181 58 31MB
English Pages 128 [130] Year 2022
WELL BUILT
MYCENAE FASCICULE 14 TSOUNTAS HOUSE AREA
WELL BUILT
MYCENAE
The Helleno-British Excavations within the Citadel at Mycenae, 1959–1969 W. D. Taylour, E. B. French, K. A. Wardle
FASCICULE 14
Tsountas House Area Kim S. Shelton
Archaeopress Publishing Ltd Summertown Pavilion 18-24 Middle Way Summertown Oxford OX2 7LG www.archaeopress.com
ISBN 978-1-80327-154-5 ISBN 978-1-80327-155-2 (e-Pdf) © Mycenae Publication 2022
All rights reserved. No part of this book may be reproduced, stored in retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners. Requests for permission should be sent to: [email protected] This book is available direct from Archaeopress or from our website www.archaeopress.com
For Lisa who inspired and encouraged me
CONTENTS TEXT Foreword (E. B. French and K. A. Wardle) xi Acknowledgements xiii
INTRODUCTION
Note on the Polygonal Tower Historiography of Excavation and Publication Chronological Theories
1 3 3 5
PART 1. THE HOUSE
11 Description and Excavation 11 The House 11 The Entrance 12 Court A 13 Room B 14 Megaron C 15 Rooms D1 and D2 16 Stairs G 17 Basement (Rooms F1–3 and Corridor E) 17 Forecourt H 19 Walls J, K, and M 20 Wall N and Offset O 21 The Forecourt Deposit 22 The Finds 22 The House 22 Forecourt H 22 Wall K 25 Summary 27 Use of the House 27
PART 2. THE TSOUNTAS HOUSE SHRINE (G) Terminology Description and Excavation The Shrine Room G The Shrine — Lower Floor The Shrine — Upper Floor Area Q v
31 31 31 31 34 36 40 41
The Finds Room G The Shrine Area Q Summary Use of the Shrine
44 44 48 51 52 52
PART 3. ACCESS TO THE AREA 55 Description and Excavation 55 Stairway and Central Drain K 56 The Culvert 59 Lower End of Stairway and Drain K 59 The Sacred Way 60 The Upper Ramp from Threshold m to the Middle Ramp 60 The Upper Ramp 61 The Middle Ramp 68 The Lower Ramp and Passage J 74 Area Z 75 The West Cyclopean Wall 76 The Finds 77 Stairway and Drain K 77 The Upper Ramp 80 The Middle Ramp 85 The Lower Ramp and Passage J 87 Area Z 89 Summary and Use 90
PART 4. CONCLUSION
93
Endnotes Bibliography
99 109
vi
TEXT ILLUSTRATIONS FIGURES Fig. 1 Tsountas House Area, detail state plan (C. K. Williams, 1960). Fig. 2 Tsountas House Area, letter key (C. K. Williams, 1960). Fig. 3 Tsountas House Area (W. Dörpfeld; after Tsountas 1886, pl. 4). Fig. 4 The Cult Centre, Greek excavations plan (after Iakovidis 1997). Fig. 5 Plan of Shrine G (E. Olympios; after Mylonas 1983b, fig. 103). Fig. 6 Plaster plaque from Room G: (a) After Tsountas 1887, pl. 10.2. (b) Drawing by M. Reid; after S. Immerwahr, Aegean Painting in the Bronze Age, pl. 63 (1990). Fig. 7 Plan of Mycenaean citadel in LH IIIA2 with position of Shrine G (after French 2002, fig. 16). Fig. 8 The Cult Centre, showing access routes (after French 2002, fig. 33). Fig. 9 Elevation and section of Upper Ramp surfaces (Trench T2C). Fig. 10 Plan with Lower Ramp indicated (after French 2002, fig. 31). Fig. 11 The Cult Centre, with access and circulation routes in LH IIIB2 (after Iakovidis 1997).
6 7 8 9 32 35
55 56 63 75 91
PLATES (for Key to image reference numbers see Supplementary Data p. 492) Pl. 1
(a) Aerial view of the Cult Centre (after Mylonas Guide, fig. 34; photograph: Whittlesey Foundation, 1975). (b) Tsountas House Area, showing the Shrine on the upper terrace of the Cult Centre. From SW. 60/35/17/26A. Pl. 2 (a) The House, ground floor showing Megaron C, Court A and basement stairs. From SW. 50-W18. (b) The House, showing Megaron C, Court A, basement rooms and stairway. From SE. 50-W21. Pl. 3 (a) The Forecourt deposit: pottery, bones and teeth in situ. 1781. (b) The Forecourt deposit: terracotta figurines (60-584, 60/35/13/18A; 60-585, 60/35/13/20A; 60-589, 60/35/13/21A), reworked pot scraper (60-586, 11/DIG/6322) and clay counter (60-588, 60/35/11/19A). Pl. 4 (a) Tsountas House complex, general view of lower Shrine, ashlar altar and House. From N. 50-W12. (b) Room G and Shrine, with the House on terrace below. From SE. 50-W16. Pl. 5 (a) 1950: Room G and Shrine looking towards Tsountas’s dump, Passage S–N to entrance. From SE. 50-W15. (b) 1960: Tsountas House area, ramp and Shrine, with South House Annex excavations in background. From S. 60/35/17/18A. Pl. 6 (a) Lower Shrine with plaster altar. From N. 50-W23. (b) Lower Shrine with plaster altar and boulder in situ. From E. 50-W24. vii
xiv
10
23
30 33
38
Pl. 7
(a) Pots associated with the Lower Shrine and plaster altar (50-234, 50-E32; 39 50-322, 50-E34). (b) Pots from Drain (50-184), Shrine (50-287) and Wall K (50-314). 50-E9. Pl. 8 (a) Tsountas House complex, Lower Shrine and view of Area Q with 43 ashlar altar. From N. 50-W14. (b) Tsountas House, general view from S with Area Q’s terrace substructure and Middle Ramp in background. 59/35/5/3A. (c) Area Q, with ashlar altar and terminus of Middle Ramp. From N. 59-Q2. (d) Shrine floors with intermediate fill. Ashlar exterior altar. From SW. 1474a. Pl. 9 Objects from Room G: 46
(a) Animal figurines (50-109, 11/DIG/6549–50; 50-98, 60/35/1/13), Psi figurine (50-99, 60/35/1/14), ivory helmet crest (50-94, 50-E77). (b) Inlays, appliques and beads after Tsountas 1887, pl. 13. Pl. 10 (a) Objects from the Shrine: plaster wall fragment (50-313, 11/DIG/6422),
clay counter (59-505, 11/DIG/5857), faience fragment (50-579, 60/35/14/29), bronze pin fragments (50-274, 11/DIG/6151). (b) Objects from Area Q: terracotta figurines (50-219, 57/35/2/24; 59-501, 59/35/7/15A), stone conuli (50-236, 11/DIG/6147; 50-237, 11/DIG/6024), lead clamp (50-231, 11/DIG/5887). Pl. 11 (a) Drain and Stairway K. From SW. 50-W10. (b) Lower part of Stairway K. From SE. 60/35/17/22A. (c) Upper Ramp, S end at threshold m, showing ramp floors 1–3. From S. 60-R27. (d) Threshold m and culvert at top of Stairway K, showing ramp floors 1–2. From NW. 60/35/17/36A. Pl. 12 (a) T2C test trench below Ramp 2 with earlier surfaces (3–5) to bedrock. 1831. (b) Upper and Middle Ramps from S. 60/35/17/13A; (inset) Upper Ramp, detail of square poros base n for post. From NE. 1800. Pl. 13 Turn of Upper Ramp to Middle Ramp with bench h. From SW. 59/35/5/7A. Pl. 14 (a) Area Q and Middle Ramp with poros steps e. From W. 1476a. (b) Middle Ramp with steps to Lower Ramp. Area Q with exterior altar. From W. 59/35/5/7. Pl. 15 (a) Area Q and Middle Ramp with poros steps e. From SW. 59-Q4. (b) Lower ramp from stairs to Forecourt. From SW. 1543b. Pl. 16 (a) 1950 fine ware sherds from Street K and Drain. 60-R20.
(b) 1950 fine ware sherds from Street K and Drain. 60-R21. Pl. 17 Tau figurine (50-193, 60/35/1/15), steatite bead (50-6, 11/DIG/6051) and miniature jar (50-185, 50-E9) from Drain; obsidian point (60-179, 60/35/10/28A), animal figurines (60-168, 60/35/12/31; 60-178, 60/35/14/13) and stone loomweight (60-61, 98-R3) from Upper Ramp; horse figurine (59-172, 59/35/6/4) from Middle Ramp; unusual figurine head (59-510, 59/35/7/24A) from Lower Ramp.
viii
50
58
62
69 72 73 78
83
SUPPLEMENTARY DATA PDF https://doi.org/10.32028/9781803271545-data CONTENTS Room Sheets, Materials Lists and Illustrations The Entrance Court A Room B (prodomos) Megaron C Room D1 Room D2 Corridor E & Stairs G Basement Room F1 Basement Room F2 Basement Room F3 Forecourt H Room G Shrine Additional Illustrations
101 103 106 109 113 115 117 120 122 124 127 133 136 140
Catalogues Registered Pottery, Other Materials and Finds Non-Registered Pottery and the Painted Plaster Bagged Sherds
150 397 415
Concordances Registered Find numbers/Mycenae Museum (BE) numbers Non-Registered Pottery & Painted Plaster/Mycenae Museum (BE) numbers Bagged Sherds/Mycenae Museum (BE) numbers
481 483 484
Mycenae Museum (BE) numbers/Pottery, Finds & Bagged Sherds
485
Indexes Index of Registered Pottery, Other Materials and Finds (in Catalogue) Index of Site Photographs (in Supplementary Data)
488 491
General Bibliography
493
https://doi.org/10.32028/9781803271545-data
ix
FOREWORD It was clear from Tsountas’s first reports on his work to the SE of the Grave Circle (Tsountas 1886) that this area contained important buildings and unusual finds including an Egyptian scarab of Queen Tiy, appliqués in ivory or gold leaf and the painted plaster plaque that seems to depict the worship of a goddess. Wace recognised their importance and revisited this area with further excavation in 1950. The full publication of this work was entrusted to Barbara Craig in 1958 who, with the resumption of excavation in 1959 by Taylour in cooperation with the Archaeological Society represented by Papademetriou, conducted further tests while C.K. Williams prepared a new detailed plan. Mylonas continued excavation between 1966 and 1972 (see below, p. 3). All these excavations contributed more to the story of the development of the area and of the Cult Centre, as Mylonas christened the complex following Taylour’s discovery of the Temple and the Room with the Fresco (Mylonas 1972b). However the full significance of Tsountas House and the Shrine, their role in the origin of the Cult Centre outside the Citadel Wall and its subsequent history, have only become fully apparent with the painstaking study presented in this fascicule by Dr Kim Shelton. She has carefully assembled and exactingly analysed all the available information in order to trace the histories of construction, repair and alteration of Tsountas House, situated on the two lowest terraces of the slope, of the Tsountas House Shrine, on the upper Terrace, and of the interconnecting ramps which form the ‘Processional Way’ which descended to the complex from the East. On the basis of the reports and studies by Tsountas and Wace and analysis of the finds made by Craig and Mylonas, Dr Shelton has been able to compile the first comprehensive and detailed account of these discoveries. In the House itself, there were few finds to show its function, but its quality and unusual internal features both demonstrate that it was a better than average residence, whilst its location, opening on to the same courtyard which gave access to all the cult rooms, makes the hypothesis that its occupants had some ritual responsibilities entirely possible. Most significantly, Dr Shelton has established beyond reasonable doubt that the Shrine (also known as Shrine G) was the earliest cult installation constructed in this part of the site, outside the circuit wall. She has shown that it continued to be used when the Temple and the Room with the Fresco Complex were added to the N. These lay beyond an open courtyard which could readily be reached by anyone from the town of Mycenae outside the Citadel and which gave access to all the different parts of the cult complex. During this period, from early in the 14th to the middle of the 13th century BC, the Shrine, equipped with a curiously elaborate hearth or altar, provided an interior space for ritual activity and received valuable, even exotic offerings. This situation changed radically at some point after the middle of the 13th century BC when both the Room with the Fresco and the Temple can be shown to have suffered serious damage, perhaps as the result of an earthquake, and the West Circuit Wall was extended to the W to enclose both the Grave Circle to the N and the whole Cult Centre (Wardle 2003; 2015). As a result of these events, the whole character and focus of the Cult Centre changed, as can be seen most clearly from the history of the Tsountas House Shrine set out in this fascicule. The Room with the Fresco ceased to be used for cult purposes. The Temple was repaired but seems no longer to have been important. The altar in the Tsountas House Shrine was covered by a new xi
floor and replaced with a stone-built altar in the courtyard. An elaborate ‘Processional Way’ with a series of monumental gateways was constructed to lead up to (or down from) the Palace on the summit of the Citadel and decorated for at least part of its length with a processional fresco to match its function. The orientation of those cult rooms which remained in use, including the Shrine, was now, in the terms introduced by Hägg (1981), palatial and official rather than popular. Access was restricted to those permitted entry to the Citadel and this arrangement continued until the destruction of the whole of the area at the end of the 13th century BC. These conclusions will require a radical reassessment of the nature and history of Mycenaean cult practice, since much of the current discussion is still erroneously based on the belief that the Mycenae Cult Centre was, from its creation, intra muros. Dr Shelton has demonstrated convincingly that the Cult Centre was created for the wider Mycenae community and that the manner of its use changed when it was incorporated into the Citadel. E. B. French and K. A. Wardle
Very sadly, Lisa’s death in June at the age of 90, came before we could bring Tsountas House to press. As in so many ways with younger colleagues, she had encouraged and supported Kim Shelton’s work in the initial study, the elaboration of the text and the final stages of editorial oversight. Like her father before her, she was always generous with her knowledge and her time. She is already sorely missed as we continue the project she devised — the Well Built Mycenae publication. No small thanks are due to her and to Linda Witherill for the financial support which they have provided in recent years towards the completion of this and other fascicules. KAW October 2021
xii
ACKNOWLEDGEMENTS This research was undertaken with financial support first from the British Excavations at Mycenae and the Mediterranean Archaeological Trust and in the later stages from a Humanities Research Fellowship and the Department of Classics, University of California Berkeley. The manuscript would not have been completed without the tremendous intellectual and practical support I received from Elizabeth French, especially while I was on sabbatical in Cambridge. The proximity to the Mycenae Archive and her expertise was invaluable. I am also extremely grateful to Lisa, together with Ken Wardle, for allowing me to research and publish this material, and for being so incredibly patient. I hope it has been worth the wait since this has been one of the most challenging and complicated puzzles with which I have worked. I am honoured also to have been entrusted with the completion of the work begun by Barbara Craig, whose detailed study and insight created a solid foundation on which I could build using all the ‘building blocks of information’ we have amassed in Mycenaean archaeology and in archaeology at Mycenae since the 1960s. Much of the research was completed in the Mycenae Museum and I would like to acknowledge the important assistance of the D′ Ephoreia in Nauplion and thank the staff at Mycenae, especially Nikos Katsoulieris in the apotheke, for his unending dedication, resourcefulness, and photographic memory. My drawings were inked by several of my students at UCB who also helped with the object catalogue and its initial formatting. Thanks are due especially to Samantha Alford and Elizabeth Niespolo. Additional cataloging of non-context sherds from Tsountas’s dump was tirelessly compiled by Lynne Kvapil and Francesco Iacono during a summer season at Mycenae with additional photographs taken by Debra Trusty.
xiii
PLATE 1
A. Tsountas House B. Processional Way G. Shrine D. Exterior altar K. Courtyard with circular altar T. Temple Complex
(a)
(b) (a) Aerial view of the Cult Centre (after Mylonas Guide, fig. 34; photograph: Whittlesey Foundation, 1975); (b) Tsountas House area, showing the Shrine on the upper terrace of the Cult Centre. From SW.
xiv
INTRODUCTION The Tsountas House Area is the southern extent of the excavation within the citadel of Mycenae conducted by the British School at Athens. The area is located on the W slope of the acropolis S of the Citadel House Area. This part of the fortified palatial site of Mycenae has become known as the ‘Cult Centre’, following Mylonas (Mylonas 1972b, 124–26; see also French 1981), due to the concentration of ritual characteristics in the architecture, features and moveable objects in buildings such as the Temple and its associated rooms, the adjacent ‘Room with the Fresco Complex’, the courtyard altar, and the Shrine of Tsountas House to the S. Segments of the area have been under investigation since 1886 and the sheer complexity of the excavation historiography (see below) together with the complicated depositional sequences and associated finds has meant that study and publication is still underway. The Tsountas House Area consists of two independent but adjacent and parallel buildings, the House and the Shrine upslope to its E, and the various routes of access to, and circulation routes within, the Cult Centre, including the processional ramp system from the E, the ‘causeway’ ramp and corridor from the N, and the Stairway ‘K’ on the S. No less than four excavation programmes have been conducted here: the earliest and last are two projects by the Archaeological Society of Athens that bookend the work of Wace and Taylour in 1950, 1959, and 1960. This publication will address primarily the material from the British Excavations but also incorporates whatever is known or deduced from Tsountas’s excavation, including finds recovered from his excavation dumps, and the accounts published from the Greek excavations of the 1970s. My objective is the complete presentation of all the factual details of the excavation of the buildings together with their stratigraphic context and the spatial association of the finds and features, as well as an analysis of the implications for a consideration of their spatial, chronological and conceptual contexts in the history of the Cult Centre, the site of Mycenae and for the place of this area and its moveable artifacts in a better understanding of Mycenaean religion and culture. Many readers will be familiar with the area generally and also have some knowledge of specific finds or details of evidence, much of which is published fully here for the first time. This definitive account of the excavation, stratigraphy, and finds along with the detailed interpretation from the original excavation materials supersedes all previously published accounts including the summaries presented in WBM 1 by Lord William Taylour. The chronology of the area is not very well understood by many scholars, especially the various phases of construction, use, and destruction. The excavation of the Tsountas House Area includes the investigation of the earliest known religious installation at Mycenae, the Shrine, and therefore, it represents the origin of the Cult Centre. The excavation is also important in the discussion of Mycenae’s socio-political development as it contributes to further understanding of the design and creation of the citadel that encompasses and isolates by fortification wall a previously accessible area of ritual and thus directly impacts on the dialogue between popular and official religion and on the role of a palatial administration. The text is divided into three principal parts, one on each of the independent structures within the area and the third on the access and circulation routes, supported by detailed information and catalogues on the supplementary PDF. The discussion of the buildings or areas begins with a detailed presentation of the excavation and the relevant architecture with measurements and construction characteristics and illustrated by a selection of the excavation photographs. 1
Further information for each area or room is given on the Room Sheets in the Supplementary Data PDF, together with additional photographs and other supporting materials. The overall plan and function of each building or feature is then considered along with a room-by-room or area-by-area discussion of the pottery and other artifacts that determine the construction and destruction dates as well as the interpretation of function. Materials Lists, which accompany the Room Sheets, can also be found in the PDF. The first section presents the House, the largest structure in the Cult Centre and the centrepiece of the initial excavation by Tsountas in 1886. Most of the building was cleared to its floors during the earliest excavation with little physical material kept. The finds, however, from the 1950, 1959, and 1960 excavations provide important evidence for the lifespan of the structure. The use of the House is also discussed and this study is especially important because of the structure’s intimate physical connection to surrounding buildings that are clearly of exclusive religious use. The second section focuses on the Shrine, including its complicated architectural development and the evidence for its religious nature. The study presents primarily the excavation and finds of the British Excavations but it also includes the finds published by Tsountas from his excavation of the adyton Room G. This building contains unique and complex built features as well as moveable artifacts; both categories of finds are potentially religious and cultic in nature. The evidence for the chronological history of construction, destruction, rebuilding and final collapse, gained through the building plans and artifact study indicates the complex life-history of the structure and a many faceted and changing function. The use of the building, the earliest in the Cult Centre, is discussed by phase, as are details of the significant changes in the architectural composition and focus which occurred whilst a continuity of ritual was maintained. The third section presents an account of the access routes to the Tsountas House Area and the Cult Centre more generally, as well as a diachronic understanding of the development and nature of the circulation routes within the Centre. The structural make-up and excavation details relating to these routes are outlined along with the interpretation of the evidence for the stratigraphic sequences of each area. The finds of all materials, especially pottery, are detailed by category and discussed in terms of their chronological and functional significance. Finally, a summary of the chronological development and use of the access points, routes, and physical boundaries to the Cult Centre is provided. The final part of the text is a conclusion that brings together the interpretation of the individual structures, features, and areas to provide an overall analysis of the development of the Cult Centre and its potential ritual function over several periods in an external, ‘popular’ context and an internal, ‘official’ context. In light of what we now know about the history of the Shrine it is important to recognise its position in the sequence of activity that created the Cult Centre and to clear up some important misconceptions about the area and its use. The finds in the text and in the individual catalogue entries on the supplementary PDF are identified first by their individual excavation number (for example, 50‑146), that describes the excavation year (1950) and the unique number in the excavation register (146). Most finds also include their museum registration number (e.g. BE 7023), meaning the ‘Biblivo Eisagwghv~’ (Accession Book) of the Mycenae Museum followed by a unique number. A few objects have registration numbers from the Nauplion Museum (NM) but no BE numbers. This means that the artifacts were registered in Nauplion storage after the excavation but were not moved to the Mycenae Museum, probably because they were not found or recognised prior to the move 2
of the collection in 1998. A very few items have only an excavation number and seem to have been lost or disintegrated before they could be registered in Nauplion. Finally, Tsountas’s finds are in the collection of the National Archaeological Museum in Athens (NMA), and if their registration number is known it will be included here. It is hoped that continued study of material from the 1886 excavation will allow for a fuller publication of the relevant finds in the near future.
NOTE ON THE POLYGONAL TOWER The Polygonal, or Hellenistic, Tower is a dominant feature in the West Cyclopean Wall (WCW) of the citadel of Mycenae. It is not really a tower in the traditional sense, a fortified extension of a circuit wall used primarily at weaker or more vulnerable points in the circuit, especially to provide additional surface area often in a flanking position to protect against attack. The Polygonal Tower (H. 18 m) is actually an extensive patch of polygonal masonry on the exterior of the Mycenaean fortification wall. There is no interior face to the tower, only a perimeter on the top of the wall creating a platform area that includes a cistern. Following the general destruction in LH IIIB2 late, in the area within the WCW, a great deal of debris accumulated against the inner face of the wall.1 A long series of habitation phases follows on top of the eroded mass and this is followed by further debris of the wall itself when it was partially dismantled by the Argives following their attack on Mycenae around 468 BC in retaliation for the Mycenaean participation in the Battle of Plataea (Pausanias II.16.5–7). This dismantling or disabling of the fortification wall seems to have been inflicted on importance points in the circuit, such as the tower and adjacent wall W of the Lion Gate and in the NE Extension. The walls were repaired in the 3rd century BC using a contemporary masonry style when Mycenae became a kome of Argos and the citadel was largely reoccupied and built over. This reoccupation included the construction of a Doric Temple probably dedicated to Hera in the area of the palace (Klein 1997).
HISTORIOGRAPHY OF EXCAVATION AND PUBLICATION In 1886 Christos Tsountas began excavation at Mycenae for the Archaeological Society of Athens (Fig. 3).2 One of the first places he investigated was the lower western slope of the citadel, just inside what is now referred to as the Polygonal or Hellenistic Tower. He dug to various depths in different areas and revealed almost entirely a structure (the House)3 consisting of two terraced levels that resembles a house in plan with living / working areas and organised storage facilities. He also partially cleared a building (G) of two rooms just upslope that contained high-value material and objects and some with religious iconography, such as ivories, beads, and goddesses in glass and on painted plaster (Fig. 6). Tsountas did not collect sherds or broken pottery vessels, although he does mention that there were many pieces in the fill he dug but none whole (Tsountas 1886, 78). He only removed upper levels above the N room of the building while he cleared the smaller, southern room to bedrock. Tsountas’s work was never completely published although preliminary reports appeared in Arcaiologikhv Efhmeriv~ (1887)4 and Praktikav (1886). He draws on this material for general statements on Mycenaean architecture and domestic life in his book, Mukhvnai kai mukhnai>kov~ politismov~ (1893),5 later published in English with Manatt as The Mycenaean Age: A Study of the Monuments and Culture of Prehomeric Greece (1897). Tsountas describes the state of the W slope and the deep fill inside the Polygonal Tower, where at a depth of eight to ten metres he uncovered the ruins 3
of a Mycenaean era house (Q on his plan), and over that, ruins of later houses but all earlier than the tower.6 The potential importance of this area of the Mycenaean citadel and its material remains was only fully realised during further excavation of the Tsountas House, now named after its original excavator, by Wace in 19507 and continued by Taylour in 1959 and 1960 (Fig. 1).8 Wace reopened and examined the House in detail,9 while the structure on the terrace above, G, was discovered not to have been completely cleared. Therefore, it was more fully investigated and identified as a cult building, or Shrine, one of several excavated by the British Excavations during their campaign. Under Taylour and Papademetriou, the areas to the N and E of the House / Shrine area were explored beginning in 1959 (Papademetriou and Taylour 1961, 490). There was much work done on the ramp, the plastered approach to the Shrine and the platform to its N in addition to parts of the House and the poros steps between terraces. In 1959, only the western part of the ‘Tsountas House Area’ was excavated.10 It was undertaken at the end of the season to investigate the area W of the poros steps but according to Taylour it proved disappointing (WBM 1, 24). Excavation was expanded in 196011 to include large segments of the upper ramp and part of the middle terrace that would prove to include the southern part of the Megaron building and the N part of Tsountas House, including its approach and entrance (WBM 1, 29). The Helleno-British expedition used the ground floor of the House for pottery sorting, partly under a temporary structure, and used the three basement rooms as headquarters (see also below, Part 2).12 Cooking was done in the northern alcove. The middle room was the headquarters with storage for supplies and a covered area for meals. The southern room was known as ‘Billy’s bower’; roofed with Agnus Castus, it was used by Lord William Taylour as his office (EBF email 10/07/10). After Wace’s death in 1957, Taylour asked Barbara Craig, who had joined the Mycenae team in 1956, to complete the publication of the complex known as Tsountas House dug in 1950. She completed supplementary excavation and study as part of the 1959 and 1960 seasons. She produced copious notes and a preliminary manuscript in the mid to later 1960s that was never published, but was given to Mylonas as he began excavation in the same area. Work continued on the lower W slope of the citadel, during the 1970s (with initial soundings in 1966 and 1968)13 once again by the Archaeological Society of Athens under the direction of Mylonas, who first called the whole area ‘the Cult Centre’,14 and research came full circle with the reopening of Tsountas House ‘Shrine Gamma’ and the closer inspection of its surrounding features.15 This work too remains to date essentially unpublished. However, Mylonas and, later, Iakovidis provide a great deal of information, some of it contradictory, in a number of publications (Fig. 4). Mylonas in the English summary of The Cult Centre of Mycenae says that the excavations of Tsountas and Wace proved that religious ceremonies were held in the so-called ‘House of Tsountas’ excavated in 1886 (CC, 36). He clearly refers here to the House itself in addition to the Shrine since he goes on to describe Wace’s work in 1950 in ‘the section of the house that occupies the uppermost level…’ He follows with a list of his contributions to the uncovering of the area including the entrance vestibule with wall paintings on the ramp and the upper part of the approach ramp and staircase. His Greek text explores the evidence for Mycenaean religion from excavation prior to the Cult Centre and then describes in detail its buildings and features (CC, 11–15 and 15–35). Iakovidis in 1983 says this is clearly the cult centre of the citadel ‘despite its remoteness from the palace’ (Iakovidis 1983, 48). 4
CHRONOLOGICAL THEORIES Tsountas was able to distinguish between palatial Mycenaean and Hellenistic remains and colour-coded his 1886 plan to illustrate the different periods of construction. The post-palatial material (LH IIIC) over the ruins of the House was identified simply as post-Mycenaean. He believed the House to be contemporary with the palace and to have belonged to the Mycenaean elite with a plan similar to that of the palace, and the one at Tiryns, where he believed that ‘there are two apartments (for women and men) that communicate’. The Mycenaean strata of the Shrine’s N room and the remainder of the building were first cleared by Wace in 1950. He expressed the view that the re-excavation of Tsountas House proved exceptionally interesting. Although he states that the area was earlier believed to be two buildings, he suggests that it is one building and goes on to describe the stepped ‘street’, a lane running N to ‘a large hall with stuccoed floor’ and benches for cult objects along the E and S walls (Wace 1951, 254–55). This Shrine was ‘the first of its kind to be discovered at Mycenae.’ The House is interpreted as belonging to the ‘priest-in-charge’ because it is so closely connected to the Shrine and access to it seems only to be through the Shrine.16 Taylour suggests a date17 for the Tsountas House complex (Shrine on terrace above and House) together with the Megaron, Temple and Room with the Fresco, although he admits that ‘the construction deposits from these buildings are slight and do not provide such a clear indication of date,’ while the repairs carried out to the Shrine may be contemporary to others in these buildings following a catastrophe. In his 1977 Guide, Mylonas dates the first Cult Centre destruction to the end of the 13th century BC, and suggests that rebuilding in part then occurred with continued use until the end of the 12th century BC, ‘when the final destruction of the site took place’ (Mylonas Guide, 24). Iakovidis sees two building periods, separated by a destruction shortly after the middle of LH IIIB, which was caused by an earthquake. ‘Already in LH IIIB the area [was] delineated to the N by the South House, to the S by stairs and its drain, to the E by a processional way, and around 1250 BC, the second phase of fortification wall closed the area on the W. The area was only systematically filled in the 13th century BC (Iakovidis 1997, 148–49). During LH IIIC, house walls were built over the courtyard of the House and the processional way, blocking its use, but indicating ‘no interruption in the use of the area’ (Iakovidis 1983, 48). All of the chronological interpretations up to now have been based on very preliminary accounts of the excavations without any detailed study of the ceramic evidence, as scanty as it may be for some areas or phases. This publication presents for the first time the complete analysis of all of the evidence for dating the Tsountas House Area, its phases of construction, use and destruction, based on a close study of the original material and the copious records and archival documents from the British excavations.
5
Fig. 1. Tsountas House Area, detail state plan (C. K. Williams, 1960).
6
Fig. 2. Tsountas House Area, letter key (C. K. Williams, 1960).
7
Fig. 3. Tsountas House Area (W. Dörpfeld; after Tsountas 1886, pl. 4).
8
9 Fig. 4. The Cult Centre, Greek excavations plan (after Iakovidis 1997).
PLATE 2
(a)
(b) (a) The House, ground floor showing Megaron C, Court A and basement stairs. From SW; (b) the House, showing Megaron C, Court A, basement rooms and stairway. From SE.
10
PART 1. THE HOUSE18 References: Tsountas 1886, 74–79; 1887, 156, 160–72, pl. 10; 1893, 15, 41, 43–44; Tsountas and Manatt 1897, 7, 67; Wace 1949, 66–67; 1950; 1951, 255; BSA 51, 122; Wace and Stubbings 1962, 395 and 496; ILN 1961; Wace Guide, 35; WBM 1, 9–10, 15, 19–20, 24, 28–29; Mylonas CC, 15–37; 1972b, 121–22; 1974, 90–91; 1977b, 22; 1983a, 308, 315–16; 1983b, 140–41; Iakovidis 1983, 47–48; 1997, 154–59.
