the same time tied to the J^™" o*f" stout South boundary wall of the basement to give additional security to support- the central pier. This wall, since removed, was somewhat skewly and badly built, as was not unfrequently the case with basement work executed at this epoch.

Other traces of the same late plaster floor about half a metre above the level of the M. M. II b gypsum slabbing are visible on the walls of the Central basement chamber, notably on that to the South.

Apart from the characters of certain objects found on this floor-level Further and marking it as a work of the New Era, crucial evidence as to its relative supplied place was supplied by the small adjoining space containing the ' knobbed by pithos ' of the exceptional stratification of which something has been already Knobbed said 2 (see Fig. 12). We have already seen that the pithos itself rested on the rough limestone slabs of a floor representing the earliest palatial stage and laid immediately on the upper part of the inner walling of the ' Keep '. At that time the space was a mere cell, walled wholly round, and accessible only, like the basement rooms of many early houses, by means of a ladder. At a later date an opening was made at a higher level in the South wall of the cell, and the iron-stone slabs of a ' mosaiko ' pavement of the M. M. II b class were laid down, the pithos itself being cut off at its level about 40 cm. above its base (see the ideal sketch, Fig. 12). Next, with its surface 15 cm. higher, or 55 cm. above the original ' kalderim ', there are well-marked traces on the borders of the cell of a clay and plaster floor 10 cm. thick associated with roughly made cups and small vessels, found in part standing on this level, of typical M. M. Ill a fabric.

The outline of some of these is sketched in Fig. 12. It is noteworthy m.M. that the cup, e. showing a brown band on a buff slip, resembles in shape and IIIrt

r & , pottery:

height — 10-5 cm. — the higher of the 'ink-written' cups of the Linear below Class A,3 from the deposit — clearly of M.M. Ill a date — found on the South- paavg. East Palace border. Its base, too, is pinched in in the same way. With ment- these smaller vessels was part of the rim of a jar presenting white spirals on a blackish or red-brown ground of which the section is given in Fig. 12,/!

This floor with the pottery upon it was in turn earthed under and, at a time when the deposit within the cell was 15 centimetres higher — or about

1 See above, p. 18. edge of the chamber by the subsequent ex-

2 See Vol. i, pp. 234, 235 and Fig. 177. ploration.

The results there shown were incomplete, 3 P. of M., i, p. 588, Fig. 431, a, and cl.

more vessels being brought out along the p. 613, Fig. 450.

24

'ROOM OF KNOBBED PITHOS'

70 cm. above its original floor, — a new opening was made for it, this time in its East wall, and communicating with the Central Basement room. The

GROUND ELEVATION BEFORE EXCAVATION

-«- w

LEVEL OF CENTRAL COURT

IN R£5TOREP PALACE (M-/VHII.6)

Fig. 12. Diagrammatic Section' of East Part of f Room of Knobbed Pithos ', etc.

level of this plaster floor is thus shown to correspond with the M.M. \\\ 5 stage, in all probability its post-seismic stage, contemporary with the great restoration of the Palace.

The threshold of this new opening lay at the same level as that of the

STONE FONTS AND LAMPS FOR RITUAL USE 25

later plaster floor which has left its traces so clearly on its central pier and Exten- which corresponded with a floor-level that had a general extension through- ^j°nM0 out these basements in the restored building. It is of special interest to l}l/'

& l . floor.

note, moreover, that, on the actual threshold of the new opening corre- sponding with this M. M. III^ floor, there lay some fragments of the Miniature frescoes, found elsewhere among the deposits on this floor-level.

It will be seen that the stratigraphic data afforded by the remains in the neighbouring cell fit in well with the relative place of this widely extended floor-level. It is definitely shown to be later in date than the Continu- earlier M. M. Ill phase. On the other hand, its association with some frag- tnis floor mentary inscribed tablets of the Linear Script B indicates a long period of m re~ use. The three rearranged gypsum steps corresponding with a gap in the Palace, plinth on the borders of the later Central Court stood in relation to this floor-level, and close to these the remains of the ' Spiral Ceiling ' had been deposited on it.

These steps led down to the threshold level, continuous with this floor, of an opening through the wall of division — dating, as we have seen, from M.M. II b — between the Central and Eastern basement space, and conform- ing at a higher level with an earlier arrangement.1 The floor as continued Minia- in this Eastern basement was of the same clay and plaster formation, and it Frescoes was, as will be shown, on the surface of this that the bulk of the fragments fgpiral of the Miniature Frescoes and of the Spiral Ceiling- came to light. Ceiling'

r to & fallen

on this

Fonts and Lamps of Ritual Use. level-

Fragments of a L. M. 1 1 amphora of the finest ' Palace Style ' occurred near the S.E. corner of the room and, in a central position, a good deal de- Stcne cayed, but with its base resting on the plaster floor, was a font-like basin of probably alabaster or fine gypsum of the kind already referred to.2 Another similar for.ritual

bJ l J spnnk-

basin of the same material, also a good deal disintegrated, was found in situ ling. in the West section of the adjoining basement.

The best example of these basins, from its good preservation, came to light — obviously removed from its original place — in the neighbouring corridor named after it (Pig. 13). It is finely cut out of a kind of purple gypsum much in vogue at the beginning of the New Era, to which it may be presumed to have belonged. It was about the same size as the other similar basins, the dimensions being 90 centimetres in diameter and 20 cm. in height.

1 The threshold of this opening was finished on the threshold of the earlier M. M. II b off on the East side by a re-used gypsum opening, block, showing a dowel-hole, set up sideways 2 See above, p. 9.

26

FONTS FOR RITUAL SPRINKLING

Numer- ous stone lamps on later

basement floor.

Fig. 13.

Font' of Purple

The close resemblance of the material to that of the fine ' lotus lamp ' found beside the alabaster basin of the central basement chamber suggests that it originally belonged to the same area.

The recurrence of these font-like basins in this basement region border- ing the North-West angle of the Central Court is itself a suggestive circum- stance. Though the old ' Lustral Basin' and 'Initiatory Area' connected

with the North-West Portico had ceased to function, and the entrance itself was narrowed, a certain number of votaries may have still used the ramp passage on this side on their way to the Central Palace Sanctuary and would have entered the Court at a point where these base- ments with the font-like recep- tacles were handy. The same may be said — especially with regard to the Eastern base- ment — in the case of those entering the Court by the North Entrance Passage.

That there was at most a ' dim religious light' in these basement rooms

o o

appears from the occurrence on the floor-level answering to the great Restoration of a series of stone lamps. Of these the finest was the ' lotus' lamp, of purple gypsum, the upper part of which with its quatrefoil flutings, lotus buds, and flowers and foliated rim was found beside the alabaster font of the central crypt (Fig. 14, a 1, a 2). Two others of steatite were found in the adjoining space North (see Plan, Fig. 9). One, Fig. 14, 6, also showed a quatrefoil pedestal. The other, with only a low base, was surrounded with sea-snails in relief, a recurring type.1 Though this bordered on the ascending Entrance Passage, it is probable that for obvious reasons it, like the basement West of it, only received a partial light, perhaps through the doorway, from the Central Court.

From the fine character of several of the stone lamps found in these basement chambers, as well as their recurrence, it is evident that there was something more than a casual and intermittent need for their illumination, such as would have been the case had they been mere store-rooms. It seems

Large Basin or ' Gypsum.

1 The reliefs in this case being a good deal worn, another specimen of this class (from the Royal Tomb at Isopata) is given in Fig. 1 1, c.

STONE LAMPS FROM BASEMENT AREA

27

to be a reasonable explanation that they were used for religious ceremonies connected with the stone ' fonts '. The identification of ' holy-water

Fig. 14. Stone Lamps : a 1, 2, Purple Gypsum ; l>, Black Steatite ; c, Purple Gypsum (see Arch, lix, p. 150, Fig. 127).

sprinklers ' akin to the aspergilla of Roman Pontifices and to those used by the early Christian Church in a series of Minoan representations x has

1 See Vol. ii, Pt. II, p. 792 seqq.

28 MINOAN ASPERGILLA AND RITUAL SPRINKLING

certainly added to the probability of such a conclusion. That some kind of ' baptism ' with the aid of such a stone basin may have entered into the lustral ceremonies ex hypothesi performed in the sunken areas, such as that to the North-West of the Palace, is probable enough and might help to account for their bath-room-like form. It is at the same time equally fitting that such a ritual sprinkling should have been carried out in the crypts — one of which was a kind of ' pillar-room ' — where the stone basins were found.

Fig. 15 a. .Restoration ok Painted Plaster Relief in Knossian Style from Pseira

(see [>. 38 and cf. p. 45, Fig. 27).

§ 6g. Discovery of 'Spiral Ceiling' and 'Miniature Frescoes'

derived from corner sanctuary date and comparative materials j

Embroidered Designs on Holy Robe.

Small corner Sanctuary containing remains of ' Spiral Ceiling ' anal Miniature Frescoes ; These fallen from Upper Chamber ; The ' Spiral Ceiling ' — Egyptian analogies; Parallel from Tomb of Senmut ; Mature L. M. I decorative style; The 'Miniature Trescocs' — -triple group; Chronological materials ; Fragments found on M. M. Ill floor — lata governing discoveries of fresco remains; M. M. Ill date of frescoes ascertained ; Fragments from Thirteenth Magazine ; Characteristic specimen beneath base-blocks of later facade ; Fragment from ' Ivory Deposit' ; True ' Miniature ' style obsolete by L. M. I ; M. M. Ill dale of ' Miniature ' fragments from Tylissos ; Boxers as on ' rhyton — unique bronze vessel; Frescoes from 'Ramp House' at Mycenae; Fragments from ' Threshing-Floor heap' ; Miniature designs from embroideries on female robe; Embossed bands ; Comparison with painted reliefs from Pseira ; Pairs of flutes ; Flutes in sacrificial scene on H. Triada sarcophagus ; Bull's head trophy between pair of Sphinxes ; Embroidered swallows on robe of Melian fresco ; ' Miniature' fragments of Threshing-floor heap, perhaps from robe of Goddess ; Commanding position of Shrine, at angle of Central Court and N. Entrance; Its small dimensions ; Miniature Frescoes set over gypsum dadoes, on line of vision.

