Volume

1972

CONTENTS Council: Obituaries: Director's

Governing

Report

Cyrus the Great (558-529 B.C.), by Max Mallowan Sir Robert Ker Porter-Regency Artist and Traveller, by R. D. Barnett

A Royal Qajar Enamel, British and French Diplomacy

by B. W. Robinson
i8oo-I8io,

in Persia,

by R. M. Savory at Tikistan,

Saljiiq Monuments Who were the Chihilgani, Excavations

in Iran: II-The "Pir" Mausoleum by Robert Hillenbrand

the Forty Slaves of Sultan Shams al-Din Iltutmish of Delhi?, by Gavin Hambly Report, by David Whitehouse Borderlands,

at Siraf: Fifth Interim

and the Indo-Iranian Tepe Yahya Ig971-Mesopotamia C. C. Lamberg-Karlovsky by Elamites, Excavations Achaemenians at Haftavin and Anshan,

by John Hansman Report,

Tepe 1969: Second Preliminary by Charles Burney Shorter Notices

Survey of Excavations

THE

BRITISH

INSTITUTE
Price: ?5.oo

OF PERSIAN

STUDIES

c/o The British Academy, Burlington House, Piccadilly, London, WIV ONS

STATEMENT OF AIMS AND ACTIVITIES

I. The Institute has an establishment in Tehran at which British scholars, men of learning versed in the arts, friends of Iran, may reside and meet their Iranian colleagues in order to discuss with them subjectsof common interest; the arts, archaeology, history, literature, linguistics, religion, philosophy and cognate subjects. 2. The Institute provides accommodation for senior scholars and for teachers at British Universities in order that they may refreshthemselves at the source of knowledge from which their teaching derives. The same service is being rendered to younger students who show promise of developing interests in Persian studies. 3. The Institute, whilst concerned with Persian culture in the widest sense, is particularly concerned with the development of archaeological techniques, and seeks the co-operation of Iranian scholars and students in applying current methods to the resolution of archaeological and historical problems. 4. Archaeological excavation using modern scientific techniques as ancillary aids is one of the Institute's primary tasks. These activities, which entail a fresh appraisal of previous discoveries, have already yielded new historical, architectural, and archaeological evidence which is adding to our knowledge of the past and of its bearing on the modern world. 5. In pursuit of all the activities mentioned in the preceding paragraphs the Institute is gradually adding to its library, is collecting learned periodicals, and is publishing a journal, Iran, which appears annually. The Institute aims at editing and translating a series of Persian edited by the late ProfessorA. J. Arberry, has already texts, the first of which, the Humay-Nama, appeared. 6. The Institute arranges occasional seminars, lectures and conferences and enlists the help of distinguished scholars for this purpose. It will also aim at arranging small exhibitions with the object of demonstrating the importance of Persian culture and its attraction for the world of scholarship. 7. The Institute endeavours to collaborate with universities and educational institutions in Iran by all the means at its disposal and, when consulted, assists Iranian scholars with technical advice for directing them towards the appropriate channels in British universities.

MEMBERSHIP OF THE INSTITUTE

Anyone wishing to join the Institute should write to the Honorary Secretary, J. E. F. Gueritz, Esq., M.A., 85 Queen's Road, Richmond, Surrey. The annual subscription for Membership of the Institute is fI, while the total sum of ?4 entitles the subscriber to receive the Journal. Application Forms at back of Journal.

IRAN
Volume X I972

CONTENTS

Page
Governing Council Obituaries Director's Report Foreword Cyrus the Great (558-529 B.c.), by Max Mallowan Artist and Traveller, by R. D. Barnett Sir Robert Ker Porter-Regency A Royal Qajar Enamel, by B. W. Robinson British and French Diplomacy in Persia, I8oo-I8IO, by R. M. Savory " Pir" Mausoleum at Takistan, by Robert Saljtiq Monuments in Iran: II-The Hillenbrand Who were the Chihilgdni, the Forty Slaves of Sultdn Shams al-Din Iltutmish of Delhi?, by Gavin Hambly Excavations at Sirdf: Fifth Interim Report, by David Whitehouse Tepe Yahya 197I1-Mesopotamia Karlovsky and the Indo-Iranian Borderlands, by C. C. Lamberg89 IoI
127

ii iii ix xii I 19
25

31 45 57 63

Elamites, Achaemenians and Anshan, by John Hansman Excavations at Haftavan Tepe 1969: Second Preliminary Report, by Charles Burney Shorter Notices: A small bronze tripod-stand from Western Iran, by P. R. S. Moorey A possible source for the sculptures of the Audience Hall, Pasargadae, by Trudy S. Kawami Neutron Activation Analysis of some Obsidian Samples from Geological and Archaeological Sites, by A. Mahdavi and C. Bovington The Jamshidi of Khurasan: An Historical Note, by Andre Singer Staircase Minarets on the Persian Gulf, by David Whitehouse A recently identified fragment of the Cyrus Cylinder, by C. B. F. Walker Un Objet Cultuel (?) d'origine Iranienne, provenant de Nippur, by P. de Miroschedji An Inscribed Bronze Vessel from Luristrn, by W. G. Lambert and P. R. S. Moorey Survey of Excavations

143

165

THE

BRITISH

INSTITUTE

OF

PERSIAN

STUDIES

c/o The British Academy, Burlington House, Piccadilly, London, WIV ONS

BRITISH INSTITUTE OF PERSIAN STUDIES GOVERNING COUNCIL President *Sir MAX MALLOWAN, C.B.E., M.A., D.Lit., F.B.A., F.S.A. VicePresident BASIL GRAY, Esq., C.B., C.B.E., F.B.A. Members DEREK ALLEN, Esq., C.B., F.B.A. ProfessorSir HAROLD BAILEY, M.A., D.Phil., F.B.A. R. D. BARNETT, Esq., D.Lit., F.B.A., F.S.A. ProfessorJ. A. BOYLE, B.A., M.A., Ph.D. MICHAEL BROWNE, Esq., Q.C., M.A. Professor W. B. FISHER, B.A., D. de l'Univ., F.R.A.I. Dr. ILYA GERSHEVITCH, M.A., Ph.D., D.Litt., F.B.A. Professor A. K. S. LAMBTON, O.B.E., D.Lit., Ph.D., F.B.A. Professor SETON H. F. LLOYD, C.B.E., M.A., F.B.A., F.S.A., A.R.I.B.A. LAURENCE LOCKHART, Esq., Litt.D., Ph.D., F.R.Hist.S. RALPH H. PINDER-WILSON, Esq., M.A., F.S.A. BASIL W. ROBINSON, Esq., M.A., B.Litt. *Sir MORTIMER WHEELER, C.H., C.I.E., M.C., T.D., D.Lit., F.R.S., F.B.A., F.S.A. Sir DENIS WRIGHT, G.C.M.G., M.A. Professor R. C. ZAEHNER, M.A., F.B.A. Hon. Treasurer Sir JOHN LE ROUGETEL, K.C.M.G., M.C. Hon. Secretary JOHN E. F. GUERITZ, Esq., M.A.

Joint Hon. Editors Mrs. GEORGINA HERRMANN, D.Phil., F.S.A. Professor C. E. BOSWORTH, M.A., Ph.D.
OFFICERS IN IRAN

Director DAVID STRONACH, Esq., M.A., F.S.A.

Assistant Director ALEXANDER H. MORTON, Esq., B.A. P.O. Box 2617,

c/o The British Academy, Burlington House, Piccadilly, LONDON, WIV ONS

Tehran, IRAN
*DenotesFounderMember

SIR

MAURICE

BOWRA

OBITUARY
SIR MAURICE BOWRA

Maurice Bowra died on 4 July I971 at the age of 73 in Wadham College, Oxford, where he had served as Fellow and Warden for just on fifty years, with unswerving devotion. He was President of the British Academy at the time of our foundation, ten years ago, and by virtue of that office was one of the chief instruments for bringing us into existence. He was proud of our Institute. His love of Iran was reflected in the links between his College and that country: a munificent gift from H.R.H. Princess Ashraf was intended for Wadham College library and for the addition of Persian books to it. On his last visit to Iran as the guest of the Minister of Science, he took the opportunity of revisiting Isfahan and was deeply moved by its incomparable beauty. A three column obituary written with sympathy and brilliant understanding of his achievements has already appeared in The Times and the full length account of his life and works will in due course be published in the Proceedingsof the British Academyof which he was President from 1958-1962. Here it is our intention once more to express the sense of deep sorrow common to all who knew him, and particularly the indebtedness of the British Institute of Persian Studies, of which he was a Founder Member serving on Council, for all that he did on its behalf. Without his initiative and inspiration this new Foundation, now in its eleventh year would never have been born: with his practical judgment and perspicacity he succeeded in associating with it the scholars and men of the world required to make it a dynamic organization. In 1959 there was evidence in Tehran that many distinguished persons would favour the setting up of a British Institute of Persian Studies there. In Maurice Bowra, then President, and Mortimer Wheeler then Secretary of the British Academy, there was available an exceptional combination of dynamic opportunism to take advantage of this situation. The story of the first ddmarches has been " and there is engagingly told by Sir Mortimer in his book entitled " The British Academy 1949-1968 no need to repeat the account here. The official Persian welcome for this project coincided with a state visit to Tehran by Her Majesty Queen Elizabeth II and H.R.H. Prince Philip to H.I.M. the Shahanshah in March 1961-a visit anticipated by Bowra and Wheeler in December 1960. A generous offer of a house by the University of Tehran then under the Rectorship of Dr Ahmad Farhad, and munificent support from various sources in this country was followed by the ceremonial opening of it in Tehran, at an Inaugural Meeting on i i December 1961 for which Bowra in his characteristically confident manner arrived from Oxford no more than half an hour in advance, and proceeded to deliver a spirited and entertaining address on Edward Fitzgerald, subsequently published in the first volume of the newly established Journal of the Institute, to which he thus gave his distinguished patronage at the outset. The controversial problems which have confronted the name and work of Fitzgerald continued to exercise a fascination for him and he returned again in April 1968 to lecture in Tehran, Isfahan and Shiraz, most eloquently, on " Omar, Fitzgerald and Graves." This paper was followed by an address to Tehran University on " Some Peculiarities of English Poetry." From the store of his prodigious memory he quoted poems in many languages and at the same time expressed his appreciation of Persian poetry. The response from the young Persian students, who had assembled in large numbers to hear him, was enthusiastic. In that year he also delivered a talk on " Sir William Jones, Father of Oriental Studies " a hero after his own heart. For the last time he returned to Iran in February 197I, the year in which he was honoured by the Queen with the conferment of the Order of the Companion of Honour. He was also deeply gratified when through his Persian friends and admirers he received at that time an honorary doctorate from Tehran University. It was however late in life, when the Institute was founded, that he came to take a special interest in Persian studies, although it is true that as a don teaching for Honour Moderations at Oxford and

JOURNAL

OF

PERSIAN

STUDIES

devoted to classical Greece, he was fully conversant with the historical relationships between that country and Iran. It is characteristic of the man that when he came to Tehran he chose to speak, not about that subject, but on other relatively modern themes relating to Persian studies, for to the end of his life he was fully open to new sensations. One wonders therefore what he would have extracted from Persian literature, and especially from Persian poetry if he had found the time and the opportunity, with his brilliant gifts, to master the language. That was not to be, instead he chose to write about the impact of Persia on the two remarkable men, whom we have already named, both of them born within the British Isles-Fitzgerald the poet, and William Jones jurist, Sanskritist and author of the Persian about grammar through which the poet learnt the language: " Fitzgerald carried the grammar... with him for a year " and wrote in January 1853: " the Persian is really a great amusement to me ... As to Jones' Grammar, I have a sort of love for it." One senses that both of these remarkable men, so different in character, held an extraordinary attraction for him. The mixture in Fitzgerald of seriousness and frivolity, the doubts that assailed him, the deep-seated sense of poetry which could only be satisfied by translating, mistranslating, and even inventing and improving upon a Persian original made a congenial appeal to Bowra's character which in some ways came near to the subject of his study-one to which he returned again and again. In spite of the brilliance and quickness of his scholarship Bowra had a boundless compassion for all, even the less gifted, who strove in the race, and the more unexpected their methods, the more they pleased him. " Myself when young did eagerly frequent Doctor and Saint, and heard great Argument About it and about; but evermore Came out by the same Door as in I went." Very different was the appeal of William Jones, pioneer of Sanskrit studies in the West, versed in Persian, founder of the Asiatic Society of Bengal in 1784. Jurist, who gave us the first English translation of Sakfintala, the finest Sanskrit drama, translator also of the Code of Manu, the most important of the Sanskrit law-books, the first man also who printed an edition of a Sanskrit text-a short lyrical poem. Here was a great scholar and man of affairs who in the course of eleven years held high office in Calcutta, ambled up and down India in an ox-drawn cart and came to know in person the kaleidoscope of men, monuments and tongues. Alas, I have never been able to set eyes on Bowra's paper, which seems to be lost-one hopes only temporarily: perhaps these lines may assist in bringing to light what he said about this man of genius who was interested in all things, including zoology, who could express himself with rapidity and elegance and loved conversing with the humble peoples of the land. It is not surprising that Maurice Bowra, a natural admirer of William Jones, was an enthusiastic traveller and to be with him in Iran on one of his excursions to see the excavations that were being currently conducted by the Institute, was a journey of delight. He was a dedicated supporter of the work at Pasargadae, the splendid and beautiful new capital founded by Cyrus the Great, where David Stronach the first director of the Institute so successfully conducted three expeditions: the sight of the rugged and powerfully built citadel in its lovely setting may well have recalled for Bowra memories of Tiryns and he felt thoroughly at home as he gazed at the noble, gabled tomb of Cyrus, strangely so for a monarch so much admired by the Greeks. Nor is it Hellenic in appearance-appropriately surprising that at Persepolis his effervescent enthusiasm was aroused when he was confronted by processions of sculpture which represented the king's courtiers wearing the Greek inspired candys. The range of his interests likewise commended to him the work of David Whitehouse-also directed on behalf of the Institute-at the medieval seaport town of Siraf on the Persian Gulf where so much has been added to our knowledge of early Islamic archaeology through an architecture overlying extensive Sasanian remains. Such excavations were a joy to him because, as The Times obituary notice has well said: " He was always at pains to see literature in the context of the society in which it was written ". Understandably he was anxious to see the important results of such work get into print without undue delay-a besetting archaeological sin-for " he despised perfectionism and pedantry alike as forms of laziness, wasting on minutiae time better spent on grappling with larger tasks ". Originality and

OBIT'rUARY

genuine learning weighed more on the scales than exhaustive thoroughness: young and old loved him for the vitality and joie de vivre that ran through him and his work. The days that he spent in Persia during the last decade of his life were a fitting close to a triumphant career. All who met him succumbed to the flash of his wit, his ebullience, his enthusiasm, the warmth of his heart, and were sensitive to his excitement at discovery and his appreciation of historical archaeology. He was deeply interested in the monuments of Iran and the architecture which gave expression to the aspirations of the peoples associated with it. His epitaph has been aptly and perhaps half consciously written by himself in his last, comprehensive work on Pindar (I964) wherein he concentrated forty years of reflection about a subject dear to his heart.
yvEoGat dTvTr' XetS, Pl? Ad-cuE Zebi) JOV iToas oZCO lp' i(pKOtZO Kcaxkv. Oval Ovcvrolaot Inpnst.

(I. 5. I4-I6) Seek not to become Zeus. You have everything, If a share of these beautiful things comes to you. Mortal ends befit mortal men. Maurice Bowra in his own fields won the olive wreaths of victory and would have deserved an ode pitched in the key of Pindar's exalted excitement. Perhaps the poet considering his life work would have numbered him among the elect:
v K1hCUOCpK Evea ~scpav EXovwes oIKaG8

ob vtovr' dvu oUrcq6dvOv.

(N. 4. 76-77) There they were put to the test, And came not home without garlands, whose fruit is fame. M.E.L.M.

OBITUARY
ERIC SCHROEDER

Eric Schroeder, a vital figure in the study of Iranian art since he first went to Persia in 1931, died on the 27th March I97I in his sixty-sixth year. His strong attraction to this subject arose, as with many of his contemporaries, of whom I am one, from exposure to the unparalleled richness of the International Exhibition of Persian art at the Royal Academy in the first three months of I931. Schroeder had already gone some way towards launching himself in the area of Islamic studies through his participation in two seasons of excavation with the Oxford and Field Museum, Chicago, expedition to Kish in 1926/7 and 1927/8, the first while he was still an undergraduate at Corpus Christi, Oxford, of which he was a Scholar. He enjoyed the Arab scene, desert and town, probably more than the digging, but he would have continued had he not fallen foul of the British administration in Iraq by unauthorized crossing of the delicate Kurdish frontier with Turkey. He had however made friends with Henry Field as well as with David Talbot Rice, and might have waited for further chances of experience in archaeology had he not needed to make a living. And so, with characteristic impulsiveness and interest in experience he accepted a temporary teaching job at Milton academy in Massachusetts and held it for two years, until his marriage to Margaret Forbes of Boston. It was on their extended wedding tour that they were in London for the Persian exhibition. Schroeder had meanwhile been working up his notes from his Syrian travels and had something more than enthusiasm to show when he met Arthur Upham Pope, to whose inspiration, energy and faith was due not only the 1931 exhibition but also the great Survey of Persian Art which was at this moment being planned with the Oxford University Press in London. Pope invited Schroeder to join his team of surveyors in Iran. He and his wife went out by boat from Venice and on from Istanbul through the Black Sea. After completing his assignment of planning the Masjid-i Juma' in Isfahan, a most complex structure, for the Institute of Persian Art, the Schroeders settled for a time with Shiraz as a base and made several tours of standing monuments, making a photographic and measured survey. Then they returned to their home in Milton. Early in 1935 they were back in Persia, and in September that year he travelled to Leningrad for the Third International Congress of Iranian art and the spectacular exhibition arranged in fiftyeight galleries of the Hermitage museum. Schroeder presented a paper on the " Gumbad Jabal-i-Sang at Kirman " and then returned to continue his systematic studies of the early Islamic monuments, which formed the basis for the two chapters which he contributed to the architectural section of the Surveyof Persian Art, covering the " First Period " and the "Seljuk period ". These volumes did not appear until I939. They were marked by an independence of judgment and a freshness of treatment which did not commend them to all scholars, but were extremely characteristic of Schroeder's work, which was integral with his character. He was essentially a craftsman, in words and also in thought. He did not care for the barren wastes of scholarship but sought the heart of the matter, both in primary sources and in a study of the building or work of art which was before him. He had learnt his craft in the field, as had his contemporaries, for lack of any formal instruction in Islamic art. Later he himself contributed to courses at the Institute in New York an imaginative approach which was entirely stimulating. It was his good fortune meanwhile to work for two years in the Department of Asiatic art in the Boston Museum of Fine Arts when Coomaraswamy was still seated there as a presiding sage and he fell under his wise influence. Henceforth he sought the essential behind the work of art or rather within it, always looking at it as a product of someone in a particular situation at a particular time and involved in a whole cultural context. These qualities appear in the major contributions to Islamic studies which he made during the twenty years 1935-55, especially in his Persian Miniatures in the Fogg Museum of Art, a beautiful and substantial monograph based upon the then rather small, though choice collection of Harvard

OBITUARY

University, published early in I943; and the thick, 8oo-page volume of 1955 descriptively entitled : and and Muhammad's science, ribaldry finance People:a tale by anthologythereligionandpolitics,poetry violence, the Muslimsfrom the Age of Ignorance-to the EleventhCentury.The tremendous labour involved in of the fabricationof this great undertaking extending over seven years was rewarded by the warm appreciation of those best able to judge, such as Sir Hamilton Gibb and ProfessorG. E. von Grunebaum, both of whom commended it as a brilliant and successful achievement and work of art. Schroeder was a poet of distinction, too seldom heard, and able to construct a mosaic picture of the formative and classical periods of Islam from the primary sources. Unfortunately for students in Britain, both these books were published in very small editions and are hardly to be found here, even in University libraries. It is to be regretted that Schroeder did not produce a second edition of his Fogg catalogue with all the splendid additions since made to that collection. He was Keeper of Islamic Art at the Fogg Museum from October 1938 until the time of his death; in later years accepting no salary but still taking much care over the improvement of the collection and in the organization and the hanging of special exhibitions in the gallery. Miniature painting was his special interest; he was responsible for the book-painting section of the New York Persian exhibition of 1940, for which he wrote a booklet on this subject. His profound reflections upon the early classical periods of Persian and Mughal painting are to be found respectively in articles on " Ahmed Musa and Shams al-Din-a review of fourteenth century painting " in Ars Islamica (1939), and " The Troubled Image: an essay upon Mughal painting " in Art and Thought, VI a volume in honour of Dr Ananda K. Coomaraswamy (I947). Even for those who do not altogether accept his conclusions, these two writings present a serious challenge to deeper thinking. Since 1955 he gave the major part of his time and energy as a writer and student to a major work on astrology, a subject very much akin to the field of Islamic thought and its artistic expression. His friendsregretted that he ceased to travel and to write on the subjects of which he had previously acquired so deep an understanding; but he was convinced of the rightness of his choice and it is to be hoped that it will prove possible to publish at least a substantial part of his massive work in this new field, which combined his lifelong sense of the suprasensual with his interest in eastern ways of thought. This is not the place to evoke the stimulation of his infectious appreciation or his penetrating assessment. For a vivid personal appreciation readers may be referred to the notice in the Fogg Art Museum special bulletin, Acquisitions 1969-70, by Cary Welch.
BASIL GRAY

DIRECTOR'S REPORT
NovemberIst 1970 to October31st 1971 The present volume of Iran marks the international celebrations held in honour of the 2,5ooth Anniversary of the Foundation of the Iranian Monarchy and the Ioth Anniversary of the Institute's foundation. The past year has also seen the opening of a major Exhibition entitled " British Contributions to Persian Studies ", which was organized by the British Council in close collaboration with the Institute.

Lectures
Those lectures that were held at the Institute during the winter of 1970/71 included " The Influence of Sasanian Art on Umayyad Architecture" by Mr. Robert Hamilton, "Safavid Lacquer Doors" by Dr. Gdza Fehdrvari, "Sasanian and Islamic Siraf" by Dr. David Whitehouse, "The Relationship Mr. Stronach. between Persian and Mogul Jades" by Mr. John Irwin and "Tepe Nfish-iJdn, 1970" by Mrs. Stronach's film of the excavations followed this last lecture. In addition to the lecture programme, the Institute sponsored a series of informal seminars at which members of the University, the Archaeological Museum and other Institutes in Tehran were able to participate. Contributors to these discussions included Dr. Rafique Mughal of the Pakistan Archaeological Service, Mr. Ch. Azami, Dr. Gaza Fehdrvari, Dr. David Whitehouse and Mr. William Sumner of the American Institute of Iranian Studies.

TheHouse
Extensive redecoration has been undertaken in different parts of the Institute's premises and all sections of the Library are now housed in the Common Room and two adjoining rooms. A second short wheelbase Land Rover was also acquired early in the year.

Bowra Sir Maurice

Elsewhere in this Journal Sir Max Mallowan pays tribute to the memory of Sir Maurice Bowra, a Founder of the Institute, who died on July 4th. Sir Maurice's significant role in bringing the Institute into being was but one reflection of his great love for Iran and its varied accomplishments. Happily, his receipt of an honorary Doctorate from Tehran University brought him back to Iran last February, when he again visited the Institute and had the pleasure of revisiting Isfahan.

Fellows
Following a period of over a year in the field Mr. Peter Andrews returned to England in January. Apart from continuing his study of Nomad Tents, Mr. Andrews completed the notes for an illustrated Catalogue of Turcoman objects for an Exhibition sponsored by Northern Arts and arranged by the Abbot Hall Art Gallery, Kendal.' Mr. Andre Singer, Fellow for the session 1970/71, has been engaged in library work in Oxford and New Delhi and has taken part in the summer migration of the Teymuri, one of the four tribal groups from Khorasan that he is currently studying. Sound progress has been made and it is gratifying to know that the Social Sciences' Research Council has promised to support Mr. Singer's work in Iran for a further year.
x Copies of the Catalogue may be obtained on application to Miss Mary Burkett, Director, Abbot Hall Art Gallery, Kendal.

ix

Mr. Stuart Swiny spent part of last winter studying Museum collections in North America before visiting the excavations at Lumbini in Southern Nepal on behalf of the British Academy. His mound survey in western Iran has thrown new light on the Iron Age III period, notably on the historical geography of the central Zagros region. In August and September both Mr. and Mrs. Swiny took part in the excavations at Shahr-i Qilmis. Mr. Andrew Williamson continued his extensive archaeological survey of settlement and trade routes in Southern Iran throughout the period under review. Of particular interest was the discovery of a Sasanian port, covering an area of 450 hectares near Bushire, and a concentration of fourth to fifth millennia B.C. settlements near Tepe Yahya. Mr. Williamson's excavations at Tepe Dasht-i Deh-carried out under the auspices of the Tepe YaihydExpedition-are described on pp. 177, below. Mr. Julian Baldick, Fellow of the Institute for 1971/72, is studying thirteenth century mystical literature with scholars of the National University and Tehran University. Among former Fellows of the Institute, Dr. Edward Keall has joined the West Asian Department of the Royal Ontario Museum and Mr. Robert Hillenbrand has been appointed to a new Lectureship in Byzantine Studies at Edinburgh University. Assistant Director Mr. Alexander Morton has continued to work on his thesis on early Qajdr historians and has completed an article on "An Abbasid Hoard of Forged Coins". He has also visited Gorgan to examine historical documents in a local collection and has begun to make a study of two well-preserved caravanserais near Siah Kuh, south-east of Veramin. Director At the Eighth Annual General Meeting of the Institute, held at the British Academy on November 6th, Mr. Stronach gave an illustrated talk on " Cyrus and Darius at Pasargadae ". In outside lectures, he spoke on " The First Achaemenians " during the Centenary celebrations of the Ancient Iranian Cultural Society and on " Recent Archaeological Discoveries in Iran " at the Institute of Archaeology, Jerusalem. Mr. Stronach returned to London late in July, in order to deliver material connected with his book on Pasargadae, beforejoining Dr. Hansman in a second season of excavations at the largely Parthian site of Shahr-i Quimisnear Damghan. During the festivities at Persepolis in October Mr. Stronach had the honour of conducting Their Royal Highnesses Prince Philip and Princess Anne on a tour of the excavations at Pasargadae. Excavations The fifth season at Sirif yielded significant discoveries and provided final proof of the existence of a Sasanian port underneath the Islamic city. The most striking remnant of the Sasanian settlement is part of a fortified Palace with round towers at the angles and a monumental entrance, partly buried beneath the Great Mosque. In the Islamic levels, further work continued in the area of the hill-top Palace and new soundings were undertaken in the vicinity of a presumed naval compound near the shore. At Shahr-i Qflmis the excavation of a three-storied mud-brick building of Parthian date yielded a good range of pottery together with a number of fine seal and bezel ring impressions. Horse skulls from
this and other Parthian structures on the site are also thought to provide the first osteological indication of the Nicean Horse. Twenty-fifthCenturyCelebrations: InternationalIranologists' Congress A large number of Iranologists were privileged to attend the mid-October celebrations held in honour of the twenty-fifth centenary of the Persian empire. Those scholars who travelled to Shiraz, and who attended the International Iranologists' Congress as members of the British delegation, included Dr. R. D. Barnett, Professor J. A. Boyle, Mr. H. S. G. Darke, Dr. Ilya Gershevitch, Mr. Basil Gray,

Dr. John Hansman, ProfessorBernard Lewis, Dr. Laurence Lockhart, Mr. George Morrison, the Rev. Norman Sharpe and Mr. David Stronach. The main theme of the Congress dealt with the continuity of Iranian culture although, not unnaturally, those papers that referred to the Achaemenian period received special attention. In this context it is pleasing to note that one of the chapters of the forthcoming second volume of TheCambridge Historyof Iran-the volume concerned with the foundation and rapid expansion of the first worldempire in history-was released as a separate publication of the Cambridge Middle East Centre at the time of the celebrations at Persepolis. The fascicule in question contains Professor R. T. Hallock's chapter on " The Evidence of the PersepolisTablets" together with a Preface by the Editor of the second volume, Dr. Gershevitch. TenthAnniversary theInstitute of after the conclusion of the Centenary celebrations, on the morning of October I9th, the Directly Archaeological Service of Iran arranged an attractive Symposium in the New Building of the Iran Bastan Museum in order to mark the tenth anniversary of the Institute's foundation in the spring of Professor Sayyid Hossein Nasr, Dean of the Faculty of Letters of Tehran University, opened the programme with a wide-ranging review of the Institute's work, dwelling particularly on the impact that has been made in Tehran by the public lecture series and the library. Dr. Wolfram Kleiss, Acting Director of the German Archaeological Institute, spoke on behalf of the different foreign Institutes and Expeditions now active in Iran, while Mr. Stronach provided a report on the Institute's activities during the past decade. Dr. Richard Barnett, Keeper of the Department of Western Asiatic Antiquities at the British Museum, spoke on " The Achaemenian Horned Lion ", Mr. Morton on "A Qajdr Historian " and Mr. Basil Gray, Vice-President of the Institute, on " The Earliest Illustrated Manuscript of Khamseh of Nezami ". The formal proceedings of the morning were wound up by Mr. Firuz Bagherzadeh of the Iran Bastan Museum, who paid warm tribute to the role played by the Institutenot least through the medium of its publications. BritishCouncil Exhibition The Centenary celebrations were marked, finally, by a British Council Exhibition, "British Contributions to Persian Studies ", which opened in the New Building of the Iran Bastan Museum on the afternoon of October I9th. A most handsome presentation, made up of a select book display and a seriesof fifty separate panels that trace the evolution of Persian Studies in Great Britain over a period of some four hundred years, the Exhibition was designed and prepared by the Aids and Displays Department of the British Council. In the panels, photographic reproductions of early prints, manuscripts and portraits complement numerous photographs and drawings from the records of recent expeditions, and an excellent, lively text informs the whole. A Brochure that was prepared for the Exhibition includes an up-to-date bibliography together with brief notes on all those British University departments engaged in teaching Persian or offering other subjects connected with Iran. Tied to the theme of the Exhibition, the Brochure contains short essays on the Fine Arts, Archaeology, History and Literature by Mr. Basil Gray, Mr. Stronach, Dr. Gavin Hambly and Mr. David Blow. The Introduction was contributed by Sir Max Mallowan, President of the Institute.
The opening ceremonies in Tehran were made memorable by Dr. Isa Sadiq's address in which he evoked memories of E. G. Browne and other noted scholars. The Exhibition will continue to travel in the provinces of Iran until the spring of I972. 1961.

xi

FOREWORD
The following essay was written on behalf of the British Institute of Persian Studies, in honour of Cyrus the Great, at a time when his memory was being revived internationally, in order to mark the lapse of approximately two thousand and five hundred years since his death in trans-Oxiana. He has long been honoured in history as a man of vision and clemency, a model prince of whom Iran is justly proud, and during the month of October I97I his In Memoriam was celebrated in that country with brilliant pomp and ceremony on a most generous scale. Those who were privileged to be present in Tehran, Persepolis and Pasargadae enjoyed a wonderful spectacle and have paid tribute to its organization. This evocation of ancient Iranian history has been stimulating to scholarly research. Our Institute thanks His Imperial Majesty the Shahanshah for his generous hospitality of which we have been the beneficiaries ever since our Foundation ten years ago in 1961. It is pleasant to recall that in the course of the celebrations our director, Mr. David Stronach, had the honour of conducting Their Royal Highnesses Prince Philip and Princess Anne on a visit to the scene of his notable excavations in Pasargadae, and that the final instalment of his manuscript, which will be the permanent record of those excavations, will be deposited and ready for printing in March 1972. This new volume will mark the latest stage in a series of expeditions to Pasargadae. It is obvious that for our knowledge of Cyrus we depend much on the contemporary art and architecture which has been revealed by successive archaeologists at that site. Earliest of all came the late E. Herzfeld, and later Ali-Sami, one time Director of the Archaeological Institute of Persepolis: his little book, published in Shiraz in March 1956, entitled Pasargadae, The Oldest Imperial Capital of Iran, translated into English by the Rev. R. N. Sharp, has been used with gratitude by many a visitor to the site. Ali-Sami was followed by David Stronach, who between 196I and 1964 directed three fruitful campaigns on behalf of the British Institute of Persian Studies and thereby considerably enriched our knowledge of Achaemenian art, architecture, history and archaeology. I wish to acknowledge the help and advice which he has always generously given when consulted. I also wish to record my debt to the deep and penetrating study of Cyrus by Sidney Smith entitled Isaiah ChaptersXL-LV, being the SchweichLecturesof the British Academy 194o, published in 1944. His second lecture in that series, entitled "History of the years 556-539 B.C." is a masterpiece of historical perspective replete with recondite learning. Finally I should like to thank both Sir Edward Robinson and G. K. Jenkins for generous guidance over numismatic problems. Our progress in the understanding of the Achaemenian period in general and of Cyrus in particular is however entirely due to international co-operation: scholars from a multitude of countries have contributed their solutions and in so doing have been indebted to the generous spirit of collaboration which has been encouraged by the Iranian Government and the Iranian Antiquities Service over many years. It was a particular pleasure to recall in these pages the recent remarkable discovery by Ali Akbar Safaraz of a Palace built by Cyrus at no great distance from Bushire. M.E.L.M.

xii

CYRUS THE GREAT (558-529 B.C.) By Max Mallowan

From a lonely pillar at Pasargadae the phantom of Cyrus the Great flits across the ruins of the long deserted city and beckons us to consider the remains of one of the world's greatest imperial dynasties: by a strange freak of archaeology we have a fleeting glimpse of a royal image arrested for eternity in stone. Many will be familiar with this great winged figure (P1. II) naming the king, the sole survivor of four which once stood on opposite sides of two doorways in the hypostyle building known as Portal R at Pasargadae.1 The top of this monument, now vanished, was once inscribed in three languages, Old Persian, Elamite and Babylonian, and posterity must be grateful to Ker Porter2 who, just before 1820, copied the inscription, and likewise to Flandin and Coste who left another record of it twenty years later. The inscription itself makes a simple statement: "I (am) Cyrus the king an Achaemenian ", an authentic and contemporary record of the style used by the early forerunners of the dynasty, before the reign of Darius, when titles became pompous and elaborate. The crown worn by the king is a remarkable contrast to the simplicity of the inscription, and was perhaps intended to signify imperial majesty: a strange Persian version of a concept of the divine Pharaoh. The splendid splayed horns are those of the Ovis longipespalaeoaegyptiacus,a variety of ram apparently common during the Middle Kingdom in Egypt, but rare thereafter.3 It is clear that this unique crown must have come to Pasargadae from some unknown source on the coast of Phoenicia, and that it carried with it the prestige and authority of some quasi-Egyptian god which had thus travelled far beyond the Nile, in a form appropriate to Ba'al. A convincing explanation of this strange transference has recently been made by Dr. R. D. Barnett: he sees in it an expression of the oecumenical attitude of the Achaemenian kings who, from the time of Cyrus onwards, adopted a liberal policy of tolerance and conciliation towards the various religions embraced within their empire.4 I find this interpretation of the winged Cyrus the more attractive because in the nearby " Palace of Audience " to which Portal R gave access there were the remains of other carvings, including a god or priest clad with a fish cloak, clearly Assyrian in origin, and derived from the protective magical figures which had once adorned the portals of Nimrud and Nineveh, (P1. III). On another portal the foot of a raptorial bird reminds us not only of the legs of a divine guardian on a doorway of Sennacherib's Palace at Nineveh, but also of the claws of the dragons on the Ishtar Gate at Babylon.5 Here indeed at Pasargadae, in
1 I owe this information to David Stronach whose forthcoming book on the excavations which he directed on behalf of the British Institute of Persian Studies at Pasargadae will demonstrate the probability that four such figures, not more, existed and were in the symmetrical fashion of Achaemenian art placed in pairs. 2 Ker Porter, Travels in Georgia, Persia etc. (London 1822) II, pl. 13; C. Texier, Descriptionde l'Arminie, la Perse et la MisopotomieII, (Paris, 1852), pl. 84. For an illustration of Ker Porter's original water-colour sketch, now in the British Museum, see R. D. Barnett, " Sir Robert Ker Porter, Regency Artist and Traveller ", p. 22 and P1. IV, below. ' See F. L. Griffiths, Beni Hassan II, 15 and pl. 3-35 and A.A.A. IX (1922), pl. XXXII; F. E. Zeuner, A History of Domesticated Animals, pp. I54, 178. This type of sheep, long extinct in Egypt, is said to be represented by the modern Abyssinian, maned sheep. 4 R. D. Barnett "Anath, Ba'al and Pasargadae" in Milanges de l'Universiti Saint-Joseph, Tome XLV, fasc. 25 (1969), 407-22

Guardian TheFour-Winged Figure

6The

and figs. 5, 6. Pasargadae reliefs depicting a pair of human feet followed by the clawed feet of a mythical beast are illustrated in E. Herzfeld and F. Sarre, Iranische Felsreliefs(Berlin, 191o), fig. 84. Brick reliefs on the Processional Street at Babylon in E. at Koldewey, The Excavations Babylon,translated by A. S. Johns (London, 1914), figs. 32, 33-leg of a firrushand raptorial bird on p. 48. C. J. Gadd, The Stones of Assyria, pl. 17, divine See also E. guardians of Sennacherib at Nineveh. Strommenger, The Art of Mesopotamia(1962), pl. 226, which illustrates a four winged Assyrian figure facing right, carrying bucket and cone, and wearing a royal helmet with divine horns. This protective genius set up in the Palace of Sargon at Khorsabad, probably in 70o6 B.C. is in many respects iconographically in line with that of Cyrus and separated from it by not more than one hundred and sixty years. See also note 8, and below, T. Kawami, "A possible source for the sculpture of the Audience Hall, Pasargadae." Carl Nylander, lonians in Pasargadae (Uppsala, 1970), figs. 42a-b, p. 123.

JOURNAL

OF PERSIAN

STUDIES

these quasi Phoenician, Assyrian and Babylonian images, we have a forerunner of the Gate of All Nations which later on Darius was to erect at Persepolis. Let us return to our rather sinister winged figure (Pl. II) which may have been remembered by Herodotus who tells us that Cyrus saw in his sleep the eldest of the sons of Hystaspes (Darius) with wings upon his shoulders, shadowing with the one wing Asia, and Europe with the other." Herodotus therefore, as I surmise, may have known of the close connection between this type of winged figure and the image of Iranian majesty, which he associated with a dream prognosticating the king's death before his last, fatal campaign across the Oxus.7 The building in which originally four of these magnificent figures stood must have served as a processionalway and portal of access to the larger hypostyle palaces, P and S, which were in the vicinity, and we can imagine the intention, namely that the king should pass on his way to the state apartments and ceremonial halls under the cover of his guardian angel. This image at Pasargadae reminds me of a magical winged guardian which at Nimrud (Calah), the ancient military capital of Assyria, three centuries earlier had been set up in the N.W. palace of Ashurnasirpal and watched over the king as he passed along the corridorsfrom his private apartments to the throne-room and of another, about one hundred and fifty years older, erected by Sargon II in his Palace at Khorsabad.8 Authorities concerned with the history of Assyrian and Iranian art have been careful to point out that these winged prophylactic genii are not portraits of the king himself, even though the Pasargadae image has a superscription identifying the figure with Cyrus. But it is worth remembering that the corresponding Assyrian winged genii often carry inscriptions naming the king, and recounting his prowess and military exploits. It is clear to me, and perhaps I am heretical in expressing this opinion, that such figureswere directly associated with the magical and charismatic powers ordinarily attributed to a king in the Orient, and that these winged phantoms corresponded with a concept, as so often best expressed by Shakespeare, "There's such divinity doth hedge a king ".9 Much has been written about the simple formula of the trilingual inscription which once ran across the top of the image of Cyrus at Pasargadae, and it has been well demonstrated by R. Ghirshman that this is an authentic and contemporary record of the style used by the early forerunnersof the Achaemenids; it was from the reign of Darius onwards that titles became pompous and elaborate.10 The simple style of Cyrus was undoubtedly a reflection of Elamite royal custom and recalls the curt inscription at Choga Zanbil, I3th century B.C., which simply names the founder as " I Untash-Gal". None of the inscriptions at Pasargadae describe the monarch as "King of Kings". Nor do they refer to his paternity, only to the name of the clan, Achaemenian, as with Zoroaster who was never known by the name of his father, but as "the Spitamid". In this way these Pasargadae legends of which there were on probably not less than twenty-four,1" the Palace portals, probably in three languages, are a remarkable contrastto the one hundred and ten or more royall2 inscriptionsof the later members of the dynasty, scattered throughout Persepolis and elsewhere. Inconsistencies in the style of these titles which simply mention the king, sometimes the great king, and variations in orthography belong to a time when the royal house was groping to establish itself with the aid of formulae that later would become set. Many and elaborate dissertations have been written about whether or not the Old Persian writing on these monuments was actually inscribed during the lifetime of Cyrus or after his death. The reason for this apparently strange hesitation is a passage written by his second successor,Darius, in paragraph 70 on the great rock of Bisutun which has generally been taken to mean that Darius himself claimed to
6 Hdt. I, 209. 7 Hdt. I, 21o: " Thus Cyrus spoke in the belief that he was plotted against by Darius; but he missed the true meaning of the dream, which was sent by the deity to forewarn him, that he was due to die then and there, and that his kingdom was to fall at last to Darius." 8 Mallowan, Nimrudand its Remains,I, 103, for the reference to the winged figures along the passage ways. Op. cit. 12o and Folding Map III for position and description of passage P. The Assyrian relief was on the wall of the corridor between P and N. See also note 5. The later figure in Sargon's palace at Khorsabad is even more closely related to the winged one at Pasargadae, Pl. III. 9 The quotation is taken from the mouth of Claudius in Hamlet and in full runs, " There's such divinity doth hedge a king, that treason can but peep to what it would ": singularly apposite to the dream of Cyrus which foretold the transference of his dynasty to a collateral branch of the family, an event which occurred after the death of his son Cambyses. 10 R. H. Ghirshman, J.N.E.S. XXIV (1965), 246. 11 Carl Nylander, " Who Wrote the Inscriptions at Pasargadae ? ", OrientaliaSuecanaXVI (1967, Uppsala, 1968), 156. 1s Nylander, op. cit., p. I58.

CYRUS

THE

GREAT

have been the first to write in the Old Persian language,13that is, that he had invented the alphabetic cuneiform used to express his native language. Hitherto court writing had perforce used Elamite, Babylonian or Aramaic. Not all scholars accept this interpretation of the Bisutun inscription and many think, and probably rightly, that Darius was not denying the existence of prior inscriptions by Cyrus in Old Persian; on the contrary he was merely claiming that he was the first to make his proclamations through the medium of Aramaic as well as other languages. Nylander, after some hesitation, favoured the thesis that the Old Persian legends were added later, over the Elamite and Babylonian.14 But more recently Hallock applying methods appropriate to a computer has brilliantly and briefly demonstrated the probability-it cannot amount to certaintythat the simplicity of the signs and their restricted number are to be ascribed to Cyrus himself, whose scribe was in an experimental stage of writing, and that Darius added to and elaborated the system.'15 Archaeology comes to the support of epigraphy in elucidating the problem, for as a result of the excavations conducted by David Stronach on behalf of the British Institute of Persian Studies it has now been possible to establish the fact that while Pasargadae ceased to be the main imperial capital after the death of Cyrus, it was continuously inhabited down to Seleucid times, and moreover, from the use of the claw chisel, an instrument unknown to Cyrus' masons, it may be deduced that work was continued and completed in some of the palaces and other buildings during the reign of Darius, who added more inscriptions naming Cyrus after that monarch had died. We know from Bisutun16that Darius did nothing to recall the past achievements of Cyrus, neither did he name him among the eight rulers of his family whom he claimed as predecessors. Indeed the name of Cyrus is only mentioned inevitably, in a single context which tells us that the usurper Smerdis claimed to have been his son. No doubt Darius was jealous of the achievements of a major branch of the clan to which he was alien. It is interesting that this bitternessbetween the two separate branches of the royal house persisted over the centuries, for Ctesias the Greek physician who served for many years birth but a commoner who had ingratiated himself with Cambyses and rebelled against him.'7 We may however safely discount this statement for we know that Ctesias was an unreliable historian fed on tittle-tattle and harem gossip, and his story was merely aimed at discrediting the line of Cyrus and justifying the seizure of the throne by Darius. It seems likely that the occasion for this falsification of history would have been the recording of the unsuccessful revolt of Cyrus the Younger, killed at the battle of Cunaxa, whereupon Ctesias entered the service of Artaxerxes II who may well have been gratified at being exhibited as a true scion of a branch of the Achaemenian line, which needed support for its legitimacy. The tendentious pseudo-history of Ctesias as reported by Nikolaos of Damascus thus strengthensthe general credibility of Herodotus, and of Xenophon, who record the legitimate claim to the throne of Cyrus the Great as the son of Cambyses I, and therefore grandson of Cyrus I.18 We have seen that on the great trilingual inscription carved to the order of Darius in 520 B.C. on the precipitous rock at Bisutun there is but a bare mention of the name of Cyrus. But paradoxically there is in that same inscription a wonderful testimony to the extent of Cyrus' power, for in the sixth paragraph we have a record of the twenty three provinces which Darius the king proclaimed as having come to him by the favour of Ahuramazda-and he might well have added by the legacy of Cyrus. This we must deduce, for the inscription was completed in the third year of Darius' reign, and the first
two years were wholly occupied in repressing rebellions by pretenders to the throne-the
13

at the court of Artaxerxes II, 405-359 B.C., claimed that Cyrus the Great was not an Achaemenian by

false Smerdis

14

Ghirshman, op. cit., has well exposed this fallacy. On the Cyrus inscription there are two lines of Old Persian at the top, then below one line of Elamite, and finally at the bottom one line of Babylonian. 16 R. T. Hallock, "On the Old Persian Signs", J.N.E.S. XXIX (1970), 52-5. Hallock counted the frequency of occurrence of 2, 3, 4 and 5 wedge signs and their distribution in the various inscriptions. The apparent early occurrence of the lowfrequency values ku (attached to the two wedge sign) and ru (attached to a 3 wedge sign) both employed in the writing of the name (Kurug) strongly suggests that CMa or some lost

Cyrus inscription served as basic text. This is not incontrovertible evidence, nor does any such evidence exist. 16R. G. Kent, Old Persian Grammar Texts, p. I59and 17 Ctesias apud Photium, ed. by R. Henry, " Soci&t6d'Edition Les
Belles Lettres" (Paris, 1959), Tome 18 Whose first in I, Ch. 72, p. Io6.

reign appeared history through the discovery at Nineveh by R. Campbell Thompson of this earlier Cyrus in the annals of Ashur-bani-pal who exacted tribute from him, see A.A.A. XX (1931-2), 95. Parentage in Hdt. I, 46, Xenophon,

I, Cyrop. II, i.

JOURNAL

OF PERSIAN

STUDIES

among others who tried to establish that he was the son of Cyrus and brother to Cambyses II. Now we know that Cambyses' reign must have been largely occupied with his conquest of Egypt, that he was troubled by internal rebellions-an unbalanced, perhaps insane monarch"9who in his comparatively short reign of seven or eight years could not have acquired the vast empire bequeathed to Darius. We may thereforehave every confidence in the later Greek and Roman historians who have left us a record of Cyrus' domains. Thus we may accept the list of Darius' provinces on Bisutun as a more or less accurate presentation of what Cyrus had first achieved for the Achaemenian empire. The Old Persian version records these countries as follows: Persia, Elam, Babylonia, Assyria, Arabia, Egypt, those who are beside the sea, Sardis, Ionia, Media, Armenia, Cappadocia, Parthia, Drangiana, Aria, Chorasmia, Bactria, Sogdiana, Gandhara, Scythia, Sattagydia, Arachosia, Maka: in all XXIII provinces. We may therefore deduce that territorially Cyrus had first acquired for a united Iran tracts of land which extended from the Greek cities on the western seaboard of Asia Minor, Palestine, Syria and Babylonia into the Caucasus and Transcaspian provinces, through what is now Afghanistan into the distant territory of Bactria and the vast tract of land between the Oxus and the Jaxartes rivers. There is little doubt that he had set himself the task of conquering Egypt, but death intervened and this target was left to his son Cambyses who fulfilled the grand design. It is instructive to compare the list on the rock at Bisutun with the more detailed list in Herodotus20 which applies to the state of the empire controlled by Darius at the end of his thirty-six year reign when he had substantially added to the legacy at his disposal at the beginning of it. Here we have the empire for administrative purposes divided into twenty provinces, each a separate satrapy paying a specified tribute: the notable and distinctive additions were of course Egypt and India, or rather the Indians who paid "a tribute exceeding that of every other people, to wit, three hundred and sixty talents of gold dust" which was reckoned at thirteen times the worth of silver or 4680 talents. This would have been the most preciousjewel of all for Cyrus' crown, but although in encompassing Bactria, Sogdiana and Gandhara he may well have gazed at the Pamirs and the Hindu Kush, he could not quite achieve that more distant goal.21 But it is known from historical and topographical evidence that he established a powerful frontier fortress named Cyropolis (Kurkath),22which was identified as his foundation by Alexander the Great, and later by the Arabs, on the river Jaxartes. In this city he established for the first time an Achaemenian frontier post on the very boundaries of Central Asia-a bulwark against the hordes of migrant tribes who were perpetually threatening Iran from as far afield as Outer Mongolia. We may however safely accept the testimony of Alexander's historian Arrian who says " but no one else ever invaded India, not even Cyrus, son of Cambyses, though he made an expedition against the Scythians and in all other ways was the most energetic of the kings of Asia ". But he had the thrill of discovering new and unknown peoples, of incorporating them in the comity of Iranian nations, of exacting gifts in his honour and doubtless of laying down and initiating the lines on which his enormously stretched civil service was to operate under the Persian system of satrapies. Herodotus tells us:23 "During all the reign of Cyrus, there were no fixed tributes, but the nations severally brought gifts to the king. On account of this and other such like doings, the Persians say that Darius was a huckster, Cambyses a master, and Cyrus a father; for Darius looked to making a gain in everything; Cambyses was harsh and reckless; while Cyrus was gentle and provided them with all manner of goods." From Darius' tribute list we are able to compare the amounts paid in silver talents by each province and we may note that the ninth province of Darius, Babylonia and Assyria, which contributed a
thousand talents was by far the richest prize: the acquisition of these Mesopotamian territories occurred in the last decade of Cyrus' reign, 539 B.c. It is also noticeable that as soon as Sardis and Lydia fell
19 Hdt. III, 30. He was an overbearing monarch; killed the Apis Bull and commited suicide when his throne was in danger. 20 Hdt. III, 90-94. 21 Arrian, Indica 9, Io. 22 E. Benveniste, " La Ville de Cyreschata ", Journal Asiatique CCXXXIV Ann6es 1943-5 (1947), for a summary of the evidence and references to the classical authorities. Alexander, according to Strabo XI, 4, and Quintus Curtius VII, 6, 20, had desired to spare the city out of respect for the memory of its founder, Cyrus. The position of this site has been marked on the map taken from the Nonesuch Herodotus, but there omitted, see p. 5. Hdt. III, 89.

23

aIA ",OXINE'S

KALE
NUOR AREL KME K

s-

YG aPIs DAIs c s E1r1 IURZ

on ARME9

CA css
O
Ass]A A
'1I
?

...

.... ,.

SINU oPE
S'o
ACIA MIEPA
AIDAA ISS
kP

s WI
o0 OIAUKH ta
PASS"OF MD MA

DA RIA OR SA
"
4k

BA

OALBAEA
SARDUT-S0MAUt oPAA o.lEPAIN SINI RAT1
oUF
ERS

0I
sepMs

7U

SYCIA oTAK.
"EMADAN

NI
DASH7-J,.AU..
R 'psi
"--D'

E MAN N
Y

EC

LEP Co "4~Vt

13"IL

I.E
o YIFi

D E SJE
R-MIDI
A-N.S:A" A-D
aTHIOPIANE

BELIRUTL ETMIPIANRUB SAMRAGDOM U

BA

G
EL IK-"

Ao
MASAANDA

oo

JDU

with theaddition Cyropolis.(By courtesy T under Herodotus, of of empire CyrustheGreatandDarius, aftertheNonesuch of Map, showingtheextent theAchaemenian

JOURNAL

OF PERSIAN

STUDIES

into his lap, probably about 545 B.C., or possibly a year or two later,24 together with the wealthy Greek cities of Ionia, he must have been enriched by a great accretion of gold and silver indispensable to the financing of his campaigns. Much money must also have been added to the Persian treasury through Cyrus' incorporation of the provinces that stretched along and behind the southern Iranian seaboard, particularly the hinterland which contained the lines of caravan routes to India-Parsa, Drangiana, Arachosia and the territory called Maka, probably the ancient Makkan which at this period presumably referred to territory on either side of the Persian Gulf and may have included Oman. We have only to turn to Darius' tribute lists to appreciate the outstanding value of these satrapies. And here archaeology comes tells us that there were other palaces besides Persepolisand Pasargadae to the aid of history, for Strabo25 and he mentions one "on the coast near TaokEas it is called". We may ask if this is not to be identified with a remarkable discovery appropriately made in Cyrus' twenty fifth centennial year by an archaeological expedition under the direction of Ali Akbar Safaraz: two lines of beautifully and simply carved column bases found in the course of excavating a palace of Cyrus the Great, some 30 kilometres from the Persian Gulf off the highway connecting Bushire and Borazjan, in an area which had previously turned up Achaemenian artefacts.26This, we may be certain, is a herald of other discoveries to come, and in the course of time, remains attributable to Cyrus himself will surely be found in more than one outpost of his empire. In assessingthe accretion of wealth that came to Iran from the time of Cyrus onwards we should not forget the great influx of men, animals as well as timber and other commodities-gold from Sardis and Bactria, which together with Sogdiana supplied lapis lazuli and carnelian: in the time of Darius, Cilicia, according to Herodotus, contributed 360 white horses and Babylonia 40ooboy eunuchs. AncientTravel These considerations lead us to broach another topic, namely ancient travel. How long did it take for the armies, embassies, officials and traders to crossfrom one end of the empire to another? We know that relays of post horses supplied at regular stations spread out at a day's interval over the royal roads and the Royal Mail were elaborately organized, as indeed they had to be if so vast an empire was not to fall apart.27 How long did it take Cyrus to march across to Sardis in the campaign which resulted in the capture of that city ? According to Herodotus28the distance from Susa to Sardis over the approved route worked out at about 1700 miles and was accomplished in 90 days-at the rate of about 19 miles a day-an exhausting rate of travel-it is likely that the preliminary subjugation of Western Asia Minor and the intervening territory must have taken Cyrus at least a year. In the opposite direction we have remarkable evidence of envoys travelling from Iran to Afghanistan, a distance which I would reckon at approximately 12oo miles over one of three possible routes. On one of the Persepolis Fortress tablets published by R. T. Hallock29we have a record of a guide named Zivandu' and his five boys who escort a lone woman, perhaps a princess all the way from Susa to Kandahar (KandaraS). This is one of the set of texts of the fifth century B.c. which give evidence of distinguished couriers who escorted Indians, Cappadocians, Egyptians, men of Sardis and others. " Nothing mortal travels so fast as these messengers" says Herodotus in another context-speaking about the Royal Mail-" and these men will not be hindered from accomplishing at their best speed the distance which they have to go, either by snow, or rain, or heat, or by the darknessof night".30 These distinguished courierswho escorted special parties were the ancestors of those who conduct Swan's Tours and the like--they had to know every
H. T. Wade-Gery in J.H.S. LXXI (1951), p. 219, note 38, deduces from Herodotus that Sardis did not fall before 544 B.C. -battle of Pallene 546, fall of Sardis perceptibly later; he therefore prefers Herodotus' evidence to the conjecture based on the Nabonidus Chronicle that Cyrus defeated Croesus in 547 B.c. See also note 34. 25 Strabo XV, 3. 26 " Palace of Cyrus Unearthed ", A. T. Zand in TehranJournal, June I, 1971, with illustration of three column bases. 27 Hdt. VIII, 98 and note in the Nonesuch Herodotus on the Persian posts. See also Esther VIII, Io, Ahasuerus (Xerxes)
24

485-464 B.C. sent letters all over the empire by means of swift horses that were bred from the royal stud. 28 Hdt. V, 53 and note in the Nonesuch Herodotus. 29 R. T. Hallock, PersepolisFortification Tablets (O.LP. XCII, 1969), PF 144o and 1550. The ration documents, written in Elamite, date from the thirteenth through the twenty-eighth year of Darius I, 509-494 B.C. 30 Hdt. VIII, 98. The mounted couriers of the express service to whom Herodotus refers are named as pirradazil in the Elamite texts.

CYRUS

THE

GREAT

inch of the road and to be persona grata in the potentially hostile or friendly territories through which they passed. Perhaps the most distinguished of all these messengers was the angel Raphael who in disguise and for wages offered Tobit to escort his son all the way from Nineveh through Ecbatana (Hamadan) to Rhages (near Tehran). " Do you know the way to Media ? " Tobias asked. " Yes " he said " I have often been there. I am familiar with all the routes and know them well."31 Journeys of this kind had of course to be provided for by the house of Tobit, Jewish bankers, who during the late Assyrian empire under Sennacherib and Ezarhaddon had wisely dispersed their interests both in Assyria and in Media, in anticipation of a clash between the two. Vitae Curriculum A consideration of the vast distances that had to be covered by the armies of Cyrus for the acquisition of the empire must lead us to reflect, if only briefly, on the order and sequence of his campaigns. We know that he was about 40 years old when he came to the throne and about 70 when he died on campaign in Transoxiana. He must have been a very tough old man: so far as we know, his curriculum vitae runs approximately as follows: born 598 B.c., son of Cambyses I and of Mandane daughter of his Astyages king of the Medes; married, probably not later than 578 B.c. to Cassandane32 best loved wife who bore him a son and heir, Cambyses II; when she died he ordered all his subjects to go into mourning; married secondly to Amytis; ascended to the throne 558 B.c; conquest of Medes, capture of their king Astyages, and of the Median capital Ecbatana (Hamadan) 550 B.c.; submission of Hyrcania and Parthia33549 B.c.; submission of Lydia, capture of Sardis and of king Croesus about 545 B.c.; capture of Babylon and king Nabonidus 539 B.c.; death beyond the Oxus fighting the Massagetae 529 B.C.34 of Chronology Conquests The situation which confronted Cyrus at the beginning of his reign has been admirably and succinctly expounded by Sidney Smith in his Schweich Lectures of 194o as follows: "The Assyriansby hard fighting had kept open the passes into Media, Armenia and Cappadocia for three centuries. After the fall of Nineveh (in 612 B.c.) the eastern and northern passes were held by the Medes, the northwestern were only open by favour of the Cilicians and Lydians, and subject to a treaty. The traffic from the Phoenician ports was interfered with by pirates from the Lydian coast, and by Greeks, the allies of
TobitV, 6-8. Long distance couriers in the reign of Artaxerxes, see EstherVIII, 9-12. 32 Hdt. II, I for Cassandanewife of Cyrus. SeealsoHdt. I, Io7, io8 for Mandane mother of Cyrus who was married by Astyages to a Persian of good family " but much inferior to a Mede of even middle condition "; but his father was none the less king of Anshan, at that time no doubt a petty vassal state of Media. CtesiasapudPhotiumed. R. Henry, Persica, Chapters I, 2 pp. I0o13, Collection Lebegue, (1947) differs from Herodotus but is unreliable. -3 Hdt. I, 130, 153, 177, and notes in Nonesuch Herodotus on the sequence of these campaigns. 34 I accept the general chronology proposed by Sidney Smith, SchweichLectures,op. cit., p. 29. " According to Dinon, Cyrus the Great was 40 years old when he came to the throne, and reigned 30 years. Dinon is not reliable, and the figures are suspicious, but they fit the probabilities. Cyrus died in 529; his reign may well have begun in 558, and he may have been born in 598 for his grandfather was already king of Parsumash in 640." Smith also recalls, op. cit. 123 note 35, that Herodotus gives 29 years for the length of Cyrus' reign against Ctesias and Justin 30o and that the difference may be due to the months of the Babylonian year, " beginning of Kingship ". The date of the fall of Sardis however is still uncertain. It seems improbable that it can be Lydia that is mentioned in a corrupt passage in the Nabonidus Chronicle which states that in Iyyar (that is May) of the 9th year of Nabonidus Cyrus marched to the land of Lu(?) ... -the reading of the cunei31

form signs is uncertain-and fought its king, for there would, here, in any case be a discrepancy between the Chronicle and the account in Herodotus I, 77 as regards the timing of the capture of Sardis, which tells us that Croesus had decided to summon his allies in the Spring in the 5th month after his indecisive encounter with Cyrus. He was taken by surprise, for Cyrus proceeded immediately to the attack which must have happened in November and not in May, the month mentioned in the Nabonidus Chronicle for Cyrus' march. See also note 24 and note 6o. I am indebted to Professor O. R. Gurney for informing me that H. T. Wade-Gery had written to him, calling attention to these discrepancies; Wade-Gery has referred very briefly to the fact that the date of the capture of Sardis is lost, in J.H.S. LXXI (1951), p. 215 note 15, but on p. 229 note 38, on the basis of Herodotus' evidence in connection with Greek historical events, proposes 544 B.c. as a more likely date. We may safely conclude that Sardis did not fall in 547 nor in 546, the date given by Eusebius, for there is no mention of Cyrus in the Nabonidus Chronicle for that year, but the city could have fallen at any time between 545 and the attack on Babylon in 540. A date as near to 545 as possible on grounds of historical probability is acceptable, for Cyrus must have required some years to consolidate his conquests in Asia Minor before proceeding to attack Babylonia, his most valuable prize. For dates proposed by the Greek Chronologers on see How and Wells, A Commentary Herodotus 98. Note also I, that Apollodoros apudDiog. Laert.I, 37 f gives a date which may Chronik(1902). be fixed at 546/5, see F. Jacoby, Apollodors

JOURNAL

OF PERSIAN

STUDIES

the Egyptians." Further, Nabonidus, the last king of Babylon, "had attempted to secure that trade from the Persian Gulf was not diverted westwards; it was desirable to control the increasingly prosperous trade of Southern Arabia and the Red Sea ".3 Cyrus, in his ambitious attempt to assume empire in western Asia was governed by such considerations. The cycle of trade had to penetrate these political and geographical barriers, and we may watch him in pursuit of these objectives during the thirty years of his reign. The first decade, or most of it, was occupied in consolidating his position under the king of the Medes, whose daughter, Mandane, was his mother, that is to say, Astyages was his grandfather on the maternal side. These years must have been devoted by Cyrus to increasing his authority over the confederation of Persian tribes and to the intrigues which led to the capture of Ecbatana. Thereafter he inherited a not inconsiderable Median empire about which we know little except in so far as scattered references in the later Greek histories36 bear witness to its extensive ramifications in the east. The year after the defeat of Astyages, Hyrcania and Parthia in Transcaspia acknowledged a new master, readily enough perhaps, and this allegiance secured Cyrus' eastern flanks and enabled him to conduct the first of his great imperial campaigns-against Sardis; but probably that had entailed a distant march to the Caspian in order to establish his authority there. It seems unlikely that he would have attempted the campaigns further east against Bactria and the Sacae or Sogdiana at so early a stage in his career. The take-over of Parthia and Hyrcania, formerly Median dependencies, provided him with a bulwark of security before embarking on his distant marches into Asia Minor where, according to Herodotus, "Cyrus in person subjected the upper regions, conquering every nation, and not suffering one to escape ".3 He also adds: "of these conquests I shall pass by the greater portion, and give an account of those only which gave him most trouble, and are the worthiest of mention "-proof that much information was available for which he could not find room in his histories-an assurance that we need not wholly discard the amplified accounts in the later Greek histories of Cyrus' activities here and elsewhere.38 The second decade must have been largely spent in the conquest and consolidation not only of the whole of Iran, but also of those distant flanks which were to put him in touch with the countries controlled by the great cities now named Merv and Samarcand. How many campaigns were involved we do not know, though as more excavations are to be conducted on these eastern confines, I do not doubt that we shall gradually come into possession of many wonderful clues. We must however accept the fact previously mentioned that Cyrus never reached India, a design no doubt frustrated by his death in the field some hundreds of miles distant from its frontier. Arrian's testimony in this respect is certainly sound.39 How many campaigns in the field did Cyrus conduct in the course of this second decade 545-539 B.c. that is between the capture of Sardis and of Babylon ? It is difficult to believe that he would have been away from home during all that time, for continued absence from the seat of government for so long would have constituted a danger to his dynasty. We know that his great imperial predecessors the Assyrians had conducted annual campaigns, but such marches to and fro rarely exceeded some five or six hundred miles, whereas Cyrus had to undertake the moving of armies up to five times that distance. It is reasonable to think that in the course of this second decade, approximately, he must have returned to his bases in Elam and Media at least three times,40 however confident he may have been in his satraps and allowing for the fact that his kingdom was sustained by the continuous acquisition of wealth that surged to the homeland on the tides of victory.

accept historicity of some Median control in the east--at least as far as Hyrcania and Parthia, but Arrian was not a very reliable historian, for he believed that the Indians living between the Indus and the Kabul rivers had been subject to the Assyrians. Indica 9, Io, also I, 3. 37 Hdt. I, I77. " I know besides three ways in which the story of 38 Hdt. I, 96. is told, all differing from my own narrative." Cyrus 39 Indica9, IO. Pliny, N.H. VI, 92 recorded that Kapisa (modern

85 Sidney Smith, op. cit., p. 39. 36 We may the

40

Kafshan) perhaps not much more than 250 miles from the upper Indus river was destroyed by Cyrus. The place is thought by A. D. H. Bivar to have been north of Kabul. See his chapter in CentralAsia, ed. Gavin Hambly (1969), p. 20. Xenophon, Cyrop.VII, 1-3 states that before his death, Cyrus now a very old man, returned for the seventh time in his reign to Persia. This statement may well be a true record of the facts--long absences punctuated by visits to the seat of government at home, about three times during each of the last two decades when he was acquiring his empire.