DESCRIPTION AND EXCAVATION (Figs. 1–2) THE HOUSE The House is built on two levels on the slope of the hill (Pls. 2a, 4a–b); it has a basement on the low ground to the SW and to the NE a higher floor created by the rising slope — the megaron area.19 It was the primary focus of both excavation and publication for Tsountas of his work on the W slope of the citadel in 1886 (Tsountas 1893, 15, 43–44; 1886, 74–79; 1887, pl. 10; Tsountas and Manatt 1897, 7), uncovered within the bounds of the Cyclopean fortification wall. Tsountas reports that the great depth of fill against which the Polygonal Tower was built, was made up of successive ruins of houses, all of them earlier than the tower itself and that the Mycenaean House remains were as deep as eight to ten metres (Tsountas 1893, 15). On the upper level are Megaron C (d),20 porch/room B (g), and Court A with two narrow rooms, D1 (e) and D2 (z), flanking the megaron and porch to the W. A flight of stairs (stone built and covered with short wooden planks),21 G, leads down to the lower level of the basement, with its three rooms, F1, F2, F3, giving on to a long corridor E. Of the basement rooms the most southerly (F3) is double with an intervening doorway and the most northerly (F1) is L-shaped. At a strange angle to both the stairs and Court A is Forecourt H (b). The W, exterior wall of the basement is very substantial. The S wall and the walls dividing the three basement rooms are also heavy. Tsountas had no doubt that they originally supported an upper storey and that one continuous roof ran over this and rooms C, B and D of the terrace above.22 He did not though think it possible to locate the position of the entrance to the house because in later years other constructions in the area disturbed it (Tsountas 1886, 77). This would have been a fair-sized structure. Tsountas, however, characterises the house as small but goes on to describe it in some detail.23 For the most part, he describes the house as constructed like the palace, the other area of the Mycenaean Acropolis he excavated in 1886: the walls are made of small stones and clay (Tsountas and Manatt 1897, 67) and covered with lime plaster, usually coloured (frescoed), the parastades (jambs) are wood on a step/base of psammitou (poros limestone), the surface of the court and megaron are plastered.24 Tsountas compared the house plan to the similar plans of the palaces of Mycenae and Tiryns and believed the house, belonging to the Mycenaean elite, to be contemporary with the Mycenae palace.25 In 1950, the British expedition re-cleared and re-planned the House. Wace suggested that the house belonged to the ‘priest-in-charge’ because it was interpreted as being closely connected to the Shrine through internal and perhaps exclusive access (Wace 1951, 255; 1956, 122; Wace and Stubbings 1962, 395 and 496). This was re-evaluated during the intensive study of the architecture in 1959 and 1960 for the Williams plan (Fig. 1) and found not to be a viable reconstruction. A few tests were made in Megaron C, Forecourt H and Passage J
11
(the Lower Ramp). The area to the N of the forecourt was excavated for the first time since 1886 and cleared to a far greater depth than by Tsountas; part of the forecourt itself was excavated to bedrock.26 The Greek excavations under Mylonas cleared some of these areas again, such as in the forecourt of the house.27 He suggested that the ‘character of the building’ is as yet undetermined even if Wace suggested it as a priest’s house (Mylonas 1983a, 316), but then goes on to suggest that the ‘almost inaccessibility of the building’, the pithoi and lead vessels in the main room, the division into small compartments indicating special use, can be interpreted as a ‘treasury’ of the Cult Centre for the storage of rich offerings, such as clothing and scented oil.28 On structural grounds, the upper level of the House (megaron, porch and court) was built after the original terrace of the Shrine but before the pseudo-Cyclopean terrace wall, flanking G on the W, was added. The E wall of the House conforms and limits its orthogonal shape to the line of the pre-existing upper terrace wall that is also incorporated into its rubble foundation. Later, the terrace addition impinges on the megaron’s E wall. It is not easy to understand why the terrace on which the megaron and court are built is orientated as it is and why it was necessarily so cramped. It may be simply that the position of the underlying bedrock dictated absolutely the placement of the ground floor even if it could have been substantially levelled. Or the layout was required to fit into an existing space; it actually respects and avoids the area to the SW beyond the basement at the base of the stairway (K). In the narrow E–W trench (~ 4.5 m × 2.5 m) excavated by Tsountas between the back wall of the basement and the Polygonal Tower there was no sign of any building contemporary with the House and which therefore might have constricted the space available,29 nor was the fortification wall or any potential point of access through it the reason, since it was not yet built (see below in Part 3). The architect did his best with the irregular platform at his disposal, but the result is a little odd. Tsountas excavated the entire structure to its floors; those of the court and, with one exception, of the upper rooms were all plastered, as confirmed by the re-clearance of 1950. Gaps in the floor plaster show the original positions of thresholds and cross-walls even where these have completely disintegrated or vanished; there were poros bases for the wooden doorjambs. The existence of ‘chases’ in the rubble masonry for a timber framework indicates that the upper walls were of mudbrick strengthened in wood as usual for Mycenaean construction. The basement walls were covered with mud-plaster while those of the megaron had a coating of lime plaster over the mud, probably once decorated with frescoes (Wace 1949, 67). The Entrance The entrance to the House was at its N end, from the Lower Ramp through Passage J, and into Court A. When the upper terrace was extended northwards, at the time of the extension of Area Q to the N of the Shrine (see Part 2 below), the approach and entrance into Tsountas House through Passage J was considerably narrowed; so was the Lower Ramp on its E side, where the NW corner of the terrace wall supporting the platform overran the upper floor of the ramp in Passage J. When the large masonry pier/buttress l was built on the surface of the ramp in order to provide additional essential support for the terrace above, the entrance to the House was almost blocked altogether. The ramp floor runs under both the terrace wall on the E of Passage J and the buttress, and reappears at the N end of the court from beneath a stretch of coarse thick plaster between l and a rectangular rubble pier to its W. When this ramp floor was laid it is clear that Passage J led into 12
the court of Tsountas House, as well as beyond to descend to the W to the lower courtyard. The entrance into Court A may have been less restricted and been positioned about 1.0 m further S than it was in its final phase. The creation of Forecourt H and the addition of three steps to Stairs G (see below) required to reach the higher floor level of the forecourt, also transformed the area and made the continuation of traffic past the entrance to the W unlikely. The addition of a rubble pier parallel to the buttress to its W both defines a narrow, restricted entrance to the court but also positions the entrance further N, in line with the new head of the stairs. Buttress l was incorrectly interpreted and reconstructed in 1950, and therefore was drawn on the plan and section by the architect Hobbis, as the foundation for a stairway.30 This large pier or buttress was thought to be the only logical place for access between the upper and middle terraces of Tsountas House, when the two terraces were believed to be part of a single structure31 and before the excavation of the poros steps e in 1959, further N along the same terrace wall. It is one of the last features constructed in this area, certainly later than all of the buildings, walls, and ramp surfaces. Its purpose is structural and it was probably built after the destruction in mid LH IIIB as a repair or support for a heavy, and possibly sagging, terrace platform. No firm dating evidence has been found that will give a closer date for its construction. Court A Court A is a large open space, roughly orthogonal with a truncated NE corner because of the position of the pre-existing Shrine terrace, the W wall of which is also the E wall of the court, and does not continue the line of the E wall of the megaron and porch. The E wall was originally plastered. The architect had not been at pains to contrive neat rectangles, where symmetry presumably mattered less than indoors. For the E side of the court he was content with the line of the terrace wall. Bedrock does not project above the level of the court floor on this side as in the megaron and porch, but it cannot have been much beneath the floor and accounts for the appreciable slope up of the court floor towards the E. The floor itself is lime plaster over a layer of stones and earth on bedrock. By 1950, much of the plaster was very worn, but two distinct floor surfaces were observed. The entrance to Court A is a threshold between a stone buttress l and a rectangular pier of rubble and timber on the W side. Evidence for a vertical wooden beam was found at the SW corner of the buttress and indicates a doorframe or pivot beam. The pier was built over the plaster floor and represents a later phase of the House entrance. Remains of a drain entering from Passage J were found cut in the rock in a shallow channel and passing below the pier from E to W. The channel becomes deeper and roughly square and was roofed with small stone slabs. The drain runs into Court A where the bottom of the channel is 0.26 m deep and then dives further underground, presumably emerging as the drain in Corridor E at the foot of Stairs G. Towards the N end of the court, and S of the rubble pier, was found a column base, which suggests the existence of a porch roof or stoa coming into the yard. The base is a roughly hewn squarish block of limestone of which the upper part has been cut to a circular shape rising slightly above the floor level. The wooden column had a diameter of c. 0.40 m. The base is not on the central axis of the court but in line with the W anta of the entrance at the N end. Tsountas does not mention, though his plan shows, part of a circular, cement lined, Hellenistic cistern sunken into the E wall of the court. Presumably he cut away the greater part of the pit, which would have projected far into the court from its E wall. 13
In the NE corner of the court, formed by the junction of its E wall with the large pier/buttress l, Tsountas’s plan indicates two small jars on a stand or platform; none of this is mentioned in his report. In this same NE corner in 1950 was found a pithos (H. 0.40 m, Di. rim 0.47 m) held in position by clay and stones and beside it, fragments of a very large ovoid jar (50‑337, BE 5816) (Di. of flat base c. 0.20 m) of coarse reddish ware. Inside it was the raised base of a pot (50‑338, BE 5815) of similar fabric but well smoothed outside. The pots were burned, as was the soil around them that included fragments of chipped stone. Just E of the centre of the court there are two worked poros stones fitted together and hollowed out to form a shallow basin (Di. 0.40 m) with a central trefoil shaped opening measuring at the most 0.16 m. The ground under the basin is hollowed to a depth of c. 0.35 m.32 Tsountas’s first interpretation was a drain opening for run-off of water in the court, however, the earth in the hollow under the basin contained ashy soil mixed with charcoal and some bone; therefore, he came to believe this feature to be a bothros for sacrifice.33 This basin lies well below the final floor level; the plaster all around it has been worn away, but presumably it originally sloped to meet the edges of the stones. Tsountas suspected that the basin led to a drain under the court and in 1950 the basin was found to be in line with the mouth of a drain discovered in the E wall of the basement Corridor E just S of the foot of Stairs G. However, this connection has never been confirmed. Tsountas’s suggestion that the basin was used for burnt offerings should be considered.34 This view is perhaps supported by the carved trefoil opening, which seems unusually elaborate for a drain. It would then be a domestic altar. Although the ashy earth in the hollow is more likely due to damage of the House by fire, or remains of domestic rubbish, the basin could still be for offerings: perhaps it could be a receptacle or conduit for libations. Room B From the court one enters a small porch or vestibule Room B (prodomos), which is almost square in plan (2.30 m × 2.0 m), and from there enters the megaron room to the S over a wooden threshold, now represented by a gap in the plaster floor of the room. At the N end of the room the floor plaster merges without a break from that of Court A, showing that there was no threshold here and, therefore, no door. Another gap in the floor plaster along the W side of the room suggests a second wooden threshold leading into Room D1. The floor is a coarse plaster ‘pebble-mosaic’. Two poros anta blocks define the entrance to Room B, the eastern at the end of the E wall and abutting the terrace wall (0.76 m × 0.74 m, H. 0.36 m above floor). There are five square to rectangular dowel holes in the upper surface of the anta, two on the N and three on the W, for the attachment of wood facing over the rubble wall core. The E wall is unusually thick and widens to the S with a tooth extending W at its S end, flanking the threshold to the megaron. This creates a nearly square room and covers bedrock projecting above floor level along this side of the room. The dressed blocks on either side of the entrance once supported piers, presumably of rubble. The western anta block (0.73 m × 0.65 m, H. 0.43 m above floor) stands in isolation and has a rectangular dowel hole in its inner face near the N end. The corresponding hole in the E block is not opposite that in the W but some 0.40 m further S near its S end. The W anta has five rectangular dowel holes on the upper surface at the N, E and S sides. Wace noted, in 1950, that the positions of the dowel holes in the upper faces of the blocks were marked by fine incised lines (as they are also in the ashlar blocks c and dd in Area Q on the terrace above). The cutting on the blocks suggests that the rubble piers had been faced with wood. 14
Tsountas’s plan shows a wall running from the W face of this block to the W wall of Room D1, but although there is a gap in the floor plaster visible in the 1950 photograph, no trace of a wall was found. In this instance, the gap is obviously shallower than the other two and it appears that the wall here was erected over the plaster floor. Tsountas expressly says (Tsountas 1886, 76) that it was impossible to tell whether Room D1 (e on his plan) was entered directly from Court A or indirectly from Porch B, which shows that all he found at the N end of Room D1 was a gap instead of the wall his architect has drawn. The absence of dowel holes on the western edge of the W anta block also could indicate the existence of an E–W wall and suggest that entrance into the western rooms D1 and D2 was from Room B rather than Court A, at least for part of the lifespan of the House. In 1950 it was noted that the anta blocks were cracked and discoloured by fire and that the eastern one had traces of plaster moulding round its base. 1950 photographs suggest that around the base of the W block was a small stone packing, no doubt once set in clay and plastered. Megaron C Megaron C (domos) is a modest room of 4.40 m × 3.55 m, with a rectangular central hearth. The walls of the room, made of the usual stone rubble packed with clay, faced with a layer of mud plaster and that covered with a layer of lime plaster, are preserved to a height of 0.45 m. The small amount of painted wall plaster Tsountas recovered from the house had band decoration only,35 with the exception of one piece from the megaron which had a small black and white checkerboard pattern, such as Tsountas found in the Palace Megaron also. The floor is of fine lime plaster, preserved in large patches, between which can be distinguished an earlier pavement indicating two phases (Mylonas 1983a, 316; 1983b, 141). In order to achieve a neat rectangle for the inner face of the walls, the architect made the E wall thicker at its S end than at its N, similar to the E wall of Room B but with a different line of the wall even while maintaining the same orientation. The foundation of the NE corner of the room had to be cut in part out of bedrock. The continuation of this bedrock further NE was concealed by the extreme thickness of the E wall of Porch B. The E wall of the megaron seems to be of stepped construction, with the top of the first step at the same height, about 0.40 m, as that of bedrock at the NE corner. This step makes a narrow ledge (0.15 m) along the foot of the E wall and was once the seating for a horizontal beam (see Pl. 2a). Beyond the first step of the E wall of the megaron a part remains of the second step, but none of the wall above that point is preserved, although it is clearly built against the older terrace behind it. Above the megaron wall rises the heavy pseudo-Cyclopean terrace wall of G, built of much larger stones, that was added in support of the terrace and building above probably in the second half of the LH IIIB period, ‘stepping’ on the exterior wall of the megaron in a manner that shows it was already, and remained, in existence. The S wall of the megaron increases in thickness from only 0.50 m at its E end to a maximum of 0.70 m at its W. The thin stone masonry probably did not support a second storey. As will be discussed in detail later, Stairway K that runs up the slope along the exterior wall of the house does not bond with the S wall of the megaron. It does bond with the later, heavy terrace wall between Megaron C and G. On either side of the megaron doorway were rubble piers sitting on bedrock and not on hewn bases. In 1959, what little remained of the W pier was re-cleared; probing in its stone packing produced part of a female terracotta figurine and only four painted sherds. The threshold of 15
the megaron, as found in 1886, was originally of wood and there were originally wooden doorjambs also. Tsountas tells us that the fire that destroyed the house, turning the walls into a calcined mass, had blazed most fiercely where there had once been a large amount of wood, as for example by the doors (Tsountas 1886, 75–79). In the centre of the megaron was a square hearth (L. 1.10 m, W. 0.88 m, H. 0.09 m); made from mudbrick with a straw admixture and fine asprochoma clay plaster coating. In the SW corner of the megaron, lying on the floor, Tsountas found a pithos (H. 1.32 m, Di. (rim) 0.35 m). Fragments of this jar were uncovered in 1950, and although his plan shows two vessels, Tsountas mentions only one. He does say (Tsountas 1886, 75–76) that practically all of the remainder of the megaron floor was strewn with sheets or slabs of lead, probably from vessels that had melted in the conflagration, and that preserved, stuck on the surface, were grains of wheat. This recalls the state of the floor in the West Room of the South House: ‘Towards the south the greater part of the floor was covered with a thin layer of lead. On close examination this appeared to be the remains of two or more leaden vessels, which had stood towards the south-west and north-west corners of the room. These had melted in the conflagration which destroyed the house and molten metal had run all over the floor’ (BSA 25, 92). Rooms D1 and D2 As mentioned above, it was impossible for Tsountas to discern if Room D1 was entered from the court and/or the vestibule (Tsountas 1886, 76). There is one gap in the floor plaster at the N end of D1 and another between D1 and Room B. That from Room B is clearly a break in between the plaster floors of each room and indicates a wooden threshold and probably a door, but the gap between the court and Room D1 is really a worn mark in the plaster that continues from the court into the room. The lack of dowel holes and finishing on the W face of the anta at the E end and the remains of stone and clay at its base suggest that a wall rather than a threshold originally filled the gap and that it was a later addition over the pre-existing floor. Tsountas assumed that Room D2 was entered from D1 even though his plan indicates the remains of a wall on the E side. Between D1 and D2 there was a cross wall, a fragment of which only remains, projecting from the N end of the party wall between the megaron and Room D2. There is no evidence of a threshold or doorway, however, unless it is to be reconstructed further W over the area where the terrace wall has not survived up to the floor level. D1 is a square room a little smaller than Room B and may have been an anteroom to the long and narrow room D2, or it may have been an anteroom to other areas originally built over the basements on the W. The high stone wall, some 0.70 m thick, that forms the E side of the basement corridor below and the W wall socle of the court and in part of D1, is completely denuded further S, alongside D2, and is not preserved to the floor level of the room, making it impossible to tell whether there was once a doorway from the SW corner of D1 or D2 to the W. In Tsountas’s plan, however, an unbroken stretch of wall is shown on the W of D2, perhaps suggesting that there has been much deterioration here since 1886 and that there was originally no doorway. The E wall of D2 was plastered in the same way as the basement with an undercoat of plaster mixed with chaff and a surface layer of a superior plaster without chaff apparently unpainted. The quality was better on the upper storey. As the S wall of the megaron continued W, forming the S wall of Room D2, it broadened; its maximum extant width is 0.80 m but the level of damage means it could once have been much thicker. In the 1886 plan the S wall of D2 and that of the basement are of equal width, but by 1950 too little was left of the D2 wall for accurate measurement. 16
Stairs G Stairs G lead down from Forecourt H to the basement 2.80 m below. Tsountas described thirteen steps made of small, irregular stone slabs once covered by wooden boards, of which there were still traces in 1886 (Tsountas 1886, 76; Wace and Stubbings 1962, 496; Mylonas 1983b, 141, repeats Tsountas’s description but in English). By 1950 the stairs had so seriously deteriorated that they had to be largely rebuilt,36 and it should be of no surprise that one step is now formed by a convenient slab of Hellenistic concrete floor. The 1886 plan shows that the stairs were originally in two parts, ten lower steps divided by a small landing from three upper, which run diagonally between the staircase walls and are aligned with the N wall of the forecourt, not with the rest of the house. Mylonas believed there to be two periods of use in the stairs (as well as in the megaron and elsewhere) indicated by the top three stairs he says are a later addition as evidenced by their orientation.37 For structural reasons as well, to be described more fully with Forecourt H below, the upper segment of stairs is a later addition along with the northern ends of the flanking walls on E and W. There is a heavy block at the beginning of the stairs on the E and on the W is a larger construction (tower/pier foundation). In 1950, the ‘square’ pier to the W of the stairway was found to be full of fill as well as based on fill and stone packing to a large extent. This too is connected to reorganisation and rebuilding of this part of the middle terrace and House over the earlier Lower Ramp (see below). Basement (Rooms F1–3 and Corridor E) The lower terrace was designed for storage of grain and oil and similar commodities, pithoi being found in two of the rooms. The basement has three rooms (F1, F 2, F 3) opening on to long, narrow Corridor E (9.35 m × 1.13 m) reached from Forecourt H by Stairs G. It is clear from his plan that Tsountas uncovered the full breadth of the back (W) wall of the basement at its N end only, and that the tip of its S corner and the whole of its outer face were not cleared in 1886. Nor was it cleared by the British, either in 1950 or in 1959/1960. Mylonas was the first excavator after Tsountas to investigate the extremely high levels inside the Polygonal Tower and to excavate the space outside of the House to its W, which he did in 1974 showing that a passage of unknown width (> 4.2 m) provided access from the court at the lowest level of the Cult Centre to the base of Stairway K, S of the basement of the House (Mylonas 1974, 90–91; see areas DΤ, A and B2). The basement is deep (< 2.80 m) with very thick stone walls. The ground level and the corridor slope down rapidly southwards. The floors of the corridor and all three rooms are of beaten earth only, without plaster, and bedrock is almost at floor level at some points in the corridor. The walls have an undercoat of clay plaster with chopped straw and a top layer of mud plaster alone. These basement rooms were clearly intended for storage, not habitation, and must have been very dark, unless windows, such as were found in the W and S walls of the ground floor of the Granary, existed high up in the walls and have not been preserved. No trace of thresholds or doors were found, indicating that they are unlikely, if they ever existed, to have been of any other material other than wood. The existence of the upper storey, which Tsountas had no doubt once ran over the basement, is indicated by the thickness of the main walls and by their structure of rubble and clay with traces of horizontal, vertical and transverse chases for a timber framework. This structure continues for the full height of the unusually high basement walls and would have made an 17
excellent support for an upper storey of mudbrick. In Corridor E there are preserved ‘chases’ for timber framing on the W sidewall only,38 and only between the doors of the middle room F2 and F3. Horizontal beams ran originally at a height of 0.90 m–0.95 m above the corridor floor and a second one at c. 1.93 m above floor level, very close to the top of the surviving wall but very clearly preserved as an impression in the mud plaster. Lateral or transecting beams could only be detected in two places: first running in from the first horizontal chase in the corridor and is thought to continue as the first horizontal chase in the S wall of F2 but slightly higher than the first horizontal chase in the corridor, and the second running through to the first horizontal chase in the N wall of F3. Vertical chases are visible in the corridor too; the first to the N (0.09 m wide), c. 1.25 m S of the doorway into F2, starts at the level of the lower edge of the first horizontal chase and extends up to at least the second horizontal one, while the second to the S, 0.95 m N of the doorway to F3 and c. 1.02 m from the first vertical chase, also join the two horizontal chases. There is possible evidence for horizontal chases in the S wall of the corridor in the white burned lime rubble and carbonised material but no firm indication. The most northern basement room, F1, is L-shaped, 3.72 m from N to S and a maximum width of 3.58 m from E to W. The irregular shape of the room is due to the flanking support wall on the W of the stairs and the unusual wall intersections next to the head of the stairs that is a result of various building phases. There is a drain opening in the E corridor wall at the foot of the staircase with drainage continuing along a rock-cut channel that crosses the corridor 0.80 m from the stairs before it then runs under room F1 and its S wall, along which the drain was covered with stones, to the SW and under and out of the building. This drain may originate below the basin in Court A indicating that Tsountas’s initial conclusion was correct, or it may be the continuation of the drain in Passage J that crosses under the pier in the House entrance heading S parallel to the court’s W wall. It is also possible that both drains converge under the court floor and continue W along this channel. The drain was thought to cross the S wall of F1 and the NW corner of F2 diagonally and to disappear in a southwesterly direction through the W wall of F2. Since the area on the far side of this wall has never been excavated it is impossible to trace the drain further. The middle room, F2,39 is roughly square, 3.65 m × 3.60 m. On the E half of the S wall is a horizontal chase, marking where a wooden beam was located, 1.15 m above the floor and continuing as a transect chase into the corridor, probably at the junction of the vertical and horizontal chases. The N wall exhibits two lines of ‘white carbonised material’ probably indicating two vertical chases: the first begins 0.95 m above the floor and is c. 1.05 m W of the room’s E wall, and the second is 1.18 m further W, also 0.95 m above the floor, the potential height for a horizontal chase although none can be recognised now. Tsountas found a pithos built into the floor near the NW corner of the room that was uncovered again in 1950. F3,40 the southernmost room, consists of two chambers, an inner room (2.54 m × 1.95 m) and an outer anteroom (2.54 m × 1.20 m). The S wall of the basement corridor and that of Room F3 appeared to have been the most heavily calcined of all the basement walls. The outer face of the basement wall, just below the foot of the stairs of Street K, gave similar evidence of a very fierce fire. The E half of the N wall had a chase from a horizontal beam 1.19 m above the floor (0.20 m deep, c. 0.08 m high), that continued as a transect beam through the thickness of the room’s E wall to join the horizontal chase in the corridor. On the projecting piece of wall to the W, between the inner and outer chambers, a hard smooth clay layer suggests a chase extending across the entire thickness of the wall, at least at its S end, with a certain chase on the W face of this wall (1.29 m above the floor), while the E face is hidden by plaster. The W 18
half of the N wall is too ruined to tell, while the E half of the S wall may preserve a horizontal chase at 1.04 m above the floor. The W wall of the inner part of Room F3 has unworked bedrock projecting from it into the room and is not the neat construction suggested by the plans. Traces remain to show that this rock face was once plastered. Near the S wall of the outer room of F3 was found another pithos. Following the destruction of the building, another structure was founded on the ruined basement’s W wall dating to the first building phase of LH IIIC.41 Forecourt H The forecourt has an exceptional interest and importance since the material recovered from the deep stratigraphic fill in this area, almost untouched in the excavation of 1886, allowed a reconstruction of the history of the house. It is clear that Tsountas cleared the forecourt and extreme N end of Court A only very superficially. The trapezoidal forecourt, which was defined to the N by a preserved mudbrick Wall K on its stone socle, is at the top of Stairs G to the basement. Tsountas’s plan shows no features of any kind in the forecourt. It must be concluded that he cleared only the surface, so that the features found in 1950 were not exposed in 1886. This is supported by Tsountas’s description of the floor as of trampled earth, unplastered, on a layer of common stones (Tsountas 1886, 75), which remained in 1950 only at the head of the stairs. Two fine, pebbled plaster floors were distinguished, one directly on the other. The W wall of Court A ends at the head of the stairs in a large roughly square stone block that was perhaps the base of a pier.42 The height of the block above the coarse plaster floor of the forecourt at the head of Stairs G is 0.50 m. This block, and the wall end or pier it supported, was very likely added in a later phase when the top three steps of Stairs G were added and the forecourt was restructured. The N end of the west-flanking wall of Stairs G had to be connected also with the W wall of the forecourt. In 1960 a heavy wall, M, of very large stone construction was found to underlie43 the scanty remains of the W wall of the forecourt and to run southwestwards into the N wall of basement room F1. The junction of these walls, the W wall of the stairs, the N wall of F1, and the W wall of the forecourt was concealed by a projection westwards of the W wall of the stairs to join that of the forecourt. A small rectangular space at the junction of these walls, cleared out in 1950, yielded nothing of interest but a part of a female terracotta figurine of LH IIIA date (50-194, BE 5942). At the E end of the forecourt is the rectangular rubble pier base at the entrance to Court A. The plaster floor curved up to meet it. The rock-cut drain from Passage J, covered with flat stones, was found just to its W after having passed under it. This wall or pier would have created a narrow corridor to its E that should be the entrance to Court A and the House. The thick, coarse plaster floor in between the pier and buttress l is slightly raised from that of the court, with a good plaster surface underneath extending further S. A burnt post-hole and gap in the floor at the southern end suggest a wooden doorsill and threshold. Under the floor is a fill of stone and plesia. Between the stairs and the N wall of the forecourt was a broken block of friable poros shaped like the letter ‘P’. It was probably a re-used drain slab (L. 0.445 m; W. 0.27 m; H. 0.55 m). It is in situ on the forecourt floor, since in 1959 the coarse plaster was still found running up on to its S face. It cannot have had anything to do with drainage in the position in which it was found, but might have supported something light, such as a wooden post. Its long sides are 19
roughly parallel with the top step of Stairs G and the N forecourt wall. A little E of the E face of the W wall of the court and midway between its N end and Wall K, is a small limestone block (c. 0.30 m2, H. c. 0.16 m) that might also have been a foundation for a light wooden post. Beneath the stone are two layers of floor plaster. On the S side, a thick layer of coarse plaster runs up to the block face and also runs from the NE corner up to Wall K; from the NW corner a line curves out and better plaster runs under it.44 The good plaster floors are earlier and probably equate with the floors in the House, while the thicker coarse plaster is the same as that above the drain in the narrow passage into Court A and reinforces the interpretation that the entrance to the House was restructured in its final phase, following the extension and buttressing of the upper terrace to support Area Q. Projecting from Wall K near the W end of the forecourt was a thick rectangular plaster construction (c. 0.50 m × 0.40 m, H. 0.16 m), that was the support for a large pithos of the flat-based type, found in situ in 1950. When this platform had to be removed in 1960, in the course of the investigation of the levels beneath the forecourt, it proved to be made of at least two layers of very thick coarse plaster with large pebbles embedded in its base. In the fill covering the floor at the head of the Stairs G were found in 1950 numerous fragments of plaster with black bands painted on a white ground. It is correct to suppose that they were under the trampled earth floor found by Tsountas. They possibly still lay in 1950 where they had fallen at the time of the destruction of the house. Where they fell from is unknown. In 1960, further clearance of the forecourt found two layers of thin hard plaster of high quality, markedly superior to the remains of the thicker, softer, pebbled plaster of the higher floor.45 The two surfaces follow closely upon each other and it is unlikely that more than a short interval of time separated them. The thickness of both and of the trampled earth between was only about 0.13 m at the E end of the forecourt cutting, though the depth of earth was slightly greater towards the W indicating a sloping surface from E to W. These two high quality floors correspond in number and composition with the floors of the Lower Ramp and there is little doubt that they are contemporary. WALLS J, K AND M Excavation in 1960 revealed Wall J,46 a massive substructure to Wall M and a stone socle under the mudbrick of Wall K. Wall J is unique in the area; it is built entirely of quite small stones with neither an admixture of larger stones nor clay packing. It is a terrace wall, with a breadth of 1.00 m. The ground, on which it is built, slopes steeply down from E to W and more gently from N to S. It is structurally the oldest terrace wall at this level and was built against the slope to create the terrace itself and the foundations for the Megaron and the N end of the forecourt. To the E, upslope side of Wall J was found a massive bedding of large stones without clay packing that is the core of the building terrace for the structures above. To the S of Wall K the width of the bedding is about 0.90 m at most. The bedding begins about 0.40 m below the forecourt and about 0.94 m above the sloping bedrock. This stone fill was put in place after the construction of Wall M to close the gap between Wall J and the House. The bedding under the forecourt is later than the corresponding bedding behind Wall J, but how much later is not certain. Wall J, in its existing state, ends where the stone socle of Wall K abuts its E face, with seepage of earth between the two on the N side. Wall M continues the line of Wall J southwards, but its W face is c. 0.10 m further W than that of Wall J. Wall M differs from the other walls it abuts at its N end in that it is of a pseudo-Cyclopean construction with particularly large boulders along 20
the western limit of the forecourt. Similar in construction are the walls that converge on the S end of Wall M, at the NE corner of the pier W of the stair head and N of basement room F1, and the S wall of a room on the lowest level immediately W of the forecourt.47 All three walls and the addition to Stairs G are in fact contemporary. The substantial patch joining Wall J and Wall M is of the same small stones as Wall J. Wall M is c. 0.80 m broad, narrower by 0.20 m, than Wall J. The top of Wall M at the SW corner of the forecourt is 0.10 m above the upper of the good plaster floors and 0.14 m below the coarse plaster floor. Wall K was still topped with an unknown height of mudbrick in 1886. By 1950 no more than the base survived and had flattened out from erosion. A deep trench cut in the forecourt in 1960 showed that the stone socle of Wall K was some 0.70 m wide and 0.57 m high. As already mentioned, its W end is built up against the E face of Wall J and the patch at the junction of Wall J and Wall M. Wall K defines the N limit of Forecourt H of the House, but by design and structurally it belongs to the foundation of the Megaron to the N of the House. It was built, like both the Megaron and the House, in the early part of LH IIIB. It became part of the forecourt very likely during that same period but was already standing when the later floors were put down over the bedding of stone and the reorganisation of the House entrance was undertaken later in LH IIIB. WALL N AND OFFSET O Another wall, N, was found below Wall K, projecting beyond its S face. Wall K, though built above Wall N, does not follow it exactly. The projection of Wall N to the S is at its broadest (0.50 m) on the W, where it curves southwards into the stone bedding, of which it appears to be a continuation. At its narrowest, so far as it was excavated on the E, it projects only 0.30 m. Since it is not visible on the N side of Wall K, its maximum breadth must be less than 0.80 m. For a little over 1.00 m E of the point at which it runs into the stone bedding its top is level with that of the bedding. The remaining 1.00 m cleared includes a course or two of fairly small stones. Here the base of the stone socle of Wall K rests directly on the top of Wall N, while to the W there is a fill some 0.20 m deep between Wall K and the top of Wall N or the bedding, meaning that it descends to the W and is part of an earlier configuration contemporary with or earlier than the stone fill. Wall N has an offset, O, starting some 0.44 m below its top. O is not more than 0.20 m broad and some 0.50 m high; its W end runs up to the stone bedding. The depth of bedrock in the cutting ranged from 0.83 m to 1.08 m below the top of O. The greatest depth of bedrock here has an elevation of 236.20 m as compared to 236.07 m for the lowest tread of Stairs G. In a deep cutting to the N of Wall K the greatest depth of bedrock had an elevation of 236.72 m. These figures show a pronounced slope down of the rock from N to S and the great quantity of fill required to bring this area to a level at which it could be used as part of the House and the neighbouring structure. The filling in of the area, later used as the forecourt, began much earlier than the building of the House. There are the two thin, hard plaster floors that descend to the W and are associated with the Lower Ramp access to the courtyard area at the base of the slope (Fig. 10),48 and there are traces of at least two earlier layers, perhaps the remains of surfaces, which were found lower in the deep cutting. Neither extended over more that a part, and both were largely destroyed by the later building. One layer was a bed of plesia with patches of asprochoma at about 0.20 m below the top of the highest part of Wall N; the other a layer of thick plesia some 0.40 m lower at 0.20 m below the offset O and above earth fill with a considerable quantity of large rubble going down to bedrock. The bulk of the sherds from under these plesia layers and down to 21
bedrock were of MH date, but the rubble fill also included pieces dating to LH IIIA49 and may likely indicate the period in which the filling for a ramp was begun. THE FORECOURT DEPOSIT (Pl. 3) In the 0.20 m of fill over the top of Wall N and under the base of the stone socle of Wall K, at the W end of the forecourt up to the E face of Wall M c. 2.00 m westward, there lay on Wall N and the stone bedding a number of broken pots, or large pieces of pots, a quantity of animal bones and teeth, and other objects suggesting household rubbish. To the N, this deposit ran into the earth under Wall K, although not beyond it. To the S, the deposit continued over a square patch of asprochoma, 0.05 m thick, and into the S face of the cutting. A piece of one of the pots of the deposit, animal bones and quantities of sherds also continued into this S face. They were clearly part of the deposit, the southern limit of which was not reached in 1960. In the metre the cutting extended to the E of the asprochoma patch, bones and large pieces of pots were absent; but at the depth of the top of the higher part of Wall N, plesia, carbon flecks, pieces of asprochoma and plaster and patches of very burnt earth appeared. The fill here was hard-packed and, some 0.24 m lower, was found the bed of plesia with asprochoma patches already mentioned as an early surface.
THE FINDS THE HOUSE In general, very few finds can be associated with the House, primarily because of the complete clearance by Tsountas in 1886. Although he reports on the architecture and a few large jars built into several of the rooms, he does not mention small finds at all. He does say that many vase pieces were found but few whole, none however, are pictured or described (Tsountas 1886, 78). All of the finds date to LH III at least. A significant number of diagnostic and interesting sherds were collected in 1950 from a ‘general cleaning’ of Tsountas’s dump to the N of his excavation. Many of these sherds certainly come from the House but have no context whatsoever.50 The reclearing of the House in 1950 resulted in a few stray finds, mostly from walls and piers, the exception being the forecourt where exploration continued in 1959 and 1960 to depths below the House and recovered dating evidence for several phases of use in the area as well as a deposit of rubbish that included a number of vases and small finds. Although the House was buried under its own debris and one might therefore have hoped to recover much from the rooms, extremely little was found in the house or its neighbourhood.51 The finds from Forecourt H are described here in some detail while the finds from other areas of the House, which are for the most part of secondary nature, are catalogued in the Materials Lists and Object Catalogue (Supplementary PDF). Forecourt H Pottery The condition and source A deposit of rubbish, consisting of pottery, animal bone and other objects, was dumped with soil fill and very little stone over Walls N and O and the early levels of the access ramp to 22
PLATE 3
(a)
60-584 60-586
60-589 60-585
60-588
(b)
The Forecourt deposit: (a) pottery, bones and teeth in situ; (b) terracotta figurines (60-584, 60-585, 60-589), reworked pot scraper (60-586) and clay counter (60-588), all scale 1:1.