The fact that the Western of the two larger basement chambers Small described in the preceding Section was provided, according to its original sanc- structure, with a central pier leads to the conclusion (adopted by Mr. W. G. tuary Newton in the Restored Plan C) x that there had been a room with a central column above it, such as elsewhere seems to have been the arrangement above pillar crypts. We may even assign to it a certain religious destination.

This square, presumably columnar, chamber, opened by its S.W. angle into a smaller oblong space, having, apparently, a cell-like recess at its Northern end. Despite its narrow dimensions, this structure held one of the most conspicuous positions in the whole of the building. It occupied, in fact, the corner space at the point where the Northern Entrance Passage entered the Central Court, and, standing as it did well above the ascending gang- way, it would have received light from that side as well as from the Court.

1 At the end of Vol. ii.

THE SPIRAL CEILING

Contain- ing

' Spiral Ceiling' and

' Minia- ture Fres- coes '.

Ceiling and fres- coes fallen from upper chamber.

The

Spiral Ceiling

From the evidence of the finds made on the basement floors it clearly appears that the principal series of Miniature Frescoes which embody such an unique development of Minoan Art were derived from this little chamber. The bulk of them had fallen on the later floor belonging to the Restored Palace, in the Northern Section of the basement underlying it. Some other fragments occurred in company with the fallen remains of a painted stucco ceiling showing spirals and rosettes on the same floor-level, and in the South-West corner of the same area, beside the steps leading down to it (see Plan, Fig. 9). In the central basement chamber beyond, only a few isolated fragments were found, but two fair-sized pieces lay on the same M. M. Ill b level on the actual threshold of the opening at that time made into the walled cell containing the ' Knobbed pithos '. Apart from the strong presumption that these frescoes had decorated the walls of an upper chamber, their sporadic distribution in more than one basement may be taken as decisive evidence that this was the case.

The Spiral Ceiling.

It would, in the same way, be quite unreasonable to suppose that the painted stucco ceiling of which remains were found on the floor of the Eastern compartment had belonged to the basement itself. It may itself have covered only a small space, but its whole character proclaims that it be- longed, like the associated Miniature Frescoes, to a more important structure connected with the piano nobile of this Palace angle. The spiral reliefs themselves with their central rosettes — repeated in the interstices — belong Egyptian to a class of ceiling decoration that reflects the fashions of contemporary Egypt. A parallel type of a simpler character, with rosettes only in the intervals and presenting spirals on a larger scale, occurred in connexion with the high reliefs derived from the decoration of what was once the great East Hall of the Palace.

The highly decorative pattern of the present example is remarkable for the quatrefoil medallions with large rosettes in their centres attached at intervals to the surface. The splendid effect of this ceiling is well brought out in the Coloured Plate XV 1 : the divergent coils in relief are white, the rosettes red and yellow, outlined in black on a brilliant ' kyanos ' blue ground. It is probable that the pounded glass used for this was imported from Egypt, and the ceiling, like its fellow from the East Hall, and the fine spiral and papyrus pattern from the ' Queen's Megaron ' illustrated

1 From the restored drawing by Mr. Theo- Decoration :, R.I.B.A.Journ., PI. I, and pp. n8, dore Fyfe. See, too, his Fainted Plaster 119.

analogies.

PLATE XV

V2ME.TR6. _ »

<-

2 TEET

10 INCHES

CEILING PATTERN IN PAINTED STUCCO RELIEF

THE MINIATURE FRESCOES 31

below,1 is evidently copied from an Egyptian class of which fine examples have Parallel been found belonging to the early part of the Eighteenth Dynasty.2 It is tomb of to be observed that a near parallel, though of somewhat simpler type, occurs Senmut- in the tomb of Senmut,3 the wall-paintings on which supply us with some of the earliest illustrations of the tributaries from Keftiu.4 As separate features both the spirals and rosettes of similar forms occur on the cups of Vapheio type that they bear as offerings. These fit in again with the widespread designs on friezes of spirals and rosettes much in vogue in the ' Domestic Quarter' and in all probability belonging, as pointed out below,5 to a re- Mature decoration of that quarter of the building about the close of the earlier phase decora- of the First Late Minoan Period. tive style.

This conclusion would make the execution of the painted ceiling distinctly later in date — as may well have been the case — than that of the Miniature Frescoes on the walls, fragments of which had fallen on the floor of the same chamber. It is at any rate clear that at the epoch when it was designed the influence of Egypt under the New Kingdom was beginning to make itself felt. As has been already shown, the floor-level on which these remains came to light presented deposits covering a considerable length of time and ranging, indeed, from the close of M. M. Ill to that of L. M. II.

The Miniature Frescoes and their Relative Date.

As a class, the Miniature Frescoes themselves fall into several groups. The A group found in a different connexion, in which small designs are incorporated ture in the larger subject of a ' Holy Robe', will be described at the end of this Frescoes

. . . triple

Section.0 Another group, of which some specimens occurred in the structure group, referred to above, though it has the greatest claim to represent a historic tradition, is unfortunately the most scantily preserved. It consists of throngs of light-armed warriors for the most part hurling javelins upwards, clearly at the defenders of some stronghold or fenced City, of the architectural features of which we have only a few indications. The subject, however, is of the greatest interest both in its relation to the ' Town Mosaic ' of an earlier age and to the more or less contemporary silver ' rhyton ' from the Fourth Shaft Grave at Mycenae dealt with below in this relation.

In contrast to this the other scenes, of which more is preserved, illustrate festal celebrations in honour of the Minoan Goddess. One of these, indeed, shows her miniature .columnar shrine between grand stands crowded with

1 See below, pp. 371-4, and Fig. 247. 4 See P. of M., ii, Pt. II, p. 737, Fig. 470,

2 E.g. Wilkinson, Ancient Egyptians (1878 &c.

ed.), ii, PL VIII, 31 (facing p. 363). 6 See below, p. 28 r, &c.

G. Jequier, Decoration e'gyptienne, PL XXI. 8 See p. 40 seqq.

3

32

DATING OF THE MINIATURE FRESCOES

Chrono- logical materials.

Frag- ments found on M.M. Ill floor.

Law

govern- ing dis- coveries of fresco

remains.

M.M. Ill date of frescoes ascer- tained.

spectators, again of both sexes, looking out on what in all probability were the sports of the bull-ring in an arena beyond. The theme of the other com- position is ' The Sacred Grove and Dance ' in which standing men and seated women look down from under the olive-trees on gaily dressed dancers within a walled space, in which we may indeed recognize the ' Choros of Ariadne '.

How far, it may be asked, do the find circumstances of these Miniature Frescoes throw a light on the chronological place of this style of wall- painting ?

The evidence as to the position in which they were found is, so far as it goes, consistent and satisfactory. The fragments lay on the clay and plaster floor-level extending throughout these basements, the bulk occurring within their Easternmost compartment, but some scattered pieces on the floor of the central room, and two on the actual threshold of the little cell named from the 'Knobbed pithos'. At this point the stratigraphic data are quite clear. The surface of the floor-level on which the Miniature fragments were there found is 15 centimetres above that on which small vessels and other pottery rested, representing the earlier phase (a) of M.M. Ill (see above, Fig. 12). That the same basement level on which the Miniature Frescoes rested continued in use to the last days of the restored Palace may be gathered from the sparse occurrence on it of parts of clay tablets belonging to the Linear Class B. But, as one now realizes, this does not by any means involve the conclusion that these frescoes belonged to that late epoch — though it was natural, at the time of their first discovery, in view of the almost rococo appearance of the seated ladies, their elaborate toilettes and highly polite gestures, to assume that they represented the most advanced and almost decadent stage of Palace life.

The invariable law, repeatedly illustrated in the course of the excavation of the Knossian Palace, is that, while the smaller clay vessels and other lesser objects found on a floor-level belong to the last period of its use, painted stucco decoration on the walls or fallen from them may go back to a considerably earlier time. They may even belong to an earlier epoch than the floor-level on which they lay.

Thus in the present case, though it is clear that the frescoes were not later than the last L. M. II elements on the floor, and were probably at least as old as the floor itself, it is impossible to arrive at more than a presumption. The plaster floor, as there is good reason to believe, dates from the restora- tion of the Palace late in M. M. Ill b and the wall decoration may have been of the same date. But the possibility always remains that it dates from the time of a M. M. Ill a floor.

M. M. Ill DATE OF MINIATURE FRESCOES

33

Of the relatively early date of this Miniature class— at any rate well within the limits of M. M. Ill — we have, however, some quite conclusive evidence.

The painted stucco fragments found between the ' Kasella ' floors in Frag-- the Thirteenth Magazine have been already cited as giving evidence of early j£

ments om

Fig. 15 b. Painted Stucco Fragment showing Crowds of Spectators from below Kasella Floor of Thirteenth Magazine.

date. They were covered by the floors of more superficial cists at an epoch which in view of our present knowledge must be taken to correspond with the time of the great restoration after the seismic catastrophe towards the end of M. M. Ill, and belong therefore to the pre-seismic stage of that Period. But these, it has been shown, included pieces depicting the crowded heads of male spectators, above a wall, like those of the Miniature Frescoes, slightly larger in scale but executed in the same style, and by means of the same artistic ' shorthand ' and, in fact, inseparable from those of the present group (Fig. 15 b).1

A minute but striking piece of evidence was acquired, moreover, by the very careful analyses carried out in 1925 and 1926 of untouched elements beneath the later West facade of the Central Court.2 The latest sherds there found showed that the ceramic contents, like those found elsewhere

Thir- teenth Maga- zine.

1 Repeated from Vol. i, p. 527, Fig. 384. III. D

See Vol. i, pp. 442, 443.

Charac- teristic specimen beneath base- block of later facade.

34 M.M. Ill DATE OF MINIATURE FRESCOES

under the walls of the restored building, did not come down beyond the later phase (b) of the M. M. Ill Period.1 There was no intrusion, indeed, of anything that could be ascribed even to the beginning of the First Late Minoan Period. But among the sherds thus defined there occurred — under the second base-slab near the S.W. corner of the Court 2 — a small fragment of a miniature stucco painting, showing the stepped outer edge of an entablature, with impressed thread lines and disk such as might well have belonged to the same deposit as the ' Temple Fresco '. A reproduction of it by Monsieur E. Gillieron, fils, is given in Fig. 16.