CYRUS

THE

GREAT

We have unfortunately no knowledge of the order in which he conducted his campaigns during this period, but perhaps we may be justified in postulating that Alexander, whose historians must have had intelligence about Cyrus' military logistics, trod on much the same tracks, and for this reason I would suggest that although at the outset he must have marched due eastwards below the Caspian, at least one march was conducted by him down from the direction of Bactria (Afghanistan) through the Helmand basin, Arachosia, Gedrosia, Kerman and Makran. Again we may consult Arrian, as I think, with some confidence; Nearchos reported that an attempt to reach India through South Baluchistan resulted in the loss of the greater part of Cyrus' army.41 However that may be, Berosus follows Herodotus in asserting that Cyrus possessed all the rest of Asia at the time when he first attacked Babylon in 540 B.C.;42 as is confirmed by the famous inscribed cylinder in which Cyrus says " The kings of the Westland dwelling in tents, all those brought heavy tribute to me in Babylon and kissed my feet ".43 This climax of his military career was, according to Xenophon,44 immediately preceded by the reduction of Arabia where Nabonidus for ten years exiled to the oasis of Tema' had perhaps been attempting to establish control over the caravan routes and trade on the Persian Gulf. At all events the year 539 B.C. marks the triumphant beginning of Cyrus' third and last decade. Precisely what Cyrus achieved militarily during the last, third decade of his reign we do not know, although the consensus of Greek historians is that he died somewhere at the north-eastern end of the empire, between Oxus and Jaxartes, according to Herodotus slain in battle by Tomyris, queen of the a Scythian tribe. It is logical to assume that this was a period of consolidation, Massagetae-perhaps as is attested by the fact that he appointed his son and designated successor Cambyses as religious ruler of Babylon, while, according to Ctesias, before his death he designated Cambyses' younger brother as ruler over the eastern provinces.45 We may, therefore, be certain that Cyrus conducted at least one great march on his eastern frontier during that last stage of his life.

Capture Babylon of
Whatever he may have achieved militarily between 539 and 529 B.c. it is clear that the moment of his greatest triumph was in 539 B.c. when according to the Nabonidus Chronicle " In the month of Arahsamna, the third day, Cyrus entered Babylon. Green twigs, doubtless reeds or rushes to smooth the path of his chariot, were spread in front of him. The state of' peace' was imposed on all the city. Cyrus sent greetings to all Babylon."46 The event marked a critical shift in the balance of power held by the contending forces of Western Asia. The peoples of Iran had for three quarters of a century been allied to Babylon, ever since the Medes made a combined attack on Nineveh in 612 and overthrew the empire of Assyria. Even as late as about 550 B.c. Cyrus had an understanding with Nabonidus that enabled this Babylonian monarch to invest Harran which had suffered Median occupation, and rebuild the temple of the Moon God. But when once Lydia had been overthrown the balance of power was upset and the interests of Babylonia and Iran were in conflict. The capture of Babylon, richest of all the Persian satrapies, inevitably brought in its train hegemony over the rich cities of Syria and Palestine, as well as the Phoenician coast which had formerly fallen within the Babylonian orbit, if not entirely under its control. There followed the return of the Jewish exiles from Babylon, and the charter granted by Cyrus which we shall consider briefly later. Maurice Dunand has recently well demonstrated that the tolerance and liberal help granted on this occasion for the return of the Jews to Zion was the corner stone of a policy which was designed to take over the
41

42

Anabasis VI. 24. Indica 9. " Cyrus son of Cambyses had got through with only seven survivors, for Cyrus did come into these parts intending to invade the country of India; but before he could do so he lost the greater part of his army by the barrenness and difficulty of the route". He entered the city on 3rd Marcheswan, 539 B.c. (corrected date). Sidney Smith, " A Persian Verse Account of Nabonidus ", Babylonian Historical Texts (1924), p. 28. It should be noted that Sidney Smith, SchweichLectures,op. cit. p. ii9 note 18,

drew attention to the fact that all the years given in his BabylonianHistoricial Texts are one too late-an error in the use of the Canon of Ptolemy. Corrected dating has been used here. 43 See in general the note in Nonesuch Herodotus I, 177 and Dougherty, Nabonidusand Belshazzar (1929), p. 161. VII, 4. 44 Cyropaedia 45 Hdt. I, 208. Nonesuch note and Ctesias, Persica, 8. 46 Nabonidus Chronicle III, 12-22, and see Xenophon, Cyropaedia
VII, 5, 20-26.

10

JOURNAL

OF PERSIAN

STUDIES

remnants of the old Babylonian empire in Phoenicia and Palestine. There is much archaeological evidence that the rebuilding of the temple of Jerusalem was followed by the repair and foundation of a chain of fortified sites which ran down from the Gulf of Issus at about the latitude of modern Alexandretta to the marches of Palestine where the Jews were doubtless expected to co-operate in sustaining a defensive bulwark against Egypt. The prophet who is usually referred to as Deutero-Isaiah leaves us in no doubt about his loyalty. " Thus says the Lord to his anointed Cyrus, whose right hand I have of Jewry and Iran thus went hand in hand and although we do not yet know to what extent Cyrus himself took a hand in engineering these defences, his impetus must be discerned behind the great quadrangular blocks of dressed masonry which the Achaemenians encouraged on the margins of river beds and rocky hill sides, as well as in the temples and fortresses at Sidon, Byblos, Banyas, AmrithMarathus, Jerusalem, Lachish and other sites on the way to Egypt.47 Records:TheCyrus Cuneiform Cylinder Nothing could be more interesting historically than the examination of Cyrus' campaigns in Babylonia and of his final peaceful entry into the capital. For our reconstruction of these events we depend not only on Greek sources, principally Herodotus, who wrote less than a century later, and Xenophon, but on the contemporary cuneiform records themselves-both the laconic and incisive Nabonidus Chronicle relating to these last days and the vivid records of Cyrus himself, first the famous Cyrus cylinder which he must have deposited in Babylon and then a highly coloured and prejudiced verse account. These two latter documents are masterpieces of political propaganda and although not the first of their kind in the ancient world are skilled instruments of tendentious history: and in addition there are the official proclamations, inscriptions on bricks and the ordinary day to day business records from which we may strike a balance and arrive at a proper appreciation of the situation. It is significant that in many of the documents which relate to Cyrus' activities in Babylonia the titulary describeshim as king of Anshan which, at this period and in this context, may well have denoted the extensive tract of territorysouth east of Elam where Pasargadae and later, Persepolis,were situated.48 The Cyrus cylinder in an appropriate Babylonian form, and in contrast to the simple titulary of the Pasargadae inscription says: " I am Cyrus, king of the world, great king, legitimate king, king of Sumer and Akkad, king of the four quarters (of the world), son of Cambyses,great king, king of Anshan, descendant of Teispes, great king, king of Anshan, of a family (which) always (exercised) kingship; whose rule Bel and Nabu love, whom they want as king to please their hearts."49 In this way Cyrus characteristically and with diplomatic astuteness assimilated the acceptable style of titulary to which Babylonia, heir of Sumerian kingship, had by long tradition been accustomed. We cannot doubt that as with all great conquerors the elements of luck and good timing enabled Cyrus to triumph over Babylonia, which proved to be the richest of all his satrapies. Nabonidus, the last king of Babylon, who ascended the throne in 556 B.c., was tainted by his northern ancestry; he was
41 Maurice Dunand, "La D6fense du Front Medit6rranean de l'empire Ach6menide" in The Role of the Phoenicians in the Interactions Mediterranean Civilizations, ed. by William Ward, of Beirut American University Centennial Publication (1968). The full list of sites is therein mentioned: doubtless Phoenician masons were often used in their construction, but plans and siting have in many cases an Achaemenian impress. 48 Sidney Smith, Schweich Lectures,op. cit., p. 28 and map No. I opposite, " both Parsumash and Anzan designate the province round Pasargadae" and pp. 120-23. In Proc. Soc. British AcademyLV (1970), " Elamite Problems ", p. 256, I took the view that Anshan was "approximately coterminous with the present day territory of the Bakhtiari ", a theory which perhaps finds support in the Sumerian Epic entitled Enmerkar and the Lord of Aratta, translation by S. N. Kramer (U. of Pa. 1952) introduction page I and lines 7o-82, io6-io, 165-67, where Anshan appears clearly to have been associated with mountainous territory in the reign of Enmerkar, third millennium B.c., E.D.II period. Dr. Georgina Herrmann has also kindly recalled for me a passage translated by S. N. Kramer in The Sumerians (1963) P. 273 with reference to the Epic entitled Lugalbanda and Enmerkar. Lugalbanda a henchman of the latter volunteered to seek help for his master by making a dangerous journey to the city of Aratta. "He takes up his weapons, crosses the seven mountains that reach from one end of Anshan to the other "-or, as the poet puts it, from the "shoulder" of Anshan to the "head" of Anshan-and finally arrives with joyful step at his destination. The identification of frontiers, never firmly fixed in antiquity is however always difficult and boundaries frequently changed in the course of time, so that we need not accept evidence which holds good for the third millennium as applicable to later periods. I understand that John Hansman's view coincides more nearly with that quoted from Sidney Smith qv. supra, see " Elamites, Achaemenians and Anshan ", pp. Io1-25 below. J. B. Pritchard, A.N.E. T. (1950), p. 316, mention of Anshan in the fourth line.

grasped, to subdue nations before him and ungird the loins of kings .. ." (Isaiah 45.1). The interests

49

CYRUS

THE

GREAT

11

the son of the great high priestessof the temple of the moon god Sin, named E-Hul-Hul in Harran and reigned for seventeen troubled years. He soon fell foul of the priesthood by introducing an alien theology and an unacceptable image of the moon in eclipse that amounted to heretical practice. So uneasy did relations with the priests become that he went into self-imposed exile in the oasis of Tema' in N.W. Arabia, a ten year Odyssey, leaving behind him as regent his son Belshazzarwho, according to the Old Testament,sowitnessed the final writing on the wall of the palace at Babylon, on the eve of its downfall. During the many years of his exile the New Year festival at Babylon which required the personal presence of the king could not be celebrated, and the people of that city and of the surrounding country did not forgive him for this disgrace. Thus when Cyrus began to undertake his campaign in 540 B.c. Babylonian affairs were at a low ebb and there was little loyalty to the ruler of the country. Gobryas (Gubaru), governor of Gutium,51who had been the principal general of Nebuchadrezzar, defected to the side of Cyrus, who within a year was able to make a peaceful entry into Babylon having first, according to Herodotus, used a cunning stratagem, namely the diversion of a main canal in order to overcome a formidable obstruction to entry. The Cyrus cylinder tells us of the restoration of the derelict city and of its sanctuaries, of a return of and of the re-institution of the New the gods to their ancient enclosuresboth in Babylon and elsewhere52 Year festival, for Marduk in his mercy had granted forgiveness on account of the probity and right conduct of the new prince, Cyrus, who respected the ways of the gods of the country. Not only did Cyrus burn the false images of his predecessor, but he instituted within the city a new slum clearance scheme and: "brought relief to their dilapidated housing (thus) putting an end to their (main)
complaints ".

It seems most probable that Cyrus' entry had been assisted by the large community of Jews which had been deported by Nebuchadrezzar II fromJudah underJehoiakin53sixty years before; these exiles believed that in spite of their captive prosperity they would receive more liberal treatment at Persian hands. In this they were right and, as we learn from the book of Ezra, Cyrus gave a charter to the Jews for the restoration of the temple in Jerusalem and the return of the utensils sequestered by Nebuchadrezzar.54 It is interesting that no images or statuary came into question for these would have been anathema in the Jewish Temple. However that may be, part of the community under Sheshbazzar (probably a Jew) and Zerubbabel returned to Zion and joined the small remnant that still held Israel together.55 The remainder were loath to leave the commercial prosperity which they must obviously have acquired under Babylonian rule, however intolerant in religious matters their Babylonian masters may have been. Policy Cyrus'Toleration:Foreign Religious toleration was a remarkable feature of Persian rule and there is no question that Cyrus himself was a liberal-minded promoter of this humane and intelligent policy. Many other examples of Cyrus' help in restoring Babylonian and alien shrines could be quoted, for this was part of a well thought-out policy. At Ur of the Chaldees, for example, a great centre for the worship of the Moon god Nanna, Cyrus installed a new gate in the great Temenos wall which had first been built by Nebuchadrezzar as a sacred enclosure for the principal temples in that city. Within the lining of the gatebox socket, Cyrus' bricks proclaimed his might and liberality and he himself restored one of the

5o Daniel 5, 5; wrongly therein described as the king. 51 Sidney Smith, Schweich Lectures, op. cit., 47 considers that

Gubaru is to be identified with the Gobryas in Xenophon's account and is not to be confused with the Persian Gubaru who was appointed by Cyrus satrap of Babylon immediately after his own entry. The latter was probably the Gobryas mentioned by Herodotus, one of Darius' trusty officers. 52 U.V.B. I, 48. He is known to have contributed to the restoration of E-anna in Uruk (Erech).

4 Ezra 6, 1-5. 55Ezra 2 gives a register of those who returned but there is a discrepancy between the figures given there and those recorded in Nehemaiah and Esdras. None the less, between 30,oo0 and 40,000 Jews took advantage of the offer. A much smaller contingent of 1,5oo persons returned later from Babylonia under Ezra in the seventh year of Artaxerxes (458 B.C.).

5' II Kings 24, o0.

12

JOURNAL

OF PERSIAN

STUDIES

temples.56 At the neighbouring city of Uruk, one of the principal urban centres of S. Babylonia, a command was given to four of the king's officers for the provision of bowmen to guard the shepherds: "in accordance with the yearly arrangement in the barracks which are upon the great river", that is on the Euphrates. This edict delivered in the very first year of his occupation, shows that Cyrus was determined to keep order in the administration of Babylonia immediately after occupation, and that his Civil Affairs Officers had in readiness a properly prepared and well thought out plan for taking over the administration of a newly conquered country.57 From these official proclamations, as well as from many other documents, we know that business went on as usual, for the selling of date groves, the acquisition of land, the renting of ships, the transactions of goldsmiths and the like. Eight out of the twelve Achaemenian business documents found at Ur were written in his reign.58 One remarkable characteristic which many historians have attributed to Cyrus is his clemency to fallen rulers, in the true fashion of mediaeval chivalry. We may consider the treatment of three of his chief opponents: Croesus, Astyages and Nabonidus. It is Herodotus' story that Cyrus condemned Croesus to be burnt on a pyre and there follows the legend of the miraculous intervention of Apollo to save him, when the flames could not have been extinguished by human hands. But historians have rightly objected that the pollution of fire by human sacrifice would have been anathema and contrary to Persian religious practice. Bacchylides59 who lived nearer to the time of Croesus-he was born about forty years after the fall of Sardis-preserved the truth: that Croesus attempted suicide. We may infer that Cyrus saved him from the flames-a more probable story, and one that accords with Greek tradition, namely that Cyrus used conquered princes to advise him in the administration of their former domains. There is no need to invoke a mutilated passage in the Nabonidus Chronicle which some scholars have interpreted as meaning that Cyrus marched against the country Ly ... possibly Lydia60 and killed its king, for it is now recognized that the Akkadian word iduk can mean fought, not necessarily killed.61 Moreover the sign read as Ly is almost illegible. Even if the sign may be read Ly ... another interpretation is possible: that the country was Ly(cia) conquered by Cyrus before its neighbour Caria which capitulated without striking a blow because the Carians had seen the fearful fate that had befallen its neighbours. Incidentally Lycia is mentioned in Hittite records and once in an Akkadian text from the ancient Syrian city of Ugarit.62 It may be however that the Nabonidus Chronicle in this passage refers to some other country, neither Lydia nor Lycia, and in any case this text as it stands cannot be taken as evidence that Cyrus killed Croesus: we may still accept the testimony of the Greek historians who reckoned that his life was saved. As regards Astyages there is no question that Cyrus treated him honourably after the fall of the Median Empire and the investment of Ecbatana. Indeed Cyrus was his grandson, and grandsons do not kill their own grandfathers. The case of Nabonidus the last king of Babylon was different, even though Abydenus according to Eusebius and Josephus alleged that the captive king was honourably
56 Cyrus adopted the pompous style of titulary used by royalty in Babylon. Burnt bricks of Erech (Uruk) are inscribed " Cyrus builder of Esagila and Ezida, son of Cambyses, Great King am I ". In this manner he honoured Marduk and Nabu under a titulary used previously by Nebuchadrezzar. See George Smith, T.S.B.A. (1873), opp. p. 146, Weissbach, Die Keilinschriften der Achiimeniden (i91 1) pp. 8-9. In the same city, Uruk, he also contributed to the upkeep of E-anna. Similarly at Ur bricks were inscribed "Cyrus King of all, King of Anshan, son of Cambyses, King of Anshan. The great gods have delivered all the lands into my hand; the land I have made to swell in a peaceful habitation." UET I (1928), No. I94.; U.E. IX (1962), 7-8; Antiqs. Journal III, No. 4 (Oct. 1923), P- 315, pl. XXV. from Erech, Neo-Babylonianto Persian 57 R. P. Dougherty, Archives Periods (1933), p. 34, No. o02. The document was dated " Ist year of Cyrus, king of countries ". 58 Dougherty, op. cit., Nos. 92, Io1, Io9. r9 Bacchylides, ed. by R. C. Jebb, EpinikionIII, lines 28 f., dated 468 B.C. Commentary on this episode, see A. R. Burn, Persia and the Greeks,p. 42 and note 9. There was good oriental precedent for the burning of a defeated prince: best known is the case of the Assyrian Shamash-shum-ukin, in Greek legend known as Sardanapalus who, when defeated by his brother Ashur-bani-pal, perished in the flames of his own palace: see Mallowan, Nimrud and its Remains I, 246; the record of that event was preserved on a fragment of a prism found in "the library", room N.T. I2 of the building known as Ezida, in Calah. Other examples of self-immolation on a funeral pyre are: Boges, Hdt. VII, Io7; Hamilcar, Hdt. VII, 167; Zimri, I Kings I6, 18. 60 A.N.E. T. (1950), p. 306. Sidney Smith, BabylonianHistorical Texts, p. 112, line 16. See also note 34 above. 61 H. Tadmor, " Historical Implications of the Correct Rendering of Akkadian ddku", J.N.E.S. XVII (1958), 129. 62J. Garstang and 0. R. Gurney, The Geographyof the Late Hittite Empire, see index under Lycia and Lukka Lands, particularly p. 82 for various references and discussion of topographical problems in the Hittite records. See also Ugaritica V, 87, letter from the king of Ugarit to the king of Alashia mentioning mdt lukkaa,line 23, and note 5 on pp. 88-9 on geographical identifications at this period.

Pl. L The Tombof Cyrus expedition) showingthe entrance, from the west. (Photo: OliveKitson,Pasargadae

with thenameof Cyrus. (Photo: OliveKitson, Pasargadae expedition) Figure,onceinscribed P1. II. The Winged

Pl. III.

of Winged genius from the Palace of SargonII of Assyria,Khorsabad. (By kindpermission Messrs. Thames& Hudsonand from The Art of Mesopotamia, by E. Strommenger Max Hirmer.)

Pl. IV. The greatstone of platform the Takhtor Citadel from thesouth-west.(Photo: Pasargadae expedition)

CYRUS

THE

GREAT

13

exiled to Carmania. The virulence of Cyrus' propaganda against Nabonidus however and the deadly hostility of the Babylonian priesthood allowed of no generous solution: Cyrus was nothing if not a diplomatist and knew that here mercy would have been dangerous. Xenophon knew better and has left us a dramatic picture of the king dagger in hand awaiting death at the hands of two of Cyrus' nobles Gadatas and Gobryas, probably in the great Throne-Room of his palace at Babylon.63 Sidney Smith has commented aptly: " The stories of the invariably merciful treatment of conquered kings by Cyrus are propaganda material in the legends, and also testimony to a new conscience in international affairs, for no conqueror would previously have desired such a reputation."64 The Gadatas who is mentioned as one of the assassins of Nabonidus may possibly have been the officer who was satrap in Ionia under Darius,65 who gave him a sharp remand for not having respected the privileges accorded to the priests of Apollo in a sanctuary near Magnesia. The only predecessor who can have rewarded the priests of Apollo in this district was Cyrus for whom " a favourable oracle was worth more than a battle ".66 This instance of intelligent diplomacy towards foreign priesthoods is one that is wholly in accord with what we know of Cyrus' policy when laying down the foundations of the Achaemenian empire.

Religion
In Babylonia as in Judah and elsewhere we have seen ample evidence of Cyrus' toleration in religious matters and there is no trace of national fanaticism. What then were the beliefs of Cyrus himself and of the Iranian State ? It is clear from many sources that polytheism was practised in Iran, and no doubt the fire cult played a prominent part in religious ceremonial. At Nush-iJan, an ancient Median site of the eighth century B.C., not far from Hamadan, David Stronach has recently unearthed a fire tower in which the sacred fire was extinguished, most probably if we are to follow the later testimony ofDiodorus67 on the death of the king, and in Achaemenian times there are many representations as well as actual monuments of fire altars. It may also be recalled that about one generation before Cyrus, north eastern Iran had come under the influence of the great religious teacher and prophet, Zoroaster, whose main scene of activity and preaching was Khwarezm (Chorasmia) in territory which today includes Merv and Herat. It was perhaps in or about 586 B.c that Zoroaster at the age of 42 made a notable convert in king Vistaspa, apparently the last of his royal line.68 Perhaps Cyrus who about fifty years later incorporated that king's territory within his empire may have come under the influence of Zoroaster's teaching-he may even have been a Zoroastrian himself, although there is no evidence yet for saying that Zoroastrianism became a state religion before the time of Darius and his successors. But it seems probable that the noble teachings of this prophet who, for the first time in history, preached the doctrine of free will, would have found a kindred spirit in the liberal minded Cyrus. In this powerful new religion it was man who held the balance between good and evil; the eternal combat between good and evil, strikingly represented in Iranian religion through the contrast between the powers of light and darkness. The doctrine harmonizes well with the part played by fire in the older polytheism of Iran. One may sense that Cyrus' new concept of mercy and justice may have emanated from such beliefs.

"6 Xenophon, 65

The objection to this assumption is that if the Gadatas of Darius is to be identified with the satrap of Cyrus he would have been a young man for office under the latter and an old man under the former. The identification is possible but not probable. 66 Sidney Smith, op. cit., 41. 67 Diodorus Siculus XVII, I14, 4. s8 W. B. Henning, Zoroaster, Ratanbai Katrak Lectures i949, (1951) makes a good case for three possible dates of Zoroaster; 630-553, 628-551, 618-541, remarkably supported by a Syrian writer, Theodore bar QOni, 628 years and seven months before Christ: the latter probably arrived at this date by using

4" Sidney Smith, Schweich Lectures,op. cit., 36.

VII, v. 29, 30. Cyropaedia

the book which Theodore of Mopsuestia had written against the Magian religion. More recently in B.S.O.A.S. XXXIII
(1970), Mary Boyce, " On the calendar of Zoroastrian Feasts ",

has argued for a date of 665-588 B.C. on the assumption that it was his death that was thus calculated in the Sasanian calendar-to which must be added 77 years for the known length of his life. Cogent arguments are used in accordance with more recent Zoroastrian practice which involves a religious duty to keep the days of remembrances for the dead of his own family, hence to record the years from the death of an ancestor. But in my opinion, Henning's lower dates are more appropriate to the historical setting, which seems to require a date for Zoroaster not more than a generation before Cyrusperhaps less. The debate is likely to continue.

14

JOURNAL

OF PERSIAN

STUDIES

Portraitof Cyrus Great military achievements speak for themselves, but when we seek to receive a glimpse of his character and qualities we inevitably turn to Xenophon's Cyropaedia, which is an artist's portrait of the Ideal Ruler and the best form of Government. This is a picture of Cyrus the Great seen through the form of Cyrusthe Younger, the hero slain on the field of battle-so greatly admired by the mercenary Greek who served him. In this work, as Gilbert Murray has aptly said "Truth is subordinate to edification". Let us recognize immediately that Xenophon does violence to the historicalfacts. " Media was subdued by force and treachery in the lifetime of Astyages, not voluntarily ceded to Cyrus by Cyaxares as the dowry of his daughter "-" the beautiful account of the peaceful passing of Cyrus is wholly out of accord with the well-established record of his violent death in the battle against the picture of a great hero could not have been painted had there not been a credible memory of such a Cyrus-Cyrus the Great, addressedin the Old Testament as the " Lord's anointed ". " The Lord, the God of heaven " has given him " all the Kingdoms of the earth "-equally lauded by Ezra, and by Isaiah,70who says of Cyrus, " He is my shepherd, and he shall fulfil all my purpose." We should therefore recognize that although the account which Xenophon has left us of Cyrus' campaign in Asia Minor is not strictly historical, it gives us an insight both into the Achaemenians' military mind and diplomatic practices: Cyrus II, the Great, the model prince, may have initiated many of these. Thus we read of his extraordinarygeneralship, his lightning Napoleonic thrusts and the way in which after a preliminary skirmish at Pteria in Cappadocia he made an immediate unexpected attack on Croesusand thereby worsted him." Most interesting, in addition to the accounts of ruses and stratagems for capturing fortresses,are his exceptionally intelligent handling of his soldiers, his understanding of their psychology and his ability to make them fear, respect and love him: the mark of a true general. We also have a remarkableaccount of the training of Persiansoldiers through lion hunting, a picture which corresponds very well with what we see on the Assyrian reliefs, and there can be no doubt that this formed part of the specific military training for young officers in the Persian army: a practice which no doubt they had received from their imperial Assyrian predecessors.7' Xenophon is singularly ignorant about the northern peoples concerned, confuses Syrians, Cappadocians and Arabs and seems to be unaware of the role played by Babylonia and Assyria. But he has one extraordinarilyinteresting passage concerning the north Syrian frontier which, if properly followed, would help solve one of the intractable Palestinian problems today. Cyrus realized that some of the northern frontier forts were an insoluble bone of contention between the frontiers of Asia Minor and Syria. He persuaded the warring parties concerned to disarm and to let him hold the key forts in his supra-national hands, while flocks on both side could cross the frontier unmolested and likewise farmers could have freedom of movement. This today would be the answer to the Golan Heights and many other similar problems.72a It was entirely owing to the imaginative insight of Cyrus that through his military and administrative skill Iran was for the first time brought into a close political relationship with the rich trading cities of the East Greek world and in touch with her merchants and bankers, many of whom were ready to accept Persian suzerainty rather than the cut-throat and spiteful competition of rival Greek cities, and for that reason Miletus, with its powerful fleet, a rival to Sardis and Ephesus, welcomed Persian intervention and did nothing to support Greek resistance.
Choiceof Pasargadae as Capital We conclude, as we began, with a brief reference to the site of Pasargadae itself. Why did Cyrus choose the place for the building of his new capital city ? How much of the architecture and sculpture bears the authentic impress of his work ? The answer to the first question must in my opinion rest on a clue provided by Herodotus,73 the relevant passage runs as follows: "Now the Persian nation is made
69

Massagetae (529

B.C.)

".69

It was his son Cambyses, and not Cyrus who conquered Egypt. But this

W. Miller, Introduction Xenophon's to Cyropaedia(Loeb edition, 1914), pp. IX, X. Death of Cyrus in Cyrop.VIII, 7, a contrast to the historical account in Hdt. I, 214. 36, 7o II Chronicles 23; Ezra I 1-2; Isaiah 44, 28; 45, 1.

71

71 Xenophon, op. cit. I ii, 9-11; VIII i, 38; I iv, 16-2472a Xenophon, III, 21.
78

Hdt. I, 76, 77.

Hdt. I,

Cyropaedia

I25-

CYRUS

THE

GREAT

15

up of many tribes. Those which Cyrus assembled and persuaded to revolt from the Medes were the principal ones on which all the others are dependent. These are the Pasargadae, the Maraphians, and the Maspians, of which the Pasargadae are the noblest. The Achaemenidae, from whom spring all the Perseid kings, form one of their clans." It seems a legitimate inference that Cyrus was a member of the "noblest tribe" which frequented this district and, like Sargon II of Assyriawhen he came to the throne, built his capital in the heart of his true homeland. Moreover, the omens for that foundation were no doubt deemed to be good, for later Greek historians74 certify that it was here that Cyrus won the decisive victory over Astyages the Mede that resulted in the submission of the Median peoples and the transfer of power and the seat of authority to the Persians. Some scholars, but not all, would derive his name from the river Kur, the principal waterway of this district.75 Finally, in answer to the second question which seeks to know what at Pasargadae are the authentic marks of Cyrus, we are indebted to David Stronach who, aided by Carl Nylander, has rendered a notable archaeological service in demonstrating the technological differences between the art and architecture of Cyrus and that of his successors. This evidence may be examined in detail in various journals76as well as in the large and final volume on the excavations at Pasargadae now in the press. The Citadel First of all we may notice the splendid Takht or citadel (P1. IV) with its massive fortifications of rusticated ashlar masonry" containing masons' marks which are Lydian in character, as well as Lydian-style lead78and iron butterfly clamps which are markedly different from those used in the time of Darius. Under Cyrus the wide dove-tail type was the only form used, but this shrunk under Darius and had become more or less straight-sided by 450 B.c. The employment of Lydian masons in Iran no doubt followed the capture of Sardis after 545 B.c. and must have been increasingly encouraged by the use of silver coinage under Darius. We may however note that Lydian craftsmen had long been in demand, for Lydian names figure in the issues of rations at Babylon shortly after 600 B.C.79 Introduction Coinage of Cyrus' conquest of Lydia resulted in yet another important innovation in so far as Persia was concerned, namely the introduction of coinage into his realm-an innovation usually attributed to Darius. Herodotus80recorded that it was Croesus who introduced the first coinage of gold and silver side by side, and the very scarce heavy lion and bull coinage is attributable to him.81 But there are, as Sir Edward Robinson has informed me, two known Croesid " lion and bull " issues the second of which has been discovered in various hoards, all later than Croesus, and lighter than the old standardconforming more closely in weight with the first darics and silver sigloi. It is therefore tempting to infer that it was Cyrus, not Darius, who first introduced current coins into his empire, a medium of payment which became indispensable as skilled foreign labour was increasingly attracted to employment in the capital cities of Iran. Perhaps therefore we may be justified in crediting Cyrus with the far-seeing ability to adopt a monetary innovation which was destined to revolutionize the older methods of fiscal procedure, as well as commerce, in Western Asia. The Zendan theKa'bah and It was Cyrus too who built the great tower named the Zendan82whereon there are no traces of the claw chisel frequently, but not invariably, used by Darius' masons. The purpose and function of this
Polyaenus VII, 6, 7; Strabo XV, 3, 8. The final victories were gained in Persian territory at the frontier pass of Pasargadae. See also Nicolas of Damascus in Jacoby. Frag. d. gr. Hist. Ha, p. 367, and note in Nonesuch Herodotus, I, 128, I. 75 Strabo XV, 3, 6. This appears to have been the ancient name of the river Polvar. 78 See especially David Stronach in Iran I-III (1963-6), and C. Nylander, op. cit., and Ionians in Pasargadae. 77 Iran I (1963), pls. II-III. 78 The clamps at Pasargadae are of lead and iron. Each has a central iron bar, turned down at each end. The main socket however, is in the shape of a double dove-tail; this was filled
'4

with molten lead to protect both the iron and the stone from corrosion. There are direct parallels at Sardis. Information kindly supplied by David Stronach. 79 Mlanges Dussaud II (1939); E. F. Weidner, "Jojachin, K6nig von Juda ", in Babylonischen Keilschrifttexten f., and 934, for 923 the four Lydian names, one of which however may have belonged to the royal house of the Mermnadai. new issues of Croesus superseded the older 80 Hdt. I, 94-the Lydian electrum currency. 81 Barclay V. Head, Historia Numorum(1911), p. 646. 82 Iran III (1965), pl. I and fig. 2, 3, opp. p. 13.

16

JOURNAL

OF PERSIAN

STUDIES

tower and its replica the Ka'bah at Naqsh-i-Rustam remains a problem. It has often been suggested that both buildings were fire temples related to the coins which depict models of fire altars, and Ker Porter was convinced from the many traces of smoke in the tower of Naqsh-i-Rustam that this was its purpose. But nothing of the kind has been observed in the tower at Pasargadae and David Stronach has informed me that in his opinion the gently sloping pyramidal roof there makes it unlikely that anything can have been placed on its summit, nor is it likely that any fire burnt therein, for there is no trace of any ventilation whereas all modern Zoroastrian fire temples are ventilated. In seeking an interpretation for their use we should perhaps be guided by the long Sasanian inscription at Naqsh-i Rustam. The writing on the outside wall of that tower83 names Kartir, the celebrated High Priest and developer of Sasanian Zoroastrianism, the Magian chief, and refers to the sacred imperial fires which were kept alight (not necessarily within the building) in memory of the deceased members of the royal family-as well as for the living. Here, it seems, memorial services were held and masses said for the souls of the blessed departed. There were also endowments for the foundation of a fire-temple for the sacred fires in various parts of the realm. Daily offerings of lambs and kids, bread and wine were also instituted. In this connection it is tempting to suggest that a great rectangular enclosure " The Sacred Precinct "84 not far from the Takht at Pasargadae and its associated fire-altars may have been a reserve dedicated to fire-rituals and fire-worship, but we must await Stronach's verdict on the subject: the association of Precinct and altars with Cyrus cannot be excluded. It is moreover interesting that the style of tower in the Zendan,85 with its dentil cornices and blind windows, finds parallels in the architecture of Urartu (later Armenia) and may perhaps be traced also at Altin Tepe in Cappadocia-again a fruit of Cyrus' conquest. Common sense tells us also that so powerful a building, once endowed with massive doors must have been a repository for guarding the sacred paraphernalia and royal relics associated with the throne-no more practical building could have been devised for that purpose-in close touch with the royal hypostyle halls, wherein the king held his imperial ldv6es. Character Architecture Sculpture and of We have clearly seen that in the hypostyle halls P and S, near the royal portal R, the architecture reveals both the old oriental style initiated by Cyrus as well as evidence that Darius inscribed his own name and was content to record that of the founder Cyrus, member of a collateral and alien branch of the royal line. In Palace P above the Portals of the Main Hall, above the sculpture, there appears to have been a trilingual inscription which began with the name of Darius who was possibly claiming the credit for finishing the Palace.86 In the same building, the sculpture on the robes bears the same simple Cyrus titles which we observed in the Great Winged figure, but the sculpture also betrays the influence of Greek artistry, an indication that here we see the hand of Darius superimposed on that of Cyrus. Stronach has informed me that Palace P has yielded traces of the original plan designed by Cyrus, a plan that was later modified, no doubt by Darius. It was Darius who was the first to introduce logograms in the inscriptions. Cyrus' Tomb and Epitaph We come finally to the most memorable relic of all, the tomb of the Founder himself (P1. I).87 This wonderful memorial, built of a fine white limestone, in the shape of a gabled house standing on a six stepped podium needs no detailed description here. Together with the citadel it is the outstanding monument in the plain of Pasargadae and from the time of Alexander the Great, who ordered its repair and the restitution of Cyrus' desecrated bones, no traveller could remain unmoved by the sight of it. Both Strabo and Arrian have recorded the inscription which, as they allege, was on the tomb, but no
83

M. Sprengling, A.J.S.L. LVII (I940), p. 197; C.A.H. XII, I09-37; A.J.S.L. LVIII (I940), p. 169; E. Honigmann and A. Maricq, Recherches les Res GestaeDivi Saporis sur (i953) II, pp. 98-I Io. 84 Iran III (1965), p. 24 and for location on site plan fig. I.

85

86 Borger and

87

Op. cit., pl. I and fig. 3 opp. p. 13 for plan and elevation. Hinz, ZDMG CIX (I959), 117-25; CXV (1965), 396. References in Nylander, Orientalia SuecanaXVI pp. 149, 150. Iran II, figs. I, 2, pls. I-III.

CYRUS

THE

GREAT

17

modern historian believes that the form of wording given by them is anything but Greek. " O0man, I am Cyrus, who acquired the Empire for the Persians and was King of Asia; grudge me not thereafter this monument."88 That was reported by Aristobulus, and Strabo adds another attributed to Onesicritus-one in Greek, carved in Persian letters-" Here I lie, Cyrus King of Kings".89 It does not need much imagination to penetrate this travesty of the truth. Greek travellers to the site must have been shown the simple bilingual and trilingual inscriptions of Cyrus in cuneiform script to which they added their own comment, or the guide's, that Cyrus had indeed acquired the Empire and that none should sort of banal remark that any tourist might utter. It may well be begrudge him this monument-the that originally there was an inscription within the park and copse which according to Strabo was the setting for the tomb. Alexander on his visit was reported to have seen a slightly different version of the inscription in Persian letters. Both Greek historians tell us of the elaborate and expensive golden furniture and other trappings with which the tomb had originally been endowed. There is no need to dilate on this subject; but one recent discovery, due to the new and detailed scrutiny made by David Stronach, deserves mention. On the gable over the top of the doors of the tomb, Stronach was the first archaeologist to observe an elaborate raised disc with twenty four outer rays surrounding a double concentric rosette. Perhaps, as Stronach has suggested to me, this ornament may have been regarded as the "sun-disc", an early symbol of Ahuramazda the supreme god whom Cyrus may well have revered in this guise. It remains only to consider the origins of the form of this most unusual tomb, venerated by the Greeks and classical in appearance. Stronach believes that the " essential character of the monument accords with an indigenous Achaemenian tradition in which a primitive type of gabled house served as a prototype for all major free-standing tombs ".90 This surmise may well be correct, but I am tempted to suggest an even closer and doubtless heretical opinion. I look on the tomb chamber as an expensive copy in stone of the original wooden coffin in which the body of Cyrus was carried over a thousand miles from the battle-field to be laid to rest in his own home. I find support for this conjecture in the rather more rudimentary and primitive tomb of the same type known as Gur-i-Dokhtar91 in the southern Zagros, south-west of Kazerun, which looks to me even more wooden in character: perhaps the steps beneath both tombs reproduce in more elaborate form the raised biers on which the coffins were carried. At Pasargadae we may sojourn with Cyrus in word and in deed and be conscious of the master mind and true architect of the Persian empire. For all the glories of Persepolis, many perhaps will feel more moved by the wonderful strength of the lonely akropolis (P1. IV), standing as it does on its great and massive foundation-a base laid by Cyrus for three centuries of continuous Achaemenian power. The man himself, happily remembered in history as a merciful and humane conqueror, has remained an inspiration for ruler and ruled down to the present day.

Anabasis VI, 29, 8. XV, 3, 7. 90 Iran II (1964), P. 27.

88 Arrian,

89 Strabo
91

Iran II (1964), pl. III and fig. 3; discussion on pp. 28-30. This tomb lies in a bleak valley of the southern Zagros mountains Ioo km. south-west of Kazerun. David Stronach, op. cit., adduces reasons for assigning this tomb to the earlier portion of the sixth century B.C., but a more recent examination by him has detected the use of a late type of clamp probably not earlier

than the fifth century B.C. We may assume perhaps that this was a late, provincial version of an older type. Various opinions about the character and date of this monument have also been expressed by L. Vanden Berghe and C. Nylander; references and discussion in RLA, ed. E. Ebeling et al., Dritter Band, Achte Lieferung (Berlin, I97I), article entitled "Grab", p. 5e9, by E. Strommenger. Nylander, lonians in Pasargadae, p. I45, has however good reason for asserting that "the mouldings and the entablature of the Cyrus tomb are Ionic ".

SIR ROBERT KER PORTER-REGENCY

ARTIST AND TRAVELLER

By R. D. Barnett
One of the most neglected but most remarkable pioneers in the field of Near Eastern archaeology was Sir Robert Ker Porter, painter and traveller (P1. I). He claimed, possibly with truth, that he was descended from an old family, one of whom fought at the Battle of Agincourt, but this is unimportant. He was born in 1777,1the third son of a family of five children of William Porter of Durham. His father, after twenty-three years as an impecunious army surgeon in the 6th Inniskilling Dragoons, died when he was two. His mother who brought him up in Edinburgh certainly contrived to get him a sound education, involving a good knowledge of Latin and Greek, in spite of poverty. He was brought to London in 1790 and admitted as a student at Somerset House to study painting under Benjamin West in the Royal Academy where he made rapid progress. A book of sketches2 dated I797, and now in the Department of Manuscripts of the British Museum, made mostly in Hampshire and the Isle of Wight, shows conventional competence. It contains some pleasant views of Carisbrooke and German soldiers and French prisoners. But in I8oo he astonished the public by his "Storming of Seringapatam", a panorama 128 feet in length. This sort of picture was carried about on rollers, a style of work then popular. It is stated on his mother's authority to have been painted in six weeks and contained hundreds of figures, but in any case must have been a quite sensational exhibit. It was, alas, destroyed in a fire, but sketches and a print illustrating it survive in the John Rylands Library, Manchester.3 To meet the anxieties of the Napoleonic War, he became a captain in the Westminster militia in I803-4. While his military interests always continued, his artistic career was paramount and in I804 he was appointed "historical painter" to the Tsar of Russia-a kind of post for which it is impossible to think of a modern equivalent, but which belonged to the quasi-scientific, quasi-romantic spirit of the age. In I8o6 he started on the task of painting the Admiralty Hall at St. Petersburg. While there, he fell deeply in love with a Russian princess, Princess Mary von Scherbatoff, but before the necessary royal consent could be obtained to his marriage, he was forced, as an Englishman, to leave Russia by that country's political alliance with Napoleon under the Treaty of Tilsit. He withdrew to Sweden, again got on well at court and was knighted by Gustav IV of Sweden in I8o6. After travelling to Germany, he came home and had his portrait painted.4 He then went to Spain with the British army. He published in I809 two sumptuous quarto volumes of his sketches in Russia and Sweden,5 and a volume of letters from Portugal and Spain, written during the march of the troops under Sir John Moore.6 The story of their retreat and Dunkirk-like evacuation in the dark days of the Peninsula War is famous, and the death of Sir John Moore is described in the well-known poem by Charles Wolfe. One of his sketches covering this subject is in the British Museum Department of Prints and Drawings.7 In 1811 he revisited Russia, married his Russian princess, then attended and described the campaign against Napoleon of 1812.8 Returning to England, he was knighted by the Prince Regent in I813.
' See article on him, Dictionary of National Biography XLVI, pp. 190-2 (though that is subject to some corrections). A full length biography by Mr. M. D. Ancketill is in preparation. 2 Add.MS.I8283 (drawings in Hampshire, Farnham, Isle of Wight, x797). In I799 he also drew three anti-Napoleonic "atrocity" scenes for publication as etchings in AugustSeptember 1803, one showing Buonaparte allegedly ordering 580 wounded soldiers at Jaffa to be poisoned (1799). In i8oI he made nine sketches of the Battle of Alexandria and a huge painting of the expulsion of the French from Egypt and of the Siege of Acre (Rodney Searight, The Middle East: A Catalogue and of an Exhibitionof watercolours drawingsby British andforeign nos. travellers, 75o-900oo, Leighton House 27.1.71-20.2.7I, Io8-9). Of these paintings he published descriptive accounts. pictureof the Stormingof Seringapatam, painted 3 The grandhistorical by R. K. Porter, a Descriptive Sketch. (London, n.d. Formerly B.M.I865 c.9.4, but destroyed in the last war). The print is illustrated in Dupouy, see note x5 below. 4 Stipple engraving by Antony Cardon, painting by J. Wright. in 5 TravellingSketches Russia and Swedenduringtheyears18o5, 18o6, 1807, z8o8 (London, 18o9; 2nd ed. 1813). 6 Lettersfrom Portugal and Spain written during the march of the troopsunderSir John Moore (18O9). 7 Sketchbook, Prints and Drawings Department, I98a.I7, folio 42. 8 Sketches: A Narrative of the Campaignin Russia duringtheyear 1.82 (London, I813?; 4th ed. I8I5).

19

20

JOURNAL

OF PERSIAN

STUDIES

In St. Petersburg his restless energy, and quickness of eye and hand were noticed by A. Olinen, then President of the Russian Academy of Fine Arts and a cousin of his wife. To understand what then happened we must cast our minds back to the state of almost total ignorance about the Ancient Near East which prevailed in Europe, though the classical world had already received attention. The veil had, it is true, been slightly lifted over Egypt by the expedition of Napoleon to the Nile Valley, and the extraordinary group of scholars and scientists who accompanied him. But no single ancient monument of Western Asia was really known, except the magnificent ruins of Persepolis, the ancient Persian capital of the Achaemenid kings near Shiraz, which had been "discovered " in the fourteenth century and frequently, though inadequately, reported on by travellers in the intervening period.9 The cuneiform inscriptions to be seen there had been copied, and indeed had already given the name of " Persepolitan" script to the cuneiform texts, which a German, G. F. Grotefend, was already trying to decipher by the beginning of the nineteenth century. But as yet nothing was known of the smaller antiquities of the Near East except from the work of a single pioneer, Raspe,xo a gifted but eccentric German chemist who, while employed in the coin-cabinet of the Landgraf of Hesse fell under a cloud, fled from Germany to England and made in I778-8o the first catalogue of Mesopotamian cylinder seals ever published, namely those in the Tassie Collection."1 This included seventeen already stated to be in the British Museum, of which we have no other contemporary record or register, and needless to say, they were described by Raspe in the manner of the day, inevitably, as Persepolitan. Raspe is, however, better known for a different achievement, namely as the anonymous author of Baron Munchausen's Travels (I784), a still popular fantasy, which has gone through numerous editions and translations. Ker Porter was encouraged to study and reproduce the Persian sculptures at Persepolis and Murghab (now Pasargadae) by Olinen, who pointed out how inadequate were the previous illustrations of these monuments, though Persepolis had been visited and drawn since I674 (P1. IXa). Olinen added the following advice: " In conclusion I repeat, draw only what you see! Correct nothing; and preserve, in your copies, the true characters of the originals. Do not give to Persian figures a French tournure, like Chardin12 nor a Dutch, like Van Bruyn13, nor a German, or rather Danish, like Niebuhr14, nor an English grace, like some of your countrymen, in your portraits of the fragments at Nakshi-Roustam. Adieu. Your faithful Friend and Cousin, A. Olinen."'5 In his published book Sir Robert comments on this somewhat turgidly as follows, speaking of himself in the third person: " What refinement of taste in some of Sir Robert Ker Porter's predecessors may have caused them to change, almost unconsciously, a scrupulous curiosity in observing the progress of the art, determined him to copy line by line, defect or beauty; whatever he saw, to portray; transcribing shape of person, character of feature, and fashion of apparel, to the minutest particle. An ardent lover and sedulous practiser of the arts, from childhood to the present time, his eye and his hand being alike familiar to every detail of the pencil or the chisel, precision in these respects is as natural to him, as embellishment to those, who may be amateurs of the arts, without having studied the principle of its schools. Therefore, to such an undeviating fidelity of portaiture, the writer of this work avows his claim. A similar experience in military objects, assisted by his observation on the arms of the people, ancient and modern, of the nations through which he travelled, has greatly facilitated his measurements and plans of the sites and elevations of certain cities and places of renown and interest."'" How well he succeeded in his task has been judged by most of the world from the printed results. We have, however, to remember that before photography was invented, there was an inevitable
9 See G. N. Curzon, Persia and thePersian Question, vols. (London 2 and New York, 1892), Vol. I, pp. 16-7, for the list of travellers, and Vol. II, pp. I48-9. 10 See Dictionary of National Biography XLVII, pp. 301-3, art.
11 R. E. Raspe, A DescriptiveCatalogueof Engravedgems, ... cast ... byJ. Tassie (Edinburgh, 1791). 12J. Chardin, Voyagesen Perse et autres lieux de l'Orient. Cf. his

Raspe (1737-94).

1686; Amsterdam, 1711). E. de Bruin, Voyages la Moscovieen Perse (Paris, 1718). par 14 Karsten Niebuhr, Reisebeschreibung nach Arabien und andern Ldndern(Amsterdam, 1774-8). umliegenden 15 Quoted in Sir R. Ker Porter, Travels in Georgia,Asia, Armenia, Ancient Babylonia, during the years 1817, 1818, z819 and z820 (London, 1821-2), preface,pp. 6-8. 16 Ibid.
13

Journal du voyage du chevr. Chardinen Perse, 3 vols. (London,

SIR

ROBERT

KER

PORTER-REGENCY

ARTIST

AND

TRAVELLER

21

tendency for the works of artists, when recopied by an engraver, to lose something of their force, or too often to be completely falsified, when seen through the engraver's eyes. In the case of Sir Robert's work, we are fortunate to be able to compare his published drawings, as engraved by others in his book, with his originals. The Department of Manuscripts in the British Museum has possessed since his death three volumes of his drawings, two of which"7 represent the raw material of his published work, though in a very advanced state. His work, as finally published in 1822, contained 87 engraved line blocks, maps, and aquatints. The manuscript volumes contain over 200 pictures, but to them is added the contents of a sketchbook illustrating racial types and costumes, with which we are here not primarily concerned. There are also Ioo more unpublished archaeological drawings in Leningrad.xs From these manuscript volumes it is possible to recognize Ker Porter as a landscape artist of power and splendid vision. Though certainly apt to give free range to a romantic imagination, nevertheless, when faced with an antiquity, he appears as an archaeological draftsman of high quality, if we consider the times. It would appear that, as his drawings and water colour sketches were too expensive to reproduce in their original form and size, their number was drastically reduced from 200 to 87 in the published book as it finally appeared in 1822; in the case of the archaeological subjects, these were restricted to pen and ink outline sketches, and the water colours to monochrome aquatints. The aquatints give only a poor idea of the magnificent sweep and dramatic, indeed romantic, vigour of his art. His method of work seems to have been in any case to start with a pencil outline drawing, then, in the case of the archaeological drawings and plans, to work them up with pen and ink or sometimes re-copy them in a sepia wash. But for publication of the archaeological subjects, mainly the pen and ink drawings were used. I select only a sample of the originals at random. On Plate II is " The Cascade in the Valley of Tombs at Tiflis " which does not appear in his published book. This is, however, how he describes this view of the river Terek, in the printed text: "Besides the peculiar pleasure, to a military taste, in viewing the remains and situation of the citadel and other works, the valley behind the public baths, which leads to the most considerable of the ruins, possesses picturesque and interesting objects in itself. In following the windings of this wide mountain cleft, for some distance, we were imperceptibly led into a deep chasm, whose dark granite sides were broken into abrupt shelves, over which rush the waters of a lofty cascade, tumbling, with great noise, into a bed of rocks beneath. Thence it flows, murmuring along, by the base of the fortress, till it unites with the broader stream of the Kur. The immediately-surrounding objects mingle more beauty with the sublime, than the first approach to Tiflis had given us to expect in any part of its adjacent scenery. Many of the cliffs are richly covered with trees and shrubs, and carry the delighted eye through rocky and umbrageous intricacies, to the shining promontory, over which shoot the waters of the fall. Still we look upward, and see the mountain of the citadel, crowned with its mouldering towers."19 After visiting Etchmiadzin and Ani, the two famous mediaeval cathedral churches of Armenia, he crossed into Persia at Erivan, now capital of the Armenian S.S.R., but which then stood on the frontier of Persia with Georgia. His manuscript also contains a long extract from his diary in which he describes a visit on October 26th 1819 to the ruins of the Palace of the Armenian King, Tiridates, of the Roman period. It includes some drawings of its remains which he made, including that of a part of the basalt ornament, a fragment of which (top left in the plate) is now in the British Museum (P1. IIIa).20 I pass over several beautiful drawings from his album and come to one of Tehran (P1. IIIb) which reached in May 1818. His view of it is published as the plate opposite page 312 and shows it viewed he from the road to Isfahan as a walled city with an external fort. The town walls then had a circuit of yards with 4 gates, each with a fort in front of them. In the foreground of his original drawing (P1. 8000ooo IIIb) is an elephant (omitted in the publication), military elephants being used in the Persian army till about I850. In the background is Mount Demavend. Several more beautiful drawings must be passed by until we come to Pasargadae, the city of Cyrus the Great, then called " Murghaub" after a nearby stream. Ker Porter describes the ruins of Pasargadae in detail and included in his book an engraving of what is now known as the Tomb of Cyrus the Great, a small gabled building on a flight of steps. It
17Add.MS 14758.I&2. These are supplemented by MS 18281 (drawings made in Persia). 18 I owe my knowledge of these to the kindness of Academician
3A 19
20

B. B. Piotrovskii, Director of the Hermitage Museum. Travels, opposite f. 16. Add. MS. I4758, opposite f. 31.

22

JOURNAL

OF PERSIAN

STUDIES

had been already discovered a few years before, in I8O9, by James Morier, the immortal author of Hajji Baba of Isfahan,then a secretary in the British Embassy, but Morier had not known what the building really was though he published a sketch of it in his book of travels.21 Ker Porter was able correctly to identify it from ancient descriptions as the Tomb of Cyrus. He then made an important contribution by recording the remains (many now lost) of the Palace of Cyrus nearby (P1.V), of which the surviving doorway slab bears the carving of a remarkablefour-winged deity. A drawing of it is also engraved in his book (P1. 13) but his water-colour sketch (P1. IV) is far more impressive. It is interesting to notice, sketched below the drawing, the Egyptian headdress,which he aptly compares. The sculpture was discovered by him on June 13th 181 Above it was a then visible inscription in Persian cuneiform I8. script of which he made a correct copy and sent it to the German scholar Grotefend, who was attempting to decipher the cuneiform. Grotefend read it correctly from Ker Porter's copy as bearing the name of Cyrus the Great, King, and reported his discovery to Ker Porter, who quotes it in his book.22 He proceeded to Naqsh-i Rustam, a few miles from Persepolis and reached it on June 17th. Here the tombs of the Achaemenid kings of Persia, whose palaces are found at Persepolis, are carved on the vertical surface of a steep cliff. He calls Naqsh-i Rustam " the Mountain of Sepulchres ". There were dangers to be met on the road: at Mianna there were myriads of poisonous bugs, but further the road was not considered safe, as a Mr. Browne had been murdered at Iraq-el-Ajem six years before. Much space is devoted to these tombs in his publication. These were not new discoveries, for they were known from 1545. As he himself mentions:23 " Naksh-i Rustam or the Mountain of Sepulchres (P1. 16) was the next object of my investigation. The village in which I had taken up my quarters, is distant from the mountain nearly three miles; and my daily visits were performed at sun-rise, where I remained, fully occupied, until the heat of the day drove me away, by becoming insupportable amongst those unshaded rocks. The face of the mountain is almost a perpendicular cliff, continuing to an elevation of scarcely less than three hundred yards; the substance is a whitish kind of marble. In this have been cut the celebrated sculptures and excavations, so long the subjects of discussionwith the traveller, the artist, and the antiquary. These singular relics of Persiangreatness are placed very near each other, and are all contained within the space of not quite the height of the mountain." The line drawing in his book illustrates one of the four tombs, but without the entertaining detail which we see on the water-colour sketch, showing the artist suspended on a rope in mid-air (P1. VIa). The tomb is carved in the form of a fa<ade of a house or palace, bearing above it a vast bed or pedestal supported by the figures of the subsidiary satrapies of the Persian Empire. On the pedestal stands the Persian king worshipping the god Ahura-Mazda before a fire altar. Below it are carved eight Sasanian triumphal reliefs of some 8oo years later. Ker Porter drew seven of these. One shows the investiture of King Narsah by the goddess Anahita, and belongs to the late third or early fourth century A.D. It has now been excavated to the ground level, but at that time was half covered up (P1. I9). Another famous relief (relief 6) which he drew (P1. 21) was that showing the triumph of Shahpur I over the Roman Emperor Valerian, when the Roman army was shattered and the proud Emperor himself was captured in 260 A.D. Beside the Roman Emperor kneels Philip the Arab, while Valerian stands. (According to Ker Porter himself, however, the kneeling figure is Valerian and the standing figure is Cyriades, whom Shahpur wished to raise to the Imperial throne.) He also drew (P1.24) the relief at the side of the Achaemenid tombs, carved over an Elamite relief many hundreds of years earlier, and two of the Sasanian reliefs at Naqsh-i Rajab on the opposite side of the valley, showing Kartir and Ardashir (P1. 27).
On June 23rd Ker Porter reached Persepolis, the great ritual centre and summer residence of the Achaemenid Persian kings. This magnificent monument was built by Darius I, the invader of Greece, about 500 B.c., and completed by his son Xerxes, but contains other smaller additions by later monarchs. It is the expression of the might of the Persian Achaemenid Empire at the height of its power, when it spread over the whole of the Near Eastern world, to be thrown back only by the heroism of the Greeks in 490 and 480 B.c. Persepolis, the prime objective of Ker Porter's expedition was, as we have seen, no new discovery, unlike most of the ancient sites of the Near East familiar today. It had survived above
Persia (London, 1812), pp. 144-6 21James Morier, Journeythrough and plate opposite, and pl. 29.
22
23

TravelsI, p. 505Ibid., pp. 515-6.