23
the Lower Courtyard. In the fill below the plaster floors of the forecourt, almost complete pots, or large pieces of pots, lay chiefly on, or only just above the stone bedding, to a depth of 0.15 m–0.25 m. The stone bedding was used to fill the triangular section of space between the closing Wall M at the W end of the forecourt and the earlier surfaces and their support walls to the E. This fill was put in at one time, when the forecourt was to be built. Throughout the deposit were found small pieces and larger patches of carbon, small bones of terrestrial and marine fauna, terracotta objects and fragmentary stone tools, together with the pottery. This dump of domestic debris was the final stage in a filling of the forecourt area.52 The source of the rubbish is not certain and it could have been scooped up and dumped in the terrace from somewhere in the neighbourhood, however, it is possible that it originated from the earliest stage of the House, perhaps already constructed in part without the forecourt, or from the use of the earlier Shrine on the upper terrace. On bedrock was a packing fill of exclusively pre-Mycenaean sherds indicating some level of early occupation on the slope. Succeeding packing fills are associated with attempts at levelling for potential surfaces of clay in periods before the construction of the ramp. Pottery types and dates It may be assumed that a forecourt connecting the court and the basement stairs was an integral part of the house from its beginning and that the date of the earliest forecourt floor is also that of the building of the rest of the house. However, the orientation and structural execution of the forecourt indicates that it may have been completed after both the House to the S and the Megaron to the N were already built, at least in part, and that the area under the forecourt was enclosed, filled and a raised floor level achieved afterwards, along with structural additions to the N end of Court A and the top of Stairs G. The forecourt deposit, under the floors, gives the date after which the forecourt floor was laid, a terminus post quem for the forecourt of LH IIIA2. Only one of the pots, a wide-mouthed piriform jar (FS 39) decorated in a deep zone with two parallel rows of plump foliate band decoration (FM 64), can be dated closely to LH IIIA2 (60‑395, BE 17027). The rest are all undecorated fineware vessels, and date more generally to LH IIIA, although most certainly will be IIIA2. With the exception of some 23 MH examples, the datable painted sherds closely associated with these pots appear to belong entirely to LH IIIA, apart from one with an early panel pattern (not on a deep bowl), which might be LH IIIA/B. Unusual for LH IIIA2 is the small number of monochrome pieces, albeit in good red lustrous paint in and out. Linear decoration occurred more frequently. One example is a small conical cup (FS 206) with fine lines all over, very much like the example found in 1950 at the W end of Wall K, also dating to LH IIIA. Another is from the base of a kalathos (FS 300/301). Considering the relatively small area and depth of the deposit, there was a fair amount of patterned sherds. This included none later than LH IIIA, and very little obviously earlier.53 The only patterned sherd close to the pots was a kylix (FS 257) decorated with an octopus (FM 21). The unpainted sherds included a fair quantity of good polished ware as well as typical examples with a more roughly smoothed surface, such as pieces of carinated kylikes (FS 267) like the two complete examples with a straight upper bowl profile and a rounded rim (60‑391, BE 17016 and 60‑393, BE 17023) and shallow angular bowls (FS 295) in addition to two whole vases of the same type (60‑392, BE 17015) and (60‑398, BE 17019). The profile and lightly polished surface treatment would indicate early examples, certainly by the later part of LH IIIA2. Examples of 24
the conical cup (FS 204), very common in LH IIIA2 contexts, have a slightly incurving rim; three examples were complete (60‑394, BE 17026; 60‑396, BE 17014; and 60‑397, BE 17024). There are only a few pieces of heavier wares including cooking ware (tripod vessel and scoop with supporting handle), semi-coarse, and coarse (pithos fragment with impressed decoration). Other pottery units were excavated below the deposit level in association with several early surfaces and in packing down to bedrock. Limited fill under the ramp surface was recovered and contained mostly MH sherds with only a very small number of plain LH III examples — a few likely also of LH IIIA date and indicating the construction of the ramped access from the Lower Court to the Shrine terrace during this initial use of the area.54 The sherds in the level from the asprochoma patch down to just above the depth of the top of the Offset O differ markedly. The proportion of MH is much higher; the LH has very few kylikes and very little painted ware. Monochrome is confined to one good red kylix stem and one goblet with a dark brown interior. Deeper still, the sherds in the earth underlying the plesia layer with asprochoma patches included very few that were certainly LH: the greater number that could be dated were MH. On the other hand, the relatively few sherds in the earth with large pieces of rubble at a depth some 0.20 m below the top of O were certainly not pure MH. At least two, including one decorated with what was probably an octopus in lustrous red paint with added white, appeared to be as late as LH IIIA. Many of the sherds in this layer were extremely worn and battered and some may actually be associated with the stone bedding fill that was packed behind, to the E of Wall M. Pure MH is found only below the level of the bottom of Offset O, or in a maximum depth of 0.58 m of fill. Wall K In 1950, in the remains of the mudbrick of Wall K, at its W end, was found more than half of a small linear conical cup (FS 206, 50‑314, BE 7122; Pl. 7b). It is decorated in and out with parallel narrow bands and dates to the LH IIIA period, perhaps even to the first half of it. Sherd evidence, however, from both the remains of Wall K and the fill under its foundation suggest that it was constructed at the very end of the LH IIIA2 period or more likely, the early part of LH IIIB.55 Also in 1950, the remains of a pithos were uncovered in the NE corner of the forecourt on a square patch of thick course plaster set on a foundation of pebbles. No other information is available for this vessel or its current location. It was not in situ in 1959. Other Materials The number of small finds other than pottery is unusually large for the Tsountas House area, which actually produced so little. This is of course due primarily to the complete coverage of the earlier 1886 excavation in most of the structure and related areas. It is certainly, though, not the only explanation. Even with a larger number of finds, only one object (animal figurine 60‑589, BE 10179) can be dated independently of the sherds in the deposit. Stone A single steatite conulus (60‑199, BE 10141) of the shanked type was found at the juncture/ patch of Walls J and M. An obsidian flake (60‑581, BE 10185) was found in the deposit over Wall N together with a jawbone, two bivalve shells, and kylix 60‑391. From a deeper deposit came a chert blade (60‑582, BE 10119) with a saw edge. 25
A stone object (60‑583, BE 10173) found just under the forecourt floors is particularly interesting. It is in the shape of a Triangular Pyramid (tetrahedron) with two opposite flattened edges. Each triangular face measures 0.03 m × 0.03 m × 0.03 m. Its exact use is unknown, however, a groove on one side suggests the object may have been suspended as a weight. Metal Two items of metal were found in the forecourt. The first is a fragmentary piece of gold jewellery (60‑587, BE 10100), probably a pendant. A strip of thin sheet gold with raised soldered edges, it has four cloisons for inlay (now empty) and is pierced at one pointed end. The second object is a fragmentary lead clamp (60‑590, BE 10171) with two small prongs for pot mending. Terracotta A single object was found between the two floors of the forecourt — the head of a female figurine (50‑239, BE 7088) wearing a polos. Although fragmentary, it is likely to be of the Tau type because of the mass at the break and therefore, dates to the LH IIIB period. Another Tau type female figurine (59‑506, BE 9725), in this case a miniature example, was found in the forecourt, in the mudbrick at the W end of Wall K. It too probably dates to LH IIIB. From the rubbish deposit under the floors of the forecourt come three fragmentary figurines (Pl. 3b): the body and hindquarters of a Spine 1 type animal figurine (60‑584, BE 10222); the leg only of a Wavy 1 type animal figurine (60‑589, BE 10179) dating to LH IIIA and found in direct association with the pottery of the deposit, and a ‘Toreador’ figure (60‑585, BE 10118). This last example is an unusual type of presumably male figurine that would have been applied to the back of the neck of a bull figurine. The forecourt Toreador is only a fragment of the original combination type figurine — only the head, arms and upper torso are preserved. The face is pinched with a flattened head (almost a polos) and body; the arms are spread to grasp the horns of the bull (now missing). The figurine is decorated on all sides and preserves application marks on its front and inside the arm. The context suggests a date of LH IIIA2. A single example of a clay counter (60‑588, BE 10170; Pl. 3b) was found in the forecourt deposit and is similar to the example found in 1959 between the upper and lower floors of the Shrine (59‑505, BE 9858; Pl. 10a) and the 156 such discs, found in 1960, from debris of the destruction at the end of LH IIIB to the W of Wall J (60‑194).56 An amphora handle (60‑586, BE 10163; Pl. 3b) was reworked and reused as a scraper, one end sharpened to a chisel shape. Dating probably to the MH period was a ceramic spindle whorl (60‑591, BE 10221) fashioned from a coarse ware sherd with a central hole pierced for suspension. It was found in fill well below Wall O and almost on bedrock. Plaster Two fragments of painted plaster were discovered among the fill in the deposit below the forecourt floors (F152, BE 24283). The larger piece is decorated with alternating red and blue horizontal zones; the smaller has a blue curved band on a white background. Organics Throughout the forecourt deposit were found small flecks and occasional large patches of carbon, quantities of small bones, including two fish vertebrae, but otherwise mainly from pig, 26
sheep, goat and a type of fowl or bird; an animal jawbone complete with teeth (found with kylix 60‑391); horn cores, one bovine and two probably goat; shells of molluscs.57
SUMMARY What little evidence there is from the House supports that from the forecourt in suggesting a building date early in LH IIIB. There are at least two phases of construction exhibited in several areas of the building both in design and in the details. The slight additional evidence from the period of use does not run counter to this. The two good floors of the forecourt, though apparently laid within a short time of each other and the two similar floors in the ground floor of the House suggest a fair period of use in its original form, before alterations were made perhaps in the first half of LH IIIB but also certainly in the later part of the period as well. There was no real trace of the earthquake destruction at the end of Phase VII, witnessed throughout the Cult Centre. No part of the structure seems to suffer substantially enough to require rebuilding or major repairs. The cutting off of access to the lower courtyard through the restructuring over the western end of the Lower Ramp and the construction of the forecourt seems to take place much earlier than mid LH IIIB, since it would be very difficult to explain a total lack of LH IIIB sherds from under the floors and in the terracing. Subsequent alterations to the N end of the court, resulting from the extension of the upper terrace for the building of Area Q, may have been after the middle of LH IIIB, while the necessity to buttress the upper terrace is very likely a result of the earthquake. So far as the evidence goes, the fairly long life of the house, after the early LH IIIB alterations, was uneventful until the great disaster that overtook the entire neighbourhood at the end of Phase VIII late in LH IIIB2. From no part of the house can pure destruction level pottery be produced since, with the possible exception of that of the forecourt, the surface in use at the time of the destruction was cleared by Tsountas. Nevertheless, the sherds recovered from the 1950 general reclearing of the House, while including a little later and much earlier pottery, include also a great quantity attributable to the end of LH IIIB. Very little may actually have been originally found in the House, besides pottery, when it was first excavated, as was the case across the Citadel House Area. There is the real possibility that the House had declined in use before its destruction, together with the structures that surrounded it. This might be evidenced in the lesser quality plaster of the upper floor in comparison with the better made and higher quality plaster of the earlier floor, especially in the narrow entrance during the final phase. The possible use of the megaron for grain storage when ample basement space seems to have been available is also odd. There remains the possibility that the building was abandoned before it was destroyed and therefore only immovable objects were left behind, or that it was accessible to some extent after the destruction and was cleared of recoverable objects.
USE OF THE HOUSE The simplest and most straightforward interpretation of the purpose of the House is habitation. It may be a domestic structure, albeit one of high status, that just happens to have been built in close proximately to a cluster of buildings that were used primarily, if not exclusively, for cult. It may have been the first in a whole series of large, well-built houses in the neighbourhood that grew up along the lower SW slope of the citadel, especially after the construction of the WCW, after the middle of LH IIIB, that enclosed this area of the hill safely and exclusively within 27
the citadel. This House was built, however, prior to the construction of the WCW and well before the other houses were built further to the S. Another important difference from the other structures further S is the orientation of the House. It was built parallel to and in physically close proximity to the pre-existing Shrine on the terrace above it and the connection may be significant. Like the Shrine, the House is orientated to the N and was entered exclusively from that side. They frame the S end of the complex of buildings we call the ‘Cult Centre,’ just as the other three structures frame the N side of the area. The remainder of the houses built beyond the stairway (K) to the S of the House are all orientated in various directions towards pathways, lanes, and other staircases providing access along and up and down the slope. Access is another issue in the function of the House (Fig. 8). When it was first constructed in LH IIIB it was potentially accessible both from below and above by way of the Lower Ramp but access would always have been somewhat restricted by the position of the Megaron just to its N. The S side of the building was always more accessible and the lack of an entrance here must be significant. In the later phase of use, access to the House became increasingly restricted and this may indicate that it was no longer a destination for significant or important traffic. Yet, the House is an important structure within the context of the Centre and should have continued in its relationship to the other structures and their various functions through most of its lifespan. So, the House may be interpreted as ‘religious’ by association but little in the physical makeup of the building, its features, or moveable artifacts would suggest that the building was used for cult. The one exception being the trefoil basin in the court floor that may have been used for some ritual action rather than as simply a drain. Everything about the structure suggests it is more than simply a house. It was suggested by Wace to be the residence of the ‘priest-in-charge’ due primarily to his understanding of the Shrine and House as a single structure and therefore with an intimately connected function. Although the House should be considered separate in physical structure and function from the Shrine, the interpretation may be a valid one. The building may still be considered domestic and yet ‘religious’ if it is the habitation for the official residents of the Cult Centre (Iakovidis 1991, 1044; 1997, 154 and 158–59). It is unlikely that a single individual resided in the building but it may have housed a number of priests, more likely priestesses,58 or a combination of the two. There was ample living and possibly administrative space, but the amount of basement storage seems generous for the immediate needs of what must have been an exclusive and limited group of people. Mylonas suggested that the extent and nature of the basement storage of this building was one indication that it was used as a ‘treasury’ of the Cult Centre.59 None of these rich offerings or possessions of the Centre were found here however; whereas caches of valuable artifacts and materials have been found in other Cult Centre buildings, such as the Temple deposits in the Alcove and Room 19, and in the Room with the Fresco complex, especially Room 32, not to mention Room G, also excavated by Tsountas, some in secondary contexts to be sure, but indicating the different character of the House remains. Something else about the House that differs somewhat from the other buildings and the other deposits of pottery and artifacts in non-ceramic materials, found either collected and disposed of in a systematic way or left relatively undisturbed under the ruins of the destroyed buildings, is the noticeable lack of finds. We cannot blame this entirely on Tsountas I believe. If he had found interesting and important finds he would have mentioned them in his account of the House, which he certainly felt to be an important and noteworthy structure. He mentions in details the finds from Room G for two reasons: because they were remarkable in number and material and because they were 28
the only really significant finds from his excavation in this area.60 There may well have been large quantitites of pottery found in the House, especially in its basements, but Tsountas was not particularly interested in pottery, especially this early in his career, and he would have kept only unbroken vases, very rare in an earthquake destruction, and probably only decorated examples, something not necessarily expected in great numbers from a domestic context, even one on the citadel. Tsountas does mention pottery though as the primary artifact type from his excavation, generally remarking on the ‘innumerable vase pieces,’ however, ‘whole almost none were found’ (Tsountas 1886, 78). As suggested above, the building may have been abandoned before it was destroyed so few movable objects were left behind, or its purpose may not even have included that many objects in the first place. Alternatively, after the destruction the ruins may have been accessed and recoverable objects removed. Although a possibility in light of the clean up witnessed in the Cult Centre after the mid LH IIIB destruction, it would be remarkable for the final destruction at the end of LH IIIB when it seems that other structures were left in ruins with their relatively few contents undisturbed. The difference in post destruction attitude towards the remains would also signal a different conceptual understanding and function of the House. It is a general characteristic of all these buildings destroyed at the end of LH IIIB, however, that very little was actually found in use from their final phase (VIII) and marks a significant change from the quantity of finds in use and damaged during the earlier destruction event in mid LH IIIB. Another possibility for the interpretation of the use of the House is for a type of hall or ‘megaron’ that could be used for the congregation of visitors to and participants in the celebration of rituals or festivals in the Cult Centre complex. The variation of structures, symbols, and rituals understood from the many buildings and features along the slope suggest a hodgepodge of cult needs and a scattered schedule of ritual alongside more regular events. In this way the domestic characteristics of the building can be understood in the context of a complex of cult buildings and ritual. It would also explain the size of the structure able to accommodate a number of individuals comfortably but yet still be exclusive for those privileged enough to gain access first to the Cult Centre area itself, then to the centre of the complex. The storage facilities also indicate that they were meant to provide for a significant number and perhaps status of individuals. The Megaron on the same general level of the slope but rising higher to the N could have been something similar with a megaron suite on its upper floor complete with rectangular hearth and basement storage areas. That feasting is an important corporate and potentially ritualised event in the Mycenaean culture is increasingly recognised among scholars (Wright 2004 and Hitchcock et al. 2008), even if the location of the gatherings is not always known. The Tsountas House may have been such a place, used as necessary for a variety of domestic and religious functions. The nature of the material from the forecourt deposit, perhaps rubbish originating from the early stages of this building or from the area before it was built, suggests a degree of specialised consumption, perhaps even feasting with a preponderance of drinking and serving vessels, several of them of high quality, mixed with the remains of the meal.
29
PLATE 4
(a)
(b) (a) Tsountas House complex, general view of lower Shrine, ashlar altar and House. From N; (b) Room G and Shrine, with the House on terrace below. From SE.
30
PART 2. THE TSOUNTAS HOUSE SHRINE (G) References: Tsountas 1886, 78–79 and pl. 1; 1887, 162–64, 169–72; ILN 1950, 1041, figs. 4–6 (1042); 1951, 254–55; BSA 51 (1956) 122; 1962, 395; Wace Guide, 35; ILN 1961, 490–92; WBM 1, 9–10, 14–15, 19–20; Mylonas 1971, 154–56; CC, 21–26; 1972b, 116–21; 1977b, 19, 21, 41, and 92; 1983a, 308, 310–15; 1983b, 133–40; Iakovidis 1983, 45; 1986, 243–44; 1996, 1044; 1997, 150–51 and 154; 2004, 14–15, 17–21, 23–24.
TERMINOLOGY The lone structure on the upper terrace of the Cult Centre had an equally complicated archaeological history as it did lifespan. Excavated, at least in part, on three separate occasions, by three different excavators, over almost a century, with intermittent cleaning and clearing for study and photography. The nomenclature for the structure and its component parts is therefore also complicated. Tsountas labelled the southern room G on the 1886 plan (Fig. 3), and it retained that designation from then on. In 1950, the main room to the N of G was called Room I and the area/room further N, with the ashlar feature was called Room II. In 1959 (Fig. 2), the different stages of construction were distinguished from one another by designating the later version of Room I with a higher floor as Room P, and Room II as Room Q (here referred to as Area Q). Room I was usually referred to simply as the Shrine with Room G to the S. In the 1970s, the Greeks (Figs. 4 and 5) labelled the main room G1, and its later version with higher floor was called G2.61 The area further N was referred to simply as the ‘rectangular altar’. In this publication, I refer to the structure itself as the Shrine or Shrine G and refer to its constructional phases, and lower and upper floors as appropriate. Room G remains the same.
DESCRIPTION AND EXCAVATION THE SHRINE After gaining access to the Cult Centre, one descended the long plastered ramps of the Sacred Way (Fig. 11) to arrive in front of a two-room structure at the S end of the upper building terrace, a Shrine (L. 13.4 m) (Fig. 5 and Pl. 4). Its function has been identified by the features associated with it, in at least two different periods or phases: a plaster feature at the back of the main, long room (Pl. 6a), and a construction with an ashlar stone base to the left of the entrance (Pl. 8c–d). Both are identified as altars. The lifespan of the building presents several phases of construction and use over a period of perhaps 150 years. The inner, southern room, G, was the oldest component of the structure and the first to be excavated; its finds reinforce the importance of the place and its potential ritual nature. As may be seen in the plan of 1886 (Fig. 3), Tsountas excavated the terrace above the House, where he uncovered Room G, a room adjoining it on the N, and an area outside these rooms to the E and N. He did not fully publish his work, only reporting briefly on Room G and the finds he made there (Tsountas 1886, 79; 1887, 162–64, 169–72). Tsountas described G as an area (L. 4.25 m, W. 2.85 m) that did not belong to the House that he had excavated on the lower terrace. He indicates that the floor of the room was earth and 31
Fig. 5. Plan of Shrine G (E. Olympios; after Mylonas 1983b, fig. 103).
not constructed, and since finds continued below that level, he dug the fill in the room down to the underlying bedrock. A carved wing-shaped ivory fragment, for example, was found almost on bedrock that was not levelled and rises high on the E side of the room where Tsountas distinguished the lines of two parallel walls. He believed the defined space had been used for storage, not habitation, and he describes many of the interesting and valuable objects he found: beads of glass, stone, and amber, an Egyptian scarab of Queen Tiy, and several appliqués in ivory or gold leaf (Tsountas 1886, 79; 1887, 169–72, pls. 10 and 13). Most significant was the painted plaster plaque that seems to depict the worship of a goddess with a body made of, or standing behind, a large figure-of-eight shield (Fig. 6). Tsountas’s plan shows that he cleared some of the room to the N, Area Q, and the plastered court and ramp still further to the N, at least exposing the tops of walls and the ashlar blocks in Q in order for them to be planned. Much of the excavated area to the N of the Shrine building was re-covered with a dump (see Pls. 2a, 5). In 1950, Wace re-cleared the greater part of the area covered by Tsountas’s plan and dug deeper into the room N of G, below what turned out to be an upper floor, revealing the features of the Shrine (Pls. 4, 6).62 In 1959, work in the Shrine was confined to cleaning for re-examination and measurements; and to a small cutting in the strip of the upper floor left at the N end of the room in 1950. At that time the enormous mass of stones and earth dumped by Tsountas upon the northern part of his excavations was removed, and excavation was carried beyond Tsountas’s northern limit. Mylonas re-cleared the Shrine in 1971 and 1972 to conduct a full study of all the remains, test the stratigraphy, and produce the first stone by stone plan of the structure (Figs. 4 and 5).63 He went on to include his observations and interpretation of the structure and its unique features in published preliminary reports, public lectures, and popular books.64 32
PLATE 5
ated
cav nex
u
Passage S–N
G (a)
Ramp surfaces: 1 2 3
(b) (a) 1950: Room G and Shrine looking towards Tsountas’s dump, Passage S–N to entrance. From SE; (b) 1960: Tsountas House area, ramp and Shrine, with South House Annex excavations in background. From S.
33
Room G The southern component of the Shrine is Room G, which is a small irregular rectangular structure (4.60 × 3.20 m) surrounded by a very heavy platform which supports it, with earlier walls of fairly small stones and a good deal of mud. Later, G’s walls were enclosed and incorporated into heavier walls that also define the main shrine room to the N, on the W side creating a strong terrace and support wall.65 G becomes an adyton. The S wall of the N room is parallel to and abuts on the original N wall of G where the entrance was located. The threshold was originally wooden and covered in plaster, both had burned substantially. The E and S walls are founded on exposed sloping bedrock that rises high enough in the SE corner so that it would not have been covered by a floor that should have been just below the level of the threshold. As mentioned above, Tsountas had dug through the floor that was only of earth because finds continued in the fill over most of the low-lying bedrock in the W and NW parts of the room. Lying almost on bedrock was the ivory wing. The S wall is made of two parallel walls: an earlier, interior wall of small stones, built partially on rock and partially on an oblique sub-foundation across the SW corner that fills the greater depth to bedrock on the W, and a much larger (W. 0.72 m) and heavier wall of several pseudo-Cyclopean blocks and smaller stones. The W wall, also of small stones and a little offline from the rest, was incorporated into the later, larger stoned, W wall and the W end of the N wall was extended to reach it. The outer E wall was built at the same time as the S wall of the northern shrine room and later than G’s S wall on which it steps. Much later, the terrace to the W and S was reinforced in a pseudo-Cyclopean manner, creating a substantial platform, possibly for a tower, through which ran the central drain (K), and impeding on the E wall of the House below and extending to the point where an offset is visible, just W of the N wall of G. The SW terrace corner also creates a wall line close to that of the House for the continuation further upslope of stairway (K) that seems to bond with the terrace extension while abutting on the House wall to the W. In total there may be as many as four different construction phases in and around Room G. The finds made by Tsountas, and their relation to an earthen floor, indicate three phases: in the fill over bedrock under the floor, on the floor itself, and in the fill over the floor. The location of Room G on the slope and in relation to the surrounding structures and features, indicate that it is the oldest structure in the area. The nature of its construction on sloping bedrock and the style of its masonry using small stones without clay packing also indicate a very early date. Actual dating evidence is derived from the room to its N, which on structural grounds is nearly contemporary to or later than Room G and indicates a date of LH IIIA, and very likely to the early part of that period (Mylonas CC, 24; 1972b, 118; 1977b, 19; 1983a, 311; 1983b, 134). The finds from Room G, both by Tsountas and the British excavations in an unexcavated strip along the N wall, were bits and pieces of fine craft items, many fragmentary that could have been votives or votive debris. Not knowing the find spots of the objects makes their original use difficult to assess fully. Their discovery generally either in fill below an earth surface or in fill above that surface, a few possibly on the surface itself, could signal breakage and disposal during times of rebuilding, in any of the last three phases, or like several other fill deposits in the Cult Centre (i.e. Megaron basements), this material may simply have been tipped in with a good deal of fill after the mid LH IIIB destruction. In either case, the interpretation of Room G either as storage for the Shrine, or as an adyton cult room, like Room 34
(a)
(b)
Fig. 6. Plaster plaque from Room G: (a) after Tsountas 1887, pl. 10.2; (b) drawing by M. Reid; after S. Immerwahr, Aegean Painting in the Bronze Age, pl. 63 (1990).
35
32 in the Room of the Fresco complex, is impossible to evaluate for certain at this point. The painted plaster plaque (Fig. 6), however, could have been an object of veneration rather than a votive and could lend more weight to the inner room being an adyton, and of course, the longterm use and maintenance of this architecturally oldest component of the Shrine also speaks to its importance. The Shrine — Lower Floor The northern room of the Shrine was not entirely explored by Tsountas and only a feature of the upper, later floor is shown on the 1886 plan (Fig. 3). He may only have followed the walls, clearing their top to put on the plan. It is clear that much of the W, N and S walls were missing or below the level he reached. It was not until 1950 that the room was fully cleared and that through missing segments of the upper floor, especially towards the centre to S parts of the room, was discovered a lower, earlier floor and several features on or in it (Fig. 1). The original dimensions of the northern room of the Shrine are not able to be accurately reconstructed because of the damage to the floor, made of a thin layer of plaster with a few pebbles on a layer of tamped earth, and especially to the walls from erosion, such as the majority of the W wall, remains of which exist only at the lowest terrace level, and the disturbance of a Hellenistic cistern sunk into the W terrace and the E wall of the House below to the W. The estimated length of the lower floor is just under 6.0 m from the S wall to the preserved edge of the plaster floor at the N end. Beyond this point, for 0.40 m, is plesia packing that may be the foundation for a N wall although no actual traces remain. The length of the room would be about 6.2 m. The plesia could also have been the sealing layer for a wooden threshold at the entrance, but again no direct evidence was found. The exact width of the lower, earlier Shrine also no longer can be determined. The plaster floor was preserved from a low outcropping of rock, with a height of 0.45 m and a depth of 1.0 m W of the later preserved wall along the E side, to a maximum distance of about 3.5 m to the W where the floor edge is missing and roughly preserved. The original W wall may have existed at about that point, in line with the earliest phase of the W wall of G. Following on that reconstruction, an earlier E wall may have been built in line with the earliest E wall of G, and would have been built on the low rock outcropping at the edge of the plaster floor, on which was found plesia packing over a hammered flat surface.66 No evidence of masonry was preserved from either of the supposed early flanking walls of the N room, however. The rubble walls that are preserved, on the E (H. (pres.) 0.70 m–0.80 m, W. 0.90 m) and S at least, are from a later, second phase of construction, equivalent to the later phase of the E wall of Room G, with which they bond. In the earliest phase of use, there would probably have been only a single party wall between the two rooms, G’s N wall, and the plaster remains on its outer face is a further indication of that. Abutting against this in the second phase is the S wall of the N room, where in 1950 a layer of burning was recognised but not explored.67 In that same spot the Greek excavation revealed a threshold (Fig. 5), originally covered with a wood plank, opening inline with the earlier threshold to Room G (MPW 1950 Notebook (Mycenae Archive x040), 17; Mylonas CC, 24–25; 1972b, 119; 1983a, 313 and 1983b, 136–37; Iakovidis 1997, 151). The entrance to the earlier Shrine is not certain but the architecture and the features suggest it was at the N end, as was also the case in the slightly larger, later Shrine. A series of flat stones below the floor level of the later Area Q and ramp, may be a step onto the threshold, marked by a long stone a little to the W and possibly the W anta of the Shrine’s entrance. 36
The reason Wace identified this room as a Shrine is the remarkable feature on the lower floor near the S end of the room (Pl. 6a). This was described by Wace as ‘a large, curiously-shaped altar-hearth of stucco with, on the W, a stand for a ritual vessel with a pointed base (rhyton) and a runlet draining into a two-handled jar set in the stucco floor (Pl. 7a).68 Beside this, a part of the altar-hearth rises like a bolster, the object of which is uncertain. Close to the bolster was found a large dish or plate with a flat base and low vertical sides, probably a vessel for carrying offerings’ (Pl. 7a) (ILN 1950, 1041, figs. 5–6). Although it is all too easy to ascribe to religion, objects and features for which interpretation is difficult, this plaster construction seems likely to be a candidate to be considered sacred and used in ritual. The fine, thin plaster surface is too delicate to have had any frequent domestic use. No exact parallel to it has so far been found in Greece or abroad. Located c. 0.80 m from the S wall of the room, the wide tongue-shaped69 feature (a) is 1.20– 1.25 m long and 1.15–1.25 m wide, and is slightly concave on the centre top surface with a raised edge 0.02 m higher and 0.19 m above floor level. Three sides are vertical while the back edge is more oblique and sloping. The core of the feature is earth and plesia clay covered by two layers of stucco, a second layer of plesia, and finally four layers of fine quality lime plaster. Only very faint signs of burning were recognised in 1950 on the upper surface of the feature and other areas in the Shrine, indicating that the feature was likely not a hearth, more likely an altar for specialised ritual use, and the discolouration from burning may be due to burning in the room, perhaps during a calamitous event.70 Added to the W side of the main feature were two plaster extensions (Pl. 6a). The first, toward the back, is shaped much like the main altar (0.60 m × 0.47 m) but with a deep circular opening in its centre (Di. 0.16 m, 0.12 m deep) with plastered sloping walls. It may be a receptacle or stand for a vessel, such as a rhyton.71 The other extension is a long ‘bolster’ of plaster-covered clay (0.65 m × 0.23 m). Between the two is a plaster channel or runnel (W. c. 0.08 m) that slopes down for c. 0.40 m to the W to the mouth of a heavy kitchenware jar (50‑234, NM 5375)72 that was sunk beneath the floor and secured with plesia clay.73 The installation is certainly designed for the controlled flow and collection of liquids and suggests that the altar was used at least in part for libation offerings. Also to the W of the feature was found a large ceramic circular tray (50‑322, BE 8046) in a tan burnished fineware that dates to early in the LH III period. The floor surface between the back of the altar and the S wall of the Shrine is almost entirely missing. A packing of plesia (Th. 0.12 m) was found here in 1950, bringing the surface up to a height of some 0.11 m above that of the rest of the room and creating a long step or raised platform, just before the threshold in the S wall. The space here seems limited either for circulation around the altar or for access to Room G. This could indicate that the expansion of the building to the E and the construction of the S wall may have taken place when the upper floor was to be installed and the altar retired from use. The plesia layer might represent an intermediate stage of preparation and filling for the later floor. At a distance of 0.60 m N of the plaster feature, and just W of its centre axis, was found a large roughly circular cutting in the plaster floor of the Shrine (W. 1.43–1.5 m). In the centre of the cutting was a large, unworked boulder (e) (1.15 m × 0.60 m), its roughly flat upper surface 0.24 m on average above the lower floor and 0.05 m above the level of the plaster feature. The stone (Pl. 6b) was let through the floor and set on bedrock at the E side (0.60 m below) and on fill at the W. The space surrounding the stone, within the cutting on the N, 37
PLATE 6
(a)
(b) (a) Lower Shrine with plaster altar. From N; (b) Lower Shrine with plaster altar and boulder in situ. From E.
38
PLATE 7
50-322
50-234
(a)
50-287
50-314 50-184
(b) (a) Pots associated with the Lower Shrine and plaster altar (50-234, 50-322); (b) pots from Drain (50-184), Shrine (50-287) and Wall K (50-314).
39
W, and S, is packed with smaller stones and mixed clays and earth. On the E side, a hard cement-like soil fills the gap between bedrock and the floor. Among the stones were found fragments of the floor plaster broken when the cutting was made for the stone, a few sherds, and tiny fragments of bronze (50‑220)74 and gold leaf (50‑244, BE 7087), remnants of the Shrine’s contents when the stone was added. The purpose of the large stone is unknown; Wace suggested that the stone was the base for a wooden column or pillar75 and Mylonas believed it to be a ‘slaughtering stone’ for the sacrifice of ritual victims.76 It is clear that it belongs to an intermediate phase of construction and use of the Shrine, after the laying of the lower floor, through which the cutting had to be made for the emplacement, and before the room was filled and floor level raised to that of the upper floor. Although highly unlikely that the stone would have supported a column while under the upper floor level, if the building expansion took place during the lifespan of the lower shrine, an internal central support may have seemed necessary. It is possible, though by no means certain, that the end of the lower Shrine was accompanied by fire, though this cannot have been severe. Signs of burning were found near the plaster altar but nowhere else for certain in the room. Fragments of burned plaster and carbon were a feature of the fill that was put over the lower floor and probably originated from debris of the earlier phase. There was no general destruction, at least by fire, when the Shrine was destroyed, although structural damage might be indicated in the later buttressing and reflooring of the structure. The cause of the significant alteration of the function of the Shrine and the complete covering of the altar is not yet fully understood. The Shrine — Upper Floor Tsountas may have reached the upper floor of the Shrine in 1886, at least in spots, even though he does not report any work here. His plan (Fig. 3) clearly indicates a circular feature in the NE corner that remained in situ on the upper floor to be investigated in 1950.77 He may also have dug through the upper floor in the centre of the room, where the plaster surface was found very disturbed and the boulder beneath could have been interpreted as bedrock. Clearly few interesting artifacts were found to warrant the excavation further across or below the floor. In 1950, the full extent of the Shrine was cleared, including wherever the upper floor survived, especially along the N and E side of the room. During the investigation of the earlier, lower floor, much of the upper floor was removed, along with the fill underneath it, in order to follow the lower floor, especially in a wide swathe from N to S across the centre of the room and to its E side.78 The plan of 1960 (Fig. 1) shows the remains of the upper floor following the 1950 excavation and a small test cutting at the N end in 1959. The northern room of the Shrine79 had a length of at least 6.45 m and an interior width of approximately 4.5 m. The later, upper floor (Th. 0.04 m) was certainly associated with the later phase of the E and S walls, up to which it runs, and it may in fact be contemporary with them.80 Wall plaster was preserved in spots but is undecorated. The W side is of course not preserved at all. The entrance must have been on the N but no indication of its position or the existence of a wall can now be confirmed. The sidewalls may have ended in antae, or squared ends at any rate, and the E side at least would have been blocked by the ashlar feature in Area Q (Pl. 8c–d) immediately outside the Shrine to the N (see below). The space between the feature and the floor edge is only c. 0.15 m so any wall would have to have been a screen or fence and likely of ephemeral materials. Much burning was evidenced here including bits of carbon, 40
black earth and burned plaster, a line of which marks the inner face of the N end from the N end of the E wall and extending to just N of the circular vat/bin (see below). On the W side, a small return wall may have once existed before the wide, open entrance. There also may have been a threshold or step, either way, most likely in wood. The upper plaster floor is 0.35 m–0.40 m above the surface of the lower floor (Pl. 5a) and between the two is fill of earth and many small stones containing carbon flecks, plaster fragments, mostly unburned, a few sherds and a miniature goblet (50‑287, BE 7060, Pl. 7b). The height of the new functional surface means that the features found on or in the lower floor would have been entirely covered and therefore, were no longer in use. The limestone boulder is 0.06 m–0.16 m below the upper floor level and the plaster altar is 0.20 m–0.25 m below. The threshold in the S wall that communicates with Room G presumably remained in use during this phase. However, the photographs of the S wall in 1950 show constructed rubble above the thick line of the burned threshold and would suggest that before the final destruction of the Shrine, the doorway to Room G had been walled up, and the remaining material in it sealed off from access.81 Little was found in the later, upper Shrine to provide any confirmation of its function. One exception is the circular feature in the NE corner mentioned above.82 It is a vat or bin of very soft disintegrating clay walls set into the upper plaster floor (H. (ex.) 0.10 m, Di. c. 0.75 m). It is held in place with small stones and plesia; the base is the plaster floor itself covered with a thick layer (0.035 m) of plesia. Inside were excavated ash, burned earth, and other burned material, possibly from the destruction, as the entire area around it was burned, including plaster from the walls.83 What is clear from the lack of evidence, the lack of features and finds in the upper Shrine room, is the complete change in the use of space — the obvious emphasis on internal ritual focus with the plaster hearth is literally removed from sight and use, as the upper floor transforms the Shrine into a large room, carefully appointed but seemingly empty, with the only permanent feature probably related to cult, and almost certainly replacing the function of the now retired plaster altar being the ashlar feature (dd) just beyond the N end of the Shrine building.84 AREA Q (Fig. 1; Pl. 8c–d) Q is the area immediately to the N, and across the front of the Shrine that contains a possible porch-like structure about 2.0 m deep, the culmination of the Middle Ramp of the Sacred Way, and a feature or aedicule. Tsountas cleared much of this area and the area further N (see Middle Ramp below). He includes on the plan of 1886 (Fig. 3) the worked stone blocks dd.85 The third block c is not distinguished and its location is represented as a wall that is only partially preserved. The plaster platform is not indicated and may not have been fully cleaned. Tsountas represents Area Q as a porch, showing a dotted line to represent where he believed the northern limit to be, and we will see below that he was likely correct. The full width of Area Q is not preserved as the W side is mostly eroded away. The substantial terrace that supports the area does remain at a deeper level so estimates can be made. It is significant that the western terrace wall was extended to the NW, and reinforced with buttressing, to provide the necessary space and foundation for a 2.0 m deep area across the entire width of the Shrine building. It is built on a heavy fill of earth and stones, plainly visible along its western edge. 41
In 1950 and 1959, only one plaster floor, belonging to the terminus of the Middle ramp, was identified. It was described as a thick plaster with pebbles and a plaster undercoat, likely to be identified as a ramp level resurfaced in tougher pebble-plaster. It is 0.19 m above the upper floor of the Shrine. In the excavation of 1972, Mylonas’s probe under the uneven southern edge of the ramp uncovered an earlier plaster surface app. 0.20 m below the later ramp floor. He ascribes this surface to the use of the earlier Shrine and identifies two steps of small poros slabs in situ that would cover the c. 50 cm height differential (Mylonas 1972b, 121). Significantly though, the height of Mylonas’s lower surface would match that of the later, upper floor of the Shrine building and very likely indicates the floor level of Area Q prior to the construction of the ashlar feature dd and the end of the ramp. At the E end of Area Q is a feature made of cut ashlar blocks dd and a rectangular step platform of plaster beside/before it to the W (Pl. 8c–d). The feature has been called an ‘aedicule’ and due to the religious interpretation of the Shrine building in front of which it sits, it too has been interpreted as a ritual installation. The plaster step platform has a maximum preserved length (N–S) of 1.80 m, and its width (E–W) is 1.10 m. Its core is of well-tamped earth, clay and small stones. The top surface and W face are coated with four thin layers of a fine delicate plaster and indicate frequent resurfacing (ILN 1961, 23 490–92; WBM 1, 15). The upper surface is 0.36 m above the floor of the Ramp to the W. The W end of the platform is a straight vertical edge that sits on the ramp surface and was constructed after it. A test cutting into the platform near the damaged SW corner showed that the ramp floor does not continue under the platform.86 The eastern edge of the step platform abuts against two ashlar, poros blocks dd. These two carefully cut blocks fit very well and tight along their the western face but have a gaping joint on the eastern side and are wedged with fill. Both the bedrock surrounding them and the blocks themselves have been cut to fit one another. A cut and sawn ledge (0.16 m–0.20 m) runs along the N edge of the N block and the W edge of both. The northern, the larger of the blocks (L. 1.0 m, W. 0.60 m), has two rectangular dowel holes near its N edge and one near its W, centrally placed. The NE corner of the block is cracked across the end of the dowel hole. The smaller block to the S (L. 0.70 m, W. 0.45 m) has one rectangular dowel hole only, near its W edge and its S end is chiselled, not finished, to receive a third block.87 The bedrock is also cut for a third and last slab that would have extended up to, and abutted on, the N end of the Shrine, for a total length of about 2.5 m. A fourth block (c) was found in 1950 adjoining the S end of the step platform already mentioned, although it was interpreted simply as an anta block for the Shrine’s N wall. The block was basically in situ but tipped out of position (L. 0.80 m, W. 0.42 m, H. 0.47 m).88 It was chiselled on its E face and drilled on its W face. Its association with the plaster platform is evidenced by the top layer of plaster extending from the platform’s W face onto the NW corner of the block. Like the other two ashlar blocks dd, block c had a cut and sawn ledge along its northern edge with a single rectangular dowel hole, more or less centrally placed. As mentioned above, the 1886 plan does not show block c as an isolated worked block and it must not have been cleared enough at that time to be seen in full. The number and positions of the dowel holes suggest that they were used for securing wooden planks or facing on a core structure. The rough surface of the blocks beyond the cut ledge indicates a core of rubble or mudbrick. The northern block rests directly on bedrock, which has been cut to receive it. Both blocks E of the platform are wedged on the E with a 42
PLATE 8
(a)
(b)
(c)
(d)
(a) Tsountas House complex, Lower Shrine and view of Area Q with ashlar altar. From N; (b) Tsountas House, general view from S with Area Q’s terrace substructure and Middle Ramp in background; (c) Area Q, with ashlar altar and terminus of Middle Ramp. From N; (d) Shrine floors with intermediate fill. Ashlar exterior altar. From SW.