Fragment A further important confirmation of these

'"ivory stratigraphic results was supplied by some exquisite Fig. 16. Miniature

Deposit', fragments of Miniature Fresco described below3, Fragment from under ° ' Base-slab.

depicting part of a Double-Axe shrine and of a

bull, found in the 'Ivory Deposit' of the Domestic Quarter, the date of which is shown by the ceramic evidence to go well back within the limits of M.M. III.4

It will be seen from the evidences above referred to that cogent proof exists of the Miniature style of fresco painting having been fully developed by the early part at least of M. M. Ill b. On the other hand, the fact that the later M. M. Ill style was still prevalent in the early days of the restored Palace makes it quite possible that those who continued to manufacture pottery of the older class may in the same way have adhered to the practice of Minia- ture painting in decorating the walls. It is on the whole most probable, in view of the position occupied by their fallen remnants, that the ' Temple Fresco' and its companion pieces date from the early days of the restored Palace. True What, however, is certain is that there are no examples at Knossos of

ture style fresco design of the true 'Miniature' class, with its multiplicity of small obsolete fioUres and ' shorthand ' technique, that can be attributed to the Late Minoan

by L.M.I. & 1

Age. The compositions of the concluding Palace periods show larger figures, more sparsely distributed, as we see in the case, for instance, of 'Taureador' and 'Camp-stool Frescoes'. The seated Miniature ladies associated with the groups above described, if in an upright position, would be about io centimetres in height, the restored figure on the balcony (Fig. 35, below) is o-8 cm.: those standing on the piers only 0-55 though we can hardly assume that this was an attempt at perspective. The female

1 P. of M., ii, Pt. II, pp. 802, 803. 3 See below, p. 207 seqq.

2 See test 17 (1925): revised, 1926. 4 See below, pp. 402, 403.

MINIATURE FRAGMENTS FROM TYLISSOS

35

Taureador, on the other hand illustrated below1, is 42 cm. high or at least four times as large as the Miniatures, and the figures seated on the camp- stools would work out to about the same scale.

Such Cretan evidence as we possess of the true 'Miniature' frescoes M. M. Ill outside the Palace at Knossos limits itself to some fragments found in «Minia-

Fig. 17. a, b, Boxers on Miniature Fresco, Tylissos ; <r, on Steatite Rhyton,

Hagia Triada.

a mansion of the not distant Tylissos. Specially important among these, ture' from their close parallelism with a class of small reliefs, were parts of from a row of boxers, advancing left 2 with a tree beyond, and all, it would appear, Tyllssos- in the same characteristic attitude as those from the ' Boxer ' rhyton found Boxers at Hagia Triada,3 the figures on which, indeed, and that on the steatite rhyton. relief and sealing from Knossos 4 make it possible to restore it (Fig. 17, a, 6, c). All these parallel examples may be safely referred to the closing Middle Minoan Period, and the building in which these painted stucco remains occurred also dates from the same epoch.

Together with these were found pieces showing the skirts of a series of

1 See below, Coloured PI. XXI facing p. 216.

2 J. Hatzidakis, TvXto-a-os Mu^i/a; (Ap%. 'E</>., 191 2), PI. XIX. Unfortunately the impress of guiding thread line has been repro- duced in such a way as to look like a rope in front of Fig. 17, a.

3 P. of M., i, p. 690, Fig. 511 ; A. Mosso, Escursioni nel Mediterraneo e gli Scavi di Creta (ed. 2), p. 176, and Fig. 89; H. R.

Hall, Aegean Archaeology, p. 60, PI. XVI. See on the discovery F. Halbherr, Rendiconti del/a P. Accad. dei Li/icei, vol. xiv (1905), p. 365 seqq. The helmet in Fig. 17, c, is restored from this. On the fragment of the steatite vessel from Knossos, however, the boxer's head is bare.

4 P. of M., i, p. 689, Figs. 509, 510.

D 2

FRESCOES FROM TYLISSOS AND RAMP HOUSE

Unique bronze, vessel.

Frescoes from ' Ramp House ',

Mycenae.

Frag- ments from ' Thresh- ing Floor Heap '.

female figures and another on which is preserved the upper border of a crowd

of male spectators (Fig. 18), some of whom raise their arms in the same

way as those on a Knossian fresco where they are looking on a spectacle

below. An architectural fragment found

here (see p. 84, Fig. 47, below) may connect

itself with the class of ' Sieee ' scenes de-

scribed below. Another curious fragment

(Fig. 1 9) l exhibits a large spouted vase,

apparently of bronze, with a handle of unique

form, surmounted by a white object resem-

1 t ..j 1 1 m .1 • , 1 Fig. 18. Painted Stucco Fragment

bhng a tied package, while at the side hangs FROM Tylissos SHOW1NG Part of

a skin vessel — such as are used for native Upper Row of Spectators, as on

cheese— here with a mammiform end. ' Grand Stand ' AT Kn°ssos.

Is it possible that we had here provisions for those engaged in the athletic contest — beer or wine, cheese, and perhaps a barley cake ?

Among frescoes that may be included in the ' Miniature' Class are those

of the Votive tablet from Mycenae2 and of the 'Ramp House'3. The

latter, some of which have to do with the bull-ring, also include parts of the

same group as that above described of women looking out through a window

opening, the posts of which, in close conformity

with the columnar shrine from the Thirteenth

Magazine, show the blades of inserted double axes.

The borders of these panels — where a white band

is succeeded by a yellow band with red bars and

a blue band with black — exactly correspond with

that of the 'Temple Fresco', Here the figures

answer to the Miniature scale : those of the

' Frieze of Warriors ',4 on the other hand from

the Mycenae Megaron — a composition which may

itself be safely placed well within the limits of pIGi 19, Miniature Frag-

L. M. I 4 — are about twice the scale of those seen ment with Vase and Uncer-

,T. . , tain Object : Tylissos.

on our Miniature panels.

The ' Temple Fresco ' and the ' Sacred Grove and Dance ' clearly stand

in a religious relation. This conclusion is confirmed, moreover, by a re-

1 Hitherto unpublished. Drawn for me by Monsieur E. Gillieron, fils. The vessel is coloured red.

" See below, p. 135, Fig. 88.

3 Miss W. Lamb, B. S.A., xxiv, pp. 19 1-4,

and PL VII.

4 Found by Tsountas in 1886 ; Roden- waldt, Ath. Mitth., 1922 and Der Fries des Megarons, 6-v. ; B. S. A., xxv (Miss W. Lamb), p. 164 seqq. : restored, PI, XXVII.

EMBOSSED BAND AND MINIATURE EMBROIDERIES 37

Fig. 20. Embossed Band (§).

markable group of fragments found in a large painted stucco deposit that came to light beneath a later threshing floor, about 15 metres North of the area where the ' Miniature Fres- coes' were found.1 As a series of these frag- ments evidently belonged to the same subject we may infer that they were in this case not mere disjecta membra, but derived from the borders of the region in which they were found and connected probably with the same group of sacral structures as those depicting the Grove and Temple.

Miniature Designs representing Embroideries on Dress.

As being of an exceptional character, the Minia- group of fragments referred to is described in sjgns this place. The Miniatures in this case did [rom,em~

r > broideries

not stand by themselves, but were subsidiary on female to larger figures, and a study of their remains established the conclusion that they repre- sented the embroidered designs on the robes of one or more seated female figures, partly shown in slight relief. All the lines connected with miniature designs were slightly waved or curved, evidently belonging to the flounces or aprons of feminine attire.

Associated with these were sections of Em- slightly embossed bands, 4-1 centimetres in bands. diameter, of very hard plaster with a smooth surface, in one case coloured a brilliant azure blue, in the other a deep red and showing a series of chevrons and spirals dotted in with

1 This great fresco heap — or, rather heaps — lay for the most part under a later threshing floor (aXwviov) above the area of the N.W. Portico. It contained miscellaneous pieces of painted stucco evidently stripped from the walls of the neighbouring Palace region in the course of their demolition to supply building

materials elsewhere. That this took place at a late epoch may be gathered from two circum- stances. Not only did the deposit itself con- tain some L. M. Ill fresco patterns, but it was superposed on wall-stumps of ' Reoccupation ' date.

MINIATURE DESIGNS OF EMBROIDERIES

Com- parison with

painted reliefs from Pseira.

a thick white engobe (Fig. 20). The whole technique was of very fine execution and the chevron and spiral motive itself suggests an interesting- parallel with that of the engaged columns of the ' Atreus ' facade at Mycenae. These, however, as can be seen in the fragment, Fig. 21, and the restored sketch, Fig. 22, formed parts of the borders of ladies' jackets, and the variation in their ground colour is explained by the fact that we see here the spring of the shoulder out- line of two figures. We may con- clude that there

Fig. 22. Partial Restoration of the Figures to which the fragment, flg. 21, belongs. (Compare the Painted Stucco Relief from Knossos, Fig. 27, at the end of the Section.)

were here two female personages seated side by side after the manner of the ' Ladies in Blue '.

An even closer comparison both in style and date is suggested by the remains of two female figures found by Seager in one of the principal houses in the Island of Pseira,1 where the rebuild- ing was carried out towards the close of M. M. III.2 These were also partially

executed in relief and the chevron pattern also forms in one case part of the

sleeve decoration. A smaller fragment of a similar relief (Fig. 27, below) 3

was found at Knossos in the same heap as the above.

The miniature designs on the Knossian fragments under discussion

seem, as already noted from the waving lines that contain them, to belong

Fig. 21.

Fresco Fragments of Two Female Figures.

1 In Mr. Seager's original publication (JEx- shows two seated figures. The fragments had

cavations on the Island of Pseira, Crete, 1910, fallen from an upper floor {op. at., p. 15). PI. V and pp. 32-4) the fragments were re- 2 Seager, op. at., p. 10. He also extends

stored as belonging to a single figure. The the process of rebuilding into L. M. I. revised restoration of the fresco by M. 3 As restored by M. Gillie'ron, fils.