SIR

ROBERT

KER

PORTER-REGENCY

ARTIST

AND

TRAVELLER

23

ground and been known from the reports of travellers since the fourteenth century when it was first visited by an Italian monk, Friar Odoric of Pordenone (c. 1325 A.D.). Twenty-four visitors to it successively reported on its amazing ruins in publications stretching from the date of Odoric, through the fifteenth, sixteenth and seventeenth centuries. The attention of scholars was already closely engaged upon it, as at that date the monuments of Assyria and Babylonia had not yet been discovered, it was the most important, nay the only known monument that represented the glory of the lost ancient world of Western Asia. But Ker Porter's drawings speak for themselves. His publication has been criticized, but it must be remembered that his critics did not have the benefit of seeing his original manuscript volumes, only the engravings. Booth writes:24 " It was with the professed object of giving a final and authoritative representation that would satisfy the curiosity of the minute student that Sir Robert Ker Porter undertook to go over the old ground once more. He was an accomplished artist and he consequently possessed qualifications many of his predecessors were without. He arrived at Murgab on June 12th, 1818, and left Persepolis on July Ist, so that he was not more than eighteen days engaged in the study of the numerous antiquities in the neighbourhood. At the conclusion of his stay, he congratulates himself upon finding that: " I had drawn nearly every bas-relief of consequence, had taken a faithful plan of the place and copied several of the cuneiform inscriptions"! His industry He surveyed the sites of Murgab and during the time must certainly have been extraordinary. former, was now made for the first time, Persepolis, and made two ground-plans of both places. The.... but the latter had been taken as early as the days of Kaempfer.25 He took two drawings of Murgab, four of the Achaemenian remains at Naqsh-i Rustam, six of the Sassanian sculptures, and two of the same period at Naksh-i Rajab. In addition to this he made twenty-four drawings of the monuments of Persepolis-some of them upon a large scale-and copied inscriptions that occupy four plates; that is to say, he accomplished in eighteen days work that now fills 42 plates of engravings. This is certainly wonderful, but if he had executed less than one-quarter, the result would perhaps have been more satisfactory. In addition to the drawings, he measured the various buildings and the more important objects in each; and took notes for his very elaborate description of the dress, the arms, and other minute details of the various figures in the numerous bas-reliefs. His visit to Naqsh-i Rustam resulted in the earliest drawings we possess upon an adequate scale of the fa<ade of a tomb, and it was completed by a ground plan of the interior. He expressed his conviction that the third tomb with the long cuneiform inscription was in all probability that of Darius, an opinion afterwards proved to be correct." I show some illustrations of his drawings at Persepolis (Pls. VIb, VII, VIII and IXb), some unpublished, some published (Pls. 31, 34-36 and 49), but often in inferior form. Since his visit, many other scholars and travellers reached and recorded its monuments. In fact, in 1891, a British expedition was actually sent out with expertformatori to Persepolis to make a series of plaster casts of the principal sculptures, and brought the moulds home to the British Museum.26 Unfortunately, many of these were somewhat damaged during the last war. The climax of work at Persepolis was the excavations conducted by the University of Chicago in 1931-4, which have now been published in two large volumes, one illustrating the ruins of the palaces, the other the small objects. An important third volume dealing with Naqsh-i Rustam has now appeared.27 On leaving Persepolis, Ker Porter made his way via Hamadan and Sanna to the now famous mountain site of Bisitun (P1. X), 30 kms. east of Kirmanshah, which he calls Be-sitoon, and reached it on September 19th (his P1. 59). The famous sculpture, 500 feet above ground level, showing Darius the Great (522-486) and the twelve enemies over whom he triumphed, had been first noticed by a French traveller named Otter in 1734,28 but Porter was the first person to climb sufficiently close-at great personal risk-to sketch the figures on the mountain face. He could not, alas, go further and record the all-important cuneiform inscription of Darius in three languages, Old Persian, Babylonian, and Elamite, which was to provide the key to the cuneiform writing for subsequent generations. But he
A. J. Booth, The Discovery and Decipherment the Trilingual 28 Photographs Casts of Persian Sculptures of of mostly from Persepolis, 12 Inscriptions(New York and Bombay, I902). Cuneiform pls. (London, I932). fasc. 5 (Lemgo, 27 E. Schmidt, PersepolisIII, The Royal Tombsand otherMonuments 25 E. Kaempfer, AmoenitatumExoticarum ... & 7I12), also his work, Overde Printrerbeeldingen Aanmerkingen (O.I.P. LXX, Chicago, 1969). 28 Jean Otter, Voyages Turquieet en Perse (Paris, 1748), 2. vols. en ... van... Persepolis(1714).
24

24

JOURNAL

OF PERSIAN

STUDIES

drew attention to the presence of this text and to the presence of texts above each figure giving their names. This task of copying was reserved for Colonel, later Sir, Henry Rawlinson, the true decipherer of the cuneiform, who in 1834-47 copied the texts amid great dangers and difficulties; he owed much to Ker Porter's example. Porter published in his book a neat line drawing of the sculpture of Darius triumphing over his enemies. It is a little unfortunate that at this point his usual sound intuition deserted him and he conjectured that this relief represented the triumph of the Assyrian King Sargon over the Israelites of Samaria in 721 B.C.29 Of exceptional interest are the original drawings which he executed at the important Sasanian site that he discovered at Taq-i Bustan, 5 kms N.E. of Kirmanshah, two days after leaving Bisitun. Line drawings of these in the form of engravings were included in his published book (Pls. 62-64),30 but those from his unpublished manuscript book are more life-like. The first is a beautiful depiction of the investiture of a Sasanian king, Peroz (457-83 A.D.) or Khusrau (590-628). Below is seen Khusrau on horseback in full armour, like a mediaeval knight. The second is a superb pair of sketches of the rock relief, showing royal boar and deer hunts from the two sides of the same grotto. Porter continued his journey onwards through Hamadan, recording but not publishing two Persian texts (P1. XI), to Baghdad where he stayed with Claudius James Rich, another still more important figure in the history of Mesopotamian archaeology. Rich, a very young and extraordinarily brilliant linguist, was the British East India Company's Resident at Baghdad, where he rapidly commanded considerable influence. Assisted by his wife Mary, Rich led an active and hospitable life in Baghdad, using his leisure to study and understand his surroundings. This is not the place to describe Rich's remarkable but all too brief career,31 except to mention that he collected a library of about 800 Oriental manuscripts, when such things were hardly known and less appreciated among the western world. Still more remarkably, he took note of the quite unknown antiquities of Mesopotamia; he collected and copied, with the help of his gifted German secretary Karl Bellino, cuneiform texts, some of much importance (though no one could as yet read them), formed a valuable collection of cylinder seals and other antiquities, and took great pains to discover and identify the true site of Babylon, then unknown, and to publish his discoveries.32 In 1822 he was to die of cholera when on a visit to Shiraz to succour the sick, and his collection of manuscripts and antiquities was purchased by the British Museum from his widow. Naturally, in 1818 Rich was delighted to receive as a guest a fellow scholar such as Ker Porter, who described and drew not only some of Rich's coins and cylinders (a valuable record of the collection) but also the tamed onager presented to him by the Arab natives. It was one of the last of a species once abundant in the Syrian desert (seen on the Assyrian reliefs33) but today long extinct. From Baghdad, Ker Porter was able to visit Hilla (Babylon) and Birs Nimrud, where he made some more excellent drawings, then returned home via Scutari and Istanbul and made arrangements for the publication of his book.34 He had now achieved his appointed task, and was decorated by the Shah (whose portrait he drew -P1. XII) with the order of the Lion and Sun. But from now on his career lay elsewhere. In I824 he chose to be sent as consul to Venezuela where his career was linked with Bolivar in the formation of Bolivia.35 Two years later his wife died of typhoid in St. Petersburg. In 1832 he was invested as a Knight Commander of the Order of Hanover. He died in I842 in retirement and was buried with his wife in St. Petersburg. A final painting of him, in Russian uniform, painted by G. Harlowe, and engraved by W. O. Burgess, was published in I843 (P1. I). With this farewell, we salute this great scholar and artist, expressing the hope that some benefactor may make it possible one day to publish his original water-colour sketches in their entirety.

29 TravelsII, pp. I6o-i 30 Ibid., Plates

and Plate 60. 62-4. 31 See C. M. Alexander's biography of C. J. Rich, Baghdad in in BygoneDays (London, 1928) and Seton Lloyd, Foundations the Dust (Oxford, 1947). 32 Rich, Memoiron the Ruins of Babylon (1818).

33 On the sculptures of Ashurbanipal, BM.124873-I24882, Gadd, The Stonesof Assyria, pp. I84-5.

C. J.

34 See note 15. 35 Walter Dupouy, Sir RobertKer Porter'sCaracasDiary (Caracas,

I969).

Harlowe(i8o8) : engraved and Ker in in Pl. L Sir Robert Porter Russianuniform.PaintedbyGeorge by JW. Burgess published 1843. O.

Pl. II. "The cascade the Valleyof Tombsnear Tiflis." in

Pl. IIIa. "Fragments at Tackt-i-Tigridata" (modernGarni).

Pl. IIIb. "View of the City of Teheranfrom the Isfahan Road."

Pl. IV. "Bas-reliefat Moorg-aub"(modern Pasargadae).

P1. V. "Remains of the Valley of Moorg-aub" (modernPasargadae).

the tombsat Nakshi-Roustam." Pl. VIa. "Oneqof ancient

Pl. VIb. "Sculptures Perse at

at P1. VII. "Sculptures Persepolis."

at P1. VIII. "Sculptures Persepolis."

Travellers." of Pl. IXa. "Comparison the5th Bas-reliefsentto Sir R. K. Porter,by Mr. Olinen,drawnby different

Pl. IXb. "Scu

P1. X. "The mountain rock at Be-sitoon" (modernBisitun).

P1. XI. "Persian inscriptions at Alvand" (i.e. Elwand near Hamadan).

Pl. XII. "FathAli Shah."

COVEREDBOWL, SAUCERAND SPOON. Gold,decorated painted translucent with and enamels: thePlanets,MajorConstellations and Signsof the Zodiac,andverses dedication Fath 'Ali Shah. SignedGhuldm-Khanaztd to of Bdqir. Early z9th century.

A ROYAL QAJAR ENAMEL By B. W. Robinson

With a few conspicuous exceptions the best miniature painting of the early QOajdr period is to be found in lacquer-work and enamel; but while good examples of painted lacquer are by no means uncommon, finely painted enamels are of considerable rarity. Not unnaturally the largest range of high quality pieces is to be seen in the Crown Jewels Museum in the basement of the Bank Markazi in Tehran. So when an enamel of royal quality turns up unexpectedly in the West it is worth remarking, at least by those who do not dismiss post-Safavid painting en bloc as beneath the attention of a serious art-historian.' The subject of this article is a bowl or cup, provided with a cover, saucer and spoon (facing p. 25); the bowl is 5 5 cm. high, the diameter of the cover is cm., that of the saucer 12-5 cm. and the 8.7 cm. long. All are of gold and all the exterior surfaces are covered with painted or transspoon is 13"4 lucent enamel, the latter of green, orange, blue and crimson. With the exception of the spoon, which is adorned with two heads of young princes on the handle and two human-faced suns on the bowl (P1. IVc and d), the design consists of a number of compartments of circular or arabesque outline, of different sizes, divided by scrollwork and flowers, each containing a figure or symbol of astronomical significance. In the central field of the saucer appears the signature Ghuldm-khdnazddBdqir. The under-side of the saucer is covered with green translucent enamel over an engraved design of the sun in splendour. auction rooms on June I3th 1956 as lot 97, " the property of a Lady ", when they were bought by a Hatton Garden jeweller. They came from the collection of the vendor's father, but nothing is now though they were rumoured to have spent some time in a private collection in Italy; but in 1966 they

The bowl and its companion pieces made their first recorded public appearance at Messrs. Christies'

known of their previous history. For the next ten years after the sale their vicissitudes are unknown,
the North, whence they passed into an English private collection in which they are likely to remain. It is sad to have to record that between 1956 and 1966 the pieces evidently suffered somewhat from rough handling, as is clear from a comparison of their present state with the photograph taken in 1956

appeared at the Antique Dealers' Fair at Grosvenor House for sale on commission from an owner in

for Messrs. Christies' catalogue. In particular the inscription round the saucer has been damaged and
clumsily repaired here and there, and one or two small pieces of enamel have been chipped off parts of the design. But mercifully all the main features remain intact, and this set remains an impressive masterpiece of the art of the Persian enameller.2 There are three poetical inscriptions, (i) round the border of the saucer, (ii) round the inner rim of the bowl, and (iii) round the base of the domed boss of the cover; they are as follows:3

For Persian enamels in general see B. W. Robinson, " Qatjr Painted Enamels " in Paintingfrom IslamicLands (Oxford 1969), pp. 187-204; V. B. Meen and A. D. Tushingham, The Crown Jewels of Iran (Toronto 1968), especially pp. 42, 43. See also Hotel DrouBt, Salle No. IO, Paris, Sale Catalogue 25/26. v. 1964, lots 1-54 (the Jambon Collection, probably the best private collection of Persian enamels ever formed). 2 It has been illustrated (i) in Messrs. Christies' sale catalogue
1

for June I3th 1956, (ii) in OrientalArt (Spring 1964), p. 36 and (iii) in Paintingfrom Islamic Lands (see above, note I), p. 196, all from the same photograph. SI should like here to record my grateful appreciation and gratitude to Mr. Souren Melikian of Paris for help received in the reading and interpretation of these verses; and to Mr. Leonard Harrow for much valuable assistance in interpreting some of the astronomical designs.

25

26

JOURNAL

OF

PERSIAN

STUDIES

(i)

By order of the King of Kings of the World, Fath 'Ali Shah, The Sultan, the son of the Sultan, Fath 'Ali Shah, This gold-bodied figure, ravishing the soul, and the heart's desire, In the Assembly of the Sovereign of the Earth is as the Moon in the Sky; Nay, Heaven itself, happiness obtained from a favourable ascendant. In the Assembly, he is the cause of Creation and the Way of Heaven, The seven Planet Stars have attained their stability. (ii)

The golden Cup in the hand of Khaqan Consider to be superior to the Goblet of Jam(shid); In the silvery Tower with its five Supports Consider the Moon as it glows; Nay, the Nine Heavens and the Twelve Signs, Consider them to be in the grasp of the Sun. (iii)

Reckon the Mouse and the Ox and the Cheetah and the Hare, And, passing to another surface, the Crocodile came, and the Snake; Look upon the Horse and the Sheep, and reckon the Monkey, Nine in all: the Cock and the Dog and the Boar complete the whole. The first set of verses makes it clear that these pieces were made for Fath 'All Shah, and this is confirmed by the second; Khaqan was the name under which Fath 'All Shah wrote poetry. The " five Supports " are the King's fingers and thumb as he holds the bowl by its base in the normal oriental manner; he is also, of course, " the Sun " in the last line. There is a pun on the word burj which means both " tower " (line 3) and " Zodiacal Sign " (line 5). The third set of verses seems to be a

A ROYAL

QAJAR

ENAMEL

27

mnemonic doggerel for remembering the twelve animals depicted on the cover in their correct order; we may compare our own old rhyme for remembering the Signs of the Zodiac: The Ram, the Bull, the Heavenly Twins, The Crab, and next the Lion shines, The Virgin and the Scales, The Scorpion, Archer, and He-goat, The Man that bears the Water-pot, And Fish with glittering scales. The remarkablefact is that the animals on the cover are the Twelve Signs of the Zodiac of the Chinese and " Dragon " for " Crocodile "-no very serious adjustment. Incidentally, " passing to another surface " in the second line of the poem refers to the fact that the first four animals are depicted on the domed boss of the cover, and the remaining eight on the flat surface surrounding it. The relationship in Persian astronomy or astrology (the dividing line between them has always been a nebulous one) of the Far-Eastern Zodiac to the normal Twelve Signs, which are also portrayed on this enamel, is not yet clear; but the decoration of this piece makes it obvious that such a relationship did exist, and that the choice of the animals depicted on the cover was by no means fortuitous. It may be of interest here to list the subjects in the various compartments, together with the Persian names attached to them, numbered according to the accompanying diagrams (Figs. I and 2):
and Japanese4 in their proper order, if we read " Rat " for " Mouse ", " Tiger " for " Cheetah ",

\3

1I2d

10

Fig. I.

(The Signs of the Zodiac)

I. The Ram, Aries. 3J (hamal) 2. The Bull, Taurus. j) (thaur) 3. The Twins, Gemini. (jauzd) ~.4. The Crab, Cancer. ZUt (saratan)

5. The Lion, Leo. aI (asad) 6. The Virgin, Virgo. 4-. (sunbula)

4

9. The Archer, Sagittarius. V.. (qaus) Io. The Goat, Capricornus. ,) (jady) ,s* i 1. The Water-carrier, Aquarius. h. (dalw) 12. The Fish, Pisces. (hat)
.

8. The Scorpion, Scorpio.

7. The Scales, Libra. 01

(mizdn)
('agrab)

Chinese, Ti Chih;Japanese, Jta-ni-shi.

28

JOURNAL

OF

PERSIAN

STUDIES

23
30 24

30
2.2.

04

C)

.20

5
2S?

31

32

37

34

35

Fig. 2.

A ROYAL

QXJAR

ENAMEL

29

(Six Major Constellations or Stars)5

I3.

I4.

15. 16. 17. 18.

the name of the star Pfin Cassiopeia). A young woman seated in a chair (P1. Ia). . Perseus: JJI1 hamil-i rds al-ghfil, " the bearer of the head ? (perhaps read JJl 9t .Lt,. of the demon "; J; might be an alternative to J;, or simply an error in spelling. It is ,,, worth noting that Algol, derived from J?lI " the eye of the demon ", is a bright star in the constellation of Perseus. l is the normal name for the whole constellation-a transliteration of Perseus). A helmeted youth carrying a demon's head by one horn (P1. Ib). Arcturus: (simdk-iramih,literally " the support of the spearman "; Arcturus is an 5,b star of the first magnitude in Bodtes). An armed youth in a leopard-skinjacket (P1.Ic). ?'-T. orange " Andromeda: 4L. (al-musalsila, the chained "). A young woman holding a golden chain U in both hands (P1. Id). ;x (shi'rd-yi Sirius, the Great Dog Star: yamdni). A young woman extending her left arm (P1.IIa). jiL. A bare-headed youth holding a serpent in both hands (P1.IIb). Serpens: .4I (al-hayya).

Cassiopeia:

(perhaps read

" the dyed palm ", :d(, kaf-i khazdib,literally

(The Sino-Japanese Zodiacal Animals: these are not labelled in Persian)

I9. The Mouse [Rat] (two are shown) (Pls. IId and IIIa) 20o. The Ox (head and shouldersonly) (Pls. IId and IIIb) 21. The Cheetah [Tiger] (Pls. IId and IIIc)
22.

The Hare (Pls. IId and IIId)

(Pl. IId)

26. The Sheep 27. The Monkey 28. The Cock

29. 3o.

25. The Horse (head only)

(Pl. IId)

23. The Crocodile [Dragon] 24. The Snake

The Dog The Boar

(The Seven Planets)

31. The Moon: ) (qamar) 32. Mercury: Ual ('u.tdrid) Venus: (zuhra) (P1. IVa) 33a..L IVb) (shams) 34. The Sun:

37. Saturn: ,j. (zuhal)

35. Mars: .6 (mirrikh) 36. Jupiter: (mushtari)

o.

(P1.

j-

(Central design of saucer) 38. This extraordinary design of a tricorporate man entwined with a dragon remains a complete mystery (P1. IIc). It obviously has some importance, in view of its central position, but it cannot so far be identified as having any astrological or astronomical significance. The artist's signature, ghuldm-khdnazdd Bdqir, is at the base of the design. The signatures ghuldm-khdnazdd B=qir and ghuldm-khdnazdd appear on the most beautiful 'Ali
enamels in the Crown Jewels Museum in Tehran," and the work of these two artists is very similar both in style and quality. Btqir has signed a magnificent gold tea-pot, enamelled with a portrait of Fath 'Ali Shah and bearing a dedication to him (Case IV, No. 21) and an oval snuff-box also with a
prominent in the northern hemisphere around the ecliptic, and I can see no better reason why they in particular should have been chosen for the series round the bowl." 6 Crown Jewels of Iran (see note 1), pp. 70, 71, Io2, 103. 7 Paintingfrom Islamic Lands (see note i), p. 197.

5 Mr. Harrow, in a letter to me, commented on this group in general: " I am sure you will have noticed the mixture of stars and constellations around the bowl. Possibly the Persian eye did not choose to discern greatly between the constellation and particular stars found in it. All the stars, however, are
4

30

JOURNAL

OF PERSIAN

STUDIES

portrait of the King (Case IV, No. 33).8 Unfortunately the name Baqir, or Muhammad Bdqir,9 was a popular one in eighteenth and nineteenth century Persia, and paintings of various types bearing this signature or attribution cover the period from the mid-eighteenth century down to quite recent times. The first seems to have begun his career under Nadir Shah. The American Minister Benjamin, who recorded traditional accounts of Persian painters current in the I88os, speaks of " Abah Ger " (sc. Aqa Baqir) as a celebrated flower painter under Nadir Shah;1o he must be the Muhammad Baqir who contributed flower paintings and other decorations to the celebrated Leningrad album."1 There is also in Leningrad a lacquered mirror-case signed Aqd Bdqir and dated 177/1765. I In the following year he executed a remarkable miniature of a sleeping nymph by a stream, now in the Chester Beatty Library (282.vi),12 copied from another dated Io84/1674 by 'Ali Quli Jabbadar, now in a private collection in Holland. This latter in its turn must surely have been copied or adapted from a European original. Our (Muhammad) Baqir comes next chronologically. Then in the Victoria and Albert Museum is a lacquer panel of good quality (L.1226-1921) showing a young prince hawking, the figure and other details built up in low relief with gesso.13 This is signed Muhammad Bdqir Td'is (" the Peacock ") but undated; stylistically it could hardly be earlier than the mid-nineteenth century. This must be a third artist of the same name. Finally there is Mirza Baqir naqqdsh-bdshiof Isfahan, who died in 1933. Apart from the enamel under consideration, and those already mentioned in the Crown Jewels Museum at Tehran, the most important surviving work that may be attributed to our Baqir is the back cover of the Nizlmi of Shah Tahmasp in the British Museum (Or.2265). This celebrated manuscript was magnificently rebound in covers of painted lacquer under Fath 'Ali Shah, the front cover being the work of Sayyid Mirza. Both covers show Fath 'Ali Shah hunting with princes and attendants. As noted elsewhere,'4 Muhammad Baqir's cover exhibits a rather hard style of drawing and a stippling technique of shading such as might be expected in the work of an artist whose chief metier was painted enamel. It is signed banda-i da'if (" the feeble slave ") Muhammad Bdqir, but neither cover is dated. The versatility of Fath 'Ali Shah's court painters is worth remarking: such men as Mihr 'Ali, Mirza tbbt,and our own Baqir could turn their hands to large-scale oil-painting, eglomis6 painting on glass, lacquer painting, painted enamel, book-illustration, and illumination as required, and with equal success. This set of gold enamels, then, is an outstanding example of a delicate art in which the Persians excelled. Persian enamels even called forth enthusiastic praise from the highly critical Comte de Rochechouart, one of our best authorities for the minor arts of the Qjar period.15 The colours are superb and the execution meticulous. Persian painting of this kind and especially in this period always seems to be striving to capture the ultimate in youthful beauty. So in this enamel both male and female figures display the boldly curved eyebrows and large dark eyes, the luxuriant tresses and tiny mouths celebrated by innumerable Persian poets. The whole scheme of decoration, with its elaborate deployment of astronomical figures covering the whole firmament, is itself a poem-a sustained panegyric to the Lord of the Conjunctions (Sdhib-i Qirdn), as Fath 'Ali Shth was invariably styled by the poet laureate Saba in his epic Shdhanshdh Ndma, the Book of the Kings of Kings, in which his real or imaginary exploits, and those of his forebears, are fulsomely celebrated. Even his vanity must have been flattered by so precious, so elegant, and so beautiful a tribute.
title Ghuldm-khananad (literally, "slave born in the house ") may be taken to mean " craftsman in the royal workshops ". The epithets ghuldm and khdnazdd are not infrequently found attached to the names of artists during the Qajdr period, but 'Ali and Bdqir seem to have been the only ones authorized to use both together. 9 Mulhammad was often omitted from signatures where it formed the first part of the artist's name, e.g. (Muhammad) Ismd'il and (Muhammad) KAzim, both eminent painters in lacquer, and the latter also in enamel, during the middle years of the nineteenth century. 10 S. G. W. Benjamin, PersiaandthePersians(London 1887), p. 326. 11 Leningrad, USSR Academy of Sciences, Institute of the Peoples
s The

of Asia, E. I4. Published as Albumof Indianand PersianMiniatures of the XVI-XVIIIth Centuries(Moscow 1962). Paintings signed by Muhammad Bdqir are reproduced on pls. 3 and 46 (signed Bdqir only); these are not dated, but one on pl. 4 bears the date 1172/1759 12 B. W. in Robinson, Persian MiniaturePaintingfrom Collections the British Isles (Victoria and Albert Museum 1967), pl. 35. I3 E. G. Browne, A LiteraryHistory of Persia, vol. I (London 19o8 and later editions), frontispiece. 14 B. W. Robinson, " A Pair of Royal Book-covers " in Oriental Art (Spring 1964), P. 3515 M. le Comte de Rochechouart, Souvenirsd'un Voyageen Perse (Paris 1867), p. 255.

Pl. Ia. Cassiopeia.

Pl. Ib. Perseus.

P1. Ic. Arcturus.

P1. Id. Andromeda.

P1. IIa. Sirius.

Pl.lIb.

Serpens.

Pl. I&c. Saucer: central groupandsignature.

from above. P1. lId. Cover

P1. IIIa. The Mice (Rat).

P1. IIIb. The Ox.

(Tiger). P1. IIIc. The Cheetah

Pl. IIId. The Hare.

Pl. IVa. Venus.

P1. IVb. The Sun.

Pl. IVc. Bowl of spoon(back).

Pl. IVd. Handleof spoon.

BRITISH AND FRENCH DIPLOMACY IN PERSIA, I800-I8Io* By R. M. Savory

In the year 1807, Napoleon was at the height of his power. The Third Coalition, formed in I805, was in ruins. After Austerlitz, Austria had sued for peace. Prussia, which had destroyed the Third Coalition " by her selfish neutrality ... proceeded in I8o6 to commit suicide by taking on Napoleon single-handed."' In two battles fought on the same day, October I4th 18o6, at Jena and Auerstidt, the Grand Army had smashed the Prussians, whose Junker generals still adhered to the rigid tactics laid down by their leading military theorist, Bulow, in his book Geist des Neueren Kriegssystem-tactics which showed little change from the time of Frederick the Great. The Russians were still in the fight, and had given a good account of themselves at the indecisive Battle of Eylau in February 1807, but there was a strong peace party in the Russian camp, and Tsar Alexander himself was angered and discouraged by the inaction of Britain. The Portland administration, which took office there in March 1807, discovered that "no more than twelve thousand men of the regular army were available for action on the Continent."2 The Continental System, instituted in I8o6 by Napoleon with a view to preventing Britain from trading with Europe, though not yet exerting its maximum pressure on the British economy, was nevertheless to lead to a serious drop in British exports in the second half of I807 and the first half of i8o8. Against this background, Napoleon's plan to invade India by marching through Persia was not as chimerical as it may appear at this remove of time, and Napoleon's own character made it a distinct possibility. " The urge to dominate, to dare, to play for the highest stakes was as instinctive to him and as irresistible as the urge of the mountaineer to climb Everest. This element of the irrational, the unlimited, the daemonic is as fundamental to the nature of Napoleon as it is to the character of Mozart's Don Giovanni,"3and Napoleon himself, reminiscing at St. Helena, said: " I consider myself the boldest of generals."4 As early as 1795, Napoleon had been urged to attack India by way of Egypt, but the mere idea of a French expedition to Egypt appeared so fantastical to the Admiralty that, even when the English agent at Leghorn reported that the French fleet was on its way to Alexandria, their Lordships remained sceptical. From Egypt, Napoleon had written to the Shah, asking permission to establish supply dumps on Persian territory, so that France could assist Tippoo Sahib, the sultan of Mysore, in his struggle against the English.5 A more definite plan of attacking India through Persia seems to have germinated from diplomatic exchanges between France and Persia in 1804-05. Since the dominance of the British fleet frustrated Napoleon's desire to invade England itself, the idea of attacking India, the source of British wealth, was most attractive to him. " It cannot be denied that, if Napoleon had succeeded in his design, he would have inflicted a severe blow on England. England would have been deprived, in fact, of more than Ioo million docile (sic) subjects, [deprived] of those vast, magnificent regions so profitable to its commerce, its industry, and its exchequer, and so valuable to its navy, and [deprived] of those countries from which it continually derived inexhaustible riches. Quelle diminution pour elle de puissance matirielle et de prestige dans le monde! "6 When Russia invaded Iran in 1803, the Q~tj r monarch, Fath 'All Shah, had turned first to Britain for help, and then to France. Napoleon immediately despatched to Persia Amid~e Jaubert, a secretaryinterpreter at the Court. While Jaubert was still at Constantinople, he was overtaken by a second
* In a much shortened form, this paper was read to the Oriental Club of Toronto, on March 4th 1969. x Felix Markham, Napoleon (London 1963), p. II. Ibid., p. I 15 3 Ibid., p. 41. * Ibid., p. 25.
2

6Gen. Georges Spillmann,

6 Alfred de Gardane (ed.), Mission du Ge'neral Gardaneen Perse

et Napole'on l'Islam (Paris 1969), p. 22 I.

sous le premier Empire (Paris 1865), pp. 5-6 (hereinafter cited as Gardane).

31

32

JOURNAL

OF

PERSIAN

STUDIES

envoy from Napoleon, Adjutant-commandant Romieu, who pressed on towards Tehran. Romieu reached Tehran in October 1805, but died shortly after his arrival; he had time, however, to despatch a Mbmoireto Napoleon. Jaubert, disguised as a merchant, reached Trebizond in May I805, but was unmasked by British agents en route to Erzerum, was imprisoned for a while in the fortress of Bayezid, and did not reach Tehran until June 18o6, weakened by the hardships he had suffered, and critically ill. Fath 'All Shah, terrified at the prospect that a second French envoy might die on Persian soil, sent Jaubert back post-haste, in the care of the Court physician. Fath 'All Shah also sent secret orders that the Court physician was to be executed if he allowed Jaubert to die before he reached the Turkish border. Jaubert survived, and delivered the Shah's reply to Napoleon. Shortly afterwards, Sir Harford Jones, the British Resident at Baghdad, got hold of the text of this document and transmitted it to London. These first two French envoys, Romieu and Jaubert, were followed by four more in quick succession. Jaubert, on his return from Persia, had been accompanied by an ambassador from Fath 'All Shah, Mirza Riz~ Khan. Napoleon received them both at the castle of Finkenstein on April 26th I807. Jaubert, who knew Turkish, Arabic and Persian, and is described as " orientaliste de premier ordre," delivered to the Emperor an extremely shrewd and accurate account of the political and military situation in Iran and Turkey. Persia and Turkey, he said, were ready to enter into an alliance against Russia, but he was not certain whether they would be prepared to join an alliance against Britain. Even if Persia declared war on Britain, what could it effectively do ? Persia lacked " even the suspicion of a fleet," and so could not harass the English by sea. As for the Persian army, declared Jaubert bluntly, it contained men, but no soldiers; a large Persian army in the north was kept at bay by a Russian force only one-sixth its size. In any case, Persian territory did not border on that of the British. However, Persia could inflict great harm on Britain by denying passage to its traders, and by interrupting land communication with India. In conclusion, MIrza Riza Khan promised Napoleon the right of passage across Persian territory for French troops, should Napoleon decide to invade India. Having digested this information, Napoleon decided to send a political and military mission to Iran, and Talleyrand actively supported the idea of a Franco-Iranian alliance.7 The outcome of all this activity was the signature, by the representatives of Napoleon and Fath 'All Shah, at the Imperial camp on May 4th I807, of the Treaty of Finkenstein. The most important clauses of this Treaty were: Article 7: " The Emperor of France undertakes to furnish the Shah of Persia with artillery officers, and engineer and infantry officers, in such number as shall be deemed necessary by the Shah of Persia to enable him to fortify his fortresses, and organize the Persian artillery and infantry according to the principles of European military science." Article 8: " The Shah of Persia undertakes to sever all political and commercial relations with Britain, to declare war immediately on Great Britain, and to act in a hostile manner without delay. The Persian Minister at Bombay is to be recalled. All consuls, factors, and other agents of the East India Company resident in Persia and the Gulf ports, are to be expelled. The Shah of Persia is to seize all British merchandise and to deny to Britain all communication with Persia, both by land and by sea. All ministers, ambassadors or agents of Britain are to be refused admittance, for the duration of the war." Article 1o: " The Shah of Persia is to use all his influence to persuade the Afghans and 'other peoples of Qandahar ' to take up arms against Britain, and to invade British possessions in India." Article I2: " If the Emperor of France wishes to send an army against India via Persia, the Shah of Persia is to grant right of passage to such an army. [Arrangements are to be made in advance regarding transportation, and any auxiliary troops which the Shah of Persia might wish to add to the expedition.]" Article i4: " The privilege allowing the passage of an army across Persian territory is to be granted to France only, and is not to be extended to Britain or Russia." Article 15: " A treaty of commerce is to be negotiated." Article I6: " This Treaty is to be ratified in Tehran within four months."
'

Spillmann,

op. cit., p. 223 if.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

I800-I8IO

33

In compliance with Article 7, energetic steps were taken to despatch a French Mission to Persia, and a number of " brave, intelligent and young " officers were assembled under the command of General Claude Mathieu, Comte de Gardane, who was himself barely forty years old. His family had a tradition of service in the East. The names of fourteen officers and N.C.O.'s are given in the sources. In addition, the names of fifteen civilian members of the Mission are mentioned; these men were to constitute the staff of the French Legation in Tehran. General Gardane's elder brother, Paul Ange Louis de Gardane, was appointed First Secretary of the Legation. It was hoped that " the exceptional reputation for integrity which France enjoyed in the Orient " would stand the Mission in good stead. This reputation, it was thought, was due in large part to the care which the ancien had always taken to appoint for service there only the most honourable men. Napoleon, heedful re'gime of anything which might ensure the success of the Mission, continued this wise policy.8 The detailed nature of Napoleon's instructions to General Gardane, dated May Ioth 1807, leaves little room for doubt that, at that moment, the invasion of India via Persia was for Napoleon a very real possibility, and he clearly relied on the Gardane Mission to provide him with the military intelligence necessary for the formulation of his plans. Gardane was instructed that his first despatches from Persia should contain information designed to make known a country " on which no definite information was available." In this regard, France was badly handicapped by its lack of previous contact with Persia. The military personnel of the Mission were to reconnoitre possible routes across Persia, starting either from Aleppo or from the Persian Gulf. Napoleon gave Gardane four months in which to furnish him with detailed maps and reports on routes, fortresses, and ports (both on the Gulf and the Caspian). A corps of 12,000 men was to be levied by the Persians. If the war with Russia continued, France undertook to supply four or five battalions and two or three companies of artillery to form a reserve for the Persian army. The size of the French army for the invasion of India was envisaged as 20,000 men, but this would be adjusted in the light of Gardane's reports. Gardane was to inquire about the possibility of obtaining assistance from the Mahrattas. Napoleon's instructions were specific as to the type of geographical and logistical information he required. He wanted itineraries of the main highways, to include information on anything of note up to a distance of 3 kilometres on each side of the highway. The scale of the maps was to be I : Ioo,ooo or approximately 2 miles to the inch-no mean undertaking in a country the size of Persia. In addition, special maps and sketches were to be made of towns, villages, forts, camp-sites, positions of strategic importance, etc., on whatever larger scale was considered necessary by the officer on the spot. The location of chief places was to be fixed by astronomical observation wherever possible. Memoranda should contain information on the condition of the roads, the type of soil, the water supply, the availability of food, and the prospects of obtaining remounts and transport facilities for the artillery. Those officers who followed the coasts should observe the configuration of the coastline; should obtain data on normal sea conditions, periodic and occasional winds, and tides; should send details of bays, roadsteads and ports which could afford anchorage to a military or mercantile fleet; should note the possibilities for constructing defences against enemy naval forces; and should investigate the sweet water supplies and revictualling possibilities. Finally, the members of the Mission were to note the manners of the inhabitants, and their attitude towards Europeans. Particular attention was to be paid to their military methods, and an opinion was to be formed as to how readily they might be reconciled to the French military system and tactics. All memoranda were to be despatched in duplicate, and were to be transmitted through the French Legation in Tehran. The Gardane Mission arrived in Tehran on December 4th I807. On December 24th, General Gardane wrote to Talleyrand announcing the ratification by Fath 'All Shih of the Treaty of Finkenstein (December 2oth I807), and the negotiation of a commercial treaty. The latter, said Gardane, was similar to the commercial treaties of 1708 and 1715, but contained a clause which provided for the cession by Persia to France of the island of Karek (Karak, Kharak, Kharg, Kharq), an island of great strategic importance thirty miles from Bushire, after the fulfilment by France of Article 4 of the Treaty of Finkenstein, the article which stipulated the restoration to Persia of Georgia.
8 Gardane,

p. 26,

34

JOURNAL

OF

PERSIAN

STUDIES

General Gardane's early despatches to Talleyrand contain various complaints: the Mission did not have enough senior officers (apart from Gardane himself, there was no officer of higher rank than captain); there were not enough translators; Talleyrand did not reply to his (Gardane's) letters with sufficient promptitude; the climate was terrible, and if the men were to be expected to endure it, they should be given extra pay; as it was, they all longed to return to France. Nevertheless, the Mission addressed itself to its tasks with a zeal which was wholly admirable. Captain Verdier began energetically to train Persian levies. By the time the Gardane Mission left Tehran, fourteen months later, he had 4000 men equipped and drilled in the European manner. Lieutenant Fabvier established a cannonfoundry at Isfahdn and, during the same period, despite all difficulties, produced twenty pieces of field artillery which were superior to anything possessed by the Persians. Barely three weeks after his arrival in Tehran, Gardane sent to Napoleon preliminary observations on the Persian military situation and on probable French requirements, and suggested two alternative routes for the invasion of India. Both routes entailed disembarkation at Alexandretta (with Ottoman permission), and a march via Aleppo to Birejik, at which point the Euphrates would be crossed, and thence to Baghdad. From Baghdad the routes diverged: Route "A " proceeded via Kirmdnshdh, Isfahdn, Yazd, and thence south of Qandahdr to Sind. Route "B " struck north-east via Urfa, Mdrdin, Jazira, where the Tigris was to be crossed on a bridge of boats, thence through Kurdistan Qazvln, Tehran, (eight to nine days of tough marching through mountainous country) to Mashhad and Harat, then south to Kabul, Peshawar and Lahore.9 A French army of 40,000-50,000 Sult.niyya, men would be required. It is interesting to note that this is twice the number envisaged by Napoleon, and it is also interesting that Gardane progressively reduced his demands as the chances of implementing the plan for the invasion of India grew more remote. Initially, however, Gardane considered such a force the smallest possible, in view of the great distance from the point of disembarkation, and the advisability of leaving garrisons at points along the route. The troops selected should be from the South of France, because the Persian climate was dangerous for Frenchmen. Gardane gave the cost of basic foodstuffs in Persia: bread, rice, meat and brandy. Brandy cost 45 sous a bottle, and Gardane suggested it should be stockpiled in advance at Isfahan. The daily ration, he suggested, should be one drachma per man-two in winter and rainy weather. The Persian soldier, thought Gardane, was basically good material: " le soldat persan est sobre, obeissant, supporte les fatigues et a besoin d'&tre conduit et d'exemple."10 His equipment, however, was poor: his matchlocks were ill-made (this contrasts unfavourably with Safavid times, when Persian hand-guns were often superior to European ones). The artillery was in particularly poor shape: many cannon were unserviceable, and there was a shortage of cannon-balls. The " most passable " balls were those taken from the Russians either after battles or captured in towns. The ones cast in Persia were rough and ill-cast, full of flaws and gravel, and they frequently burst on leaving the mouth of the cannon. The falconets, beloved by the Persians, had, in Gardane's view, very little effect. These guns were carried by camels, and were mounted on a pivot placed in front of the camel's hump. The firer sat on a saddle placed behind the hump. At a given signal, the camel squatted, and the gun was aimed and fired. Persian gunpowder was badly mixed, and Gardane thought it advisable that European powder be used for priming. Nevertheless, the position was not hopeless; copper could be imported from Baghdad; saltpetre was in abundant supply and was of good quality; cannon-balls could be improved by using iron-ore from the mines at Amul and at NUr, 140 kilometres from Tehran. Fabvier at IsfahIn had found that the wood of the plane-tree was suitable for most of his purposes, and Persian leather was actually preferable in the Persian climate to European leather, which dried out and cracked. Gardane had one final word of advice for French commissary personnel: " never believe information given to you by the locals!""1 In the midst of this industrious activity, news reached Tehran which rendered all the efforts of the Gardane Mission nugatory, and called in question the value of its continued presence at Tehran. Tsar
g

According to Spillmann, op. cit., p. 237, the detailed information transmitted by Gardane on the possible trans-Persian routes convinced Napoleon that they were too difficult, and he turned his attention to an alternative route via Syria and Iraq to Basra.

10 Gardane, p. 128.
"I

Gardane, p. 138.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

1800--I8IO

35

Alexander, badly defeated at Friedland, had made peace with France at Tilsit in July 1807. Owing to the fact that a letter took four or five months from Europe to Tehran, Gardane had proceeded to Persia in blissful ignorance of the signature of the Treaty of Tilsit, which " converted, in an hour, the Emperor of France and the Autocrat of Russia into sworn friends and active allies."12 Napoleon's " direction politique " had changed, and Gardane was ordered to work for a rapprochement between Persia and Russia. His real difficulties now began. Persia was ready for peace with Russia but only if Georgia were restored to her and Persian territory evacuated by Russian troops. But Marshal Gudowitch, the Russian C.-in-C. in Georgia, had orders from the Tsar not to cede any conquered territory. For Gardane, it was a diplomatic game in which the opposition held all the cards. In the summer of i8o8, however, Fath 'All Shah was still prepared to believe that Napoleon would eventually fulfil his obligations under the Treaty of Finkenstein, although he could not understand why Napoleon had not made the restoration of Georgia a condition of the Treaty of Tilsit, when the Tsar would have had to accept any terms Napoleon chose to impose. At the same time, Gardane's morale was raised by a resounding diplomatic victory over Britain when Brigadier-General John Malcolm, the splendour of whose first mission to Persia in I80oo was still a vivid memory, arrived in the Persian Gulf on his second mission to the Persian court, and was denied permission to proceed to Tehran, Fath 'All Shah adhering faithfully to his obligations under Article 8 of the Treaty of Finkenstein. In order fully to appreciate the nature of Gardane's triumph, we must retrace our steps at this point. British diplomacy in Iran during the nineteenth century was bedevilled by the rivalry between the Home Government and the Government of India. In I8oo Lord Wellesley, the Governor-General of India, took the initiative by appointing Capt. John Malcolm envoy to Persia, "whither no accredited envoy from the British Government had proceeded since the reign of Queen Elizabeth, more than two centuries ago."13 The object of Malcolm's Mission was threefold: (I) to relieve India from the annual alarm of invasion by Zaman Shah, the ruler of Afghanistan, by occasioning a diversion upon his Persian provinces; (2) to restore British and Indian trade with Persia; (3) "to counteract the possible attempts of those villanous but active democrats, the French." From the outset, Malcolm determined to make his visit a memorable one. He was accompanied by an enormous entourage: "six European gentlemen, two European servants, two surveying boys, 42 troopers of the Madras native Cavalry, 49 Bombay Grenadiers, 68 Indian servants and followers, 163 Persian attendants, and 326 servants and followers belonging to the gentlemen of the Mission."14 Malcolm shrewdly perceived that "the two great necessities of diplomacy in Persia were the giving of presents and the stickling for forms."15 Accordingly, although ceremonial was abhorrent to him, he resolved "to stickle as manfully for forms as any Hadjee in the country."16 He was as good as his word. After a lengthy and inconclusive dispute with the Governor-General of Fars as to whether the letter sent by the latter to Malcolm should be in the form of a hukm or a murdsala, Malcolm successfully argued that he and his suite should be allowed to sit rather than stand during audiences with royal personages, and followed this by haggling over the status of his own seat. At the end of all this, Malcolm observed that he was equipped to write a book on Qayd-i Nishast u Barkhast (lit.: " the Rules of sitting down and getting up ").17 Malcolm, unlike the French, found the climate of the country " delightful,"s1 and, after a pleasant journey, reached Tehran and was received by Fath 'All Shah on November I6th I8oo. At a second audience on November 27th, he presented his gifts: " Watches glittering with jewels, caskets of gold beautifully enamelled; lustres of variegated glass; richly chased guns and pistols of curious construction; marvels of European science, as air-guns and electrifying machines; besides a diamond of great value, and the mirrors, which had been brought up with so great toil."19
12J. W. Kaye, The Life and Correspondence Major General John Sir of Malcolm, G.C.B. (London I856), 2 vols., I, p. 399 (hereinafter cited as Kaye). 13 Kaye, I, p. 89. This is incorrect. Sir Dodmore Cotton, the first accredited British ambassador to Iran, had been dispatched by King James in 1627.
14 Kaye, I, p. I16. 16

15 Kaye, Kaye, 17 Kaye, 18 Kaye, 19 Kaye,

I, I, I, I, I,

p. I I2. p. I 13. p. 120o. p. 124. pp. 132-3.

36

JOURNAL

OF PERSIAN

STUDIES

Zamin Shdh's preoccupation with internal problems having removed the need for action against him, Malcolm was able to proceed rapidly with negotiations for a commercial and political treaty, which was signed on January 28th 18o0. Article 2 of the commercial part stated: " the traders and merchants of the kingdoms of England or Hindoostaun that are in the service of the English Government shall be permitted to settle in any of the seaports or cities of the boundless empire of Persia ... and no Government duties, taxes, or requisitions shall ever be collected on any goods that are the actual property of either of the Governments." An additional article stipulated that English iron, lead, steel, broadcloth and perpetts should be admitted free of duty.20 Under Article 2 of the political part, the Shah was to invade Afghanistan if the Afghan ruler attacked India; under Article 4, Britain undertook to supply the Shah with arms if he were attacked by the Afghans. Under Article 5, Britain undertook to supply both arms and troops to Persia if a French army attempted to establish itself" on any of the islands or shores of Persia."21 A clause proposing the cession to Britain of certain Gulf islands caused consternation in Tehran, and was hastily withdrawn by Malcolm.22 His mission accomplished, " the prudent, able, and penetrating, the greatest of the exalted followers of Christ " (as he was designated in the Treaty), returned to India in high good humour, wrestling en route with various problems of translation arising from the Persian text of the Treaty: " the preambles are the most difficult papers I ever read. How often have they made me curse (the) S. . munshial-mamdlik,who, in defiance of reason and remonstrance, persisted in writing such bombastic nonsense.''23 On arrival at Bombay, he was only slightly abashed by the storm of criticism which was levelled at his extravagance, and he made a spirited defence of his conduct. In I802 the Shah sent to India as ambassador HjajjIKhalIl Khan, to obtain the formal ratification of the treaties which had been negotiated by Malcolm. The mission ended in disaster. Iajjl KhalIl Khan, emerging from his ambassadorial residence in Bombay to investigate a brawl between a group of sepoys and a number of his own servants, was accidentally shot by one of the guards. Malcolm was dismayed: " To me it brings the most severe distress. I see in one moment the labor of three years given to the winds ... just when it was on the point of completion."24 The Government of India hastened to make amends for this untoward incident, and granted to Hjajjl Khalll's son, Ismd'11 Khan, a pension of 2000 rupees for life. Ismd'll Khan promptly repaired to Paris, where he continued to draw his pension for sixty-five years. The Shah, however, was not particularly upset by the affair, and observed that " the English might kill ten ambassadors if they would pay for them at the same rate."25 The Persian Government did not send a replacement until 18o5, and the new ambassador, Muhammad Nabl Khan, the brother-in-law of the deceased H~ijjI KhalIl, did little more than use his mission " as a pretext for advancing certain pecuniary claims of his own,"26 and was coldly received. In July 1805 Lord Wellesley had been replaced as Governor-General of India by Cornwallis, who proposed to abandon all the territory west of the Jumna which had been acquired by " Damn your writing, mind your fighting " Lake and Wellesley. Cornwallis died in October 18o5, but it was not until the appointment of Lord Minto in June I807 that Malcolm, the consistent advocate of a positive policy towards Persia, was again able to get a hearing. On November 23rd 1807, Malcolm sent a long and closely reasoned letter to Minto. " The ambassadors of Buonaparte," he noted, " are said to have been very successful in establishing an influence at the Court of Persia." Although their success was no doubt chiefly attributable to the war between France and Russia, the French Emperor, even after peace had been made at Tilsit, would doubtless " discover other ways of improving a connexion through which he is known to cherish hopes of striking a blow at the power of Great Britain." " This danger, though prospective, is very serious, and I am satisfied it will require the most early and spirited measures on the part of the British Government to defeat it. The first measure that would suggest itself is the deputation of a mission to the Persian Court ... ." Britain had to consider " how far it would be politic to abandon Persia to the avowed ambition of Russia, or to encounter the greater evil of allowing that State to throw itself into the arms
21

0oKaye, I, pp. 519-20.

Kaye, I, pp. 524-5. 22 Kaye, I, p. 141. 23 Kaye, I, p. 148.

25 26

24 Kaye, I, p. 179. Kaye, I, p. I95.

Kaye, I, II, 8n.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

I8OO-I8IO

37

of France to avoid subjection to Russia." The rulers of Persia, considered Malcolm, " can hardly be so blind as not to perceive that the complete subjection of their country must be the first step towards an invasion of India, by either Russia or France, as, without that preliminary measure, these nations would be at the mercy of Persia, a change in whose politics would destroy their line of communication, and cut off all hopes of retreat." " Should the King of Persia ... be ever so far deluded as to give openly and decidedly a preference to the enemies of Great Britain, that nation should ... take such steps as were calculated to awaken the Persian monarch to a just sense of the importance of its friendship." In conclusion, Malcolm alluded delicately to the cost of financing a Second Mission to Persia: " the support of our political influence in Persia must and will be attended with both trouble and expense; but what is that to the evils which it is our purpose to avert, and which, though distant, are of a magnitude that must make every reflecting man alarmed for the best interests of his country ?"27 This question was not intended by Malcolm to be a rhetorical one, and Lord Minto did not take it to be such; shortly after receiving the letter, Minto despatched Malcolm to the Gulf. But this time he was not to have a free hand in Persia. Malcolm was not popular either in Leadenhall Street or in Whitehall. He had the reputation of being an able and energetic, but an " unsafe " man. London wanted someone " with less magnificent notions of the greatness of England and the dignity of an ambassador." A suitable man was ready to hand-Sir Harford Jones, the British Resident at Baghdad. Malcolm's biographer says sourly of Jones: " He was not without a certain kind of cleverness, but it had never obtained for him any reputation in Persia, and among the Persians themselves his standing had never been such as to invest him with any prestige of authority, or to secure for him general respect. that he was in What it was that particularly recommended him to the authorities at home-except almost every respect the very reverse of Malcolm-it is difficult to say.""28It is only fair to add that the French had the same opinion of Jones; to them he was " connu par des talents qui valaient mieux que ses principes."29 Not only did Malcolm have a reputation for extravagance, but his patron, Lord Wellesley, was temporarily out of favour, and Malcolm, who was regarded as the " very cock and captain " of the " dangerous Wellesley school," suffered accordingly. Lord Wellesley's younger brother, Sir Arthur Wellesley, who had returned to London from India in order to defend the GovernorGeneral against the attacks of his enemies in Parliament, did his best to get Malcolm appointed in preference to Jones. He wrote to Malcolm in his own inimitable manner: " Government have some thoughts of sending an embassy to Persia; Baghdad Jones as the ambassador. I put a spoke in his wheel the other day, I think, in conversation with Tierney,30 and urged him to get Lord Howick31 to appoint you. God knows whether I have succeeded in the last object, although I made it clear that Jones was an improper man, and that you were the only one fit for the station."32 Sir Arthur (the future Duke of Wellington) had not succeeded, and Jones, newly created a Baronet, was heading for a direct confrontation with Malcolm. Lord Minto instructed Malcolm: " Your commission is framed in such a manner as not to clash with a diplomatic mission to the King of Persia, if you should find Sir Harford Jones at that Court."33 Minto was prepared to meet the French with men to face a French army in Persia, but, he told Malcolm, force; he would furnish 20,000-25,000 " I should be glad to find that less would be sufficient." If the French, instead of sending a large army, sought gradually to establish themselves on the Persian Gulf coast, Minto had 4000-5000 men ready to " push off on the first summons." Indeed, Minto had considered sending such a force with Malcolm, but wisely desisted. Instead, Minto proposed to place on board the King's ships and the Company's cruisers in the Gulf as many Marines as possible; " a force very superior to anything French actually in that country will thus be collected at little expense, and may very much retard, if they do not frustrate, the first projects of the enemy. The only fly in the ointment was Sir Harford Jones," wrote Minto, who " is, I confess, rather a Marplot (since I am writing confidentially) in our play." " I have great confidence, however, in... your conciliatory talents and your magnanimity."34
31 Later Earl Grey, at the Foreign Office. 32 Kaye, I, p. 376. 33 Kaye, I, pp. 402-03. 34 Kaye, I, pp. 410-I.1

Kaye, I, pp. 396-8. Kaye, I, p. 401x. 29 Gardane,p. 63. 30 Rt. Hon. George Tierney, Board of India Control.
27

28

38

JOURNAL

OF

PERSIAN

STUDIES

Malcolm sailed from Bombay on April I7th I8o8, with an impressive retinue of 300 Marines, 1oo cavalry, 50 sepoys, and 2 six-pounder cannons. He intended, he tells us, to address the Persians in the language of " temperate remonstrance and offended friendship."35 Nine days later, on April 26th, Sir Harford Jones reached Bombay. As his vessel, the Sapphire, was working into Bombay harbour, Jones actually saw some of Malcolm's fleet standing to the north, on course for the Persian Gulf.36 Had Jones arrived before Malcolm's departure, Lord Minto would have had an awkward decision to make. As it was, he decided to keep Jones at Bombay pending news of the outcome of Malcolm's mission. As Sir Harford saw it, he had three courses of action open to him: " to go immediately, not to go at all, or to delay going to September." The first course he rejected on the ground that if he followed hot on the heels of Malcolm, this would interrupt the measures which Malcolm must have taken before his own arrival; the second course was impossible unless he were to disobey his orders from the King. To bide his time therefore seemed to him " the best mode," as by such a delay he would show a respect for Malcolm's feelings, and leave the latter full time to conclude his negotiations. All this was communicated to Malcolm by his friend Sir James Mackintosh, the Recorder of Bombay.37 Malcolm reached Bushire on May Ioth I8o8, and found Persia " writhing in the iron grasp of the Muscovite usurper," as his biographer puts it. " The Persians believed that we had deserted them. We had, at all events, looked unconcernedly on, or purposely turned our backs upon them, whilst they had been spoliated by the Northern conqueror. .... Who were we, that we should now expect a single word from us to dissolve a promising alliance and to disperse a magnificent embassy (i.e. the Gardane Mission), strong in all those external attributes best calculated to rivet the confidence of the Persians in the military strength and national greatness of their energetic allies? "38 Who were we indeed! As we have seen, Gardane's influence at the Persian Court was still such that Malcolm was rudely rebuffed, and told that he should communicate with the Governor-General of Fars. In high dudgeon, Malcolm re-embarked on June I2th I8o8, and sailed direct to Calcutta to report to Lord Minto. For once, Malcolm's biographer permits himself a word of criticism, and declares that Malcolm was at fault " in assuming too dictatorial a tone at the outset."39 No sooner had Malcolm left the stage than Jones made his entry. He was released by Minto from " diplomatic quarantine " on August I2th I8o8, and sailed for the Gulf on September I2th, one day before the arrival of a despatch from Lord Minto desiring him to remain at Bombay. His departure was signalled by a salute of guns, " a ceremony," observed his secretary with satisfaction, " which always excites a powerful feeling of respect in the minds of natives."40 A second despatch from Minto, ordering Jones to return to India, arrived at Bushire two days after Jones had left there on December I3th I8o8. Minto wrote to Malcolm in despair: " I doubt very much the power of controlling him from Bengal, although I certainly possess the right to do so, and shall assert it."41 The plain fact was, however, that Sir Harford was now beyond Lord Minto's control. Minto's despatch, ordering Jones to return to India, finally caught up with Sir Harford near Tehran on January 4th I8o9, at a time when he was importuning the Shah to grant him an audience. Sir Harford refused to accept Lord Minto's authority. "I will not retire from Persia," he said, " for many reasons, but principally for this, that by doing so, at this moment, I should proclaim to the Persians, that the Governor-General is superior to the King; and this my Welsh blood will never suffer me to do."42 The unexpected arrival of Sir Harford in the Gulf gave the coup de grdce to General Gardane, whose influence had waned steadily during the summer and fall of I8o8. Indeed, had Malcolm been more patient, he would undoubtedly have achieved his object. Gardane's straightforward, military mind baulked at diplomatic chicanery, and he felt keenly the dishonour of his position. Napoleon had signed the Treaty of Tilsit without making any attempt to fulfill Article 4 of the Treaty of Finkenstein. In September 1808, when St. Petersburg insisted on the immediate acceptance of its peace terms on
35 Kaye, I, pp. 412-13.

36 Sir Harford Jones (Brydges), Bart., K.C., LL.D., An Account of the Transactions His Majesty's Mission to the Courtof Persia in of -7 (hereinafter referred to theyears r8o7-II (London 1834), P. as Jones). 37 Kaye, I, p. 440.

38 Kaye, I, pp. 417-19. 39 Kaye, I, p. 41940 J. Morier, A Journeythrough Persia, etc., p. 2 (hereinafter cited as Morier). 41 Kaye, I, p. 438. 42Jones, p. I28.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

I800-I8IO

39

pain of the resumption of hostilities, and in October, when Marshal Gudowitch laid siege to Erivan, the French Government refused to mediate on Persia's behalf. " This decision," wrote Gardane's son, " placed in jeopardy everything which the Mission had achieved up to that date. France's traditional policy was being sacrificed for the sake of Russian friendship, which could change tomorrow. But the most serious fault, in our eyes, was that France had broken its pledged word."43 Gardane was not even supplied with official guidance from Paris as to how to counter British propaganda and to reply to the reproaches of Fath 'All Shah. On November 23rd I8o8, Fath 'All Shah summoned General Gardane into his presence, and upbraided him bitterly. Gardane, he said, had assured him that Russia could not resume hostilities against Persia without Napoleon's consent. Now, Russian troops were at the gates of Erivan, and the Persian reserves which might have been sent to the aid of the hard-pressed 'Abbas MIrza had been sent on leave, because Persia had relied on the guarantees of her security given by Gardane. Fath 'All Shah was particularly incensed by the fact that Gardane, on October I2th I8o8, had ordered the French officers in Azarbayjan not to take part in military actions until the situation was clarified. Gardane had sent M. Lajard, the Third Secretary at the French Legation, to 'Abbas Mlrza's camp, " to try and convince the prince of the need (for the French) to preserve their neutrality in the present circumstances." From Azarbayjan, Lajard was to proceed to the camp of Marshal Gudowitch, and to attempt to persuade the Russian C.-in-C. to modify his peace terms and adopt a less intransigent attitude. Gardane pleaded that Napoleon's attention had perhaps been diverted from the Orient by the revolt in Spain; he was confident, however, that when the Emperor was apprised of the strange conduct of Russia, he would fall on his enemies like a thunderbolt and annihilate them. Fath 'All Shah rejoined that he had been waiting for the thunderbolt to strike for the past ten months. " It must be obvious," continued the Shah, " that Russia has no intention of submitting to French mediation. Was Napoleon's good name to be lost, and to be replaced by the ill-repute which attaches to a person who does not keep his most sacred promises ? " Even the 20,000 muskets ordered from France had not arrived. " Hardly had you arrived at our Court," continued the Shah, " when you demanded that we break off all relations with Britain. They have been broken off. You demanded that we reach an understanding with Russia. We have done this. You demanded that we conclude an armistice with Russia, and give Napoleon full powers to negotiate a peace. We agreed. You demanded that we send back Malcolm. He has not been received. You demanded that we make no move against the Ottomans, despite our numerous grounds for complaint against the Pasha at Baghdad. We heeded your representations." " Voila, General, votre conduite! ", concluded the Shah.44 Fath 'All Shah's arguments were irrefutable, and the General pleaded for sixty day's grace (i.e. until January 2oth 1809), in the hopes of receiving during that period instructions from France, and also a reply from St. Petersburg to his dispatches delivered to Marshal Gudowitch by the hand of Lajard. Gardane learnt later that Gudowitch had impounded his dispatches, and had sent an express courier to the Tsar urging the latter not to accept the mediation of France in the negotiation of peace between Persia and Russia. The Tsar needed no urging to reject Gardane's request. While Gardane was still reeling under the Shah's onslaught, Sir Harford's arrival in the Gulf made his position at Tehran even more precarious. As the French account puts it: " Sir Jones 6tait a bord, charge d'une mission du roi d'Angleterre pour Feth-Ali Sha."45 Whereas the Shah in May had been prepared to rebuff Malcolm, the Shah in December, despairing now of assistance from France, was ready to accede to Sir Harford's request to be allowed to proceed to Tehran. Gardane notified the Shah that, if the British ambassador were admitted, he (Gardane) must regretfully take his leave, as such action by the Shah would constitute a contravention of Article 8 of the Treaty of Finkenstein. On February 8th 1809, Gardane had an audience with the Shah, at which there were further recriminations, and on February 12th took his leave. The French officers and the Shah parted with mutual expressions of regret. Fath 'All Shah repeated several times, with emotion, and as a form of farewell: " G6ndral, et vous messieurs les Frangais, vous avez 6t6 les bienvenus, les trbs-bienvenus."46
43 Gardane, p. 54. 44 Gardane, pp. 230-44.