43
packing of fairly large stones and plesia. Bedrock to the N is only a little lower than the surface of the blocks; and to the E it slopes up fairly steeply and soon becomes higher. Here, the rough surface of the bedrock was filled with plesia and covered at the time of excavation with a good deal of reddish earth, perhaps dissolved mudbrick rubble from the core structure. In 1950, it was clear that this area had been subject to a great heat, certainly fuelled by the presence of wood, and in addition to the earth and plesia, calcined stone and fragments of baked mudbrick were also prevalent. At the N end of Area Q, a rectangular cut groove 0.18 m wide, 0.12 m deep and filled with charred remains of a beam, was found running E and W for c. 3.0 m through the thick plaster and pebble floor, in line with the northern face of the plaster platform. The walls of the cutting are smoothed and coated with white clay. The function of a thin wooden beam just under and across the floor here is obscure. The alignment with the N edge of the ashlar feature must not be coincidental and should mark the boundary, real or conceptual, of Area Q, a porch or vestibule to the Shrine. From the NW corner of the step platform eastwards to Wall R, there may have been something more substantial. No remains of wood or stones were recovered but there is a gap of some 0.60 m between the northern face of the large poros block and the coarse plaster floor of a court or platform to its N. Since there is no obvious architectural or other practical use for the ashlar feature dd and the delicate step platform, a religious use may be hesitatingly proposed. This was Wace’s original suggestion. The platform may have been a stand for offerings or a kind of altar. And although evidence of burning was found along the feature to its S and behind it to the E, there was no burning found on the step platform or damage from fire on the ashlar blocks. The extensive use of wood in the structure itself and potentially as the material of the N wall would explain the burning without the need for a burning altar, especially this close to the structure and probably under a light roof. The life of the Shrine ended with fire as evidenced by the burning in the NE corner, the burned plaster on the E and S walls and the burned wood in the N and S ends of the room.
THE FINDS
ROOM G Pottery The condition and source Due to the fact that Room G was excavated almost entirely to bedrock by Tsountas in 1886 and he does not seem to have collected sherds,89 we now can say almost nothing about the ceramics from this room, and therefore, are unable to date accurately its construction and phases of use in isolation. Architecturally, it is the earliest component of the Shrine so based on the dating of the other elements, Room G will be either earlier or contemporary. In 1950, a pocket of fill was excavated near the N wall of the room. The small size and mixed nature of the sherds suggest that they do not originate from the contents of the room itself but are more likely to have been part of fills that entered the room during construction phases, as part of the destruction and through later occupation and discard. Of course some of the sherds would have come from the mudbrick walls after their collapse. 44
Pottery types and dates A number of sherds were recovered from the 1950 cleaning of Room G. Unfortunately, the sherds were very mixed and included several post-Mycenaean pieces. A large percentage was MH that included a wide variety of Minyan and matt paint types. LH sherds occurred from most periods but were primarily LH IIIA and IIIB; several pieces were easily identified as belonging to LH IIIB2, such as a Type B deep bowl (FS 284) and a krater (FS 281) fragment that is monochrome on its interior, and LH IIIC middle, including a Close Style stirrup jar body (FS 175) with two body zones and a shallow angular bowl rim (FS 295) that is undecorated out and linear inside. These sherds are only the remnants from an area already cleared to bedrock and are unlikely to be truly representative. There is a greater proportion of MH than was encountered elsewhere except in the lower levels under the forecourt of the House and in the lowest phases of the upper ramp. The lack of material from early LH IIIC is not unusual in the context of the Cult Centre (WBM 16/17), nor is the post-Mycenaean mixing. Other Materials 90 Tsountas of course did collect finds of other materials that he says were found in the fill above the floor, on the floor itself and in the fill below the floor down to bedrock.91 As discussed above, these objects may have come from storage in the room itself, some on shelves, potentially from different periods, from fill dumped into the room during a rebuilding phase, even possibly as debris from the main Shrine room to the N. Since no trace remained of the floor dug through by Tsountas, it is impossible to relate the objects found in 1950 to this floor. The ivory helmet crest came from the N wall and the gold fragments and other objects from in or near it. The more or less datable finds from 1886 and 1950 are not all of the same period. The total accumulation92 might well be scraps of damaged offerings or a collection of valuable trinkets; the plaque strongly suggests an icon or cult object.93 Most of the ivories and beads may date to LH II94 or early in LH IIIA while the glass pieces could be a little later into LH IIIB. None date later than a destruction date at the end of LH IIIB2. Metal A gold elaborate palm/flower-shaped bead (NMA 2500) (pl. 13.27;95 Pl. 9b) has the border and petals outlined with granulation and originally had cloisonné inlays; traces of red paste remain. It probably dates from LH II–IIIA.96 From the area of the N wall come small fragments of pressed gold leaf, including at least three rosettes (XS 87), one of which is pierced by four holes (50‑89, BE 7017 and 50‑95, BE 7016). Bone/ivory Tsountas published four ivories, including two lily-shaped inlays, one large (XS 105) (NMA 2737) (pl. 13.14, Pl. 9b) and one small (XS 106) (pl. 13.15), and a lantern bead (XS 14) (NMA 2515) (pl. 13.6) although he identifies them as bone. The fourth is a large ivory wing (NMA 2726) (pl. 13A) that must belong to a sphinx or griffin. The carved details are similar to the smaller scale wings of the heraldic sphinxes on the plaque from the House of the Sphinxes. The wing was said by Tsountas to have been found almost on bedrock and may date to the earlier phases of the building, certainly not later than early in LH IIIB. Ivory carving, however, is a very conservative craft and therefore particularly difficult to date. The wing must have 45
PLATE 9
50-98 50-109
50-99
50-94
(a)
13.8
13.3
13.14
13.23
13.24
13.27
(b) Objects from Room G. (a) Animal figurines (50-109, 50-98), Psi figurine (50-99), ivory helmet crest (50-94), all scale 2:3; (b) inlays, appliques and beads after Tsountas 1887, pl. 13: ivory lily (13.14); glass rosette (13.8), female figures (13.23–24), and nautilus (13.3); gold bead (13.27).
46
formed part of a fairly large piece of carving. Its base is scored for attachment; its back is absolutely smooth. Another large piece of carved ivory (50-94, NMA 7103, Pl. 9a) was uncovered in 1950 in the N wall of the room. It is perhaps the crest of a helmet and is likely to be of LH IIIA or IIIB date. Glass and beads Most of the objects from Room G illustrated by Tsountas are of glass (pl. 13.1–5, 7–12, 16–18, 20, 22; Pl. 9b). Most are blue and a few are white; the majority are beads or appliqués and a few are inlays, all from the typical range. Designs include rosettes (XS 86, 87, 63 and a fragmentary 77), ivy leaf (XS 96 and quatrefoil XS 103, NMA 2514), nautilus (XS 117), lily (XS 104), pendant spiral curves (XS 59 and a double 62), pendant wave (XS 138 and 139), and a large encircled star (NMA 2512). Exceptional are two glass appliqués with female figures (pl. 13.23, 24; Pl. 9b), possibly goddesses, with rectangular schematised lower body and female upper torso and head; both have arms encircling their breasts. A third example was found but not illustrated. Beads of glass were also mentioned, presumably spherical, along with beads of amber. Of particular importance is an imported Egyptian scarab of faience, with the name of Queen Tiy, the wife of 18th Dynasty pharaoh, Amenhotep III. The scarab would have been manufactured, at the latest, in the early 14th century BC but may have arrived at Mycenae well after that and could have been kept as an exotic curio for even longer. Terracotta From near the N wall were found in 1950 three terracotta figurines: two unusual animals and an unusual female figurine (50-99, BE 7015, Pl. 9a), a very small example of the Psi type without articulated breasts. This might be the latest of the finds from Room G dating to LH IIIB. To early LH IIIA probably dates the monochrome hindquarters of an animal (50-98, BE 5970, Pl. 9a) with a very stocky body, short rounded legs and a short straight tail. The other animal figurine is probably a dog (50-109, BE 29085, Pl. 9a). It has a small head and short ears and tail, is decorated in the Linear 1 style and dates to LH IIIA2 or IIIB. A biconical conulus (50-100, BE 5990) in terracotta was also found in 1950 and could date as early as MH or to LH III. Plaster Among the significant finds from the Shrine, and specifically from Room G, is the famous painted plaster plaque (NMA 2666) (Tsountas 1887, 162–64 and pl. 10.2 and 10.2a), showing two female worshippers in flounced skirts on either side of a large figure-of-eight shield with traces of a figure, presumably divine, either behind it or extending from it (Fig. 6). The surface is very damaged in spots so the central figure is not fully visible. What can be distinguished seems to be arms with white flesh on the blue background, suggesting a goddess. On the ground between the shield and the figure on the right, is a small concave-sided altar. The scene is framed with a yellow band marked with black cross lines and black cross lines on a blue background around the perimeter. The sides of the plaque are decorated with painted stripes of alternating light and dark hues of blue and red on a white background, and must have been meant to show, so the plaque would not have been framed and may have been handled or manipulated — a moveable icon. It suggests an ex-voto or possibly a kind of icon. There is no way to date this object with any accuracy, although it probably belongs to LH III. 47
THE SHRINE Pottery The condition and source Very little pottery can be associated with specific features and phases in the Shrine, however, there are some important whole vases that provide evidence for the ritual use of the structure, at least in its earliest stage. No sherd material was collected from the walls or under the lower floor in undisturbed areas in order to provide the date/s of construction. The earliest material comes from under the lower floor and around the limestone boulder that was let into the lower floor in a later phase. Many of the pieces may be from the fill packing under the lower floor but certainly include material contemporary to the emplacement of the stone, potentially fragments from the use of the Shrine. The bulk of the sherds excavated in the Shrine come from the 1950 investigation of the central and eastern parts of the Shrine when c. 0.40 m of fill was removed from between the upper and lower floors. The few joins among the sherds suggest that these are pieces that mixed with the soil during the raising of the level for the laying of the upper floor and not remains of the original contents of the shrine although a few scraps may be. Another unit between the floors was excavated in 1959 near the N end of the room and was of similar depth and character to the other larger unit. Due primarily to Tsountas’s excavation of the upper levels, no pottery was recovered, or at least kept, from above the upper floor that would have included the final contents of the Shrine room, material from the collapsed mudbrick walls, and evidence of the destruction, presumably like the remainder of the Cult Centre at the end of LH IIIB2. Pottery types and dates From the lower floor of the Shrine and the unique plaster feature in the southern part of the room no sherd material appears to have been collected or kept.97 Two whole vases were found, however, that provide significant indication of the use of the plaster feature as a libation altar and focus of ritual, even if they do not provide tight dating evidence. The first vase is a large jar (FS 67) (Pl. 7a) in a heavy kitchenware fabric (50‑234, NM 5375).98 It had been placed under the lower floor to the W of the altar with its mouth flush with the floor at the end of a small plaster channel.99 It was clearly intended to collect liquid run-off from the altar. The shape and fabric type indicates a probable date of LH IIIA.100 The second vase is a large (Di. 0.33 m) straight-sided cylindrical tray (FS 322) (Pl. 7a) with two or three handles on top of the rim (50‑322, BE 8046). It is in a pink-tan fabric that has been burnished extensively, a type of fabric tradition that remains in production into LH IIIA1 at Mycenae and elsewhere, the period to which this pot should date (Mountjoy 1981, 74). The tray is neither a common shape nor commonly found in domestic contexts and here, found just to the W of the altar near the bolster, was likely used for the transportation and presentation of offerings. The sherds from around the boulder in the cutting into, and below, the lower floor are mostly early, especially MH, both Minyan and matt paint examples, but include early LH pieces in coarse and fineware. The latest sherd is the handle of a monochrome LH IIIA2 kylix (FS 264). There is no way to know if this sherd was under the lower floor or in the context of the stone fill around the boulder that was installed after the lower floor was built.101 The lack of sherds from the soil fill under the floor in other areas might suggest that this material was a later intrusion at the time of the cutting of the floor and the stone setting. 48
The sherds from between the upper and lower floors provide a wider selection of examples and a wider date range. Pre-Mycenaean pieces include Neolithic, EH, and many (> 38) middle to late phase MH sherds. Early Mycenaean sherds are present but the bulk is LH IIIA with some certainly LH IIIB, possibly mid IIIB. The additional 1959 test between the floors at the N end of the room confirmed the primarily earlier LH III character of the sherds including mostly undecorated kylikes and conical cup fragments together with a handful of LH IIIA decorated pieces. The installation of the upper floor and the covering of the features on/in the lower floor happens sometime after the early LH IIIB period, possibly towards the middle of the period, or even slightly later, and could be associated with the phase VII destruction and rebuilding in the beginning of the next phase.102 A single whole vase was found in this material, from between the floors and in the NW area of the room, preserved intact because of its diminutive size (H. (ex.) 0.03 m). It is a miniature handmade goblet or kylix (50‑287, BE 7060, Pl. 7b) with an irregular conical profile; unusual horizontal handles near the rim and decorated with a row of vertical buds. Neither the decoration, which seems early, or the unusual shape can be dated closely. However, most miniatures, especially drinking vessels, date to LH IIIA (like wheelmade examples and shapes in miniature) or LH IIIB (mostly handmade and with generic wavy line decoration). As was the case of the vessels associated with the lower floor, this miniature is probably connected to the function of the Shrine and may be one of the few identifiable remains of its use for ritual, since it is likely to be a votive, a symbolic representation or gesture of a full-sized functional example with intensified ideological properties (Shelton 2009c and Tournavitou 2009, 229–30). Other Materials In addition to the miniature goblet (50‑287, BE 7060; Pl. 7b), a few objects found in the Shrine might be scraps of damaged votive material, even if there was nothing other than the bin built into the upper floor that can be associated with the latest use phase. There were only very fragmentary pieces of gold leaf, faience, and bronze, including parts of a heavily corroded pin. These were mainly from the fill between the upper and lower floors of the Shrine and do not necessarily belong to an earlier use of the Shrine. It is interesting and perhaps fairly significant that no figurine fragments were identified in the fill associated with either floor of the Shrine or with its remodelling in a later phase. Metal Several objects in bronze must originally have been in circulation in the area since a number of tiny fragments were found: in the fill between the upper and lower floors was a single fragment (50‑195) c. 0.10 m below the upper floor and part of a bronze pin (50‑274, BE 5993; Pl. 10a) at the N end of the room; from the earth and stones around the boulder were two small bronze fragments (50‑220) on the W side and three more (50‑232) elsewhere. Also from the earth and stone packing around the boulder were four tiny fragments of gold leaf (50‑244, BE 7087), one on the W side. Glass and beads A fragment of faience (50‑579, BE 7117; Pl. 10a), either an inlay or vessel piece, was found in among the sherds from the fill between the upper and lower floors in the Shrine. 49
PLATE 10
59-505 50-579 50-313
50-274
(a)
59-501
50-219
50-237
50-231
50-236
(b) (a) Objects from the Shrine: plaster wall fragment (50-313), clay counter (59-505), faience fragment (50-579), bronze pin fragments (50-274); (b) objects from Area Q: terracotta figurines (50-219, 59-501), stone conuli (50-236, 50-237), lead clamp (50-231). All scale 1:1.
50
Terracotta A single clay counter (59‑505, BE 9858; Pl. 10a) was found at the N end of the room between the floors. It is similar to one found further N still in a mixed level over the Middle Ramp (59‑170). Plaster Several plaster fragments were found mixed in fill and still on the walls, those showing signs of burning, but only two decorated examples were registered: a small wall fragment with pale blue and red areas found in the earth around the altar (50‑172, BE 5701); and a white wall fragment with two black stripes and red tinting from the fill between the floors (50‑313, BE 24284; Pl. 10a). AREA Q Pottery The condition, source, pottery types and dates Tsountas’s excavation in Area Q cleared most of the ceramic material that would have been associated with the ashlar feature and the entrance porch. The area was cleaned again in 1950 and the LH II–IIIB pottery was mixed with post-Mycenaean pieces including two Megarian bowl sherds. Further exploration was done in 1959 along the S edge of the plaster platform to the E of the third ashlar block and in a small square test (0.30 m × 0.30 m) under the platform immediately N of the same block in burnt soil. Only eight sherds were found and five were unpainted body pieces. There were two MH pieces and the others were clearly LH III but not more narrowly dateable. Another area cleaned in 1950 was in the NW part of the Shrine into the heavy terrace that supported Area Q, and was extended and strengthened during the last building phase. Other Materials A few small finds were found during the cleaning of Area Q, all of them from a mixed ceramic context but certainly of Mycenaean date. From the test cutting was found a painted plaster fragment and the late LH IIIB female figurine (59‑501, BE 9747; Pl. 10b) that dates the construction of the ashlar feature and the step platform, the final components of the Shrine and built after the final Middle Ramp surface in front of the structure. Stone Two stone conuli were found: one ovoid to the E of the ashlar blocks (50‑236, BE 7079; Pl. 10b) and the other conical near the W wall of the terrace (50‑237, BE 7104; Pl. 10b). Metal A fragmentary lead clamp (50‑231, BE 5981; Pl. 10b) was found between the ashlar feature and Wall R in a mixed context together with the ovoid conulus (50‑236, BE 7079; Pl. 10b). Terracotta 50‑219 (BE 7042; Pl. 10b) is an unusual figurine head from mixed fill and sherds in Area Q. From burned soil N of the third ashlar block and c. 0.19 m under the plaster step platform was 51
found a Psi female figurine of a late LH IIIB type with a thin crescent body, hollow back, and uneven pellet breasts (59‑501, BE 9747; Pl. 10b). It is discoloured by fire. Plaster Also from the test cutting below the platform was a burnt plaster fragment decorated with a black band similar to a fragment from the forecourt of the House (50‑218, BE 24289).
SUMMARY Room G is the earliest component of the Shrine, built probably during the LH IIIA period. It may have stood alone or more likely as the back/inner room of a two-room structure, with the original walls preserved only in G itself. The area had previously been inhabited during the MH period and MH sherds are found in quantity under the floors down to bedrock. The small amount of early Mycenaean material, probably from the disintegrated mudbrick, indicates that construction here probably began towards the beginning of LH III and possibly just a little before that. The finds potentially associated with the room, whether as part of a constructional phase or the room’s function, provide dates from later LH II up to LH IIIB. The plaster altar and the lower floor of the Shrine were in place and in use by LH IIIA2 at least, the earliest date when the boulder could have been placed through the floor cutting. Both the kitchenware jar and the tan burnished tray, in use during the lifespan of the earlier Shrine phase, indicate a date in LH IIIA and very likely beginning in the earlier part of that period. The walls and their terraced support exhibit several phases of construction but can be dated in sequence only on architectural grounds. Sherds from between the two floors suggest that the major alteration of the Shrine with the covering of the earlier ritual features with fill and the laying of a new plaster floor took place at the earliest during the middle of LH IIIB but may have been in part a result of damage from the earthquake at the end of Phase VII, as was also likely the cause for the addition of heavier terracing and buttressing of that terrace on the W and S. The extension of the building terrace to the N allowed for the construction of Area Q and the ashlar-based feature with a plaster step platform in front of the Shrine during the late LH IIIB period and replacing the function of the now covered altar inside the building. The Shrine building and Area Q were destroyed by fire, presumably in the general disaster at the end of the LH IIIB period, although due to the early excavation of this destruction debris no direct evidence of the date of this event was preserved.
USE OF THE SHRINE There can be no doubt that Wace’s interpretation is correct that Shrine G was a Mycenaean cult building and that for part of its history, the older, smaller room to the S perhaps was used as an adyton and here might have been stored the ritual paraphernalia and votive offerings. The N room of the Shrine has clear cult use with the plaster feature built into the floor as a permanent focus of ritual action (Pl. 6a). Other possible installations in the room have not survived. The altar was set close to and in front of the entrance to Room G emphasising the importance of the inner part of the Shrine. The accommodations for catching liquid run-off such as the runlet and large jar sunk into the floor indicate libation offering as at least one of the primary uses of the altar, and by extension, of the Shrine in its early phase (Pl. 7a). Burnt 52
offerings were unlikely to have occurred inside the building on the altar since there was only a very light indication of burning on the surface of the plaster that is more likely connected to a destruction event rather than a ritual act, although this latter cannot be entirely ruled out. The surface and walls of the altar are covered by many layers of fine plaster, and the delicate nature of the surface would not have stood up well to repeated use and therefore were occasionally renewed. The floor, on the other hand, consists of only one layer of thin and carefully made plaster. This would suggest that the interior of the Shrine was not subject to substantial traffic and that entrance into the room was permitted to only very few, possibly only the religious functionaries. So even while the altar’s surface was renewed probably because of wear and tear over a quite long period of use, the floor was not as worn. The rituals performed on the altar may then have been somewhat destructive or that the periods of use of the altar or the surface and walls of it required a renewal. Further evidence for the use of this cult building has not been preserved. The very denuded state of much of the N and W sides of the building has contributed to this. Besides the plaster altar, no other cult installations were found. As was discussed above, the boulder that was let through the floor of the Shrine and set in place on bedrock with a packing of smaller stones cannot be interpreted with certainty either as a feature or focus of ritual or simply as a structural element. It does seem very unlikely to have been used as a spot for animal sacrifice, as suggested by Mylonas, because of the very low level of the stone, the limited interior space, the lack of any faunal remains or other sacrificial paraphernalia from the immediate area or in the fill over the lower floor, and the uncertain nature of any evidence for burned sacrifice within the structure. A clearance of objects, especially votives, may have occurred prior to the reworking of the Shrine in its later phase and the covering of the altar. It is curious though that the tray (50‑322, BE 8046, Pl. 7a), clearly associated with the use of the room and the only object found on the lower floor, was the only item left behind. Some of the objects originally from the Shrine may have been among those found by Tsountas in Room G although it is also possible that most votives and ritual paraphernalia were housed for safety and convenience in the inner room. Unfortunately, the nature of Tsountas’s excavation has made a more accurate interpretation of the source and use of the materials in G impossible, even though the assortment of small objects in glass, ivory, stone, and gold could have been at the least the remains of votives or scrap items as collected in other cult buildings such as the Temple and Room of the Fresco complex. The exception is the painted plaster plaque (NMA 2666) showing the worship of a goddess with a large figure-of-eight shield that is very likely an icon for religious veneration (Fig. 6). What is clear from the lack of evidence and the lack of features and finds in the upper Shrine room, is the complete change in the use of space — the obvious emphasis on internal ritual focus with the plaster hearth is literally removed from sight and use as the upper floor transforms the Shrine into a large room, carefully appointed but seemingly empty. The only permanent feature probably related to cult, and almost certainly replacing the function of the now retired plaster altar, is the ashlar feature just beyond the N end of the Shrine building (Pl. 8c–d). Nothing was found in Area Q that is contemporary with the earlier, lower Shrine, when the ritual focus is located in an interior space with probably limited access and visibility by anyone other than the closest participants. This seems to change when the upper Shrine floor in installed and no contemporary cult features are installed within the structure, other than the clay bin in the NE corner. The physical focus is now on the exterior of the structure while the 53
purpose and use of the Shrine building itself it unknown. It is also unclear how and if Room G is used in this later phase since the doorway between rooms was walled up before the destruction of the building. The ashlar altar may also have been used for libation or as a ‘table of offerings.’ Burnt offerings seem unlikely here too as it was with the earlier plaster altar because of the extensive use of wood as a facing on the feature and the total lack of burning on the limestone base and plaster platform. The original height is unknown as is its original form beyond the use of facing slabs or boards and a core of other materials, probably mudbrick which later melted and solidified during the destruction by fire. The plaster platform to the W of the altar indicates that the ritual functionaries would probably have stood to that side of the structure, perhaps looking E and upslope, while additional participants could have stood in the small open court to the N, especially on the slightly raised soft plaster platform. Others in the area may have been able to view the proceedings from various vistas on the Upper Ramp, Middle Ramp and from the roofs and windows of the Megaron and House on the next terrace down to the W. Until at least the middle of LH IIIB the Shrine was located outside of the citadel of Mycenae and for more than a half century it sat relatively alone on the slope (Fig. 7). During this phase the ritual was enclosed and visibility and participation was extremely limited. Only in the beginning of the LH IIIB period are other structures built nearby including impressive domestic structures like the South House to the N and the other structures immediately associated with this neighbourhood but on terraces below that of the Shrine (Fig. 8). Access became increasingly more limited with these constructions but still possible from almost any direction. It is only after the construction of the WCW and the organisation of access routes by way of Stairway K and the Ramp system from upslope in later LH IIIB that access to the area of the Shrine is truly restricted, but the ritual focus is shifted to the exterior and becomes more physically and visually accessible to a larger potential number of participants (Fig. 11). At the entrance to the Shrine terminates the long processional approach to the Cult Centre, ‘the sacred way of Mycenae’ (Pls. 8c, 15a). The ashlar altar beside that terminus belongs to the upper level of the ramp, both constructed on the extended and strengthened terrace and all constructed in LH IIIB2. This external focus replaced at that time the internal ritual of the earlier, lower phase of the Shrine built and in use by at least the LH IIIA2 period and perhaps even earlier in LH IIIA, and were then taken out of use and covered over sometime after the early LH IIIB period.
54
PART 3. ACCESS TO THE AREA References: Tsountas 1886, 77; 1887, 160–62; 1893, 22; Tsountas and Manatt 1897, 34; Wace 1949, 67; ILN 1950, 1041, fig. 7 (1042); Wace 1951, 254; BSA 51 (1956) 122; ILN 1961, 490–92; WBM 1, 15, 19, 24, 29; Wace Guide, 35; Mylonas 1966, 107–10; 1971, 151–55; 1972b, 121–22; 1974, 90–91; CC, 18–21, 23, 36; Mylonas Guide, 24; 1977b, 19; 1983a, 308–10, 315; 1983b, 128–32; Iakovidis 1983, 45 and 48; 1986, 243–44; 1996, 1044; 1997, 148–50, and 154; 2004, 14–15, 17–21, 23–24.
DESCRIPTION AND EXCAVATION (Figs. 1–4, 7–11; Pls. 1, 8c, 11–15) This section will discuss the ‘auxiliary areas’ of the Tsountas House excavation, i.e. the areas surrounding the structures and features and directly associated to them. These areas, many of them are features in fact, define the borders, artificially determined, or perhaps intentional, of the Tsountas House complex. At the same time, and of greater significance, is the definition of access and circulation in to, out of, and within the complex itself and the Cult Centre as a whole (Fig. 8). The House and the Shrine are the southern edge of the Cult Centre complex as we now understand it and their southern limit was defined by a long staircase along the edges of the two buildings. Upslope to the E of these buildings began a long and monumental approach to the area through a series of descending ramps, the lower portions of which separate the House from the remainder of the complex to the N. Finally, the W Cyclopean fortification wall, now
Shrine
Fig. 7. Plan of Mycenaean citadel in LH IIIA2, with position of Shrine G (11) (after French 2002, fig. 16).
55
9. Processional Way
13. Central Court 14. Temple 15. Room with the Fresco Complex
10. Megaron 11. Shrine 12. Tsountas House
Fig. 8. The Cult Centre, showing access routes (after French 2002, fig. 33).
including the later Polygonal Tower, encloses the area to the W just beyond the basement of the House and incorporates it into the citadel. All these areas, or features, were explored more or less by all three excavation campaigns: Tsountas, Wace/Taylour, and Mylonas. The least amount of clearance was done early by Tsountas, so much more recent work centred on the periphery of the complex, especially in 1959/60, and under Mylonas (1966–1974), some of which has been described in print.103 STAIRWAY AND CENTRAL DRAIN K In spite of the great mass of cultural accumulation inside the Polygonal Tower, Tsountas’s tests led him to believe that between the W, exterior, wall of the house and the West Citadel Wall (WCW) there was initially a narrow passage (D) and from that point access was gained from the W wall to a steep stairway (K) made of small stones for easier ascent to the higher parts of the citadel. He reported 32 steps and believed the stairway to predate the house.104 The 1886 plan (Fig. 3) 56
shows only 31 steps, 25 on the N side and six higher up on the S. In 1950, Wace investigated the stairway, or stepped street, and an open drain in the rocky slope running down its side (Wace ILN 1950, 1041, fig. 7; 1951, 254; briefly in 1949, 67). Although not mentioned by Tsountas, his plan suggests that he recognised the drain in part, at least along the long stretch of steps, but probably did not clear it to any extent.105 His plan also shows a low stepped wall bordering the drain on the S side and protecting the adjacent building foundations. Both stairway and wall are built on bedrock with a fill of large stones. The treads are made of small to medium roughly squared and unevenly levelled stones, two to four across to make a step, and they decrease in width along the length of the stairway, from 1.40 m at the lowest end, level with the House’s basement corridor (E), to 1.10 m in line with the SE corner of the megaron room (C). Above this corner the stairs change direction slightly for the highest five or six steps on the N side, conforming with the line of the platform to the S of G. The cleaning and examination of the stairway in 1950, and again in 1959 (Fig. 1), countered Tsountas’ opinion and confirmed that the steps abut against the S wall of the House and must therefore be contemporary or later.106 At the level of Shrine G’s W wall, the steps to the N of the drain stop and then continue up on the S side, flanking the heavy drain wall of the upper terrace, in a short but carefully made flight of six steps, of approximately three stone width.107 From the top of the N steps to the bottom of the upper flight of S steps there must have been a bridge or pass over the open drain course, presumably of wood. In 1960, only four steps were left on the S side and 24, of roughly equal breadth, on the N.108 The stairway is not of a uniform breadth. At the point where its S flanking wall ends and is met almost at right angles by another wall from the S, it measures 3.00 m across, but in line with the outside face of the SE corner of Megaron C of the House, it measures only 2.30 m. The drain was open for the length of its lower course (ILN 1950, fig. 7), from the top of the stairway on the N side. It runs over sloping and undulating bedrock that has not obviously been cut or shaped into a channel, although the outcroppings form a zigzagging pattern for a descending channel and this may have been enhanced through minor cutting. In addition, large stones had been set at intervals down its course to break the violence of the torrent that would have rushed down after heavy rains.109 This drain, like many others on the citadel, was designed to channel storm run-off that would otherwise have caused flooding and serious erosion as it poured off the steep and rocky hill. Its open form may be evidence that it is early, possibly even taking advantage of a natural watershed, and therefore was not built into or under the later (LH IIIB) structures that hide a sophisticated drainage system planned and implemented during the widespread construction phase on the acropolis, from the Palace on down the slopes and out through the fortification walls. Above the top of the stairs on the N side of the drain, and above the bridge that must have led to the S flight of stairs, the drain ran between thick, high walls of very large stones (see plan, ILN 1950, fig. 7; and ILN 1961, fig. 6). The lower end of the heavy wall on the N is formed by the jog in the wall to the S of G. The size and construction of the walls could have contained/controlled a large force of water but may also have held a superstructure of some sort, a tower perhaps, even simply a covering roof of the drain itself, just W and downslope from the gated entrance to the Cult Centre, where the drain passes through a culvert (see below) over which runs a ramp to the N from a fine, conglomerate threshold (m). The massive walls could even have been part of the gateway itself as support or protection. Possibly in 1886, one or two cover slabs may still have been in situ and might explain why Tsountas’ plan (Fig. 3) shows a wall running from the G platform to the lower end of the S wall of the drain. Today the SW corner of the jog of 57
PLATE 11
(a)
(b)
Ramp 1 Ramp 3
Ramp 2
(c)
(d)
(a) Drain and Stairway K. From SW; (b) lower part of Stairway K. From SE; (c) Upper Ramp, S end at threshold m, showing ramp floors 1–3. From S; (d ) threshold m and culvert at top of Stairway K, showing ramp floors 1–2. From NW.