(Jillieron, fils, in the Museum at Candia,

MINOAN FLUTES

39

to the flounces of one or more of such seated figures. The narrow almost Pairs of imperceptibly undulating band, Fig. 23, exhibits on its blue ground a row "utes-

Fig. 23. Embroidered Band with Flutes.

of three pairs of flutes attached to one another by some kind of strings.

What seem to be parts of a bone flute were found at Mycenae,1 but the Flutes in

best evidence of its ^iai"

Minoan use is sup- scene on

i- j i i tt • H. Tnada

plied by the Hagia sarco-

Triada sarcopha- phagus- gus. A youthful ministrant with long locks is there seen blowing pipes

double

of

more elongated form above a sacri- ficed bull, laid on a table with his limbs closely bound together in the Egyptian manner, while his life-blood pours into a vessel placed below,2 Fig. 24. The sacrifice in this case would have had to do with funereal rites : on the other hand in the Iliad we find avXoi played, together with lyres, at

1 Schliemann, Mycenae, pp. 78, 79, and 2 R.Paribeni, // Sarcofago dipinto di Haghia

Figs. 128, 129, 130 a. Triada (Jfon. Ant., xix, 1908), PI. II, and

Fig. 24. Sacrificial Ministrant playing Flute : H. Triada Sarcophagus.

40

MINIATURE DESIGNS OF EMBROIDERIES

Bull's

head

trophy

between

pair of

sphinxes.

Em- broidered swallows on Melian fresco.

marriage dances.1 How early the use of the double-pipes was known in the Aegean World may be seen from the marble image of a flute-player found in a grave of the Island of Keros 2 and going back to an epoch corresponding with the last Early Minoan Period.

Another inset, in a slightly arched space of the larger pattern, consists of a kind of bull's head trophy apparently between a pair of sphinxes (of which only one is preserved), antithetically arranged. The bull's or ox's head figure is completed by another somewhat larger example (Fig. 25, a) already illustrated as supporting elephants' tusks.3 A figure of a seated griffin (Fig. 25, e) may have formed part of a similar opposed pair on the same zone. Fig. 25, f, shows part of one or other of these sacral monsters with its front facing and expanded wings. There are other enigmatic fragments.

As to the meaning here of the curvilinear arrangement of the larger part of the decorative framework in which these miniature insets are in- cluded there can, as already said, be little question. The whole forms part of the front of the robe of a female personage. That the curving bands are in fact quite appropriate in this connexion may be seen, for instance, from the remains of figures belonging to the ' Procession Fresco' at Knossos, and those of Thebes and elsewhere. What is singular in the present case is the introduction of these miniature designs, which have nothing to do with textile art and must certainly be regarded as elaborate pieces of embroidery. A close parallel to this, however, is supplied by the front of the skirt of a female figure, part of which is preserved on some painted stucco fragments found at Phylakopi in Melos (Fig. 26) in the same room that contained the ' Flying Fish ' fresco,4 and, like it, a work of the Knossian School going back to the earlier phase of M. M. III.5

A richly dressed lady seated on a rock is here seen engaged in fishing with a net of a kind still used in those waters, and the beautifully outlined arms of a companion, who bends forward,6 may be thought to be pulling one up.

Here, on the part of the robe corresponding with the lap of the first lady, is an embroidered design of two conventionally rendered swallows set

pp. 51-4, where the subject is fully discussed. 3 Cf. P. of M., ii, Pt. II, p. 742, Fig. 475.

Cf. Mosso, Escursioni tie/ Mediterranco e gli * See R. C. Bosanquet, Phylakopi, pp. 73- Scavi di Crela,pp. 260, 261, and Fig. 146.

1 II. xviii. 491 seqq.

2 U. Koehler, Ath. Mitth., ix (1884), PI. VI, and pp. 156-8. Cf. Perrot et Chipiez, Hist, de I' Art, vi [Grece primitive), p. 760, Fig. 357.

5 and Fig. 61 (wrongly described as a man).

6 See P. of M., i, p. 544.

0 Phylakopi, p. 74, Fig. 62, and cf. P. of J/., i, p. 544, Fig. 396.

/

Fig. 25. Miniature Subjects from Fresco representing Embroidery on Robes (f).

42 PHYLAKOPI FRESCO WITH EMBROIDERED SWALLOWS

back to back. A restored drawing of the whole design is given in Fig. 2(5. l Though the birds — which are of miniature scale — are on a somewhat larger scale, it will be seen that their symmetrical, antithetic position answers to that assigned to the sphinxes and griffins on the present fragments. It will be observed, moreover, that they show the ' notched plume ' ornament on their wings under its earlier, well-formed asoect.- This feature — seen on the bone arrow-plumes from the Temple Repositories at Knossos, as already noted — has specially religious associations, and not only appears on the wings of the sacred monsters, but forms a recognized ornament on the skirts of the Minoan Goddess and her votaries.3 Minia- On the whole it does not seem unreasonable to suppose that the robe

coratioii to w^ich the present group of painted stucco fragments belonged, with its richly perhaps embroidered bands, on which the miniature sphinxes and griffins repeat of God- themselves in confronted groups, actually belongs to a figure of the Goddess herself, perhaps in company, as she often appears, with a companion or double. This finds an analogy with the lower borders of the skirt of the facing figure recognized as that of the Goddess in the Procession Fresco, where, though in that case the designs are simply decorative, they are exceptionally rich in ornament. It may be fittingly described as a ' holy robe '. The style of this large-scale work, the details of which link it on to the miniatures, fully corroborates the conclusions as to their relatively early date. Close comparisons have been established above between it and the group of paintings executed by a Minoan artist at the command of some Melian Prince to adorn the walls of the pillar sanctuary of his residence. But these Phylakopi frescoes in some respects present the finest characteristics of the earlier phase of M. M. Ill and it is impossible to bring down the date of this Knossian work later than at most the concluding phase of that Period.

The Small Sanctuary Structure containing the 'Miniature ' Frescoes.

Com- Apart from what seems to have been a little nook at its Northern end,

corner"8 tne small Sanctuary Chamber, to which the ' Spiral Ceiling ' and the purely position • Miniature' frescoes belonged, occupied one of the most important vantage

of shrine. .. ,.

points in the whole building. Its smaller, Southern wall faced the Central Court while that on its Eastern side looked down on the upper section of the Northern Entrance Passage and thus on to the paved ramp that gave the most public access to the interior of the Palace.

1 Executed for me by Monsieur E. Gillieron, blade of King Aahmes (1587-1562 B.C.) this fils, who first recognized this feature. Minoan feature already appears in a secondary

2 See on this P. of M., i, p. 547 seqq. and shape {ibid., p. 551, Fig. 402). Figs. 399, 400, 401. On the griffin of the axe- 3 Ibid., pp. 549, 550.

Fig. 26. Figure of a Lady drawing up Net showing Embroidered Swallows on Skirt. (From Phylakopi : partly restored.)

44 MINIATURE FRESCOES FROM SMALL SANCTUARY

Though small di- mensions.

Frescoes set over gypsum dadoes on line of vision.

The structure itself was, as already remarked, of quite narrow dimen- sions, about 4-70 metres (c. 15! feet) from N. to S. and 2-70 m. from E. to W. These general proportions correspond very well with those of the little Shrine that forms the centre of one of the Miniature scenes themselves, or of the analogous structure that occupied a central position on the West facade of the Court, which was about 5 metres (i6| feet) in breadth. In this case the facade would have been on the East side, but the interior arrange- ment must have been different if we are to interpret the nook at the North end as a cella.

The Miniature Frescoes themselves were specially adapted for a little sanctuary chamber of this kind. Some small but very beautiful fragments of such, moreover, found in association with the 'Ivory Deposit' of the Domestic Quarter on the East, were probably derived from an exiguous shrine of the same kind.1 Such frescoes, owing to the minutiae of their designs, were obviously intended to be placed close to the eye, and it is there- fore difficult to suppose that they had been set above the upper borders of the high gypsum dado slabs such as were usual in the Third Middle Minoan Period, and the earlier phase at least of L. M. I.

The upper part of these slabs, according to the regular arrangement — so well illustrated in the ' Domestic Quarter ' — corresponded with the lower line of horizontal beams forming a continuation of the lintel of the door- ways. Like the doorways, the gypsum dado slabs were almost exactly two metres in height, and fresco bands, as was shown for instance by fragments found in position in the Bath Room by the ' Queen's Megaron', covered the wooden band, and themselves rested on the upper borders of the dado slabs. In the ' Caravanserai' the ' Partridge Frieze' had been set above this lintel line,2 and in this case, though the subject was of the natural size, it has the appearance of being somewhat ' skied'. On the other hand, in the ' House of the Frescoes ' 3 the friezes of painted plaster with their rich details show traces of a horizontal beam immediately above them, and they were probably set below the lintel line, starting therefore, if we allow for a mean height of about 70 or 80 cm.,4 from a line a metre above the floor level.5

1 See below pp. 208, 209, Figs. 142, 143.

2 P. of M., ii, Pt. I, p. 109 seqq., and see p. 108, Fig. 49. The lintel line in this case ran i-8o metres above the floor.

3 Ibid., Pt. II, p. 431 seqq.

4 The preserved height or vertical width of the fresco representing the Sacred Grove and Dance is 49 cm. : if we add to this some

additional width for the lower part of the scene, as well as the lower border, the original width would probably have been about 70 cm. The maximum height of the ' Temple Fresco ' as preserved was only 42 cm., but the whole of the front scene with the performance is wanting.

6 Ibid., p. 460. In the basement room of

POSITION OF MINIATURE FRESCOES ON WALLS 45

In a small structure, such as is here presupposed, it is better to take the suggestion supplied by this private house rather than the analogies of palatial corridors and halls, and we may well conclude that Miniature fresco bands, which were probably of about the same width as the panels of the ' House of the Frescoes ', started from the same height of 70 or 80 cm., and would in that case have been admirably set on the line of vision.

this house with the double-axe vases (ibid., beam) was set only 61 cm. above the pave- p. 435 seqq.) a decorative band (superposed ment. on a painted imitation of a narrow horizontal

Fig. 27. Restoration of Painted Relief from Fresco Heap of N. Threshing-Floor Area, Knossos (c. |).