45 Gardane, p. 56.
46 Gardane, p. 280.

40

JOURNAL

OF

PERSIAN

STUDIES

Two days later, Sir Harford entered Tehran in triumph and presented to the Shah his gifts, which included: His Britannic Majesty's picture set with diamonds, and a diamond of 61 carats valued at ?C20,000.47 On March 12th I8O9, Sir Harford Jones, in the words of his Secretary, " succeeded in his great object; and concluded a treaty with Persia (where the French influence had already baffled and driven away one English agent) by which the French, in their turn, were expelled, and our influence was restored; at a time when, instead of co-operation, he experienced only counteraction from the British Government of India, and encountered all the rivalry of the active and able emissaries of France."48 The signature of the Preliminary Treaty was indeed a joyous occasion. While the seals were being affixed, the Prime Minister, MIrzd Shafl', kept shouting " Be-zan! Be-zan! " (" Strike! Strike! ") and " in the act of signing and sealing the parties made frequent exclamations, such as 'God grant the friendship between the two states may be binding!' 'May this prove a fortunate day! ' ' Let us hope that nothing may ever break this bond! ' To which every one present emphatically and repeatedly resounded 'Inch Allah! ' "49 The negotiations leading up to the signature of the Preliminary Treaty had not been without their dramatic moments. In the draft treaty, one article had been left indefinite. MIrzd Shafl', the Prime Minister, demanded that it be made definite. When Jones demurred, Mtrzd Shafl' said to him (in obscene terms), " Do you come here to cheat us ? " Jones then " pushed him with a slight degree of violence against the wall which was behind him, kicked over candles on the floor, left the room in darkness, and rode home without any one of the Persians daring to impede my passage." The Persian negotiators chased after Jones to recover the copy of the treaty which had been signed by them. Jones said that he could not be disturbed. "By God," said the Persians, " this Fringee is either drunk or mad." Next day, there was a reconciliation between Jones and the Prime Minister, after the Shah had counselled the latter that, when he was dealing with Europeans, he should take care not to use words which he frequently used to persons of great rank in Iran.50 The news of Jones's success was, as may be imagined, gall and wormwood to Malcolm. Lord Minto opined, however, that " it is too late ... to tell the King of Persia that Sir Harford Jones was subject to instructions which he has exceeded." " My opinion ... is, that we are bound to execute the principal and leading conditions of Sir Harford's Treaty." In Minto's view, however, the payment of the subsidy provided for by the Treaty required the presence of a " person of confidence " in Tehran. " I need not tell you," wrote Minto " all that has been done through the zealous ministry of Sir Harford Jones to lower the rank and estimation of the British Government of India within the sphere of his influence. I entreat you, therefore, to go and lift us to our own height and to the station that belongs to us once more."51 Jones later asserted that his crime " of lowering the estimation in which the Supreme Government of India had, hitherto, been held in Persia," was forever held against him, and debarred him from holding office in the service of the Government of India. " This my scarlet sin," he exclaimed bitterly, " lost me the fruit of nearly thirty years arduous service, just at the moment in which I might have rationally hoped to gather it."52 What made matters worse, his arch-enemy Malcolm gained the reward which was denied to him. " Now the comical part of the story is, that the person whose embassy to Persia to turn out the French, completely failed (Jones is referring to Malcolm's Second Mission), had afterwards the Government of Bombay; and that I, whose embassy was completely successful, was not only refused that, but when the narrowness of my circumstances made it convenient to me to ask for employment, I was refused offices, which persons far my juniors in the Company's Civil Service, might have asked for without being accused of presumption."53 With Lord Minto's backing, Malcolm had his chance to redeem himself, and he made the preparations for his Third Mission to Persia in his usual lavish style. There was one snag: Sir Harford was still at Tehran. He was the representative of the King; Malcolm was only the representative of a Viceroy.
4' Morier, p. 186. 48 Morier, p. 199.

49 Morier, p. 202. 50Jones, pp. I95-9.

51 Kaye, I, pp. 507-1o. 52Jones, p. 2o9n. 53Jones, p. 8n.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

1800-1810

41

If Jones chose to humiliate Malcolm and his Mission, this might be considered not only " proof of the weakness of human nature, but the duty of the Ambassador from the Court of St. James to uphold the dignity of that Court, and to assert, on every occasion, the supremacy of his ambassadorial character."54 For his part, Lord Minto did all he could to discredit and humiliate Sir Harford. He struck a shrewd blow when he sent a despatch to Tehran dishonouring Jones's bills, then amounting to some C65,ooo, which sum Jones had borrowed from bazaar merchants. This action threw Fath 'All Shah into a rage, which he vented on MIrzd Shafl', the Prime Minister, Amln al-Dawla, the Minister of Finance, MIrzA Buzurg, the vazir of 'Abbas Mlrz~, and on Jones himself. The Shah complained of the illtreatment he had received from the English, and lamented " the mighty things he had lost by dismissing the French Embassy, whose praise he kept chaunting alto voce."'55 Throughout the second half of the year I809 the struggle for precedence at the Persian court went on between the two rivals, Jones and Malcolm. Jones compared his position to that of a fully-accredited Persian ambassador to the Sublime Porte, whereas Malcolm was merely a beglerbeg.56 Jones faced his creditors, and succeeded in satisfying them that their accounts would eventually be paid. He wrote at great length to Lord Minto, justifying the stand he had taken in Tehran. He argued that a Government of India representative, " whose credentials and powers run in the name, and by the authority, of a commercial company and their Governor-General," could not have the same status as an ambassador from London.5" His sole motive, he declared passionately, in insisting on his superior status, was to uphold the dignity of his royal master. It was Minto, he charged, who forced him to make the odious comparison between his own position and that of Malcolm. " I had in honor no choice of path left. I may stand on a very slippery footing in this dreadful breach,-power will be exerted to crush a person humble as I am,-I may be on the point of being hurled into the when lying there lacerated and wounded, I will still cry out,-' I have maintained the ditch,-but honor of my King! ' "58 It is clear that Sir Harford possessed at least a measure of that hwyl which is so highly prized by his fellow-countrymen. Finally, Jones unfavourably compared the cost of Malcolm's lavishly-endowed Missions, which he referred to scornfully as Malcolm's " Travels in Persia," with the more modest cost of his own embassy. Basically, however, Jones felt that Lord Minto would never honour him with his confidence, and that his views and those of Minto " of what were the real interests of England and the East India Company in Persia " would never agree.59 Indeed, despite Sir Harford's explanatory despatches, Minto continued to hold an unfavourable opinion of him. At this point Sir Harford made a fatal mistake; he requested permission from King George III to return to England. It was the view of Henry Lord Melville, at the Board of India Control, that this was " the only false step Sir Harford has made in Persia."60 Jones's object, no doubt, was to be allowed to plead his case in Whitehall, but his request had the unexpected result of bringing about his replacement by Sir Gore Ouseley. Sir Harford had powerful enemies in London. He later declared that, " at the India House, out of the 24 directors, there were only two who viewed my appointment from the Crown with common temper, and not many more who showed me the civility commonly due from one gentleman to another."61 There was a further insult to come: Sir Harford had sent ahead to London his private secretary, James Morier, to represent his interests there. Imagine Sir Harford's chagrin when " he, who before he went out with me [to Persia] had never been in any public employment, was all at once, from being my private secretary, taken into H.M. regular diplomatic service, and appointed Secretary of the Embassy in Persia; and the consequence to me was, that I who had been near thirty years in a regular service, and employed in several arduous and delicate affairs ... was taken-into no service at all." No wonder that the ill-fated Sir Harford mused: "Can such things be, And overcome us like a summer's cloud, Without our special wonder? "62
"4Kaye, 5 Jones, 6 Jones, 5"Jones, 58Jones, II, p. 13. pp. 204-07. p. 209. pp. 225-6. p. 219. 59Jones, 60Jones, 61 Jones, 62 Jones, pp. 235-6. p. 235n. p. 356. pp. 228-9.

42

JOURNAL

OF

PERSIAN

STUDIES

This shattering news had not yet reached Sir Harford, however, and he spent the winter of 18o9-1o at Tabriz, where he was harassed by still further despatches from Lord Minto. Minto again ordered Jones immediately to leave Persia, and informed him " that he had appointed Surgeon Jukes, of the Bombay Establishment, to take charge of the British interests at Taeheran, until the arrival there of General Malcolm." Minto also made a strong request to Fath 'All Shah that " his Majesty would, from that time forward, cease to hold any communication with me (Jones), or to consider me as a person accredited to him." Jones commented acidly that " It would have been as well for my lord and his adviser, if they had looked into Vattel or Wiquefort, and instructed themselves, as to what justifies a Minister leaving the territories of a Sovereign to whom he is accredited, and by whom he has been once acknowledged, before he issued these orders.""63He went on: " though I held an appointment under my Sovereign's sign manual, and under the great seal of England, his Lordship [i.e. Minto] forgot himself so much, as to treat me as if I was solely under his orders,-as if I had committed to use language which, some enormous crime,-or had been guilty of some egregious folly;-and if I had compromised my Sovereign's honor, or injured the interest of his kingdom, his gracious and noble mind would never have commanded faults, heavy even as these, to be rebuked with. I confess, without reserve, my patience was now exhausted; and being a native of that part of the King's dominions where that virtue is perhaps less exercised than in some others ... in as plain English as I could make use of, I gave his Lordship my opinion both of his style and his measures."64 It seems to have been about this time that Sir Harford tendered his resignation, which was accepted with alacrity. Meanwhile, Malcolm had sent Jukes on ahead to tell Jones that he " hoped he would see the propriety of their meeting as countrymen, and saving at least outward appearances." Malcolm assured Jones that he would not " assume with the Persian Court any duty beyond that of being the officer empowered by the Indian Government to execute such parts of the preliminary treaty as related to the employment of its resources."65 but Jones imposed impossible conditions. Condition No. 6, " that Jukes met Jones at neither the Persian Ministers themselves, nor any person upon their behalf, shall enter into any disSult.niyya, cussion on affairs connected with the British Government with any other person except the Ambassador " [i.e. Jones], clearly precluded the possibility of co-operation. Condition No. 2 was merely petty: " that the General (Malcolm) shall sound no trumpets on his entrance into the King's camp nor carry any flags." It was Fath 'All ShTh who broke the deadlock, by announcing that he intended to receive Malcolm with the same honours as in 18oo.66 On April 8th 18Io, Malcolm received the royalfarmdn to proceed to Tehran. Jones capitulated, and received Malcolm " in a gracious and affable manner." Malcolm was received in audience by the Shah on June 23rd, and the Shah inquired why he had retired from Shiraz in dudgeon in i8o8. Malcolm replied: " How could he who had been warmed by the sunshine of His Majesty's favour, be satisfied with the mere reflection of that refulgence through the person of His Majesty's son? " " Mashallah! Mashallah! " cried the King, " Malcolm is himself again." " Mr. Jins (sic!)," said the Shah, " is a good man. I have a regard for him, and those with must do everything for me now.""67 him have labored hard in my service-butyou received Malcolm again. The Shah was especially delighted by the gift of two On July Ist the Shah galloper guns, and at the sight of Lieut. Lindsay, a giant of a man who stood 6 ft. 8 in., whom the Shah likened to Rustam. Sir Harford, however, had one trump card left, though it must have cost him a great deal to play it. Shortly after this audience, he handed Malcolm a despatch from Whitehall, which declared that London still intended to regulate British diplomatic relations with Persia, and had, " in prosecution of this intention to repudiate the power and authority of the Governor-General (of India) in that direction, appointed Sir Gore Ouseley ambassador to the Court of Tehran." Malcolm " saw at once that it had become his duty to bring his mission to a close. . . . He could no longer remain at the Persian Court in a recognised official position." Jones had won, but only at the cost of finding himself supplanted by Ouseley.
63Jones, p. 345 and n. 4 Kaye, II, p. 15-i6. 6 Jones, pp. 346-7.
66 Kaye, II, p. I8 ff. 6 Kaye, II, pp. 23-5.

BRITISH

AND

FRENCH

DIPLOMACY

IN

PERSIA,

I800--I8IO

43

Malcolm asked the Shah for permission to leave. The Shah besought him to stay, but, finding Malcolm adamant, desired to invest him before he left with the Order of the Sun, which he had conferred on General Gardane. Malcolm refused to accept a title which had been " instituted for the benefit of an enemy." The Shah thereupon created another new Order, that of the Lion and the Sun.68 At the investiture on June 15th 1810, Malcolm was also made a khdnand sipdhddrof the Persian Empire. " You are now," said the Shah, " confirmed in my service, in which I know you have been faithful for ten years. I can do no higher honor to any one than at this moment I have done to you. You will wear this star on your breast as a proof to all the world of the royal favour of the King of Persia."69 Not long afterwards, Sir Harford also left Tehran. At his final audience with the Shah, Jones was in tears; Fath 'All Shah said, " Go,-God go with you, prosper you and protect you! "70 Lord Minto's hope that British relations with Persia would once again be placed under the charge of the Governor-General of India, and that Malcolm would remain in Tehran as Resident Minister charged with the duty of giving effect to Jones's Preliminary Treaty, had been dashed. To this extent, Malcolm's Third Mission had been a failure. But although the political results were meagre, literary and scientific information of inestimable value was brought back by the Mission. Malcolm's History of Persia, with its accurate insights into the Persian way of life, remains a classic. There were intangible benefits, too, from Malcolm's Mission. Sir Robert Ker Porter, who visited Persia in 1818, wrote: " Everywhere that I went in the empire his mission had led him, still I found his remembrance in the hearts of the inhabitants. In many of the villages the people date their marriages or the births of their children from the epoch of his visit amongst them; for wherever he appeared his goodness left some trace of himself, and the peasants often said to me, that if the rocks and trees had suddenly the power of speech, their first word would be ' Malcolm '."71 The principal actors had left the Persian scene, but offstage, echoes of the rancorous strife between Jones and Malcolm were occasionally heard. Malcolm was very proud of the fact that he had introduced the potato into Persia. Some had been planted at Shlraz, and some thirty-odd bags of seed potatoes had been given away in different parts of Persia, together with a memorandum on the method of cultivation. Jones, however, could not allow his rival to take credit even for this: " Long before Sir John Malcolm visited Persia," he wrote, " I gave roots of this plant to several Persians." In the opinion of one of the recipients, Mlrzd Buzurg, " boiled, they were eatable, but only as something to keep body and soul together; in ragout, they were detestable; and roasted, not half as good as the chestnut. .... You see God Almighty provides the greatest of all delicacies and the wholesomest of all food [rice], for the faithful, and leaves you what is fit only for badgers, porcupines and squirrels." Malcolm promptly investigated the situation, and came to the conclusion that Jones's importation was the dla-yi zamin, " not unlike the potato in form, but (having) no resemblance in taste, being a bitter, useless root." " I was much relieved," he wrote, " as I desire the good fame of introducing potatoes into Persia, and look to immortality in the name they have received in that country-dla-yi Who emerged the winner from this decade of intense diplomatic activity ? Britain and France had each expended much, and achieved little. Russia had expended little, but had occupied large tracts of Persian territory, and, by the Treaty of Gulistdn in 1813, occupation was to be turned into permanent annexation. Napoleon had abandoned his project too soon, and had allowed himself to be diverted by other affairs. He had dreamt of " la M6diterranfe frangaise d'abord, la conquete des Indes ensuite." But he was a realist. " Il ne soutiendra les pays musulmans que dans la mesure ou ils lui seront utiles."'73 As matters got progressively more difficult for him in Europe, he gave up his dream of following in the footsteps of Alexander the Great. British policy was frustrated by the rivalry between Whitehall and India, and by the " blow hot, blow cold " attitude of successive governments. Russia's policy of imperialist expansion at the expense of Persia had been consistent since the time of Peter the
Malcolm's " immortality " in this respect was short-lived. 73 Spillmann, op. cit., p. 388; pp. 391-2.
72 Kaye, II, pp. 47-8.

Malcolmiyya.'"72

68

Kaye, 69 Kaye, Jones, "0 71 R. K.

II, pp. 27-31. II, p. 34. p. 377. Porter, Travels, quoted in Kaye, II, p. 53n.

44

JOURNAL

OF

PERSIAN

STUDIES

Great, and was quick to exploit any weakness in its rivals. As Alfred de Gardane justly observed: la " Qui done a gagn &' retraite de la France de Perse ? La Russie seule."74 Sir Harford Jones's Preliminary Treaty was eventually ratified in I814, but the British Government failed to discharge its commitments under this Treaty. In 1826 Canning, in a letter to the Duke of " Wellington, referred to the Treaty as a most unlucky effort of negotiation,"75 and it seemed to Malcolm that Canning was " most anxious to shake off Persia."76 We may allow Malcolm the privilege of the last word on this: " In point of policy I believe him to be wrong; but, supposing him to be right, he must take care that he does not, by injuring our reputation for good faith ... destroy that strength on which we must trust for every stand we may hereafter have to make, from the banks of the Araxes to those of the Ganges, against the encroachments of Russia. And with respect to all questions of faith, as connected with Asiatic states, we must decide them according to their understanding of them when the obligation was contracted. Better for our character to break a treaty at once, than to fritter it away with nice distinctions drawn from Puffendorff, Grotius and Vattel, familiar to our diplomatists, but unintelligible to Courts like that of Tehran. Such a proceeding would add to the belief of our bad faith an impression of our art and meanness! ,7"" Malcolm and Gardane may have been implacable enemies, but they spoke the same language.

P4 Gardane,

p. 67. Letter dated November 22nd 1826; 56 p. 454.

76

quoted in Kaye, II,

Letter, dated December x2th 1826, to the Duke ot Wellington; quoted in Kaye, II, p. 455. 77 Ibid.

SALJUQ MONUMENTS IN IRAN: II. THE " PIR " MAUSOLEUM AT TAKISTAN1 By Robert Hillenbrand
In the thirty years since the Survey PersianArt was published,2 knowledge of the early Muslim of architecture of the eastern Islamic world has improved considerably. This is partly the result of a more rigorous examination of already known monuments, but to a greater extent it is due to a mass of hitherto unknown buildings discovered in Iran, Afghanistan and especially the U.S.S.R. Indeed, one is approaching the state where regional histories of eastern Islamic architecture in certain limited periods could profitably be written. Perhaps the most interesting group of buildings are the mausolea of Soviet Central Asia, which display a series of variations on the domed chamber of square ground plan. Few of them are paralleled elsewhere in the architecture of the time. The development of domed commemorative structures in Islam was studied recently by Professor Grabar.3 A glance at the catalogue he has drawn up will suffice to show that the domed square so common in Central Asia was far from being the typical Iranian mausoleum. From the Gunbad-i Qatbiisonwards,4the preferredtype of mausoleum in Iran was the tomb tower, and this continued until well into the Il-Khanid period.5 This at least is the logical conclusion to draw from the evidence of surviving buildings. But a few domed mausolea of square plan were built in Iran simultaneously with the first tomb towers. The present article adds a little known, probably Saljiq, monument to the list.6 and The Name,location historical background. monument is situated on the outskirtsof Takistan, a small town on the road from Qazvin to Tabriz and Hamadtn. Although this is one of the busiest roads in the country, the building has apparently escaped notice until quite recently; it seems first to have been mentioned by M. T. MustafAvi, who published a photograph of it and assigned it to the eleventh century A.D.7 The site of the building, in an extensive cemetery some 2 km. south of the main road, is not easy to find. Local tradition identifies it as the tomb (buq'a)of a pir, literally " old man "; but the term is also used to designate a spiritual teacher.8 If the latter meaning is intended, one might S.ifi's
For the first article in this series, entitled " The Masjid-i Jdmi' at Qurva ", see OrientalArt N.S. XVIII/I (1972). 2A Surveyof Persian Art, from PrehistoricTimes to the Present, ed. A. U. Pope (London and New York I939). Grabar, " The Earliest Islamic Commemorative Structures, 30. Notes and Documents ", Ars OrientalisVI (1966), pp. 7-45. 4 Or even earlier. I hope to be able to show elsewhere that the octagonal mausolea at Samirin may well be of tenth century date. 5 It was also common for people of note to be buried in mosques, and khdnqdhs. madrasas 6 Leaving aside the buildings which have been variously classed as Saljfiq or Mongol (Tfis; the 'Alaviyydn at Hamaddn; Zavan; Khwa-ja Sa'd), and the pre-Mongol mausolea of domed square type in Afghanistan, which number at least five, the earliest Islamic domed square mausoleum in Iran known at present is the Davdzda Imam at Yazd, dated 429/1037. The mausoleum at Sangbast may be a little older. The Tdkistdn mausoleum, by virtue of its much smaller size, belongs to a different sub-category. It is related rather to the Mazdr-i Khusraubdd, near Tabas, which is probably of the twelfth " century and may be even earlier (see my note Mosques and Mausolea in Khurdsdn and Central Iran ", Iran IX (1971), pp. 161-2). The question of dating the mazdr is complicated
1

by the lack of related monuments in the general area. The mazdr may originally have resembled Tdkistdn more than it does now, for it has traces of what may have been a pish.taq. Mongol tombs fairly similar to TakistAn are Langarak (W. M. Clevenger, " Some Minor Monuments in Khurisdn ", Iran VI (1968), pp. 59-60, pls. IIIa-Vd) and the Imimzdda 'Is~ at Kanak near Vardmin (pl. VIIIb: I am grateful to Mr. J. W. Allan for drawing my attention to this building). In his PersianArchitecture a Glance(Tehran 1966), photograph at on p. 54. The entry in the Fihrist-i bindhd-yitdrikhiva amdkin-i Irdn (Tehran I345 A.S.H.,p. 181; Mustafavi, op. cit., bdstdni-yi p. 7o, n. I, mentions that the author of this book is Nusratallah Mishkati) dates the building in the sixth/twelfth century. See also my note " Islamic Monuments in Northern Iran ", Iran VIII (1970), p. 205. I recorded the building in October I969 and September 1970. I would like to thank Mr. R. H. Bishop for his drawings of the plan and section (Fig. I) and Miss D. M. Haigh and Mr. D. H. Gye for their assistance in recording the building. I am also grateful to them all for their helpful comments. 8 R. A. Nicholson, Studiesin Islamic Mysticism (Cambridge 1921), pp. o10, 22-3. The term buq'a is commonly used for the mausolea of dervishes.

45
5A

46

JOURNAL

OF

PERSIAN

STUDIES

relate the building to a similar, late Saljfiq, structure at Ddmghdn.9 But no further details are available. The building is anepigraphic, and Tdkistdn does not appear to be mentioned by medieval geographers, at least not under that name. Thus very little historical data exists to throw any light on the " Pir " mausoleum. There is not even a name to serve as a clue. The problem is much the same as for the Kharraqtn tomb towers, where names are given but cannot yet be related to any personages known from other sources.10 The mausoleum is not of a size and impressiveness appropriate for a ruler, and its site calls for no special comment. The general area was no doubt a convenient grazing ground for the herds of Saljiiq and later nomads, but there is nothing to connect the building with nomadic chieftains, as was tentatively and most plausibly suggested for the nearby Kharraqdn towers."1 Its site by, or even in, a large settlement would militate against this theory. In fact, the interest of the Takistan mausoleum is very largely architectural and decorative, and it is an inadequate peg on which to hang historical disquisitions. The most one can say is that local tradition is probably correct in identifying the building as a mausoleum. It is true that no medieval mausolea or tombstones survive in the area of the monument, but its funerary purpose must have been recognized, or a cemetery would not have come into being there at a later date. It is of course possible that later generations took the building to be a mausoleum even though there was no evidence for this; but it is difficult to suggest any other function for a building of this size and plan. The mausoleum contains several modern gravestones. Exterior. The exterior appearance of the mausoleum offers a contrast of simple and sophisticated elements. Only the northern side, which is that of the entrance faqade, is highly ornamented (Pl. I). were a frame for the doorway-a feature found also at the Indeed, the effect is as if the entire Gunbad-i Surkh and the Gunbad-i facade Ghaffiriya, both at Marigha. As an architectural element, the faqade is not a logical development from the rest of the building. Rather is it a decorative revetment intended to emphasize the principal side of a simple square plan. Inside the mausoleum the north wall is no longer of special importance. For this reason it mirrors none of the complexities of the facade. The south side of the tomb has a pattern in decorated brick which occupies about a fifth of the wall surface; the rest of the wall is plain (P1. IIa).12 It seems possible from the placing of this decoration-it is slightly awry-that the building may have been abandoned unfinished.13 Unplugged scaffold holes are found in two rows on the east, south and west sides, with three holes per row; in addition there are irregularly placed scaffolding holes on each side.14 The east and west sides are quite blank; their sole decoration is a simple moulding, which projects three bricks and runs along the top and down the corners of each side. Each of the four exterior walls of the tomb have this moulding, though on the facade engaged columns replace it on the corners. In this comparative emptiness of three out of four sides, with the fourth side richly decorated, one sees a distinct similarity to the Central Asian mausolea, for what distinguishes these from other tombs of similar ground plan is their elaborate fagade. The three mausolea at Uzgend, dating from 1012-13, 1152 and I186-87, illustrate this feature.15 It was
* Illustratedin M. 'A. TdhiriyA, Ddmghdn: shishhazqrsila " " excavation in the immediate vicinity of the TakistSn tomb to look for traces of buildings contemporary with it. p. (Tehran 1347 A.S.H.), 125. See also Fihrist,p. 213, where it is called ImdmzddaMuhammadand dated in the Timfirid 10 The same is true of the Gunbad-i Surkh at MarAgha, despite its wealth of inscriptions (A. Godard, " Notes compl6mentaires of period. D. N. Wilber, in The Architecture IslamicIran: the sur les tombeaux de Mardgha ", Athdr-d Period(Princeton 1955), p. Iii, also dates it in the II Khdnid IfrnI/2 (1936), p. 135). fifteenth century; but for the argumentsagainst this, and a xxOn the basis of the names given in the inscriptions, and the fact that the area is mentioned twice in medieval sources as the translationof the inscription, see my note " Mosques and Mausolea", p. 16i. The building is known locally as the camping ground of princes (S. M. Stern, " The Inscriptions of for the Kharraqdn Mausoleums ", Iran IV (1966), p. 26). of khdnqdh Shah Rukh, a ratherunlikelyidentification such a small structure; but it is also referredto as a buq'a. Its form 12 The panel is 209ogm. high and begins 2 -58 m. above ground level. does indeed suggest that it was intended to be a mausoleum. to It was not at all unusualforSfifi shaikhs be buriedin or beside 13 Cf. the tower of Chelebi Oghlu at Sultnilya, where the decorative brick bond on the exterior stops abruptly at about onetheir khdnqdh the exampleat Gizze, and a brief discussion (for third of the height of the blank arches. of similar monuments, see G. Gropp, " Bericht tiber eine Reise in west- und Siidiran", A.M.L N.F. III (1970), p. 195). 14 These unplugged scaffolding holes do not offer corroborative flourished evidence for the theory that the building was left unfinished; For a discussionof how Sfifis and their khdnqdhs underthe Saljfiqs,see C. Cahen," The TurkishInvasion: The they are a feature shared with many completed structures. " Selchiikids in K. M. Setton and M. W. Baldwin, eds., A 15 E. Cohn-Wiener, Turan. Islamische Baukunst in Mittelasien I (Berlin 1930), pp. 35-6, pls. XI-XVI. Historyof the Crusades (Madison, Milwaukee and London 1959), P. 155. It would be interestingto carry out a trial

SALJUQ

MONUMENTS

IN

IRAN.

II

47

equally pronounced in the tenth century.16 Without forgetting the originally monumental aspect of the iwdn, as displayed in Parthian and Sasanian architecture, one is perhaps not being too fanciful in which was to form such an important part of Mongol and especially postsuggesting that the pishttdq,

Fig. I. Groundplan and section of the " Pir " mausoleum, Tdkistdn.

Mongol architecture, developed to an equal extent from these sumptuous Central Asian facades. The ftaade of the Takistan tomb carries the principle of recession, already faintly suggested in the square panels of the east and west sides, to unusual lengths. On the fagade, the central panel is recessed five times, and each of the side panels three times. The aim is clear-in the central panel, to create a
16

It occurs on the " Tomb of the Simdnids " at Bukhara (L. I. Rempel, " The Mausoleum of Isma'il the Samanid ", Bulletin Institute PersianArt and Archaeology (1936), IV of the American for
5B

pp. 198-209; Grabar, op. cit., p. 17) and on the mausoleum of 'Arab-ata at Tim (G. A. Pugachenkova, " Mazar Arab-ata v Time ", SovietskaiaArkheologia(196x, iv), pp. 198-211).

48

JOURNAL

OF PERSIAN

STUDIES

sense of progression to the entrance; and in the side panels, to balance this complexity.17 Similar effects are known in Armenian architecture.18 is the panel of three-tiered over-riding Stalactites. The single most notable feature of the facade stalactites over the door (P1.IIIa). This panel is on an almost flat plane; the stalactites scarcely project and are not corbelled over each other, though by a trompe l'oeil effect this seems to be the case.'9 In where the composition has been damaged one can see that the cut brick segments of which they places are constructed are set into a thick, deep mortar bed.20 These stalactites are therefore solid and closely linked to the structural core, which makes it likely that they are of Saljfiq date. In the Mongol period and later, stalactites were usually built up on a wooden framework and consisted merely of a thin plaster shell.21 The construction of such stalactites can be appreciated easily enough by studying a photograph of them in decay.22 Solid stalactites of the Tdkistan type were to be found in the now destroyed Masjid-i Jdmi' at Demavend.23 Related forms are not rare in the Saljfq period, and may also be found on entrance faCades.24 On close inspection one is struck by the coarse workmanshipof the individual details in the stalactite panels (Pls. IIb and IIIb). The brick segments are not separately moulded but have been cut in a rather haphazard way from larger bricks. The sharp, sometimes jagged, edges of some of the bricks suggest that the cutting was done after firing, possibly by the mason as he built the structure; bricks which had been prepared in advance-such as those on the corner columns of the Gunbad-i Surkhwould have had smoother edges. The very thick mortar beds serve to disguise some of the inevitable unevennesses. About six different brick shapes are used, of which the smallest is less in area than the mortar which binds it.25 That the mortar is of low quality is demonstrated by the severe weathering which some of the joint plugs have undergone (P1. IVa). Comparison with other exposed plugs, such as those at the Varamin Jimi', proves that some mortar is hard enough to resist the ravages of the weather. Only the plugs immediately under the main cornice, and therefore in the most sheltered position, have survived intact. Arch profiles.A distinctive feature of the fagade is the form of the central stalactite system and of the subsidiary ones in the side panels. The basic unit here is an arch which is technically trilobed, though the cut brick which is used makes the cusps more angular than rounded. These arches, which are often found in pairs, occur with such frequency in the Saljfiq architecture of northern and north-western Iran (and rarely at other places and times) that they serve to date a building within about Ioo years. They are to be found in two mausolea at Samirdn, the IHaidariya,the Kharraqan and Demavend mausolea and the Demavend Jami'.26 The TikistTn arches are among the more developed examples of the type. Through their diminuitive size they may lack the purity of the IHaidarlya arch, the finest of the
17

Cf. the mausoleum of 'Alambardara at Astdna-BgbS (G. A. Pugachenkova, " Puti Razvitiya Arkhitekturi Iuzhnogo Turkmenistana Pori Rabovladeniya Feodalizma ", Trudi VI Arkheologicheskoi luzhno-Turkmenistanskoi Ekspeditsiy (Moscow 1958), pp. 270-0). The doorway of the so-called tower of Shaikh Shibli at Demdvend also once had similar multiple recessions, as has that of the Gunbad-i 'Alaviyyin, Hamadin. Spatial experiments of this kind became increasingly popular in the Mongol period, as the portals of the caravanserais near Marand, at Sin and at Sarcham show (Wilber, op. cit., figs.

23

Arminiennes Revue des N.S. V (1968), pl. ,tudes XLVII, fig. 2. Here too the purpose is to mark the entrance by architectural as well as decorative means. 19The earliest stalactite system so far recorded, that at Tim, is also on an almost flat plane. 20 Cf. the solid stalactites in the portal of the Gunbad-i Qabfis (A. Godard, " Les Stalactites", Athdr-d Irdn IV/ (1949),
21 J.
22

18s E.g. at Soradir. See P. Cuneo, " L'6glise de Ste. Echmiadzin

55-7).

A Soradir ",

Rosintal, Pendentifs, Trompes et Stalactites (Paris 1928), pp. 89-92. E.g. that of the Masjid-i Shah, Isfahdn (E. Diez, " Persien. Islamische Baukunst in Churisan " (Darmstadt 1923), pl. 30).

fig. 281).

M. B. Smith, " Material for a Corpus of Early Iranian Islamic Architecture. I. Masdjid-i Djuma', Demdwend ", Ars Islamica II/2 (1935), PP. I2-14, figs. 9-10, 13-16 and I9-20. 24 E.g. at Seh Gunbad, Rijd'iya, and Khwaja Atdbeg, Kirmin. 25 One cannot be certain of the original number because of the damaged state of the facade. Some of the bricks still have a shallow incision, perhaps marking the spot where the brick was to have been cut. 5 x 6 cm. and 7 x 6 cm. are common sizes. The mortar is usually cm. thick. 3"5 26 Related forms of small trilobed arches, but without the distinctive keel shape, and not occurring as twin arches, are to be found in the portal of the Gunoad-i Qabfis and in the cornice of the Lajim tower (for the latter, see A. Godard, " Les tours de Ladjim et de Resget ", Athdr-eIrdn I/I (1936), fig. 74). Even closer is the form of a subsidiary mihrdbin the Gulpdyagdn Jdmi' (Smith, op. cit., fig. 12). The keel shape in recurs in the lower arch of the miihrdb the Marand Jami' (Survey,pl. 398), and in a stalactite wall panel in the Sujas Jami'. The Saljfiq trilobed squinches of Central Iran are simply too large to be used for purposes of comparison in this context, though it could be argued that they developed from such small beginnings.

SALJfJQ

MONUMENTS

IN

IRAN.

II

49

series,27but they offer a variant on the theme, for the arch on each side panel encloses three small facsimiles of itself.28 Brickwork.Common bond is used for most of the exterior.29 Vertical bond is used for most of the framing bands of the faqade. It consists either of a course of stretchers laid vertically or of a vertical course of stretcherswith headers alternately above and below them. Some of the framing bands of the fagade are of headers laid in common or vertical bond. That the architect was familiar with the techniques of decorative brickworkused elsewhere is shown the similarity of the lozenge pattern on the south side to those on the Demavend and Kharraqan by towers30and from the use of decorative engaged corner columns. But comparisons with the three most elaborately decorated of eleventh century towers are uniformly to the disadvantage of the Thkistin mausoleum. On the south side, the application of decorative brickworkflush with the structural core is slightly unusual.31 The decoration of the fagade is all applied to a structural core, as is the case with the Demdvend and Kharraqdn towers and many other Saljfiq buildings. A combination of mortar and brick in equal proportions, as in the engaged columns, is to be found in the lower registers of the side panels.32 The design, which is executed in doubled stretchers, is basically that of a reversed Y; it is a pattern which was popular throughout the Saljfiq and Mongol periods (Fig. 2).33

Fig. 2. Decorated brick panel on the entrancefa;ade.

A surprisinglysubtle touch, in view of the poor quality of much of the faqade, is the use of half-size bricksin the stalactite panels. This enriches the pattern. Curved, presumably moulded, half-size bricks are used in the outer framing band of the central panel, another device which displays familiarity with advanced brickworktechniques. There is a serrated brick moulding just below the cornice. columns.The engaged columns flanking the doorway have capitals which approximate to a Engaged lotus shape and are c. 12 cm. high. No close parallel for the motif in this material and at this time suggests itself.34 It seems doubtful that it had a symbolic significance.35
27

Probably also the latest. Strictly speaking it should not be included, for it is of multi-stepped profile; but its crown is keel-shaped. 28Also found in simplified form on the so-called tower of Shaikh Shibli, Demavend (D. B. Stronach, " The Demdvend Tomb Tower ", pl. IIIb, apudDavid Stronach and T. Cuyler Young, Jr., " Three Seljuq Tomb Towers ", Iran IV (1966), pp. I-2o). The arches of the central panel at Tdkistdn are of triangular plan, like most of those at the Demdvend Jami' and those at the HIaidariya. 29 The typical bricks are pinkish or yellow and measure 20o 5X 5 cm. to 21 X 5 cm. The bricks at the corners measure 19 xio x 5.5 cm. Normal horizontal joints are 3-4 cm., rising joints also 3-4 cm. o0Stronach, op. cit., fig. 4, nos. 6 and 7; Stronach and Cuyler Young, Jr., op. cit., passim. 31 Wilber gives several Mongol examples of decoration applied as the wall rose (op. cit., p. 55). In these cases, as at Tdkistdn, the pattern was deliberately designed to leave spaces for the

32 The mortar is ii cm. deep on the east panel and 23 cm. deep

scaffolding holes.

on the west panel. "aThe design is similar to a panel at the Jurjir mosque in Isfahin, probably of Bafyid date (A. Godard, " The Jurjir Mosque in Isfahdn ", SurveyXIV (1967), fig. I162 on p. 31o2). Other examples are to be found on the fagades of the Gunbad-i Surkh, Mardgha; the Gunbad-i 'Alaviyydn, Hamaddn; at and the khdnqdh Natanz. 34 The capitals on the fagade of the southern mausoleum at Uzgend (dated i I86-87) have a related form, which continued in use in Central Asia for centuries (D. Hill and 0. Grabar, Islamic Architectureand Its Decoration (London 2964), figs. I15-16 and i18; cf. figs. 75, 8o and 88). 35 For general remarks about the significance of the lotus in preIslamic Persian iconography, see P. Ackermann, Survey, p. 867, especially n. 3. It is difficult to assess the validity of these remarks for later periods.

50

JOURNAL

OF

PERSIAN

STUDIES

Little is left of the engaged columns at the corner of the faqade. The poor integration of these columns with the rest of the fagade betrays the builder's immature handling of architectural form. They are used only in the spirit of applied ornament, as is demonstrated by their small size and by the fact that they do not appear on the south-east and south-west corners. The engaged columns of octagonal tomb towers, such as those at Samirin and Kharraqdn, offer a strong contrast. They are placed at each angle, and this, together with their greater size, allows them to play a much more decisive role in articulating the structure. The columns at Takistin afford another link with the architecture of Central Asia, where similar columns occur in mausolea from the so-called " Tomb of the Sdmdnids" onwards.36 In that area they are usually of brick, closely integrated with the building, whereas at Takistdn mortar is applied very lavishly to a brick core, and bricks set in it to form a rough pattern. The bricks are in vertical bond, with two coursesof bricksin common bond dividing the vertical lays (P1.IIc). This recalls the structure of Parthian, Sasanian and early Islamic columns.37 The technique bears a slight resemblance to some of the corner buttressesand engaged columns at the earlier Kharraqdn tower, but the mortar is much thicker and the pattern cruder in both conception and execution. The pattern should rather be compared to that on minarets at Niriz and 'Ala; the only difference is that on the minarets the horizontal, not the vertical, courses are dominant.38 Brickdecoration: summary.The facade contains a surprisingly wide range of decorative motifs for a composition which at first strikes one as rather heavy. The cumulative effect of many small touches which only register on detailed examination is one of unexpected richness. It must be admitted, however, that the decorative brickwork is equally far both from the sophistication of the Gunbad-i Surkh (dated I I48)-where bricks of twelve39different shapes interlock like pieces in a jigsaw puzzle, and the joints are of minimum thickness-and from the bold contrast of various brick shapes on the " Tomb of the Simdnids " at Bukhara, probably built in the first half of the tenth century. The evidence available suggests that it would be unwise to attempt to trace a coherent chronological development of decorative brickwork techniques from the early tenth century onwards in the eastern Islamic world. One of the finest examples of the style, the " Tomb of the Sdmdnids ", is also commonly acknowledged as the earliest which survives. Developments in the various provinces of the Iranian world were of uneven quality. The Takistdn mausoleum demonstrates only that in the Saljfiq period the standard of decorative brickworkwas not uniformly high. The quality of its decoration is thus no guide to its date, though individual decorative elements can be used as dating controls within a local context. Dome. The dome is of unusually steep profile, and the first parallels which spring to mind are again Central Asian mausolea.40 A further example of the lack of imagination displayed by the architect is the way in which the dome is placed in the middle of the roof, with no attempt made either to raise it to a commanding position by means of a high drum or to effect a smooth transition between it and the lower walls by a gallery or by an externalized squinch zone. Both methods were well established by the eleventh century.4' Perhaps the rough juxtaposition of dome and base is a further indication that the building was abandoned unfinished, but this is doubtful. The building has been restored more than once, at indeterminate dates.42 Traces of this Restoration. work remain on all four sides. The upper half of the dome appears to have been rebuilt. The entrance
38 The

mausoleumof Bayan QQli Khan at Bukhdrd(Rempel, op. cit., pp. 204-05 and fig. 12) showsthat this featurewas still currentin the fourteenthcentury. 87 E.g. at Ashfir,Tepe IIisdr, Ctesiphonand the Tari KhAnaat Dqmghdn (0. Reuther, Survey,pp. 422-3). Cf. also the Atashkfih of Nimwar, though this is of stone (A. Godard, " Les monumentsdu feu", Athar-ie Irn III/x (x938), fig. I4); and, for a medievalIslamic example, the engaged columnson the faqadeof the so-calledHarfiniyaat Tifs (Survey, 380). p. 38 For Niriz, see A. Godard," Le Masdiid-6Djum'a de Niriz ", Irdn Athdr-e' I/I (1936), fig. I112; for 'Ala, see A. M. Hutt, " Recent Discoveries in Iran, 1969-70; A Major Islamic Monument", IranIX (1971), p. 16o, pl. Va.

39 Not ten, as stated by E. Schroeder (Survey,p. Io042) and A. U. Pope (PersianArchitecture (London 1965), p. 144). 40 Representative monuments are the Mazdr Shir Kabir; the mausoleum of Sarakhs-BibS; the mausoleum of Abu Sa'id; two mausolea at Mashhad-i MisriyAn; and mausoleum no. I at G6k Gumbez (Pugachenkova, op. cit., pp. 169, 277, 279, 298 and 372-4 respectively). 41 The externalized squinch zone appears on both the Saljfiq dome chambers of the IsfahSn Jimi', while the gallery is used at the " Tomb of the Sqmdnids ", the tomb of Sanjir and possibly also on the Sangbast tomb. 42 The Fihrist (loc. cit.) mentions repeated restorations.

Pl. Ia. " Pir " mausoleum, Tdkistan: exterior, north side.

Pl. Ib. Exterior, north and west sides.

Pl. Ha. Decorated brickpanel on south side.

Pl. IIb. Detail of stalactite panel over entrance.

fafade. Pl. IIc. Detail of entrance

Pl. IIa. Stalactite panel over entrance.

P1. HIb. Detail of stalactite panel over entrance.

Pl. IVa. Detail of joint plugs over entrance.

P1. IVb. Interior: zone of transition, east wall.

Pl. Va. Interior: dome.

P1. Vb. Interior: zone of transition, looking west.

Pl. VIb. Interior: detail of squinch on E.N.E. side.

Pl. VIa. Interior: zone of transition, north side.

Pl. VIIa. Interior: detail of stucco work on east wall.

Pl. VIIb. IntErior:detailof stuccoworkon west wall.

zi

&

~s~er

a,

SALJUQ

MONUMENTS

IN

IRAN.

II

51

faqade has remained close to its original state. Even there, traces of stucco in the topmost panel show that it once bore an inscription or decoration applied on to a brick ground.43 The most recent restorations are easily identifiable because a brick of lighter colour has been used and because the mortar, which is grey instead of buff, is more lavishly applied. Dating controls.The discussion so far has been founded on the assumption that the Takistdn mausoleum is of Saljtiq date. The evidence for this is not incontrovertible. Few elements of the fagade can be dated even approximately. An example is the serrated moulding below the cornice; dogtooth patterns of this kind are known at least as late as the fifteenth century, when decorative brickwork was in full decline.44 Similarly, the decorative brick panel on the south side has a pattern which was popular for so long that no conclusions can be drawn from its use. Joint plugs seem at first sight to be another example (Fig. 3). The X-shaped plug, with a central disc, found in the central stalactite panel is

Fig. 3. Joint plugsfrom the interior the mausoleum. of

extremely common; it occurs all over Iran, though the specimens in the Masjid-i Jdmi' at Qazvin and the Haidariya are the most relevant in this context. But the very fact that joint plugs are used at Ttkistdn is of some value as a dating control, for to my knowledge the earliest datable joint plug occurs in the south dome of the Isfahdn Jimi' (c. 473/Io8o-8i).45 Coincidentally, it is the same plug as that ante quemfor the mausoleum;46 a on the Tdkistin facade. This provides an approximate terminus is terminus quem more difficult to establish. Other typically Saljfiq elements are the engaged columns post and their decoration, the flat stalactite composition, the trefoil niches and the form of the trefoil itself. The last two features, in particular, suggest a date in the late eleventh century.47 Having said that, one should strike a note of caution; studies of Islamic ornament are still not sufficiently advanced for style

4 The size of the panel is I - 55 m. by 0. 26 m. Bibi Sakina at Bdbolsar. 4 The joint plugs at the so-called tomb of Shaikh Shibli at Demavend (Stronach, op. cit., fig. 3) may also be of this date. Earlier eleventh century buildings use finger-impressed joints only. 46 It is quite possible, in fact even likely, that joint plugs were used earlier than c. i o8o in Iran. But enough eleventh century buildings have survived to permit the assumption that joint
,4 E.g. at the Imimzqda

plugs were a development of the later eleventh century, and that those in the Isfahan Jami' were therefore among the first examples. 47 Although trilobed arches are found in Iran throughout the medieval period, the type found at TAkistan, where the top lobe is a tiny keel-shaped arch, does not seem to occur after c. I Ioo. The latest example I know occurs in the Sujds Jdmi', which Wilber (op. cit., p. 118) places " close to the year A.D. I IOO ", but which may well date from later in the century.

52

JOURNAL

OF

PERSIAN

STUDIES

to be a truly reliable dating control.48 Moreover, it is well known that medieval Iranian craftsmen were content to repeat established decorative schemes for long periods, varying them only slightly. These two factors suggest that a relatively secure chronology would depend on an accumulation of small details rather than on one or two apparently significant variations. In the present case the combination of elements seems to justify a date close to 493/ I oo for the exterior of the mausoleum, but a dating later in the twelfth century could also be defended. Interior. The interior threshold is I 2 m. below modern ground level.49 The floor is of brick. One is immediately struck by the lack of brick decoration, though this was the principal feature of the entrance fagade. The implications of this omission will be considered later. Below the zone of transition, the plan is of the simplest. On the east, south and west sides a large pointed arch is recessed into the wall. An elegant plaster moulding outlines the profile of each arch,50 at a distance of 23 cm. from the intrados of the arch; the extrados of the arch is also slightly recessed, to a depth of 5 cm. The deeper recession, which varies from 33 to 36 cm., ends at the base of each arch, thus creating a ledge at I *2 m. above the floor.51 As far as I .9 m. above the floor the brickwork, in commond bond, remains exposed. From there until a little above the springing of each arch (the level varies) a thin coat of whitish plaster covers the wall.52 A wide band, which must once have contained an inscription but is now empty, encircles the interior at the base of the zone of transition. The dome, which at window height is three bricks thick, is undecorated; once again, changes in brick colour betray the repairs (P1. Va).58 Deep cracks in its inner surfaces show that conservation measures are urgently needed. At least three scaffolding holes are visible in the dome, and the lower blank arch on the south side is also flanked by holes placed just above the brick base.54 Two windows, facing south and west, let in the only light. The one which faces south is larger and rests on the collar of the dome, while the other is placed about a metre higher. At the base of the zone of transition on the south side is a blocked hole. It is impossible to say whether this was originally a window or a scaffolding hole. Zone of transition. The zone of transition contains deep cone squinches of the traditional Sasanian type, each flanked by an arch of identical profile flush with the wall (P1. Vb). So far, the pattern is one familiar from many Saljiiq buildings-a squinch and a blank arch alternate.55 But just above these arches occurs a significant departure from the norm-small keel-shaped arches fill the spandrels which separate each squinch from its neighbouring arch. They number eight in all. By virtue of their depth and their position these arches function as little squinches, replicas of the bigger ones but for their more angular profile (P1. VIb).56 The salient apex of each abuts on the base of the dome, which is corbelled out over them. These little squinches are merely one of several devices used in the eleventh and twelfth centuries to fill an otherwise obtrusive gap in the spandrels of the squinches and their neighbouring arches. At the Haidarlya relieving arches are used. At Ardistan a double row of squinches, one on top of the other, solved the problem in another way. At the Isfahan J mi', in the two dome chambers, a similar solution is employed though the upper arches are much smaller than the lower; the same is
48 More than thirty years ago M. B. Smith made a similar 53 The section (Fig. I) gives the mausoleum a dome thinner at

comment (" Imam Zade Karrar at Buzfin, a Dated Seldjfik Ruin ", A.M.I. VII (1935), p. 72) to the effect that the similarity between the stucco at Buzfin and that of Sasanian and Abbasid buildings greatly increased the difficulty of dating early medieval Iranian buildings from their ornament alone. Had there been no inscription at Buzfin, there is little doubt that many would have dated its stucco two centuries earlier. 49 Cf. Kharraqdn I (Stronach and Cuyler Young, Jr., op. cit., p. 7) and the examples cited by Wilber (op. cit., p. 4I). Accurate measurements of the difference in levels are difficult to make in the case of Takistin, because the ground slopes down towards the entrance from some distance before the threshold. 50 The plaster of the outer moulding is 4 cm. thick. 5xThese inconsistencies are another sign of the poor craftsmanship noted on the exterior. 52 The plaster coating is cm. thick; it reaches to a height of 2"5 1.62 m.

the top than at the base. Other thicknesses could equally well have been drawn. It might be possible to determine the true profile with the aid of a theodolite, but the practical problems in so small a building would be formidable. Shells of uniform thickness, as well as ones thicker or thinner at the peak than at the base, are all recorded in medieval Persian architecture (Wilber, op. cit., pp. 66-7). Since the upper half of the dome is a modern reconstruction it is now impossible to determine the original profile. 4 These lower scaffolding holes are 13 X 7 cm. and 15 cm. square; one of them is I -14 m. deep, an extraordinary depth for such holes (Wilber, op. cit., p. 54, says that scaffolding holes in Mongol buildings are generally c. 40 cm. deep). 15 Cf. the dome chambers of Ardistdn, Zavdra and Gulpdyagdn. 56 Keel-shaped arch heads are, as already noted, a recurrent feature on the fagade.

SALJUJQ

MONUMENTS

IN

IRAN.

II

53

true of the Gulpdyagdn Jdmi'. At Khwaja Sa'd, Isfahdn, the dating of which is disputed, the upper lobe of the trilobed squinch is made to reach right to the collar of the dome. No close parallel to the unobtrusive device at Takistdn comes to mind, though it would be surprising if none existed.57 The base of the dome is flush with the apices of the large arches, so the device of breaking up the octagon into a sixteen- or thirty-two-sided figure is not used.568 Each of the major arches and squinches is framed by a band of bricks which projects several centimetres from the wall. The bases of these bands, set at an angle to the wall and thus of triangular ground plan, are located above the upper moulding of the empty inscription band. It is possible that the bases once rested on corbelled wooden shelves, like those of the nearby HIaidariya and Qazvin Jdmi'. There is no trace of a mihrdb. Incisedplaster decoration. The decoration of the interior is confined exclusively to the zone of transition, but its richness dominates the chamber. The brickwork of the whole area is covered by a thin coat of plaster, though in the secondary squinches it must have been applied fairly thickly to achieve the necessary concave curve. The plaster now has an attractive honey colour. Sometimes, as in the spandrels and the small arches, imitation bricks are cut into the plaster, usually with joint plugs.59 The incised plaster patterns inside the four main blank arches do not, however, counterfeit brickwork; two form interlocking lozenges, one a series of stepped squares which fit inside each other, and one-the finest-a combination of the two.60 A joint plug marks each right-angle turn of these patterns. In the centre of each pattern is a square, cross or rectangle of carved stucco (P1. VIIb).61 Several of them, such as the swastika or nine-star designs, are of greater complexity and size than most joint plugs so far recorded (P1. VIIa).62 In the zone of transition as a whole, the most common type of joint plug is X-shaped with a central disc, i.e. the same type as is favoured on the exterior of the building. Applied stucco. Perhaps the most striking feature of the decoration is the use of applied stucco, richly carved to form medallions, leaf clusters and pendants. These are placed at points of key architectural importance: at the crown of the intrados of the large blank arches, at the apex of the squinches, and at the crown and springing of the small keel-shaped arches (Pls. IVb and VI).63 Most of the larger pieces have survived in good condition, but of the twenty-four small motifs which once adorned the eight secondary squinches eleven have disappeared and five are in poor condition. Some of the motifs display foliated compositions which are part of the established repertory of twelfth century stucco ornament. The bold relief, the clotted forms and the preference for a few large elements instead of

s"The nearestparallelis the still insecurelydated Sultan Sa'dat ensemble at Tirmidh (B. P. Denike, Arkhitekturniy Ornament Srednei Aziy (Moscow and Leningrad 1939), fig. 5). This has the same small archesbetween the largerarches,but they are placed about a metre higher, i.e. at the base of the dome. In Iran, the nearestrelated form may be that in the zone of transition inside the Seh Gunbad at Ritd;'iya (Godard, "Notes compl6mentaires fig. II ). In this tower the ", betweeneach squinchand blankarch are filled by an spandrels arch whose crown is corbelledout to support two half arches on which rests the moulding at the base of the dome. This device, like that at Tdkistdn,can be regardedas an embryonic stalactitesystem. The same is true of the Gunbad-i Seh Tan at Amul, where the upper angles of the octagon are bridged by a blank arch whose corbelledhead touchesthe base of the dome. This arch is flanked,well above springinglevel, by two tiny replicasof itself, their apices also touching the base of the dome. 58 The twenty-four-sided figure is very rare. It occurs in the Masjid-i Jdmi' at Ridd'iya (A. Godard, " Les Coupoles", Irdn Athdr-i IV/2 (1949), P. 273, n. 2, and fig. 229). 59It is interestingto note the occurrenceof this feature so long
before the Mongol period, the time of its greatest popularity. 60 There are a good many examplesof this kind of decorationin

medieval Iran. The Haidariya parallel is the most relevant in this context. Anotherexampleis in the mausoleumof Salar

Khalil (A. S. Melikian Chirvani, " Remarques prdliminaires sur un mausol6e ghazn6vide ", Arts Asiatiques XVIII (1968), figs. 2, 7 and 8). It also occurs in Soviet Central Asia, e.g. on the wooden shelter over the grave of one Khazret (sic) Posho in Chorku, Tajikstan (V. L. Voronina, " Reznoye Derevo Chorku ", in Arkhitekturnoe Nasledstvo, no. 6 (Moscow 1967), p. 177, fig. 2). Cf. also the floors at Khirbat al-Mafjar (R. W. Hamilton, Khirbatal Mafjar, An Arabian Mansion in the Jordan Valley (Oxford 1959), pls. LXXXIII/viii and XCVIIId). 61 Cf. the similar design at the Demavend Jami' (Smith, " Material for a corpus ... I ", fig. 27). 62 The swastika pattern recurs at Qurva, Sujas and the north zwan of the Qazvin JRmi'; the nine-star pattern is found in a squinch of the tomb of 'All ibn Ja'far at Qumm, and on a stucco panel in the Mongol part of the shrine at Turbat-i Shaikh Jam (A. U. Pope, Persian Architecture, pls. XXI and fig. 191 respectively). 63 The best parallel to this is at Qurva (see below); but the Mil-i KIshmar also has a similar feature. At the Imdmzdda Kishlaq, near Vardmin, a slightly different method is used to accentuate the arch heads in the hexadecagon. The entire arch is set in a rectangular frame, and all the spandrel area is carved with foliate patterns. The element at the crown of the arch is painted white and thus stands out against the dark brown ground used for the surrounding area.

54

JOURNAL

OF

PERSIAN

STUDIES

many small ones are all typical of the period.64 Parallels in the Qazvin Jdmi' make an instructive study.65 The surface of the forms is covered with the complex pierced and hatched designs which were used in stucco-workthroughout the Saljiq and Mongol periods.66 The stucco is of greyish white colour. No trace of applied colour remains on it. Related buildings.The key to the date of this decoration is best sought in the Saljfiq buildings of the area. Although the links cannot be denied, there are difficulties in using these buildings for Qazvin purposes of comparison. The fact that the Qazvin J~mi' is dated-it was erected from 507/1113 onwards-gives it special importance in a discussion which involves undated or problematically dated buildings.67 Its great size, however, and its function as the main congregational mosque of a large city, makes it likely that the best craftsmen worked there. This renders it a rather unsatisfactory dating control for much smaller buildings in the outlying villages. To a lesser extent the same applies to the Haidariya, which in the past has been dated in the early twelfth century, on the basis of less evidence than is now available. Happily there exists, at only about 30 km. from Takistan, a building much more closely related to it than either of the two structures at Qazvin. This is the Masjid-i Jdmi' at Qurva. Most of its decorato those at Takistdn, so much so that one is tempted to believe that they are the work of the same hand (Pl. VIIIa). It is rare in a country as varied in architectural expression as is Iran for two schemes of decoration to resemble each other as much as this. There can be little doubt that one copies elements of the other. But which came first? And how many years separate them? These questions are too difficult to answer at the present time. One may not go far wrong, however, if one dates the Takistan The interior decoration of Saljiq domes was executed mainly in brick until well into the twelfth century.69 In the very few cases where stucco is the principal means of ornament, as in the Btiyid Davazda Imim at Yazd, the technique bears little relationship to the work at Tdkistmn. A dating of c. twelfth century architecture in Khurisin: he found that it was precisely the first half of the twelfth century which saw the gradual replacement of brick by stucco decoration.70 Conclusion. seems worthwhile to comment briefly on the place of the Tdkistdn mausoleum in It architecture. Mausolea of the domed square type are rare in Iran at this period. The surviving Saljfiq examples are concentrated in Jib~l and Khurdsmn; the location of the Thkistan mausoleum in JibWlis thereforeespecially noteworthy. Most of the Iranian examples are a good deal bigger than the Takistan tomb and also differ from it in details of plan, structure and decoration. Some of the comparative material therefore has to be drawn from mausolea of the domed square type from outside Iran, and in particular from Central Asia. But it would be unwise to lay much emphasis on the similarities which undoubtedly exist. There is not enough evidence to prove direct influence from Central Asia, let alone to suggest that an architect from Transoxiana might have erected the building. At least as early as the tenth and eleventh centuries Muslim architects were experimenting to discover the most suitable type of building to serve as a mausoleum, and similar solutions in widely scattered areas call for no special explanation.71 The form of the Tdkistan tomb, though unusual for Iran at that time, can well be
is also quite compatible with the conclusions drawn by Godard on the basis of his work on 570/i 174-75 interior to about 570/I 174-75. Nor is it simply the resemblance to Qurva which suggests such a dating. tion dates from 575/1
79.68

The stucco medallions there are astonishingly similar in design and location

In the Survey,p. 996, n. 3, Pope substituted 507/1113 for the Cf. the similar, but more swollen, shape over the entrance arch at the Chehel Dukhtarin tomb tower, DAmghAn (F. Ripertoirereading of 509/11 6, but did not give his reasons for Sarre, DenkmalerpersischerBaukunst, Textband (Berlin i9io), doing so. 68 On this Abb. 156). date, see the article cited in n. I, above. 69 65See A. U. Pope, " Notes on the Stucco Ornament in the E.g. the Ijaidariya (Survey,pl. 314) and, as late as 1140-41, the Masjid-i Gunbad at Sangin-i PA'in in Khurisan (A. Sanctuary of the Masjid-i Jami', Qazvin ", Bulletin of the American Institutefor Persian Art and ArchaeologyIV (1936), Godard, " Khorisdn ", Athdr-i Irdn IV/I (1949), fig. 6). 70 Godard, op. cit., pp. 59-60. This is not to say that conclusions pp. 209-16. 66 See the representative designs (not all from Saljfiq buildings) drawn with respect to Khurisdn are equally relevant for illustrated in Survey,figs. 473-4. north-west Iran; but the broad outlines of stylistic development do have a wider frame of reference than Khurdsdn alone. 67 t. Combe, J. Sauvaget, G. Wiet, et al., Ripertoirechronologique d'epigraphiearabe VIII (Cairo 1937), nos. 2960 and 2965-7.
64

SALJUQ

MONUMENTS

IN

IRAN.

II

55

explained as a chance survival of a once widespread building type,72 as a representative of an old and at that time unfashionable tradition, or as a deliberate return to an archaic tradition in preference to the then current fashion for tomb towers. It seems most profitable to relate the Takistan mausoleum to the Saljfiq style of north-west Iran. The Saljfiq buildings in and around Qazvin show that this was a major centre of architectural activity. The decoration at Takistin is also firmly within this local school. The dating controls provided by these buildings; the different systems of decoration used for the interior and exterior; and the traces of at least two building campaigns all combine to suggest that the mausoleum was erected, and its external decoration completed, around 493/I 100, while the stucco ornament of the interior could be dated c. 570/1174-75-

71

The same period saw the evolution not only of the tomb tower but also of the four-fwdn plan in its classical Islamic form, and of the distinctive Iranian minaret.