58
the platform is 1.64 m lower than the end of the S wall of the drain opposite it; but originally it must have been equally high, unless the drain roof ended higher up opposite the junction of the jog and the N drain wall. The most probable explanation of the heavy superstructure, carried by the drain walls and roof, is a tower or barrier to restrict direct access at this point to the N. The stairway that skirts the walls upward to the S emphasises this circulation route. The Greek excavations to the S of the upper flight of steps uncovered a route to the S (Fig. 11) that provides access to the Southwest residential quarter of the citadel, to a prothyron structure at the entrance to the ramped ‘processional way’, and to a stairway leading even further up the citadel slope (Mylonas 1966, 107–09; Iakovidis 2004, 15–16). The Culvert At the top of its course, the drain passes through a culvert under the threshold and ramped entrance to the Cult Centre (Pl. 11). The walls of the culvert are of medium stones in shallow courses with much small-stone packing. The succeeding courses are arranged with a slight slope inwards for an almost corbelled arrangement, but as the channel narrows, the stones are more vertically aligned and do not meet at the top. Rather, the roof is of large flattish stones, or slabs. Some, like the conglomerate cover slab, overlap the N wall of the drain and are supported to the S by resting against threshold m, while others span the channel, overlapping, and supported on, its walls on both sides. The floor of the drain in the culvert is formed by bedrock along its entire length and so far as it is visible beyond the inside of the Mycenaean wall. To begin with, the rock is more or less horizontal, giving the culvert a maximum height of c. 1.20 m. The width in the culvert is 0.70–0.75 m. Further up, beyond the constructed culvert, the bedrock slopes up steeply and the height and width of the drain channel diminishes rapidly. About 3.00 m in from the outer edge of the culvert cover-slab, the channel becomes too narrow for further exploration110 and at that point divides into two branches and vanishes into deep unexcavated fill.111 Lower End of Stairway and Drain K The lower course of the drain and the bottom of the steps ran through unexcavated ground between the House basement and the Polygonal Tower. This area was excavated by Mylonas in 1974 and revealed a few more steps and a further segment of the drain on bedrock (Fig. 4) (Mylonas 1974, 90–91). The stairway is not preserved W of the basement corridor (E) although stone packing on the bedrock might indicate that it did continue, at least for the full length of the House’s S wall. The exterior of this wall, just below the stairs, still has traces of its original plastering. The gaps between the stones were filled with clay and pebbles covered by a plaster of mud and chaff. The plastering remains due to heavy burning, evident from the blackened plaster, hard burned mud plaster, and partially calcined stones. There still remains a good deal of unexcavated fill on the inside of the Polygonal Tower (see Introduction above) so the drain course cannot be traced further, however, the mouth of a drain in the outer face of the tower112 might show the position of the original exit through the Mycenaean circuit wall, even though the stone with a channel cut in it forming the drain floor, like the tower, is post-Mycenaean. The exit is c. 2.00 m S of a line due W from the culvert. Some time after 1886 Tsountas excavated S of Stairway K, ‘a complex of walls, passages and buildings’, belonging to LH III.113 These remains of building complexes, alleyways, and further stairs, from S of Stairway K to the southern edge of the citadel, were fully explored by Mylonas and Iakovidis from 1968–1974 and 1987–1989.114 59
THE SACRED WAY (Figs. 1, 4, 7, 11) The approach from up the slope and the E, to the Cult Centre was by a series of paved ramps, in a switchback arrangement, beginning with steps and transitioning across a landing to a gentle down sloping ramp to the N, at least at the time of the major destruction of the complex and its entrance. In the early stages, the approach from up slope was less well defined and certainly less monumentalised and less restricted. The construction of an enclosed and gated prothyron created a barrier and interruption in the natural flow produced by a street, or ramp-way.115 Access to the various structures in the complex, including the Shrine G and the House on the terrace below it, was achieved through further descending ramps, here called the Middle and Lower ramps. The Upper Ramp from Threshold m to the Middle Ramp Tsountas seems to have cleared enough of the Upper Ramp, its supporting walls, and the postMycenaean terrace wall built over the area, just to the N of the culvert, to include on the plan of 1886 (Fig. 3). He certainly recognised the different chronological periods of the various walls, indicating the terrace wall that was the eastern border of his excavation in pink to correctly suggest a Hellenistic date of construction (like the Polygonal Tower, for instance). Tsountas, however, did not clear the entire area of the ramp that he includes on the plan, as the condition of much of the ramp between Walls A and R, still covered with calcined debris in 1960, plainly indicates. Wace did not further investigate this area to any real extent in 1950, so that a good deal of fill, destruction debris, and post-destruction reuse remained to be excavated in 1960 (Fig. 1). This ramp, or road, has been traced for the whole length of the terrace on which sits the Shrine and the Middle Ramp that leads to it from the switchback turn at the N end of the Upper Ramp (Pl. 13), forming the eastern boundary of the area and the Cult Centre more generally. In 1960, a test (Pl. 12a) to the S of threshold m indicated that the ramped approach appeared to carry on into territory that was outside the permit area, and which was later entirely cleared of debris and Hellenistic over-building by Mylonas (1966–1972) to reveal a covered and gated prothyron, and the continuation of the ramped approach to the S and then a switchback to the N and E further up slope (Figs. 4, 11) (Mylonas 1971, 152–55; CC, 19). At the S end of the upper ramp investigated in 1960, it passes over the culvert and across the conglomerate threshold m. There are several phases of the ramp, represented by levels of plaster surface renewed or resurfaced, both associated with the gated entrance and the drain culvert and earlier phases prior to their construction (Pls. 11, 12). Threshold m The cover-slab of the culvert, of rough-hewn conglomerate, is supported on the N wall of the drain and by leaning its S end against a great conglomerate threshold m. A heavy postMycenaean terrace wall, determined to be of Hellenistic date, conceals more than half of threshold m, and much of the ramp to the immediate N and S of the threshold. Until the Greek excavations of the 1970s cleared the post-Mycenaean terrace wall and the fill it supported to the E (CC, 19), the length of m could be measured only from inside the drain, where the lower northern end of the great slab was visible along the S side of the drain passage. The threshold is a massive cut conglomerate block 2.52 m long and 0.93 m wide, with a cut ledge on the upper surface along the length of both sides. The northern ledge is set back c. 0.28 m from 60
the N edge of the slab and has a pivot hole, 0.12 m in diameter, at its W end. Both the width of the ledge and the pivot indicate the existence of a door that opened to the N and closed against the edge of the doorstop ledge. The southern ledge is only 0.04 m from the S edge of the slab and was cut to accommodate the edge of a floor to the S. The surface of the slab between the two ledges and to the E of the pivot hole has been sawn and polished to a very smooth even surface. The western edge has been cut but unpolished. It was suspected in 1960 that the size of the threshold suggested space for a double door. This was confirmed when fully cleared by Mylonas and a second pivot hole was revealed at the E end of the northern cut ledge. He also found burned remains of the wooden doors on the threshold (CC, 19 and Mylonas Guide, 24). Threshold m belonged to an imposing structure on the ramp (prothyron) and would have supported a substantial doorframe and heavy wooden door that were meant to protect as well as impress. It is similar to other examples at Mycenae, in which conglomerate was used for appearance as well as strength, such as at the Lion and Postern Gates, the House of Columns, and in the Palace.116 To the S of the threshold, a narrow strip of plaster floor was found projecting from beneath the post-Mycenaean terrace wall in 1960 (Pl. 11c). Above it was a very hard burnt layer, and the plaster itself had been seriously damaged by heat. Below it no earlier floor was found, only a lump of calcined poros immediately beneath the plaster. The strip of the floor revealed was some 1.05 m long and ran from the threshold on the N to a small block of conglomerate to the S at the limit of the excavation. Investigation below the plaster surface did not uncover any earlier/lower plaster layer. Sherds both under and over the plaster surface date to the LH IIIB2 period, those above were found within the hard burnt destruction debris.117 Also from the layer on the plaster floor were two pieces of blue painted wall plaster originating from the covered and enclosed prothyron structure that was the main entrance to the Cult Centre from its construction in LH IIIB2.118 The Upper Ramp To the N of threshold m, a ramp gently descends for 30.5 m to the N where it turns sharply to the left in a switchback to the W and then S. The width of the ramp fluctuates from about 1.5 to 2.5 m wide originally although the surfaces are very damaged and the full width is preserved in very few spots. The plaster surface of the ramp or road was rebuilt or renewed several times over the use-life of the area and is defined by a tall terrace wall upslope to the E, Wall A, and a low supporting wall along the W side, Wall R. Although the ramp can be traced very clearly, it is poorly preserved in many places and more than half of it together with its retaining wall R has all but disappeared. At its lower, northern end, where the Upper Ramp transitions to the Middle Ramp, it also meets the S entrance of a long corridor (Room 4) in the Citadel House Area (Fig. 8) (ILN 1961, 490, cols. 2, 3; figs. 2f; MT III, figs. 85, 87 and plan VI). The threshold was apparently only of wood and is evidenced by the rectangular gap across the width of the floor plaster. Room 4 was entered from the N through another doorway at the end of another intersection of two ramps, a small ramp119 descending steeply on the E and the ‘Causeway’ ramp, made of two rows of dressed poros blocks, leading up from the NW.120 Although little is known about the function of the corridor, it is clear that it could control access to the N end of the Upper and Middle ramps. 61
PLATE 12
Ramp 2
Ramp 3
surface 4
surface 5
bedrock
(a)
(b) (a) T2C test trench below Ramp 2 with earlier surfaces (3–5) to bedrock; (b) Upper and Middle Ramps from S; (inset) Upper Ramp, detail of square poros base n for post. From NE.
62
1 2
2
2 3
1. Calcined debris 2. Rubble fill
3. Fine grey layer Plaster floor
Fig. 9. Elevation and section of Upper Ramp surfaces (Trench T2C).
Wall A The wall is founded on bedrock and consists of large stones for the most part, Cyclopean in size, along the lowest course especially. Its primary purpose may well have been to retain a building terrace of unknown nature on the still unexcavated slope above the area of the Cult Centre, on which was at least one structure of architectural and administrative significance, as the source of the massive conglomerate column base121 and Linear B tablets that fell into Room 4 (Corridor) to the E of the South House Annex during, or just after, the destruction (Chadwick 1962, 35–46). Wall A becomes the eastern boundary of the area because of its height and ultimately acts as the curb wall for the Upper Ramp. It is not built in a straight line but rather is a series of straight segments that follow the contour of the rock on which it is founded and makes subtle adjustments of orientation with a jog in its line at points along its length. At the time of excavation, the W face of the wall still bore traces of surface treatment with rough mud plaster in several spots, a few still covered with a c. 0.015 m layer of hard white lime plaster, especially at the jog c. 10 m N of threshold m (trench T2B/C) and at a point 8.0 m further N (trench T2A). It is likely that much of the face of the wall was once plastered. Most of the middle section of Wall A is preserved to a height of about 2.0 m. Its height dwindles southwards and is quite low where it disappears into the post-Mycenaean terrace wall c. 5.5 m 63
N of threshold m. Wall A’s continuation southwards is obliterated by the post-Mycenaean terrace wall that crossed threshold m and most of the area to its S, where the Greek excavations exposed another segment of it that supports the staircase of the ‘Sacred Way’ further up the slope to the E (Mylonas 1970, 119; 1977b, 24; 1983a, 309). In order to clear the E end of threshold m, Wall A cannot have continued along the line of its last visible stretch, but must have made another jog to the E; yet another small jog is made at the S end of the prothyron above an ashlar anta block (Mylonas 1970, pl. 164b). Wall A was crossed by another stretch of Hellenistic terrace wall about midway along the ramp at c. 6.0–7.0 m N of the jog, and the two run parallel to one another, the Mycenaean wall behind the Hellenistic wall, for c. 25.0 m to the E of Room 4 / ‘the corridor’ where it continues to be founded on bedrock. The E wall of Room 4 is actually a rather high rock outcrop that was squared off with stones and plaster to form a ledge before the higher face of terrace wall A above and behind it. This lower ledge wall continues S beyond the doorway of Room 4 to become the E curb wall of the Upper Ramp at its N end. This wall was made of small to medium stones and its surface covered both with an under layer of mud plaster and a surface layer of hard white lime plaster (trench G4). It is preserved to a height of c. 0.80 m (the height of the ledge wall in Room 4) and for a length of c. 7.0 m where it is disturbed by later cross walls (LH IIIC) and the Hellenistic terrace wall. Presumably it masks the rock outcrop as it did in Room 4, and presumably it originally intersected with wall A at a point no longer surviving; this is unfortunate since the direct relationship of the two walls could have provided welcome evidence for their chronological as well as functional relationship.122 Wall R The Upper Ramp is supported along its W side by Wall R that is built in four segments (Fig. 2), each with slightly alternating orientations; of these the most southern is now largely missing. The first and second segments, meeting behind Bench h (see below under the Middle Ramp), are built in one piece incorporating only a change of direction. There is, however, a break at either end of segment three;123 and at its S end the wall bulges out to the W, restricting the already narrow lane between Wall R and the Shrine to only 0.50 m, while following the jog in Wall A at the same point. The N end of segment three, on the other hand, meets segment two at an inward, eastern curve of the wall where Wall A is straight, therefore unnecessarily narrowing the width of the ramp surface. In this case, Wall R seems to allow for an opening of space to its W in Area Q with the ashlar altar (dd) and plastered platform. Wall R is founded on the bedrock but does not slavishly follow the edge of a long N–S outcrop at this point as might have been expected so its position has more to do with support for the ramp and maintaining its consistent width within the restrictions of pre-existing features on both the upper and middle terraces. The most southern, fourth segment of Wall R was in an extremely destroyed state, where little more than its bedding remains, largely missing apart from a stretch some 2.40 m long built of much larger stones than the rest of the extant wall. The smaller stones just N of the culvert are in part a remnant of Wall R when an early phase of ramp surface passed over them and before larger stones ascending to the drain capstone were put in place. As mentioned above, the junction of segments three and four in a westward bulge impinges significantly on the width of the lane between the Ramp and the Shrine, suggesting that this lane was not a viable circulation route.124 The high drain walls to the S of G apparently would 64
have blocked the S end of the lane anyway, while an additional low blocking wall ran across to Wall R from the SE corner of the G platform, a corner built of massive stones.125 The route of Stairway K to the S of the heavy drain walls suggests that circulation veered to the S and around to the N on the ramp, once the culvert was built with restricted access from this point once the prothyron was in place. In an earlier phase access to the passage alongside the Shrine may have been possible from the stairway. An early ramp surface, predating the culvert, may also have descended towards the stairway or back into this lane. The Ramp surface The ramp surface N of threshold m is of a good, almost fine, hard, white plaster. The fine quality of the plaster suggests that the ramp may have been roofed, even beyond the covered and frescoed prothyron, and also was more vulnerable to damage and erosion during and after the destruction. The patchiness of its preservation, however, led to the identification of several plaster surfaces at varying depths and in different segments of the ramp. The uppermost and final ramp was covered with a thick, hard layer of destruction debris resulting from the great fire that destroyed the whole area at the end of LH IIIB. Two phases of ramp (Pl. 11c) seem to have led up to the culvert, but only one of these clearly passed over it to reach the N edge of threshold m. This final plaster surface can be traced only as a line of plaster immediately below thick destruction debris on which is built the post-Mycenaean terrace wall. The plaster line extends for only c. 3.40 m N from the threshold,126 where it disappears before reaching the visible stretch of Wall A. Ramp 1,127 the uppermost surface, may have existed only in this segment to connect the lower, slightly earlier level of Ramp 2 to the N with the increased level of threshold m when it was built in LH IIIB2. This final ramp surface should correlate with the single surface identified S of threshold m. The penultimate plaster surface, or Ramp 2, is the one most clearly visible along most of the length of the Upper Ramp, at least along the E edge against Wall A where it remains best preserved, built up against that wall. The plaster of Ramp 2 can be traced to the N edge of the culvert capstone, where it is c. 0.23 m below the final plaster surface. Along the W edge, the surface is quite worn and three earlier phases of plastered or pebbled surfaces can be recognised at various depths. A short stretch of the plaster of Ramp 2, uncovered below Ramp 1 plaster, in a small test burrowing under the post-Mycenaean wall revealed a well preserved surface where the packing between the floors was of hard earth and poros rubble with sherds. Elsewhere, further N, mudbrick matrix material had been observed at some points in the packing of the floors. The third surface from the top, Ramp 3, cannot be traced beyond the N wall of the culvert (Pl. 11c). It is visible there as a line of plaster running through the stones of a wall to the N of the culvert, with fairly small stones beneath it and considerably larger above. Its end is about 0.65 m below the cover-slab of the culvert, showing that the level of the ramp was raised by this amount when the culvert was built. Further to the N (6.75 m), Ramp 3 is only 0.30 m below the surface of Ramp 2 indicating a descent of 0.30–0.35 m towards the culvert rather than ascent. Ramp 3 is certainly earlier than the construction of the threshold and the culvert capstone and its supporting stones that are built on top of the plaster surface. Ramp 3 runs over smaller stones beneath, which may be remnants of support Wall R before the N culvert wall becomes a facing for it. The descending plaster surface may have originally, pre-culvert and large drain walls to 65
its W, led from/to the area S of G and Stairway K, as well as to an early lane to the S, just above Mylonas’s Building A (Mylonas 1966b, 109). A 1.0 m cut was made through Ramp 2 along its very broken W edge at c. 6.0 m N of the threshold to test for earlier ramp surfaces (Pl. 12a). The cutting exposed four floors below Ramp 2. It also revealed bedrock, rising from c. 0.30 m from S to N, revealed a levelling of the rock and packing with small stones, sherds, and earth to a height of 0.45 m. Upon that the first surface, Ramp 5, was constructed of pebbles alone and 0.16 m above that the second surface, Ramp 4, had a considerable quantity of pebbles in a thick, rough plaster matrix, unlike the third that was mainly of plaster while the fourth and final surfaces were of a finer and smoother plaster. From the eastern jog in Wall A to S of the join between segments two and three of Wall R (trench T2B), the ramp is very poorly preserved; over considerable stretches no plaster floor remains. In a very few spots, especially at the eastern edge along Wall A as before, one or two roughly plastered surfaces were found (max w. 0.55 m). To the immediate E of the stone base n, two floors remained, neither in good condition, with a fill 0.13 m deep between them. The upper floor seems to have been resurfaced with only 0.04 m between two plaster lines. Much of this part of the ramp had been over built both in post-destruction periods as well as postMycenaean times. Both preserved plaster surfaces are more level than in the southern part of the ramp and represent a flattening out of the ramp slope. Under the upper preserved plaster surface, there is a bedding of small stones packed with a mudbrick matrix. Resting on this bedding was a rectangular cut poros block, stone base n (Pl. 12b). Still further N, from the join between Wall R segments two and three to the S scarp of the section unexcavated in 1960, parts of four distinct ramp floors, in addition to resurfacings, are preserved, with a total depth of only 0.15 m. The W edges were progressively less eroded downwards, so that the lowest floor extended almost to Wall R. Of these surfaces, the uppermost and lowest are of fine hard plaster, while the second and third are much inferior. The uppermost plaster surface, which inclines noticeably to the S, had calcined debris to a maximum depth of 0.30 m directly upon it, or separated by a thin layer of fine, black earth. In patches, the intense heat of the destruction event destroyed the plaster surface. Above the calcined debris lay a mass of disintegrated mudbrick and stone. To the N of the unexcavated section128 (between trenches T2A and G4), although the ramp surface was blackened by fire, its condition was very good with the full width preserved from the low curb wall on the E to Wall R on the W. This E curb wall also narrowed somewhat the width of the ramp, now at its narrowest at this point, only 1.50 m, just S of its convergence with the Middle Ramp. Wall R is also well preserved at this point culminating in a cut stone block (L. 0.33 m, W. 0.26 m, H. 0.13 m) where the Upper Ramp curves sharply in a switchback to the W and then S onto the Middle Ramp (Pl. 13). Remaining to be added to this account of the Upper Ramp is the excavation of the area of the overhanging post-Mycenaean wall, near the N end of the ramp by the Greek excavations in 1971 (Fig. 4). There Mylonas uncovered evidence of domestic buildings from several phases of LH IIIC over accumulated fill covering the destruction debris on the ramp. Also uncovered was a cut poros block, similar to stone base n (see below), and a wide, rectangular gap (0.83 m) in the plaster ramp floor across its width that he compares to the gap in the doorway to Room 4 for the placement of a wooden threshold (0.85 m) and just 5.0 m S of it. Mylonas also found charred remains of wood in the gap and suggested it had been covered by a wooden step 66
(c. 0.30 m high).129 It is more likely that there was another doorway installed across the ramp at this point, one that would check access to the Upper Ramp from below or from the N, or to any point further into the Cult Centre via the Middle Ramp. From all this detail it will be clear that the Upper Ramp, a stretch of some 30 m is not homogenous. Its E Wall A ran in straight lines with jogs. Its W Wall R meanders. The third segment may have been rebuilt and the S end of the fourth was built up much higher when the culvert was constructed. The number of phases, the width of the ramp and the kinds of surfacing and packing used are not uniform from one stretch to another. The state of the ramp surface when uncovered in 1960 was also varied. The extreme N end and a stretch for some 3.50 m S of the baulk were relatively well preserved because they were covered to a considerable depth by the collapse of burning buildings. At the S end, the line of the fifth floor visible in the section and the short stretch of the one plaster floor uncovered S of threshold m, were similarly deeply buried. The middle stretch, however, on which the stone base n stands seems to have escaped severe burning and though fallen stones were found on it, there was nothing like the quantity of collapsed stone and mudbrick found further N. Presumably because it was less well buried, the ramp here is exceedingly ill-preserved. On the other hand, this section of the ramp also received the greatest extent of overbuilding during the post-palatial period and much of that directly on the surviving ramp surfaces. The location for these constructions may have been chosen because of the low level of destruction debris or more likely, the builders removed much of the debris to create a level building surface. Where this clearing ran deep, disturbance of the upper ramp surface was also a result. Stone Base n (ILN 1961, fig. 14) (Pl. 12b) On the ramp near its western edge and c. 15 m from the threshold was found a rectangular cut poros block sitting on a layer of small stones and crumbled mud matrix over a well-preserved part of the third segment of Wall R. It was believed to be in situ. The block (H. 0.28 m, L. 0.47 m, W. 0.37 m) has a circular cutting on its upper surface (Di. 0.20 m, D. 0.03 m) and a small circular opening (D. 0.05 m) centrally placed on the S face. It must be the base for a wooden post or slender column that sat securely in the circular cutting. On the S face of the block fragments of plaster were found in the small central depression, the purpose for which is unclear. As mentioned above, the part of the ramp on which the block stands has two floors only preserved. The lower surface appears to have been cut to accommodate the block and the second is level with its top. It is not known how the plaster of the second ramp floor fit around the base since the floor is largely missing in this area. Only on the E side of the base does the plaster come very near, but even here there is a gap. This suggests that, unless the base went out of use, this second floor was also the last in this stretch and not merely the last preserved, with later floor/s worn away. The sides of the stone would have been hidden by the second floor, suggesting that the central depression and the fragments of plaster on the S face belong to an earlier period of use with either the first floor, or elsewhere. If the first floor had to be cut to make room for the base, its use in that position looks like an afterthought. Taylour believed that it ‘could not have been an isolated example and one can visualise a porticoed walk climbing the side of the hill’ (ILN 1961, 490 and WBM 1, 15 and 161d). In fact, further clearing of the same area in 1966 by Mylonas uncovered another stone with a cutting for a wooden post, in this case an irregularly shaped stone, just 1.5 m to the N in line with stone base n. In between the two was said to be ‘a hearth inside a ring of plaster’; the entire 67
installation was said to date to LH IIIC, well after the destruction of the ramp.130 This segment of the ramp and the next segment further N as well, had a good deal of disturbance and overbuilding in several periods of LH IIIC including several light walls and a later floor surface at the E edge of the ramp. In most areas there was some accumulation of fill over the LH IIIB2 end debris to potentially indicate some lapse in use131 before building in LH IIIC (WBM 1, 15 and 28–29; Mylonas 1966b, 109–10; 1971, 152–53). The question remains if stone base n, an identical block found by Mylonas in the steps E of the prothyron (Mylonas, PAE 1970, 119–20 and pl. 165a), and perhaps even the second rough stone base found in 1966 and cut block in 1971 were originally part of the ‘Sacred Way’ or were used here for the first time after its destruction. The cut blocks are very unusual and the slender posts they would have supported are also not typical,132 but it is not hard to imagine that part of the covered ramp could have been open to the W, especially in the segment to the N of the obstruction of the Shrine building and overlooking the ashlar altar in Area Q. There was no doubt in 1960 that stone base n was in situ and was probably installed in the second half of LH IIIB together with the upper floor surface associated with it;133 possibly this stretch of the ramp with only two floors was rebuilt in the second half of LH IIIB expressly for the suggested colonnade. The Middle Ramp The stretch of the plastered approach from the hairpin turn to the W and then S at the N end of the Upper Ramp (Pl. 13), southwards to the entrance of the Shrine (c. 13 m) (Pls. 14, 15), is termed the Middle Ramp. Elements of the Middle Ramp include a plastered bench (h) on the E side against the curb Wall R and on the W side, in the curb Wall C, poros steps (e) that provide access between the terrace on which the Shrine is built and the ground floor level of Tsountas House, as well as to the Lower Ramp at that same level (Figs. 7 and 11). The northern limit of Tsountas’s excavation, as shown in the 1886 plan (Fig. 3), crosses the Middle Ramp not far to the N of the Shrine and just S of the bench on the E side and S of the features in Wall C, the curb wall for the ramp on the W side. His clearing N of Area Q was only superficial and did not expose the ramps or the destruction debris upon them. The same crumbled mudbrick and areas of fine black earth were found to the S as to the N of the limit of his excavation. His plan indicates as well that he cleared in the area of the poros steps, however, only one partial tread may be represented and the remainder were still buried under destruction debris in 1959 when totally excavated. In fact, Tsountas also must have dumped backdirt and stone from his excavations over part of the Middle Ramp that he had cleared, as it was covered once again to a significant height in 1950. The ramp widened somewhat as it descended to the S and a little to the W, its narrowest point having a width of c. 2.20 m where the Upper and Middle ramps converge in a flattened area (2.40 m × 3.55 m). From the corridor (Room 4) southwards the hard pebbled plaster surface of the ramp was burnt and covered with patches of fallen wall plaster, usually stuck hard, and with a layer of fine burnt black soil with small pieces of charcoal. This plaster originally belonged to the surfaces of the curb flanking walls on the W and E, especially from the upper parts, as did also the layer of orange-red earth, clearly crumbled and burnt mudbrick, that had collapsed over the layer of burnt remains of wood. The western edge of the ramp had cracked and was partly missing along much of Wall C, especially for a stretch of about 2.60 m from the SE corner of blocks g northwards. The broken 68
PLATE 13
Upper Ramp
bench
Middle Ramp
Turn of Upper Ramp to Middle Ramp with bench h. From SW.
edge showed that this part of the ramp had two superimposed layers of a mosaic-like plaster, hard and thin, with very fine stones set in it and a highly polished surface that had burned to a very dark blue-grey colour. Bedrock beneath it was packed with plesia clay. Bench h On the E side of the Middle Ramp is a long low bench (Pl. 13) built against Wall R. The betterpreserved part of this bench is at its N end, which stops a little short of the N end cornerstone of Wall R (1.5 m). Further S the line of the front of the bench is clearly marked by the turning up of the floor plaster of the ramp but the bench itself is not preserved. Tsountas may have uncovered more of this bench; his plan shows a feature of some kind along the foot of Wall R. If this is the bench, in 1886 it extended much further S to a point behind Area Q than it does now. The bench was in two segments with a short jog to the E (0.09 m) between them where the bedrock outcrop it covers slants to the SE, just S of the intersection of Wall R segments one and two. The N segment measures 2.70 m and the S is preserved for c. 1.00 m. Its depth varies with the distance of its front edge from Wall R, from about 0.30 m to 0.53 m, and is partly determined by the bedrock it covers. The height of its seat above the ramp is 0.40 m to 0.43 m and it has a good plaster back against the face of Wall R. Mudbrick on clay-packed bedrock forms the core of the bench. There is an undercoat of thick rougher plaster. All of the surfaces were very burnt. 69
Court and platform (Pls. 8c, 14a, 15a) About 8.0 m from its N end and just S of the jog in bench h, the Middle Ramp widens to a little over 3.0 m and then divides into two sections, on the W side descending further to the S it narrows to 1.43 m, and on the E side flattening out onto a stuccoed platform c. 0.10 m higher than the ramp to the W. This area just N of the ashlar feature dd and Area Q may act as a court for the Shrine. The E–W channel cut into the ramp floor, from the NW corner of the step platform, is the only potential boundary between Area Q and the Middle Ramp. The two sides of the plastered surface, the ramp and platform, are divided by a line of straight-faced stones running N for c. 4.20 m from the step-platform W of the ashlar feature (dd), continuing the line of its W face. A number of stones are missing from this row. They were not wedged into position, nor covered any longer, so many were dislodged by the weight of Tsountas’s dump over them and some were removed by mistake during the cleaning in 1959. The plaster of the ramp turns up over the W faces of the stones that remain. At the N end of the row, the plaster is not broken and there makes a rising turn to the NE comparable with that around the cornerstone at the N end of Wall R, where the Middle and Upper ramps join. The sloping ramp to the W of this row of stones is not preserved along much of its W edge partially through erosion and partially from the very denuded state of the terrace wall that would have supported it at this point. Otherwise, the plaster surface remains for the most part up to the line of stones and onto their W face. The hard plaster is thick and mixed with pebbles, would be appropriate for an exterior paving, and likely represents a final resurfacing over an undercoat of plaster alone. A square (50 × 50 cm) test trench N of Area Q revealed bedrock at 0.20 m below the floor level. The ramp plaster at its S terminus (Pls. 14a, 15a) in Area Q, in front of the Shrine, is of good quality; though not as excellent as the thin pebbled plaster found further N between Wall C and the bench. This superior plaster, of which there were two coats, apparently began roughly in line with the cornerstones g of Wall C; no clearly marked division was noticed, however, and the hard thin pebbled plaster merged imperceptibly into the thick plaster extending southwards to Area Q. The difference in quality may be a chronological factor, or may indicate areas of higher traffic, or more prolonged exposure. The plaster surfaces in the court, both ramp and platform to the E, may be later additions at the time of the extension of the support terrace for Area Q (see below under Lower Ramp) and the construction of the ashlar feature (dd) and the step platform, up to which the ramp extends (Pl. 8c–d).134 The final plaster coat of the platform, however, goes over the ramp floor and was the last element to be renewed in this area. To the E of the line of stones, the plaster of the raised and flattened platform is of an even lesser quality and runs eastwards to the bedrock outcrop at the foot of Wall R. It is soft, thick, and poorly preserved. This surface ends in an approximate E–W line about 0.50 m to the N of the ashlar feature and it may have been the edge of the platform. To the E of ashlar blocks dd, no evidence of plastering was observed over the unworked bedrock packed with earth and clay. The platform is the same level as the step platform W of the ashlar feature, the surface of which is 0.49 m above the ramp floor. Wall C Wall C is the curb wall that retains and encloses the Middle Ramp along its upper W side. It is also, and more importantly, the E wall of the Megaron. For most of its length, only the 70
foundation of the wall remains, made of rubble masonry, although to some depth as it is required to support the terrace above it to the E. At its N end, some mudbrick of the superstructure remained, especially where it bonds with the substantial mudbrick remains on the N wall of Room 2. Both the E and W faces were originally plastered, with more remaining on the E face where it merged with the plaster of the ramp. The S end of Wall C is formed by three cornerstone blocks of ashlar poros (g) that are very closely joined so as to appear as a single block. The force of the final destruction in this area, as well as the weight of the debris mass, has tilted them downslope to the W, though their bases are still in situ. The total length of the three blocks is 1.10 m, the width of each is 0.62 m, and their tops are 0.46 m higher than the floor of the ramp and slab f immediately to the S, originally having a height of 0.64 m (1.05 m above bedrock with the addition of their foundation course). The W, S, and E faces of the blocks were plastered. Slab f sits immediately to the S of the Wall C cornerstones (g) and 0.90 m N of the poros steps (e) (see below). It is a long, flat, hard limestone slab (L. 0.90 m, W. 0.50 m, H. 0.10–0.12 m) with a high polish on its upper surface. From the destruction by fire of the structure around and partially under it, the slab had slipped southwards c. 0.05 m from the cornerstone of the wall to its N, with which it originally belonged. Traces of plaster were seen on the upper surface where pieces of wall plaster had also adhered, just as they had done on the ramp floor surface in spots. The rough underside of the slab had a 0.07 m thick coat of plaster with crumbled mud matrix and a little earth beneath it. At further depth and to the W of the slab were many small stones and burnt black earth, ash, and pieces of carbon, remains of a good deal of wood or a wooden structure in this area. Slab f is very likely a threshold for access into the Megaron building, and perhaps via the Megaron, for access to the lower level of the Cult Centre.135 Most of the ramp surface, especially between Wall C and the bench, was covered with a layer of burnt black soil under building debris from the destruction at the end of LH IIIB2. Between the slab f and the poros steps e to the S, the ramp edge had cracked away with the heat of an intense fire and a large piece had fallen away to the W. The remains on the spot were reduced to a white powder of calcined lime over a charred mass of wood that was probably a turning post attached to a wooden substructure of the steps to the N and on which may have been set a further flight in wood starting from the threshold, slab f. Very fierce burning had occurred here indicative of a good deal of wood originally, some extending perhaps under the final ramp floor where burning is visible in the section. Steps e (Pls. 14, 15) Access from the Middle Ramp and the Shrine to the Lower Ramp and the House beyond (Figs. 7 and 11) was by way of five steps of cut poros limestone located about 10 m from the N end of the ramp.136 The total height of the steps is 0.87 m. The height of each riser is 0.17 or 0.18 m. Steps e, though rather worn now, are of good ashlar masonry; each step is made up of two or three rectangular blocks. The length of the top step is 1.28 m, of the second and fourth from the top is 1.21 m, and of the third and fifth is 1.14 m. The steps are aligned and the blocks flush at the S end (on the right) but end at different lengths on the N side (on the left). On either side they were packed with small stones and plesia holding the blocks firmly in place. The irregular N end of the steps was not meant to be seen in this way and originally would have bonded with a support of some type to the N. I suggest that this support was wooden based on the burning and carbon remains and that it formed a stair spine, railing, or wall, as well as a 71
PLATE 14
(a)
platform
Shrine
(b) (a) Area Q and Middle Ramp with poros steps e. From W; (b) Middle Ramp with steps to Lower Ramp. Area Q with exterior altar. From W.
72
PLATE 15
(a)
(b) (a) Area Q and Middle Ramp with poros steps e. From SW; (b) lower ramp from stairs to Forecourt. From SW.