§ jo. The Miniature Frescoes, i, The Temple and Grand Stands.

The ' Temple Fresco' : its technique; The ' Grand Stand' and Spectators; Artistic shorthand in delineation of figures ; Great numbers; The men; The women, more carefully represented — 'Court Ladies; Freedom in Art, without decadence; Relation of figures to earlier 'Ladies in Blue' — close dependence; The Seated Ladies — Groups A— E ; Groups compared with those of the rococo Age, as depicted in the boxes of theatres, &c. ; Separate groups of women in front scats mark of Matriarchal stage of Society ; Dramatic action; Prominence of gesture language — that of Naples compared; Ges- tures in scenes on signet-rings — primitive elements ; Women segregated in front seats but mixing freely with men below ; Isolated fragments zvith parallel groups; Window scenes — Cypriote, Assyrian, and Biblical parallels ; Con- trast between Knossian Ladies and Oriental ' Hierodulcs' ; Presumption that scenes of the Bull-Ring were depicted in lower part of panel ; The Central Columnar Shrine; 'Superposed Pillars' of Grand Stand — parallel examples ; Upivard tapering posts and their analogies ; Theatral signifi- cance of single pillars in agonistic scenes.

The Of the two miniature fresco subjects brought to light in the basement

Fresco*'.6 area above mentioned, that called from its central feature the ' Temple Fresco' is represented by the greater number of fragments and may have filled more than one panel. Although only a certain proportion of the frag- ments could be actually united, the main lines of the composition to which they belonged were fairly clear, while architectural symmetry has supplied a further guide. It has, therefore, been possible to propose a general scheme of restoration, admirably worked out for me by Messieurs Gillieron, father and son, of which the central section is reproduced in the Coloured Plate XVI.1 A restored section of the continuation of the scene to the left is shown in Fig. 28. In contradistinction to a corresponding" space to the right of PI. XVI, if rightly placed the female figures, thirteen in number, are here seen in a standing position. They seem, too, to be placed on a wall-top rather than a plinth.2 It is noteworthy that there were no figures in the 'pit' below this wing. Was it possibly the side of an enclosure ?

1 The shrine itself has been also repro- interpreted as steps, but rather, as in the case

duced above in Vol. ii, Pt. II, p. 594 seqq. of the adjoining piers, as courses of masonry,

and Fig. 371. See on this, p. 63, note 1.

2 The horizontal lines on this are not to be

PLATE XVI

llllllltlllllllllllllllllllllllllllllllltlllllll

-WM&M^^immifo* war

nd

f*a

PANEL OF 'MINIATURE FRESCO' RESTORED, SHOWING CENTRAL PILLAR SHRINE WITH GRAND STANDS ON EITHER SIDE

AND CROWDS LOOKING ON AT SPORTS

THE TEMPLE AND GRAND STANDS

47

The artist, or rather artists, who painted the original panels were them- selves guided in their general arrangement by incised lines set out in the plaster. A remarkable feature of these lines is that they were produced by means of fine strings or twisted threads strung in parallel lines, the twisted

$mmmmi

Its tech- nique.

Fig. 28. Restored View of Grand Stand and Spectators to left of Shrine.

texture of which is clearly impressed on the plaster surface. The border, the outlines of the shrine, and the divisions of the theatre are clearly marked in this way. The broad colouring of the shrine, with its alternating areas of red, black, blue, and yellow, was then rapidly filled in, and also the broad washes of red background to indicate the crowds of men. The details were finally added, evidently with a fine brush.1

The whole composition clearly centres round the little Temple, which, The though here seen in fuller detail, belongs to the typical Minoan and Mycenaean §t™d > class. On either side are the terraces and tiers with curiously constructed and Spec- supporting pillars of a kind of Grand Stand or Theatre, crowded with spectators of both sexes.

1 These technical details are from a note kindly supplied by Mr. Noel Heaton.

48 GREAT CROWDS OF SPECTATORS

In front again of the central shrine and adjoining wings of the Grand

Stand is a rectangular Court enclosed by low walls that divide it from

similar enclosures on either side. These spaces, like the tiers behind, are

Artistic entirely filled with a dense crowd of men and women. Very remarkable is

hand in tne artistic shorthand here brought into play for the rapid delineation of

delinea- these multitudinous figures. The original ivory-white background and the

lion of p , .

figures. broad zones of Venetian red washed on it supplied the conventional colour

for the two sexes, alternating in groups ; the individual details being then summarily sketched in. It has already been computed that, in the parts of the frieze that it has been possible to restore above and in front of the central shrine and in the first section only of the stands on either side, the number of persons amounted to about six hundred. But parts evidently belong to further panels, perhaps centring round other architectural features, and we can have the actual evidence of only a fraction of the concourse of people gathered from all parts of ' broad Knossos ' to look on at the great religious spectacles. In the ' Dance ' fresco, when complete, there would have been some 1 ,400 figures. What an earnest this of the Homeric tradition of the great 'City' ! ' What a still living witness to the 'countless' population of Minoan Crete, here chiefly concentrated ! Great From the fact that the proportions of these vary in the groups in

numbers. different parts of the scenes and contain variations in style, we may even infer that, owing to the rapid execution necessary in the fresco process and the complexity of detail, more than one artist had collaborated on a single panel. The men. The men, of whom only the busts are shown with white collars round

their necks, have curly hair, the locks of which fall down from their temples in front of the ears — like those of the women, and at times, like the latter, they wear a kind of band or diadem.2 The male heads of the upper row, however — as is shown by the regularly smaller scale on which they are drawn — are probably, like the smaller figures on the companion fresco, intended to represent those of young boys, and these display the peculiar feature of top-knots with a curl in both directions. They are pointing excitedly at some performance which is evidently being enacted in the arena below.

What was the character of the spectacle that thus thrilled the lookers- on ? Unfortunately the whole of the lower field of the design on which this was certainly set forth, like the ritual dance on the companion fresco, is in

1 Od. xix. 173-4, 178. Cf. P. of M., ii, artists have at times inserted the outlines Ft. II, pp. 564, 566. of female heads and coiffure on the red

2 It would appear, however, that in the patches. rapid execution of this work the artist or

PLATE XVII

C 5-9

MINIATURE FRESCOES OF SEATED LADIES ON GRAND STAND

{T^estored Drawing by E. Giilirron, fits)

THE COURT LADIES 49

this case entirely to seek. But a welcome parallel from Mycenae and a portion of a fresco in the same Miniature style found in the Domestic Quarter of the Palace itself supply, as will be seen, a probable clue.

It is clear that, though the male spectators were the most numerous, the The artist's attention was really concentrated on the female figures. The men more6"' are treated in the most summary way, only the head and neck with the carefully

rep re-

surrounding collar being rendered, while their eyes are indicated by mere sented. white dots. In the case of the women, on the other hand, their com- plete figures are reproduced, whether seated or standing, their eyes more- over are duly outlined, and full details are given of their brightly coloured robes.

The fine drawing of some of these ladies on a white ground inevitably The recalls the white Athenian lekythoi x executed some eleven centuries later, ladies. But what a contrast here in style, in movement, in the character of the figures ! We are very far away from the restrained pose of Classical Greece. At a glance we recognize Court ladies in elaborate toilet. They are fresh from the coiffeur's hand with hair frisd and curled about the head and shoulders ; it is confined by a band over the forehead and falls down the back in long separate tresses, twisted round with strings of beads and jewels.2 In some cases the locks above the forehead curve down in a curious way above the shoulder.3 The sleeves are puffed, and the constricted girdles and flounced skirts equally recall quite modern fashions. A narrow band appears across the chest, which suggests a diaphanous chemise, but the nipples of the breasts indicated beneath these — in one case the pendent breasts them- selves— give a decollete effect. The dresses are gaily coloured with bands of blue, red, and yellow, showing white stripes and at times black striations indi- cative of reticulated and scaled designs like those of the Processional Fresco. (See Coloured Plate XVII.)

A curiously artificial atmosphere of social life pervades these highly Freedom polite groups of Court ladies with their puffed sleeves, their wasp waists, and without elaborate hairdressing. This impression, which smacks rather of Versailles <]eca~ than Florence, made it natural, when they were first discovered, to bring down these productions to the latest phase of the Knossian Palace art, and

1 See my original appreciation of these art would have rewarded the Minoan artist,

figures, Knossos, Report, 1900, p. 47 {B. S. A., 2 The meaning of the white bands crossing

vi). By a fitting chance one of the first to see the large tresses of hair is supplied by the

them at the time of the excavation was Dr. ' Ladies in Blue ' fresco, of somewhat earlier

Wilhelm Klein who had done so much to illus- date, executed on a larger scale (see P. of M.,

trate the white-figured Athenian lekythoi. His i, p. 545, Fig. 39 7).

surprise at such a revelation of pre-Hellenic 3 E. g. in Group B, 3, Fig. 30. III. E

5o DEPENDENCE OF GROUPS ON 'LADIES IN BLUE'

Relation of figures to earlier ' Ladies in Blue'.

Close depen- dence.

to an epoch on the brink of decadence. But conclusive evidence, to which attention has been already drawn, is now at hand to demonstrate that the Miniature style in wall-painting was itself fully evolved well before the close of the Third Middle Minoan Period. At most it can be said that the ' Temple Fresco' before us marks the beginning of the New Era.