72

The preponderance of tomb towers among surviving Saljriq and Mongol mausolea weakens this theory.

WHO WERE THE CHIHILGANI,THE FORTY SLAVES OF SULTAN SHAMS AL-DIN ILTUTMISH OF DELHI ? By Gavin Hambly
The three decades which separate the death of Sultan Shams al-Din Iltutmish2 in 633/1236 from the seizure of the throne by Ghiyath al-Din Balban in 664/1266 form a period of quite exceptional political instability in the early history of the Delhi Sultanate, especially during the years prior to the accession of Ndsir al-Din Mahmfid in 644/1246. One feature of this period, emphasized by all the dominant role assumed by a group of prominent slave commanders known as the historians, isSult.n the chihilgdni, Forty Slaves of Sultan Shams al-Din Iltutmish, who throughout these thirty years appear to have exercised a more or less uninterrupted control over a succession of relatively feeble rulers who were, for the most part, mere puppets in their hands.3 were and precisely what role they fulfilled in this It would be useful to know who the chihilgdni Muslim state so recently established upon the very fringes of the Ddr al-Isldm,but unfortunstruggling ately the sources for the period say very little about them, although this has not prevented historians from giving free rein to their imagination when approaching the subject. One of the most cautious, Stanley Lane-Poole, wrote at a time when the study of the Delhi Sultanate was still in its infancy, but his view of the chihilgdni exemplifies the attitude of the older generation of historians of Muslim India which eventually found its way into TheCambridge Historyof India.4 In Lane-Poole's view, the chihilgdni Praetorians: were, quite simply, overweaning
The slave system had grown stronger by the successful careers of Aybek and Altamish. The latter had formed

a corpsof Turkishmamliksknownas " the Forty ", and these men, profitingby the removalof the master's hand, sharedamong themselvesthe wealth and power of the kingdom. The free-bornmen who had served Altamishwith great abilityin variousofficeswere removed,and all controlwas in the hands of " the Forty".5 Over thirty years later the young scholar, Ibn Hasan, reacted strongly against this traditional oversimplified assumption, and offered a most interesting alternative theory-although one for which, unfortunately, direct evidence is lacking: .... Shams-ud-dinalso created a body of loyal supportersto the throne and kept it at the centre. It was
intended as a check upon the powers and ambitions of the military chiefs, who divided the resources of the

countryand the army among themselves. This body of loyalistsis known as " The Forty ".
1 An expanded version of this paper was read at a meeting of the
4

Pacific Coast Branch of the American Historical Association at San Diego, California, in August 1969. 2 Controversy on the subject of the form of this Sultan's name has surely been finally laid to rest by Simon Digby in his paper, " Iletmish or Iltutmish? A Reconsideration of the Name of the Dehli Sultan ", in Iran VIII (1970), pp. 57-64. 3 The order of succession of the Shamsi Sultans was:
(I)

The CambridgeHistory of India, edited by Sir Wolseley Haig (Cambridge, 1928), III, p. 61. I. H. Qureshi maintains substantially the same view in The Cambridge History of Islam, edited by P. M. Holt, Ann K. S. Lambton and Bernard Lewis (Cambridge, 1970), II, pp. 6-7. 6 S. Lane-Poole, MediaevalIndia under Mohammedan Rule (London, 1903), p. 76.

Shams al-Din Iltutmish (607-633/121 x0-1236)

(2) Rukn al-Din Firfz (633/1236) 'Ald' al-Din Mas'fd (639-644/1242-1246)

(3) Raiiyya (633-637/1236-1240)

(4) Mu'izz al-Din Bahrdm (637-639/1240-1242)

(6) Nasir al-Din Mahmad (644-664/1246-1266)

(5)

57

58

JOURNAL

OF PERSIAN

STUDIES

It was a very useful device, and it worked very well under the Slave dynasty. It gave full support to the dynasty, and, in spite of the weakness of Shams-ud-din's successors, the throne remained in his family. The changes were made in the interest of the kingdom and the throne, and weak and worthless kings were always replaced by stronger and more capable ones. The experiment limited the ambitions of the Muslim element of the kingdom to a very small group, and a tradition of dynastic rule was established. But the success of the device and the experiment depended upon the unity of the party at court, together with the provincial military chiefs. This became impossible, and the domination of the Forty was later on resented and opposed by other sections of the ruling class.6 More recently, A. B. M. Habibullah has given an added dimension to this theory although, again, without any data to support it. A large progeny was not favourable to a king's interests, while a number of tried and efficient slaves having no other interest than to serve the master's family, was a sure asset. Iltutmish had no illusions about the capacity of his sons and the only way to counteract the opposite tendency seemed to lie in organizing his personal retainers into a party who would stand by his family and thereby uphold his absolutist monarchy. Like the Muizzi and Qutbi slaves, the Shamsi slaves were thus allowed to form themselves into a political group which, after his death, received the collective name of the " Forty ". By absorbing or destroying the adherents of former kings they were enabled to reign supreme after his death. .... Thus there came into being a curious phenomenon, a party of bondsmen pledged to support the power of their master's family who considered the state a vast household in which outsiders could have no place. The Sultanate was converted into a kind of household polity.' K. A. Nizami has adopted an altogether different position. He apparently sees the chihilgdni as a cohesive group of slave commanders within a larger ruling l1ite, bound together by mutual concern lest their monopoly of office be threatened either, on the one hand, by the intrusion of free-born Central Asian immigrants fleeing into India in the wake of the Mongol holocausts, or, on the other, by the emergence of an indigenous Muslim elite born and bred in the sub-continent. " It was probably the growing assertion of the Indian elements in the body politic ", he suggests, " which led the Turkish maliks to organize themselves into a corporate body known as Chahlganian".a We have here three distinct theories regarding the chihilgdni: that they were a band of Turkish slave-commanders who formed a selfish oligarchy of which the raison d'etre was the exclusion of outsiders from those offices in the state which they claimed as their monopoly; that they were a cohesive group deliberately brought into being by Shams al-Din Iltutmish to impose upon the infant Sultanate a specific institutional character; that they were a tightly-knit faction bound together by their need to protect the monopoly of office hitherto enjoyed by the Sultan's Turkish slave commanders from interlopers (e.g. free-born refugees from Central Asia and Iran, and the expanding class of nativeborn Muslims). In fact, however, all three interpretations, although neither mutually exclusive nor lacking in general plausibility, rest on a very thin foundation of fact. Jiizjani, the chronicler of both the Shansabanid dynasty of Ghir and of the so-called " Slave Kings " of Delhi, at whose court he wrote the Tabaqdt-i JVisiri, makes no mention of the chihilgdni. It seems probable, therefore, that the earliest known reference to them occurs in the Ta'rikh-i Firiz Shchi written by Ziya' al-Din Barani and completed sometime between the accession of Sultan Firfiz Shah Tughluq in 752/1351 and his own death around 758/1357, or, in other words, almost ninety years after the accession to the throne of Ghiyvth al-Din Balban. It is from Barani, or from later writers of the Mughul period who drew upon Barani as a source, that modern historians have derived what little is known regarding the chihilgani. That Jtizjani, their contemporary and an intimate with at least one
6

Ibn Hasan, The CentralStructure the Mughal Empire (Oxford, of I936), PP. 44-45. 'A. B. M. Habibullah, The Foundationsof Muslim Rule in India (Allahabad, i961), pp. 345-346. The assumption that the members of a ruler's family were, on balance, less to be trusted than his personal slaves was clearly stated by the Saljiiq vazfr, Nizim al-Mulk, in the eleventh century: One obedient slave is better than three hundred sons;

For the latter desire their father's death, the former long life for his master. Siyar al-Mulak (Siydsat-ndma), edited by H. Darke (Tehran, or 1962), p. I50, tr. idem, The Book of Government, Rulesfor Kings (London, g96o), p. 121. K. A. Nizami, SomeAspectsof Religion and Politics in India during the 13th Century(Aligarh, I961), p. 127.

WHO

WERE

THE

CHIHILGiNI?

59

of them (Balban), should have avoided using the term is perhaps not altogether surprising. It may well have been in origin a popular nickname, bandied about the streets of the capital but not on the lips of those who associated with the ruling elite, especially if, from the outset, it was an expression of opprobrium, as it may well have been. It is probably as an expressionof opprobrium that Barani uses it, conscious of its effect upon his own contemporaries who, like himself, would have derived much of their knowledge of the events of the preceding century from oral traditions preserved from generation to generation among the old Delhi families. Dr. Hardy's warning regarding the Ta'rikh-i Firaz Shdhi,that it " is not an annal or chronicle; it is an important example of didactic historiography in Islam ", needs to be taken constantly into consideration.9 Whether writing of his own lifetime or of an earlier period, Barani was outraged by what he regarded as the scandal of rulers entrusting offices of responsibility to low-born upstarts, a category elastic enough to include both infidel Hindus and perhaps all but a handful of Muslim ghuldms. Not a great deal is known about the history of Barani's family but they had certainly been people of some prominence in Delhi at the turn of the century: his maternal grandfather had been vakil-iddrto one of Balban's most trusted amirsand, later, shahnaof Lakhnawti in Bengal; his father had been nd'ibto the second son of Sultan Jalil al-Din Firiz Khalji and later was in charge at Baran (now Bulandshahr) in the Dtab; and an uncle had been kotwdlof Delhi itself during the reign of Sultan 'Ala' al-Din Muhammad Khalji.lo Presumably there was no background of slavery in the family and a free-born Central is Asian emigre' origin may be supposed. His Fatwd-yi Jahdnddri strewn with denunciations of the kind of of obscure origin favoured by the late Sultan, Muhammad ibn Tughluq, and no doubt he people in viewed with jaundice Sultan Firfz Shah Tughluq's employment of ghuldms the administration on a
scale perhaps exceeding anything known since before the rule of the Khaljis. With this background in century earlier assumes a deeper significance, for in castigating the misrule of the chihilgdnihe surely had in mind-as would also his readers-the implications for their own time of this tale of oppression by all-powerful ghuldms. Barani makes six separate references to the chihilgdnd,of which two refer specifically to individual ghuldms. Of the remaining four, the first comes at the beginning of the section dealing with the reign of Ghiyath al-Din Balban where he states that the Sultan himself had been

mind, Barani's sweeping condemnation of the abuse of power and office by the slave-commanders of a

... one of the Shamsi slaves [e.g., a slave belonging to Sultan Shams al-Din Iltutmish] and among the forty Turkish slaves freed. .xx ... Further on, referring to the decay of the Sultanate following the death of Shams al-Din Iltutmish, he describes how under young and inexperienced rulers, devoted to self-indulgence, power and wealth had accumulated in the hands of the Shamsi slaves who had been made khdns, and during the reign of those Sultans " those Turkish slaves who were called the chihilgdni became all-powerful in their control of the government " at the expense of the free-born amirs and maliks.12 Following the death of Sultan Shams al-Din Iltutmish the chihilgdniwere successful in everything which they undertook.13 Barani also stresses their unwillingness to submit to any leader, even from among their own number, a fact which accounts for much of the hostility provoked by Balban's rise above the level of his peers. Barani writes: ... because the Shamsi slaves were all officers of the Household [the term khwdja tash refers to personal attendants of the Sultan] and because all forty were trained [together] in the same way, not one would take orders from or obey another.'4
9 P. Hardy, art. " Barani ", Encyclopaediaof Islam, new edn. (Leiden, I960), II, p. Io36. Barani's place in the historiographical tradition of Muslim India is treated at length in Peter Hardy's Historians of Medieval India (London, 1960). K. A. Nizami adopts a rather different point of view in " Ziyaud-Din Barani " in Historians of Medieval India, edited by M.
Hasan (Meerut,
10 11

of Encyclopaedia Islam, loc. cit. 2iyv' al-Din Barani, Ta'rikh-i Firaz ShUhI, edited by Sayyid (Calcutta, 1862), p. 25: Balban ki banda-yi az A.hmad Khan

1969), pp. 37-52.

bandagdn-ishamsi bad va dar maydn-ibandagdn-iturk-i chihilgdnf dzdd shuda. 12 Ibid., p. 26: Bandagdn-i turk fshdn-rd chihilgdni miguftand bar umar-i mamlakatf mustaulf shudand va bd quwwat va shaukat gashtand. 13 Shams al-Din bandagdn-i Ibid., p. 27: Va ba'd-i naql-i Sul.tdn turk-i chihilgdnia kdmydb gashtand. 14 Ibid., shamsikhwdjatdsh bada va har p. 28: Va az dnki bandagdn-i chihil bandaba-yakkarratbuzurgshudandyakimardigarl-rdsarfura va nakardi. naydvardi i.ta'at

60

JOURNAL

OF PERSIAN

STUDIES

This is, in effect, virtually all the information available to us and it does not amount to very much. were a group of forty Turkish ghuldms(Barani uses the term banda)who had To sum up, the chihilgdni to Sultan Shams al-Din Iltutmish, who had all served together in the Sultan's originally belonged entourage and who all attained the rank of khdnso that each regarded himself as being the equal of the remaining thirty-nine.15 It is significant that Barani makes mention of the bandagin-iShamsiin general as well as of the in differed from chihilgdni particular, and it is perhaps useful to enquire in what respect the chihilgdni the other Shamsi slaves, some of whom, in the years following the death of Sultan Shams al-Din Iltutmish, also attained positions of great authority. The only clue appears to be that phrase which for Barani uses when he mentions the chihilgdni the first time, including Balban " among the forty India, Iran and Central Asia the contemporary sources have little or nothing to say with regard to the or question of manumission and whether, as slave commanders were promoted to the rank of .hejib amir,their formal manumission was more or less automatic. For most practical purposes, of course, the question of legal status was irrelevant if a ghuldmcommanded the loyalty of a sufficient number of troops or had at his disposal the resources of a rich iq.td'. Precisely what slave status meant under such circumstances is still far from clear. Apparently a slave could own another slave (Shams al-Din Iltutmish was the slave of Qutb al-Din Aybak who was in turn the slave of Sultan Mu'izz al-Din Muhammad ibn Sam of Ghiir) and a master sometimes gave one of his daughters to a favourite slave (as Qutb al-Din Aybak did with Shams al-Din Iltutmish). Perhaps the time when servile status mattered most to a high-ranking ghuldmwas at his death when there arose the question of how his property should be disposed of. Jtizjani, for example, never makes any reference to Ghiyath al-Din Balban's manumission although he describes his career in great detail from the time when he was first sold as a slave in Bagdad down to the time when he became the father-in-law of Sultan Nasir al-Din Mahmfid and defacto master of Delhi. Apparently formal manumission was not a matter of great concern so far as high-ranking slave-commanders were concerned. When, therefore, Barani describes the chihilginias freed Shamsi slaves he is surely indicating what set them apart from other slaves who belonged to Sultan Shams al-Din Iltutmish. It is possible-in my view, most likely-that all forty were manumitted at the same time, perhaps on some famous occasion of state long remembered by the inhabitants of Delhi, and thereby accounting for the fact that the sobriquet was still in circulation during Barani's lifetime.17 If so, the event may have occurred during the closing years of the reign of Sultan Shams al-Din Iltutmish or, conceivably, may have marked the accession to the throne of either Rukn al-Din Firfiz or Ratiyya in 633/1236. Since, however, the manumission of slaves was regarded as an act of piety, frequently authorized by a master on his deathbed, it is not unreasonable to suppose that this celebrated band may well have obtained their freedom upon the death of Sultan Shams On his own admission, Barani derived much of his knowledge of events in the preceding century from hearsay of his father and grandfather,'8and we therefore cannot be quite certain that he is correct
15

Turkish slaves freed. ....

"16

Throughout the entire period when ghuldms were extensively employed in

al-Din Iltutmish in 633/1236.

It should be noted that the historians of the Mughul period of the Rise of the MahomedanPower in India, London, 1829, 4 vols., I, p. 249. appear to have been content merely to elaborate upon Barani's In the reign of Shums-ood-Deen Altmish, Forty of his original statements. Thus Khwaija Nizam al-Din Ah.mad, Toorky slaves, who were in great favour, entered into a describing Balban's accession to the throne in the Tabaqdt-i solemn covenant to support each other, and on the King's Akbarf,edited by B. De (Calcutta, 1913), p. 78, simply states: death to divide the empire among themselves. Jealousies Sultan Ghiydth al-Din was the slave of Sultan Shams aland dissensions afterwards arose among them, and preDin, one of the band of slaves known as the chihilgdnf. vented this project from being carried into execution. Sultan Shams al-Din had forty Turkish slaves each one of whom reached the rank of amirand this band was spoken of 16 Barani, op. cit., p. 25. 17 A famous occasion upon which one thousand slaves were as the chihilgdnf. manumitted was the feast given by the 'Abbisid Caliph alThis is clearly taken straight from the Ta'rikh-i Ffriz Shdhi, Mutawakkil (232-247/847-861) to celebrate the circumcision which Khwaja Nizdm al-Din Ahmad lists among his sources, of his son al-Mu'tazz. See C. E. Bosworth, The Book of but the same is true of 'Abd al-Qadir Badd'fini in his Curious and Entertaining Information. The Latd'if al-ma'drif of Muntakhabal-Tawdrikh. At the beginning of the seventeenth century Firishta in his Gulshan-i Ibrdhifmexpanded this brief Tha'dlibf (Edinburgh, 1968), pp. Ioo-loI. statement, at the same time taking considerable licence with his 18 Barani, op. cit., p. 25. sources. The translation here used is that of J. Briggs, History

WHO

WERE

THE

CHIHILGXNI?

61

when he asserts that the chihilgdni were a group originally composed of forty slaves, and that the term was not used merely in the Biblical sense of meaning a large number. But assuming for the sake of argument that there was in existence around 633/1236 a band of forty recently-manumitted slavecommanders who constituted a more or less cohesive group, it is surely unrealistic to suppose that a decade or so later the group was still numerically intact and was collectively stage-managing the palace-revolutions which, following the disappearance of Rukn al-Din Firriz and Raiiyya, resulted in Wolseley Haig in The Cambridge Historyof India presupposes that even after Balban's sucession the still chihilgdni existed as a cohesive group and that, indeed, most of the great nobles of the court were to be counted among their number.19 But on grounds of mathematical probability alone, this would seem were already high-ranking slave comhighly unlikely. If it be assumed that the original chihilgdni and amirs,around the time of the death of Sultan Shams al-Din Iltutmish, this premanders, .hdjibs supposes that they were already in their mid-thirties. Two centuries earlier the Saljfiq vazir, Nizim
al-Mulk (408-485/1oI8-IO92), had expressed the opinion in his S3ydsat-ndmathat thirty-five was a the elevation first of Mu'izz al-Din Bahrdm (637-639/1240-1242), then of 'Ala' al-Din Mas'iid and finally of Nisir al-Din Mahmaid (644-664/1246-1266) to the throne. Sir (639-644/1242-1246),

suitable age for promoting a ghuldm outstanding ability to the rank of amir,20 of and although it would be dangerous to apply without qualification Nizam al-Mulk's cursus honorum the training of ghuldms for in eleventh century Iran to the circumstancesof the thirteenth century Delhi Sultanate, the fact remains that the little we know about the early career of Balban and his contemporaries suggests that it was indeed around this age that they obtained independent commands and high office. In 664/1266, when Balban-by murder or by default of a candidate with a better claim-at last succeeded his son-in-law, he was already a sexagenarian, albeit an active one, but such other members of the original chihilgdni who were still alive cannot have been any younger. But in any case, given the probable expectation of life of a thirteenth century Turkish warlord, it would seem unrealistic to suppose that more than a handful of the original band of forty survived to witness the elevation of the most ruthless and most successful of their number to their former master's throne. In his account of Balban's reign Barani names only three amirs as belonging specifically to the chihilgdni-Balban himself; the Sultan's cousin, Shir Khan Sunqur, governor of Lahore and Bhatinda; and a certain Tamar Khan, who was granted some of Shir Khan Sunqur's after the latter's death. Shir Khan Sunqur is, of course, a comparatively well-known figure, but iq.td's is known of Tamar nothing Khan beyond this single fact, and in designating him one of the chihilgdniit is not impossible that Barani has confused him with the great Qipchaq amir, Malik Tamar Khan-i Qiran, a governor of Awadh who died in 644/1246. Apart from these three, no other amirsare described as belonging to the chihilgdni,although reference is made to former Shamsi slaves as well as to slaves belonging to Balban himself-all of whom, for no very obvious reason, Sir Wolseley Haig assigns to that select band. Had, for example, Barani,
on the basis of the information available to him, reckoned Shamsi slaves such as 'Adil Khan and said so. On the other hand, given the nature of the sources available to him, it is probable that while Barani may have heard a good deal concerning the chihilgdni collectively, he may only have ever known the names of a handful of them. Thus he mentions two amirs who, in my view, should probably be included among the chihilgdni: Arslan Khan Sanjar, the rebel governor of Bengal who died in 662/1264 and whose son, Tattr Arslan Khan, promptly submitted to Balban upon the latter's accession; and Balban's own brother, Malik Kishli Khan. However, neither of these are allocated a place among the Forty. Mention has already been made of the fact that never uses the term chihilgani at all. What Jiizjmni he does do, however, is to include in the Tabaqdt-i NJcsiri a section, approaching perhaps one-tenth of the whole work, which is a tadhkirat or collection of short biographical sketches of the twenty-five principal Shamsi maliks. All these twenty-five were Turks-Qipchiqi, only one Ilbari, etc.,-with
History of India, III, pp. 74-78. "9 The Cambridge
20

Tabar Khan, who were on terms of intimacy with Balban, among the chihilgdni, would have surely he

NizaIm al-Mulk, op. cit., p. I34, tr. p. Io7. See also C. E. Bosworth, The Ghaznavids(Edinburgh, 1963), pp. Io02-10.

62

JOURNAL

OF PERSIAN

STUDIES

possible exception, Malik Hindi Khan Mehtar-i Mubdrak, who may have been of Indian origin; all without exception had been slaves of Sultan Shams al-Din Iltutmish and had owed to him their initial rise to prominence; and all, in one way or another, had played a part of some consequence under one or more of the Shamsi Sultans. It is not clear upon what basis Juizj*niselected for his tadhkirat lives the of these particular maliks, but the choice cannot have been wholly arbitrary. Although he himself makes mention of gratitude to benefactors (among whom Balban was unquestionably one, and perhaps the principal one) it would probably be wrong to assume that the selection was made, at least wholly, as a means of returning past favours or seeking future preferment. At the time when Jfizjani was writing (in the early part of the twelve-sixties), eighteen out of the twenty-five malikswere already dead, although conceivably their families would have been honoured to have had their names so conspicuously recorded. One hypothesis worth considering is that Jiizjani selected these twenty-five maliksbecause whose deeds were still remembered in Barani's day but whose they had been the core of the chihilgdni, names, except for one or two, had long since been forgotten. Presumably there were others-the remainder of the original Forty-but some would have died or disappeared from the scene during the late twelve-thirties or early twelve-forties, leaving these twenty-five who were the ones who really counted. They certainly fit Barani's broad description of the chihilgdni:Turks, Shamsi slaves, holders of high offices and titles. It is interesting to note that of the three persons specifically named by Barani as belonging to the chihilgdni,two-Balban and Shir Khan Sunqur-appear in Jfizjani's tadhkirat. Tamar Khan, it is true, is missing but his omission may simply mean that he was a relatively insignificant figure. Moreover, as I have suggested above, Barani may have erred in including him among the at chihilgdni all. Juzjani's list certainly includes a number of the most prominent figures of the Shamsi period, a veritable Who's Who of the leading ghuldms,including Malik Tamar Khan-i Qiran, Malik Arslan Khan Sanjar and Malik Kishlfi Khan, already mentioned. If this hypothesis is correct and the then at twenty-five maliksof Jfizjani can be identified with the most prominent of Barani's chihilgdni, least a tentative advance will have been made in bringing this elusive group into clearer perspective.

EXCAVATIONS AT SIRAF Fifth Interim Report By David Whitehouse

The fifth season of excavations at SIrif took place between November and February I97o-71. During the season we extended the survey by planning the remains of buildings to the west of SIraf at completed our work at the Great Mosque, establishing beyond reasonable doubt the Sasanian date of the early enclosures. We excavated two buildings near the western defences (Site J), examined the defences themselves (Site L) and began to explore a palatial complex overlooking the bazaar (Site K). We investigated a ruined imdmzddeh (Site H) and conducted minor excavations at Sites C, D, E, F and G. The object of this diverse activity was to bring to a close the excavation of all sites except the palace. We hope to return to SIraf for a final season in 1972-73, during which we shall complete our work at the palace and on the finds stored in the expedition's house. We are indebted to H.E. the Minister of Culture, Mr. Mehrdad Pahlbod, and the Director General of the Archaeological Service, Mr. A. Pourmand, for permission to excavate at StIrf. The Deputy Director General, Mr. S. M. Khorramabadi, and the Director of Excavations, Dr. Naimi, gave us most valuable assistance. Mr. Enayat Amirli and Mr. Ghollam-Reza Massoomi accompanied us to SIrdf as Representatives of the Archaeological Service and we thank them for their help. The excavations at Siraf are sponsored by the British Institute of Persian Studies. During the season we received continuing support from the British Museum, the British Academy, the Russell Trust and a munificent anonymous Trust. The British Museum released Mr. Nicholas Lowick to join us for three weeks, while sabbatical leave from the Ashmolean Museum, Oxford, and study leave from the School of Oriental and African Studies, London University, enabled Mr. Robert Hamilton and Dr. Geza FehervArito spend most of the season at StIrf. We are deeply grateful for this generous support, without which we would have been unable to work on a scale sufficiently large to establish the Sasanian origin of SIr~f and elucidate so much of the Islamic city.2 The expedition staff was as follows: David Whitehouse (director), Peter Donaldson and Jan Roberts (site supervisors), Robert Hamilton and Geza Feh6rvari (site and survey assistants), Jonathan Erskine (surveyor), Jeanine Erskine, Imogen Robinson and Elliot Fine (assistant surveyors), Sarah Jennings, Ruth Whitehouse and Peter Farries (finds assistants), Patricia Cathcart and Eileen Flynn (conservators),Joseph Cloutman (photographer) and Graham Price (quartermaster). The team was augmented by three volunteers: Nicola Kisch (draftsman), Sydney Robinson (surveyor) and Warwick Ball (assistant surveyor and draftsman). Renata Holod-Tretiak and Andrew Williamson visited the site and gave us valuable help. The photographs published here were printed by Giles Sholl.
1 For a preliminary report, see Iran IX (197I), pp. 176-7. A summary of results obtained in the first four seasons appeared in World Archaeology no. 2 (1970), pp. 141-58. For more II, detailed accounts of previous seasons, see " Excavations at Sirdf: First Interim Report ", Iran VI (1968), pp. 1-22; " Excavations at Sir~f: Second Interim Report ", Iran VII " Excavations at Sirdf: Third Interim (1969), pp. 39-62; ", Iran VIII (1970), pp. I-18; and " Excavations at Report Sirdf: Fourth Interim Report ", Iran IX (1971), pp. 1-17. Elsewhere in this paper, the reports will be referred to as Sfrdf I, II, III and IV. * We thank also Prof. Stephenson and the Department of Civil Engineering, Imperial College of Science, London University, for lending survey instruments. As before, the staff of Decca Services, Bushire, gave us hospitality and assistance, as did the staff of Philco Ford, Bushire, and the crew of M.V. Aggie, Price International, Abadan. In addition to direct financial support, our sponsors again offered bursaries for competition among members of staff and lent us a short wheelbase Land-Rover. Bursaries were awarded to Sarah Jennings, Jonathan Erskine and Jan Roberts. Under the terms of her award, Miss Jennings spent the spring and summer of 1971 helping to prepare the results for publication. Finally, we are indebted to the Curator of the Pitt-Rivers Museum, Oxford, Mr. Bernard Fagg, for providing storage facilities for the finds,

a scale of I : 500 and defensive works on the east side of the site at I : Iooo. At the same time we

63

64

JOURNAL

OF

PERSIAN

STUDIES

Area mappedat 1:500

BAGH I SHAIKH " KUNARAK

J~

,Om

.-..D

500

DOBAND

Si 1000
its E - iA H B

VALLEY i"
Earthwork

C
1500 ' 2000 K TANG I LIR

i
Defensive Walls

" TAHERI

TAHERI. .......

Aqueducts

Modern Settlements -,
SEastern Defences

Modern Road

RUOyster MiAddens
",
Fig. I. Siraf. (Scale

RUDKHANE KUCHEK
: 25,oo000.)

EXCAVATIONS

AT

SIRAF

65

Finally, I am grateful to Mr. David Stronach and Mr. A. H. Morton for their help and advice on practical arrangementsin Iran. The report which follows contains three sections: The Excavation. 3. Discussion: Sasanian SIrif.
2.

I. Topography.

I. TOPOGRAPHY

In 1968-69 we mapped the visible remains of SIraf at a scale of I : 500 revealing, often in detail, a palimpsest plan of the main urban area.3 The following season we added contours, usually at 2 5 m. vertical intervals, but at intervals of 5 m. or even Io m. on the steepest slopes. Last year we extended at : the survey to include traces of extra-mural settlement to the north and west of SIrTf,planned I 500, and of defensive works on the east side of Taherl village, planned at I : Iooo (Fig. I). At the same time we planned at a large scale one of the city's principal cemeteries and the footings of a funerary mosque, 325 m. north of Site F. Lastly, we explored a series of aqueducts which supplied Siraf with water collected in the hinterland to the north. The City (Fig. 2). The planning of a large cemetery at a scale of I : 200 provided an opportunity to study in detail the form of the graves. Several of Slrdf's largest cemeteries occupied the floors of abandoned quarries and so presumably belonged to a relatively late stage in the growth of the city. While the majority of graves was aligned approximately north-south, with a significant minority pointing east-west, numerous examples simply filled the space available and orientation was not of o m. deep, but one group with paramount importance.4 Most graves measured some I .7x 0o6 0.5 an east-west alignment was considerably larger, with a maximum length of 2-8 m. The rock-cut graves had vertical or splayed sides, the latter giving greater width at the bottom. Many graves possessed offsets or horizontal grooves in the sides to support stone slabs which covered the body. Above the slabs, the grave was backfilled with earth or plaster and, although most are empty today, it is clear that originally many were surmounted by a plinth or mortared rubble. In addition to the rock-cut graves, SIrdf possesses numerous chambers cut in vertical rock faces (P1. Ia). The chambers, which often measure barely 2 m. across by I -5 m. high, are manifestly too small and inaccessible to be dwellings and, as several contain human bones, it is clear that they were tombs for collective burial. The tombs occur either singly or, more often, in small cemeteries, usually situated in secluded ravines. When well preserved, the tombs have small rectangular openings, sometimes retaining traces of white plaster. The interior may contain a simple rock-cut bench. Obviously the tombs are not Islamic, for Qur'dnic law prescribes single burial in a grave. The most likely alternative is that they are the ossuaries of a Zoroastrian community, intended to receive the bones after exposure of the corpse. While it is possible that a Zoroastriancommunity may have existed at SIrdfin the Islamic period, the presence of Sasanian structures elsewhere on the site suggests that many, if not all, of the tombs are Sasanian.5 Elsewhere in the city we examined several-systemsof catchment, designed to conserve storm water and prevent it from sweeping down the escarpment and flooding the town. A gully between Site G and the western defences, for example, was dammed at three points, with a conduit leading from the highest barrier towards an aqueduct which entered SIrPfthrough KunTrakgorge. Associated with the
conduit were two or more well-like structures, which presumably served as settling tanks to remove silt and other debris from the water. One of the large quarry-cemeteries in Shllau valley contained conduits associated with three large cisterns designed to collect storm water from the quarry floor.
WorldArchaeology no. 2, p. 145, contains a simplified version II, of the plan, together with a verbal description. 4 The orientation of graves was discussed in SfrdfIII, p. I6, where I suggested that graves aligned north-south were Shi'ite and those aligned east-west were Sunnite, in accordance with reports of recent practice in the Taheri region. However, the
3

Shi'ite mullahof Tdheri states that all Muslims in the area are buried in graves aligned east-west today and the problem of the medieval graves aligned north-south requires further investigation. p. 5 Stein, ArchaeologicalReconnaissances, 218, reports similar chamber tombs at Haraj near Galehdar, 33 km. E.N.E. of Sirif.

66

JOURNAL

OF

PERSIAN

STUDIES

15
30

STEJ

BAGHISHAIKH

A3
60

-i

\*: GATE
CITY WALL

%UEDUCT AQUEDURY

-, -

Kunarak gorge

*,

Om

9&10c

Si
/ )

au.*All
.-
,o

_ ..... .... __, ,, Q. ~ ,', ..... ~..\o ,"-,t,,.' .... ~ ..... :1 : :::
:: ::: : : : :
..
.

,,,, ,,, // 5c H USE : ' ::::::-: -::: :!? : -,:!i Sounding I ,,= ] j .T =! , / ' . , , " .F 'i -: " ....... MOSQUE ::: :: : MOSQUEFUNERARY : : " .)_k_ . ....SE SIla ' , ZA !,,, 00"i HME.ERY E 'os"" .......... G EATIM ,., ....
........

.. ....

// . L

......

:
...........

:: ....... ...... ......... LF

.. .

.........F :: :::~ PERSIAN

.

.I ... .I ........... . OU.,, :TR

~.
.

~,IdX
BZ

i / K , ITE

, .' : . e(/ r kt ( ~ . 1i .- 7

, :

l )..
~
.

".. _ .. ,. ..... :

'-J

. "

EMET

/
.:

..

:"

u "

: --::-

----:

:I -

45 5oR(4i'"'"(l :: :

"

--

~ ,4, -"" ~:]F~,I.: ~II'

-

'"'z/ : "/.,,'.k:TANGILI \t ? ' ( .

-:

MAI

.,'G =

tT 7.:: "<
? o

l.K

"

cg. ,

:' _,

'"' ::i;'.. : . . ' .

..l,

JRA
~

..~ .. ~ ...
::~:1
]: .. .... ........~ . .... . . .

-SQ?
20

I 1! II
..

,,,,

('../

..

,".SET

~~~[I
ot

-/I_

~~~~
,

D-;
, (
1.
605

.= -< %. '" "L

".

":,-

.,

. " ,,

"

:,

;,_
-

_,-, ' TLE--"

::

- ., ... " , .% _+, .

--:'

?"

TaheriVillag MEERY ,, .,:'. C QUEDL! F120. EXTA, URA

/f

Fig. 2. Siraf. The mainurbanarea,

EXCAVATIONS

AT

SRRAF

67

Bdgh-iShaikh. To the west of SIrafthe coastal plain continues for several kilometres towards Akhtar as a sandy tract 5oo-1ooo m. wide, which today supports scattered palm gardens and a hamlet known as Bagh-i Shaikh. The topography closely resembles that of SIraf itself, with a steep escarpment, dissected by gullies, to the north. Traces of extra-mural settlement exist on both the plain and the escarpment. On the plain eroded mounds of rubble, mortared footings and cisterns cover an area more than 300 m. across. On the ridge, ruins extend for some 500 m. westwards from Kundrak gorge and include cemeteries, two small mosques and three quadrilateral enclosures up to 33 m. across. With rare exceptions, the graves are aligned north-south. The mosques are rectangular buildings with perhaps resembling the mosque at Site C (see below, p. 82). Little survives beyond projecting mihrdbs, this area, with the exception of a quarry containing a cemetery of chamber tombs and a group of wells, each approached by a flight of steps. TheEastern Defences.The remains of SIrtf are obscured on the east side by TaherI village. Although parts of Taherl appear to rest on several metres of artificial deposit and at least one small cemetery existed on the escarpment behind the village, we believe that dense occupation of the type found on the west side of the shaikh's fort did not exist to the east. Indeed, it is possible that the eastern defences of SIraf reinforced the spur on which the fort now stands. Nevertheless, traces of buildings and a wide scatter of medieval pottery indicate occupation to the east of Taherl and clearly suburbs, either within or outside the walls, extended along the shore to the eastern arm of TaherI bay. The slight promontories which define Ttherl bay are composed of alluvium washed from the hinterland by winter rains. Thus, the west promontory is formed by alluvium from Kunarak and the east spit by alluvium from a watercourse known simply as Rild-khdne kuchek.Like Kundrak, the R1dkuchek khdne approaches the coast through a ravine in the first escarpmentand has an irregular, boulderstrewn bed extending across the plain to the sea. The west bank was reinforced with a curtain wall, closing the eastern approaches to SIraf. The wall has been destroyed almost totally and survives intermittently as a single course of boulders and mud, up to I m. thick. At the south end, the wall approaches a small fortified post, 21 m. square, with towers at the angles. Three of the towers appear to be square, while the fourth is circular, a combination which suggests that the fort is post-medieval in date. The wall itself was reinforced by at least four semi-circular towers with an external diameter of 4-4'5 m. These defences have little in common with the buttressed, sdruj-bondedwall on the west side of SIrdf (see below, p. 78) and, while excavation offers the only means of establishing their date, it seems likely that they formed the eastern boundary of Shtllu, enclosing not only the inhabited area but also the fields and gardens.6 At 350 m. beyond the eastern defences stands a group of eroded middens of oyster shells. The middens are at the top of the beach, just above the high water mark. It is tempting to regard the middens as the debris of pearl fishing, for Istakhrl comments on the pearls marketed at SIrdf in the tenth century and Tisti, writing in the thirteenth century, states that pearls were fished at Shll~u. It should be noted, however, that shell middens occurred in fourteenth to sixteenth-century contexts at Sites A and F7 and that we found oyster middens in the post-medieval levels at Site J (see below, p. 76). Clearly, therefore, shellfish were consumed at ShIliu and it would be premature to regard an oyster midden as positive proof of pearling. Diband Valley. The northern edge of SIrtf is marked by an escarpment I8o m. high. The escarpment is roughly parallel to the coast and beyond it, again parallel to the shore, lies a valley known as Dibandwhich extends from Akhtdr to Tang-i LIr. The whole valley contains scattered traces of medieval
occupation, presumably gardens and hamlets on the outskirts of SIrtf. The largest surviving structure stands near the east end of the valley and is a rectangular enclosure measuring some 75 x 60 m. (P1. Ib). The enclosure has eroded earthen banks up to I 5 m. high. It is divided by a central partition and a second partition divides the northern part. No trace of stone survives and it appears that the banks
6 For Shiliu, the post-medieval successor of Sir~f, see Jean
7 In Period 3 at Site A (Siraf I, p. 8) and in the post-medieval

Aubin, " La Survie de Shiliu et la Route du Khunj-6-Fil ", IranVII (1969), pp. 21-38; and World II, Archaeology no. 2, pp. 154-7.

occupation of Site F (Sfrdf III, p. 15). Clam-like shellfish are said to be eaten in Taheri today and Mr. Williamson tells me that clams are eaten at MinAb, east of Bandar 'Abbis.

68

JOURNAL

OF

PERSIAN

STUDIES

conceal mud walls. The building may have been a caravanseraion the route to Jam, but the complete absence of datable pottery makes it impossible to suggest its date. Aqueducts. SIrdfreceived part (perhaps the major part) of its water from the hinterland, as Istakhrl reported. Aqueducts entered the city at three points: Kundrak gorge, Tang-i LIr and the Rfid-khane kuchek. In each case the conduit, usually o*35 m. wide, was either rock-cut or carried in a sarqj-lined channel, as local conditions required. Where necessary, the conduits were carried across gullies on bridges (Pl. Ic) and in Tang-i LIr a U-shaped conduit transferredthe aqueduct from one side of the gorge to the other, passing beneath the torrent bed.8 The longest aqueduct brought water from the upper Kunarak valley to the western defences, a distance of 5 km. The head of the valley was closed with a dam, below which stood a group of cisterns and other buildings. At the foot of an escarpment called Bard-i Par, I km. south of the dam, traces of two conduits survive at different heights (one of .5 which is shown on Pl. Id), and 750 m. farther south conduits are visible on both sides of the river bed. In the Rfid-khane kuchek no fewer than five separate conduits survive and last year we found a total of nine aqueducts supplying the city.
2.

THE EXCAVATION

The excavation lasted fourteen weeks, during which we employed up to 150 men. In the first half of the season we continued work at the Great Mosque with the object of investigating further the early enclosures. Excavation established that the inner enclosure, if symmetrical, was 62 m. square, with circular towers at the angles. The larger, outer enclosure had a long history, beginning in the Sasanian and period and continuing until the eighth century. Meanwhile, we examined a ruined imdmzddeh two structureson the promontory which forms the west side of TaherI bay (Site J): a large building and a hammdm. also examined a gate in the city wall (Site L). In the second half of the season we transWe ferred the labour force from the Great Mosque to a palatial complex (Site K), in which the largest building measured at least 37 x 38 m. with more than thirty rooms on the ground floor. At the same time, workmen from Site J carried out minor excavations at Sites C, D, E, F and G. By the end of the season we had brought the investigation of Sites B-J and Site L to a satisfactory conclusion, so that only Site K demands further attention. The sites are described in the following order: I. Site B. The Great Mosque and Sasanian enclosures. 2. Site K. A palatial residence. 3. Site J. A large building and a hammdm. 4. Site L. The west gate. 6. Minor excavations at Sites C, D, F, G and E. I. Site B. The GreatMosqueand Sasanian Enclosures9 Work at the Great Mosque began in 1966. In 1967-68 we excavated the main building, a courtyard mosque measuring 44 m. wide by 55 m. deep, and began to uncover the extension. The following year we investigated the history of the main building, completed work in the extension and excavated the external ablution facilities. In 1969-70 we showed that the earliest mosque was built shortly after
I88/803-04. It rested on a platform 2 m. high, filled with earth and rubble. During the third season we began to remove the filling to reveal the floors and footings of an earlier complex. The following winter we concentrated our resources on the early buildings and found that they comprised two enclosures: a fortified inner enclosureand a larger, lightly defended outer enclosuremore than 70 m. across (Fig. 3). Last year we continued to examine the two enclosures and established their Sasanian
8

5. Site H. The imdmzddeh.

Stein, Archaeological Reconnaissances, 210o, suggested that the p. vertical sides of the U might be connected with water mills. However, a similar construction carries the modern conduit which supplies the gardens of Bagh-i Shaikh beneath a torrent

bed 3 km. west of Tiheri and the function of the water works in Tang-i Lir cannot be doubted. Sirdf I, pp. 9-11; II, pp. 41-8; III, pp. 2-8; IV, pp. 2-5.

EXCAVATIONS

AT

SIRAF

69

..

--

'

"2\

S"

Embrasure ?

Sm

o0Postern
IS2.8

O I OUTER

- - - - , ----_metrs

metres

Fig. 3. Site B. The earlyenclosures.

70

JOURNAL

OF

PERSIAN

STUDIES

origin. Excavation outside the Great Mosque involved removing part of the bazaar, which yielded further evidence for the chronology of the earliest mosque (Pl. IIa). The Inner Enclosure. This was an imposing structure, aligned approximately north-south. We have excavated parts of the south and east sides; the centre lies beneath an embanked road and is inaccessible. However, we have uncovered enough to show that the structure was a fort or fortified palace. The south side contains an entrance and, if this occupied the axis of the building, the fort was 62 m. across. The entrance consisted of an opening 2' I m. wide, flanked by towers. The towers were hollow, with a slightly stilted semi-circular plan 4'9 m. across. The only features on the ground floor were single arrow slits covering the door. The entrance gave access to a rectangular gatehouse, measuring m. internally. Beyond the gatehouse was a gravel yard containing several small buildings. Io02 x At the 3"8 south-east angle of the fort was a circular tower, again hollow at the base, with a diameter of 4' 9 m. (Pl. IIc and d). Inside the south and east walls, and presumably running round the whole building, was a range of rooms 2-2 m. wide. The rooms had plaster floors and each was entered from the yard. Presumably they were barracks or magazines, the roof of which served as a fighting platform 3 m. deep. In general terms, the building recalls the smaller Umayyad fortress-palaces, such as Qasr al-IHair ash-Sharql, 70 km. north-east of Palmyra, which was built by Hisham inI I o/728-29.10 The OuterEnclosure. The outer enclosure had a different character: a rambling warren of buildings traversed by narrow earth streets and protected by a simple curtain wall. We have excavated only part of the enclosure to the south of the fort, and its full extent is unknown. A curtain wall I m. thick protected part of the south side, running for 30 m. along the top of the beach. We found only two features: a gate I 7 m. wide and an archers' embrasure. At the east end of the excavation were traces of a second curtain wall, again I m. thick, apparently extending from the south-east angle of the fort towards the sea, and possibly forming the east side of the enclosure. Within the enclosure were buildings of mortared rubble (Pl. IIIb). Although they presented a disorderly plan, the structures were soundly built with mortared walls resting on foundations of rubble and clay. Plaster floors were used throughout the complex and in several rooms pottery vessels had been embedded in the floor during construction. One structure (Pl. IIIa) was particularly well built, with two adjoining ranges of rooms 14 m. long. The complex was traversed by narrow streets (P1. IIIc) and contained several small rectangular yards. Chronology.The outer enclosure had a long history and a section through the curtain wall showed that it was rebuilt on three occasions (P1. IVa). The latest rebuilding (Period 4) was contemporary with the penultimate phase in the interior; the latest building of all (Period 5), which is indicated by solid black lines on the plan in Sirdf III, fig. 7, was a small structure beneath bays KIo, L9-I I and M Io of the Great Mosque. The floors of Period 4 sealed several lead coins comparable with the eighth- and very early ninth-century coins from the platform of the mosque. Periods 4 and 5, therefore, belong to the eighth century, although part of the curtain wall collapsed between Period 4 and the construction of the mosque. Period 3 lacked the lead coins, but yielded a lightweight solidus of Constans II (641-68), struck at Byzantium in 651-59 (P1. XIIa). The coin is almost in mint condition. A single coin, however, is insufficient evidence for chronological deductions and Period 3 is datable only within the wide bracket of 65I-c. 750. Periods I and 2 were associated with Sasanian bronze coins and, although these cannot be attributed to individual rulers, we are confident that the earliest occupation is Sasanian. Among the debris from Period I was a pair of gold ear-rings, each embellished with a pearl (P1. XIIc). Other Sasanian finds from the outer enclosure included chalcedony seals depicting an ibex and a human figure (P1. XIId and e). Throughout its history, which extended from before c. 650 to c. 8oo, the enclosure had the same aspect; Fig. 4 shows part of the complex in Periods 2 and 3: a group of rooms adjacent to the postern gate (P1. IVb) and associated with a courtyard containing a well. The well (P1. IVc), which was rebuilt at least twice, was approached by a flight of steps. No similar sequence or evidence of date is available for the fortress-palace. However, the fort and the outer enclosure so clearly form a unit that it is logical to regard the nucleus as original. The presence of a central fort explains the flimsy defences of the outer enclosure. Moreover, by the eighth
10

K. A. C. Creswell, Early Muslim Architecture, edn. (London I969), I, pt. 2, pp. 522-8. 2nd

EXCAVATIONS

AT

SIRAF

71

century the fortress was almost redundant; the buildings of Period 4 in the outer complex had encroached on the walls to an extent which made it barely defensible, for not only did they obscure the defenders' field of fire, but also afforded cover to attackers approaching the gate. If the first phase in the outer enclosure is Sasanian, then I suggest so is the fort, despite its resemblance to an Umayyad palace.

PERIOD3

<

Dr/in

Outer Wall

05

metres

J&JE

Fig. 4. Site B. Periods2 and 3 in the outerenclosure.

The Great Mosque. The finds from the platform of the mosque confirmed the chronology suggested in SirdfIV, pp. 3-4. The filling yielded additional lead coins, the latest of which were struck in or about 188/803-04, as were the latest coins found in 1969-70. Other finds included a fragmentary " Dusun " stoneware jar, painted stoneware and an earthenware jar with a green alkaline glaze and barbotine decoration (P1. Xb). Removal of the steps beside the minaret did, however, elucidate one point which was not clear in 1969-70: the significance of the coin of I99/814-I5 noted in Sirdf IV, p. 3. It is clear now that in their surviving form the steps were a secondary feature. They overlay a series of deposits and one structure: a stone footing bonded with the earliest minaret, but begun after the minaret-the side of the tower had a flat face from which bonding stones projected (P1. IIb). The footing, therefore, was planned from the outset, but not necessarily built initially. Further excavation showed that the coin of 814-15 came not from the make-up for the steps, as originally thought, but from the rubble core of the footing. Thus, while the platform of the mosque probably was completed shortly after 803-04, some features, the footing included, were not finished until 814-15 or later. Observations near the minaret also cast welcome light on the problem of the earliest tin glazed pottery. It was clear that the earliest structures in the bazaar outside the mosque were built shortly

72

JOURNAL

OF

PERSIAN

STUDIES

after the mosque itself. Thus, a deposit associated with one of the earliest shops which abut against the footing of 814-15+ yielded a small group of coins, most of which were lead pieces struck no later than the early ninth century; the other coins were either illegible or earlier in date. An adjoining shop of the same period yielded two undecorated sherds of tin glazed pottery. It is difficult to escape the conclusion that the sherds are of the ninth century and that the traditional ninth century date proposed for the earliest tin glazed wares is correct. Site K. A Palatial Residence To judge from surface remains, the ridge dividing ShIlu valley from the coastal plain contains at least four buildings, or building complexes, of unusual size. Although the complexes appear to have a domestic character, they differ profoundly from the houses at Site F. They stand on or near the crest of the ridge at an altitude of at least 40 m., separated from the commercial and industrial quarters and occupying the coolest parts of the city. They are exceptionally large; while the largest building at Site F, House W, has an area of 540 m.2 and the largest house planned from surface remains is 25 m. square, the smallest complex covers at least 900 m.2 The complexes have irregular plans. Unlike the houses at Site F, which do not always possess a well, each complex has at least one cistern. It seems reasonable to suggest that the complexes are the palatial residences of leading merchants and officialsthe buildings in the mind of Istakhrl, when he writes that an inhabitant of STIrfmight spend 30,000
2.

dindrs on a house.

?j

.00

00.-0-

i65.00O
-I'

135.00 I e

3750 - ..00.0
.40

, .

42.50
.

.
0.-

? /".'I
.i, o

is/

V.

.-I

- ._.. .
"

.
?

. .. "
"
,

47 50 .0 ..

, "

/?

00

l.0

6 7.50

__I0-1
i"

I.,0"0
. .%

".,r

.-

"

II
-".m,..e /-...

, ..,, CI;

i
- -,

r ci~ternciste600
?,
I

.. ......... , !., 60nncser ,

,". ,"
. . . '

I. _..- -. - ? L ..e I-I ---- ,, Ii

,/
I.

"\e i

?I

,.- .-- -I!I., .,l-m.

m

-- - -

m.,

. " " .,
"

" "I I .
. .\

"

"

'- - .

__.0

\50.00

'r

,"""

!,
.0
L

, .-...
L

0062.50

.
"'\
'l "1

1
50
52.
"*""

,
I

.45.055.0

OOP52750

, ,.
0?-?

"-.

.., ,J ....
ou-0.0 ?

,,

0%505100

50

100

-' ?

Ii ~

I~

I
met res

Fig. 5. Site K. (Scale i: wooo.)

EXCAVATIONS

AT

SIRAF

73

"#45.oo00
well
47.50

"drain
/

?well Upper Platform

.1

50.00

cistern,

152.50

55.00

'

. . -.. --------------

50.00

rock outcrop

53.40

10 I

20

-Walls

fully excavated

I metres

Walls partly excavated

WB& PDm DWd

and Fig. 6. Site K. The upper lowerplatforms. (Scale ":4oo.) abovesea level. are Note: spot heightsand contours shownin metres
7

74

JOURNAL

OF

PERSIAN

STUDIES

Site K is the largest palatial complex. It stands on the crest of the ridge with commanding views of the city and bay; to the south, it overlooks an area of courtyard houses, probably comparable with Site F, and the bazaar, while on the north side the ground drops precipitously into Shillu valley. The full extent of the complex is unknown, but surface remains suggest that it covered at least one hectare (Fig. 5). At the west end of the enclosure stood two large buildings, most of which we excavated last season. Farther east are the footings of another large building, while a curtain wall more than Ioo m. long encloses the south side of the complex. Last year we examined the west end of the site and exposed the curtain wall. The west end of the ridge had been levelled to form two " platforms ", each of which supported a building (P1. Va and Fig. 6). The upperbuilding measured at least 37 x 38 m. and contained more than thirty rooms. On the north and south sides the outermost rooms were built at a lower level than those in the centre. Thus, the south range had floors 2 m. below the floors of rooms situated immediately to the north (P1. Vb). Each room in this range was entered from a courtyard or passage outside the south wall (P1. Vc and d). On the north side of the building the outermost range was supported on a row of massive piers, with a difference of 6 m. between the base of the piers and the floors in the central rooms. On the west side of the building was a cistern. The upper building contained several periods of construction. We found two types of masonry and both semi-circular and rectangular pilasters. As at Site F, where semicircular pilasters were characteristic of the earliest phases of construction in Houses S, W and Y,11 so here the round pilasters were sometimes encased in later masonry. The lower building, which measured more than 28 x 20 m., was considerably eroded and disturbed by post-medieval graves. The plan is not yet clear; indeed, it is possible that the structure adjoins the upper building at the west end. As was customary at SIrMf,the walls of the palace were plastered. Decorated stucco was common; we found several pieces with Qur'anic inscriptions, fragmentary panels bearing star and polygon ornament comparable with Sirdf III, pl. XIa, and a broken roundel, again with moulded decoration. Scratched on the wall of a room in the upper building was a three-masted boat (Fig. 7), evidently an ocean-going vessel of considerable size; Prins notes a two-masted dhow of 238 tons,12 which required a crew of at least thirty. The graffito was made some time before the building collapsed, perhaps in the eleventh century, since late sgraffiato and Saljuq frit wares were absent from the fill. The upper platform also yielded two exceptional finds, both of Chinese stoneware: a four-lobed bowl and a jar. The bowl (P1. XI) has a harsh white paste, an ivory glaze enlivened with spots of green and moulded decoration. It is a rare type at SIr~f and previous finds consist of small fragments with simple, often geometric, ornament. The jar (P1. Xa) has a brown glaze which ceases a short distance above the base. It had two handles (now missing) and is embellished with applique reliefs. Several varieties of applique ornament associated with brown glaze occur at SIraf; another fragment is illustrated in Sirdf I, pl. VIa. Both vessels were found in deposits which postdate the abandonment of Site K. Other Chinese ceramics from the palace include green wares (comparable with Siraf IV, pl. IXa), painted stoneware (cp. Sirdf IV, pl. VIIIa and c) and white stoneware and porcelain.

3. SiteJ. A LargeBuildinganda HIammdm

We excavated two adjacent buildings on the promontory which forms the west side of Taheri bay: a large structure resembling a caravanserai and a hammcm (Fig. 8). The promontory is an alluvial fan deposited by Kunarak water course, which reaches the sea near to the tip. For much of its length, the city wall reinforced the east bank of the water course and in effect the buildings stand on a narrow spit between the water course and the sea, with the city wall running along the western edge. The large building stood on the neck of the spit, only 5 m. inside the city wall. The hammdmwas 15 m. farther north and abutted on to the wall. In neither case was the date of construction established, although it was clear that both buildings had been abandoned before the introduction of late sgraffiato ware (i.e. SirdflII, pl. VIe and SirdflIV, pl. VIIId), c. Io50. Deposits associated with the construction of the large
III, pp. 13-13. xxSirWf
12

A. H. J. Prins, " The Persian Gulf Dhows ", Persica II (1965-

66), pp. 1-18.

EXCAVATIONS

AT

SIRAF

75

0o
I

o
I I

20
I
Cms PDm DWd
Fig. 7. Site K. Graffitoof a boat.

76

JOURNAL

OF

PERSIAN

STUDIES

building yielded only coarse pottery and monochrome green glazed ware, while similar deposits in the hammdm contained tin glazed sherds of the ninth century or later. After it had fallen into decay, the large building was re-occupied from time to time, perhaps by fishermen, who left shell middens associated with fourteenth- or fifteenth-century pottery. TheLargeBuilding(Fig. 9). The large building, which is known locally as Bang-i Sdr, was thought by Stein to be a fort.13 Its massive construction, revealed by marine erosion, and its position near the city wall supported this view, which I followed in previous Interim Reports. Excavation, however, showed that this identification was wrong; the absence of towers and the plan of the interior suggested an alternative explanation. The building was rectangular, measuring 51 x 38 m. (P1. VIa). One corner had been destroyed by the sea, but the rest survived to a maximum height of 2 25 m. The standing remains belonged to a single period, but overlay the foundations of an earlier building, perhaps equally large and with a similar plan (P1. VIc). The standing remains were built on an imposing scale; the outer walls were I - 35 m. thick at the base, while internal walls and most of the piers varied in thickness between I - 22 boulders, again saraj, rough and mortar. The superstructureconsisted of mortared bonded with sdra-j, capped with layers of sdraj courses of dressed boulders. Where examined, the internal walls usually had foundations of rubble and below a superstructureof mortared mortar, poured into trenches, with one course of boulders and sdraj rubble. Junctions were bonded with timber and stones. Each row of piers rested on a continuous foundation and individual piers had one course of boulders and sdrajbelow the mortared rubble. The building had two distinct, but connected, parts: a group of rooms surrounding a courtyard to the east and an arcaded area containing a massive foundation to the west. A single entrance gave access to the courtyard, from which one entered the area to the west. The east and west parts were divided by a wall with five openings, which provided ready access from one end of the building to the other. The eastpart had a central courtyard, 14'5 m. long and m. across, surrounded by an arcade 15"4 - 6 m. wide. At the east end, bisected by the axis of the building, stood the entrance passage. Although 2 partly destroyed, the passage evidently measured 9x 3"44 m., with shallow pilasters at the inner end. Flanking the passage on either side were two rooms, measuring 6-04 x 2"75 and x 2"75 m. res7"5 pectively. To the north and south of the courtyard were rows of rooms. Originally, each row contained and 2 7-2-75 m. wide; The rooms in the north range were five identical rooms. m. long 4.-65-4.'7wide. Each had a m. long and those in the south range were m. single opening 2.72-2.75 m. wide; from the courtyard, I 23-I 4.74-4"'78 in some cases thresholds and sockets for the door frame "27 survived. Fallen masonry from the arcades included fragmentary arches, but none was sufficientlylarge reconstruction. Despite the thickness of the walls, which could have supported at least one to permit upper storey, we found no direct evidence that such existed. No trace of a staircase survived, although stairs may have existed on the east side of the yard, outside rooms 2 and 15-. The westpart contained six rows of piers, with neither courtyard nor rooms. Although no pier survived to a height of more than I m., we may make several deductions about the superstructureif we assume: (I) that the presence of a continuous foundation between two piers indicates that they were linked by an arch; (2) that the width of an arch correspondswith the thickness of the wall it supports;
and (3) that the thickness of a wall is related to its height. Granted these premises, it appears that the m. wide, massive foundations in the centre supported a rectangular structure 12 m. long and 9"5 borne on piers 15 m. thick. The exceptional thickness suggests that the structure may have risen 2" to an appreciable height. Projecting northwards from piers 33 and 35 and southwards from piers 28 and 30 were rectangular pilasters I *2 m. wide. Each of these was linked by a continuous foundation to one of the L-shaped piers 22, 25, 38 and 41. It appears, therefore, that arcades supporting walls approximately I 2 m. thick formed three sides of rectangular features abutting on to the north and south sides of the central structure. Finally, the outermost rows of piers were linked by a foundation
i3 Archaeological Reconnaissances, p. 20o5.

and I. 28 m., with one group of piers 15 m. thick. The outer walls were slightly battered. They had boulders and 2" above which was one course of trimmed a foundation of

II / /

II
""

:::3II

II

i!

KUNARAK '
II
II Ii i1 II

BATH HOUSE

II ''

CIY ?.

1 /? LARGEBUILDN ""--.-,

,._.. ?!

? ..

?'_C

BA,.**

/\
I
_ _ _

i"

IX

LARGEBUILDING .,. 0
.-33 z
V~

50
I I ,

;s
2/VALLG3\rL/
/ ?

mere
__

FGA D3 J&J~m

..-05

metres
FGA J&JEm DWd
Fig. 8. SiteJ. (Scalei :ooo.)