73
roof support over the steps. As mentioned above, the poros Steps e may have been the lower flight of a stairway into or over the Megaron, the upper flight of which was wooden and began from threshold slab f. The plaster remaining in superficial irregularities on the poros steps shows that they were once plastered over their surface. The top tread has preserved three layers of thin plaster, above which runs the final plaster and pebble floor of the ramp. It seems clear that the steps, the single ramp floor covering the top tread, and the retaining wall of the ramp, which abuts on the S side of the steps at the level of the fourth riser, were contemporary. The single ramp floor corresponds in number and quality with the final ramp floor ending in Area Q. The lower level ramp floor to the N of the steps must have ended or descended at the point of the steps to the lower level to the W. The steps were covered with heavily burnt black earth, fused mudbrick, stones, and pieces of carbon. The lowest step sits 0.75 m above bedrock, at its highest. Small stones and a thin slab were wedged under the bottom tread, making it likely that no further steps are missing, despite the apparent height differential to preserved levels below and the lack of any sign of a plastered floor surface or path leading up to them. An earthen floor was first recognised just 0.04 m below the bottom of the fifth step but was later dug through and believed not to have been a floor. The multiple phases of building and use in the area (see below) allow for the possibility that the earthen floor was part of the Lower Ramp at the base of the steps. There is no direct connection between these steps in their present position and the two floors of the Lower Ramp (see below). The Lower Ramp and Passage J Passage J is the narrow access route from the Steps e to the Forecourt H and entrance to the House (Figs. 7 and 11). It passes between and is limited by the foundations of the basement rooms of the Megaron on the W and the terrace to the E that supports the Middle Ramp terminus, Area Q, and the Shrine. The Lower Ramp is located, in part, in Passage J and served in its later stage as a plastered surface sloping down to the House entrance. The plaster surface is preserved beginning only at a point 1.22 m to the S of the Steps e, and it consists of an early surface and two later floors, and it originally turned to the W to descend to the lowest level of the slope, the area of an open court, later enclosed by the WCW. The Lower Ramp is largely overlaid by the terrace wall supporting Area Q, especially its heavy extension to the NW, by the buttress/pier l, both clearly of later construction than the floor, the latter even blocking much of the entrance to the House, and the earlier levels by the foundations of the Megaron (Room IV) and the Forecourt H of the House (see above). The ramp area between the steps and the preserved start of the plaster surface is now exposed and uneven bedrock. At the base of the poros steps, the bedrock is 0.75 m below necessitating a much higher surface to transition from the steps to the preserved part of the Lower Ramp, 1.22 m to the S. As mentioned above, in 1959 an earthen floor was identified 0.04 m below the fifth and last step, and although it was believed later at the time to not have been a floor, that possibility should be considered. In Passage J, the plaster ramp floor, supported by fill and plesia over bedrock, was cut along its W edge for a drain that runs N to S in a bedrock channel just to the E of, and parallel to, Room IV’s E wall. The drain course crosses through the entrance of the House, crossing under the pier base in a stone lined channel, then heads S under the floor of Court A (see above). In the entrance area the ramp floor and the drain channel are covered with a surface of thick, coarse 74
Lower Ramp
Fig. 10. Plan with Lower Ramp indicated (after French 2002, fig. 31).
plaster. The finer plaster ramp surface extends under it to the S between the pier and buttress l, where remains of a burnt wooden threshold were found in 1950, associated only with the thick plaster in the doorway and the buttress, both from the final phase of construction, probably in the second half of LH IIIB2. The Forecourt H of the House is described in full in Part 1. Its floors are built together with the construction of the forecourt itself, with the closing of the area to the W with Wall M, and in the last phase with the expansion of the basement stairs G. Two earlier surfaces originally allowed access to the W and lower level (Fig. 10), from an open court, until some time in late LH IIIA2 or early IIIB. Several phases of wall segments and a mass of stone fill, associated with terracing and Wall J, cover the earlier access to the lower level (Pl. 15b). Two layers of the Lower Ramp were recognised also under Room IV of the Megaron that was built early in LH IIIB1 over the western sloping part of the ramp floor. AREA Z As his plan shows (Fig. 3), Tsountas excavated extensively in Area Z, the expanse to the NW of and outside basement Room F1 of the House. A number of wall segments are indicated, 75
most of which have not been reinvestigated since 1886. It was here that he found a tub-shaped lead vessel and near it, the famous piece of plaster painted with three donkey-headed figures (daimones) carrying on their shoulders a striped pole (Tsountas 1887, pl. 10.1). In 1950, Area Z was investigated again. It seems that Tsountas had excavated quite deep and little was found. The exception was a very burnt stone block in the SE corner with three drilled holes forming an equilateral triangle, believed possibly to be a tripod stand. A stone covered drain was also found running roughly NW to SE, which may be the continuation of a drain found in 1960 in the W face of Wall J. It was also confirmed that the cross wall between the exterior W wall of the basement of the House and the mass of accumulated and unexcavated fill inside of the WCW is indeed of Hellenistic date, like the Polygonal Tower, and was appropriately coloured pink on the 1886 plan. The wall was removed in 1974 revealing a corridor, > 4.2 m wide, between the House and the WCW, therefore allowing access from the lower courtyard area to Stairway K, ascending along the S of the House (Mylonas 1974, 90–91). THE WEST CYCLOPEAN WALL One factor that makes the Cult Centre a ‘palatial cult complex’ is its location within the citadel of Mycenae, enclosed by the West Cyclopean Wall, together with the rest of the lower western slope of the acropolis. The fortification wall acted as both a physical boundary as well as a social boundary, separating the complex from the wider community and restricting access to those individuals inside the citadel, presumably of elite status, or at least allowing them the control of access (Wright 1994, 51, 61–63). This situation is a basis for many interpretations of Mycenaean religion and cult, especially the division between ‘official’ and ‘popular’ ritual and belief (Hägg 1981 and Whittaker 1997, for example). The situation, however, is naturally more complicated because of the construction date of the WCW. The majority of the structures and features in the Cult Centre were built early in LH IIIB1 or certainly by the middle of the LH IIIB period, whereas the Shrine was built as early as LH IIIA. The WCW, on the other hand, was not constructed until after the middle of LH IIIB, possibly even in the last half of the period (Wardle 2003, 320–23; cf. Mylonas 1962, figs. 53 and 55; WBM 16/17). So the Cult Centre existed before it was enclosed within the citadel and would potentially have been accessible to the wider population of the Mycenae settlement and could be redefined as ‘popular’ cult. Originally an extramural cult complex, once the wall goes up, the accommodations for controlled access from upslope, within the citadel, become a reality. Following the destruction in the middle of the LH IIIB period, use of the Room with the Fresco, the Lower Courtyard altar, and the Temple is altered and in some cases ceased altogether. This may also be the result of the restricted or difficulty of access to these areas once the WCW is built, whether before or after the destruction, is hard to determine. A gated access to the citadel itself may have existed at the base of Stairway K and would have allowed some entry into the Lower Courtyard by way of the narrow passage to the W of the House’s basement into Area Z, or up the stairway to be directed up to the entrance of the Sacred Way, or further up slope. The inside of the fortification wall at this point has never been completely cleared of the LH IIIB destruction rubble, the LH IIIC levels or the Hellenistic overbuilding, so the possibility of a gate is mere speculation.
76
THE FINDS STAIRWAY AND DRAIN K Pottery The condition and source The material recovered from the drain in 1950, rather than the stairway, indicates that Tsountas probably did not excavate very deeply in the area.137 What was recovered, therefore, should represent the last Mycenaean periods of use of the drain and certainly any later use as well. The sherds are individual pieces with joins but are in general medium pieces and not particularly beaten up. Many sherds may have washed downslope during the lifespan of the drain or may be the result of disintegrated destruction debris over the area. The only whole pots are also miniatures, no bigger than the smaller sherds. The almost entire lack of burning on the sherds also argues for wash. The absence of preMycenaean and early Late Helladic sherds suggests that the area of the drain and Stairway K was heavily cleared prior to construction, possibly down to bedrock upon which both were founded. Pottery types and dates Among the 1950 sherds from ‘Street K and the Drain’ (Pl. 16), no certain MH or early LH examples were found and LH IIIA examples were very scarce; a kylix decorated with horizontal flower and stirrup jar shoulder with multiple stem. Nothing certainly LH IIIC was identified either. Post-Mycenaean pieces (7–8) include the miniature Hellenistic black glaze juglet (50-185, BE 7025; Pl. 17) and a piece of a Megarian bowl. Little was kept, especially from unpainted Mycenaean shapes; a few kylix fragments with rounded rims (FS 265 and 267) and several lipless rims from conical kylikes (FS 274) are exceptions. Among the LH IIIB decorated sherds, the following shapes were identified: piriform jars; amphoras or jugs; stirrup jars; kraters; cups; bowls, deep and stemmed; and a stand. The number of deep bowls was noteworthy; many of these had panel decoration, which occurred also on stemmed bowls and a krater. Some pieces can be dated to LH IIIB1, such as the stand (FS 336) but the majority belongs to LH IIIB2. The miniature stemmed bowl (50‑184, BE 5976; Pl. 7b), the only whole pot from the Mycenaean period, was found in the drain fill near its upper part and may be debris from Shrine G. Being handmade and only 2.7 cm high it is, like many miniatures, difficult to date very closely and can only be defined as LH IIIA2–IIIB.138 Only one sherd exhibits any sign of burning, from the body of a krater with pictorial decoration showing the hind legs of an animal, probably a horse. Another fragment from a different krater also shows the ‘knobby-kneed’ forelegs of a horse. A second group of sherds was collected in 1960 from the drain (T2C’60/48, BE 23273), in this case higher up, within the culvert. 70 sherds were collected from the drain floor and walls, from the mouth of the culvert up to a point 3 m to the E within the drain itself, beyond which access was impossible since the drain narrowed substantially.139 The majority of the sherds were unpainted Mycenaean, including several kylix rims (FS 265, 267, and 274), a shallow angular bowl (FS 295), and several dipper bases. Pre and Post-Mycenaean pieces were found in very small numbers, including two MH matt-painted and eight Hellenistic domestic sherds.
77
PLATE 16
(a)
(b) (a–b) 1950 fine ware sherds from Street K and Drain. Includes sherds from BE 23292, BE 23297 and BE 23296 (krater with legs of ?horse).
78
The decorated Mycenaean sherds date predominantly to the LH IIIB period, mostly to LH IIIB2 in the same way as the 1950s sherds from the lower drain course, although possibly earlier in the period than its end, but with a more significant number of LH IIIB1 sherds than recovered in 1950. Shapes represented include predominantly open shapes with one or two sherds only from a stirrup jar and a jug. Interesting examples include two LH IIIB2 deep bowls (FS 284), one with a flaring dotted rim with elaborate Sea Anemone and another unusual example, a very worn body sherd, monochrome inside with four vertical whorl tails reaching the upper part of a band; and a large, thin and flaring rim sherd of a polished mug (FS 225) with ridges at the waist, almost certainly LH IIIB1 in date. Among this sherd group, from a part of the drain that was covered, there was again no trace of LH IIIC, which suggests that the upper course may have been out of use or blocked following the destruction. The slightly earlier phase of LH IIIB2 represented by this group may also indicate material related to the construction and use of the culvert rather than the destruction phase in the area and above the threshold and drain covers. Other Materials Only a few fragmentary objects of materials other than pottery were found in association with the Stairway K and the drain, and all come from the 1950 excavation. Stone Conuli were found in two areas: near the top of the flight of stairs, N of the drain, was a conical, steatite conulus (50‑101, BE 5923), while in the area of the upper structure wall, S of the drain, were found three matching conical conuli (50‑169, BE 5943) in a red-purple steatite. They come from burnt soil together with two figurine fragments (50‑167 and 168, see below). Metal In the fill near the top of the stairs was found a lead object, possibly a ring (50‑114, BE 7023), somewhat irregularly shaped. More generally said to be in the area of the ‘Stairs/Drain (K)’ was a piece of lead sheet, rolled at one corner (50‑235, BE 7185), the function of which is unclear. Bead A single black steatite bead, incised on all sides (50‑6, BE 7010; Pl. 17), was found near the drain top, by the S side of the House. Terracotta An almost complete Tau figurine (50‑193, BE 7041; Pl. 17) came out of the drain fill, and a fragmentary animal figurine torso (50‑312, BE 7043) was said to have been found in the more general ‘Stairs/drain (K)’ area. A fragmentary Tau figurine (50‑167, BE 5978) and the festooned polos head of another female figurine (50‑168, BE 7040) were found together with the three stone conuli, in the structure wall to the S of the drain. Both showed signs of burning and the soil matrix in which they had been found was also described as burnt (MPW notebook, (Mycenae Archive x040), 24). 79
Other Two other very interesting finds were said to have been discovered in fill over the area of Stairway K, both made of painted plaster. The first is a fresco fragment with a helmeted goddess and griffin, discovered by Mylonas in the unexcavated fill at the bottom of the stairway, which he attributed to backdirt that originated from room G140 providing for him further proof that the building had a sacred nature. Mylonas also stated, ‘It is also possible to assume that the head of the so-called Mycenaean sphinx, found by Tsountas in 1896, was in the debris thrown out of the adyton.’141 Tsountas’s diary apparently identifies the find spot in the fill of the stepped way. THE UPPER RAMP Pottery The condition and source For the most part the pottery from all of the excavated segments of the Upper Ramp consists of small and battered sherds, most used with earth in the filling of bedrock and the packing under new or renewed floor surfaces. The quantity was not significant. Only one vessel was able to be mended up to a greater extent (60‑461) from a potentially disturbed context in a poorly preserved segment of the ramp. There was also a good quantity of sherds retrieved from heavily calcined debris covering much of the surface of the ramp and on which later building is founded. The debris results from the higher walls and likely roof over the ramp along with remains of structures and material upslope that burned hard in the massive fire of the destruction at the end of the LH IIIB2 period and then collapsed onto the ramp below. The resulting debris was extremely hard, like concrete or rock, and was very difficult to excavate with sherds very hard to extract.142 The combination of the destruction in this area together with a good deal of overbuilding in post palatial and post-Mycenaean periods has destroyed a large proportion of the stratified sequence of floors and ramps that became the main entrance and approach to the Cult Centre, at the latest in LH IIIB2. Pottery types and dates In 1959 only upper levels at the turn to the Middle Ramp (G4) were excavated. Units from the Upper Ramp were recovered in 1960 and the lowest levels were accessed in segments of the ramp which were damaged or missing entirely. In addition, a 1.0 m test cutting, opened along the W edge of the ramp at c. 6.0 m N of threshold m, exposed four floors below Ramp 2 and bedrock, and allowed the collection of sherds associated with the various layers and surfaces. The first phase: In the very small area of the main cutting (Pl. 12a) (1.00 m long in the fourth phase, then progressively reduced downwards in order to leave parts of the three earlier phases showing), bedrock was packed with 0.45 m of earth beneath the earliest floor or surface made of pebbles only. The sherds in the packing, although almost entirely MH, included a few of LH date, linear and monochrome, with examples very likely of early LH III. Just S of the main cutting, an investigation of bedrock W of the eroded edge of the ramp also produced mainly MH pottery (including late pieces: fine orange Minyan and fine polished pale buff) and one transitional or 80
LH I piece (matt white on lustrous dark). There was also LH IIIA2 including good monochrome kylikes (FS 264), a stemmed bowl (FS 304), and two domed and banded kylix bases (FS 257), but nothing obviously later. Elsewhere also LH IIIA sherds were found just W of the ramp in conjunction with plentiful MH on or near bedrock. Many sherds were very small and battered, as though deliberately pounded before being used as packing.143 LH IIIA2 red monochrome kylix fragments occurred with MH sherds immediately W of the jog in Wall A in an area where none of the ramp floors was preserved. Though this evidence is inconclusive, the finding at several points of LH IIIA2 pottery sherds on, or a little above bedrock, in the area later covered by the Upper Ramp as well as immediately to the W, suggests that the first phase of construction of a road or access point, possibly even a ramp, may be as early as LH IIIA. The character of this material is strikingly similar to that of the sherds found in the Shrine, under its lower floor and in the units around the large stone emplacement. It includes predominantly MH sherds and a few of LH IIIA date, and may provide further evidence for the first phase of construction in the area. The MH sherd material recovered from almost all pottery units excavated in the area of the Upper Ramp, especially in the lowest level of packing, is remarkable for its quantity, variety of styles and high quality. The sherds most likely represent domestic refuse from affluent households in the neighbourhood and upslope to the E, such as the remains of a MH grain store excavated by Mylonas in 1970 (Mylonas PAE 1970, 120). The later phases of the Ramp, culvert, and threshold m: The other floors identified in the test cutting were very difficult to date due to the very small number of closely dateable sherds. In the fill and packing between the lowest surface (5) and the fourth floor of plaster and pebbles, only MH sherds were found, most of them of late MH with good examples of grey and yellow Minyan. Above Ramp 4 was yet another packing fill with a preponderance of MH material and again, sherds from LH IIIA, such as the banded body sherd of a small piriform jar (FS 44 or 54) and the rim of a one-handled bowl (FS 283 — LH IIIA2). Because sherds of LH IIIA date were associated with both the lower level of packing under floor 5 and the packing for Ramp 3, the absence of LH sherds associated with Ramp 4 is likely due to the small sample of the cutting. A handful of sherds was recovered from over an area where the Ramp 3 surface was identified, 2.0 m N of where it descends below the stones of the culvert, and from the packing for Ramp 2 above. These include LH IIIA, IIIB1 sherds and a single deep bowl body sherd (FS 284) with brown monochrome paint on the interior and surely dating to LH IIIB2. This would date the construction of Ramp 2 to LH IIIB2 and its possible association with the culvert would indicate the date of the building of the culvert in its final form, in which the penultimate phase of the ramp was raised c. 0.65 m by the addition of fairly large stones above Ramp 3 at its S end beside the drain channel (see above). To this same period, but a little later, belongs the laying of the conglomerate threshold m and the construction of the prothyron as well as the one plaster floor that leads to the threshold. The test made under the post-Mycenaean terrace wall between the second and the final ramp floors, overlaid by the destruction debris of the end of the LH IIIB2 period, revealed that MH sherds continued to be present in floor packing, with 14 pieces among the predominantly LH IIIA sherds with a only a few LH IIIB examples and none later than the middle of the period. There is no doubt, however, that Ramp 5 belongs to the second half of LH IIIB and was built 81
with threshold m, following the building of the culvert in its final form. Further confirmation comes from the sherds below the single plaster surface S of the threshold where among the 22 sherds was a deep bowl handle (FS 284) with very worn, dark monochrome paint on the inside of the attached piece of bowl. Elsewhere from the area of the Upper Ramp, few undisturbed or uncontaminated units were found. The many gaps in the various ramp surfaces led to post-Mycenaean material penetrating quite deep in the stratigraphy. The few sherds excavated from seemingly undisturbed fill below the plaster surface near the N end and a group associated with Wall R along its second segment provide dates in the first part of LH IIIB. Trench T2B contained several layers and features from phases of the LH IIIC period under post-Mycenaean construction and produced a good deal of confusion and mixing of the stratigraphy. At the N end of the Upper Ramp (G4) was an earthen floor, that dates to LH IIIC, built over the destruction debris which marks the end of LH IIIB. This debris had not been disturbed down to the ramp surface, which is remarkably well preserved at this point, although burned, with a thin black soil layer remaining on the surface. Stone base n: Additional dating evidence for a final ramp surface in the second half of LH IIIB was provided by the test around the stone base n. Only two phases, or surfaces, of the ramp at most were found for the stretch on which the base stands. It was suggested by the excavators that both ramp and retaining Wall R were rebuilt here (segment three that does not bond at either end with the rest of the retaining wall) at the time of the construction of the gated entrance to the S and for this reason the ramp has only two phases. The upper plaster surface is level with the top of stone base n; the lower floor (c. 0.13 m deeper) was apparently cut into to accommodate the base. The packing between these floors yielded a few sherds that are all LH. These included a deep bowl (FS 284) of general IIIB date and the base and lower body of a linear basin (FS 294) that belongs certainly to the second half of LH IIIB with very worn brown monochrome paint inside apart from a reserved centre. Just S of the stone base, below the level of the upper plaster surface, in a spot where it was not fully preserved, several LH IIIB2 sherds were identified but so were a few post-Mycenaean and a possible LH IIIC piece indicating some contamination. Clearance around the base itself yielded sherds that were all LH and included a stemmed bowl rim (FS 304) and two deep bowl body sherds (FS 284), all with linear decoration and belonging to LH IIIB, but not more closely dateable. It may be imagined that the installation of the base belongs to LH IIIB, probably its second half, and may belong to a rebuilding or redesign of this segment of the Upper Ramp. The only restorable pot from the Upper Ramp, a collar-necked cooking jar (60‑461, BE 21758) of Handmade Burnished Ware (HBW) was found scattered at the broken W edge of the mud matrix packing under the upper of the two surfaces at this point and wedged in stones along the E side of Wall R. It is fragmentary with about 1/3 missing including part of the convex base, much of the neck and possibly a handle. The burning on the lower part shows that this jar had been used. When discovered it was associated with the preparation for the laying of the final floor in this segment, probably during the second half of LH IIIB2, however the evidence for LH IIIC building and disturbance, including a potential hearth (Mylonas 1966b, 109–10; Fig. 4), would allow this dating as well. Even though found scattered and fragmentary, the fact that it is the only 82
PLATE 17
50-6, scale 2:1
50-185
50-193
60-179, scale 2:1
60-178
60-168
59-172
59-510
60-61
Tau figurine (50-193), steatite bead (50-6) and miniature jar (50-185) from Drain; obsidian point (60-179), animal figurines (60-168, 60-178) and stone loomweight (60-61) from Upper Ramp; horse figurine (59-172) from Middle Ramp; unusual figurine head (59-510) from Lower Ramp. Scale 2:3 except where indicated.
83
restorable ‘whole’ pot from this area would argue for a domestic context, even if a secondary deposit or discard. There was much disturbance, however, of surfaces and fills, often at or below the level of the stone base and the plaster ramp surfaces. Much of this may date to phases of LH IIIC, both early (further N along Wall R) and late (N end of Trench T2B, below plaster surface) but postMycenaean contamination is pervasive. Features above the level of the plaster floors and stone base n include a short wall segment perpendicular to Wall A and a plesia patch c. 0.25 m higher than the plaster floor that dates also to a later phase of LH IIIC. Further excavation by Mylonas in this area uncovered additional cut stone bases, short walls, floors and hearths of plesia; all are said to date to several phases of LH IIIC including stone base n, its counterpart to the N, and the remains of a hearth ringed in plaster between them (Mylonas 1966b, 109–10; Fig. 4). Other Materials A number of small finds in various materials was recovered in the excavation of the Upper Ramp and the many fills above it. None were from a primary context and most were from mixed levels. Stone A stone loom weight dating to the Hellenistic period (60‑61, BE 17028, Pl. 17; WBM 36, Area Q) was found in trench G4 in the initial level. Two steatite conuli were found (60‑63, BE 9914 and 60‑177, BE 10003), both conical shaped, and both from mixed contexts. Other ground stone objects include a smoothed grey stone, pierced near one edge that might be a pendant or tool (60‑161, BE 9999); a stone vase fragment, possibly the stem of a lamp (60‑169, BE 10006; WBM 27, #31); and half of a dagger pommel in lapis lacedaemonius (60‑172, BE 9994; WBM 27, #42). All three came from levels with post-Mycenaean materials but are certainly of LH date. All three ground stone objects were found towards the N end of the ramp in trench T2A. Finally, two chipped stone objects were found in the deep test cutting in trench T2C to the S end of the ramp. An obsidian point (60‑179, BE 9998; Pl. 17) from the packing below Ramp 3 and an obsidian blade fragment (60‑180, BE 10122) in the fill on bedrock under the lowest pebble floor 5 could date to as late as LH IIIA. Metal Lumps of lead were found in several areas of the Upper Ramp, especially towards the northern end in trench G4. Three small amorphous pieces turned up in levels of mixed date (60‑171, BE 10000; 60‑175, BE 9997; 60‑176, BE 9905) and their original source and form are unknown. Lead dating to the Mycenaean period and probably in use on or near the Upper Ramp surface consisted of 53 melted lumps (60‑65, BE 10028) found at its N end and within a thin layer of black soil on the ramp’s plaster surface, believed to be the remains of burnt wood. The lead may have been part of the walls, ceiling, or other feature of this area. Other metal objects found were likely of post-Mycenaean date and include the fragmentary head of a bronze pin (60‑162, BE 9908) and an iron ring or link of a thin chain (60‑173, BE 9904). Clay Bead Half of a round bead (60‑170, BE 9907) was found in a mixed topsoil level in trench T2B.
84
Terracotta Several fragmentary terracotta figurines were found, all in levels of mixed Mycenaean and post-Mycenaean date. One columnar stem (60‑166, BE 9913) and two legs of different animal figurines (60‑62, BE 10013; 60‑66, BE 9952) could not be further detailed or dated stylistically beyond LH IIIA/IIIB. Three female figurines may be further described: a proto-Phi torso and stem (60‑174, BE 10005) should date to the early LH IIIA period, the upper torso fragment of a Phi figurine (60‑164, BE 10007) may be of later LH IIIA or early IIIB date, and the torso of a Phi figurine of type B (60‑64, BE 9952), found in a mixed level near the N end of the Upper Ramp dates to LH IIIA2. This example has been identified by distinctive features, such as the brushwork on the plait, as a product of ‘Hand A’ figurine maker from the Petsas House ceramic facility in the settlement of Mycenae (Shelton 2009a, 59). Three animal figurines were also identified: the body of an unusually decorated quadruped with a transitional Wavy 1 and Wavy 2 scheme (60‑178, BE 10018; Pl. 17) probably dates to late LH IIIA or early IIIB, while the head and torso of a Spine 2 type (60‑168, BE 9975; Pl. 17) more likely dates to the LH IIIB period on stylistic grounds, the third animal figurine is that of a horse from a chariot group (60‑67, BE 9919) already published in Tamvaki 1973, #107. Other One other object was found in the fill above the ramp, an unguentarium or ‘tear bottle’ from the Hellenistic period level in trench G4 (60‑361, BE 23845; WBM 36, 331: Rudolph #56). THE MIDDLE RAMP Pottery The condition and source As with the Upper Ramp, the sherds from the area of the Middle Ramp originate for the most part from destruction debris of structures further up the slope during or more likely after the event at the end of LH IIIB2. No objects were found in direct association with the use of the ramp, only the architectural features outlined in the previous section and a possible lead vessel (59‑175, BE 9880) found near the surface in a very poor state. Here too, the debris was disturbed to some extent and built over in succeeding periods; in the case of the Middle Ramp there was little to no post-palatial evidence, while Hellenistic remains did exist, especially over the N end. Two further issues in this area involved the early excavation of Tsountas down to the floors, or very near them, in the court area of the S end of the Middle Ramp before the Shrine, and the vast dump of mixed material that was then re-deposited over much of the area. This dump on the Middle Ramp was cleared away in several seasons including 1950 and 1959. So the bulk of the sherds (found primarily in 1959) come from the destruction of the ramp and its surroundings, especially from trench G4 at the N end, with the few explorations below the plastered surfaces and features having produced very little narrowly dateable finds and little that was not contaminated by later periods. Pottery types and date The bulk of the sherd material comes from units at the N end of the ramp in trench G4. Masses of post-Mycenaean overbuilding and dump were cleared from over the earlier levels and features.
85
A few Geometric and Archaic sherds turned up among essentially Hellenistic levels down to the Mycenaean debris. The destruction debris was made up primarily of stone and mudbrick, some burned hard and more disintegrated or eroded. Some sherds surely originated from the mudbrick while other pieces had accumulated in the destruction itself. Most of the sherds date in the LH IIIA–IIIB range with a higher number of early to mid IIIB among decorated pieces. A few recognisable LH IIIB2 sherds, such as deep bowls (FS 284) with a dotted rim and others with a monochrome interior, appear in the units especially among the burnt soil on the ramp itself. The same can be said for the few sherds found at the S end of the Middle Ramp. From the square test cutting into the final ramp surface N of Area Q, only a handful of sherds was found. These date from MH to LH IIIA, as did the stem of an early Phi figurine (59‑502, BE 9853). In the area of the poros steps, while clearing to the N in an area of heavy burning that may be the remains of the wooden stair post, sherds predominantly of LH IIIA and IIIB date were identified, with no MH or early LH examples. A single deep bowl of rosette type must date to LH IIIB2 while one Hellenistic sherd was also present. From over the steps themselves, but also from the fill flanking them on either side, came material that primarily dated to the first half of LH IIIB with a single MH piece and one post-Mycenaean intrusion. Over 30 deep bowl fragments were identified, mostly LH IIIB1 in date but there were also a couple of the rosette type with dotted rim, which indicate the presence of some LH IIIB2 material. The greater part of a deep bowl (FS 284, 59‑527, BE 15946) decorated with running spirals (FM 46), which was restored from several large fragments, belongs to the same period. Joins were also made to material from just below the bottom step in an ashy soil. Ceramic evidence for the construction of the Middle Ramp and the poros steps is not secure, but seems to indicate a date in LH IIIB. Wall C will be dated by the construction of the Megaron and Area Q. The final floor of the ramp was laid before the step-platform in Q and therefore should be dated during or before LH IIIB2. The steps could have been added later in the series of remodellings here, possibly after the construction of the two buildings on the Lower Ramp (the Megaron and Tsountas House) in early to mid LH IIIB, and could be contemporary with the packing stone around them that appears to cut through the terrace wall to the S (mid LH IIIB to IIIB2). They were certainly in place before the final ramp surface was laid over the top tread in LH IIIB2. The destruction date for the Middle Ramp and steps is the same as elsewhere in the Cult Centre, at the very end of the LH IIIB2 period. Other Materials A limited number of small finds in various materials was recovered in the excavation of the Middle Ramp and the fills above it. None were from a primary context and most were from mixed levels. A few were from previous excavation dumps. Stone A single conical conulus (59‑169, BE 9835) in a green grey steatite was found in a mixed level high above the N end of the Middle Ramp. Another fragmentary example (59‑508, BE 9808) in dark blue steatite was found in the upper level of the steps. Metal Three pieces of melted lead, probably the remains of a vessel (59‑175, BE 9880) were found on the ramp close to the ramp surface and near Wall C, 2.4 m N of the slab f. 86
Terracotta Objects in terracotta found in the area of the northern end of the Middle Ramp were primarily fragmentary figurines, both female and animal types: the columnar stem and base of a female figurine (59‑171, BE 9774) from a mixed level; two heads of female figurines (59‑166, BE 9776; 59‑167, BE 9861), certainly LH IIIA in date, the first of an early Phi or Proto-Phi with applied eyes and the latter probably from a Phi, type B figurine, both from debris of mixed origin including building debris; a possible LH IIIB female figurine head with a tall polos and festoon decoration, also from building debris above Wall C (59‑173, BE 9863); and an interesting fragment of a horse figurine (59‑172, BE 9806; Pl. 17), from near the female figurine base (59‑171), where the forequarters are preserved along with the attachment mark of a rider, once applied to the animal’s back, and the tips of his legs. Another figurine stem, an early Phi or Proto-Phi type (59‑502, BE 9853), comes from the test square cutting in the Ramp just N of Area Q and dates to the LH IIIA period. Clay disc A small circular disc (59‑170, Di. 0.025 m) with a squared edge and an incised line around circumference of the type often called a counter, although it may have been a very light weight was found in a high mixed level in trench G4 towards the N part of the area. THE LOWER RAMP AND PASSAGE J Pottery The condition and source The condition of the material recovered in the area of Passage J and the Lower Ramp was greatly affected by the earlier excavation in 1886 of almost the entire area, at least to the wall tops, and the subsequent dumping of backdirt and sherds over the same area and beyond where much accumulation of eroded fills had already occurred. The re-clearing in 1950 and the completion of the excavation in 1959 and 1960 further complicated matters as destruction debris and later disturbance was dug in a piecemeal fashion that did not allow a true understanding of the depositional processes. This work was also plagued by issues of contamination from the mountainous dump that towered over and occasionally eroded into areas under excavation at deep levels. Very little was found on the ramp or in association with it that can be directly related to its lifespan and use. The bulk of the material, as was seen in the other ramp segments, was from the destruction debris dating to the end of the LH IIIB2 period. Evidence for the construction and use of Passage J in different periods was mostly lacking. Pottery types and date Pottery from the ashy debris level (Level 4) at the base of the stairs produced almost entirely LH III sherds with only a single MH piece and one Hellenistic intrusion from the upper W part of the trench that certainly trickled down from the dump. There are a few LH IIIA pieces but most are LH IIIB with a few deep bowl fragments (FS 284) dating to the second half of the period. The collapsed building debris that characterised this level probably came from the superstructure above and around the steps e just to the E. Since the Megaron was built early 87
in LH IIIB, the later sherds may indicate the date for the steps themselves, or could as easily have come from material in use at the time of the destruction. A single fragmentary shallow angular bowl (FS 295) (59‑512, BE 15943) was uncovered 0.20 m N of the steps and at the level of the fifth step riser. It is undecorated but carefully slipped and dates to the LH IIIB period generally. Level 4 sat on a tramped earth layer that was potentially a floor surface (see above) running from the base of the poros steps e over a packing fill on bedrock S to the edge of the plaster ramp surface. During excavation, the interpretation of the earthen floor was in doubt and the level was excavated as one with the fill underneath (Level 5) and without being separated from adjacent deposits of later date. Accordingly it became contaminated with a number of Hellenistic sherds. It is therefore impossible to say whether the late LH IIIB sherds provide important dating evidence for the laying of the possible floor in that period after the construction of the steps, or are simply additional contamination. When the floor level and that below were dug they included a few Hellenistic pieces but this was probably contamination from further W where Tsountas’s dump had not been entirely removed. It should be noted that a few pieces of the LH IIIB2 deep bowl (59‑527, BE 15946), associated with the steps themselves (see Middle Ramp, above), came from this level and may have fallen down onto the earthen surface. Even deeper, from 0.30 m below the steps to bedrock at a maximum depth of 0.79 m, and below the level of the Lower Ramp, only Mycenaean sherds were found, dating to LH IIIB, but early in the period which are likely to date the construction of the Megaron and House together with the initial use of Passage J to that period. Additional ceramic evidence was collected further S in Passage J, over the Lower Ramp and in association with a number of stones fallen from above, possibly from the terrace supporting the Middle Ramp to the S of the poros steps. Unfortunately, all of the units were either of mixed periods or were contaminated by dump containing ample post-Mycenaean material, including a fragment of a black glaze lamp. Also from the area of Passage J was a single unit collected in 1950 at the base of the buttress l and the terrace wall of Area Q. The sherds date to LH IIIB1 and are free of post-Mycenaean contamination. Sherd evidence from under the Megaron building (Room IV) and Forecourt H of the House indicates those two buildings were constructed over the earliest ramp surface sometime after the LH IIIA2 period following the construction of a terrace filled with stone, as well as some earth and domestic rubbish. Limited fill under the ramp surface was recovered and contained mostly MH sherds with only a very small number of plain LH III examples — a few probably also of IIIA date which indicate that the construction of the ramped access from the lower courtyard to the Shrine terrace probably took place during this initial use of the area.144 Other Materials Metal A small sheet bronze appliqué (59‑513, BE 9727) with a raised design comes from a mixed level within a fall of stones over the Lower Ramp in Passage J. It may date to the Hellenistic period. 88
Terracotta Three interesting figurine fragments were found in the dump over the general area of Passage J and the Lower Ramp: (59‑509, BE 9728) a horse head from a chariot group with reins still attached on either side of the neck; (59‑511, BE 9817) the right forequarters of a horse figurine, also from a chariot group; and (59‑510, BE 9818; Pl. 17) an unusual female head with a large festooned polos. All three fragments could date in the late LH IIIA to IIIB range. Pottery lid/stopper Also found in a mixed level in Passage J was a lid or stopper (59‑514, BE 9814) fashioned out of the base of a LH IIIA2–IIIB1 mug (FS 255). AREA Z Pottery The condition and source This area on the lowest level below, and to the W of the House, was dug extensively by Tsountas in 1886. As mentioned above, it was here that he found a tub-shaped lead vessel and near it a wall painting fragment with three donkey-headed figures (Tsountas 1887, pl. 10:1). He did not report any other finds. In 1950, Wace excavated here again in the hopes of finding further pieces of the fresco, but none were found. Only a handful of solid colour fragments of painted plaster were found and one with a red stripe (50‑298, BE 24286). This low-lying area at the base of the slope within the fortification wall would naturally accumulate eroded debris from further up the slope. Based on the depth of fill and quantity of material excavated by later excavators at the N end of this court, across from Area Z, Tsountas must have removed a substantial amount. Even so, a large unit of sherds and a few small finds were found during the 1950 season, which originated primarily from the collapsed building debris. Pottery types and date There was a good mix of coarse and fine wares, decorated and unpainted from Area Z, although only a few undecorated kylikes. All later post-palatial and post-Mycenaean remains must have been removed by Tsountas since only Mycenaean sherds were found as well as only a few MH examples. The sherds were mostly LH III, a good deal of IIIA and some IIIB, though almost entirely LH IIIB1 with only a couple of sherds (a stemmed bowl rim and an undecorated conical kylix) belonging to the second half of the period. The LH IIIA and early IIIB ceramics certainly come from the mudbrick debris and reflect the building periods, while the small number of later destruction date may suggest that this area was no longer receiving heavy traffic, something also suggested by the difficulty of access at that time. It may also be due to disturbance from earlier excavation. Other Materials Stone A poros stone block was found in the SE corner of the area of the stone-covered drain. It had three holes in its upper surface in a triangular formation and was thought to have been a tripod 89
stand. It had been damaged by fire and had fallen out of its original position, but was left on site. Nearby was found a large shanked conulus (50‑292, BE 7092) in dark grey steatite. Metal Two small fragments of lead were found in the area (50‑297).145 Terracotta Three figurine fragments were identified: the head and forequarters of a Linear 1 bovid figurine (50‑293, BE 7082); the partial body of a Spine 1 type animal (50‑294, BE 7107); and the head of a female figurine (50‑296, BE 7067) wearing a festooned polos. All three examples may date from the later LH IIIA period to LH IIIB.
SUMMARY AND USE Until the citadel wall was built after the middle of LH IIIB along the lower W slope of the hill, the area of the Cult Centre was potentially accessible from any point outside (Fig. 8). The slope itself would have had to be navigated, especially as buildings were constructed within the complex in early LH IIIB. Earlier in LH IIIA, the Shrine was fairly isolated in this area and its entrance on the N was accessed either from downslope by way of a clay and earth surface on soil-packed bedrock replaced later by a plaster-surfaced ramp leading from the lowest terrace (Area Z and the court to its N) across the middle terrace and reaching the level of the Shrine just to the N of its entrance (Figs. 8 and 9), or from the S along the slope just E of the shrine on a surface of pebbles (Ramp 5) and later by pebbled-plaster (Ramp 4) in the area of the Upper Ramp (Pl. 12a). These surfaces are not necessarily ramps but seem to bring traffic along the slope and perhaps up towards the entrance to the earlier citadel in the area of the later Great Ramp and the precursor to the Lion Gate. Access from further up the hill and from the early citadel could also have been by these routes. During the long and prolific building phase in early LH IIIB, internal circulation between terraces and among buildings becomes a priority and the area becomes a ‘centre’ as the individual structures are interconnected in, perhaps, a planned way. The routes and interconnected access ways were also altered with the construction of complex buildings and the strong terraces that supported them (Fig. 8 and Pl. 1). As the Lower Ramp is impinged on by terracing for the construction of the Megaron and closed off entirely by the construction of the forecourt of the House, direct open access from upslope to the lower courtyard, the Temple, and the Room of the Fresco complex was eliminated, at least through the middle of the Centre, while entrance was seemingly unrestricted around the perimeter. This building on the middle level and the closure of the Lower Ramp further W together created a barrier to unrestricted access from the lower slope still outside the protection of the fortification wall. During LH IIIB the poros steps e were built with the terrace supporting the Middle Ramp and entrance into and through the Megaron was achieved by way of a probable wooden staircase originating from the threshold slab c. This would have provided access to the buildings on the lowest level, which were no longer accessible from the Lower Ramp. A fine hard plaster was used on surfaces at every level including on Upper Ramp 3 that ascended from S to N and still may have provided a route towards the main entrance and ascent to the citadel further N.