These frescoes do indeed belong to a stage already removed from that represented by the class of wall-painting that characterizes the earlier M. M. Ill phase and of which the ' Ladies in Blue' are the best record.1

Still, in certain respects, this earlier group, though very different in scale, supplies a real anticipation of that before us. The figures in the former case are of life size, but yet in certain features of hair-dressing and costume present distinct resemblances to the smaller series, and they are also arranged in groups with their skirted legs drawn in sideways, like so many of the miniatures. They show, however, a more stable pose and their outlines are purer and drawn with greater artistic care. Their measured movements contrast with the vivacious gestures of those beside the little Temple. One of the seated figures is seen fingering a necklace, a sign that their conversation — as is said to be commonly the case with harem ladies at the present day— may, even under the freer conditions of Minoan women's life, have largely concerned itself with jewellery and fashions. The subject of a closely related and contemporary painted relief is a man's hand attaching a gold necklace to a lady's neck.2

We must admit, in any case, the close dependence of the Miniature groups before us on these larger works of the immediately preceding epoch, the chief theme of which seems to have been toilet scenes and intimate con- versation on subjects of female interest. This dependence, indeed, may help to explain the extreme detachment that these ladies, who occupy the front seats of the Grand Stand, show from the performances of which they were clearly supposed to be spectators. Those posted above on the piers of the Grand Stand and others who appear in a standing position have,

1 P. of 31., i, p. 544 seqq. and Figs. 397, 398. Unfortunately the very imperfect remains of this group have further suffered from the damage done to the Candia Museum in 1926. Amongst other features, the most beautiful detail, the hand fingering the necklace (ibid., p. 546, Fig. 398), has been destroyed. For analogous fragments of figures of the same style found in the Earlier Corridor from the West Porch see Vol. ii, Pt. II, pp. 680-2

and Figs. 430, 431.

2 Ibid.) i, p. 526, Fig. 383. Except for two finger-tips this painted relief was entirely- pulverized by the Earthquake. These debris have been reset by Monsieur E. Gillieron, fils, in a coloured cast already made by him at my request. It does not seem necessary to be- lieve, as suggested by me, loc. a'/., that the robing scene had to do with any special sacral function.

THE GROUPS OF SEATED LADIES

5*

indeed, the air of looking on. But the seated figures seem to be only con- cerned with themselves and their own coterie, the subjects of conversation at times provoking dramatic personal emotions.

Of the two groups, A and B, shown beside the little Temple and between The it and the first pier on either side (see Coloured Plate XVII) that to the left j^dies: (A) can be recovered in its main composition from the remains of the upper Group. A.

A 12 3 4 5

Fig. 29. The Seated Ladies : Group A. (See Coloured Plate XVII.)

part of four figures, all on one piece. The first of these (No. 2) entails the original existence of another on the left, which, according to the space at our disposal, would have completed the group.

The reconstitution of the first figure is here based on the corresponding pair of Group B (Fig. 30, 1, 2). It suggests the intimate converse of two confidantes. Nos. 3 and 5 to the right of this show a certain conformity with 3 and 4 of Group B, the raised hand of the first lady being seemingly indi- cative of surprise at what she is hearing.

Group B may be practically regarded as complete except as regards the Group B. heads of Nos. 2 and 3. Here the first pair are engaged in a very close tete- a-tete talk, the right arm of the second lady being laid on the other's side as if to arrest her attention. A special feature, moreover, in the case of the second figure may have a definite intention. On the other female figures or fragments of such in this series, the bosoms regularly show two mere dots, indicating the nipples. In contrast to this, pendent breasts are here clearly outlined — a matronly touch.

e 2

52

THE SEATED LADIES

May we venture to suppose that we have here a mother giving social advice to a debutante daughter ?

A unique feature of B 3 is the termination below of two long fore-locks in a piece of netting. A veil in such a position seems inexplicable.

The lively nature of the conversation between No. 3 — the lady to

B 1 2 3 4 5

Fig. 30. The Seated Ladies : Group B. (See Coloured Plate XVII.)

whose coiffure this net belongs — and her neighbour on the right at once strikes the eye. The latter points her statement by thrusting forward her right arm so as almost to lay her palm on the other's lap, while her confidante raises hers in amazement. ' You don't say so ! ' — the sense of the words can be supplied, though we may never decipher the language.

Meanwhile, the young woman on the extreme right of the group profits by the engrossment of her companions in their own affairs to beckon to some friend beyond.

A larger number of these ladies occupied the wider space between the pier of the Grand Stand, that rises to the right of the little Temple, and similar piers beyond. This space was of about twice the width of the interval between the first piers and the borders of the Shrine, and accommo- dated therefore about twice the number of persons. On the basis of the

THE SEATED LADIES 53

fragments a-d, Monsieur Gillieron has completed Group C of nine persons on Group C. the front border of a space of this width — (22 centimetres) — which fills the

Ua^A.V^^\aCk7sv'w^m /vA>

567 89

Fig. 31. Seated Ladies : Group C (Nine Figures). For 4-9, see Coloured Plate XVII.

broad interval to the right of the first pier on that side in the restored fresco.1 It is reproduced in two slightly overlapping sections in Fig. 31.

1 As seen in the Museum at Candia.

54 SEATED LADIES: DRAMATIC ACTION

Here again the subjects fall into pairs. Of No. i there is only a sug- gestion— the attitude being modelled on Group B, i and 2 — while merely the head and shoulders exist of No. 2. Of the second group there is only a small part of No. 4. In this case, too, the last figure on the right turns away from the preceding and looks out beyond the pier. The profile rendering of the face in all cases makes it possible, however, that she is intended to be looking straight out at the sacred sports beyond.

The composition itself, as will be seen, is very simple in all cases. The groups are largely discontinuous, the conversation being broken up into pairs, very much as an English dinner party. There, too, when the numbers are uneven the odd one is often left in the lurch.

In Group A, indeed, in Nos. 3-5 we have clear evidence of a conversa- tion in which three partake, the middle figure rather as a more or less de- Group D. tached listener. In another shape, in which two are actively concerned with the statements of a third party, this triple arrangement seems to have been twice repeated on the interesting but unplaced fragment, Fig. 32 (Group D).

The upper part of this piece of fresco has been a good deal effaced, but essential traces of three heads, a raised arm, and shoulders can be made out, sufficient to supply the key to a well-founded restoration. At my request Monsieur E. Gillieron, fils,1 has made, in Fig. 32, D, a, a careful tracing of the design as it exists in its present state and added the completions, as indi- cated, in Fig. 32, D, b. Dramatic It looks as if what we see in this case is a section of one of the wider

action. groups consisting of about nine persons. Nos. 1 and 2 may thus be re- garded as completing another trio. The first of two listeners bends forward with her right forearm half raised, as if intent on what she hears. The body of the second is thrown slightly back with her hand raised in front of her face — she is quite shocked at the scandalous tale! The second trio suggests an even more personal and dramatic interpretation. The first lady, with an emphatic down-thrust of her visible hand on the side of her thigh, bends forward so that her nose almost touches that of the second figure, who seems to raise her right forearm as if in deprecation. But the forward pose of this apparently sharp-tongued neighbour may best be taken to show that

1 The late Monsieur E. Gillieron, pere, had Apart from these, the new restoration of

executed similar copies for me some years Monsieur E. Gillie'ron, fils, made independently

back, but a minute study of the fragment has for me, corresponds almost exactly with the

since elicited some further details including earlier drawing, a proof of its essential accuracy, the evidence of the raised hand of No. 2.

SEATED LADIES: DRAMATIC ACTION

55

D,a

2 3 4 5

Fig. 32. Seated Ladies : Group D.

her winged words are aimed, across the second of the group, at the young woman on the extreme right.1

1 The greater part of this figure is wanting, is preserved by the nipple of her right breast, but a section of the forefinger of her left hand showing that her missing arm was turned this

56

THE SEATED LADIES— ROCOCO COMPARISONS

Group E.

Rococo compari- sons.

Fig. 33. Female Heads.

Another isolated fragment is inserted here in Fig. o3 as showing excep- tionally fine delineation of a female head. Here the scale is somewhat smaller.

Group E, which bears the mark of being by another artist, presents somewhat higher figures and may be reasonably supposed to belong to a

continuation of similar scenes on a separate panel. If the arm of the lady on the right is stretched down in the manner indicated she must be taken to be the terminal figure of the group hailing some friend in the 'pit' below. Similar detached action on the part of the outermost figure has been already noted in the case of Group B. Were these ladies perhaps pointing or beckoning to favourite champions in the arena beyond ? Such gesticulations would have been quite in the spirit of the Spanish Corrida.

These scenes of feminine confidences, of tittle-tattle and society scandals, take us far away from the productions of Classical Art in any Age. Such lively genre and the rococo atmosphere bring us nearer indeed to quite modern times. We recall such groups of fashionable spectators — in these cases embracing both sexes — as that depicted in Tiepolo's -Fresco of the reception of Henri III at the Villa Contarini,1 absorbed in their own social interests, gossiping and flirting under cover of fans, on the balconies above. Analogous groups were executed by Pietro Lunghi in the Palazzo Grassi.2 So, too, the bevies of gay Minoan ladies seated in animated converse between the piers and pillars of the Grand Stand seem to reincarnate themselves in Guardi's modish figures, with high perukes — feathered, bejewelled, smirking, and ogling their beaux, equally bewigged and powdered — who fill the pale hi of the Teatro San Benedetto,3 little concerned with the ball below. Similar scenes in the loges of the Theatre Francais are depicted in a print of Moreau le Jeune,4 where the company has the air of general detachment from the crowning of the bust of Voltaire on the stage before them, which they had ostensibly come to celebrate.

These Minoan ladies, indeed, seated in the interspaces of the Grand

way and pointing in the direction of No. 3. This may be taken to show that her head was turned in the same direction.

1 Molmenti, Tiepolo (191 1): Plates 196, 197. For the selection of these and other later comparisons I am much indebted to Mr.

C. F. Bell, Keeper of the Fine Art Depart- ment in the Ashmolean Museum.

2 Aldo Rava, P. Longhi, Plates 88, 89.

3 G. Fiocco, F. Guardi, Plate LXII.

* ' Coui-omiement du buste de Voltaire ' (see Gazette des Beaux Arts, 1903, p. 387).

THE SEATED LADIES: GESTURE LANGUAGE

57

Stand present a marked contrast to the others from the fact that, so far as these posts of honour are concerned, there is no admixture of the other sex. The phenomenon itself points to very different social conditions from those,

like the above, taken from the spectacles of the rococo Age. It seems natural to con- nect it with a matri- archal stage of society, such as otherwise is marked by the domin- ance of the female divinity. Neither, it maybe added, is there anything barocco in the artistic environment of the Knossian groups themselves. They be- long on the contrary to a period of great natural- ism in Art.