78

JOURNAL

OF

PERSIAN

STUDIES

running round all four sides of the west part of the building. If we assume that the outer walls were without windows and that the central structure had a roof, two principal interpretations of the west part are possible: (I) that the four aisles outside the rectangle made by piers 15, 20, 43 and 48 were open to the sky, while the rest of the building was covered, and (2) that the same aisles, and the aisles outside them, were roofed, while the rectangular areas abutting on to the central structure were open. In favour of the former hypothesis is the observation that piers 15-20 and 43-48 were offset slightly from piers 21-26 and 32-42, suggesting (but not proving) that they were not covered by a single roof; the second hypothesis, on the other hand, provides a greater covered area. The function of the large building is uncertain. Clearly, the east and west parts were interdependent and it follows that any suggestion must explain both. The east part recalls a caravanserai; if the building served as a warehouse or caravanserai for merchants transporting goods by sea, the massive foundation in the west part might be seen as piers supporting a mosque at first floor level. Alternatively, the structure could be interpreted as a tower commanding the seaward approaches to SIraf, in which case the large building might be seen as the headquarters of the port authority; its strategic position might support this view. The IHammdm (Fig. I o). The bath-house was a roughly rectangular building abutting on to the city wall, with maximum dimensions of 17 X I I 5 m. (P1. VIIa). The structure had foundations of boulders and sdrlj, with floors and other features coated with sdrzj to create a waterproof finish. The building had been dismantled or thoroughly robbed with the result that nothing survived above floor level, making it impossible to determine the precise positions of doorways and baths. The general plan, however, was clear. One entered the hammdmthrough a rectangular room (I), measuring 8 x 4 m. internally, by an opening in either the east or, less probably, the south wall. Footings near the east and west walls were all that remained of benches or perhaps baths resembling those in the hammdmat Site was a C;14 presumably the room was the apodyterium(P1. VIIb). At the north end of the apodyterium small room (2) with an earth floor, possibly used for storage. To the east was a long room (3), measuring approximately 8 5 22 m., the east end of which had been completely destroyed. The room contained a series of drains and a latrine. The bathrooms occupied the rest of the block and comprised four rooms, each roughly 3 m. square (P1. VIIc). In the south-east angle room 4 had a solid sdrzj floor, indicating that it was an unheatedfrigidarium. The rooms on each side of this (5 and 6) had a hypocaust under most of the floor, showing that they were tepidaria. The unheated areas may have supported baths. Finally, room 7 was the caldarium,with a complete hypocaust heated by a furnace in the west m. square, which formed the wall. The debris from rooms 5 and 7 included fragmentary bricks, 0.44 above the hypocausts. Outside the building, between the hammdmand the city wall, was the floors was drawn from an external stoking area (8) (P1. VIIc) and a small fuel store (9). Water for the hammdm well or conduit, which we did not find. 4. Site L. The West Gate In 1968-69 we planned the remains of the western defences, a curtain wall which reinforced Kun~rak water course. Although considerably eroded, the wall was traced from the promontory at Site J to a spur, 200 m. north of Site G; World Archaeology2, no. 2, fig. 5 contained a plan of the remains. Last year we enlarged the plan to include the northern part of the wall, from the spur to the cliff above Kundrak gorge (Fig. I ). We know now that the western defences extended for some I - 5 km. from the promontory to a vantage point above the gorge, with a clear view of the hinterland. Indeed, the wall was carefully sited to guard the approaches from both north and west. The wall survives only intermittently and we know little about the disposition of bastions and forts. On the coastal plain, the visible remains consist of a boulder footing bonded with sdriij, roughly I m. thick. At several points the footing was constructed against, not above, the bank of the torrent, presumably as a precaution against seasonal floods. The outer face was reinforced at irregular intervals with slightly stilted semi-circular buttresses I *6 m. across, again of boulders and sdrzj. Before the excavation of Site L, only one bastion was known: a massive triangular structure 8 m. across. Presumably a
14

Sirf IV, pp.

11-12,

EXCAVATIONS

AT

SIRAF

79

20

26

31

36

42

4 8

19

25

3035

41

18E

24

40E

46

29
17

34

K
39 45

23 E]

16

22

28

33

38

44

15

21

27

32

37

43

6 5

9 10
10

11

"IN1 I

3 312 2
ir-"

12 1
-

SI--1 L---------------

14

13

13

3
3
3I

, r---. .
I

-,I
I'

15

14

10

20

metres
Fig. 9. Site J. The large building.

80

JOURNAL

OF

PERSIAN

STUDIES

fort defended the south end of the wall, beyond the large building at Site J; no trace, however, survives. An enclosure adjoining the wall on the spur north of Site G (described as a " fort " on the plan in World Archaeology) be a fortified post or a barracks,despite its irregularplan. A second fort probably may existed on a re-entrant I50 m. farther north; once more the compound adjoins the wall, occupying an isolated site too inaccessible for domestic or commercial activity. At the north end the wall terminates at an observation point with footings of mortared rubble. We have determined the position of only one gate in the west wall: at Site L, although others may have existed in portions of the wall now destroyed.15 Surface remains at Site L suggested the presence of a gate, at the end of a street extending eastwards for at least 250 m. The area is considerably eroded and its excavation was complicated by the existence of a track linking Taherl and Bdgh-i Shaikh.

CITY LI WALL
drain

---2

!00 00!o6

6oo

I!

10 metres

Fig. io. Site J. The bath-house.(Scale : 200oo.)

Nevertheless, we found evidence for the following features: (I) an earthen bank behind the city wall; (2) the gate itself; (3) a second triangular bastion, of which nothing survived above ground level; and (4) an aqueduct. A section through the defences revealed that the foundation here was of boulders and sdrij, 8 m. wide and some 2 m. high. It was capped with a skin of sdrij, above which stood a wall o. of mortared rubble. Piled against the wall was an earthen bank, m. wide, which survived to a 4"*2 maximum height of m. The presence of this bank explains the flimsy construction of the wall itself, o'9 which was reinforced by a bank supporting the sentry walk. The gate survives as a break in the wall,
15

Indeed, it is likely that a second, possibly larger, gate existed farther north, near the mouth of Kunkrak gorge, by which

caravans left Siraf for the interior.

EXCAVATIONS

AT

SIRAF

81

3'9 m. wide, approached by a passage 6.7 m. long. Nothing survived to indicate the form of the gate, but a pair ofjambs were found at the inner end of the passage. The whole structure had been rebuilt on several occasions. Excavation in the torrent bed revealed the stump of a sdrij and boulder feature projecting from the city wall. Although most of the structure had been swept away, enough survived to m. deep. Deposits associated with the show that it was a hollow bastion, I I m. wide and some 5"5 wall yielded three Sasanian coins and nothing distinctly Islamic. original The aqueduct, a continuation of the conduit in Kundrak gorge which probably served the hammdm channel sealed with boulders and sdri-j. and other buildings in the vicinity of Site J, had a sdr-7j-lined We recognized three periods of construction, in the latest of which the channel was o04.5m. wide and 0.5 m. deep.

S
BASTIO

PASSAGE
I i i I

PS

i/

10 metres

Fig. II.

Site L.

82

JOURNAL

OF

PERSIAN

STUDIES

5. Site H. The Imdmzddeh

The ruins of a small mausoleum stand on the ridge which separates ShIlu valley from the coastal plain, 390 m. north of the Great Mosque. Although known today as a madrasa, the building is an imdmzddehwith a mihrdbin the west wall and three funerary monuments on the floor. The mausoleum occupies an outcrop of rock on the crest of the ridge. Cuttings and foundations of mortared stone show that the area had been occupied at an earlier date. The imdmzddehwas the nucleus of a complex comprising: (I) a platform, measuring 22-5 X 16.5 m., with a dry stone revetment; (2) the mausoleum itself; (3) traces of smaller buildings; and (4) a cistern. Other buildings abutted on to the south and west sides of the platform. In this preliminary report we are concerned solely with the mausoleum. The mausoleum consisted of a single domed chamber aligned approximately east-west, entered through a porch on the east side (P1. VIIIa and Fig. 12). The chamber was almost square. It was built of mortared rubble, faced on the outside with dressed boulders. The dome had collapsed, but sufficient remained to show that it was constructed in sixteen segments divided by plaster strips. The chamber measured 3-6x m. internally. Each wall contained an arched recess m. wide and 3"7 2"3 m. high and I 0 5 m. deep. The east recess contained the entrance, an arched opening I I15m. and wide, while the north and south recesses had rectangular windows I m. high .85 o0 6 m. wide at floor level (P1. VIIIc). The west recess contained the mihrdb,a niche with a pointed arch and a small opening which admitted the light. The axis of the niche had been rotated towards the south to conform with the qibla, which at SIrdf is approximately south-west. The mihrdband the openings in the north and south recesses had been filled with rubble, apparently when the structure became unsafe. Above the four recesses were rectangular windows measuring o *85 x 65 m. The angles of the chamber had squinches o. supporting a rudimentary zone of transition, invisible from the outside (P1. VIIIb). The chamber had a " squatters ", whose fire pits had obliterated parts of three plaster floor, considerably damaged by memorials, consisting of low rectangular plinths, such as occur in the mausoleum at Site G.16 The chamber was finished with white plaster and the zone of transition was faced with rectangular plaster slabs; no decorated plaster survived. The porch, which was badly ruined, measured m. long and I 7 m. wide and had a niche on 2"4 without boulders. Both features either side. It abutted on to the chamber and was built suggested that it was an addition to the original structure. The date of the imdmzddehis unknown. The plan of the domed chamber echoes the Sasanian chahdrtdq," while the simple zone of transition gives little indication of date. The state of preservation now datable suggests a late date and the porch recalls the porches of the funerary mosque at Site G,18s or later, and the fifteenth-century mosque at Site B.19 A single coin established to the Safavid period that the cistern collapsed in or after the seventeenth century. Although proof is lacking, we suggest that the structure belongs to the period of Shllu, between the fourteenth and the sixteenth centuries.

at 6. MinorExcavations SitesC, D, F, G andE

Apart from the Great Mosque, we returned to five sites examined in previous seasons, where problems remained to be solved. None of the excavations was large, but all provided valuable results. Site C.20 This was part of a bazaar, excavated briefly in 1966 and examined on a larger scale in 1969-70. The area contained shops, a warehouse, a hammdmand a mosque (P1. IXb and Fig. 13). The mosque stood on the south side of a street running from east to west. Although partly destroyed by the sea, sufficient of the mosque survived to show that it was a rectangular building, 8 4 m. wide and m. deep, with a courtyard attached. In 1969-70 we found that the mosque contained four periods 7"7 of construction. Outside the mosque we discovered two massive piers. Last season we re-examined the piers and investigated the courtyard adjoining the mosque. It emerged that a row of at least five piers ran along the street, linked by narrow walls. The piers varied in plan. They existed already in Period I
16

SirafIII,

17

A more useful comparison is with the mausoleum at Langarak in Khurdsdn, which contains a tomb bearing the date 818/ 1415; William Murrie Clevenger, " Some Minor Monuments

p. i6.

of Khurisan ", Iran VI (1968), pp. 57-64, especially fig. 3. is Sfrdf III, pl. VIIIa and fig. 8. 19 Sfrdf II, fig. 3. 20o Srdf IV, pp. 10-12.

EXCAVATIONS

AT

SIRAF

83

Fig. 12. Site H. The imdmzdadeh.

84

JOURNAL

OF

PERSIAN

STUDIES

and in Period 2 an entrance pierced the wall between piers 3 and 4. In Period 3 piers I and 5 were demolished. The surviving piers (2-4) continued in use in Period 4. The courtyard, which in Periods 1-3 had gravel floors, was reduced in size in the final phase and received a pavement, small parts of which survived. A street surface outside the mosque yielded a gold finger-ring with a paste setting Site D.21 Here we excavated part of a workshop to the north of the pottery examined in I969-70. The pottery occupied the whole of one city block, some 40 m. across, with streets on the north, east and west sides and the sea to the south (P1.IXa and Fig. I4). We returned to determine whether the pottery was part of an industrial quarter containing other factories and yards. The answer was: yes. Opposite the main building of the pottery stood an enclosure containing workshops, at least one kiln and a well. We uncovered the latest period only, which was contemporary with the final phase of the pottery already examined. At this period, in the tenth or eleventh century, at least two streets in the area had been blocked and the enclosure itself encroached on one of the thoroughfares. The enclosure formed a m. long. We excavated parts of three workshops, each of which quadrilateral, with a south wall 17"5 contained one or more hearths. A pottery kiln of type 222had been constructed against the south wall of the yard and in the south-west corner was a well. SiteF.23 We carried out three small excavations at Site F: (I) on the south side of House X; (2) in the mosque adjoining House W; and (3) outside the south entrance to House W. House X stood on the south side of the main street, opposite House W and immediately west of House R. On the south side, the house was considerably eroded and the removal of only m. of alluvium from an area 8 m. long, with an wall x I 5 m. revealed the plan of the fagade. The south o.5 was measuring I8.5 m. long 17" to the east of the entrance, shown in Sirdf III, fig. 4, in the centre. A single room lay 5"9 entrance passage, with the long axis running east-west. To the west of the passage were two parallel - I m. wide, with their major axes aligned north-south. The partition was unusually wide, rooms, 3 and 3 suggesting that the house was at least two stories high. In the mosque adjoining House W we removed part of the floor to check that the structures identified as Period I in Siraf IV, pp. 7-8, were indeed the earliest. The sounding revealed no trace of earlier structures. The absence of a yard or street surface below the floor of the mosque suggested that Period I might be contemporary with the earlier (semicircular pilaster) phase of House W.24 Turning to the house itself, we removed part of the pavement outside the south entrance to test the hypothesis, advanced in SirdfIV, p. 7, that a row of piers beyond the south wall already existed in the forerunner of House W. It emerged that the foundations of the piers and the south wall of the house were constructed from the same ground surface. If, as we surmise, the present south wall also existed in the foreunner of House W, then the hypothesis is confirmed. Site G.25 Our objectives at Site G were to learn more about the history of the funerary mosque and of the tomb chamber below the qiblawall. A sounding in the north angle of the courtyard revealed that in this area the standing remains were not part of the original structure. The walls and pilasters did not rest squarely on the foundations, which evidently supported an earlier structure, since demolished. Similarly, investigation of the arcade above the tomb disclosed the remains of an earlier pier beneath the existing floor. These observations suggested that the mosque had been extensively rebuilt, although we found nothing to suggest that the undercroftswere not original. In 1968-69 we recovered two coins of Abish bint Sa'd (1264-82) from a room in the north courtyard thought to be contemporary of with the earliest mosque, but noted the existence in the porch and mihrdb decorated plaster panels which appeared to be Safavid or later. The new evidence reconciled the coins and the panels; the
original mosque may indeed be of the Ilkhanid period, as previously suggested, while the panels are a product of rebuilding the earlier structure and not of redecoration, as originally thought. In the tomb chamber we removed the plaster floor and revealed that the stone cover had been placed over a rock-cut grave, a task which involved trimming the revetment of the mosque. Clearly,
21 Sfrdf IV, pp. 12-15. 22 Sfraf IV, p. 15. 21 Strdf II, pp. 48-53; 24 Sfraf III, p. 13. 25 Sfrdf III, pp. 15-18.

XIIb). (P1.

I11, pp. 9-I5; IV, pp. 5-Io; David Whitehouse, " The Houses of Sirdf, Iran ", Archaeology 24, no. 3 (1971), pp. 255-62.

EXCAVATIONS

AT

SIRAF

85

therefore, the mosque existed before the tomb. The chamber itself was constructed at an even later date, after the cover had been broken and reassembled. Indeed, it may have been built during the reconstruction of the mosque; moreover, if reconstruction was necessitated by collapse, the grave cover may have been broken by falling masonry. Site E.26 Here we enlarged the section published as Sirdf II, fig. 7, with the object of recovering additional evidence for the date of the compound and the associated pottery. At the same time we excavated 50 m.2 of the buildings adjoining the south-east angle of the site. Hitherto, the chronology of Site E had rested on coins trampled into the surface of the yard and on pottery, notably Persian wares with underglaze ornament, all of which suggested that the compound was occupied in the fifteenth century. The search for further evidence of date was prompted by a preliminary account of material from Bahrain, where types found at Site E are attributed to the twelfth century.27 Our new excavation yielded sherds of pottery painted under the glaze (two types: Bibby's type I and bowls comparable with Sirdf II, pl. VIb and VIc), unglazed painted ware (cp. Sirdf II, pl. VIf, but not Bibby's type 3, which appeared later in the history of Site E) and storage jars with an inturned rim

SN

?Entrance to Covered Bazaar

P4

courtyard

Street PublicBath

Mosque

P2

?Warehouse
:3

-\

oBlock b .

of Shops

Street, later filled with shops

r--

Row ofo
Shops Stoet
I Erodingcliff face

15
---J&JEJR DW.............

J&JE JR DW

Fig. 13. Site C.

"6 Sfraf II, pp. 54-8. 27 Geoffrey Bibby, " Bahrains Oldtidshovedstad Gennem 4000

1970), pp. 1o8-o9, to which the type numbers quoted above refer.

Ar ", Kuml (1957), PP. 128-52; Looking Dilmun (London for

86

JOURNAL

OF

PERSIAN

STUDIES

K Dr N Pi Ps :f T W

Kiln Drain Platform Preparation surface Tank Well Burnt clay

-"

with c yard, Large building well and cold bath

struturII

GS
II

PERSIAN

............. -Street --?-----/-

G U LF

Shed

PK

>S

CENTRE

YARD

WW

BUILDING

0.heStreet

m etres---......... SandJE JR MTm DWd

Fig. I-4. Site D.

EXCAVATIONS

AT

SIRAF

87

(Bibby's type 7), all sealed by the floors of the square building. Nowhere at SIraf has any of these types occurred in association with late sgraffiato ware (see above, p. 74) or Saljuq frit wares (residual sherds apart), suggesting strongly that they did not reach the site until the thirteenth century or later. 3. DISCUSSION: SASANIAN SIRAF

The high probability that the earliest buildings at Site B are Sasanian prompts a review of other evidence for Sasanian settlement at Sirtf. While none of the following provides conclusive proof of Sasanian occupation, the combined evidence suggests very strongly indeed that a considerable preIslamic settlement existed: (i) the chamber tombs are not Moslem and some, if not the majority, may be pre-Islamic; (2) at Site F, deposits beneath the forerunner of House W yielded three Sasanian coins and nothing definitely Islamic;28 (3) I km. from the fortress beneath the mosque, the earliest deposits "2 at Site L on the western defences also yielded three Sasanian coins; (4) all told, we have thirteen Sasanian coins from SIrdf, together with several Sasanian-style seals, at least one Pahlavi graffito and a Roman coin struck in 376-94 ;29 (5) I am told that a cache of at least thirty-one stone seals, comprising thirty Sasanian pieces and one engraved in Roman style, was found by chance at or near Strif after the close of the excavation in 1971. It is difficult to escape the conclusion that Slrif was a Sasanian settlement with defences more than I -2 km. across and a fortress commanding the shore. Granted this deduction, the function of the Sasanian settlement is clear; it was the port of Gtir. Gair (renamed FIrfizabdd by 'A<dud ad-Daula in the tenth century) was one of the leading Sasanian cities in Fdrs, built by ArdashIr I (224-41). The earliest itinerary between Giir and the coast is the route to SIrdf. This route was in use by the tenth century, when SIrafl caravans bound for Shiraz travelled via Jam, Pas-i Ridak and Frtfiztbdd,30 and Vanden Berghe has described as Sasanian several monuments along its course,31 although some of these, such as the dam near the plain of Jam, may be Islamic. It comes as no surprise that Gtir possessed a port, for the Sasanians were a major commercial power in the Indian Ocean, with contacts in Axum, India and Ceylon.32 IHamza al-IsfahanT, writing c. 961, lists eleven ports allegedly founded or refounded by ArdashIr I and, according to Tabari, ShThptr II (309-37) mounted a naval expedition against the Arabian coast after raiders from Balhrain and Hajar had plundered RIshahr and the coast of Fars. Cosmas Indicopleustes (XI, 337) reports that a flourishing Sasanian colony existed in Ceylon c. 522 and shortly afterwards Procopius (Bello PersicoI, xix-xx) records that a trade mission sent by Justinian to Ethiopia to buy silk failed because the Persians had acquired a monopoly in the Indian ports. SIrMf,in short, was one of a series of ports in the Persian Gulf, providing the Sasanians with a profitable share of the maritime trade which carried luxury goods from the entrep6ts of Ceylon and South India to the markets of Western Asia and the Mediterranean Sea.

28 Straf IV, p. io.

"2 Sfrdf IV,

p. 3 and pl. VIIb. " 30Jean Aubin, La Survie de Shilu Iran VII (I969), pp. FgI ", 21-38.

" R6centes D6couvertes de Monuments 31 L. Vanden Berghe, et la Route du Khunj-5632

Sasanides dans le Firs ", Iranica Antiqua I (I961), pp. 163-98. David Whitehouse, " Sirif: a Sasanian Port ", AntiquityXLV, no. I8o (i971), with references.

Pl. Ia. Chamber in tombs Kundrak in north from thecrestof the gorge. The tombon therightis unusually P1. lb. The largeenclosure Diband valley,looking well preserved, with tracesof plasterin therecess. was at ridge. Thephotograph taken sunrise; laterin the daythe site is almost invisible.

in valley. The channel P1. Ic. Remainsof an aqueduct the upperKundrak was carried acrossa gully on two masonry are piers. Tracesof the channel visiblein theleftforeground.

at Pl. Id. An aqueduct Bard-i Par. At thispoint the conduitconsisted a of tunnelthrough sandstone a spur.

Pl. Ha. Site B. Remains the GreatMosqueoverlying earlyenclosures. the of In the are overlain the and foreground theminaret part of theinnerenclosure, by bazaar.

outer bonded theoriginal with Pl. IIb. SiteB. The of foundations theminaret, seen wall of themosque, hereon theright.

Pl. IIc. Site B. The innerenclosure.The south-east angle tower,looking north-west.On the left is a drainwhichcarried waste of from theinterior the enclosure.

Pl. IId. Site B. The innerenclosure.The south-east angle tower,looking the reveals m. of superimposed north-east. thebackground section In 2 buildings in the bazaar.

south. in Pl. IIIa. Site B. A groupof rooms theouterenclosure, looking

north. Pl. IIIb. Site B. Part of theouterenclosure, looking

the lookingnorthtowards Pl. IIIc. Site B. The streetin theouterenclosure, innerenclosure.

Pl. IVa. Site B. The curtainwall of theouterenclosure.

Pl. IVb. Site B. Thepostern duringexcavation. gate in theouterenclosure

Pl. IVc. Site B. The well in theouterenclosure.

Pl. Va. Site K. Generalview, lookingeast. In theforegroundis an eroded a contained street. cuttingwhichoriginally

Pl. Vb. Site K. The southrangeof the upper building,lookingwest, the with buttresses. showing rowof openings alternating rectangular

Pl. Vc. Site K. A roomin the southrangeof the upper building. Note the runcated semi-circular buttress the vertical scale. by

P1. Vd. Site K. A roomin the south rangeof the upperbuilding. The in graffito Fig. 7 is on the left sideof theniche.

Pl. VIa. Site J. The largebuilding,looking north-east.

south. Pl. VIb. SiteJ. The courtyard the largebuilding,looking of

Pl. VIc. Site J.

The large building. An earlierfoundationbelow the floor. courtyard

roomin the lookingwest, with the undressing P1. VIIa. Site J. The bath-house, and on in foreground the suite of bathrooms the distance, the left. At the top of the pictureis the watercourse.

west. Pl. VIIb. SiteJ. The bath-house.The undressing room, looking is In theforeground Room2 which had an earth floor and may have served a store. as

Pl. VIIc. Site J. The bath-house.The heatedrooms, lookingeast, with the of stokingareaandfurnacein theforeground.At the bottom thepictureis the city wall.

Pl. VIId. SiteJ. The bath-house.Two of the heatedbathrooms, is lookingwest. In theforeground a warm roomwith a small now it hypocaust;behind is the hot room,with a large hypocaust, completely destroyed.

Pl. VIlla. Site H. The imamzddeh.

Pl. VIIIb. Site H. The imamzddeh.A squinch thezoneof and transition.

P1. VIIIc. Site H. The imdmzddeh.The northrecess.

P1. IXa. Site D. General view, looking west.

P1. IXb. Site C. General view, looking west, with the mosque on the left.

and bastionin theforeground thegate on thefar left. Pl. IXc. Site L. The city wall, with remains the triangular of

stoneware (restored), Pl. Xa. Site K. Chinese cm. jar height 18"3

P1. Xb. Site B. Jar with green alkaline glaze (restored); height 44 cm.

Pl. XI. Site K. Dish with moulded ornament, lengthT4 7 cm.

Pl. XIIa. Site B. Solidusof Constans struckin 651-59 (twicelife size). II,

iring P1. XIIb. SiteC. Goldfinger withpastesetting,diameter 8cm.

with Pl. XIIc. Site B. Pair of gold ear-rings embellished pearls. Each ear-ringis

I -5 cm. across.

Pl. XIId. Site B. Stone seal depicting a humanfigure, diameter I -6 cm.

Pl. XIIe. Site B. Stone seal depicting an ibex, i *8 cm. across.

TEPE YAHYA 1971

MESOPOTAMIA AND THE INDO-IRANIAN BORDERLANDS By C. C. Lamberg-Karlovsky

This brief account focuses on the results of our fourth consecutive summer (June 9-August 22) of field work at Tepe Yahyd.1 Excavations continue to provide a remarkable corpus of material resulting in a wholly new understanding of the 3rd-5th and Ist millennia settlement in southeastern Iran.2 Excavations concentrated on establishing a greater horizontal exposure of the important settlements, namely: Period I, IA Parthian-Sasanian ca. 0-500 A.D. Achaemenian IIA II, 500-300 B.C. IVA Elamite (?) B.C. 2200-2500 IVB Proto-Elamite 2500-3000 B.C. IVC Proto-Elamite 3000-3400 B.C.
3400-3800 B.C. "Ya.hyd The results gathered from these periods merit discussion. More restricted excavations were undertaken

VA, VB

"

in Periods III and VI.3 We have confirmed through materials gathered this season the unequivocal Late Uruk/Jemdet Nasr dating of Period IVC, the Early Dynastic I/III dating of IVB, and the late third millennium date of IVA. The direct association of distinctive Uruk-Jemdet Nasr and Early Dynastic materials with painted ceramics of the Indo-Iranian borderlands allow us to establish for the first time, with considerable conviction, a comparative stratigraphy which ties this area to the Mesopotamian chronology. The implications which arise from this comparative stratigraphy throw a new light on the process and date of West-East cultural relations and the rise of urban centres in the borderlands. Firstly, we summarize the substantive results of our 1971 season. PeriodI, IA can be divided into two distinct phases of directly superimposed levels, I and IA. The entire top of the mound is capped by a single large building complex consisting of rooms, many containing large storage vessels which are typically decorated by large bands applied in relief or by incising. Only minor changes in the ceramics can be detected between phases I and IA. The majority of the pottery consists of a coarse, grit-tempered, plain tan and buff ware, dominantly in the shape of bowls and jars. A considerable quantity of painted " Londo-type " pottery appears in I decreasing in IA.4 To date we have recovered over 20 iron three-flanged arrowheadsof varying length and io flat iron dagger blades,
none over 20 cm. in length. Selected pieces of iron are being analyzed from Periods I and II by Dr. Radomir Pleiner of the Institute for Archaeology, Prague. A considerable quantity of plain glass, We have exposed over 250 square metres of architecture in this period. It is now evident that Period I

dominantly small perfume bottles, has also been recovered.

again I would like to thank Professor Sir Max Mallowan and Dr. Georgina Herrmann for allowing me to submit this article at a late date and allowing for its rapid publication. 2 Excavations have been supported by the National Science Foundation, the Ford Foundation and private benefactors, The 1971 season included Mr. I. Yaghmai, James Humphries, Philip Kohl, Richard Meadow, Abdullah Masry, Thomas Layton, Martha Prickett, Christine Lesniak, Mina Sadegh, Raffaelo Biscione, Thomas Beale, Elizabeth Stone, Andrew Williamson, Marian Laaff (Registrar), Grace Corso, Ann

xOnce

Hechle (Registrar-Artist), Deyne Meadow, Dr. Dexter Perkins and Pat Daly (Palaeozoologists). The co-operation and assistance of the Archaeological Service of Iran is gratefully acknowledged. ProgressReport 3 See Excavationsat Tepe Yahya, Iran, 1967-1969: I, by C. C. Lamberg-Karlovsky (American School of Prehistoric Research, Bulletin 27, Peabody Museum, Harvard University, 1970) for initial results in Periods III and VI. * See Lamberg-Karlovsky, Ibid (1970), Fig. 4, pp. 8- I.

89
9A

90

JOURNAL

OF

PERSIAN

STUDIES

We publish here, for the first time, two objects of major interest: from Period I an incised (before firing) dancing girl (Anahita?) on a jar fragment (Pl. IIIb) and from IA a baboon (cynocephalus; Pl. I). This baboon is typical of the type which appears in Egypt from the 25th Dynasty through Roman times. It is of a faded, green glazed faience, standing 8 cm. high and is a representation of .2 Thot, the Egyptian god of wisdom.5 There can be little doubt that it represents a foreign import

Fig. I. PeriodII ceramics.

to Yalhyd. From Period I we have also recovered a ceramic horse (33 cm. in length) with rider, broken at the torso. The horseman is precisely paralleled in size and style to the upper-torso found in 1970.6 The horseman sits on a high saddle, the bridle is shown in relief-no stirrups are evident. Period I has not provided either an elaborate architectural complex nor evidence of the use of elaborately produced materials. The above are exceptions. It would appear that from the early centuries to the middle of the first millennium A.D. Tepe Yalhyd represents a provincial outpost of the large and important Sasanian centres known in the Jiroft, 75 kilometres directly east of Yahya.
* I would like to thank Mr. Dows Dunham of the Boston Museum of Fine Arts for his assistance in the identification of this piece.
6

For an illustration of this torso see C. C. and Martha LambergKarlovsky, " An Early City in Iran ", Scientific American224, No. 6 (197I), pp. 102-I 11.

MESOPOTAMIA

AND

THE

INDO-IRANIAN

BORDERLANDS

91

PeriodII, IIA Approximately 350 square metres of this period have been exposed. Period II can be divided into two phases of directly superimposed architecture; Period II walls resting directly above the walls of the earlier IIA construction (which are better preserved and built over stone foundations). In both II and IIA architectural units consist of large interconnecting rooms with outside spaces separating the domestic complexes. Within the rooms specialized activity areas are conspicuously absent. As in Period I, IA there are neither elaborate material remains nor architectural elaborations. There are markedly fewer iron implements and a paucity of bronze compared to Period I. Of considerable interest are the ceramic remains. In Period II there is less than i % presence of a painted bichrome (black on buff or orange slip) and polychrome (black and red paint on buff slip) pottery which cannot be readily paralleled beyond Yalhy (Fig. I: A-C). The painted pottery of Period II, nevertheless, bears a close resemblance to the painted ceramics of Period I, IA. The most elaborate designs appear in Period II (Fig. I: A-C) while in the earlier phase IIA the painted designs are more open, less complex (Fig. I: I-K). The majority of Period II pottery consists of a buff to brown medium-tempered ware with shapes tending toward the dominance of bowls and jars, without any decoration. Ceramic parallels in the plain wares can be drawn to the sites of Nad-i Ali I/II and Dahdn-i Ghulamdn both in Seistan.7 Thus the rim profiles of our bowls (Fig. I: E, L, M, P, T, N) as well as the high-neck jars (Fig. I : S, T) find ready parallels at Nad-i Ali and Dahin-i Ghulamdn.8 A study of the drawings of the excavated ceramics from Pasargadae, most kindly supplied by Mr. David Stronach, indicate that a considerable percentage of our Period II, most particularly IIA, finds direct parallels in the Achaemenian corpus from Pasargadae. Period II appears to be a mid- to late first millennium settlement with a clear Achaemenid presence (influence?). The important underlying Iron Age settlement of Period III was only briefly touched in 1971. The question of cultural continuity from Period III into IIA cannot yet be answered. We are now in position for a large horizontal exposure of this Period which in past seasons indicated close ceramic parallels to the Iron II and III materials of northwestern Iran. PeriodIVA square metres in the north step-trench, complementing the 200 square metres previously recovered in the south step-trench. We met here with a considerable surprise in the recovery of three major superimposed levels of construction-the walls of the middle level preserved to a height of 2 m. The south step-trench, on the other hand, contained in the 200 square metres but one building of IVA with phases of rebuilding.9 Each of the three levels of the 1971 Period IVA exposure consisted of several interconnecting rooms, approximately 2 *5 x 3 - 5 metres in size. The walls of all the rooms were carefully plastered as were the floors. The middle level of Period IVA was best preserved. On entering each room a hearth was located to the right of the doorway. Each hearth was connected to a chimney which rose to the roof within the wall. Dozens of replasteringswere noted on the walls, the floors and about the hearths. TI%ceramics of Period IVA continue to be largely without parallel save for ShTh-ddd,northeast of Kerman,10and perhaps Shahr-i Sokhta IV." Over 8o% of the pottery throughout Period IVA is a red slipped or plain red-orange ware.12 Little has been added to the previously identified types of IVA
pottery. The most distinctive aspect rests in the potter's marks of Period IVA. We have to date To date we have uncovered 300 square metres of Period IVA. In the 1971 season we exposed i oo

la De'ligationArchiologique Franfaiseen AfghanistanVIII, pp. 3948. U. Scerrato, " Excavations at Dahan-i Ghulaman (Seistan, Iran) ", East and West 16 Nos. 1-2 (1966), pp. 9-30. 8 See U. Scerrato, Ibid, Figs. 52-58-61. 9 See Lamberg-Karlovsky, Ibid, pp. 34-47. Directly beneath this IVA construction was Period IVB in the south step-trench. 10 Shih-dAd is a cemetery site being excavated by Engineer Hakemi for the Archaeological Service of Iran. He has re9B

SR. Ghirshman, " Fouilles de Nad-i Ali dans le Seistan Afghan", Revuedes Art AsiatiquesXIII (1938), pp. 10-22; also Memoires de

covered bronzeshaft-holeaxes, pins, cylinderand stamp seals, completesteatitevessels(over 25), lapis lazuli, and hundredsof complete vessels which can be readily paralleled at Yahy IVA and IVB. x Dr. Raffaelo Biscione, assistant to Dr. M. Tosi at Shahr-i Sokhta participatedin the excavationsat Yahby this summer. My thanksfor his very considerableassistancein confirming and correctingthe comparativestratigraphytying Yahlydand Shahr-i Sokhtatogether. is See Lamberg-Karlovsky, Ibid, Fig. I4-20; P1. 15-16.

92

JOURNAL

OF

PERSIAN

STUDIES

identified 19 distinctive classes within a sample of 278 signs, usually incised on the lower third of the vessel prior to firing but sometimes impressed on the vessel with a stamp seal. Several signs find exact parallels at Shdh-ddd. Aside from a greater architectural understanding established this summer, we illustrate two objects of fundamental importance, both recovered from excellent Period IVA contexts. Firstly, a steatite plaque (P1.IIa), depicting the architecturalfaqade of a temple-palace (?) complex, was found directly beneath the floor of the second building level of Period IVA. This steatite plaque has the typical " hut-pot " motif depicting a doorway. Within the doorway, incised in perspective is a feature which we interpret as a stepped stairway leading up to the entrance.13 To the right of the door are four columns each with an incised triangular window. The columns are bounded on each side by a zig-zag line. Atop the columns rests a gabled decoration which supports at its top a V symbol.14 Running vertically along the top of the plaque is a brickworkpattern which is precisely paralleled on a steatite block from Shih-dSd depicting a small building-also with " hut-pot " doorway and columns on three sides of the block. The plaque illustrated here is broken to the left of the doorway. Note the first stage of the gabled decoration and the zig-zag pattern exactly duplicated on both sides of the doorway. The unbroken part of the plaque has on the far right side a recessed and protruding extension, as if for insertion into or retention by wood or mud-brick. It may well have served as a wall plaque or box cover. Steatite in Period IVA is rare, rarer yet is it found with decoration. The motifs on this plaque, however, are all common to steatite fragmentswhich have been found in abundance in Period IVB. It is entirely possible that this piece is a Period IVB heirloom. It is, at any event, a piece of major importance in visualizing the elaboration evident in third millennium architecture.15 One can only wonder whether the physical structure depicted on the plaque was located at Yalhya? (not likely); made to order at Yalhyd for another city? (there is considerable evidence for the manufacture of steatite materials in Yalhya IVB); or brought back to Yahya from another city (which city ?). A single radiocarbon date is available for the uppermost level of construction of IVA. The date is of some interest, in the light of the Harappan type sealing found in the building level of IVA directly o I 0 B.C. (Tata Institute)! A date of beneath the construction which offered a date of 16I1 2320oi35 B.C. was obtained for the terminal occupation of Period IVA.e1 Of particular interest was the recovery of a sherd (broken in antiquity) having directly above its base a rectangular stamp sealing (P1.IIb). We are informed that its style and content can be compared to examples found on Harappan sites. This sealing together with several pieces of steatite of Period IVB are our only evidence for the relative contemporaneity of Periods IVB, IVA with early Harappan sites." PeriodIVB This period saw limited work this season. The ceramics have allowed us to build up a chart of comparative stratigraphy-the relations of Yahyd to other sites (Chart I). From Period IVB we have recovered over 75 Jemdet Nasr-Early Dynastic I solid-footed gobletss1and this season a full profile of a " studded ware " vessel readily paralleled in the Diyala and reported there as a type evidencing Mesopotamian (E.D. II/III)-Indus relations.19 To date we have recovered over 2 metres of Period IVB deposit consisting of surfaces and floors mostly without architecture. Indeed, some surfaces extend over Ioo square metres without a trace of architecture. This puzzling aspect continues to
13 Alternatively this feature may be interpreted as providing

corrected date would thus be 2870? 135 B.C. balance to the gabled decoration atop the columns on the right. 17 Photographs of this sealing were shown to Dr.'s B. B. Lal, B. M. 11 For an Pande, and I. Mahadevan; working independently, all three interpretation and a review of the popularity in time confirmed its Indus nature (information conveyed by Richard and space of these gabled decorations see E. Porada, " Battlements in the Military Architecture and in Symbolism of the Meadow). For parallels see M. S. Vats, Excavationsat Harappa Ancient Near East ", in Essays in the History of Architecture (Gov. of India Press, 1940), pl. LXXXVI. 18 For illustrations see Lamberg-Karlovsky and P. Kohl, " The (Phaidon) I967, pp. I-w2. 15 See also Sir Max Mallowan, " Elamite Problems ", Proceedings Early Bronze Age of Iran as seen from Tepe Yahy ", Expedition 13, No. 3-4 (I97I), fig- 33:L, and Lamberg-Karlovsky, Ibid, of the British AcademyCV (1969), pp. 255-292, for a discussion of steatite trade and its identification as an early Proto-Elamite fig. 27 :J and 30:E. For full variations of this type see H. Wright The Administration Rural Productionin an Early Mesopotamian of style. A suggestion with which we entirely agree. 16 Based on a I life of 5730 years. An additional 550 years would Town (Anthrop. Papers, No. 38 Univ. of Michigan), fig. 22, p. 73have to be added for the new calibrations, see H. N. Michael and E. K. Ralph, Dating Techniquesfor the Archaeologist 19 P. Delougaz, Pottery from the Diyala (O.I.P. LXIII, 1952), pl. I88, and pp. 143-44 for discussion of the type. (Massachusetts Institute of Technology Press, in press). The

MESOPOTAMIA

AND

THE

INDO-IRANIAN

BORDERLANDS

93

confound our understanding of IVB-above (IVA) and below (IVC) which are major architectural complexes. It is, of course, entirely possible that within the heart of the mound are the elusive structures which rest on these surfaces. We have to date only a single complete room of Period IVB; that from the earliest level, which contained the seals and ceramics excavated in I970 and previously published.20 Period IVB continued to yield numerous pieces of steatite-both decorated and undecorated, of types previously published.21 A second large purse-shaped steatite " ritual slab " was recovered from a CHART I: Archaeological Correlations as seen from Tepe Yahya Shahr-i Sokhta

Mesopotamia

Yahya VB

Bampur I II
III III

Namazga Tal-i Iblis II, III

3500 B.c.

Late Uruk

VA +

IVC .74Mundigak

IV I
III Susa C, Sialk III6-7 Iblis 5,

Jemdet Nasr
3000

Hafit, Barbar I + -V
o

Deh Morasi II Khurab Hili, Umm an-Nar Sialk IV Anjira IV, Siah III

Early Dynastic I II
2500 III

IVB

II/III

VI

Gap?

IVA
2000

IV

Shah-dad (Late) Harappan

" Lamberg-Karlovsky, The Proto-ElamiteSettlementat Tepe Yahyd ", IranIX (1971), pp. 87-96, for the publicationof the seals and ceramics. 21 The important pieces of steatite from have been published in Lamberg-Karlovsky,Ibid Yalhy. (1970-1971), and
20

Lamberg-Karlovskyand Kohl, Ibid (i97i). Mr. Kohl is undertaking a programme of neutron activation on steatite from YalhyRand other sites as well as from the local mine sources about Yahyd. His PhD. thesis will include the full corpusof Ya.hy steatite.

94

JOURNAL

OF PERSIAN

STUDIES

villager who was using the object as a weight (exactly 15 kilo). This object is precisely paralleled on a modern shrine at Ashin, 20 kilometres north of Yahya.22 Both pieces are typical of steatite objects common to Mesopotamia where they are dated to the middle of the third millennium. We illustrate here our finest single new form of steatite from IVB (Fig. 4: 1). A copper-bronze stamp seal from this period (Fig. 4: F) finds a parallel in one recovered from Shahr-i Sokhta where Period II/III ceramics also find close parallels in YahyJ IVB.23 We found in 1971 considerable pottery decorated with snakes applied in relief and readily paralleled at Susa C, Anjira IV and Umm an-Nar.24 Of particular interest are the ceramic parallels which one can draw between the Surab region, West Pakistan, and Yahya IVB. The Anjira IV snake motifs as well as incised wares and painted ceramics can be paralleled to the extent of suggesting that Anjira IV-V is roughly contemporary with Yahya IVB.25 The relations of Yalhya IVB to Mesopotamia and the borderlands are summarized on Chart I; based largely on ceramic parallels. PeriodIVC The materials recovered from Period IVC in 1971 continue to provide physical evidence necessitating a re-evaluation of the chronology of the Indo-Iranian borderlands-and a consequent new view for the process of urbanization in this area. We have recovered 200 square metres of the IVC architectural complex which yielded the Proto-Elamite tablets and cylinder sealings.26 This large complex consists of 7 excavated rooms and extends into the baulks in all areas (P1. IV). The building appears to be laid out in a pattern of small storage rooms connecting to larger rooms (at least 7 X 4 metres). Our exposure is still too small to enable us to determine a predictable outline of the entire building. An interesting baked clay water channel extends through the wall of one of the rooms (see lower left corner of P1. IV). We turn to the important evidence which continues to support the Late Uruk/Jemdet Nasr dating for Period IVC; their Mesopotamian comparisons and the associated Bampur I-IV, Shahr-i Sokhta I, Khurab, Hafit (Oman) ceramics. This past seasonwe have found one complete Proto-Elamite tablet (Fig. 4: D) and four fragments of others (Fig. 4: E). All were found in different rooms of the large building complex. The activities within and functions of this building are indicated through the materials recovered from the floors and may be interpreted as threefold: (I) storage, small rooms with large storage jars, (2) record keeping, both blank and written account tablets, (3) the receiving and shipping of goods, cylinder seal impressions on jar-stoppers of one style and actual seals of another style. Of considerable interest are two cylinder seals found in this complex; one (Fig. 4: A) has its lower portion snapped off while the other (Fig. 4: B) is broken vertically. Both seals are of " glazed steatite " and almost identically paralleled at Sin Temple IV at Khafajah.27 We have found many cylinder seal impressions of this type on jarstoppers, almost all of which can be readily paralleled in the seal corpus at Khafajah. It is interesting to note that these Jemdet Nasr seals and sealings are found on the same floors as the sealing impressions with naturalistic scenes: lions, demonic forms, procession of cattle, ibex amidst vegetation, etc., published in this journal last year. To date, however, we have not recovered a single seal of this Susa C type. Evidence for the contemporaneity of these two types at Tepe Yalhydis unequivocal-both types are found associated on the same floors together with Proto-Elamite tablets and a most interesting ceramic assemblage.
22

See Lamberg-Karlovsky and Kohl, Ibid, for illustration of the one recovered in 1970 from the local shrine. The one recovered this year was said by the owner to have been brought from Seistan by his grandfather. For a discussion of this type see, F. A. Durrani, " Stone Vases as evidence of connection between Mesopotamia and the Indus Valley", AncientPakistan I (1964), p.5I ff. 23 My thanks to Raffaelo Biscione for his careful assessment of Yahyd IVB's ceramic relation to Shahr-i Sokhta II/III. 24 For Susa C type see P. Toscanne, " Etude sur La Serpent ", Me'moiresDiligation en Perse XII (19I1), pp. 412-16. For Anjira see B. de Cardi, " Excavations and Reconnaissance in Kalat, Western Pakistan-The Prehistoric Sequence in the II Surab Region ", Pakistan Archaeology (1965), pp. 86-182.

For Umm-an Nar see K. Frifelt, "Jamdat Nasr Fund fra Oman ", Kuml 1970, fig. 7. For a single YahyA parallel (published) almost identical to that from Umm an-Nar see Lamberg-Karlovsky, Ibid, 1970, fig. 32:G. 25 For nearly identical incised wares of Anjira see B. de Cardi, Ibid, pl. IX, 4 and Lamberg-Karlovsky, Ibid, x971, fig. 3:E, while an identical snake's head on an Anjira IV vessel (pl. VIII) in 197I. The shapes and design motifs of was found at Ya.hyy figs. 17, 18 can all be paralleled at Yalhy IVB. 26 For their publication see Lamberg-Karlovsky, Ibid, 197I. Sealsfrom theDiyala Region, 27 See H. Frankfort, StratifiedCylinder (O.I.P. LXXII, 1955). Compare, for example, our fig. 4: A with pl. I2:Ioo and fig. 4: B with pl. I6:I57.

MESOPOTAMIA

AND

THE

INDO-IRANIAN

BORDERLANDS

95

Our ceramics allow us further to confirm Period IVC to be of Late Uruk/Jemdet Nasr date while making important and far reaching ceramic correlations with the Indo-Iranian borderlands. Bevelledrim bowls, grey burnished ware, as well as wide channel and drooping spouts suggest Uruk IV continuities; both types are also present at Yahya in earlier Period V. Of special interest is the recovery of

Fig. 2. PeriodIVC ceramics.

several large storagejars ofJemdet Nasr type. The large storage vessel depicted on Fig. 2: G (reduced ?) has the dotted area of the vessel slipped with a rich plum-red, while the clear area below the neck is of a buff slip, the paint is black. The second one illustrated (Fig. 2: H) has a black paint over a brown wash on buff slip. Both vessels were found in a small storage room complex together with Fig. 3: H. Both of the above polychrome vessels are similar in shape and decoration to the recently reportedJemdet Nasr of

96

JOURNAL

OF

PERSIAN

STUDIES

Oman as well as to finds in the Diyala.28 Of considerable interest are the associated ceramics which allow us to tie the pottery of the Indo-Iranian borderland to the chronological structure built up for Mesopotamia (Chart I). We can equate ceramic types of Bampur I-IV, Shahr-i Sokhta I, Khurab, Tal-i Iblis 5-6, and their established parallels, to the ceramics ofYahyd IVC.29 Such a chronological revision

Fig. 3. Ceramics, above Period IVC, below Period VA, B.

28

29

Compare K. Frifelt, Ibid, fig. I7:A and fig. I2 from Hafit, Oman and fig. 4 from a Hili tomb, Abu Dhabi with our fig. 4:A,B. The combination of suspension lugs and incised " roping " at the point of carination are also common to E. D. I ceramics in the Diyala as well as Jemdet Nasr, see Delougaz, Ibid, pl. 32,d. Identical vessels of this type are also at Tal-i Ghazir in Stake io, Room. Here they appear in what McCown called the Proto-Elamite levels. They are well illustrated and discussed by D. S. Whitcomb in " The Proto-Elamite Period at Tal-i Ghazir, Iran ", thesis submitted to the University of Georgia for the A. M., Athens, Georgia, 1971. See B. de Cardi, Excavations at Bampur, A Third Millennium in Settlement Persian Baluchistan, 1966 (Anthropological Papers of American Museum of Natural History, vol. 51, pl. 3, 1970),

fig. 34:326 and compare with our fig. 3:C; fig. 30,5 with our fig. 3:B, or fig. 25,257, fig. 22,147; fig. 30,20 with our fig. 3:F. For our fig. 2:B see parallels also at Shahr-i Sokhta II/III, M. Tosi, " Excavations at Shahr-i Sokhta, Second Preliminary Report ", East and West I9 (1969), Nos. 3-4, fig. 33,e. Our fig. 2:A-C also find ready parallels at Shahr-i Sokhta (Dr. Biscione personal communication). The animal horns on fig. 3,E are precisely paralleled at Bampur (fig. 36,81, fig. 29,300) as well as at Khurab. The complete vessel, fig. 3:F, is also directly paralleled at Khurab. Additional parallels to sites in the borderlands and Period IVC ceramics are discussed and illustrated in Lamberg-Karlovsky and Kohl, Ibid, 1971.

MESOPOTAMIA

AND

THE

INDO-IRANIAN

BORDERLANDS

97

places the above sites ca. 3000 B.C. or several hundred years earlier than the present chronological

structure has allowed.30 The presence of seals, sealings, and Proto-Elamite tablets as well as associatedJemdet Nasr ceramics all converge toward a late fourth and very early third millennium date for Yahya IVC and the related sites of Bampur I-IV, et. al. The well-worked-out ceramic parallels presented by B. de Cardi for Bampur remain the same-however, the date for the beginning of this sequence (and its related sites) must be pushed back to the end of the fourth millennium. Lastly, we illustrate a clay figurine of a human head found in Period II context which finds its closest parallel in the clay figurines of Namazga III (Fig. 4:G).31 PeriodV Period V, with approximately 250 square metres of architecture cleared, consists of three major superimposed building levels. Understanding of the upper two levels, Period VA and B, was expanded in 1971. Numerous ceramic types are continuous from Period V into Period IVC, changes being quantitative rather than abrupt. Architecture of VA, B consists of large rooms (ca. 2 5 x 2 metres) some with hearths against the walls and with adjoining smaller rooms. One room in Period VB contained large painted storage jars (black paint on red) another complex of rooms was built about a large (I .5 metres in circumference) circular storage bin with four dog-kennel type entries at the base of the " silo " providing access to the grain (?) from four different rooms. Within Period V we have also found numerous fragmentarybowls of uncarved steatite and alabaster. From the floor of a room of Period VB we recovered the largest single piece of lapis lazuli from Yahya-a large, incised and perforated bead (Fig. 4: H). This bead, together with a tubular bead from Period IVC and a single small tubular bead from the surface, complete the very limited sample of lapis lazuli found at Yahyd. Several square stone and round clay seals from Period VA and B have been found. We illustrate one here (Fig. 4: C) which finds an almost identical parallel at Tal-i Bakun III.32 From Period VB we have recovered another major piece of sculpture-again a ram (P1. IIIa). This piece together with the magnificent alabaster ram recovered in 1970 and the remarkable steatite female figurine from Period VI attest to a high artistic accomplishment in stone sculpting throughout Periods V and VI. In fact, we cannot think of another site of the early fourth and late fifth millennia in Iran which has yielded as many outstanding pieces of large stone statuary.33 The ceramics of Period VA and B can be paralleled on many surface sites throughout southeastern Iran and at the excavated sites of Tal-i Iblis, Periods II, III,34 Susa A and Tal-i Bakun II. Of special interest is the bichrome ware (less than I %) which appears to have been an import ware in Period VB. It is an extremely well-made and carefully decorated pottery which finds no ready parallel on other sites. The majority of the pottery throughout Period V is a coarse, chaff-tempered ware and a blackon-red painted ware.35
s3 See for example M. Tosi, " A Tomb from Damin and the

Problemof the BampurSequencein the Third Millennium", Mesopotamian demand centres, and that at Namazga III there EastandWest Nos. 1-2 (1970). Here the authorpresentsthe are signs on their figurines resembling Elamite, see V. M. 20, dates of BampurI-IV:2400-2 00oo, V-VI:2 o0o-1800 Masson V. I. Sarianidi, " O0Znakh na Sredneaziatskih statuetB.C. This kah epohi bronzy ", Vestnik DrevniiIstorii (1969), pp. 86-99. chronologicalscheme can only be accepted by arguing that YahyAIVC is 600 years later than its Mesopotamian Jemdet 32 Compare our fig. 4: C with A. Langsdorff and D. E. McCown, or Nasr counterpart rejectingthe numerousparallelswhich tie Tal-i Bakun A, (O.I.P. LIX, 1942), pl. 82,14; L. le Breton, " Note sur la C6ramique Peinte aux Environs de Suse et a Yahyy IVC to BampurI-IV, et. al. Suse ", M.D.P. XXX (1947), fig. 44,11; for a close parallel on 31 I would like to thank Dr. R. Biscione for pointing out this a site east of Yahyd, at Hrarj, Persian Baluchistan, see A. parallel. His directexperiencein handling the materialsfrom the excavationsin Turkmenistan, both in the museumsand on Stein, Ibid (I937), pl. XXX. the sitesallowedhim to draw parallelsbetweenthe YahyAIVC 33The alabaster ram is illustrated in C. C. and Martha Lambergceramics and those of Namazga III. Dr. Tosi has already Karlovsky, " An Early City in Iran ", ScientificAmerican224, indicated Shahr-i Sokhta II/III relations with Namazga. No. 6 (1971), pp. 102-1I1; the female figurine is published in With the relations to Shahr-i Sokhta I and Yahyd IVC the Lamberg-Karlovsky and R. Meadow, "A Unique Female relations of Yahyd to the sites in Turkmenistanbecome less 23, Figurine ", Archaeology No. I (1970), pp. 12-17. of a surprise. We have already suggestedthe possibilityof a 34 See A. Stein, Ibid, pl. IX, Sar 4, from Qal'a-i-Sardagah and its north-southas well as the expected west-eastpattern of comexact parallel in our fig. 3: M. For Tal-i Iblis see J. R. Caldmunicationin the borderlands at and well, Investigations Tal-i Iblis (Illinois State Museum, 1967). (Lamberg-Karlovsky Kohl, Ibid, x97i). We might note here that Russian archaeologists 35 For illustrations of Period V ceramics see Lamberg-Karlovsky, have argued that urbanizationin Turkmenistanwas brought Ibid, 1970, pls. 31-34, figs. 36-43.

about in part by Elamite tribes in search of resources for

98

JOURNAL

OF

PERSIAN

STUDIES

Fig. 4. Small finds from Periods IVB, C and VA, B. Scale r: r.

MESOPOTAMIA

AND

THE

INDO-IRANIAN

BORDERLANDS

99

borderlands Mesopotamia, Yraydand theIndo-Iranian In the past archaeologists have tended to explain the rise of urban centres in the Indus and in the borderlands as a result of diffusion from Mesopotamia. Furthermore, the urban sites of the IndoIranian borderlands were all dated to the early and mid-third millennium, with the rise of Harappan cities dated to ca. 2500 B.C. Excavations at Tepe Yalhyt have challenged both the adequacy of this model of diffusion and the late appearance for urban centres in this large area. Protoliterate, urban, Tepe Yahyt IVC (1500 years after the initial settlement at Yalhya) can clearly be dated to the Late Shahr-i Sokhta I in southern Iran and the closely worked out parallels for these sites with Umm an-Nar, Barbar I, Hili, Hafit in the Persian Gulf, Damb Sadaat I-II, early Kulli, Nal, Anjira IV, Mundigak I, II, Amri II (Baluchistan borderlands), Namazga III in Turkmenistan date the above sites to the late fourth and first centuries of the third millennium.36 With the presence of earlier settlements on most of the above cited sites one can see a respectable antiquity and development in the proto-urban settlements in each of the above areas. It becomes difficult now to pass off this area as a cultural backwater stimulated only by diffusion from Mesopotamia in Early Dynastic and later times. We prefer to adopt a more complex model which sees southeastern Iran, the Persian Gulf, Baluchistan, Turkmenistan areas a separate " interaction sphere "37co-ordinating its own specialized functions, manufacturing its own resources (as steatite at Yalhyd or metals at Tal-i Iblis) or transhipping merchandise and goods (lapis lazuli at Shahr-i Sokhta, ores through Bahrein) and establishing its own administrative centres and social organization. Such a view would support the multilinear view of urbanization in each of the above areas without the primacy of one over the other. There is an increasing body of data to support the contention that ca. 3000 B.C.all of the above " interaction spheres " were in some form of relationship with each other. This model supports none of the above areas as nuclear, none as peripheral. For example, Mesopotamia is not the nuclear area generating all development in the peripheral zones, i.e. southeasternIran, rather it is the economic dialectic between bothareas which stimulates the developing complexity of southeastern Iran andMesopotamia. This systemic view eliminates the false dichotomy of central vs. peripheral in an attempt to understand the similar and/or different processes which led to the synchronic development of both.38 With the present evidence for a settlement of the complexity of Tepe Yahya IVC, only 800 kilometres from the centre of the later Indus Civilization, and some 500 years prior to its assumed start, it is time to stop " explaining " the rise of urban places in the borderlands and in the Indus as resulting from the " crossing from Mesopotamia to the Indus " of the urban phenomenon. The processes (i.e. large scale inter-regional trade) which led towards urban development in the borderlands are already evident by 3000 B.c. and cannot be accounted for by the diffusion and adoption of the Mesopotamian urban " stage ". Clearly, the situation is more complex! With the initial recognition of contemporary and shared urban processes in the above areas (Mesopotamia, Turkmenistan, borderlands, Persian Gulf, et. al.) we are now in a position to turn to the study of the specific processes which led to the urban growth of each and their economic, political and social organizational configurations. Clearly, one might expect the urban process in each to be somewhat different. Equally clearly, diffusion from Mesopotamia as the single causal factor in the rise of urban centres is not satisfactory. The Yahyd Project now moves into its second phase-having established a relative chronology for the immediate area we begin intensive survey, analysis of settlement distributions, selective sondagesat
different sites, all towards the understanding of the local urban process of southeastern Iran.39
36The important comparative stratigraphy of the borderlands as worked out by G. Dales continues to stand. We might suggest that his Phase D be placed as contemporary to the Late Uruk/ IVC, see G. Dales, "A Suggested Jemdet Nasr of Ya.hy Chronology for Afghanistan, Baluchistan and the Indus " in in R. Ehrich, Chronologies Old WorldArchaeology (University of Chicago, 1965). S37For a definition of this useful concept see, J. R. Caldwell, "Interaction Spheres in Prehistory " in Hopewellian Studies, Illinois State Museum Scientific Papers 12 (1964), pp. 135-143. 38 A model is presented which sees trade as a major factor in the urban process in C. C. Lamberg-Karlovsky, " Trade Mechanisms in Indus-Mesopotamian Interrelations ", Journal of AmericanOrientalSociety (in press). 39 Limited survey was begun in 1970. In the Jiroft, Tepe Nurabad, reported by Stein (Ibid, 1937), was found to be entirely a Period IVB, C settlement. In the area about Dolatabad, 30o km. from YalhyA to the southeast, we met our greatest surprise, thirty-seven Period V localities were located, some mounds, some surface scatter only. A dense settlement distribution of Period V is indicated.

Uruk/Jemdet Nasr horizon, ca. 3400-3000 B.c. The contemporaneity of Yahyd IVC, Bampur I-IV,

the Indus and Mesopotamia as already separate polities, ca. 3500-3000

B.C., with each of the above

100

JOURNAL

OF

PERSIAN

STUDIES

CATALOGUE OF ILLUSTRATIONS
Illustration Fig. I: A B C D E F G
H

Period II II II IIA II IIA II

II

Locus' ANI Io ANI 9 ANi 9 A 3o A-ANI 9 A 23 ANi io

ANI

Description Red and black paint on buff Maroon and black paint on buff Maroon and black paint on buff Buff ware Buff ware Buff ware Buff ware Orange/red paint on buff Black on buff Black on buff Buff ware Buff ware Buff ware Buff ware Buff ware Buff ware Buff ware Buff ware
Black on buff Maroon slip over buff

K L M N O P

Q
R S
Fig. 2: A

IIA IIA IIA IIA IIA II II IIA IIA II IIA

IVC

A 23 ANi 14 AW Io A 23 A 28 A8 ANI 9 A 27 AW/ANWI 6 ANI 7 AW/ANWI 6

B 4-10

o, 1+ -2

B C D E F G H Fig. 3: A B C D

IVC IVC IVC IVC IVC IVC IVC IVC IVB IVB IVC

E
F G H I

IVC
IVC IVC IVC VB VA VA VA VA VA VA IVC IVC VA IVC IVC IVB IIA ??
VB IVC

BW TT, 5-2 XCE 4-9 XCE TT, i2e B 29 BTT, 2-1 B 4-8 B 4-8 XCE TT, 13 XC 7 XC 9 BW TT2 3A

K L M N O

XCE TT, 12B B TT, I B 4-8 XCE TTI, 4B XC TT, 5 XCE TT1 I2A XC TT2 5 XCE TT, 6A XCE 14C C I BW/CW 8-1 BW/CW TT3 XCE 14-3 BM 3-3 B 4-8 XB TT, 15 A 23-5
2

XCE TT, I2A

Black on buff Black on buff Black on brown Black on green/buff Black on grey Plum slip, black paint, buff slip Brown slip, black paint, buff slip Buff ware, brown slip Black on plain red, brown slip Black paint on buff Black paint on buff Black paint, orange slip Black paint on grey Black on buff Black on buff Black and red paint on buff Black paint on plain red Orange slip over plain red Black on buff Black on buff Black on plain red Black on buff Cylinder seal, glazed steatite Cylinder seal, glazed steatite Perforated clay stamp seal Proto-Elamite clay tablet Proto-Elamite clay tablet Copper-bronze stamp seal Clay figurine, broken at neck
Incised lapis lazuli bead Steatite bead

Fig.4:

A B C D E F G
H I

XCE TT, 15-45 BM 2-2

Area of excavation = B, ANx, XCE, etc., TT = Test Trench; followed by strata number and feature number.

", P1. L Faded,greenglazed,faience " baboon Godof Wisdom,Thot. Ht. 8"2 cm. PeriodI. AN 4-6.

Pl. Ha. Steatite complex.L. 15*8, ht. 11'3 cm. PeriodIVA?: XBE 4-2. plaque, architectural fafade of temple-palace

Pl. IIb. Plain red waresherdwith Harappan typesealing. L. of sealing1.4 cm. PeriodIVA: XB TT1 2A.