90
Fig. 11. The Cult Centre, with access and circulation routes in LH IIIB2 (after Iakovidis 1997).
Access and entrance to the Cult Centre seems open and fairly unrestricted until after the middle of LH IIIB, and probably after the earthquake destruction of that phase, when the West Cyclopean Wall is built across this part of the slope and the Centre becomes an internal part of the citadel (Fig. 11). Up to this point the buildings could been seen and approached from all directions on the hill with the most limited approach perhaps being from above, while the activities and rituals acted out within the structures were invisible and restricted to small numbers of individuals. The WCW closes free and unrestricted access to the Cult Centre from the W. Area Z and the lower courtyard may not have been entirely cut off if there was originally a gate or opening in the WCW as is often suspected. This is particularly likely because of the Stairway K, which was built later in LH IIIB along the S wall of the House and indicates a need to bring traffic from the area at the base of the WCW to further up the slope, or the reverse, traffic descending presumably to exit through the wall. An argument against a gate would be the changes in use of the lower courtyard altar and Room of the Fresco, both of which go out of use following the destruction in mid LH IIIB and could be a strong indication that access was increasingly difficult. Upslope the situation is different. Even as the upper terrace of the Shrine is expanded and buttressed for reasons of structural stability, probably in reaction to the earthquake damage,
91
Area Q is created in front of the entrance to the Shrine and an elaborate ashlar-based altar is constructed in a potentially more visible and possibly more accessible area. This occurs during the later part of LH IIIB at the time when the entire upper terrace is monumentalised with the Upper Ramp and its support Wall R, a frescoed prothyron structure with a double-gated entrance over a built culvert and pseudo-Cyclopean tower protecting the entrance area to the Cult Centre and the S end of the Shrine both from the destructive power of nature and any potential human threats. The Upper Ramp is carefully built with a wide and easy descent, covered with surfaces of fine hard plaster. This approach could have facilitated a high volume of traffic access and yet the series of doorways suggests a hierarchy of access. The external ritual seems to be more visible yet proximity is restricted. Access is never more important or controlled than in this period when the Cult Centre becomes a physical part of the palatial sphere. The focus is the long approach down into the Centre through the multiple barriers and towards the Shrine, which culminated in an open plaster-surfaced court with a long well constructed bench in view of the ashlar altar and the façade of the Shrine. The nature of the Ramp suggests that an extended and highly visible entrance was possible and that its use for procession is certainly possible. Evidence for processions is rich in the iconography of the period. The upper part of the ramp and the stairway upslope excavated by Mylonas provides further proof of a direct connection to the upper levels of the citadel physically and conceptually. The Cult Centre and ‘Sacred Way’ were destroyed at the end of that period, LH IIIB2. There is no evidence for repair or rebuilding following the destruction and in fact, there is good evidence for exposure of the ruins and erosion down the slope of debris in the large masses of hard burned structures and lenses of mixed origin. The recovery of Hellenistic and earlier post-Mycenaean remains as well as post-palatial Mycenaean material showed that this area had had subsequent occupation during several phases of LH IIIC, for the most part not immediately following the disaster, but also well into the post-Mycenaean period. No attempt had been made before this to clear the mass of destruction debris and the ramps themselves were never used again.
92
PART 4. CONCLUSION The Cult Centre of Mycenae located midway along the lower western slope of the fortified citadel consists of five complex structures, four and possibly all five of them probably religious in nature, with access to the area by enclosed ramp and stairway (Figs. 1–4 and 7). From E to W, the religious structures are: Shrine G, the Megaron, the Temple and the Room with the Fresco complex. The remaining building is Tsountas House itself — the function of which is not as easily defined. The Tsountas House Area includes both the origin of the Cult Centre, with the earliest religious installation known at Mycenae, and its fullest development as an organised and controlled precinct within the citadel. This investigation has analysed the conception and function of Mycenaean religious space along with its associated material paraphernalia and provided insight into aspects of Mycenaean cult in its chronological and cultural context. The diachronic trajectory of these structures and features presents a narrative that directly impacts on the dialogue between ‘popular’ and ‘official’ religion, and the role of the palatial administration.146 The chronology of the Centre has not been very well understood, especially the various phases of construction, use and destruction, and the general impression among scholars has been that the area was not really occupied until the 13th century BC (Iakovidis 2004, 15). The fascicules published already in this series have clarified the situation substantially by detailing both generally and more specifically the complex lifespan of the buildings, features, and materials, but it is only with the close study of the Shrine and its interpretation here that the picture is completed and the beginning of ritual architecture on the slope is identified in the 14th century BC. In light of what we now know about the history of Shrine G, it is important to recognise its position chronologically within the Cult Centre and to clear up some important misconceptions about the area and its use. When most people think of the Cult Centre of Mycenae, they think of it in exactly that way — the ‘palatial’ religious centre within the citadel acropolis of Mycenae. With four independent structures, internal and external ritual areas and the ‘processional’ ramped approach that may have originated from the palace at the summit of the hill, it is the ‘type site’ for Mycenaean palatial cult. It is also considered to be a phenomenon of the LH IIIB palatial system and of the expansion of the citadel at that period. The problem with this reconstruction is that the Centre never existed in the way that it is often pictured — especially as presented in general state plans and views. There are five complex structures, there are monumental approaches and there is a citadel wall — but they were never actually a total package, whether at the inception of the site’s use for religious practice, or at the time of the Centre’s final destruction. Following the widespread use of the hill slope for habitation and burial during the MH period, the systematic construction of cult buildings begins with Shrine G in LH IIIA in a sparsely populated area on a steep slope of unlevelled rock a bit S of the Ramp House and even further along from the old dilapidated royal cemetery — Grave Circle A — but definitely outside the walls of the LH IIIA citadel (Fig. 7). The approaches to the building were probably from below, downslope, and may have been linked to the roadways approaching the citadel from the SW. Early ramped surfaces ascend the slope towards the N entrance of the building (Fig. 10). The Shrine (Fig. 5), with a remarkable altar before the entrance to its adyton, Room G, is a two-room structure, with its original walls preserved only in G itself. The Shrine entrance lay 93
on the N and, if it had a court, this was of levelled but unplastered rock. The finds potentially associated with the adyton, or even the main room, may be the remains of offerings from later LH II up to the early part of LH IIIB. The walls and their terraced support exhibit several phases of construction but can be dated in sequence only on architectural grounds. Sherds from between the two floors suggest that the major alteration of the Shrine with the covering of the earlier ritual features with fill and the laying of a new plaster floor took place at the earliest during the middle of LH IIIB but may have been in part a result of damage from the earthquake at the end of Phase VII. Area Q and the ashlar-based feature with a plaster step platform in front of the Shrine were then constructed during the late LH IIIB period replacing the function of the interior of the building (Pls. 14, 15). The Shrine building and Area Q were destroyed by fire, presumably in the general disaster at the end of the LH IIIB period. By the time the full group of structures had been built towards the middle of the 13th century BC the Shrine was joined by two multi-storeyed structures with rectangular hearths and extensive storage basements on the next terrace down to the W and at a lower level flanking a central open space were two independent building complexes, also with clear indications of cult use (Fig. 8). The area could then be approached up the stairs (K) to the S of the House, from the N by way of the IIIB1 causeway through the roofed and plastered corridor to the E of the S House Annex and from the W through the open court with the circular altar. The House has a construction date early in LH IIIB, structurally probably a little after the building of the Megaron to its N. There are at least two phases of construction exhibited in several areas of the building, both in design and in the details, that were made perhaps in the first half of LH IIIB but also certainly in the later part of the period as well. There was no real trace of the earthquake destruction at the end of Phase VII, witnessed elsewhere in the Cult Centre (cf. WBM 13). The long life of the House, after the early LH IIIB alterations, was uneventful until the great disaster that overtook the entire neighbourhood at the end of Phase VIII, in LH IIIB2 late. The House may have declined in use before its destruction, together with the structures that surrounded it. This might explain the lack of important finds in the building when it was excavated, however, it should be stressed that very little has actually been found in any of the destruction contexts across the Cult Centre and the Citadel House Area, which is in striking contrast to the remains from the earlier LH IIIB middle destruction in the Temple and Room of the Fresco complex for instance (WBM 10, 83–84). Until the construction of the WCW after the middle of LH IIIB the Cult Centre was potentially accessible from any point outside the citadel (Wardle 2003). Earlier in LH IIIA, the Shrine was fairly isolated in this area (Fig. 7). An earlier road or ramp may have traversed this area to the E of the Shrine as a pebbled plaster surface indicates. This may, however, be a more general approach towards the entrance to the citadel and a reason to build on this slope. Access to the Shrine seems to have been primarily from downslope, whether from the W or the S. During LH IIIB1 and the building of the Cult Centre, internal circulation between terraces and among buildings is the focus of constructed access rather than an emphasis on approaches from outside. The routes and interconnected access ways were occasionally altered with the construction of complex terraced buildings, sometimes limiting easy access within the Centre while seemingly allowing unrestricted entrance from outside. Excellent plaster surfaces were used throughout in this period providing another indication that the construction was planned and organised to some extent and that a good deal of effort and expense was put into this 94
program. This may already be a sign that a level of administrative control from the palace was already present even if the complex was still very unrestricted and unprotected on the slope. As was discussed above (Part 3), access and entrance to the Cult Centre seems open and fairly unrestricted until after the middle of LH IIIB, and probably after the earthquake destruction of that phase, when the West Cyclopean Wall is built across this part of the slope and the Centre becomes an internal part of the citadel (Fig. 11). Up to this point the buildings could been seen and approached from all directions on the hill with the most limited approach perhaps being from above, upslope, while the activities and rituals acted out within the structures were invisible and restricted to limited numbers of individuals. The WCW closes off free and unrestricted access to the Cult Centre from the W. The changes in use of the lower courtyard altar and Room of the Fresco following the destruction in mid LH IIIB, both of which cease to function and are covered or filled, could be a strong indication that access was increasingly difficult. The upper terrace of the Shrine is expanded and buttressed for structural stability, however, a larger, wider area or court is created in front of the Shrine. An elaborate ashlar-based altar, or offering table, is constructed in a potentially more visible and possibly more accessible area. This occurs during the later part of LH IIIB at the time when the entire upper terrace is monumentalised with the construction of a double-gated and frescoed entrance to the processional ramp. This approach could have facilitated high traffic access and yet the series of doorways, with another near the N end of the ramp and a third limiting access from the N, suggests a progression, a hierarchy of access. The external ritual seems to be more visible yet proximity is restricted. Access is never more important or controlled than in this period when the Cult Centre becomes a physical part of the palatial sphere. The focus is the procession down into the Centre through multiple barriers and increasingly restricted access towards the Shrine culminating in an open plastered court. The use of a court recalls those of the palace that were not open to the public, where a limited capacity served as another form of control in addition to physical boundaries and barriers (Cavanagh 2001, 130). The upper part of the ramp, beyond the gated entrance, and the stairway upslope suggests a direct connection to the upper levels of the citadel and the palace. It is at this stage that the Cult Centre becomes physically connected to the palace while conceptually the monumentalising of the entrance underlines the conceptual connection as a marker of the sequence of the procession and the potential roadblocks to status and station. The power of the ruler is emphasised by the ceremonial aspects of ‘processional progress’ (Cavanagh 2001, 130–31). The Cult Centre of Mycenae (Fig. 8) should perhaps be termed a sanctuary, one within a settlement context to be sure but also an intricate organisation of ritual space with a ‘deliberate arrangement of buildings and open areas designed and furnished according to differentiated function’ (Albers 2001, 131). If the complex was built and functioned primarily as a cult facility then it is reasonable to assume that these aspects will in some way reflect the requirements of the cult. There will be features that were introduced in order to facilitate the celebration and smooth functioning of the cult. That might include the architecture itself, fixed and permanent features and moveable equipment. It might be understood as Albers has termed ‘public communal cult’ (Albers 2001, 131) where ritual is highly heterogeneous due to the variety of individual deities ‘housed’ in separate structures or even fundamentally different connotations of a single deity. The Centre is public rather than private yet official in the sense that communal festival and worship would require administrative organisation and execution 95
by priestly functionaries. It is not state cult like that performed in the palace but must be physically set apart since it is fundamentally different. Several elements of monumentality including architectural forms and building material used in a palatial context are found also in the Cult Centre, especially during LH IIIB and suggest that it functioned under the immediate control of the palace (Albers 2001, 143). Shrine G was a Mycenaean cult building that for part of its history at least included the older, smaller room to the S perhaps as an adyton and for storage of ritual paraphernalia, including the cult icon, and votive offerings. The walls were built in many stages but it remained a planned and coherent structure. The N room of the Shrine has clear cult use with the specialised and unique plaster libation altar built into the floor as a permanent focus of ritual action. The altar was set close to and in front of the entrance to Room G emphasising the importance of the inner part of the Shrine. The arrangements for catching liquid indicate libation offering as at least one of the primary uses of the altar, and by extension, of the Shrine in its early phase. The floor consists of only one layer of thin and carefully made plaster suggesting that the interior of the Shrine was not subject to substantial traffic and that entrance into the room was permitted to only very few, possibly only the religious functionaries. The room could perhaps have accommodated around 20 individuals if they stood fairly close together.147 The complete lack of features and finds in the upper Shrine room, signals an important change in the use of space, from the obvious emphasis on internal ritual, with the plaster altar and limited access and visibility by anyone other than the closest participants, to an external focus since the upper floor obliterates from view the ritual feature. Its function was very likely replaced by the ashlar feature in Area Q. The physical focus is now on the exterior of the structure, the court to its N where participants could have easily viewed the proceedings while additional participants of various socio-political status levels may have been able to view the proceedings from various vantage points on the Upper Ramp, Middle Ramp and from the roofs and windows of the Megaron and House on the next terrace down to the W.148 The ashlar altar may also have been used for libation or as a ‘table of offerings’. There was likely a continuity of function from phase to phase, even if the location shifts, a persistence of religious practice and continuity of worship. There also continues a strong emphasis on the division between inner and outer space, however reversed in focus. In a sense it may be that worship went on in much the same locality. Those in real proximity to the action would not have been more numerous than those able to fit inside the Shrine in the earlier phase. For more than a half century the Shrine sat in a lightly occupied area on the slope outside the citadel of Mycenae. The ritual was enclosed and visibility and participation was limited. After the construction of the WCW and the organisation of access routes by way of Stairway K and the Ramp system from upslope in later LH IIIB, access to the area of the Shrine is restricted but the ritual focus becomes more physically and visually accessible to a larger potential number of participants. It was not replaced by something sufficiently similar for the change to have been dictated simply by a change of fashion. There must have been more to it. Possibly the earlier Shrine offended in some way, especially following its inclusion in the cult complex, not because the rites were foreign, but as an instance of worship that was in danger of competing with the state worship of Mycenae. The most straightforward interpretation of the function of the House is habitation, a domestic structure built in close proximately to a cluster of buildings that were used primarily, if not exclusively, for cult. The orientation of the House, parallel and in physically close proximity to 96
the pre-existing Shrine on the terrace above it and facing N towards the cult structures opposite facing S. Access to the House became increasingly restricted and may indicate that it was not intended for significant or important traffic. Yet, it is an important structure within the context of the Centre and should have continued in its relationship to the other structures and their various functions through most of its lifespan. As Wright has demonstrated, architecture and building techniques ‘were powerful visual markers of the ascendance of the ruling power at Mycenae’ (Wright 2006, 18). Just as Tsountas and Wace suggested so long ago, the House emulates the palace and displays many features of palatial architecture: massive rubble masonry, systematic half-timbering, internal built staircases, cut stone elements and even frescoes. The physical similarities to the palace may indicate a real, perhaps official, connection where functionaries could have lived and worked with specialised activities (Wright 2006, 23–25). The Linear B texts indicate that there existed a priesthood but whether professional or part-time is unclear. The House may be interpreted as ‘religious’ by association but little in the physical makeup of the building, its features, or moveable artifacts would suggest that the building was actually used for cult. Everything about the structure (location, orientation, plan and construction quality) suggests it is more than simply a house. It was suggested by Wace to be the residence of the ‘priest-in-charge’ due primarily to the physical proximity to the Shrine, and although the House is a separate independent structure, the interpretation may be valid. The building may still be considered domestic and yet ‘religious’ if it is the habitation for the official residents of the Cult Centre (Iakovidis 1991, 1044; 1997, 154 and 158–59). The heterogeneous nature of the Cult Centre makes it likely that it may have housed a number of priests, or more likely priestesses, or a combination of the two. Also based on textual evidence, gods were served by priests and goddesses by priestesses. Based on our current understanding of the iconography and possible cult images in several buildings of the Centre, a number of goddesses were ‘housed’ and worshipped here. The nature of the female deities is sometimes redundant and could well be reflected in the female figurines and even female-shaped beads, such as those from Room G and Room 19 of the Temple, with a strong emphasis on fertility, while others are identified by weapons and armour. Another interpretation of the use of the House is as a hall or ‘megaron’ that could be used for the congregation of participants in the celebration of rituals or festivals in the Cult Centre complex. The variation of structures, symbols, and rituals understood from the many buildings and features along the slope suggest a hodge-podge of cult needs and a schedule of sporadic ritual events alongside more regular ones. In this way the domestic characteristics of the building can be understood in the context of a complex of cult buildings and ritual. Feasting is an important corporate and potentially ritualised event in Mycenaean culture and is one of the major activities which took place in the palaces, in their courts and perhaps also in the secondary ‘megaron’ complex (Wright 2004 and 2006, 39; Bendall 2004). The similarities to palatial features in the House may extend to some limited function. The Tsountas House may have been such a place, used as necessary for a variety of domestic and religious functions. Feasting was primarily a social aspect of a variety of ritual practices, including processions. Public feasts promote group solidarity among elite peer groups and between the ruling elite and their functionaries. ‘These practices provide opportunities for the affirmation of identity, its hierarchical structure and the legitimacy and authority of the ruling order’ (Wright 2006, 39). There is also a strong association with specific architectural forms such as the megaron and 97
court, and outside of the palace, religion can also be an essential element in the construction of social identities, while the social construction of religious identities marks the importance of religion in the definition of boundaries between different social groups (Edwards 2005, 110– 25). The architecture repeats the palatial features and mimics a setting that reminds us how political roles and political power are often inextricably bound up with ritual power. As we have seen, the Cult Centre and ‘Sacred Way’ were destroyed at the end of the LH IIIB2 period by fire and collapse, probably as the result of yet another earthquake. There is no evidence for repair or rebuilding following the destruction and in fact, there is good evidence for exposure of the ruins and erosion down the slope of debris in large masses of hardburned structures and lenses of mixed origin. Significant amounts of post-palatial Mycenaean material showed that this area had had subsequent occupation during several phases of LH IIIC, for the most part not immediately following the disaster, but also well into the postMycenaean period. No attempt had been made to clear the mass of destruction debris and the ramps themselves were never used again. The ritual landscape is usually very conservative and amazingly persistent, so the complete abandonment of a cult area is remarkable. Perhaps the deities were thought to no longer inhabit their shrines and therefore no longer protected their worshippers. Perhaps the Centre no longer had a patron. The collapse of the palatial administration would naturally have affected the Cult Centre and it is interesting that the practice of ritual does not revert to a more accessible, popular use. This could result because either the ‘Cult Centre’ had always been an official communal sanctuary, dependent primarily on the palace or its cult was by that point too far removed from the collective memory, more than 200 years beyond its origins. The early and complicated history of Shrine G must lead us to a new assessment of the Cult Centre and the implications for the nature and character of the cult practiced there. We can no longer speak of an exclusive religious precinct designed and built for the use of the privileged few. Perhaps more importantly, we can now see a shift in the character of the religion through changes in ritual design, access and with the incorporation of the Centre into the sphere of palatial control — just another sign of the changing atmosphere in the later 13th century BC at Mycenae.
98
ENDNOTES INTRODUCTION
1
Tsountas 1886, 77 and 1893, 15, describes the state of the W slope and the deep fill inside the Polygonal Tower where at a depth of eight to ten metres he uncovered the ruins of a Mycenaean era house (Q on his plan); he reports that after the destruction of the house, other structures were built on the ruins of the basement, c. 2.0 m higher than the surface of the corridor, made of small stones and clay but without any plastering. Only when these structures too were destroyed and the fill in this area had reached several metres, was the Polygonal Tower built in the citadel’s W wall, during the Hellenistic period, without an internal face because prior to its construction the entire area was filled. 2 In May of 1886 in fact (Tsountas 1886, 59–79). He was 29 years old. He uncovered a complex of buildings or remains of buildings, ‘some indistinguishable as to their different dates or even periods’ (74); Tsountas explains (1893, 7) that he was sent by the Archaeological Society to excavate Mycenae, which he ‘continues until today’ i.e. 1893. See also Shelton, 2006, 159–64 and Petrakos, 2009, 6–34. 3 Tsountas 1893, 7, katwtevrw dev th'~ korufh'~ para; to; dutiko;n tei'co~ oijkiva th'~ aujth'~ ejpoch'~. 4 The AE article (Tsountas 1887) in which he discusses the finds from G actually appeared in print before the PAE report (Tsountas 1886) in which he discusses the House itself. 5 7, 15, and 43–44; the latter pages include a description of the House in the context of Homeric houses and in comparison to the Palace that was excavated in the same campaign (1886). 6 Tsountas 1893, 15; he also adds that the accumulated depth of ruined houses meant that there was no need for an internal face of the tower. 7 Wace directed excavation that season in several different areas at Mycenae and visited each site daily making relevant notes. Also under excavation that year was: the House of the Oil Merchant (Blegen’s House), the Cyclopean Terrace Building, the tholos tombs of Clytemnestra and Epano Phournos, the Prehistoric Cemetery East under the House of the Warrior Vase and elsewhere. His two excavation notebooks for 1950 contain only brief but specific references to the Tsountas House Area. Marigold Pakenham Walsh (MPW) was the trench supervisor under the supervision of Sinclair Hood. In ILN 1950, Wace mentions the cooperation of Papademetriou and Petsas in 1950, and the loaning of British architect, Charles Hobbis, to the Greeks for on-going work at Clytemnestra (Wace 1951, 254–57); (BDC 2002 letter to EBF). 8 ILN 1961, Papademetriou said that the BSA permit lapsed with Wace’s death in 1957 and to honour his memory, the Greek Archaeological Service (he was at that time Ephor of the Argolid) undertook to complete the excavation under the joint direction of Papademetriou and Taylour. In WBM 1, 1, Taylour correctly identifies the collaborating body as the Archaeological Society of Athens, which has the exclusive right to excavate at Mycenae, rather than the Service and then begins the ‘Hellenic-British’ collaboration with Papademetriou as the Society representative. Two four-week seasons were carried out in 1959 and 1960 in the Tsountas House area. The 1962 season was confined to the areas further N. See Iakovidis 2004, 15, however, where he mistakenly suggests that the British worked from 1966 until 1974 supervised by Papademetriou then Mylonas. 9 ILN 1950. He also re-planned Tsountas House. 10 A few tests were made in Megaron C, Forecourt H, and Passage J. The area to the N of the forecourt was excavated for the first time since 1886 (in 1950 a narrow strip was excavated, to no great depth, along the N face of Wall K, between the wall and Tsountas’s dump. Apart from this, Wall K was the northern boundary of the excavation) and cleared to a far greater depth than by Tsountas; part of the forecourt itself was excavated to bedrock in 1960. 11 The architect, Charles K. Williams, corrected the 1950 plan, beginning in 1959, and included in it all new features (made especially in 1960). His Plan and Letter Key are reproduced in Figs. 1, 2.
99
12
13
14
15
16
17
In Taylour’s 1959 album (Mycenae Archive 1000), the entry for 10th August details the assignment of rooms: ‘some rooms in [the House of Tsountas] chosen as H.Q, i.e., the basement rooms with corridor to the E. The S end of the corridor used for fireplace for brewing up tea. The basement room with vestibule to E of it chosen as shelter’. Mylonas 1966b, 107–10; 1968, 9–11; 1972, 116–26, where he outlines what is known of Wace’s excavation in the Shrine and its backfilling following the excavation and remarks how ‘since 1950 the important remains of this building have remained invisible and unknown to the visitor and scholar, and practically ignored entirely.’ This name first appears in Mylonas 1972b, 124–26, called ‘to qrhskeutikovn kevntron twvn Mukhnwvn’ and is followed by the publication in Greek and English of Mylonas’s lecture with the same title to the Athens Academy on 18 April 1972. Mylonas 1972b, 118–21, where the new investigation of the Shrine (G–G1) is outlined in some detail; 121–22 mentions the House itself and potential access routes to the lowest level where Mylonas uncovered the ‘circular altar’ that year; additional commentary on finds and dating appears, 123–26; CC, 23 and 37, he explains the re-investigation of the area in order to: complete the clearance of the Shrine exterior altar and ‘the generalized statements published until then about Tsountas’ and Wace’s excavations.’ Mylonas 1983a, 308 n. 1, he states that he was unaware of the JHS 1951 report on Wace’s excavation in the Shrine until 1968 when he began to clear the area again; he also says that he worked there until 1975, meaning the entire Cult Centre, not the Shrine only. Iakovidis 2004, 15, attributes to Mylonas the excavation of the processional road in its entire preserved length and a court with an altar in front of cult rooms and cleared ‘a few places left unfinished by our British colleagues.’ BC 2002 letter to EBF: ‘Mylonas was not given the 1950 excavation photos nor did he request them.’ Wace 1951, 255; BSA 51 (1956), 122; Wace and Stubbings 1962, 395 and 496. At that time, an internal staircase had been identified based on buttress l (see Part 1 below). WBM 1, 9–10; Taylour identifies the construction Phase VII, a second stage of building, and Phase VIII following the catastrophe. Phase VIII in turn ends with a general, area conflagration; 19, he describes the complex as three terraces with the upper occupied by the Shrine, ‘an earlier construction than the remainder of the complex and on a different alignment.’
PART 1. THE HOUSE
18
19
20
21
22
It is clear from a comparison of Tsountas’s plan (Fig. 3) with that produced by Williams (Fig. 1) that the house deteriorated considerably between 1886 and 1960. The essentials remain nonetheless. Of course, the condition today, after fifty more years, is even worse. Dump from Tsountas’s excavation covered the northern limits of the area he had cleared and in 1950 extended about as far as the N face of wall K. That year a narrow passage was cleared along the N face of K, but the dump immediately beyond was not removed until 1960. Wace 1949, 66–67, where he mentions ‘cement floors’ that indicate plaster and two pithoi standing in the southern corner of the megaron as Tsountas had found. The designation of rooms and features in the Tsountas House Area by letters of the English alphabet refers to the Letter Key plan (Fig. 2) produced in 1959–1960 adding to some labels used already in 1950. The Greek letters in parenthesis refer to the room identifications on the 1886 plan (Fig. 3). The rooms in the basement were not assigned letters at that time. Tsountas 1893, 43, but Manatt translates incorrectly (1897, 67) ‘wooden staircase’; Wace remarks (1962, 496) that ‘some thresholds on the upper floor were stone and some, in the basement, of wood.’ Tsountas 1893, 43, the second storey over the basement rooms was interpreted by Tsountas as the ‘women’s quarters’ or apartment that was required by Greek custom and described in Homer
100
23
24
25
26
27
28
29
30
31
32
35
33 34
(Od.I.328–35). The first segment of rooms with a megaron was identified as the men’s apartment and would have been parallel on the same floor level as the women’s; cf. Tsountas and Manatt 1897, 67 Tsountas 1886, 74–78 with plate 4, the plan of the excavation area is by Dörpfeld. Although Tsountas described incidentally Stairway K and the finds he made in G, it was the house that he singled out for fuller publication. Tsountas 1893, 44; BC says in her scholia that this must refer only to the porch of the megaron and that parastades must mean pilasters or returns since there was no gap in the floor for a threshold. But see Tsountas and Manatt 1897, 67. Wace 1949, 80, also cites the plan/arrangement of Tsountas House in comparison to the palace, as well as to other houses at Mycenae. The area N of Wall K belongs to the substructure and basement of the Megaron, the two storey building on the middle level, N of the House, that will be published in WBM 12. Mylonas 1972b, 121–22; in CC, 15–16, he describes the house as built on three levels, on the middle of which are well-preserved foundations of a megaron with court and forecourt; by 1981 (see Mylonas 1983a, 308) he has decided that the house and shrine are ‘two different structures, occupying three rising terraces.’ Mylonas 1983b, 141, none of these rich offerings was found, however; compare artifacts from the Megaron basements, the Temple deposit, and especially Room 32 in the Room of the Fresco complex, not to mention Room G. Mylonas 1974, 90–91, further clearance in 1974 allowed the excavation of the area and revealed a passage (4.2 m wide) between the house and the WCW. It is not known what may have existed in this area before the construction of the fortification wall (Fig. 4). Apparently this identification rested partly on the advice of the architect who was later considered to have done shoddy work. Charles Williams, the architect from 1959 pointed out that ‘stairway’ l could not have existed because the bottom of the stairs, meant to be at the N end of the heavy pier construction supporting a 90° angle turn, would require a first riser of >0.50 m in height and the top of the stairs would enter the Shrine at its NW corner where no evidence existed in the plaster floor. The whole construction of stairs, floor, and roof would be very difficult and there simply was not enough space to cover the necessary height. This impacts on the significance of the Shrine more than it does on that of the House. Their independence only opens more questions about the use of the House. Taylour (WBM 1, 19) suggests that the shrine on the upper terrace is an earlier construction and on a different alignment but that access to the middle terrace ‘megaron area’ was perhaps by wooden stairs for which a stone buttress at the NE end of the room would provide a base. Showing, incidentally, the depth of fill beneath this part of the floor of the court. Tsountas 1886, 74–75, he compares the bothros to the well-known altar in the court at Tiryns. CC, 16, 23 and 37, the court of the megaron contained a bothros for sacrifices, ‘revealed by its excavator,’ i.e. Tsountas; 1977b, 22; Mylonas quotes Tsountas (1886, 74–75) but then questions whether this bothros is evidence enough to characterise the structure as sacred — a temple; same repeated in Mylonas 1983a, 315 n. 2 and in 1983b, 140, where he calls this the ‘outer court’, and quotes Tsountas’s description without attribution and emphasises that this is the ‘only recorded evidence that could indicate some religious use of the building;’ 141, reiterates and states that the purpose of the building remains undetermined. The famous plaster fragment with donkey-headed figures came from Area Z just outside basement room F1. The fragment is said to be a wall painting fragment and since I have never seen its back, I assume that it has been characterised correctly. The rough sides of the fragment as visible on exhibit in the National Museum in Athens indicate breakage and its relative thickness would rule out a plaque such as that found in Room G.
101
36
During excavation, the Greeks were certainly unaware of this and their insistence that the upper stairs were later may be a result of this. However, the upper stairs are orientated differently and are more likely to be original than the lower steps. 37 Mylonas CC, 122; 1983a, 316; 1983b, 141; see also Iakovidis 1983, 47, where he describes stone steps originally covered with short wooden planks and the first three steps of which are at a slightly oblique angle to the rest and a later addition made necessary by the raising of the level of the courtyard during the construction of a retaining wall over the ramp to the lowest level. Although the ‘later’ date of the differently orientated steps may in part be due to their rebuilding in 1950, they do appear with this orientation in 1886 and on other structural grounds are in fact a rebuilding or addition because of other structural changes as described by Mylonas. 38 Measurements and observations of the timber-frame chasing were taken in 1966 at Barbara Craig’s request by Nicholas Postgate and sent to her by letter. 39 Referred to in some 1960s excavation notes as the ‘dinner room’ or ‘breakfast room’ in both cases because it was partially roofed for shade and used for breaks and meals on site (see above Introduction and note 12). 40 Also called ‘Billy’s Bower’ in some 1960s notebook since this double-chambered basement room was used as Lord William Taylour’s study (see above Introduction and note 12). 41 Mylonas 1968, 11; the LH IIIC early floor was found 3.45 m below the surface. 42 A pier could have helped to support a roof over Stairs G and Room F1. It seems unlikely that the wooden stair-planking was unprotected from the weather. It could also have helped support roofing over the entrance to Court A. 43 The top-most stone of Wall M is 0.14 m lower than the coarse plaster floor at the head of the stairs. 44 Part of this coarse upper floor was removed in 1950. 50-239 was found between the floors. 45 The difference of height between the top of the upper hard floor and the softer coarse floor at the head of Stairs G was c. 0.24 m. 46 No trace of the wall was visible on the surface at the beginning of the campaign, although it seems to appear on the 1886 plan. The 1960 excavation suggested that it had not been previously uncovered and may have carried upper courses discovered by Tsountas and later destroyed under his dump. 47 This room is part of the last phase of the Temple complex; called T3 on the plan of Iakovidis 1997, 165, fig. 8. 48 Mylonas 1972b, 122; Iakovidis 1983, 47; Iakovidis 2004, 21, in LH IIIB2 the ramp to the lowest level was blocked off. 49 These include shallow cups (FS 219) with stipple (FM 77) decoration and examples of monochrome kylikes (FS 264). T3’60/23, 25, 27, and 28. 50 See the Supplementary Data for a catalogue of these sherds. 51 This was true elsewhere in the Cult Centre as well, cf. WBM 1. 52 Comparable domestic rubbish in the fill beneath a floor was found in the Palace, dating to several different periods: below the East Lobby — BSA 25, 157 (black earth, quantities of potsherds, a few fresco fragments, bones of sheep, pig and other animals, and oyster and mussel shells); Pithos area, under lowest floor and on the two upper — ibid. 174; Pillar basement, ibid. 181; Court, under floor on either side of main drain, ibid. 199; room N of West Portal, ibid. 218; topos 28, ibid. 264. A similar deposit occurred on rock in the Atreus Bothros, Wace 1949, 124. 53 Sherds typical of LH IIIA in fabric, finish, paint, and mode of decoration from this deposit are shown in 62-R17 (Supplementary Data p. 460). 54 This sherd material and associated small finds will be published with the Megaron, WBM 12. 55 Disintegrated mudbrick from Wall K was excavated to the N in Room IV of the Megaron. 56 WBM 10, 110 and Supplementary Data 123; 74, several examples also were found in Room 19 of the Temple (60-150). 57 See note 52 above for similar deposits.
102
58
59
60
This is based on the interpretation of the various cult places in the Centre being dedicated to primarily female deities and therefore would be attended by female religious officials. See further discussion in Part 4 below. Mylonas 1983b, 141, ‘the ‘almost inaccessibility of the building,’ the pithoi and lead vessels in the main room, the division into small compartments indicating special use, can be interpreted as a ‘treasury’ of the Cult Centre for the storage of rich offerings, such as clothing and scented oil.’ An exception is the fresco fragment from Area Z that he does describe in full because of its obvious importance.