That the effect of such pictures of arti- ficial Society life should ^^^^ be here brought, out successfully by crafts- men working in only two dimensions is itself a remarkable perform- ance. Such a result, indeed, could not have been attained had not the spoken word of the Minoans been largely supple- mented and emphasized by gesture language.

Even for a Southern people the constant recourse to gesticulation as a substitute or reinforcement of speech that we witness in these Miniature groups must be regarded as exceptional. Naples, perhaps, but certainly not modern Greece, may afford some parallel. The older gesture language of Hellas, such as we see it in many vase-paintings, has hardly received the scientific attention that it deserves, but it certainly played a very secondary

But seated ladies belong to period of natural- ism in Art.

Promi- nence of gesture language.

2 3

The Seated Ladies : Group E.

58 NEAPOLITAN GESTURE LANGUAGE COMPARED

That of

Naples com- pared.

Gestures in signet scenes : primitive elements.

Women in front seats and mixing freely with men.

part as compared with what we see here. How far it may have been de- liberately kept in the background by the reserved spirit of Greek art it is diffi- cult to say. In any case there is a strong presumption that much of it had been largely taken over from the older civilized element that, even on the Mainland side, can now be shown to have dominated the chief urban centres till at least the thirteenth century before our era,1 and whose actual tongue, in part of Crete at any rate, survived for another thousand years. It may well be the case that Naples itself — that abundant source for our modern knowledge of the sub- ject ~ — where, with Reggio, the Greek of ' Great Greece ' was last spoken on Italian soil, has preserved an uninterrupted South Mediterranean tradition of gesture language, derived on one side at least from a Minoan source.

The gesticulation, however, as seen in these groups of Court ladies, though sometimes pointed, is never violent. It does not over-pass the limits of what is permissible in good society.

Our knowledge of Minoan sign language, as seen on the frescoes, is also supplemented by many of the scenes, at times extremely dramatic, on the engraved signet-rings. We are thus, indeed, led back to a very primitive human stratum, and I have even ventured to compare the attitude of the Goddess, for whom her attendant plucks a fruit from a sacred tree, with a widespread hunger sign and pictograph of the American Indians.3

The women, as we have seen, take the front seats in these shows and the non-admission of male spectators among them may well, as suggested, be a sign of female predominance characteristic of the matriarchal stage. But it was perhaps this very feeling of social superiority that enabled them in the case of the crowd below to mix freely with the other sex. Both are there deliberately grouped together, in a conversational relation. Although, moreover, on the other fragments we see them looking on, as it were, from the boxes of a theatre, there is no sign here of their seclusion

1 Witness the Theban inscriptions, to which I propose to call attention elsewhere, actually painted on the Kadmeian jars of L. M. Ilia style (XI Vth Century B. c.) — after the fall of the Knossian Palace but demonstrating the sur- vival of the script of the Cretan Linear Class B. Similar sign-groups occur, involving identity of language. Some account of these will be given in Vol. IV of this work.

'-' Interesting information regarding the sign language of Naples is collected in the work of the Canonico Andrea de Jorio, La Mimica dcgli Antichi, invest/gala nel gestire Napoletano

(Naples, 1832). The author also attempts a comparison with gestures on Greek vases, that might well be carried farther. Some of the Neapolitan gestures given closely resemble those of the Minoan groups here reproduced. But in the former case there is greater variety in the action of the fingers, and the hand is at times brought up to the face.

3 Myc. Tree and Pillar Cult, p. 79 and Fig. 53 (J. H. S., xxi). See Garrick Mallory, ' Pictographs of the North American Indians ', Fourth Annual Report of Bureau of Ethnology, (1886), p. 236, and cf. p. 235, Fig. 155.

FREE INTERCOURSE WITH MEN

59

in the Oriental, or, one may add, in the South European sense. So far from being veiled their bosoms are at most barely covered by a diaphanous tissue.

Perhaps, however, the most significant feature of the whole compo- sition is the way in which they rub shoulders with the men in the Court below. Their busts are grouped together on the white ground — necessitated in the case of women by the rapid artistic shorthand of this fresco process — which is wedged in patches into the red ground conventionally used for the men. They are clearly meant, however, to be in varied positions — some of them indeed confronting male heads — and they must not therefore be regarded as separate and self- contained elements in the crowd, but rather as units or groups of one or two, mixing freely with the men.

How foreisfn is such free in-

Free inter- course with men.

Yei_t_ow

_feU0E. = f^ED.

Fig. 35. Lady looking out of Balcony or Chamber. (Threshing Floor Area.)

terminghng;

to the scene pre- sented by an Eastern bazaar, or to the spectacle that rises to my mind's eye, of the piazza of a little Sicilian town on a festival day — packed with men in their black jackets, without a woman's face or a streak of feminine colour !

Two isolated pieces of which the restored connexion is reproduced in Fig. 35, x were found in the ' Threshing Floor Deposit ' on the borders of the same area as the ' Temple Fresco '. These are of the same scale, but must be regarded as belonging to some other parallel design. The subject, indeed, is a good deal on the same lines. It shows a lady — doubtless one of a group — depicted in the same style as those above illustrated, standing behind

1 Repeated from P. of 31., ii, Pt. II. stated, but in the ' threshing floor ' area to the p. 603, Fig. 37G. They were not found in the North of it. 'area of the Miniature Frescoes', as there

Parallel fragments from Thresh- ing Floor Area.

Parallel group

6o

WINDOW SCENES

on iso- lated fragment.

Window scenes.

a kind of paling in front of a balcony or box of some presumably theatral structure, with a netted casement behind her, and looking forth at some spectacle in the foreground. Enough of her bent right arm and of the fingers of her raised left hand is preserved, as, combined with her bent right arm, to suggest a gesture such as clapping the hands.

The supporting post of this structure on the left, clearly intended to represent woodwork, is identical in its details and colouring with the upright pillars of the Central Shrine in the companion piece. This parallel with the ' Temple Fresco ' is, moreover, enhanced by the appearance immediately left of this post of a man's face in profile and part of another outlined on a Venetian red ground, and obviously forming part of a closely packed crowd of male spectators like that seen on either side of the little Temple.

Another fragment found in the fresco heaps to the North- West ' contains part of a closely related scene showing groups of ladies with highly elaborate coiffure looking out through casement openings formed of cross timbers. This, which is on a slightly larger scale, stands apart, however, from those above described, and belongs to a parallel class illustrated by fragments of scenes found at Mycenae in which ladies are seen looking out of windows. On the frescoes from the Megaron there, presenting the fronts of buildings • — which from the non-appearance of sacral horns maybe ordinary dwellings of a secular character — single profiles of female busts, disproportionately large in scale, appear at the windows.2 On the fragment found near the Grave Circle is depicted a broad window opening of the kind frequent in the Domestic Quarter of the Knossian Palace, divided into two by a central post, on either side of which fat women let their arms fall over the embrasure.3 Here the sacred character of the building itself is shown by the double axes stuck into the window-posts, in the same way as they are inserted in the columns of the sanctuary hall, a section of which appears in the painted stucco fragments found in the Thirteenth Magazine at Knossos. The identity of the structural features exhibited in the case of these fragments which belong to the pre-seismic phase of M. M. Ill at Knossos, combined with the other resemblances, must incline us to refer the Mycenae fragments approximately to the same Age.

These window scenes, as already shown,4 find a much later parallel in the

1 Cf. r. ofM.,\\, Pt. II, p. 602, Fig. 375.

- Op. tit, p. 601, Fig. 373, d. Cf. B. S. A., xxv (1925), p. 247 seqq. and Plates XLII, XLIII.

3 P. of M., i, p. 444, Fig. 320, and cf. ii,

Pt. II, p. 6or, Fig. 373, c. A coloured repro- duction of this was published by Rodenwaldt, Ath. Mitth., xxxvi (191 r), PI. IX, and see p. 222 seqq.

' P. of A/., ii, Pt. II, pp. 602, 603.

KNOSSIAN LADIES CONTRASTED WITH KADISHTU 61

bronze stand, from Old Paphos,1 while the Nimroud panels with a woman's head looking out of a window 2 carry the parallel still farther East. So, too, the Biblical analogies have been already cited, notably that of Jezebel who ' painted her face, and tired her head, and looked out at a window'.3

The latter description certainly recalls some of these highly tired Minoan Cypriote, ladies. The Cypriote, Syrian, and Assyrian parallels have even suggested and the question — may we have here to do with temple prostitutes, like those of panels Byblos or Heliopolis? Had the Cretan Goddess, too, her kadishtu and ' hierodules' of the Oriental class within her sanctuaries ?

The plump dames of the Mycenae fragment and the even ampler Contrast proportions of the luxuriously somnolent figures on the well-known ivory Knossian mirror handle,4 from the same site, might at least be taken to illustrate the ladies and

' . Oriental

results of pampered seclusion such as that habitual with the better class 'hiero- women in certain Southern countries. But these ladies of the ' Temple Fresco ' and its companion pieces can certainly not be regarded as the products of such a sedentary existence. Still less can they be conceived of as a sacral guild apart, such as those dedicated to the obscenities of the Syrian cult. Their elaborate toilet is sufficiently explained by the festal occasion, and they have obviously taken their seats as much for social inter- course as to see the sports. Mixing with other spectators, slim, lithe, vivacious, many of them were no doubt ready, according to the practice of Minoan girls — seemingly, even in a good position in life — to exchange their gay jackets and flounced skirts for boys' loin-cloths and to step down into the Bull-ring to take their part in the athletic and acrobatic performances.

In the case of the 'Temple Fresco' itself no remains have been Presump- preserved of the lower part of the panel — or possibly a lower band of fresco sp0rts of — depicting the spectacle that the Grand Stand was designed to overlook. bull-nng But a variety of converging evidence from parallel sources is at hand to depicted

i-i. in lower

enlighten us. part of

It has been already shown 5 that among the painted stucco remains Panel- from beneath the ' Kasella ' floors of the Thirteenth Magazine — so important

1 Ibid., p. 602, Fig. 374, and cf. A. S. the Athens Museum from the same grave Murray, B.M. Excavations in Cyprus, p. 10, described by Mr. A. J. B. Wace (B. S. A., Fig. 18. xxv, pp. 369, 370) shows a similar relief.