Pl. IIa. Stone sculpted ram's head. Max. ht. 13* 2 cm. Period VA: XCE TT2 14-30.

bqffstoragevessel,incised" DancingLady". Ht. offigure 17 2 cm. PeriodI. AN1 TT1 2-I. P1. Illb. Plain orange

Pl. IV. PeriodIVC building from the southwest. complex

ELAMITES, ACHAEMENIANS AND ANSHAN By John Hansman

ELAMITES The precise location of the city and district of Anshan in ancient Iran is an outstanding historical problem. Anshan is first attested in Sumerian and Akkadian texts of the late third millennium B.C. Elamite rulers of the second and first millennia traditionally took the title King of Anzan1 (Anshan) and Shushan (Susa). Still later Anshan became the home district of the early Achaemenids who were styled Kings of Anshan in Neo-Babylonian inscriptions. Yet despite this prominent role in the political history of early Iran, different scholars tend to associate Anshan with different regions; while the city of Anshan has never been found. The writer submits, however, that by carefully comparing the evidence of early texts with that of archaeological reconnaissance,it would seem possible to put forward a more plausible solution to these problems. The earliest associated reference to a possibly Elamite state in Iran is found in the Sumerian King List. This text records that Enmenbaragisi, King of Kish (c. 2700 B.C.), " carried away as spoil the smitten with arms and its kingship taken to the Elamite state of Awan.3 A King of Kish, in turn, is reported to have smitten Awan and to have taken its kingship to Kish.4 Somewhat later, around 2500 B.c., the First Elamite dynasty of which we have record, was founded in Awan by a certain Peli.5 The dynasty of Peli seems to have flourished for several generations until Sargon of Agade (23342279 B.c.) invaded Elam and sacked a number of districts of that region, including Awan and Susa.6 Rimush, son and successor of Sargon, continued to war in the territories of Awan, Susa and Elam.7 ninth king of Awan.8 The dynasty of Peli eventually ends with the fall of Khita's successor KutilInshushinak in about 2220 B.C.9 It is perhaps significant to note that Gudea, a ruler of Lagash (c. 2220 B.C.), claims to have conquered the city of Anshan in Elam.o1 A comparison of relative dates would indicate that the downfall of Elamite Awan occurred during the same period when Gudea, who subdued Anshan, was still in power." The possible association of the two events suggests that Anshan may indeed have been the chief city of the district of Awan. The sudden disappearance of Awan could then be explained in a perfectly reasonable historical context. This proposal, however, requires further substantiation.
9 All Elamite chronology used in the present study follows that of W. Hinz, Das Reich Elam (Stuttgart 1964). Mesopotamian of this word. Similar renderings are found in late Elamite chronology is adapted from J. Brinkman, " Mesopotamian texts but the usual Elamite orthography is anzan. The phoneme Chronology of the Historical Period ", in A. Oppenheim, indicated by z is unknown. AncientMesopotamia(Chicago I964). 2 T. 10 Barton, RISA, 184. Jacobsen, (ed), The Sumerian King List (Chicago 1938), 83-5. On the date of Enmenbaragisi cf. M. Rowton, C.A.H. I I (1970), Gudea is generally placed in the last quarter of the twentyxl third century B.c. Hinz (cf. n. I2) suggests that the Kingdom 65-6. 3 Jacobsen, ibid., 95. of Awan probably fell a victim to the anarchy brought in by the Gutian invasion of Mesopotamia. Yet it is Gudea who 4 Ibid., 95-7. invaded Anshan at about this time and who commanded 5 V. Scheil, " Dynasties 61amites d'Awan et de Simal ", Revue XXVIII (I931), 1-3. tribute from the more easterly regions of Magan, Meluhha d'Assyriologie and Dilmun. Gadd proposes that these political successes of 6 G. Barton, The Royal Inscriptionsof Sumer and Akkad (New Haven 1929), 113. Gudea suggest that a part of his reign " fell in the period after 7 Barton, RISA, 124. the final defeat of the Gutians " (C. J. Gadd, " The Dynasty 8 F. K6nig, CorpusInscriptionum Elamicarum. I Die altelamischen of Agade and the Gutian Invasion ", CAHI (3rd edn.), fac. 17, Texte. Tafeln (Hanover 1923), no. 3. On the identification of 46). However, we learn from a text of Ur-Nammu, founder of Khita in this text cf. G. Cameron, Historyof Early Iran (Chicago the third dynasty of Ur (cf. S. Kramer, " Ur-Nammu Law on 1936), 34. [continued nextpage
1 A Babylonian Akkadian inscription of Manishtusu of Agade (2269-2255 B. c.), giving an-sa-an,attests the earliest occurrence

weapons of the land of Elam ".2 A further passage states that the Kingdom of Ur (c. 2600 B.c.) was

Naram-Sin (2229-2255 B.C.), a later ruler in Agade, concluded a treaty of alliance with Khita the

101
10B

102

JOURNAL

OF

PERSIAN

STUDIES

Throughout much of its earliest history Elam constituted a political union of several separate provinces each with a governor subject to a central viceroy accountable to the Elamite king.12 The most important of these districts was Awan where the king resided. As we have seen, Awan of the Sargonic annals, is mentioned in association with Elam and with Susa; but Anshan is not attested. An Akkadian text of the late Assyrian period refers to Elam and Anshan, but not to Awan, as regions conquered by the first Sargon.13 Manishtusu (2269-2255 B.c.) son and second successor of Sargon in Agade speaks of resubjugating Anshan after the ruler of that place revolted from the empire created by Manishtusu's father; but he makes no mention of Awan.'4 During the two succeeding centuries, as we shall see, Anshan is always presented as an important district of greater Elam. It also takes precedence in the title of later Elamite rulers, i.e. Kings of Anzan and Susa. The name Awan during these more recent periods, on the other hand, is not recorded except in one instance when it is referred to specifically as a land.'5 In summary then, the written sources show that both Awan and Anshan are closely associated with Elam and the Elamites in the earliest historical phase, some times contemporaneously but never in the same text. The kings of Awan of the Old Elamite period become the kings of Anzan and Susa of later dynasties. The political divisions of Awan, Susa, Elam and Simash, which were contemporary with the dynasty of Awan, were continued in the districts of Anshan/Anzan, Susa, Elam and Simash of the succeeding dynasty of Simash (Simashki). Available evidence would seem to suggest, therefore, that Awan might represent a small political district within a greater geographical region associated with the name Anshan, and that possibly with the expansion of this state, the name of the smaller area (Awan) was gradually replaced by that of the larger (Anshan) in the political sense and when used in royal titles. The district of Simash located to the north of Susa (Fig. I) was apparently next in importance to Awan during the Old Elamite kingdom, for after the fall of the dynasty of Awan (c. 2220 B.c.), it was in Simash that a new line of twelve Elamite rulers appeared.e6 Very little is known of the first kings of Simash. Most of the surviving texts of this period refer only to adjoining Elamite regions. Besides invading Anshan, Gudea of Lagash, for example, claims that residents of Elam and Susa assisted in the building of a temple to one of his gods.'7 Shulgi (2095-2048 a ruler of the third dynasty of Ur, married one of his daughters to the ishshak or governor of B.c.), Anshan.18 The association would seem to imply vassalage status for Anshan, as was the case with Susa during this period; though in later texts Shulgi speaks of laying waste to Anshan.'9 A temporary peace was apparently re-established when Shu-Sin, son and successor of Shulgi, like his father gave a daughter in marriage to a governor of Anshan.20 This state of concord was, however, not to last. In about 2021 B.c.-a few years after Ibbi-Sin, a son of Shu-Sin succeeded to the throne of Ur- the King of Simash
Hinz, CAH (3rd edn.), fac. 19, 6. 12 OCf. " '1 A. Albright, A Babylonian Geographical Treatise on Sargon of Akkad's Empire ", JAOS XLV (1925), 244. 14 Barton, RISA, x28-30. 15 Cf. n. 21. RA V-VI; Cameron, History of Early Iran, 47, 155. '6 Scheil, 17 Barton, RISA, 218. 18Thureau-Dangin, Die Sumerischenund Akkadischen Kdnigsinschriften(Leipzig 1907), 230o. 19 Thureau-Dangin, SAK, 231-7. 20 C. Virolleaud, " Quelques texts cun6iformes in6dits ", ZA XIX (x9o5-o6), 384-

continued page] from previous Code ", Orientalia N.S. XXIII (0954), 45), that this ruler overthrew Nammakhni, the fourth successor to Gudea in Lagash. It is also attested that Ur-Nammu replaced his known contemporary Utu-Khegal of Uruk as the major political " Babylonia c. power in southern Mesopotamia (cf. Gadd, 2 120-I8oo ", CAH I (3rd edn.), fac. 28, 3-4). Before his fall Utu-Khegal had expelled Tirigan, last of the Gutians from Sumer (F. Thureau-Dangin, " La fin de la domination gutienne ", RA IX (1912), III ff.). Gadd suggests that UtuKhegal and Tirigan were active in about 2120 B.c. The association of events listed above, however, indicates that Nammakhni the fourth successor to Gudea must have been ruler in Lagash at about the time of the Gutian expulsion.

Gadd recognized the problems of chronology which would arise if Gudea should be placed after Nammakhni, for the independence displayed by Gudea contemporary with the sovereignty of Ur would be hardly conceivable (Gadd, CAH I, fac. 17, 45). The evidence seems to suggest that at some time during the reign of the earlier Gutians (described by Gadd as feeble and sporadic) Gudea must have asserted paramount political control in the areas which he seemed to have invaded -though the Gutians may have regained a nominal control over the dynasty of Lagash at a later period. A recent study of early Mesopotamian chronology suggests 2230-2000 B.C.for the Neo-Sumerian dynasty at Lagash of which Gudea was the second ruler. Cf. M. Rowton, " Chronology ", CAH I, pt. I (3rd edn.) (1970), 219.

..

........

..ISFAHAt

SUSA

s:'.::::i:~i~i~:::~il~i.............."'~ ::::dw::;;;:::x:?:-:
r

'''~riii~ii~i~i~a ii
. . . .

L~iii~i~i~~i:I::::XH
H

AZi

TEPE
W

?lP;ii~v
4

un

...............: (DUR-UNT A.-ISO- -aMA.-

l~ i~i ti~ijli~iji ili ~i lij ijl~i i ij~i ~ij rg~pil~:~:::::::':::::~i~i~i~~iBi~~~~~~li~~li i? id

CHOGA
I

AMBIL

jm1?~~?~,,

a~~?c-?~~i~~f~~~iizi'~'ijii'ii~~~ i~:i~jj~i .......... l~i?)il.' ~~ ~f'"::"::.''':''' ........ . ;A ::;::;:?~??::::?.::.::~i:/::.~/Sixjxiii..................iiiii ....... 7~j?:i i~~i~~i~l~i~i~ ~iaa~~? ...~?::;i:%::~:::~i'~ ~c:::::~ li~... ~.~~li~~iiiij ............. ii~~j......... ...~i~ ...................
A D
Pii : aX~i i ~i i~i.....

l ad- ii.??:t; ".I"OI;:LBC~E~~lg~.! -::::::::::::--:?:a~w~ "d""'." :::??:ii~i~i:.

ip e 310)AL i~:I;x

~
u:: ..............::::::~:::::-

i???ii~i?~::i~i~ii~s:...........:: ........X~j~~i~~!:iI~lii~~ii
:,~..............ISAHA

gadmii.xc

.......iiia

..............:i::;:::::: ........... .......

N ~..............
N
z!: i

':;:::i' :i ~i"' , .............~.~.;'''~: ~r:i:::~i

l
' '?: ' r:

~ l~i~i~ ~ ~ ~ ~ ~ii~~3i~ii~i' ................... ~ ~ i:~l~? . ~~~bii~i~:!:i~i~:,i:::: ii~jiX~liiiiiii~i~iji~~~~:~~~i ~ OX~C~ii

TS

......... .

...........::

...i~Bi

lj~liiiii~~iliiiiiiiiiX

::::::::::::::n::::::::::i~i.:i~~~i................L':~::

i''::~:3iii~i.................W~~ii

......

SASRAAB N...
JD~i

A...........A

Ktilkt

SACAR

RUil i~lii~i~~~i~ii~iilZ
N
la~i',
i Di~

SCALE.~iiiiiiijiiaiiij~ii~~l~liadii~j~iiii
0
50
MILES.

PR E S

100
0
MAIN
j: :I~iji~i~ i ~i~ :

150~i:il:i~l
zlal:i rli i i i~

10

w 20

40

60 10

:7=::Z::::fzz,=j
,?I

~'':'''~":ii~l
:;Ci

iiiiiiiii~iiiiiii~iiiii~iE~iiiiii ..iiBlli~li;i~:iiiiisiiiiiiiijii:

.:i~~i~: f~~~~~iili~::ii~iiii ~ ~iki~~~?::?::

5USHIRE::~i::::~iii~~~i~

C~ii
? ? . ?:

6 FI:?::?:? U
L

ii~i~is~Xiii~..........ii'a

~~

TRACK$:~:::::::::

ALE. 0 10 20 40mik60Aw0 100LIA

IABbL

PER

LIYANiaia

?Ia1

? : :; :

U L F

Iran. The insertat rightequalsareaof Fig. 2. Fig. I. Map of south-west

104

JOURNAL

OF

PERSIAN

STUDIES

had occupied the land of Awan (Anshan?) and Susa in Elam. By 201o7 B.c. Ibbi-Sin was to retake
much of this territory;21 but his success proved only temporary for we learn that within a few years the

Elamites had waged a successful military campaign against Ur. Following his defeat, the last King of Ur, Ibbi-Sin, was carried off to Anshan together with a statue of the Sumerian moon-god, Nanna.22 A generation later Gimil-ilishu, second King of Isin brought back Nanna, the God of Ur, from Anshan.23 Finally, we learn that Gungunum fifth King of Larsa boasts of military victories in Anshan24(c. 1928 B.c.). The text of Gungunum is important to our study for it contains the last mention of Anshan, Although the paramount Elamite rulers during the last century of the third millennium seem to have resided in Simash, the texts of this period continue to give a great deal of prominence to Anshan. This suggests that at that time Anshan still played an important role as a traditional political centre in Elam.
district or city, to be found in Mesopotamian sources for over 1300oo years.

Indeed it is possible to consider that the ascendancy of Simash in Elam was only occasioned because of the long occupation of the proposed earlier Elamite capital Anshan in Awan by Mesopotamian forces. It is perhaps of some significance in this regard to remember that when Ibbi-Sin of Ur was defeated by Political strife among the various Babylonian states gradually weakened the hold of Larsa in south

the Elamites he, together with the statue of Nanna, a trophy of victory, were carried off not to Simash, as one might expect, but to Anshan (Awan), the older capital.

Iran and a new line of Elamite kings was eventually able to re-establish local rule there. Eparti, the founder of this dynasty (c. 1890 B.C.), was also the first known Elamite to call himself King of Anzan
and Susa.25

or Eparti's heir Shilkhakha (c. 1870-1840 B.C.) styled himself, in addition to king, sukkal-mah grand regent, a Sumerian appellation which suggests a government not entirely independent of Babylonian suzerainty. During this period the title sukkalor regent of Elam and Simash and sukkalof Susa are also commonly used.26 The sons of the ruling sukkal-mah normally filled the office of the two on sukkal,though inscriptions show that the sukkal-mah, occasion, would hold all three titles. It is interesting to note, however, that throughout the approximately 3oo-year rule of the Eparti kings, we have no record of a sukkalof Anshan. This suggests that Anshan consisted at that time of a district and subject entirely to the jurisdiction of the sukkal-mah.Hinz proposed that the sukkal-mah the sukkal of Susa were both resident in the city of Susa (Fig. I), but allows that a relationship of this sort in the same town could have caused political tension.27 Indeed, any decree of the Elamite king which might apply to the political district of Susa required ratification by the sukkalof Susa. Nevertheless, the case for Susa would seem to be the most plausible. This is especially true when it is remembered that Anshan is hardly mentioned in Mesopotamian or Elamite texts during the whole of the Eparti dynasty A except as used in the conventional title of the sukkal-mah. political decline of the older capital together with that in Simash is strongly suggested. Susa, on the other hand, which appears not to have developed during the ascendency of earlier Elamite dynasties into a major independent political centre, was certainly to do so with the rise of Eparti and his successorsin Elam. Numerous inscriptions found at and by various of the Susa attest to the building activities undertaken there by different sukkal-mah sukkalof Susa. The extensive excavations made by Ghirshman in the " Ville Royale " at Susa have shown that much of this very large quarter of the ancient city was first built upon in the twentieth century B.c.28 We should remember that it is not long after this time that Eparti established a new dynasty in Elam, presumably at Susa. The available information therefore, suggests that the main political centre of Elam may have been moved from its traditional location at Anshan to Susa within the
period associated with the expansion there.
21

L. Legrain, Business Documentsof the Third Dynasty of Ur, Ur ExcavationTexts, III (London 1928), no. 1421. 22A. Falkenstein, " Die Ibbisin-Klage ", WeOr I (1947-52), 379 and 382. 23 Gadd and Legrain, Royal Inscriptions: Ur Excavation Texts I (London I928), no. ioo. 2 L. Matou', " Zur Chronologie der Geschichte von Larsa bis zum Einfall der Elamiter ", AOr XX (1952), 304 if. 2 Scheil, " Documents et arguments ", RA XXVI (1929), 1.

On the founding of the Eparti dynasty cf. Hinz, " Persia ", CAH (3rd edn.), fac. 19, 7. On the date of Eparti cf. W. Hinz, " Eine Altelamische Tonkrug-Aufschrift von Rande der Lut", A.M.I. IV (1971), 22-3, note 6. 26 On the terms cf. sukkal-mah Cameron, HistoryofEarly Iran, 71-2; and Hinz, op. cit., 3-5.
27

28 R. Ghirshman, " Susa Campagne de l'Hiver 1965-1966 ",

Ibid., 5.

Arts AsiatiquesXV (1967), 4-12.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

105

Different rulers in southern Mesopotamia appear to have maintained at least an intermittent suzerainty over succeeding Elamite kings until Babylonia fell to Kassite and allied invaders from the North in about 1593 B.C. Cuneiform texts continue to mention Elamite rulers until the final quarter of the sixteenth century s.c. It is not, however, known whether these later governments in south-west Iran were subject to the Kassite alliance or, indeed, if the Kassites eventually put an end to the house of Eparti. After 1520 B.C. we have no further record of the Elamites for over 200 years.29 An apparently independent Elamite dynasty reappears suddenly on the historical scene during the last half of the fourteenth century B.C. The earliest known ruler of this new line to adapt the old title King of Anzan and Susa was Attar-kittah (I310-1300 Bs.c.).30 Inscriptions of the dynastic predecessors of Attar-kittah are not known, on the evidence of published texts, to exist and no foundation or rebuilding dedications of this king or of his immediate successor Humban-numena (I300-1275 B.C.) were found at Susa. Only at Liyan located near the centre of the Bushire peninsula, an island-like extension of the coastal plain of modern Firs (Fig. i), have inscribed bricks of Humban-numena been recovered. The Liyan texts commemorate the construction by this king of a religious sanctuary at that site.31 It is important to note here that numerous inscriptions found on stelae, statues and on foundation bricks from temples at Susa and at the religious centre of DurUntash (now Choga Zambil), are attributed to the son and successor of Humban-numena, UntashB.C.), who probably established his chief residence at the latter city which bears (d)GAL (1275-1240 his name. The absence of any significant inscriptions of the father of Untash- (d) GAL at Susa or elsewhere in Khfizistan prompted Labat to submit that the capital of Humban-numena may have been located in the province of Anzan.32 The evidence, indeed, suggests that the dynasty represented by Humbannumena arose in the east of Elam, perhaps while Susa was still under foreign occupation, and that the Elamites only re-established political control in parts of western Elam, including Susa, during the reign of that king or, more likely, in that of his successor Untash- (d) GAL. Equally, the documented association of Humban-numena with Liyan, as Labat suggested, indicates that Anzan/Anshan may have included that area. We might at this point of our enquiry move a step further and consider Herzfeld's suggestion that Anshan is to be equated with much of the Ostan of modern Fdrs33 (Fig. i). Certainly we have evidence that different parts of this district remained under Elamite control for many centuries. An inscribed alabaster cylinder recovered during the excavations at Liyan is attributed to the Elamite king Simut-wartash (I772-1770 B.C.).34 At Liyan also were found dedicatory bricks of four later Elamite kings, those of the already mentioned Humban-numena and of Shutruk-Nahhunte B.C.). All B.C.), Kutir-Nahhunte (II55--II50 (1185-1155 B.C.) and Shilhak-Inshushinak (1150-1120 of these rulers styled themselves Kings of Anzan and Susa.35 An Elamite rock relief dating in part to the seventeenth century B.c., and in part to the Neo-Elamite period of the seventh century B.c., survives at KurangCin near the Fahliyiin plain (Fig. i) in the hills of western Fars.36 Inscribed bricks of the later Elamite ruler Shutruk-nahhunte (I185--1I55 B.C.) have been found at Tulaspid37 on the Fahliyin plain (Fig. i) close to where the main road from Khizistan to southern Firs passes through a series of connecting valleys. Lastly it should be noted that the relief of Bahram II at Naqsh-i Rustam (Fig. 2) has been worked over pre-existing Elamite rock sculptures. These earlier reliefs, like those at Kuranguin, date partly from the seventeenth century B.c. and partly from the Neo-Elamite period.38 We therefore seem to have firm evidence of a long-surviving bridgehead of Elamite influence, and no doubt evidence
29

On this phase cf. Hinz, Das Reich Elam, 82-3. During the past five years numerous tablets and two stone inscriptions have been recovered by Dr. Ezat Negahban from the important Elamite site of Haft Tepe in Khfizistan (Fig. i). This material has been dated to the second half of the second millennium B.C. None of these texts are as yet published. On the excavations at Haft Tepe cf. Negahban, " Haft Tepe ", Iran VII (1969), 177. " Elam c. R. 30so Labat, B.C.", CAH (3rd edn.), fasc. 16, i6oo--2oo 8. 31 F. K6nig, CorpusInscriptionum Elamicarum. I Die altelamischen Texts (Hanover I926), no. 40.

33 E. Herzfeld, The Persian Empire (Wiesbaden I968), I78-9. en 3 Mimoires de la Mission arche'ologique Perse, XV, 91. 35 G. Hising, Die einheimischen Quellen zur GeschichteElams. I. Teil: AltelamischeTexte (Leipzig 1916), 51, no. I9; 57, nos. 29-31. 36 P. Amiet, Elam (Paris 1966), 386. These estimates are based on a comparison of the reliefs with similar material and a seal from Khfizistdn which can be dated by inscriptions. 37 Herzfeld, op. cit., 177. 31 HUsing, op. cit., 84 ff., nos. 60-5.

32 Labat, op. cit., 8.

106

JOURNAL

OF

PERSIAN

STUDIES

of Elamite political control in widely separated areas of Firs. Herzfeld's attractive theory of equating Anshan with the present Ostan of Firs will be further considered in a later section of our examination. We turn now to the Middle Elamite dynasty which succeeded that one associated with Humbannumena. One successorto Shilhak-Inshushinakin this later line of rulerswas a Khutelutush-Inshushinak menirrather than sunkir(king) of Anzan and Susa.39 Menirpossibly implies a federal authority rather than a divine right to rule and may indicate a lessening of the political control exercised by the holder of this title.40 The reign of Khutelutush-Inshushinakwas to witness a devastating invasion of Elamite lands by Nebuchadrezzar I (1124-1103 B.c.) of Babylonia,41 an event which marked the effective
(I 20-11
o10 B.c.).

Unlike his several predecessors, however, Khutelutush-Inshushinak

is titled only

PosorgodueZn

,fgi TOr

04$

%rvoonn .Nr?

((
/

KM. lo MAIN RuTRA ILES.10

10

2
Tept

C.-------.0 "ItJ4
Doarvaza

10

SAIN

TRACK.

-----

Fig. 2. Map showing the location of Maliviin.

end of Elam as an independent military power for nearly three centuries. Only in 821 B.c. do we again hear of the Elamites, who are then found allied with the Chaldeans against the Assyrian king Shamshi-Adad V.42 Although Elam did not preserve its political unity during these years of obscurity, local chieftains must have maintained a semblance of control in areas where their authority had been traditionally exercised. Cameron suggests that local rule would probably have remained strongest in the remote eastern districts of Elamite territories, sc. the region of modern Fdrs.43 Even so, despite their isolation from the aggressive Mesopotamians, the inhabitants of southern Firs were eventually to suffer new pressuresfrom Iranian migrants moving down from the North. Whether or not the coming of the Iranians caused the divided provinces of Elam to support the rise of a new
39

Labat,
21.

" Elam and Western Persia CAH ", (3rd edn.), fac. 23,

40 Loc. cit. 41 R. Thompson, Reportsof the Astrologersof Nineveh and Babylon

(London 9goo), no. 200, rev. 5D. Luckenbill, AncientRecordsof Assyria (Chicago 1927), I, no. 726. 43 Cameron, History of Early Iran, 156.
42

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

107

centralized authority in south-west Iran is a question which cannot be answered on present evidence.
Whatever the reason by c. 742 B.C. Humban-nikash I had become king of an apparently reconstituted

Elamite federation.44

the old title Great King of Anzan and Susa.45 This ruler entered into alliance with the Chaldeans against both Sargon II and Sennacherib of Assyria; his military efforts, however, were not successful and he was eventually replaced in Anzan and Susa by a younger brother, Halludush-Inshushinak (699-693 B.C.). The new Elamite leader invaded Babylonia and temporarily held parts of it from the Assyrians,but Sennacherib quickly retook most of these territories. Halludush-Inshushinak,meanwhile, was deposed in Elam and replaced by his son Kudur-Nahhunte (693-692 B.C.). The short rule of Kudur-Nahhunte is of interest to our study in that from his reign onwards, no Elamite head of state is known to have assumed the ancient royal title King of Anzan and Susa.46 Such an omission would suggest a loss to the Elamites of at least the area of Anzan/Anshan either by Kudur-Nahhunte or by his immediate predecessor. Shortly after the reconquest of Babylonia, Sennacherib invaded the Elamite territorieswhich lay to the north of Susa. As a result of this campaign, the weakened Kudur-Nahhunte was forced to flee and seek refuge in the mountains of Hidalu (present Kuhgiluya, Fig. i). The succeeding king in Elam, Humban-numena (692-687 B.C.), renewed the old political alliance with Babylonia and sought military assistance from a number of neighbouring districts, including the lands of Anzan and Parsuash.47It is important to remember here that Anzan is now definitely treated as a separate territory and not subject to centralized Elamite authority. What happened in Anzan at this time is bound up with the political and territorial relationship of that place with Parsua, Parsuash and Parsamish/Parsuwash,and of those same regions with the Assyrian empire. To understand these associations better, we must establish the location of Parsua and that of the similarly-named Parsuash. Assyrian texts of the ninth century B.c. give the earliest reference to the land of Parsua. Scholars generally agree that this district, which was invaded by Shalmaneser III, is to be found to the north of Kermanshah and south of Lake Urmia.48 variant rendering of Parsua)49and show it to have become by that reign an established province of the Assyrian empire. When Sargon subdued six cities further to the south, in the area of Kermanshah and northern Luristdn, he placed these new territories under the governors of Parsuash.50 Most probably it was these former subject people of an expanded Parsuash who, under that name, were identified as allies of Humban-numena against the armies of Sennacherib. At some time subsequent to the formation of this alliance a major clash with the Assyrianstook place at Halule in Mesopotamia. Sennacherib claims a considerable victory over the joint Babylonian/Elamite forces at the battle We know very little about the immediate successorsof Humban-numena in Elam. Rival claimants to the central kingship, some favourable to the Assyriansand some opposed to them, seem concurrently to have controlled various parts of the old domains. There were local rulers in Susa and at Madaktu52 to the north; at Hidalu, which bordered on Fars; and in other areas. Yet none of these regional chieftains was able to gain a lasting political control over all Elam.
44 Ibid., I57; Luckenbill, ARA II, no. 8Io. 45 Scheil, MDP V (1904), no. 84. 46 Cf. Cameron, ibid., 158-65; Hinz, Das Reich Elam, 1i6-24. 47 Luckenbill, ARA II, no. 252.

The successor to Humban-nikash was a nephew, Shutruk-Nahhunte

II (717-699

B.C.),

who took

Two later Assyrian kings Adad-nirari III (8io-783 B.C.) and Tiglath-pileser III (744-727 B.C.) also campaigned in Parsua. Inscriptions of Sargon II (721-705 B.C.) identify the region as Parsuash (a

(c. 692

B.C.);

Babylonian texts record a more inconclusive result.51

Ibid., I, no. 581. For the location of this Parsua, cf. E. Wright, " The Eighth Campaign of Sargon II of Assyria ", JNES II, to 178, n. 26 and map 176; also L. Levine, Contributions the Historical Geographyof the Zagros in the Neo-Assyrian Period (Doctoral thesis, University of Pennsylvania 1969), 152. 49 It has been suggested recently by some scholars that Parsua cannot be the Assyrian province of Parsuash. But the sources possibly suggest otherwise. Sargon II mentions Parsua only
48

once (as a land he subdued) in a general text which deals with his overall military conquests. In other texts he identifies Parsuash as a dutiful Assyrian province which he expanded by six cities (ARA II, no. 56) and where he received tribute from the rebel leaders of neighbouring districts (ibid., no. 146). Whereas Sennacherib, the successor of Sargon speaks only of Parsuash, the predecessors of Sargon record only Parsua. The two names would seem to be the same. 50 Luckenbill, ARA II, no. 56. 'x Cf. Cameron, Historyof Early Iran for sources and discussion. 55 On the location of Madaktu cf. Cameron, ibid., i65.

108

JOURNAL

OF

PERSIAN

STUDIES

Periodic attempts were certainly made to reform the old pan-Elamite alliance, notably by TemptHumban-Inshushinak (668-653 B.C.) of Susa. However, these efforts met with only a passing success. Tempt-Humban was eventually defeated by an army of Ashurbanipal and several districts in Elam which had been overrun were thereafter placed in the control of local chiefs whose support the Assyrians claimed.53 These supposed loyalties did not last, for a year later, Huban-nikash III, vassal governor in Madaktu, supported a new uprising in Babylonia against the Assyrians. The rebel Elamite suffered an inglorious rout by enemy troops at Der and fled to the mountainous district of Hidalu54 to seek aid from that region and from the people of neighbouring Parsumash.55 A revolt in part of Elam at this time, however, caused Huban-nikash to be replaced as king by Tammaritu (651-649 s.c.), who continued local resistance against Assyria. Tammaritu urged Hidalu and the adjoining district of Parsamash (Parsumash) to send troops in support of his campaign. The people of Parsamash apparently did not respond at once and the delay seems to have proven costly to the allies. By 649 B.C. BEl-ibni, Assyrian governor of lower Babylonia, was able to report to Ashurbanipal that most of the Elamite lands up to the borders of Parsamash had been captured, presumably by Indabigash, an Elamite who was loyal to the Assyrians and whose pursuit of Tammaritu caused the latter to fall into the hands of There now follows a final confused period of Elamite history. Local chieftains continued to manoeuvre for control in the disunited realm while the Assyrians and their allies frequently intruded into Elamite lands. Still deeper advances by these hostile foreign forces could not be indefinitely warded off. By 640 B.C., Huban-haltash III, the last King of Susa, was compelled to flee to the mountains, while Ashurbanipal boasts of devastating twelve districts of Elam up to the borders of Hidalu.57 ACHAEMENIANS An Assyrian text relating to the final destruction of Elam mentions a king of Parsuwash named Kurash. Scholars recognize this Kurash as Cyrus I of the Achaemenid line. Cyrus offered submission to Ashurbanipal and sent his son to Nineveh as a testimony of good faith.58 It is with this reference that the House of Achaemenes first enters the historical record. In a Neo-Babylonian text Cyrus II (the Great) gives to his grandfather Cyrus I the title " Great King of Anshan ".59 It would appear then that the first Cyrus was political chief in Parsuwash and also ruler of the former Elamite province of Anshan/Anzan, which we have previously associated with the district of Fars. The two lands are possibly identical.60e Yet when did the Achaemenians accede to kingship in Fars? According to the evidence of relative dates and royal titles, the Achaemenids were governing Anzan/Parsuwash at least a generation before Ashurbanipal commenced his decisive invasion of Elam. As we have seen, the earlier alliance of Anzan and Elam against Sennacherib (692 B.C.) occurred during the reign of the Elamite king Humban-nimena, this being shortly after the rulers of Elam abandoned the old title King of Anzan. Cyrus I (of Parsuwash) recognized the supremacy of Ashurbanipal at the time, or shortly after Assyrian troops had overran the districts of Elam (c. 640 B.c.). As Cyrus II gives
53 A. Piepkorn, HistoricalPrismInscriptions Ashurbanipal of (Chicago 1933), 70 ff. "4 Hidalu is most probably to be identified with the mountainous district extending to the north-east of Behbehan in southeastern Khfizistdn and south-west Firs which is now identified by the name Kuhgiluya. This would fit the requirement for Hidalu to be a mountainous region located on the border of Parsumash (Firs). The suggestion by Hinz (ZDMG N.F., XXXV, 250) that the chief town of Hidalu is to be equated with the modern settlement of Behbehan would not seem feasible, as no cultural remains at the site of Behbehan nor any which are presently recorded from the surrounding plain can be dated to the period in question. However, it is entirely possible that the district of Hidalu could have included the plain of Behbehan. L. Waterman, Royal Correspondence the Assyrian Empire (Ann of Arbor 1930), II, 4x10-12. 56 On this sequence of events cf. Waterman, op. cit. II, 300-o3 and Cameron, History of Early Iran, I85 ff5 Luckenbill, ARA II, no. 8o8. " Die alteste Nachricht uiber das persische 58 E. Weidner, Konigshaus ", AFO VII (i931), 459J. Pritchard, AncientNear Eastern Texts (Princeton 1955), 316. 60 Parsuwash is possibly an Assyrian rendering of OP Parsa, which relates specifically to the district of Firs. Anshan remained the traditional name in south Mesopotamia for the region of Firs down to the Neo-Babylonian period. Herzfeld (The Persian Empire, I7o) recognizes that Anshan and Parsuwash, in the texts cited, refer to the same area.
'

BEl-ibni.56

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

109

both his grandfather Cyrus I and his great grandfather Teispes the title " King of Anshan ",61 the first Cyrus must have already been ruler in Anshan at the time of his contacts with Assyria. Thus according to the relative chronology of these events, we have a period of just over fifty years between the first alliance of Anzan/Parsuwash with Elam and the submission of Cyrus I. This length of time could fall well within the limits of the possible reigns of two successive Achaemenid kings. Although Achaemenes (Old Persian Hakhamanish) is usually recognized as eponymous founder of the Achaemenian royal house, it is his son Teispes (OP Chishpish) who is first called " Great King of Anshan ". Thus it would appear more probable that Teispes, rather than his father Achaemenes, established Achaemenid rule in Anshan/Parsuwash/Parsa and became the ally of Elam against Sennacherib.62 It is not likely, on the other hand, that the land of Parsuash, allied with the lands of Anzan, Pasheru, Ellipi and Elam in this campaign, is to be associated with Parsuwash or with the Achaemenian Persians. The lands of Parsuash and Anzan, being under the same local rule, would hardly need make separate alliances with the Elamite king. Lastly it should be noted that the presence of w/m in the toponyms Parsuwash/Parsumash which we have associated with the area of Achaemenian Parsa, suggest that the medium through which these forms developed was Babylonian. M and w are interchangeable in Babylonian and one or the other consonant would probably have been expressed in rendering Parsuash in that language. Therefore while Parsuash, Parsuwash and Parsumash may ultimately derive from a common root, there appear to have been several different geographical areas identified with this name. The first (Parsuash) is to be located in Luristdn and further north where the name possibly passed directly into Assyrian. The second (Parsuwash/Parsumash) is to be placed in south Iran where the name perhaps reached Assyria by way of a southern Mesopotamian form. It is generally agreed that Teispes divided the Achaemenid domains between his two sons Cyrus I and Ariaramnes, with the former ruling in the western part and the latter in the eastern portion. Cyrus II grandson of Cyrus I appears to have reunited the Persian realm by putting an end to the independent reign of Arsames son of Ariaramnes. It will perhaps be useful for the development of our enquiry now to consider the association of Cyrus II with Anshan and with Pasargadae. The so-called chronicle of Nabonidus, last independent ruler in Babylonia, relates that Ishtumegu (Astyages, last King of Media) was defeated (c. 550 B.c.) by Cyrus (II) King of Anshan, and that Cyrus brought booty from the campaign to Anshan.63 Nicholas of Damascus in the epitome of Ctesias states that these spoils were taken to Pasargadae.64 Strabo65 and to Strabo in commemoraQuintis Curtius66 relate that Cyrus built his palace at Pasargadae-according tion of his victory over Astyages. The ruins of Pasargadae are situated some 96 km. north of Shirdz in Firs (Figs. I and 2). It is, of course, generally accepted that the modern Ostdn of Fars represents, in part, the Achaemenian homeland of Parsa.67 Whereas the Nabonidus Chronicle refers in one passage to Cyrus II as King of Anshan, in another passage Cyrus is identified as king in Parsu (an Akkadian rendering of OP Parsa).68 Thus as we have seen in the case of earlier references to Anshan and to Parsuwash, Anshan was also considered at this later period a part of the province now called Fdrs.69 Some scholars have equated Anshan only with the mountainous regions which form the natural border between eastern Khtizistdn on the one hand and the south-western Isfahdn district and north-western Firs on the other.70 Yet while the district of Anshan possibly extended over much of this mountainous border
62

Cf. note 59. e61 It is, of course, possible that Achaemenes ruled a part of

69 70

Cf. n. 60.

Anshan without having the title " king ". On the genealogy of the Achaemenian line, cf. R. Kent, Old Persian Grammar (New Haven 1953), I58-9. 63 Pritchard, ANET, 305. 64 C. MUller,Fragmenta Historicorum GraecorumIII (Paris 1828), 406. 65 Strabo XV.3.8. 66 Quintus Curtius V.6.Io0. 67 On Parsa/Fars, cf. Frye, The Heritageof Persia, 40, 46, 75 and Cameron, ibid., 31, n. 28. 68 S. Smith, BabylonianHistorical Texts (London 1924), xo0 if.

A summary of past theories on the location of the district of Anshan are given by Cameron in ibid., 31, n. 28. Hinz
I9, Anshan lay in the mountains of " Western Persis ". Husing that the valley of the Karkheh/Saimarra river located suggests to the north-west of Susa may have been the centre of Anshan (cf. Cameron, above). This contention, however, presupposes that the Assyrian Parsuash, which is in part to be located in the same area, was the Achaemenian homeland at the time of the Persian Kings of Anshan. This can hardly be the case since the on [continued next page (" Persia ", CAH (3rd edn.), fasc. 6) supports the view that

110

JOURNAL

OF

PERSIAN

STUDIES

country, the evidence of our sources indicates that the greater area of the later district of Parsa (modern Fars) was also included. The change in name most probably occurred when the Achaemenian Persians, after their migration to Anshan, transferredthis ethnic of Parsa to their new home, in accordance with their custom of calling a country by the broadly-applied name of their own tribal association.71 This regional identification of Parsa with Anshan was no doubt gradually taken over by the native residents there. Darius styles himself" King of Parsa " on all three versions of the Behistun inscription.72 The name Anshan, however, is more tenaciously preserved at one point in the Elamite text of the Behistun account, but only as an unspecified area in Parsa.73 Classical authors do not refer to Anshan at all, but only to Persis, the Greek rendering of Parsa. By the coming of Alexander, Anshan appears to have been completely superseded, in surviving texts at least, by Parsa and Persis. What then of the old Elamite city of Anshan ? Where was it and how long did it survive as an occupied site? To begin with we know that Anshan could not have been located on the plain of Pasargadae, for only the barest cultural remains are to be found there which antedate the period of the Cyrus constructions. Nor indeed do we find any very extensive cultural debris on the plain at Pasargadae contemporary with the palace and citadel buildings there which might represent a refounded city of Anshan, of early Achaemenid date, later to be called Pasargadae. The essentially rural nature of Pasargadae is indicated by Anaximenes of Lampsakoswho interprets this name as meaning " camping ground of the Persians ".74 In a more recent explanation, however, Dr. Ilya Gershevitch has argued that the Greek Pasargadae may be a rendering of the name of the tribe to which the Achaemenian " dynasty belonged. Old Persian *pasragada- those who wield strong clubs " is given by Dr. Gershevitch as a possible solution.75 This theory suggesting a transfer of a tribal name to Cyrus's capital relies on the statement of Herodotus, that the Achaemenian clan belonged to the Pasargadai tribe.76 Indeed, it is possible to consider that the plain of Pasargadae had been a favourite camping ground of the Achaemenian clan long before Cyrus began his more permanent constructions there. This point perhaps gains emphasis when it is remembered that the early Achaemenians in Parsa were part of a larger Indo-Iranian tribal community which had migrated to this region from more northerly areas. It is well-recognized that the later Achaemenid kings were seasonally resident in various palaces at Susa, Ecbatana and Persepolis; and considering their tribal origins there can be little doubt but that this practice of moving with the season reflected a long-established tradition of the Achaemenian line. The plain of Pasargadaeis bitterly cold in the winter and often excessively hot in summer. Whereas later Achaemenian kings traditionally held the great spring festival of Nau Riiz at Persepolis, it is possible that their dynastic predecessorCyrus II resided at his palace at Pasargadae also mainly during the spring, when the weather there would have been at its best. The Achaemenian following no doubt would have established a temporary settlement of tents in their seasonal camps, as do many of the migratory tribes of Iran today. We should not then expect to find extensive Achaemenian remains of a permanent nature other than those of palace, religious and defensive structures on the plain of Pasargadae. Because of these strong nomadic traditions it is equally unlikely that an extensive Achaemenian settlement existed at the old site of Anshan.77 This point inevitably returns us to the question of the location of the city of Anshan. While certainly not at Pasargadae, the site would also not appear to be found at Istakhr, a later capital of Fars. Extensive archaeological soundings made by Schmidt in the
This view is expressed by Herzfeld, The PersianEmpire, 186. L. King and R. Thompson, The Sculpturesand Inscriptionsof Behistun (London I907), OP Parsa, I; Elamite Parsin, 93; Neo-Babylonian Parsu, I6o. " The Old Persian Text of the Bisitun Inscription ", 73 Cameron, StudiesV (1951), 50. Journal of Cuneiform "74 Stephanus Byzantinus (quoting Anaximenes), s.v. Passargadai. 75 I. Gershevitch, "Iranian Nouns and Names in Elamite Garb ",
72 71

Transactions the PhilologicalSociety(1969), 168. of Herodotus, I.125. 77 Cameron, Historyof Early Iran, 31, n. 28; I8o and 218, suggests that the early Achaemenians were kings of the city of Anshan. But the Neo-Babylonian texts, where this title is attested, give " only the form King of Anshan ". There is no occurrence of the word " city " in relation to Anshan in any text of the Achaemenian period.
76

continued page] from previous Achaemenians were already settled in Parsa/Fdrs at least from the time of Teispes, the first Persian to take the title King of Anshan. W. Nagel in Berliner Jahrbichfiir Vor-undfriihgeschichte

VI (1966), 9, and Taschenkarte I, has attempted to reconcile this impasse by proposing two Anshans, one in the Samarra plain and one in the mountains of western Firs. This is an accommodation but not a solution.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

111

great mound of Istakhr, which is located about 5 km. to the north-west of Persepolis (Fig. 2), produced no material of pre-Achaemenid date.78 Perhaps then we must look further to the south for our lost site. This possibility will be considered in the following section of our enquiry. ANSHAN It is well established that major urban settlements of Mesopotamia during the third millennium B.c. were situated in plains where adjacent agricultural lands provided crops to supply the needs of these larger concentrations of population. It would seem therefore to follow that, as was the case with Susa, the other major cities of early south Iran, including Anshan, were located on the agriculturallyproductive alluvium and not in the barren, economically-poor uplands of the Zagros Mountains which even today are sparsely populated. The most fertile regions in the Ostan of Firs at the present time are the plain of Mary Dasht and the bulaksof Tavdbi, Khafrak Sufld, Khafrak Ulyd, Ramjird and Baydd which surround it (Fig. 2). The high concentration of early sites to be found in this area, hereafter to be termed the Mary Dasht basin, indicates that it enjoyed a similar prosperity in antiquity. We have already submitted that these plains comprised a part of the former Elamite province of Anshan/Anzan in south Iran. Hence if our proposal of association is valid, we might possibly find remains of the old city of Anshan in this neighbourhood. The first map of archaeological sites of the Mary Dasht basin was completed in 1937 from photographs taken during aerial reconnaissancecarried out in co-operation with the University of Chicago.79 In 1951 and 1952 ProfessorL. Vanden Berghe completed two extensive archaeological surveys of that region. This work describes the pottery to be found at many of the sites earlier recorded and on numerous additional sites located during the ground surveys.80sArchaeological soundings in depth made at selected mounds in this area by Vanden Berghe have enabled a relative sequence of pottery forms and styles of the successive cultures of the Mary Dasht to be established.8' The largest site of the earlier periods examined by ProfessorVanden Berghe is Tepe Sauz which is located immediately to the south of the modern town of Mary Dasht (Fig. 2). This mound covers an area of approximately 9 should expect to find at Anshan. We turn next to the bulakof Baydd, an area which was not surveyed by the University of Chicago team nor by Vanden Berghe. The Dasht-i Bay<djforms a level, natural extension to the west of the Dasht-i Ramjird and thus of the Mary Dasht basin itself (Fig. 2). The plain, which measures approximately 25 by 25 km., is protected on the north, west and south by foothills of the Southern Zagros and on the east by an extensive though seasonally-dry area of marsh land. The district of Bay<d;is exceptionally fertile since it receives water from numerous springs which rise in the surrounding hills. In about 1960 members of the staff of the Fars Department of Antiquities from Shiraz put down a series of archaeological soundings at a large site called Maliyan (locally Maliyfin) which is located at the eastern end of the Dasht-i Baydd (Fig. 2). As far as the writer has been able to learn, these soundings have not yet been published and no further excavations at the site have been undertaken up to early
1971. The writer first learned of the site at Tell-i Maliyin from an entry in the Fdrsndma-iNd]siri. This reference, dated 1886, states that the former chief town of Bay~dl was Maliyain. According to our source, there exist there " scattered, numerous broken bricks, which indicate its populous state, and lofty buildings. Up till now in certain places the wall of the city still remains. Inside the ruined city they practice agriculture. The village beside the city they call Maliyain ".s2 In 1970 Mr. William Sumner,
On the excavations at Istakhr cf. E. Schmidt, The Treasuryof Persepolis(Chicago 1935), 105-09. 9 Ibid., 138-9 and fig. 97. 8soL. Vanden Berghe, " Archeologische navorsingen in de omstreken van Persepolis ", JaarberichtEx OrienteLux (195354), 395-408.
78 81

hectares (30oo X 300oo m.). Yet it can hardly be identified as the remains of the major city which we

Ibid., 400-05; Vanden Berghe, On the Trackof the Civilizations of AncientIran, Memofrom Belgium, CIV (September-October I968), 14. 82 Mirzd Ilasan Fas,'i, Fdrsndma-i Ntsiri (Tehran 1313/0895), I83 (in Persian). The section treating Maliyiin was completed in 1304/1886.

112

JOURNAL

OF

PERSIAN

STUDIES

Executive Secretary of the American Institute of Iranian Studies in Tehran, described to the writer a number of archaeological sites of the Mary Dasht basin; among these was that of Maliyfin. Between 1967 and 1970 Mr. Sumner had been engaged on a comprehensive archaeological survey of the whole of this region including the bulllk of Bayod. During the course of his survey Mr. Sumner visited Maliyiin on several occasions. The cultural remains there were described as extensive but confused in layout. Mr. Sumner confirmed the existence of the embankment walls attested in the Fdrsndma-i Ndsiri. He also stated that most of the pottery which he had collected at Maliyfin seemed to date from the Kaftari culture of the third millennium B.c. At the time of this discussion Mr. Sumner had not had the opportunity to examine aerial photographs of the site. Some weeks later the writer located two useful prints. The photographs showed the embankment walls at Maliyfin to be nearly rectangular in form (P1. I and Fig. 4). Measurements calculated from these prints indicated that the total area enclosed by the embankments was approximately 3oo hectares. The texture of the land within the enclosure, as revealed by the photographs, suggested that much of the north-eastern quarter of the site contained only isolated cultural deposits whereas the remaining three-fourths of the area hectares) showed an uneven, scarred surface which is usually to be associated with occupa(about 2oo00 tional debris (P1. I). If as Mr. Sumner allowed, the pottery of Maliyu-n was mainly of the third millennium, this would seem an exceptionally large settlement for such an early period in south Iran. Although no historical identification for the site had as yet been proposed, it was thought by the writer that the possibility of an association with the lost Anshan might bear further investigation. The potential importance of Maliyfin seemed all the more significant at the time. Indeed, the deposit of the third millennium to be found there appeared to be nearly five times as extensive as the estimated comparable occupation levels at Susa, the hitherto largest known site of this period in south-western Iran.83 Further judgement on the identification of Maliyin, however, was reserved by the writer until the late summer of 1970 when he was able to visit the site. The plain of Baydd on which Maliyfin is located may be approached from the main road which runs between Isfahan and Shiraz by either one of two routes. One road diverges to the west from the highway just south of the town of Zargan at the southern edge of the Mary Dasht basin (Fig. 2). A gravel track at this point runs north-west for 36 km. to the town of Tell-i Bayod. From here the track leads north for 6 km. to Maliyin. The second and more difficult route leaves the highway to the west immediately to the south of the Pul-i Khan, a bridge which crosses the Kur river just south of the town of Mary Dasht (Fig. 2). The first section of this alternative approach road runs 36 km. to the north-west and is paved. The traveller then turns right onto a rutted and dusty dirt track which curves in towards the Dasht-i Baycd, first to the west and then to the south. After a further 26 km. on this uneven stretch, the site of Maliyiin is reached. The remains at Maliyfin are decidedly impressive. Though heavily eroded the embankment walls still rise at many points to a height of between 3 and 4 m. above the level of the surrounding plain (P1. IIa). On the north-west and south-west the walls are well defined. This definition is less marked on the remaining two sides where the fill of the embankments seems to have been eroded into a gentle slope and because the cultural debris rises here to the uppermost level of the embankments. The average depth of deposit within the enclosure is perhaps 3 m. as measured from the plain. There are, however, higher points within the site which rise above this level (Fig. 4A, B and C). As the aerial photographs examined above seemed to indicate, about one-fourth of the area within the walls which borders on the north-eastern, and in part on the south-eastern embankments, was nearly level with the plain and did not appear to contain very much cultural material. A part of this lower area is shown in P1. IIb; the full extent is shaded on Fig. 4. It will be noted that much of the corresponding area on the aerial photograph represented in P1. I is dark in colour. This is because the land represented is low enough to be irrigated and therefore to produce the crops which give the darker hue. By contrast the higher level of the site can be used only for dry farming in the wet season because there is no practical
81The cultural 85 hectares. much of the early second deposit at Susa covers an area of approximately However, Ghirshman has shown (cf. n. 28) that " Ville Royale " was not built upon until the millennium B.C. Miss Elizabeth Carter of the

University of Chicago, who has recently studied this problem at Susa, estimated to the writer that the deposit of the third millenniumthere coversan area of approximately hectares. 55

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

113

way for the local farmers to supply water to those parts. In the foreground of P1. IIb can be seen a part of one of four isolated cultural mounds which are to be found within the lower level of the site at Maliytin. This mound is identified with the letter D on Fig. 4. Other mounds are marked E, F and G. The archaeological soundings put down by the Fars Department of Antiquities some ten years before the writer visited Maliytin correspond with the points indicated by the letters A, B and C on Fig. 4. The present village of Maliyfin from which the site seems to have taken its name, is located at about the centre of the south-eastern embankment. The village, surrounded by a mud-brick wall, is built partly upon the embankment. This location can be seen on P1. IIa. The same photograph shows a wellpreserved section of the embankment. The walled garden of the village extends for some distance into the area of the older site, mostly on low ground (P1. I and Fig. 4). The surface of the deposit at Maliyfin is thickly littered with fragments of baked brick and broken pottery (P1. IIIa and b). Most of the sherds examined there were of two successive cultures. These have been designated the Kaftari and Qal'a periods by Professor Vanden Berghe. The Kaftari culture is characterized by three distinct types of ware. Examples of each ware were recorded at Maliyfin. These are: (a) unpainted pottery of medium texture with a red slip (P1. IV A-B); (b) pottery of medium texture with a red slip decorated with geometric designs and parallel lines in black or brown paint (P1. IV C-D); and (c) pottery mostly of medium texture which ranges in colour from salmon-pink to terra-cotta and cream. This ware is decorated in brown paint with parallel lines (P1. IV F-G), and/or with different geometrical patterns. These include stylized flying birds (P1. IV E), hatching (P1. IV G), diamonds (P1. IV H), interlacing squares (P1. IV H-I), vertical lines (P1. IV I) and triangles (P1. IV J).84 The succeeding Qal'a culture is represented at Maliyiin by characteristically wellmade spherical vessels of medium texture which vary in colour from salmon-pink to yellow-cream. Decorations in brown paint are found only on the shoulders of these vessels.85 The patterns include parallel and waving horizontal lines and parallel short vertical lines (P1. IV K-L). Vanden Berghe gives the period of the combined Kaftari and Qal'a cultures as approximately 3500 to 2000 B.C. but he admits that this estimate remains uncertain.86 It is not therefore unexpected that certain modifications have already been proposed. In 1969 and 1970 Dr. Murray Nicol of the Semitic Museum, Harvard University, carried out extensive excavations at Darvaza Tepe which is located in the bulak of Tavabi immediately to the south-east of Mary Dasht (Fig. 2). These excavations have produced a number of carbon samples in a stratified sequence of the Shughd culture. This succeeded the period now identified with Qal'a ware. Dr. Nicol has advised the writer that a series of consistent carbon-14 dates which these samples produced, suggest that the horizon for the Shughd culture should be brought back to about the early nineteenth century B.c.87 This modification of a previously uncertain chronology, would also bring forward the terminal date of the preceding Qal'a culture to some time in the early twentieth century B.C. Therefore if current evidence is to be accepted it would appear that ancient Maliyiin was already largely abandoned by the early second millennium B.C. However, as Qal'a ware is far less in evidence on the surface at Maliyiin than is Kaftari ware, and as there is still a possibility that these two wares may overlap, it is conceivable that the terminal date of the Kaftari should also be advanced. Two series of qandt-shafts which cross Tell-i Maliytfin (Fig. 4 and Pls. I and IIIa) enabled the writer to study the stratification of the deposit there in depth at a number of points. A part of the site which seems to centre around area A on Fig. 4 and another small area produced sherds of Qal'a pottery in thin layers. At all other points investigated on the main site and also at the isolated mounds D and G and along much of the embankment wall, the Kaftari ware predominated. Because some of the qandt shafts were badly eroded the writer was able to climb within several of these at widely separate points. In each case the two-and-a-half to three metres of deposit studied produced sherds of Kaftari pottery. Equally an investigation of the old archaeological sections still to be seen in areas A, B and C,
84

On Kaftari ware, see Vanden Berghe, Archiologie de l'Irdn ancien (Leiden 1959), 42; for examples of painted red-slipped ware, see pl. 51 b; for unslipped painted ware see pl. 5xa. 85 On Qal'a ware, see ibid., 42 and pl. 52b and c. 86 Cf. n. 81 (second reference, 12).

87 This informationwas kindly supplied to the writer by Dr.

Nicol in a letter datedJanuary i9th 197I. Dr. Nicol statesthat his earliest carbon-14date to be associatedwith the Shughd ware is I820 B.c.i+ 6o, uncorrected for solar disturbances, which means add c. 5o-1oo years upwards.

114

JOURNAL

OF

PERSIAN

STUDIES

showed, except for the upper-most level at one location in area A, sherds of the Kaftari sequence. The earliest ceramic remains recorded by the writer at Maliyin were perhaps sherds of bevelrimmed bowls, a type of ware which continued in use for many centuries during the third millennium. Sherds of an interesting vessel of a somewhat later period were noticed at about half-a-metre below the surface in one of the old archaeological sections at area A. The sherds formed the profile of a cylindrical vessel 30 cm. in height, with a button base, a slightly bulbous body only 8 -5 cm. in diameter and a vertically-extended rim (Fig. 3). The vessel had been improperly fired for the core was burnt black. The surface colouring was terracotta. The texture appeared uniformly coarse and showed a straw temper. This jar is characteristic of a family of Elamite vessels found at Susa and at Choga Zambil in Khfizistin which, in similarlydeveloped forms, are usually dated to the late second millennium B.C.88 The example at Maliyin would seem to attest a limited reoccupation of a part of that site in later antiquity. Fragments of the bases of two more of these plain-ware vessels were recorded on the surface in area A. The question would now seem to arise: Can we relate the evidence of the pottery to be found at Maliyin and the size and location of that site with what we should expect, according to the events and indications recorded in historical texts, for the city of Anshan ? To answer this we will need to review briefly some material presented in the first section of our study. It will be remembered that Awan is the most prominent district of the middle third millennium to be mentioned in texts of that period from Mesopotamia. The kingship of Ur was taken to Awan before 2550 B.C. Some 250 years later Sargon of Agade conquered Awan. While Manishtusu, a son of Sargon, tells of resubjugating Anshan after the ruler of that place broke away from the empire created by his father, he makes no mention of Awan. Indeed, from this period onwards the name Awan disappears almost completely from contemporary texts. Now we have at Maliyain a cultural site of some 300 hectares which is surrounded by an impressive wall-system. The evidence of stratified sherds to be found in the qandtshafts at numerous points on the
88

Cf. Ghirshman,

Mimoires de la De'ligation Arche'ologique en Iran

Fig. 3. A jar of the late second millennium B.C. from Maliypn.

XL, Tchoga Zambil II, especially pl. LXXXVIII,

G.T.Z. 893.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

115

site show that a deposit which contains mostly Kaftari ware is recorded to a depth of from 2-5 to 3 m. throughout much of this area. There must necessarily be a further depth of deposit extending below these higher levels which cannot as yet be examined but which certainly relates to the earlier Kaftari and to the preceding culture characterized by bevel-rimmed bowls. A conservative dating of the Kaftari sequence (Vanden Berghe suggests 3500-2500 B.c.)89 would place an urban settlement much the size of the present site at Maliyfin by at least the middle of the third millennium B.c. A city of this size at such an early date in south-western Iran seems unique to Maliyin. Indeed, recent studies, it will be remembered, indicate that even the settlement of the third millennium at Susa covered only approximately one-fifth the area of the one of similar date with which we are presently concerned.90 It would not seem inconsistent with the evidence, therefore, to relate what is perhaps the largest early site in this part of Iran to Awan, the most important district of the same region and period. Equally, the proposed identification of Anshan as the capital of Awan would perhaps best explain why Awan and Anshan are both associated with the conquests of Sargon of Argade but never mentioned in the same text. This solution would also help to explain why the Elamite districts of Awan, Susa, Elam and Simash of the first Elamite period were apparently represented as Anshan/Anzan, Susa, Elam and Simash during the succeeding dynasty of Simash. Finally, it is to be noticed that within a gap in the hills at the eastern limit of the Dasht-i Bayodl,a gap which leads to the Kur river valley beyond (Fig. 2 and P1. IIb), there is an area and a village called Avanjan (pronounced locally Avangiin). This may represent an ultimate derivative of the earlier Awan. We have shown that the fall of Kutik-Inshushinaklast King of Awan (c. 2200 B.c.) possibly coincided with the capture of the city of Anshan by Gudea of Lagash. From this period onwards Anshan (in Awan) seems to have been subjected to a varying degree of suzerainty by successive kings of Ur. At the same time a new and apparently independent Elamite kingdom arose in Simash, a district usually placed to the north-east of Susa. The loss of the Elamite kingship from Awan and the political instability which appears to have beset Anshan during the late third millennium can be related to the evidence at Maliyin. The Qal'a sequence of pottery which attests the last major phase of occupation at that site is to be dated on present knowledge from about 210oo to 19oo00 B.c. However, Qal'a ware is far less common at Maliyiin than is the earlier Kaftari pottery. The former is found only in shallow deposits in certain areas there, especially in area A, but not at all over much of the site. The evidence suggests that the ancient city at Maliyuin was largely abandoned within the period represented by the late Kaftari culture and that the site was almost totally deserted at some time in the period represented by the Qal'a sequence. The disappearance of the dynasty of Awan and the subsequent domination of Anshan by Mesopotamian rulers would fall within the period of the major abandonment of our site. Gungunum, fifth king of Larsa, gives the last mention of Anshan to be found in Mesopotamian texts for over 1300 years. This is in c. 1928 B.c. when Gungunum boasts of his conquests there. In terms of the evidence of the pottery of Maliyfin, it is possible to consider that most of the settlement which still survived at that site in the Qal'a period could have been abandoned during the earlier phase of this culture, i.e. at about the time of the invasion of Anshan by Gungunum. It is perhaps also of significance to remember that extensive excavations by Ghirshman in the so-called " Ville Royale " area at Susa show that much of that quarter of the ancient city was first built upon during the same century as Gungunum's conquests in Anshan. This indicates that Susa sustained a considerable increase in population at that time. Could some of these new residents have come from the old Elamite capital of
Anshan and thus have further contributed to the abandonment of the latter site? If we associate Anshan with Maliyfin this would seem a possibility. Yet whatever the circumstances at Susa, our assessment does show that beginning with the nineteenth century B.c., Anshan falls into a long period of total obscurity. As we have seen, Eparti, founder of a new dynasty in Elam (c. 1890 B.c.), calls himself" King of Anzan and Susa ". It is generally agreed that the capital of his dynasty was Susa. Although sons of the ruling Elamite kings traditionally took the titles sukkal of Elam and Simash and sukkal of Susa,
89

Vanden Berghe, On the Track of the Civilizationsof AncientIran. Memofrom Belgium (September-October 1968), 12.