PART 2. THE TSOUNTAS HOUSE SHRINE
61
63
64
65
66
67
69
70
72
62
68
71
Mylonas 1977b, 19, the S room G is called G1 and the N room G1 is called G2. The Shrine was very briefly reported by Wace (ILN 1950; 1951, 254–55). An account of the excavation was put together from these publications, Wace’s 1950 notebooks, and the 1950 notebook and preliminary report of Miss Marigold Pakenham-Walsh, who was in charge of the work on Tsountas’ House, under Sinclair Hood’s supervision and Wace’s direction. The plan produced in 1950 by Hobbis is not an actual or accurate stone by stone plan; many details of walls, etc. are also inaccurate. Williams corrected the wall lines and features of the Shrine plan in 1959 and 1960, but did not redraw the wall stones. Mylonas 1971, 151–56; 1972, 121–22; CC, 21–26; 1977b, 19, 21, 41, and 92; 1983a, 308, 310– 15; and 1983b, 133–40. The intensive study of the Shrine was conducted by Iakovidis in 1972 and he presents his preliminary interpretations in several publications: 1983, 45; 1986, 243–44; 1996, 1044; 1997, 150–51 and 154; 2004, 14–15, 17–21, 23–24. The terrace of G differs from many others at Mycenae, such as those of the Ivory Houses, on artificial horizontal platforms of vast quantities of stones and earth. No such platform was made for G that was built on a steep slope. Rather, the foundation walls of G were packed with fill up to the top of the rock, much like the foundations of Petsas House, and is therefore likely to be an indication of its construction date well before the LH IIIB period when the platform terraces begin to be built. Cf. Shelton 2009d. Part of the foundations of the Ramp House (i.e. the ‘south chamber’) seem to be similar as well, reinforcing further the earlier date of this construction style, BSA 25, 75. Wace believed there to have been a bench along the E side of the room covering the bedrock outcropping, but no physical remains of a bench, other than the clay covered bedrock surface itself, were found and the lower plaster floor stops at its edge, unlike the bench on the Middle Ramp, where the plaster floor turns up over the lower part and seat. In the later phase, the outcrop was entirely covered by the upper floor. Mylonas 1977b, 21, the bench was perhaps used for the placement of sacred vessels and other objects used in the rituals; by 1983 he no longer mentions a bench. It was believed to be the location of a horizontal framing beam that had burned. Most of the floor to the W of the altar had been destroyed and the jar (50-234, NM 5375) was not complete. The shape has been described as horseshoe shaped (petalovschmo), apron shaped, D shaped, unusually shaped; tongue-shaped best describes the filled bodycore with the outward curving N end, the straight E and W sides, and the slightly concave S end. Remains of burnt plaster were found on the E and S walls. Mylonas 1977b, 21, suggests that the libation is blood from animals slaughtered on the stone to the N (see below and note 76); 1983b, 136, use for a rhyton was ruled out by actual experimentation. H. 0.33 m, Di. (max.) 0.24 m, Di. (base) 0.9 m. There is no BE number since the pot has not been brought from Nauplion storage to the Mycenae Museum.
103
73
Again, much of the floor in this area and to the immediate W was badly damaged and eroded away, so the jar was removed from its position and taken to museum storage in Nauplion where it was registered. 74 Lost before registration in Nauplion. 75 ILN 1950, 1041 and figs. 4–7; 1951, 254–55; in his 1950 notebook #1, 74, Wace queries if this is a column base belonging to the upper floor then doubts this interpretation. A week later he has reversed his assessment. 76 Mylonas CC, 25; 1972b, 120; 1977b, 21; 1983a, 313, believed the altar and stone to be contemporary because of reported traces of plaster ‘from lower floor’ visible on its vertical sides; see also 1983b, 136. The unworked and irregular upper surface, as well as the size of the stone, rule it out as a base of a column or built pier, according to Mylonas. He suggested that the stone had a role in the ceremonies conducted in this room, perhaps here were sacrificed animals brought here, the blood from which was used as libation. 77 Mylonas 1983a, 311, speculates from the plan (Dörpfeld’s) that he found a floor around the ‘elliptical/circular’ object in the NE corner at a height of 57 cm above bedrock, preserved in an area approximately 2 m × 1 m; also 1983b, 134. 78 It was in this way, that the boulder and surrounding stones were found. 79 Whether called Room P or G1. 80 In several publications, the remains of the upper floor and the bedrock outcropping along the E side of the room are described as a bench, such as Iakovidis 1997, 151 and 2004, 17–19, for example. This originates with Wace who believed a bench to have covered the rock during the lifespan of the earlier Shrine (see above for the possibility of the footing for an earlier E wall instead). No actual bench structure was found. 81 This would be similar to the walling up of the doorway to Room 19 of the Temple that occurred during the rebuilding phase VIII following the earthquake destruction in LH IIIB middle (WBM 10). 82 This vessel was most likely of unbaked clay originally and similar to others found in the area (see WBM 16/17) and elsewhere on the site. These circular clay containers were called kotselles by one of the excavation workmen and have continued to be identified with that term by the British excavation team members. The burning in and around the feature during the destruction of the Shrine may have partially preserved the clay by accidentally baking it to some extent, however, not fully enough so that it appeared to be disintegrating at the time of its excavation. 83 Mylonas 1971, 155 and CC, 22, suggests that this vessel was used to deposit the ash from sacrifices as well as libations. 84 Wace (ILN 1950) says Q is contemporary with the lower, earlier shrine, which it certainly is not and he says as much in his earlier analysis in his excavation notebook (1950 notebook #1, 64) ‘perhaps poros and stepped stucco at north end is later version of hearth on lower floor.’ 85 Shown in blue that is his convention for worked stone but without distinguishing them as two stones. 86 Iakovidis 1983, 45, identifies this cutting as a stand for a libation vessel like that of the earlier plaster altar inside the Shrine, not realising that the square was opened by the excavation in 1959. 87 Iakovidis 1997, 150, states that a third missing stone left indications in underpinnings of small stones and mud on the bedrock. He suggests the upper covering was probably of poros slabs with filling of small stones; also Iakovidis 2004, 17; cf. V. Karageorghis and M. Demas, Excavations at Kition V.I: 107 and 255 (Nicosia, 1985). 88 Block c is now missing; it was last seen in its position on site in 1966 and appears in Mylonas 1966b, pl. 93.b. 89 Mylonas 1983a, 311, Mylonas states that Tsountas did mention sherds in the fill but were apparently not studied so no date was determined, however, in 1983b, 134, he says that Tsountas found sherds of the LH IIIB2 period.
104
90
91
92
94
95
97
98
99
93
96
100 101
102
BDC consulted the original Athens National Museum register and was surprised by the large number of entries for 1886 since Tsountas himself had stressed how little he had recovered from either the house or the palace, the two areas he excavated in 1886 along with a test trench in the region just above the Lion Gate, later called Building N, which he dug in 1890. BDC notes that she was not allowed to consult Tsountas’s notebooks that were said to be too fragile (no mention of any missing at this time — see Xenaki-Sakellariou 1985). The entries in the register were apparently made by Tsountas ‘in his own hand’ in 1892, and in no particular order since the 1886 entries were scattered among others. A few from 1886 are labelled simply ‘Mycenae’ while most say ‘Acropolis’ as well. A very few had the additional label of ‘Palace’ or ‘House by the Polygonal Tower’. In PAE 1886, Tsountas emphasised that the Palace and House produced almost nothing (except for ‘a wealth of sherds’ from the house, cf. 61, 70, 78), BDC suggested that the majority of the entries labelled ‘Mycenae Acropolis’ must be objects from the Tsountas House Area, excluding the House itself. Tsountas 1886, 79 and 1887, 169; In both accounts evdafo~ was used; he intended ‘floor’ rather than ‘ground’ since it is described as unplastered and it is the term he employed for all the Tsountas House floors. Tsountas published only a selection from a number he describes as considerable. Similar groups of these types of materials and objects were found in Room 19 of the Temple and Room 32 of the Room of the Fresco complex. In both instances the votive scraps were found in association with iconic images, such as ceramic figures, that may represent the recipient, cf. WBM 10 and WBM 11 (forthcoming). See Burns 2010, 144–45, who agrees with the early date of most items, especially those believed to be foreign imports. Cf. catalogue in WBM 25 (in preparation). These references refer to the published illustrations of the objects by Tsountas (1887, plate 13). Another example of gold cloisonné (60-587, BE 10100) was found in the forecourt of the House and could date as late as LH IIIA. To the same period date 21 ivy leaf shaped gold beads with granulation from Petsas House; cf. Shelton 2007, 174 and PAE 2005, 37 and fig. 30. This seems odd and not in keeping with the collection methods elsewhere in the area in 1950 so it suggests that in fact nothing was found in the small areas investigated below the lower floor and that the fill under it was particularly clean soil. There is no Mycenae museum number (BE) because the jar was not transferred from storage in Nauplion to Mycenae. Iakovidis 1996, 1997 and 2004, reports that only either the base or mouth of an amphora was used, rather than the entire jar, as was actually found and removed to the museum. The heavy kitchenware is one of two standard fabrics used in kitchen and cooking pottery from Petsas House and dating to LH IIIA2. Only the gritty kitchenware fabric continues in use during LH IIIB with the heavy ware appearing very rarely. Iakovidis 2004, 15, in a test under the lower floor during the Greek excavation of the Shrine, only MH sherds were found in the packing on the bedrock. Iakovidis 2004, 23, dates this to the advanced years of LH IIIB2, but without specifying on what evidence this dating is based.
PART 3. ACCESS 103
Papademetriou and Taylour 1961, 490–92; WBM 1, 15, 19, 24, 29; Wace Guide, 35; Mylonas 1966b, 107–10; 1971, 151–55; 1972b, 121–22; 1974, 90–91; CC, 18–21, 23, 36; Mylonas Guide, 24; 1977b, 19; 1983a, 308–10, 315; 1983b, 128–32; Iakovidis 1983, 45 and 48; 1986, 243–44; 1996, 1044; 1997, 148–50, and 154; 2004, 14–15, 17–21, 23–24; where appropriate this evidence will be incorporated into the account.
105
104 105 106
107 108
109 110 111
112
113 114 115 116
Tsountas 1886, 77, due to the fact that the steps are of different widths and that stones from the house wall sit on the steps; the uneven width of the House’s W wall is also explained by the pre-existing stairs and the construction of the House being accommodated to ‘not make the stairs too narrow’. If he had cleared it fully, its bedrock floor would have been indicated on the plan and Tsountas would likely have mentioned it; the drain culvert was also neither mentioned nor indicated on the plan while a drain mouth exiting through the Hellenistic Polygonal Tower is shown. Wace 1950 notebook #1 (Mycenae Archive x026), opp. 71, and BC 1959 notebook (Mycenae Archive #1228), 10, CKW suggested that at the top of the N flight, the steps bond with the heavy outer terrace wall between the Shrine and the House, however, this does not appear to be the case in his plan of 1960 nor can it be distinguished in contemporary photographs. A close inspection of the current remains does suggest this; see also Mylonas 1966, 107 who agrees that the stairs are later than or contemporary with the house. MPW 1950 notebook (Mycenae Archive x040), 29–30; only 4 or 5 steps remained in the S flight in 1959 and 1960. Iakovidis 1983, 48, ‘Immediately to the south of Tsountas House a narrow staircase with rather uneven stone steps descended the slope along the outside wall of the house. It started from an earlier path just below the processional way, and ended either at the foot of the Cyclopean fortification wall or at the lowest basement in the Tsountas House.’ Iakovidis and French 2003, for examples in torrent beds around Mycenae and identified during the survey in the early 1990s; see also Loader, 1998, 205. Investigated in 1960 by Epaminondas Kolizeras. Excavation by Mylonas (1983b, 130) further upslope actually uncovered the end of the single course that was fed by two diverging drains that lead up into unexcavated areas E and N of the upper stairs of Mylonas’s processional way; see also the plan in Iakovidis 1977, 124; Iakovidis 1983, 48, ‘An open drain which passed underneath the threshold on the processional way ran parallel to this staircase and continued as far as the enciente. It was fed by two smaller conduits that met beneath the staircase at the beginning of the processional way, both of which continued to be used until the end of LH IIIB.’ The drain mouth floor is in the third course of stones above bedrock (Schliemann 1878, pl. II; Tsountas and Manatt 1897, fig. 5). The LH drain mouths (of which at least 18 are visible in the outer face of the circuit wall of Mycenae) are at rock-level only if bedrock is very high, and are otherwise two or three courses up. This is so even when in the thickness of the wall the drain is channelled on rock. Wace 1949, 67, part of this complex appears in ILN 1950, fig. 5; it had not been examined since 1886 until the Greek excavations, beginning in 1968. This residential area is published in S. Iakovidis, Anaskafev~ Mukhnwvn. III. H Notiodutikhv Sunoikiva. Archaeological Society of Athens, 278, 2013. The uppermost, eastern segment of the ‘Sacred Way’ and almost the entire area N of threshold m was excavated by Mylonas from 1966 to 1972. Cf. Mylonas 1966b, 107–09; 1971, 152–55; and CC, 19. The closest parallel to Threshold m, though wider, is probably the broken, sawn conglomerate threshold at the entrance to the anteroom at the head of the Grand Staircase of the Palace at Mycenae. This has ‘a sawn door ledge cut on each side, and at the end of one of the ledges … a hole for the insertion of a pivot’ (BSA 25, 180; pl. xxii d). Tsountas, 1886, 64, gives the diameter of this pivot hole as 0.12 m, exactly that of the pivot hole of threshold m. The comparable dimensions, material, and especially techniques of manufacture could perhaps be used as a dating tool and an indication of a later LH IIIB construction. Wright 1987, 177–84, presents a discussion of conglomerate’s use symbolically by the ruling group at Mycenae to express both power and authority; further analysis in R. D. Fitzsimons, ‘Architecture and power in the Bronze Age Argolid’, in J. Bretschneider, J. Driessen and K. van Lerberghe, Power and Architecture: Monumental Public Architecture in the Bronze Age Near East and Aegean (Orientalia Lovaniensia Analecta 156; Leuven 2007).
106
117 118 119 120 121 122
123 124 125 126 127
128 129 130
131 132
The Greek excavations also seem to have found only one layer of plaster floor and agree with a construction date in LH IIIB2 and destruction at the end of that same period, although the nature of the dating evidence is unclear, cf. Mylonas 1983a, 309; 1983b, 130; and Iakovidis 1983, 45. Figural and decorative frescoes were found by the Greek excavation under the post-Mycenaean terrace wall, Mylonas 1983a, 308; the E wall has a dado course in two pieces: one is a chariot procession and the other has multi-coloured wavy bands. This small ramp is not to be confused with the ‘Little Ramp’ further to the N, and E of the Ramp House, although the two ramps connect at the S end of the ‘Little Ramp’. Dr Elizabeth French dates the construction of the Causeway ramp to the first half of, or at latest mid, LH IIIB, from a restorable Zygouries kylix found immediately beneath it. Two additional conglomerate bases are to be associated with this unexcavated structure, one that fell into Corridor 4 after hard rains in the late 1990s and another still visible on the upper terrace. Without any evidence, I suspect that Wall A is the older wall meant to act as a terrace support and that it was added to, or connected with, later walls (low E wall of Room 4 and the E wall of the prothyron) built in connection with the redesign of the access and approach avenues to the Cult Centre, on the one hand necessitated by the construction of the WCW, and on the other hand, by the later addition of the gated entrance and upper approach in LH IIIB2. Unlike Wall A where changes in direction are accomplished with a ‘jog’ in the wall, as is usual practice in Mycenaean construction, Wall R does not include any jogs, whether segments bond or abut as in segment 3. This passage or lane may have been broader in the earlier phase of the Shrine if indeed its original dimensions were narrower than in its later phase. At that time also, there would not have been heavy drain walls blocking access at its S end. Not drawn in on the CKW plan, but see Tsountas’ 1886 plan and that of the Greek excavations in Iakovidis 1997, fig. 8. It is shown on the plan by the dots along the line of the post-Mycenaean terrace wall running from a point c. 3.40 m N of threshold m to the southern limit of excavation. The numbered Ramp surfaces should not be confused with those identified in the ‘Great Ramp’, the ascent from the Lion Gate court up to the S, cf. G. Mylonas, ‘H Akrovpoli~ twn Mukhnwvn. Oi Perivboloi, ai Puvlai, kai ai Anodoi’, j Arcaiologikhv Efhmeriv~ 1962, 128–43, 194–95 and 198. The earlier phases of the Great Ramp, several ascending from S to N, are not connected to the phases of the Upper Ramp here, but may be related as part of an approach route or routes to the upper parts of the hill. Excavated by Mylonas 1971, 152–55. Mylonas 1971, 153, near the ‘step’ was found a small drain (20 cm wide, 22 cm deep) under the corridor and running in its line to the drain under the corridor/Room 4. Mylonas 1966b, 109–10; Fig. 4; CC, 18, ‘the corridor was built during LH IIIB and its use ended towards the end of that period. Then a large part of its length and entrance were destroyed and covered by burned wall remains — fallen from above c. 1200 BC’; 1983a, 310, ‘Also revealed was a small wall P at right angles to the side wall, small stones and the floor under it of plesia over a layer of small stones. In the floor was found a second poros base and between it and the other, the large part of a circular hearth inside a ring of plaster (81 cm diameter). Lower 6–9 cm below the plesia floor was found the plastered floor of the corridor.’ ‘The corridor was abandoned at the end of LH IIIB, then in LH IIIC building on top of it in two periods and then destroyed end of IIIC;’ 1983b, 131, ‘cobblestone pavement strewn with sherds of LH IIIC 2.’ WBM 16/17, elsewhere in the area, fill over the destruction debris has been shown to be stabilisation rather than a time lapse. Column bases at Mycenae, Tiryns, Pylos and Korakou (House L) are either rough hewn stones with the upper face cut back so as to leave a neat circular protuberance projecting above floor level to support the column — cf. column base in Court A of Tsountas House; or the entire base is disc-
107
shaped and itself rests on a separate slab or a collection of smaller stones. Squared blocks such as that on the ramp usually support squared rubble piers or door-jambs. LH IIIB2 late slender posts are known in Room 38 (WBM 11) and outside Room xxiv (WBM 13). 133 Mylonas, PAE 1970, 119–20, Mylonas also believes the blocks, on the ramp and on the steps, to be in situ, but he dates them both to LH IIIC. 134 Mylonas 1972b, 121, describes an earlier plastered surface at c. 20 cm below the final ramp floor in tests along its southern edge. 135 WBM 1, 15; the height of the slab is 241.19, the floor of Room 2 is 241.15. 136 Mylonas 1971, 154; CC, 21; and 1983a, 315, mentions five steps but states that Tsountas uncovered six. However, Tsountas never mentions this feature nor are the stairs indicated on his plan; the British excavation records uncovering them for the first time in 1959. Tsountas did excavate six steps in the upper, southern flight of Stairway K, indicated clearly on his plan (1886, 77). Mylonas 1972b, 121, ‘…from the bottom poros step of the processional way a descending ramp covered with a plaster surface that leads to the forecourt of the House’; here he identifies Taylour as the excavator of the area rather than Tsountas as he had suggested in 1971; in Iakovidis 1983, 45 and 1986, 243–44, only four steps are mentioned while later (1997, 154), six poros steps return. 137 Tsountas would likely have kept the two miniature pots (50-184, 50-185) and a black steatite bead (50-6), found in the drain itself, while he often overlooked terracotta figurines (such as 50-312) that are later recovered from his dumps. 138 The thickened zigzag and narrow banding suggest an earlier rather than later date, and would therefore be more closely related to the pottery from the Shrine including another miniature, also a stemmed bowl or goblet (50-287). True miniature vessels are better known from LH IIIB, see Damm 1997, but are also made in LH IIIA2 as evidenced by those from Petsas House, see Shelton 2009c. 139 1.xi.60 basket (48) ‘Main Drain Investigation by Epaminondas’ referring to the workman, Aristeidis Epaminondas Kolizeras, then 15 years old, who climbed into the drain to investigate for as far as he could fit. 140 Mylonas 1983a, 311, ‘from fill dumped beyond the south wall of the room over the stairway presumably by Tsountas’. 141 Mylonas 1983a, 311, thrown out presumably following the Mycenaean destruction, or possibly Tsountas’s excavation? In Mylonas 1983b, 133–34, he reports that in his diary of 1896, Tsountas states that the head was found in the fill of the stepped way. The date of the diary suggests that this was found much later than his excavation of the House, etc. in 1886, but his plan indicates the steps had been cleared in the first campaign. 142 W. D. Taylour, ‘The Citadel House’, MT III, 35–40, for a description of the identical type of material further N in the Citadel House Area excavation, especially in the corridor, Room 4. 143 Similar small battered sherds were noted in the packing of the Upper Ramp and in the deeper levels of the forecourt cutting. 144 This sherd material and associated small finds will be published with the Megaron (WBM 12). 145 Missing from storage before registration in the Nauplion Museum.
PART 4. CONCLUSION 146 147 148
Hägg 1995, 387–91; 1981; Albers 2001, 131–41; Bendall 2001, 445–46; Whittaker 2001, 359–60; but see also Wright 1994, 72–76, for reasons not to consider this division as separate concepts. Allowing for approximately 1.0 m2 per person and subtracting floor space for the altar. This reconstruction recalls the fresco of women at windows in an architectural façade found by Schliemann near Grave Circle A and published by G. Rodenwalt, ‘Fragmente mykenischer Wandgemälde’, AM 36 (1911) 221–50, and to be associated with the fragments excavated from under the Ramp House, cf. W. Lamb, ‘Frescoes from the Ramp House’, BSA 24 (1919–21) 189– 94 and M. Shaw, ‘The bull-leaping fresco from below the Ramp House at Mycenae: a study in Iconography and Artistic Transmission’, BSA 91 (1996) 167–90.
108
BIBLIOGRAPHY For references and abbreviations used throughout the WBM series, which were published in the introductory fascicule (WBM 1), see Supplementary PDF pp. 493–99. A consolidated bibliography for all fascicules will be published when the series is complete. ABBREVIATIONS CC G. E. Mylonas, The Cult Center of Mycenae. Pragmateivai th~ Akadhmiva~ Aqhnwvn, 33. Athens, 1972. ILN 1950 A. J. B. Wace, ‘Exploring the city of Agamemnon: this year’s excavations at Mycenae’, Illustrated London News 23.xii.1950, 1041, figs. 4–6. ILN 1961 J. Papademetriou and W. D. Taylour, ‘The last days of Mycenae: tablets and houses revealed in the resumption of the late Professor Wace’s excavation’, Illustrated London News 23.ix.1961, 490–92, figs. 1, 14–15. MT III J. Chadwick (ed.), The Mycenae Tablets III, Transactions of the American Philosophical Society Vol. 52, no. 7 (1962). Mylonas Guide G. E. Mylonas, Mycenae: A Guide to its Ruins and its History. Athens, 1977. Wace Guide H. Wace and C. K. Williams, Mycenae Guide, 9th edition. Meriden Connecticut, 1976. WBM E. B. French and K. A. Wardle (eds.), Well Built Mycenae. The HellenoBritish Excavations within the Citadel at Mycenae 1959–1969, quoted by Fascicule number, specifically in this publication: WBM 1 WBM 8 WBM 10 WBM 11 WBM 12 WBM 13 WBM 16/17 WBM 21 WBM 24 WBM 25 WBM 27 WBM 36
W. D. Taylour, E. B. French and K. A. Wardle, Fascicule 1: The Excavations. Warminster, 1981. D. H. French, Fascicule 8: The Pre-Mycenaean Pottery. Forthcoming. A. D. Moore and †W. D. Taylour, Fascicule 10: The Temple Complex. Oxford: 1999. A. D. Moore and †W. D. Taylour, Fascicule 11: The Room with the Fresco Complex. In preparation. S. J. Aulsebrook, Fascicule 12: The Megaron. Forthcoming. E. B. French and †W. D. Taylour, Fascicule 13: The Service Areas of the Cult Centre. Oxford: 2007. E. B. French, Fascicule 16/17: The Post-Palatial Levels. Oxford: 2011. J. H. Crouwel, Fascicule 21: Mycenaean Pictorial Pottery. Oxford: 1991. O. H. Krzyszkowska, Fascicule 24: The Ivories and Objects of Bone, Antler and Boar’s Tusk. Oxford: 2007. H. Hughes-Brock and J. Phillips, Fascicule 25: Beads and Related Small Items. In preparation. D. Evely and C. Runnels, Fascicule 27: Ground Stone. Oxford: 1992. L. C. Bowkett, Fascicule 36: The Hellenistic Dye-Works. Oxford: 1995. 109
REFERENCES Albers, G.
2001
Bendall, L. M.
2001
2004
Burns, B.
2010
Cavanagh, W.
2001
Chadwick, J. (ed.)
1962
Craig, B. D. Damm, U.
2002 1997
Edwards, D. N.
2005
French, E. B.
1981
2002
French, E. B. and 2005 K. Shelton Hägg, R.
1981
1985
1990
1995
‘Rethinking Mycenaean sanctuaries’, in Laffineur and Hägg 2001, 131–41. ‘The economics of Potnia in the Linear B documents: palatial support for Mycenaean religion’, in Laffineur and Hägg 2001, 445–52. ‘Fit for a King? Hierarchy, exclusion, aspiration and desire in the social structure of Mycenaean banqueting’, in P. Halstead and J. C. Barrett (eds.), Food, Cuisine and Society in Prehistoric Greece. Oxford: 105–35. Mycenaean Greece, Mediterranean Commerce, and the Formation of Identity. Cambridge. ‘Empty space? Courts and squares in Mycenaean towns’, in K. Branigan (ed.), Urbanism in the Aegean Bronze Age. Sheffield: 119–34. The Mycenae Tablets III, Transactions of the American Philosophical Society, NS 52.7. (Abbreviated MT III) Letter to E. B. French. Die spätbronzezeitlichen Miniaturgefässe und Hohlge formten Stiere von Tiryns. Eine Analyse der Form und Funktion. Ph.D dissertation, Rheinischen FriedrichWilhelms-Universität zu Bonn. ‘The archaeology of religion’, in M. Díaz-Andreu, S. Lucy, S. Babić and D. N. Edwards, The Archaeology of Identity. Approaches to Gender, Age, Status, Ethnicity and Religion. London: 110–28. ‘Cult places at Mycenae’, and ‘Mycenaean figures and figurines, their typology and function’, in Hägg and Marinatos 1981, 41–48, 173–78. Mycenae. Agamemnon’s Capital: The Site and its Setting. Stroud. ‘Early palatial Mycenae’, in A. Dakouri-Hild and S. Sherratt (eds.), Autochthon. Papers presented to O.T.P.K. Dickinson on the Occasion of his Retirement. British Archaeological Reports International Series 1432. Oxford: 175–84. ‘Official and popular cults in Mycenaean Greece’, in Hägg and Marinatos 1981, 35–39. ‘Mycenaean religion: the Helladic and the Minoan components’, in A. Morpurgo Davies and Y. Duhoux (eds.), Linear B: A 1984 Survey. Louvain-la-Neuve: 203–25. ‘Mycenaean religion in perspective: cult practices and beliefs’, Journal of Prehistoric Religion 3–4: 48–49. ‘State and religion in Mycenaean Greece’, in R. Laffineur and W-D. Niemeier (eds.), Politeia. Society and State in the Aegean Bronze Age. Aegaeum 12. Liège: 387–91. 110
Hägg, R. (ed.)
2002
1988
Peloponnesian Sanctuaries and Cults: Proceedings of the Ninth International Symposium at the Swedish Institute at Athens, 11–13 June 1994. Stockholm. Sanctuaries and Cults in the Aegean Bronze Age Proceedings of the First International Symposium at the Swedish Institute in Athens, 12–13 May, 1980. Stockholm. Early Greek Cult Practice. Göteborg.
2008
Dais: the Aegean Feast. Aegaeum 29. Liège.
1983
Late Helladic Citadels on Mainland Greece. Monumenta Graeca et Romana 4. Leiden. ‘Destruction horizons at Late Bronze Age Mycenae’, Fivlia e]ph eij~ Gewvrgion Ev . Mulwa`n dia; ta; 60 e]th tou` ajnaskafikou` tou e]rgou. Library of the Archaeological Society of Athens 103. Athens: 233–60. ‘H Arcaiologikhv Etaireiva sti~ Mukhvne~’, AE: 339–59. ‘Mycenae in the light of recent discoveries’, in E. De Miro, L. Godart and A. Sacconi (eds.), Atti e memorie del secondo Congresso internazionale di micenologia, Roma-Napoli, 14–20 ottobre 1991. Volume 3: Archeologia. Incunabula Graeca 98. Rome: 1039–49. ‘To Qrhskeutikov Kevntro twn Mukhnwvn’, Praktikav tou E’ Dieqnou~ Sunedrivou Peloponnhsiakwvn Spoudovn. Athens: 147–68. ‘The Cult Center of Mycenae’, in I Culti Primordiali della Grecità alla luce delle scoperte di Tebe. Atti del Convegno internazionale (Roma, 24–25 febbraio 2000). Rome: 13–30. Archaeological Atlas of Mycenae. Athens.
Hägg, R. and 1981 N. Marinatos (eds.) Hägg, R., N. Marinatos, and G. Nordquist (eds.) Hitchcock, L., R. Laffineur and J. Crowley (eds.) Iakovidis, S.
1986
1987 1996
1997
2004
Iakovidis, S. and E. B. French Klein, N.
2003. 1997
‘Excavation of the Greek temples at Mycenae by the British School at Athens’, BSA 92, 247–322. Laffineur, R. and 2001 Potnia: Deities and Religion in the Aegean Bronze Age. R. Hägg Aegaeum 22. Liège. Loader, C. 1998 ‘Mycenaean drainage’, Bulletin of the Institute of Classical Studies 42: 205. Lupack, S. 2008 The Role of the Religious Sector in the Economy of Late Bronze Age Mycenaean Greece. BAR International Series 1858: 138–50. Mountjoy, P. A. 1981 Four Early Mycenaean Wells from the South Slope of the Acropolis at Athens. Miscellanea Graeca 4. Gent. Mylonas, G. E. 1966a Mycenae and the Mycenaean Age. Princeton. 1966b ‘Anaskafhv Mukhnwvn’, PAE, 107–110. 111
1968 1971 1972a
‘Anaskafhv Mukhnwvn’, PAE, 9–11. ‘Anaskafhv Mukhnwvn’, PAE, 151–56. The Cult Center of Mycenae. Pragmateivai th~ Akadhmiva~ Aqhnwvn, 33. Athens. (Abbreviated CC) 1972b ‘Anaskafhv Mukhnwvn’, PAE, 116–26. 1974 ‘Anaskafhv Mukhnwvn’, PAE, 90–91. 1977a Mycenae: A Guide to its Ruins and its History. Athens. (Abbreviated Mylonas Guide) 1977b Mycenaean Religion: Temples, Altars, and Temenea. Pragmateivai th~ Akadhmiva~ Aqhnwvn 39. Athens. 1983a ‘The Cult Centre of Mycenae’, Lecture, 1 December 1981, Proceedings of the British Academy 67. 1983b Mycenae Rich in Gold. Athens. Papademetriou, J. and 1961 ‘The last days of Mycenae: tablets and houses revealed W. D. Taylour, in the resumption of the late Professor Wace’s excavation’, Illustrated London News 23.ix.1961, 490–92, figs. 1, 14– 15. (Abbreviated ILN 1961) Petrakos, V. 2009 ‘Crhvsto~ Tsouvnta~ (1857–1934)’, O MENTWR, 91. Renfrew, C. et al. 1985 The Archaeology of Cult: The Sanctuary at Phylakopi. London. Shelton, K. S. 1993 ‘Tsountas’ chamber tombs at Mycenae’, AE 132, 187–210. 2006 ‘The long lasting effect of Tsountas on the study of Mycenae’, Mythos. La préhistoire égéenne du XIXe au XXIe siècle après J.-C. BCH suppl. 46: 159–64. 2007 ‘Foot soldiers and cannon fodder: the underrepresented majority of the Mycenaean civilization’, in S. P. Morris, and R. A. Laffineur (eds.), EPOS. Reconsidering Greek Epic and Aegean Bronze Age Archaeology. Aegaeum 28: 169–75. 2009a ‘The figurines from Petsas House and other find-spots at Mycenae’, in A-L. Schallin and P. Pakkanen (eds.), Encounters with Mycenaean Figures and Figurines. Papers presented at a Seminar at the Swedish Institute at Athens, 27–29 April 2001. Stockholm: 55–60. 2009b ‘Who wears the horns? Gender choices in Mycenaean terracotta figurines’, in K. Kopaka (ed.), Fylo. Engendering Prehistoric ‘Stratigraphies’ in the Aegean and the Mediterranean. Proceedings of an International Conference, University of Crete, Rethymno 2–5 June 2005. Aegaeum 30. Liège: 125–30. 2009c ‘Drinking, toasting, consumption and libation: Late Helladic IIIA pottery and a cup for every occasion’, in L. A. Hitchcock, R. Laffineur, and J. Crowley (eds.), DAIS. The Aegean Feast. Proceedings of the 12th International Aegean Conference. Aegaeum 29. Liège: 221–28.
112
2009d ‘Bringing down the house: changing construction techniques in LH IIIA 2 and IIIB Mycenae’, in D. Danielidou (ed.), Dwvron. Timhtikov~ tovmo~ gia ton Spuvro Iakwbivdh (A volume in honour of Spyros Iakovides). Academy of Athens: 635–46. 2010 ‘Citadel and settlement: a developing economy at Mycenae, the case of Petsas House’, in D. Pullen (ed.), Political Economies of the Aegean Bronze Age. Papers from the Langford Conference, 22–24 February 2007. Oxford: 184–204. Tournavitou, I. 2009 ‘Does size matter? Miniature pottery vessels in Minoan peak sanctuaries’, in A. D’Agata and A. Van de Moortel (eds.), Archaeologies of Cult: Essays on Ritual and Cult in Crete in Honor of Geraldine C. Gesell. Princeton: 213–30. Tsountas, C. 1886 ‘Anaskafaiv Mukhnwvn’, PAE, 59–79. 1887 ‘Anaskafaiv Mukhnwvn’, AE, 155–172. 1893 Mukhvnai kai mukhnai>kov~ politismov~. Athens. Tsountas, C., and 1897 The Mycenaean Age: A Study of the Monuments and J. I. Manatt. Culture of Prehomeric Greece. Boston and New York. Wace, A. J. B 1921/23 ‘Excavations at Mycenae’, BSA 25. 1949 Mycenae: An Archaeological History and Guide. Princeton. 1950 ‘Exploring the city of Agamemnon: this year’s excavations at Mycenae’, Illustrated London News 23.xii.1950, 1041, figs. 4–6. (Abbreviated ILN 1950). 1951 ‘Mycenae 1950’, JHS 71, 254–57. 1956 ‘Mycenae 1939–1955’, BSA 51, 103–22. Wace, A. J. B. and 1962 A Companion to Homer. London. F. H. Stubbings (eds.) Wace, H. and 1962 Mycenae. Guide. Meriden, Connecticut. (Abbreviated C. K. Williams Wace Guide). Wardle, K. 2003 ‘The “Cult Centre” at Mycenae and other sanctuaries in the Argo-Saronic Gulf and the NE Peloponnese: location and status’, in E. Konsolaki-Yiannopoulou (ed.), ARGOSARWNIKOS. Praktikav 1ou Dieqnouv~ Sunedrivou Istoriva~ kai Arcaiologiva~ tou Argosarwnikouv. Povro~, 26–29 Iounivou 1998. Athens: 317–32. 2015 ‘Reshaping the past: Where was the “Cult Centre” at Mycenae?’, in A-L. Schallin and I. Tournavitou (eds.), Mycenaeans up to date, Acta Instituti Atheniensis Regni Sueciae, 56. Stockholm: 577–96. Whittaker, H. 2001 ‘Reflections on the socio-political function of Mycenaean religion’, in Laffineur and Hägg 2001, 355–60.
113
Wright, J. C.
1994
2006
Wright, J. C. (ed.) 2004 Xenaki-Sakellariou, A. 1985
‘The spatial configuration of belief: the archaeology of Mycenaean religion’, in S. Alcock and R. Osborne (eds.), Placing the Gods: The Landscape of Greek Sanctuaries. Oxford: 37–78. ‘The formation of the Mycenaean palace’, in S. DegerJalkotzy and I. Lemos (eds.), Ancient Greece: from the Mycenaean Palaces to the Age of Homer. Edinburgh: 7–52. The Mycenaean Feast. Princeton. OiJ qalamwtoi; tavfoi tw`n Mukhnw`n ajnaskafh`~ Cr. Tsouvnta (1887–1898) / Les tombes à chambre de Mycènes. Fouilles de Chr. Tsountas (1887–1898). Paris.
114