2 A. S. Murray, ibid., p. 10, Fig. 17. From the short hair and the equally short

3 2 Kings ix. 30. skirts Mr. Wace suggests that these and other

4 From the Grave Pit discovered by Mycenaean ivories may have been imported Tsountas in the dromos of the 'Clytemnestra' from Cyprus. But short skirts themselves are tomb at Mycenae (Tsountas and Manatt, an M. M. Ill characteristic.

Mycenaean Age, p. 187, Fig. 82). Another in 5 See P. of M., i, pp. 527, 528.

62 SPORTS OF BULL-RING IN HONOUR OF GODDESS

The

central

from their ascertained M. M. Ill dating — there was found, together with the fronts of columnar halls and sanctuaries, part of a fresco design showing a crowd of men above a wall T identical in style, on a slightly larger scale, with the throngs on either side of the Miniature Temple (see above, Fig. 15 b). In the same deposit, though of still larger dimensions, was a design of the head of a charoino- bull showing beneath the horns the ends of the flying tresses of some acrobatic figure. So, too, a fragment of painted stucco found in the 'Ivory Deposit' described below shows part of a coursing bull, in connexion with a miniature entablature of a shrine marked by the Double-Axe symbols of the Goddess.2 In other cases, as has been shown, the characteristic pillars that mark the Grand Stands are associated with boxing and wrestling bouts which, like the bull-grappling scenes that they also accompany, must equally be regarded as held under divine patronage.

The central Shrine or little Temple (PI. XVI), which on the Miniature Fresco now under discussion marks the presence of the Goddess at the shnne. spectacles in her honour, answers in its arrangement to that of the little gold models from Mycenae, consisting as it does of a central cella of higher dimensions flanked by two wings. In this case, however, the fluttering doves of the wings — the symbols of divine possession — are replaced by a series of sacral horns, and the central compartment shows two columns, appropriate, it would seem, to a dual cult. The back wall of this central section is of a kyanos blue colour, that of the lateral sections Venetian red and ochreous yellow respectively, though whether these colours have a special religious significance, as they might have had in Babylonia, remains uncertain. Of the half-rosette panel below the central opening something has been already said. The woodwork framing with which this was fixed and the posts and beams of the structure are clearly marked by the brown colour.

That this little building consisted of a central cell with two shallow columnar wings may be gathered by the existing traces already described of a similar shrine on the West facade of the Central Court.3 In the present case, however, we see two columns within the central compartment, while the other structure had only room for one.

The ' Superposed Pillars ' of the Grand Stand.

It remains to call attention to some other remarkable architectural

On either side of the section, the

Super- posed

pillars of features presented by this fresco panel

Grand Stand.

1 Ibid,, p. 527, Fig. 384.

2 See below, p. 207 seqq.

plans and elevation, Figs. 526, 527. Cf. A. E. R. I. B. A. Journ., 191 1, p. 294 seqq.

See P. of M., ii, Pt. II, p. 804 seqq., and

THE 'SUPERPOSED PILLARS'

63

centre of which is occupied by the little shrine, appears a stag- ing consisting of peculiar col- umns, the shafts of which, in contrast to the usual Minoan practice, widen towards their base. The upper and more taper end of the lower shafts is socketed into an oblong block enclosino-'in its blue outer frame- work a series of red and black disks. Superimposed on this block is a column base from which rises a similar shaft, the taper of which continues that of the shaft below. It may be inferred that this, as shown in the restoration given in the Coloured Plate (opposite page 47), was surmounted by a similar block — in this case a true capital, supporting some kind of entablature. The re- maining parts of the column of this form, seen to the left of the central section, are shown in the photographic reproduc- tion, Fig. 36, which also gives a good idea of the incised lines on the surface of the fresco. It will be there seen that the shafts run up the face of what appears to be a projecting pier of masonry.1 This pier forms a platform

Fig. 36. 'Superposed Pillars' of Grand Stand.

It is true that the horizontal lines analogy. Their perpendicular borders separate on the piers show no upright divisions, them off from the similar horizontal divisions of and might therefore be interpreted as steps, the intervals, which otherwise correspond with But, if they are steps, balustrades at the sides their eight lowest lines. Considering the sum- might have been expected according to Minoan mary rendering of many details of this compo-

64 THE 'SUPERPOSED PILLARS'

a little above the oblong block, on which are standing female figures in elegant pose and attire.

The lacunas in our evidence prevent the completion of the structure, but the fresco fragments supply remains for two pairs of such supporting pillars and piers on each side of the central Shrine. Between these in either case are seen lower terraces with female figures, sometimes seated, sometimes standing. It looks, therefore, as if we had to do with the supports of some kind of roof or awning for the Grand Stands of spectators attending the ceremonial show. Parallel Since the discovery of the Temple Fresco several illustrations of pillars

examples. and imposts of the same character have come to light, all of approximately contemporary date. A small bone model of an oblong capital of this type occurred in a Knossian deposit.1 Single pillars of the same kind, as has been already shown, divide the boxing or other contests on a series of steatite ' rhytons ' and gem impressions found at Knossos, Phaestos, and Hagia Triada.2 The fullest illustration, however, is supplied by the relief on a fragment, apparently belonging to a 'rhyton', of grey steatite, found in the N.E. area of the Palace site at Knossos and illustrated in Fig. 37. Here we have a processional scene of youths bearing offertory bowls advancing on the level beneath an isodomic structure, parts of three piers of which are seen stepping up. The wooden pillars with their rectangular imposts are placed at the intervals between the piers, to the framework of which they therefore actually belong.

It is impossible in the present state of our knowledge to explain the

precise constructive value of this system of pillars and imposts. The position

between stepped piers as seen on the 'rhyton' fragment in itself recalls the

gaps once filled with massive upright beams between the bastions of the

Upward North Entrance passage. The distinctive feature of the superposed pillars is

posls'and ^le fact that, unlike the typical Minoan columns — as seen, for instance, in the

their little ' Temple ' — the circumference of which gradually grows greater as it

analogies, . r . ,. , c .

ascends, the shafts in this case become smaller as they mount upwards. Structural analogies for this system can, however, be found at Knossos itself. As we see by the example of the roof of the Pillar Room in the

sition it seems on the whole safest to suppose courses of masonry in the one case it is hardly

that the lines represent courses of masonry, or steps in the other.

possibly sun-burnt brick construction, faced ' With the incised bone inlay of vesica pisa's

with stucco imitative of stone-work. It is form in the ventilating shaft South of the

awkward for a pillar to have stood imme- Domestic Quarter.

diately at the foot of a flight of steps. If it is 2 See P. of M., i, p. 688 seqq.

THEATRAL PILLARS

65

Royal Villa1 a split tree-trunk could be made use of, retaining to a great extent its taper form. The Corridor again that runs beneath the East wall

of the great Palace Hall adjoin- ing the 'Domestic Quarter' shows the same system carried out in the case of upright timbering. Here, as will be seen from Suppl. PI. XXXV,^, show- ing a section of M. M. Ill a walling superposed on a massive M. M. II course, the taper form of the trunk is clearly preserved in the lower section.2

The visible function of these ' superposed pillars ' in the case of the Temple Fresco, as part of the Grand Stand, and apparently as supports of some kind of roof or awning above the tiers of spectators at ceremonial sports in honour of the Goddess, explains their introduction beside represen- tations of agonistic episodes on the small reliefs and intaglios. They are there the symbols of the theatre itself and are introduced as indications of the ceremonial and religious character of the sports. They find in fact an exact parallel in the single Doric or Ionic column placed beside the scenes on Greek painted vases — those, too, often of an agonistic nature— and which stand there as abbreviated indications of the Temple or the Stoa.

Theatral signifi- cance of single pillars in agonistic

Fig. 37. Part of Steatite 'Rhyton' with Offertory Scene and Superposed Pillars.

1 See Vol. ii, Pt. II, pp. 407, 408 and and at the same time to maintain the whole Fig. 235. structure, the interval originally occupied by

2 Here, as in other cases, in order to pre- the woodwork has been filled in with ferro- serve the record of the carbonized timbering, concrete.

III.

§ 7i. The Miniature Frescoes : 2, Sacred Grove and Dance.

' Sacred Grove and

Dance' : centre of interest to left.

The ' Sacred Grove and Dance '—Centre of interest to left ; Self absorption of the female groups ; The Dance — separate performers ; Ritual Dance on Isopata signet; Ecstatic figures — -Sacred Eye in background ; Ecstatic posses- sion— Philistine Prince at Dor ; 'Saul among the Prophets' '; Dancer on Vapheio gem ; Fresco of Dancing Lady in ' Queens Megarou ' — mature L. M. I work; Terra-cotta group from Palaikastro—a ' ring dance' ; Central object of the ' Grove and Dance ' religious ; Aphrodite A riadnc ; Theseus and the Delian Crane Dance ; Both sexes included in later ritual dance ; Traditional Dances of Cretan peasants — the 'Kastriuos ' ,&c; Secret Dance of the women ; Mazy course of Dances ; All ' Chain Dances ' ; ' Leaping Dance ' (tttiSlktos x°P°s) — tumbling performance, as Homeric ; 'Leaping Dance' of Cretan Apollo Delphi uios ; ' Siganos ' and choral accompaniment — Matinadas ; The ' Dancing Ground of Ariadne' at Knossos ; Lis probable position and character as illustrated by Fresco ; Level site bordeied by old olive-trees, be?ieath E. slope of Palace ; The Magic of the Spot.

While in the case of the ' Temple Fresco', described in the preceding Section, the central part, including the little shrine itself, was the most fully preserved, the bulk of the fragments of the companion piece, here entitled the ' Sacred Grove and Dance', seem to lie to the right of the point on which the interest is centred. It will be seen that most of the crowds of spectators, at least of the male sex, have their faces turned to the left. The ladies, equally closely packed, whose heads are seen in the field above the trees are nearly all gazing in the same direction. Those seated in the front row, indeed, like the similar groups of the ' Temple Fresco