90 Cf. n. 83.

~G

.A

Q~ie

W.

/
N

LlYD

/
/

KEY.
/ // AREA LIMITED OF CULTURAL DEPOSIT

SCALE.
KM. 0 0-25 0.50 0-75 1

Fig. 4. Siteplan of Malziyn.

Pl. L Aerialviewof Malyvin.

Pl. Ia. The walled village of Maliygn viewedfrom the south: to the right is a length of the south eastern embankment. The modern village is built partly upon this embankment.

Pl. IIb. A part of the low-lying quarterwithin the site at Maliypn, most of which does not appear to contain occupational debris. In the runs right and left in the foreground is one of several isolated moundsof this area (cf. Fig. 4, point D). The north-easternembankment middle distance. In thefar distance (near the village of Avanjan) is the gap which leads to the Kir river valley beyond.

nearpoint A. The total area viewed in this photograph P1. IIIa. Maliyi7n; looking south-east over the sitefrom the north-westembankment up to the far wall of the village of Maliyin, seen at the distant left, is deeply overlaid in cultural deposits. A series qofqanit channels crosses the site in the middle distance. Numerousfragments of brokenbricks can be seen littering the ground.

Pl. IIIb. Maliyin; looking north-eastfrom the south-westernembankment. Cultural debrisfills all the land visible in this print up to the walled village of Maliyin seen in the distant right. The contourswhich cross the deposit in theforeground representrecentattempts to dryfarm this area of the site.

A Pl. IV. Pottery theKaftariculture(letters to J) and of the Qcl'a culture(lettersK andL) from Maliyin. of

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

117

there was no sukkal of Anshan. This had prompted the writer to suggest that Anshan, the proposed dynastic centre of early Elam, remained a district subject only to the Eparti kings. However, the complete lack of mention of Anshan for nearly 300 years (except in the conventional title of the successive rulers of this dynasty) would appear to attest the relative unimportance to Elam of Anshan as a city or district in that period. Besides the evidence at Maliyiin the archaeological survey completed by Mr. Sumner in the Dasht-i Baydl indicates that most of the other settlements of this period ceased to be occupied during the late third or early second millennia. Hence the obscurity of Anshan during the dynasty of the Eparti kings of Susa may be understood if we associate our lost city with the site of Maliyfin in the bulak of Bayda. The name Anshan as used in a territorial sense reappears in the inscriptions of two rulers of a later Elamite dynasty at Susa. Shutruk-Nahhunte (1185-1155 B.c.) admits to having found, far up in Anzan, a stele of a king whose name he did not know.91 It would be intriguing to consider that this stele may have been shifted from the debris of the old city of Anshan, identified at Maliyiin, which had perhaps even then been abandoned six or seven hundred years before. A later king, Shilhak-Inshushinak (1150-1120 B.c.), records having camped in Anzan while he was in pursuit of rebel forces.92 Since inscriptions of both of these rulers have been recovered at Liyan on the Bushire peninsula, and those of Shutruk-Nahhunte at Tulaspid on the Fahliyiin plain, it may be assumed that by their reigns much of western Fars or Anshan was controlled by the kings at Susa. However, of the excursions into Anshan, one seems to have been partly concerned with the bringing back of an archaeological trophy and the other describes the progress of a punitive operation. These suggest that a somewhat looser political association existed between Susa and certain other areas of Anshan in that period. The obscurity of the old city of Anshan continues. Although if we are to propose Maliyfin as Anshan, the existence of late second millennium Elamite pottery in restricted areas of this site suggests a reoccupation there at that time. It is possible to consider that Shilhak-Inshushinak or more probably Shutruk-Nahhunte founded a settlement at Maliyfin, perhaps as a fortified station on a major route joining central Fars to Khtizistan which would have passed through the Tang-i Khds, discussed below, and located directly to the west of Dasht-i Bayjda. This may also have been the primary function of the foundation of ShutrukNahhunte at Tulaspid, which is located near the junction point of the Tang-i Khas route and the main road leading between southern Fars and Khrizistan. To sum up: it appears that the archaeological evidence at Maliyfin, as considered in necessarily general terms at this stage of enquiry, would manifestly suit the indications of the prosperity and decline of the city and district of Anshan which are to be found in contemporary texts. Does the location of Maliyimn, on the other hand, also agree with where we should expect to find the city of Anshan? We have already noted the hypothesis of Herzfeld that greater Anshan corresponds to the modern Ostan of Fdrs. Further, it has been submitted that our city should most probably be located in the fertile plains, as is located Maliyiin. Equally, a careful look at the available sources may afford an additional clue for this location. The evidence, which will be examined, is found in the " Key to Anshan ". We first learn of a district Khukhnur in Anshan from a text of Amar-Sin (2047-2039 B.c.), a king of the third dynasty of Ur, who claims to have laid waste that region.93 By c. 20 17 B.c. another ruler in Ur, Ibbi-Sin, is reported to have marched into Elam and to have attacked the district of Khukhnur which he calls the bolt or key to the land of Anshan.94 Hinz has proposed that Khukhnur is to be located near Malamir (now Izeh)95 a small plain enclosed by foothills of the Southern Zagros in north-eastern Khaizistan. Beginning at this point an old caravan route, the so-called Bakhtiari trail, leads across the Zagros to Isfah~n (Fig. I). This identification is apparently suggested because two tablets in Elamite which record Khukhnur were allegedly found at Malamir and because the mountains to the north-east of the plain there have been associated by some modern writers with Anshan. On this
91 " Datenlisten ", RA II (1938), Elams. I. Teil: 94A. Ungnad, Quellenzur Geschichte I46 (103). On the Hiising, Die einheimischen different readings of this phrase cf. E. Reiner, " Malamir ", AltelamischeTexte (Leipzig 1916), 52, no. 2o. 92 R.A. LVII (1963), 73-4Ibid., 66, no. 44." und 13 Thureau-Dangin, Die sumerischen akkadischen Kdnigsinschriften 9- Hinz, Persia ", CAH (3rd edn.), fac. i9, 6. (Leipzig x907), 144.
11

118

JOURNAL

OF

PERSIAN

STUDIES

hypothesis Khukhnur is the key (via the Bakhtiari trail) to the land of Anshan (a part of the Southern Zagros). However, the occurrence of Khukhnur on the tablets seems inconclusive by itself, for no indication of the location of Khukhnur is given in either case.96 Equally, although the mountains in question may well have formed a part of greater Anshan,97it seems more appropriate that the " Key to Anshan " would indicate a point of access to the city of that name and to the more valuable agricultural land of that district. Herzfeld clearly had the latter view in mind when he stated that the " key " was situated at Kotal-i Sangar (Fig. I). That defile forms a narrow passage, 12 km. north-west of Fahliyfin, on the traditional, southern route which leads between Khizistdn and Fdrs.98 The argument presented here suggests a modification of the location given by Herzfeld. The last alluvial plain of any size which is encountered by travellers moving eastward into Fars on the direct road from Khfizistdn is that of Behbehdn (Fig. I). Hinz has proposed that Behbehdn should be identified with the district of Hidalu described in Assyrian texts as being located on the border of Parsumash, i.e. Parsa/Anshan.99 In the present writer's view, the mountains of Hidalu would have included part of the mountainous district of Kuhgiluya, which borders on the Behbehdn plain to the east and north. In proceeding east from this frontier district of Hidalu/Kuhgiluya, by the straightest route available in this broken country, one passes through the afore-mentioned Kotal-i Sangar into the shallow Fahliyiin basin. These two valleys comprise part of the first district of Fdrson the Khizistdn boundary. At the southern end of the Fahliyfin basin the established road to Fdrs turns to the south (Fig. I): it then passesthrough Kazarfin, bends to the east and eventually reaches Shiraz. Alternatively there is a second, much shorter route to central Fdrs. At the end of the Fahliyfin valley where the main road to Shiraz runs southwards, a track diverges to the east along the course of the Rfidkhana-i Shul. After Io km. this track turns left into a side valley. Some I6 km. further east the trail enters a narrow defile some 2o km. in length which affords a direct passage to the Mary Dasht basin. This defile, the already-mentioned Tang-i Khds, is situated between two parallel ridges. It leads to a cross-ridge, a spur of the Kfih-i Saran, on the opposite side of which is the Dasht-i BayodIand Maliyin where we have placed Anshan. It is a reasonable hypothesis that the Key or Bolt of Anshan should indicate the important pass (the Tang-i Khds) which permits direct access from Khiizistdn to the most prosperous area of Fdrs/Anshan. While the inscription of Ibbi-Sin mentions "Key to Anshan" a Herzfeld has pointed out that " a region second version of the same text reads " Key to Elam ".xo00 which is the Key of Elam, seen from Anshan, is the Key of Anshan seen from Elam ".10x Given these geographical equivalences the most direct route of access from Khfizistdn to Fars is via the Fahliyfin valley at the Tang-i Khds. Accordingly we regard these areas as representing the Elamite Khukhnur, of which the Tang-i Khds is equated with the " Key to Anshan ". Further support for our identification of the Tang-i Khds as the " Key to Anshan " may be gained from classical accounts of the campaign of Alexander the Great in south-west Iran. According to Arrian, after Alexander had subdued the Uxians during his advance from Susa to Persepolis, he sent the
MDP XXIII, nos. 270 and 273. Hinz suggests that the name " Aiapir " (cf. n. 96, 6, n. 3) is late Elamite for Milamir. However, Milamir in the context used here is an Arabic term meaning property (mal) of the Amir. The association of this word with the site it identifies, is not attested before the fourteenth century A.D. (cf. H. A. R. Gibb, (trans.), The II (Cambridge I962), 287). The older Travels of Ibn Ba.t.tri.ta Iranian name for MNilamirin Khfizistdn was Idhaj. 9 The high country of Anshan is mentioned in a Sumerian epic which relates to the third millennium B.c. (S. Kramer, and Enmerkar theLordof Aratta (Pennsylvania 1952)). According to this text, Enmerkar crosses the seven mountains of Anshan (I7) to demand lapis lazuli and carnelian from the Lord of Aratta (27). The lapis lazuli is stated to have been brought down from the mountains in slabs from which items were fashioned in Aratta (9). It is possible that Anshan at that time included much of the present Ostan of Kermdn as did Achaemenian Parsa which was later to be identified with the same region as Anshan. It may be, therefore, that we must look to the east of Kermqn province, possibly as far as Sistdn, for the
6

site of Aratta. This suggests the very large site of Shahr-i Sokhta where hundreds of fragments of lapis (presumably brought there from Badakhshan) and of carnelian have been found in recent excavations. Many examples recovered were waster beads. The excavator of the site considers that the presence of great numbers of these and of other types of decorative stone at Shahr-i Sokhta indicates that an important local industry was devoted to the working of this material (M. Tosi, "Excavations at Shahr-i Sokhta ", East and West XIX (September-December 1969), 372-74). It was perhaps the artisans of Shahr-i Sokhta who supplied lapis and carnelian for Mesopotamia in the third millennium; if so this site may be identified with ancient Aratta. 98 Herzfeld, The Persian Empire, 176-7. 99 Cf. n. 54. 100 Gadd and Legrain, Royal Inscriptions. Ur Excavation Texts I (London 1928), no. 292. The identification of ka.mu as " Key " is attributed to Poebel; cf. Cameron, Historyof Early Iran, 58, n. 43. '0 Herzfeld, The Persian Empire, 176.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

119

baggage trains and his more heavily-armed troops along the main road which leads into Persia (Parsa). He himself took the faster and lighter-armed troops through the hills towards the Persian Gates where Ariobarzanes, the satrap of Persis, had built a wall.102 Quintus Curtius describes the route of Alexander as following a pass shut in by steep and craggy rocks.103This is an accurate representation of the track which leads east from the Rid-i Shul through the Tang-i Khas. Alexander eventually circles to one side of the defences of Ariobarzanes at the Persian Gates (Quintus Curtius calls this feature the Susian then attacks the Persiansfrom the rear and presseson to Persepolis. MacDonald Kinneir and Gates)104 Stein, who had both travelled through the Tang-i Khds, proposed to identify this narrow defile with the Persian Gates described by Arrian.105It is difficult to see an alternative to this view. Herzfeld suggests that Dionysius Periegetes uses a local idiom when he describes the Caspian Gates of north-centralIran as the " Keys to the land of Asia ".106e Thus what would appear to have been the " Key to the land of Anshan " going east and the " Key to Elam " going west in the Middle Elamite period may be identified as the Persian Gates going east and the Susian Gates going west in Achaemenian times. The analogy is strikingly similar. The Tang-i Khds is proposed as the " Key " to these "Gates " which lead directly to Maliyfin in Dasht-i Bay;d-the suggested city of Anshan in Anshan. We learn a further potentially useful fact for our study from the text of Arrian, which is that the normal carriage road between Susa and Persepolisdid not run via the Persian Gates (identified with the Tang-i Khas), but diverged to the plains before this defile was reached.107The main road can hardly have passed over the rugged mountains which are to the north-east of the Fahliyfin valley, above the Tang-i Khas. It is therefore most probable that the ancient carriage route is to be identified with the present main road which leads from Fahliyfin through the plains to Kazarfin. Dr. R. T. Hallock, translator and editor of the fortification texts from Persepolis, suggests that the mention of Tirazzish would indicate that this settlement was an important point on (Shiraz) fifty-four times in these texts108 the road between Persepolis and Susa at the time of Darius I.109 On the other hand, the occurrence of Anshan only three times in the published fortificationtexts110 led Dr. Hallock to propose that Anshan has was not located on the main east-west road in that period.11 This, of course, agrees with what we have argued above. However, we also know from the fortification accounts that Anshan was situated on a route which led from Susa to Persis. In PF 1780 a horseman named Shada, travelling " from Anshan [and] from Elam ", receives from Ushaya five quarts of wine for eleven horses.'12 Dr. Hallock has informed the writer that the probable destination of Shada was Persepolis.113Since Anshan is east of Elam (the text is in Elamite, thus Elam is used instead of OP Huja for Khrizistdn), Shada must have ridden from Elam via Anshan to Persis.114 At this date, the reign of Darius I, OP Parsa had already superseded Anshan as the name used by the Achaemenians to identify much of the region now called Fdrs. In these late Elamite texts, however, the term Anshan would appear to have survived for the home district of the old city there. If Shada brought his horses via the Persian Gates or Key to Anshan and through the Dasht-i Bayda where we find Maliyin, the proposed site of Anshan, then he would not have travelled the main route to Parsa/ Persis. This is in keeping with Dr. Hallock's contention that Anshan was not situated in the usual Persepolis-Susa route. Indeed, if Shada was driving eleven horses to Persis from Elam, it seems only practical that he should take the far shorter route through the Persian Gates rather than the lengthy and unnecessary diversion via Shiraz.
x102 Arrian, AnabasisI II.18. x-2. 109 This is given in a personal communication to the writer dated 1a03 Quintus Curtius V.3.'7-I9. June 24th 1968. 104 Ibid., V.3.x17. Diodorus (XVII) uses the term "Susian Hallock, op. cit., PF I, o1097, 1780. 11no 111Cf. n. i o9. Rocks ". 105 McD. Kinneir, GeographicalMemoir on the Persian Empire s Hallock, op. cit., PF 1780. (London I813), 72 ff. and 457 if. ; A. Stein, Old Routes in n13From a personal communication to the writer dated January Western Iran (Oxford 1940), 18-27. 24th 1969. 106 Herzfeld, The Persian Empire, 176. The text quoted is that of 114 In a letter to the writer dated February I9th I969, Dr. Hallock suggests that a scribal error may have caused the Dionysius Periegetes, 1035, although Herzfeld calls this writer unclear nature of the passage "from Anshan [and] from Dionysius of Teredon. 107 Cf. n. I1o2. Elam ". It seems possible, however, that the last point visited 108 On Tablets (Chicago tirazzis, cf. Hallock, Persepolis Fortification (here proposed as Anshan) was listed first in a kind of reverse itinerary. 1969), I62.

120

JOURNAL

OF

PERSIAN

STUDIES

The fact that Anshan is mentioned only three times in the published fortification texts suggests that the reign of Darius it was not a particularly important place. This admittedly indirect evidence, by appears nevertheless to be in agreement with our findings in Dasht-i Bayjdt. The pottery at Maliyfin suggests that the greater part of the site there had been largely abandoned for over a thousand years before the Achaemenian period. Moreover, Mr. Sumner has informed the writer that he has found little evidence of Achaemenian occupation in the whole of the plain of Bayd. One further source of the Achaemenian period should be considered for evidence of the lost site of Anshan. In the Behistun inscription the Old Persian text contains this passage which relates to the usurper Vahyazdita who revolted against Darius I in Parsa: " Thereupon the Persian army which [was] in the palace, [having come] from Yadaya previously ... went over to that Vahyazdata ".115 Cameron has shown that at the identical point where Yadaya occurs in the OP account, Anshan appears in the parallel Elamite text.11ne According to Kent's rendering of this somewhat obscure passage, a Persian army came from Anshan to the palace, presumably Pasargadae or Persepolis. The evidence of the fortification texts suggests that Anshan remained a relatively unimportant place at the time of Darius. If this were the case, it is legitimate to ask why an army was stationed there. On the present hypothesis, it may have been necessary to guard the Persian Gates, just as it was in the time of Alexander. Alternatively, an army on the march from a point further west might have taken a short cut through the Persian Gates via Anshan (Bayjd) on its way to Persepolis. In either event, the presence of a large Persian force at Anshan need not, of itself, imply the presence of a large permanent settlement there in Achaemenian times. We turn now to the question of whether either the name Anshan/Anzan or the Old Persian name Yada (of which Yaddyd, quoted above is a case form'117) survived into the post-Achaemenian period in Firs. The district where we have placed Anshan is now called Bay~d (pronounced locally Bayzd). This, of course, is an Arabic word meaning " white " and was most probably introduced there after the Arab conquest of Fdrs. Istakhri relates: " Al-Bay<dl' is the biggest city in the kira of Istakhr and it is only called al-Bay~d' because it has a fort . . . whoever stands away from it, sees its whiteness ".118 The present town of Tell-i Bay~d, today the headquarters of an administrative district, occupies the medieval site and retains its name. Muqaddasi records that Bay~d is the same as Nisd, the latter being the older Persian name.119 The substitution for this word of an Arabic one meaning " white " was due, as Dr. Gershevitch points out in the Annex below, p. 124, to the former having meant "bright, brilliant, splendid ". Situated about 750 m. to the west of Tell-i Baydld there is an extensive outcrop of whitish rock no doubt to be identified with the site of the fort mentioned by Istakhri. Extensive enquiries made on the Dasht-i Bayjd, and a search of the geographical literature which concerns this region, have produced no local names which could relate linguistically to Old Persian Yadd or to any form which Dr. Gershevitch tells me might underlie this Old Persian spelling. There is, however, an Islamic source which in one breath with BaytdI mentions a toponym Ash, one of the possible linguistic derivations of which is from the name represented by the ancient spellings Anshan and Anzan (see Gershevitch in the Annex below). It is therefore appropriate to examine the evidence concerning the location of the place referred to as Ash. Although the name Ash no longer survives in the area of our enquiry, it is attested by Ibn al-Balkhi (c. 494/1 Ioo) as being a village which existed in his day in the neighbourhood of Bayjd.120 Ibn alBalkhi, writing in Persian, gives no further description. However, in the Arabic geography of Muqaddasi (375/985) Ash is also mentioned. According to this text, it is one of the post-stations on the summer track from Shiraz which travellers to Isfahan pass on the way to Sumayran.x'2 A comparison
Kent, Old Persian Grammar,125, lines 21i-8. Cf. n. 73.117 Kent, op. cit., 204 sq. 11 Is~takhri,126, 1. I1I. 119 Muqaddasi, 432, 1. I. 120 Ibn al-Balkhi, Descriptionof the Provinceof Fars (trans. G. Le Strange), Royal Asiatic SocietyMonographXIV (1912), 30. 121 Muqaddasi, 458, 1. 8. Ibn Hauqal, 187, 1. 12, lists a Qal'at al-As immediately after Qal'a Baytd; in a section on the forts of Fars. This suggests that these two places may have been in
115

118

proximity and that As was Ash. If so, as in many instances with Persian shin, the Arabic spelling of Ash is with sin. It has been suggested that sin may represent an incorrect transcription from the Persian. However, the same orthographic change is used too consistently by different writers in rendering Persian place-names into Arabic to support this theory. A few of many other examples are Qfsmis for K6mish, Maisan for Meshdn, Kqsqn for Kishan, Tustar for Shushtar and the double change of Sfis from Shfish. In the case of Qfimis, the Persians have adopted the Arabic form with sin.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

121

of Muqaddasi's itinerary for this route with the exact distances given by Ibn al-Balkhi'22will provide a close location for Ash. The itineraries are tabulated below: Muqaddasi to Shiraz Ibn al-Balkhi Shiraz to Juwayn (5farsakhs or 24 km.)
Baydl (3 farsakhs or
I4

Ash

Nisa (one day) Mihrajandbdd (one day) Kfirad Kalldr (one day)

km.)

Burd (2 barids or 19 km.)

Tar (4 farsakhsor 19 km.) Tir Mayijiyn (5farsakhsor 24 km.) Jarmaq (4farsakhs or i9 km.) Kirad (4farsakhs or 19 km.) Kallar (5farsakhs or 24 km.)

(38 According to Ibn al-Balkhi, after two stages which total 8farsakhs123 km.), the traveller arrives at Bayjd. Muqaddasi identifies this stop as Nisd which, as we have seen, is the older Persian name for Arabic Bayod. The route of Ibn al-Balkhi continues for 4farsakhs (i9 km.) to Tir. This place is still to be identified by the Tang-i Tfir, a narrow defile located exactly 19 km. to the north-west across the Dasht-i Baydd from Bayod/Nisd (Fig. 2). Ibn al-Balkhi then takes us into the buliak Ktm Firiz, a of district name still applied to the valley of the upper Kur river to which the Tang-i Tfir connects (Fig. 2). After Nisa/Bay<dl,Muqaddasi allows 2 bards124 km.) to Ash and Burd. His itinerary, like that (19 of Ibn al-Balkhi, then leads through the valley of the Kur river where the villages of Mihrajdnabdd, Kirad and Kalldr were once located. The distance quoted by Ibn al-Balkhi between Baycd/Nisa and Tir is the same (i9 km.) as that given by Muqaddasi from Nisd/Baydldto Burd. Moreover the Tang-i Tir affords the only direct access between the plain of Baydj and the valley of the upper Kur in Kam Firiz. It seems reasonable to conclude, therefore, that Tir and Burd are the same stage on our route.125 The first name had been possibly associated with the defile and the second with a now vanished poststation which perhaps was once located near there. The point should be made here that the itinerary of Muqaddasi is given from north to south while that of Ibn al-Balkhi runs from south to north. The latter writer, who actually lived in Fars, places the stage of Kallar 5 farsakhsnorth of Kirad. Muqaddasi, on the other hand, working in the opposite direction, gives as a single stage " Kirad and Kallar, one day ". What appears to have happened is that in an ultimate source the order of names was that of Ibn al-Balkhi, but that Muqaddasi has reversed the direction of stages. Where two villages are cited in the same stage, he has neglected to reverse them. As a result the names occurring in pairs are in the wrong order, and thus appear to conflict with the reliable list of the locally-knowledgeable Ibn al-Balkhi. This error has therefore to be corrected; it has been done on the preceding list. The same correction was already made (without explanation) in the account of Schwartz.126This correction would also operate in the case of Muqaddasi's arrangement of the villages of Ash and Burd. Therefore, as shown in the table, and accepted again by Schwartz, Ash should be the halt north of Bayda/Nisd, and the next halt in a northerly direction should be Muqaddasi's Burd (the Tir of Ibn al-Balkhi). It is sufficient that even today one of the most direct tracks which leads north-west across the plain of Baydldto the Tang-i Tfir and into the Kur valley beyond, passes along the north-west embankment wall at the site of Maliyin (Fig. 2). This is where we have sought to locate the site of Anshan and since the medieval Ash is now fixed close to the same point, it provides a valuable confirmation of the earlier argument.
122 123

Ibn al-Balkhi, op. cit., 883. On this unit of measure cf. A. Houtum-Schindler, " On the Length of the Persian Farsakh ". Proc. RGS X (September I888), 586. 124 The sikkat al-barid is equal to 2 farsakhs. On this, see art. " Barid ", El (2nd edn.).

125 It

126

is not improbable that Burd represents a corruption of a place name originally written Tfar and that this is identical with the name spelt by Ibn al-Balkhi with the letter (Tair) which could be a deliberate artifice to avoid confusion .td between td and ba. P. Schwarz, Iran im MittelalterIII (Leipzig 19x2), 1i8o.

122

JOURNAL

OF PERSIAN

STUDIES

One may ask why Ash, together with dozens of other villages mentioned by Ibn al-Balkhi, has disappeared. Numerous reasons, of course, can be cited, but in particular it is probable that many of these settlements were destroyed during the years of political instability brought about by the Mongol conquest of Fars. The earliest mention of Maliyfin is found in the Fdrsndma-i N4iri (1886). It is possible that this village may be a late foundation from which, as the local inhabitants claim, the archaeological site at Maliyin receives its name. If this is correct, the name Ash, once possibly associated with these remains and with the now vanished early Islamic settlement there, would have gradually fallen from use after the founding of Maliyfin. As a final point of enquiry in this examination, it may be of interest to compare the range of pottery found at Maliyfin with similar material recovered from the Elamite site of Liyan on the Bushire peninsula. Pezard, the excavator of Liyan, has divided this pottery into two early series and one later group. The late ware is associated with the Sasanian period. Style I of the early range is characterized by cream or buff-coloured plain wares.127Vessels in this group tend to be somewhat bulbous in shape. There are tall, cylindrical jars with well-developed button bases of a type also found in Khfizistan. The latter examples date from the second half of the second millennium B.C.x128Style II of the early series is characterized by three ceramic groups. These include unpainted pottery with a red slip; ware with a red slip decorated with geometric designs and parallel lines in brown paint; and pottery showing a salmon-pink to cream surface also painted with geometric designs.129The pottery of Style II at Liyan is identical with the range of ware assigned to the Kaftari culture on the Mary Dasht basin. Pezard identified Style I as the earlier sequence and Style II as the later one.130According to our parallels, however, Style II is the earlier group. It is a Kaftari ware and should be dated to the third millennium B.C. Style I is the later ware which, on comparison to the parallels from Khizistdn, should be dated to the second half of the second millennium B.C. Of course in 1914, when Pezard completed his work, most of the pottery which is now available from other sites for comparison with that of Liyan, had not yet been excavated. The incorrect sequence given for Styles I and II, therefore, was perhaps unavoidable. The excavations at Liyan also produced a small piece of broken baked brick which contained a fragmentary inscription in Elamite written in an archaic Sumero-Akkadian cuneiform script. This inscription, which contains a part of the name of an otherwise unattested Elamite king, has been dated on epigraphic grounds to the twenty-fourth or twenty-third centuries B.C.x131 A number of more complete baked bricks showing inscriptions were also recovered at Liyan. These contain dedications material Pezard was unable to relate either the early inscription or the group of later ones to any of the pottery from Liyan. However, present knowledge of this pottery would suggest that the Kaftari ware (Style II) is comparable in date to the early inscription and that the plain ware (Style I) relates chronologically to the later texts. E. Reiner has proposed that the older epigraphic example from Liyan may date only from the later inscriptions from Mesopotamia or south-west Iran of the twenty-fourth and twenty-third centuries B.C., are either in the Sumerian or Akkadian languages or written in the Proto-Elamite script.132Dr. Reiner, however, does agree that the "early " fragment in Elamite from Liyan was written in archaic cuneiform, but argues that this was done deliberately at a later date.13s
If we are to relate the oldest Elamite inscription to the Kaftari pottery at Liyan, then this would suggest an equal cultural affinity with the same range of ware found mostly in evidence at the much larger site of Maliyfin. Because of the direct cultural association of these sites during the Kaftari period,
127

to by four later Elamite kings who ruled from c. 1300oo I I20 B.C. Because of a lack of comparative

Elamite king Humban-numena

(1300oo-1275 B.C.).

This is argued because the only other known

128 129

M. Pezard, " Mission A Bender-Bouchier ", MDP XV (1914). See pl. VI 6, 7, 8 for examples of the pottery from Liyan designated Style I. Ghirshman, " Susa au tournant du IIIe au Ile mill6naire avant notre 6re ", Arts AsiatiquesXVII (1968), fig. 6. See reference n. I29, pl. IV for examples of Style II at Liyan,

E. Reiner, " The Earliest Elamite Inscription ", JNES XXIV (1965), 337. 132Ibid., 340. 133Loc. cit.

o30 13 Ibid.,

13.

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

123

it is equally possible to maintain that the king mentioned in the Liyan inscription was in fact one of the rulers of Awan the proposed earlier name for the district of Anshan. The suggestion that no other texts in Elamite of the twenty-third century B.c. have been found may not be entirely accurate. It has been accepted by most scholarsthat the treaty of alliance concluded between Khita of Awan and Naram-Sin of Agade, mentioned in the first section of our examination,134 should be dated to the reign of the latter
(2291-2255

already been removed to Awan by the middle of the third millennium B.C., Awan should have been a place of political importance by that date. It would not thereforebe unduly surprisingif one encountered further inscriptions of the first dynasty of Awan. In view of the evidence already presented, the writer feels entitled to suggest that if such texts do exist, the place where they would probably be found is at Maliyin, since that site has been identified here as Anshan. One final point should be made with regard to the excavations at Liyan. It will be remembered that while the pottery of the earlier occupation there (Style II) is identical with that of the Kaftari sequence of the Mary Dasht basin, the pottery of the second phase (Style I) corresponds to the later plain wares of Elamite Khizistdn. Thus the cultural, and no doubt political orientation of Liyan seems to have changed from the direction of Firs and Anshan, where it is suggested that Elamite kings resided in the late third millennium B.c., to that of Khiizistan and Susa, which was the residence during most of the second millennium B.c. CONCLUSION A number of conclusions have been drawn from the present enquiry. Cuneiform texts show that the city of Anshan played a prominent role in the history of the early Elamite civilization in southwestern Iran. Examination of these sources supports an identification of Anshan with the capital of the Elamite dynasty which ruled from the district of Awan. It is further contended that the important city of Anshan would have been located on temperate, crop-producing plains rather than in the barren mountain regions suggested by earlier writers. Inscriptions and rock sculptures of different kings of Anzan and Susa have been found in various parts of Fars. Equally, early Achaemenian kings are associated with both Anshan and Parsu (Parsa) in Neo-Babylonian texts. These data imply that much of this upland country, together with the greater part of the present Ostdn of Firs, may once have constituted the province of Anshan. Also, Maliyfin on the fertile plain of Baydl is proposed as the site of the city of Anshan. Visible pottery at Maliyin, which is mostly of the Kaftari culture, shows that a very large settlement flourished there from at least the second half of the third millennium B.c., the period when our texts imply that Anshan (in Awan) developed into a major political centre. It is also a period when Anshan seems to have been a prime goal of successive Mesopotamian rulers who periodically invaded and occupied the Elamite territories. Invaders from another area perhaps brought Qal'a ware to Maliyin. Indeed, the scarcity of Qal'a ware on the site suggests the possibility that this pottery may have been contemporary with the late Kaftari ware and that Maliyfin was abandoned before the Qal'a became firmly established there. Only excavation can answer these points. We do know, however, since Qal'a ware is the terminal pottery of the earlier periods at Maliyin, that the most ancient phase of occupation there would have
ended at some time before the appearance on other local sites of the succeeding Shughi ware (in about 19OO B.C. or slightly later). This is the same relative period (c. I890 B.c.) when Eparti was created with the establishment of a new Elamite dynasty at Susa and when much of the area of the " Ville Royale " at Susa was first built upon. The last mention of Anshan in Mesopotamian texts for over i300 years occurs in an inscription of Gungunum of Larsa in about 1928 B.c. It is possible that these events reflect the final decay of Anshan as indicated at Maliyain. On the other hand, the discovery of late second millennium pottery in area A at the site attests to a limited reoccupation of a part of Maliyin at that time.
134 Cf. p. 101 and n. 8.

B.c.).

This is well within the period given above. Again, if the " Kingship of Ur " had

124

JOURNAL

OF

PERSIAN

STUDIES

The writer has attempted to show that the Key or Bolt of Anshan, the proposed Persian Gates, are most probably to be identified with the present Tang-i Khas, a trough-like defile which leads directly from the Khizistdn road at Fahliyfin to the plain of Bay<d where Maliyuin is located. By the reign of Darius the Great the name yadd appears to have been used by the Achaemenian Persians to identify the old home district of Anshan, just as the greater region of Anshan through the same process became Parsa, the modern Fars. It is possible, on the other hand, that the ancient site of the city of Anshan may have retained its Elamite name in a local pronunciation which developed in Middle Persian to Ash. A comparison of Early Islamic itineraries has shown that the settlement of Ash is to be located close to or at the present site of Maliyin. It is evident that the site at MaliyiIn was situated on the most direct route which joined Khizistan to central Fars in antiquity. Moreover, the same road may have formed a part of the main overland trade artery of the third millennium which extended from Mesopotamia through Fars to the countries of further Asia. It is therefore possible that evidence of early cultural contacts between these widely separated regions may be recovered at Maliyin.

ANNEX: NOTES ON THE TOPONYMS ASH and NISA By Ilya Gershevitch

It is characteristic of the Elamite language that place-names which in other languages end in -af often show instead an ending written -an. The most famous example is the name of the capital city o, in Greek, in O(ld) P(ersian), but Sulan in Elamite. Accordingly, Elam itself, which is written ' U0oac into account that the 1/z alternation points to the sound in question having been (thus Professor taking .Sazd R. T. Hallock in a letter), one may suppose that the OP pronunciation corresponding to what is spelt An3an or Anzan in texts other than OP, was *nzva. If this form also underlies the OP spelling Yadd (the case for so thinking will be presented elsewhere), its first vowel will have been short in Anganian pronunciation, the second vowel long. It would be only natural for Anganians in their home-territory to have continued to call it *andzaeven after the Persians had conquered it and, within Dr. Hansman's definition of Ansan, made it part of their own homeland; it would then be equally natural for the Anvanian pronunciation of Anvan's name, and not the OP pronunciation of it, to have survived into post-Achaemenian times on Anvanian territory, even though that territory would by then have been Persian, and its inhabitants speaking Persian or a Persian dialect, for centuries. Given these suppositions, Dr. Hansman's suggestion that the place-name A1 in the Baydt district might represent a survival of the name of the city of Ansan would be defensible on the following philological grounds. Once the final d of *anzd had been lost in Middle Persian according to expectation, and the initial short vowel had been lengthened before consonant cluster,1 the resulting *dn would be likely to have come to be pronounced *q, with nasalized long d. As nasalization of d was often left unexpressed in writing,2 one might expect the New Persian outcome of *anzd to have been written * Baydp T plain, viz. at Oir7z and Sivand, sporadic und. But in two dialects spoken not far from the of 4 has been noticed.3 If such unvoicing was once a feature also of the dialect spoken in the voicing here proposed would be regular. Bay~dl plain, the derivation of AT Of course LT3is a name which lends itself to a good many alternative derivations: a written New Persian d normally represents an original d and does not usually conceal a nasal; J is usually from either Old Iranian I, x3, or Oy; an r could have been lost before it, or an initial h could have been lost
1 Cf. the New Persian outcomes *apatama-, dzmdy- "to test, try" " Sogdian zm'y-), aivdn portico ", of appadan(Henning, TPS (0944), " " of OP dfdum the last of *uzmay- (cf. Christian from Middle Persian divan, 1o n. beside aspris x), dsris of *as(p)a-wraisa- (Herzfeld, Inschriften,170), etc. 2 See G. Lazard, La langue . . . persane, 156 sq. 3 See Horn, GIP, I, 9go. " race-ground " Altpersische

ELAMITES,

ACHAEMENIANS

AND

ANSHAN

125

before d. But precisely because the possibilities are so numerous, none stands a chance of being convincing unless historical geography supports it. It is only if the place whose name was written LT stood where centuries earlier the city of Anvan had stood, that there is a point in noting that the name of that city may have sounded *anzd, and that in the dialect of the region where the sources locate "As ", *anzd might have become *ip. As regards the toponym Nisd, the fact that the Arab invaders replaced it with their own word for "white ", Baydd,suggests that it, too, was a term referring to the brilliance of the nearby landmark, an outcrop of white rock. There can thereforebe little doubt that the name is identical with the Parthian adjective nys'g" bright, brilliant, splendid ". The loss of Middle Iranian final g after d is the rule in New Persian. But the formation of the Parthian adjective, unlike its meaning, has so far remained obscure. It ceases to be so as soon as one recognizes that it has a cognate in the Pahlavi word which translates Avestan a-saya-" shadeless, without shadow ". The Pahlavi word is written ns'yk,which spelling Bartholomae intelligently, but wrongly, emended to *n's'yk. Since in the Pahlavi compound sdyak" shade " corresponds to saya- in the Avestan compound, the privative a- of the latter proves the value of ni- in the former to have been privative. The use of privative ni- is found also in Sogdian.4 In Parthian it is attested in nyz'wr" weak ", lit. " strengthless ", and in nyl'm and nb'mpresently to be mentioned. If the ni- of Parth. nys'gis also privative, s'g must be the Parthian form of the word for " shadow ", agreeing with Northern Balkardi sdg and Yidya sdyo. This noun may in fact occur in the Parthian of verse [bs](t) bwynd s'g'n cygryw'nmwrdg,5 which the translation " the shadows of the dead souls will be " seems more convincing than " the parts of the dead souls " as the edition has it. [fetter](ed) The meaning " bright " of nys'g, which perhaps already Avestan asaya-was intended to convey, presumably arose from an identification of" shade " with " obscurity ". Typologically the compound would then be antonymous to Parth. nyJ'm"dark ", lit. " lightless ",6 and to the hitherto unexplained which can now be seen to contain simply b'm, " lustre ", and to mean " lustreless ". hapax nb'm,7

4 See Gershevitch, GMS, ?I148.

* Mary Boyce, The ManichaeanHymnCyclesin Parthian, I62.35. 6 See Benveniste, JA (1936), 231. Andreas-Henning, Mir. Man. III, 865.1 I; on tnd " faint ", which in this passage precedes nb'm, see Mary Boyce, op. cit.,

Glossary. The base of tnd is the same as occurs in Khotanese " pdtaund& dimmed " (see Bailey, Prolexis, i90, and Emmerick, SGS, 83) and in Sogd. prt'mch " in a swoon ", pt"m " dark(ness)" or similar (on which see GMS, pp. 248 sq.).

EXCAVATIONS

AT HAFTAVAN TEPE 1969: Second Preliminary Report By Charles Burney

The background to the excavations at Haftavan Tepe, situated near the town of Shahpur in the province of Western Azerbaijan, and the relation of this site to others, excavated or known from field surveys, in the Urmia region and beyond was briefly summarized at the beginning of the first preliminary report, and therefore need not be repeated here.x The second season was conducted during the summer of 1969, excavations proper beginning on July 13th and the camp being closed on September I8th. As before, the excavations were sponsored by the University of Manchester. They were generously supported by the Calouste Gulbenkian Foundation, the Russell Trust, the University of Manchester, the British Academy, the Ashmolean Museum, the British Museum and the Munro Trust. Many thanks are due to the Archaeological Service of Iran, especially to Mr. Pourmand, the Director-General, and to Mr. Khorramabadi, the Deputy Director. Mr. Asker Mirfatah was appointed as official representative to serve with the expedition, and performed his duties with zeal and good humour. Much help was also received from the education department in Shdhpur, through whose good offices we were once again allowed the use of the village school at Haftavan, with its ample accommodation and delightful compound. There was a staff of fourteen, of whom six had taken part in the 1968 season, comprising: Charles Burney (director) and Brigit Burney (photographer and also in charge of the health of the expedition and of its labour force); Malcolm Stephenson (architect-surveyor); George Learmonth (conservationist); Miss Gail Durbin (housekeeper and archaeological assistant); Mrs. Helga Schippmann (archaeological assistant); Messrs. Peter Burney, John Curtis, Robert Garland, Nigel Nicholson, Duncan Noble, Klaus Schippmann, Nicholas Sims-Williams and Stuart Swiny (site supervisors). Mr. Noble also had the responsibility of arranging for the cleaning and recording of pottery; and he carried out a preliminary classification. A special debt of gratitude is owed to Mr. David Stronach for his unfailing help in Tehran. We were delighted to have a visit by him and his family in August 1969. His services to archaeological expeditions in Iran can never be over-estimated. The site of Haftavdn Tepe is a mound of formidable size and height, as the site plan (Fig. i) demonstrates. This plan shows the mound as it must have appeared as recently as fifty years ago, before the commercially motivated excavations were carried out across the top of the citadel and at the foot of its south side. The site has been terribly scarred since then, both by those trenches and by very recent diggings by villagers evidently quarrying for soil to fertilize their fields. If this mound, one of the largest near Lake Reza'iyeh, had been left without investigation by properly conducted excavations, by 198o little would have survived beyond the remnants of a honeycombed landscape. Even by 1968 very extensive damage had been done, especially on the south and west sides. Only the steep north slope and the wide lower shelf on the east, extending towards the spring, remained almost unscathed. This large mound-at least 550 m. from east to west and up to 400 m. from north to south, with a maximum height of about 25 m. above virgin soil-would present by its very area of some fifty acres a considerable challenge. The extent of recent destruction aggravated the task of choosing where best
to lay out the areas for excavation. In addition, the powdery consistency of much of the mound, particularly the citadel, has resulted in extensive erosion and washing down of deep layers of silt, especially evident on the west side. So soft are many layers that the sides of the trenches became
1 C. A. Burney, Iran VIII (1970), pp. 157-64.

127

0 0 0
13
17
19

14

1 li
15

1
21

22

213

242

I---'-

-'" --

--------'--

",oooo
oooo"0
, , ,

,0o?

----'t,
t t .

.--

:-

i1/'I.

J 1 6 7 9 1 1 11 1 3 1 20 21 18 1 1 15 1 3

22

-14*

9,1,011

12131415161

1819202122

2322

about yearsago, before recent at inte appeared of damage: contours half-metre fifty Fig. I. The mound HaftavanTepeas it probably

EXCAVATIONS

AT HAFTAVAN

TEPE,

1969

129

undercut by weathering within the space of a few weeks, although the earlier levels are on the whole more compact. The scale, destruction and erosion of the mound have thus largely dictated the strategy of the excavations, once the primary objectives had been decided upon. In the 1969 season work was carried out in four parts of the mound. The addition of only three new letters (Q, R, S) to the list of areas opened in the first season itself indicates that extension of and continuing in the earlier areas was the keynote of the general plan of the second season (Fig. 2). Lack of resources to some degree prevented the achievement of all that it had been hoped to accomplish; for example, the sounding at the east end of the site (JX) could not be carried right down to virgin soil, one of the tasks awaiting the third season's work, due to be carried out in the summer of 1971. The discovery in the first season of the large stone building on the west sides led to the decision to open up an extension (P2) to Area P, with a further northward extension (Q), in the hope of finding similarly impressive structures of the same level. But here the sheer depth of deposit was daunting and the results to date are disappointing, though further work may bring the desired outcome. Some compensation was gained by the uncovering of interesting burials in PI and P2 (see below). The main effort was concentrated on the citadel, where several new areas were opened up as extensions of the 1968 excavations, eventually being amalgamated (Fig. 2): whatever the theoretical advantages of maintaining baulks as stratigraphic checks from one season to the next, the rate of erosion and the number of soft-filled pits made this here both pointless and impracticable. At the foot of the north-east side of the citadel, where a recent cut had been made into the slope by villagers, a new trench (RI) was opened, soon being extended (R2) to give a total area of 10 X 20 m. Within the Area J (20o X 20 m.) at the east end of the site a smaller sondage JX (I o Iom.) was made. An extension west-north-west from J was made in an area of I o I o m. (SI) during the later part of the season. In RI and R2 was found, not far below the existing surface, a well-preserved building level with walls radiating at the lower end of the area as if round the end of the mound as it then existed, c. in mud brick, and thus 2000-1800oo B.C. (Pl. Ia). These walls are massive in thickness, well constructed contrasting with the method of setting stones in mud which distinguishes the main level in E, which survives to this day in the surrounding villages and which presents so many problems to the excavator. The period of the settlement so tantalizingly revealed in RI-R2 must surely have been one of wellestablished urban life, its position on the I17m. contour suggesting a long history of settled occupation, extensive yet soon to be surpassed in the great expansion of the town during the second millennium B.C., in which the whole wide lower area or skirt of the mound, especially on the east, most probably originated. Four phases were distinguished in RI-R2, the latest being represented by stone walls at the lower (north-east) end, clearly stratified as being cut down into the earlier levels, or terraced against them, and thus later. Of the three earlier phases the second was the most important, only relatively fragmentary remains of the succeeding phase of secondary occupation being found. Among characteristic features of the pottery are Nakhichevan lugs, both hemispherical and pierced and of the small vestigial variety; finger-depressions, a form of ornament known widely in the Early Trans-Caucasian cultural zone in the third millennium B.C. ;3 and black ware with graphite burnish, probably intended to imitate silver and typical of certain Early Trans-Caucasian III sites (from c. 21oo including sites and cemeteries in Georgia and also Yanik Tepe, near Tabrlz.4 Occurrences of B.C.),
polychrome painted pottery, including light and dark brown on buff; may well suggest a date after c. 2000 B.C. for this level. The dating of this level in RI-R2 seems clearly indicated within limits of c. 2000-I8OO B.C., the with higher date being the more likely, by the pottery and by a polished stone axe or hammer-axe, There can be no doubt of the and elsewhere in Trans-Caucasia parallels at Shengavit (Pl. Ic-d).6

2 Ibid., pl. IIb-d and fig. 6.

3 E.g. in the group of pottery from Ernis, on Lake Van, Burney, AS VIII (1958), pp. I82-5; and at Yanik Tepe. 4 For this and other features of Early Trans-Caucasian III

pottery, Charles Burney and David Lang, Peoples of the Hills: AncientArarat and Caucasus(London x971), pp. 66-7. 5 S. A. Sardarian, Primitive Society in Armenia (Erevan 1967), p. 345, fig. XLVIII and pl. 47.

21S1

I V

I / - ---

I ---' - ----

-,-----------

.- -+

.--_

.- . -

.~2

--2,

,/2

Fig.

2.

Siteplan, showingtrenches:contours half-metre at intervals.

SL.APPS FACE [VLaL 2] M M1Y.P 'W

4Wi.Ot% SItZ11a Mt-D B4oZf

Saow t TeXc e
A5RA 1WE Ui1tN
-s
.TV.

5
MN.ED'$EU.-OWi6U B1R4IL

6k .
6G. --K' E5
asa

P3 MIT%
wi*
Some.

V.LELLOWISH

PES

s]au

$ay
IT 'RIsbWFILLOfF Vs/4.
OEZlUCIO BR-ICK
DESP4

, ..

oaMcmt-4P

flM

Ezs.

%1T 7
WNT
-

E~f

A"&(745
VT

SURFA5.

oT4
&*"o
sA.St.

MIwV.AsH .1MM"
om
Cimgsf
LC

xf
Aswt
ww

D645.

SUKFA46

FIu

RLfL'OIN

4FRLEO fPIT.x
%Tx7

OFPM.

.bw
we

Fit PI OF T
WSP#

Fr

.Wu **"-L

TEP5 MIWAFTAVAN NOR

s

Aug

U17

VEv-.L.WARp

x5A EASFig.

side of X. of 3. Section north

E1

0ME.

U%

2
4.

3A

AFTAVAN
3.

Fig-

Section

of

north

side

of JX.

132

JOURNAL

OF

PERSIAN

STUDIES

general contemporaneity of the pottery with that from the second Early Bronze Age period at Yanik Tepe and of its consequent attribution to the period which the writer now terms Early Trans-Caucasian III, the Early Trans-Caucasian I period not being represented at Yanik Tepe.6 The sondage in JX was largely intended to yield a sequence of pottery, to help to fill a void in knowledge of the Urmia region hitherto illuminated only by the material from Dinkha Tepe.7 The first task was to complete the exposure of JX3, a level partially revealed in the I968 season. Then the fill of the very large pit (JX3A) had to be removed before the excavations of the surrounding levels could begin. This pit had been given a terminusante quemby the presence overlying its fill of J I, Grave 7, the one burial so far found at Haftavan Tepe which can definitely be attributed to the Iron I period (c. 1350-1000 B.C.), best known from IHasanlfi V.8 The writer is inclined to place this burial on present evidence late in this period, towards c. 1000 B.c., although of course this dating may have to be revised. The section (Fig. 3) of the north side of JX, begun from the surface of Level 2, reveals a succession of building levels, some with little occupation deposit and mostly bricky fill characteristic of destruction debris (JX7), others with occupation layers overlaid by bricky destruction debris (JX6). Courtyard layers are suggested by part of JX5, with alternating clay surfaces and ashy bands. The bricky debris forming most of the fill ofJX4 and JX3 suggests buildings of some size in the vicinity. Pits provide their 5-74

+5.25

HEART
. 00

.,.HEARTH% .025

LEVEL 3APIT

LEVEL 4 FLOOR

Fig. 4. Plan of JX, Level4.

6 Peoplesof theHills, chapter III. For a discussion of terminology,

HAFTAVAN TEPE 1969 I AREA 3A& kx-LEVELS

p. 44. SR. H. Dyson, (a) " The Archaeological Evidence of the Second Millennium B.C. on the Persian Plateau ", CAH I-II,

fascicle 66 (1968), pp. 20-3; (b) "Dinkha (1967), pp. 136-7. 8 Iran VIII, pl. IIIc and fig. 8 (1).

Tepe ", Iran V

EXCAVATIONS

AT

HAFTAVXN

TEPE,

1969

133

usual disturbances of stratigraphy. The general impression is one of relatively peaceful development: there is no indication of any violent burning. As the plan (Fig. 4) shows, structural remains in JX4, hearths and scattered remnants of stone foundations, suggest a courtyard: there is no proof of any well-constructed buildings in this level, element though this is probably fortuitous. In a sondage of this limited area there is .inevitably.an of.-1 luck in the uncovering of buildings with a significant plan or merely of open areas. A short section of wall in the south-east corner ofJX is sufficient to show that JX5 must be reckoned a significant building level. Better fortune met our efforts in Levels JX6 and JX7 (Figs. 5-6), in both of which more of a plan of mud brick houses was distinguishable. In JX6 the buildings had walls up to 8o cm. thick,

6 FLOORS LEVEL

BRICK/PISE WALLS MUD

LEVEL 6 FLOOR

1969 HAFTAVAN TEPE

Jx AREA
3 2 1 0o5 0

LEVEL 6

Fig. 5. Plan of JX, Level6.

though mostly slightly thinner; stones were used for foundations. Most impressive of all the plans uncovered was that of JX7, the first level unaffected by the deep Iron I pit (JX3A). Damp patches in the soil suggested that here the sondage had reached a point not many metres above the water table. This is a period demanding further investigation. The development of painted pottery can be traced back to JX7, the best preserved example being the tumbler (P1. IId) from JX6, which is of fairly fine light brown ware with red wash inside and out and decorated in black paint. This shape of pot, presumably for drinking and thus termed a tumbler, is most readily distinguished by its sharp carinationjust above the flat or slightly convex base; and it is a characteristic type of this period.9 At least up to JX4 there was a continuous evolution of painted wares; but their variety must not lead to an over-emphasis of their importance in relation to the
T. Burton-Brown, Excavations in Azerbaijan g948 (London
1951),

fig. 22 (no. 1o52), for a bowl with carination just above

the base. But there is no example of the tumblers found in JX at Haftavin Tepe.

12A

134

JOURNAL

OF

PERSIAN

STUDIES

preponderant plain wares (P1.IIc). Naturally their significance lies largely in the evidence which they provide of external relations. Bowls with a lozenge-patterned painted zone along the shoulder are paralleled at the nearest contemporary excavated site, some seventy miles south of Haftavan Tepe, Geoy Tepe near Rezd'iyeh, especially in Period D.10 The origins of this painted pottery appear to lie in and around the Urmia basin rather than further afield, with a movement north into the Araxes valley. No suggestion of a direct link with the Ali?ar III (Cappadocian) painted pottery of central Anatolia seems tenable, although the tin trade through Assur doubtless may have brought some contact between the two regions."

-3-46

LEVEL

FLOORS

HAFTAVAN1969 TEPE Jx AREA 7 LEVEL

3 2 1 &5 0

Fig. 6. Plan of JX, Level7.

Two burials uncovered in PI are of intrinsic interest and of additional significance in that they supplement the data gathered from a number of graves excavated in the 1968 season, especially N I, Grave 5,12with its long bronze chain and bronze diadem or circlet. The section (Fig. 7) of the west side of PI shows the position of PI, Grave 3, which was cut down from the surface of Level I in that area, and which moreover is stratigraphicallyrelated to N i, Grave 5. The appearance of pottery found with two of these burials (Ni, Grave 5 and PI, Grave 5) and the probable parallel for the bronze chains in those attached to the bronze lion pins from IHasanlh3 both seem to indicate a date contemporary with the Hasanlil IV period (c. Iooo-800ooB.C.) for these burial groups, which can thus be classed as Iron II. The evidence from J suggests a ninth- rather than a tenth-century B.c. dating; but in a conservative area an early eighth century B.C.date can scarcely be ruled out.
Ibid., fig. 19, fig. 20o (no. 56), fig. 21 (no. 58), etc. 7. 11 R. H. Dyson, CAH I-II, fascicle 66, p. z 12 Iran VIII, pl. IVa-b; fig. 7.
10 13

R. H. Dyson, " Protohistoric Iran as seen from Hasanlfi ", JNES XXIV (1965), pl. XXXVII.

EXCAVATIONS

AT

HAFTAVAN

TEPE,

1969

135

The richest burials found in the 1969 season were PI, Graves 3 and 5, both of young girls aged about seven or eight. In PI, Grave 3 (Fig. 8 and Pls. IIIa-IVb) the body had been laid in the usual contracted position, on the left side, the hands resting on the knees. A stone slab lay close to the back of the head. This girl was lavishly decked with jewellery: the plain bronze band on her head was the centrepiece of a splendid headdress, comprising large and small bronze discs, some with a loop on the inner side for attachment by sewing, the thread and woven cloth having survived in very small fragments (P1.IVa). Seven strings of beads of brown, green and yellow glass and frit as well as carnelian
LINEOF C..ARlEAoSroZ $WSuFAE OF MouhD

4PoW-R1.-0 D--

s-

l/

.P1/iSP/ e

_
..47-. BIZJITL

eosl . ,w L1/i.

F" , WMn, pt'

pt m CAT mMEDOZ?FT

IOM=ZD&

BR.ICKNf.

Pl1/5

Pl/56

PP 51/

"AFTAVAN S1 ,

TEPE 1969
SAREA. P1
Fig. 7. Section of west side of PI.

WEST SECTION

beads and mollusc shells were probably attached to some form of woven skull cap, to which also the bronze head band would have been attached. She wore the S-shaped bronze earrings found with many burials of this general period in different parts of the site. Two simple bronze pins presumably held her garments together. On her left wrist she wore one bronze bracelet, with three such on her right wrist. Her hands were clasped together, with a bronze ring on each middle finger. On each ankle was a heavy bronze anklet, surely too massive to have been worn in her lifetime, and therefore probably pressed on after death. Beneath her head, and so certainly associated with this burial, was unearthed a cylinder-seal (P1.IVb) of frit; its possibly Mitannian style must make it several centuries older than the level in which it was found.4
14

Briggs Buchanan, Catalogueof Ancient Near Eastern Seals in the Ashmolean MuseumI: Cylinder Seals (Oxford 1966), pp. z179-85 for seals of the Mitannian style. There are none which com-

pare very closely with the seal from Haftavdn Tepe, though some comparison may be permitted with nos. 9g9igand 941-2 (pls. 57-8).

136

JOURNAL

OF

PERSIAN

STUDIES

PI, Grave 5 was largely similar to Grave 3, though there were significant differences (Fig. 9 and Pls. IVc-Va). One was the presence of pottery, two handled bowls at the feet and ajar beside the face. There were also the remains of a small sheep, a food offering placed beside the two bowls. This provides further evidence for the dating of these burials to the ninth century B.c., for a similar offering was found with JI, Grave 3 in the 1968 season, that burial being accompanied by a bridge-spoutedjug of drab red ware, unmistakably of Hasanhli IV type.15 There were similarities to the jewellery of PI, Grave 3, with the headgear of the same pattern, a bronze head band and a headdress formed of small bronze discs, here found arranged in three triangles beneath the skull, a pattern originally in all probability repeated to cover the whole of the now vanished skull cap. There were also large bronze discs and a variety of beads. Six bronze tassels, of the type found with many burials in different areas of the site, also formed part of the headdress: parallels for these are not easy to find.1e The girl's garments were
PI Burial 3 ")',and head-band 7strings of beac small discs

2 earrings discs of .head-dress 2 pins bracelets

2 finger rings

anklets

scale $Fig. 8. PI, Grave 3:

cm 5 10 15

the burial in situ.

fastened with three bronze pins. She wore at least four necklaces of carnelian and blue frit beads and two of white beads, one also including some mollusc shells. On each wrist was a coiled bracelet of iron; she wore also a pair of heavy bronze anklets. She had only one finger-ring and no earrings. The bracelets illustrate the use of iron in this period, when it was still largely confined to decorative purposes. Other burials in PI-P2 were accompanied by largely comparable beads and metal jewellery, though not all in the same level. With PI, Grave 2 were some carnelian and iron beads and a bronze stud. In P2, Grave 3 the skeleton wore two iron and two bronze bracelets and a necklace of iron, glass, carnelian and frit beads. The necklace in P2, Grave 4 comprised beads of bronze made from a flat sheet and others of iron and glass; and there was a bronze stud like the one from PI, Grave 2. P2, Graves 5 and 6 each contained a pair of bronze S-shaped earrings, while the latter included a hitherto unique string of carnelian beads with silver spacers and a pair of bronze dress pins with rolled head (P1. Vb).
1s Iran VIII, fig. 8 (2). 1 C. F. A. Schaeffer, Stratigraphie et de Comparie Chronologie l'Asie Occidentale (London 1948), fig. 217 (19, 20o) and fig. 2x9 (18) for possible parallels.

Pl. Ia. Trench north-east. RI-R2, generalview looking

RI-R2: fragmentof clay P1. Ib. Trench

7.5 Pl. Ic. TrenchRI-R2: polishedstone axe or hammer-axe, cm long: width along blade, 7 cm.; widthacross shaft-hole,4 -6 cm. 4"

Pl. Id. Trench RI-R2: pottery andbowlof Early Tra jar

south-east. Pl. Ha. JX: the beginning theseason's excavations, looking of Pl. Hlb. JX: Level7 wal

of pottery various styles. P1. IIc. JX: painted

Pl. Id. JX: paintedtum diameter rim,9- I cm of

P1. IIIa. PI, Burial 3, as found.

P1. IIIa. PI, Burial 3: grave goods after clea

with from theheaddress, tracesof wovenclothstill Pl. IVa. PI, Burial 3: bronzediscs
adhering: diameter, 2 15 cm.

Pl. IVb. Cylinder-seal found immediately Pi, bepeath Burial 3, w

diameter, o -85 cm.

Pl. IVc. PI, Burial5, asfound.

Pl. IVd. PI, Burial 5: detailof headdress beneath sk the

Pl. Va. PI, Burial5: gravegoodsaftercleaning.

Pl. Vb. P2, Burial 6: gravegoodsaftercle

citadel: viewfromthelargebuilding(4) with thecorner room(2) (rightrear) Pl. Vc. The Urartian
and the open court (12) (left rear).

Pl. Vd. The Urartian citadel: the perimeterwall, wit background.

Pl. VIa. General view of the Urartian north. citadel,looking

Pl. VIb. The Urartian citadel: thecolumn-bases, east. looking

Pl. VIIb. A pottery " salad bowl "from the Urartian citadel.

bowlsof redpolished ware from P1. VIIa. Two typicalUrartian thecitadel.

Pl. VIIc. The Urartian citadel: view looking south from the kitchen (5) in the foreground to the area of the columnedhall.

Pl. VIld. The Urartian citadel: the columned hall, looking west.

and from thecitadel: P1. VIlla. Bronzering with carnelian glass beads, 2 diameter, 8 cm. Pl. VIIIb. A groupof ironarrowheads fro

bastion: view lookingeast, with Pl. VIIIc. Part of the earlySasanianperimeter wall with a semi-circular part of the Urartianlevel (centre right).

bronzetassels,stu Pl. VIIId. Fragmentary

EXCAVATIONS

AT

HAFTAVAN

TEPE,

1969

137

The main effort of the 1969 season was devoted to excavations on the highest part of the site, the citadel. The great trench cut across it some forty years ago left the larger undamaged area to the north, the general contours of the summit giving it the semblance of a saucer, with the highest part along the rim and sloping down towards the centre. The principal aim was to uncover as wide an area as possible of the major building level found in the 1968 season.17 The attainment of this objective was hampered by problems of the disposal of soil but much more by the complex character of the later deposits. Their poor preservation, honeycombed as they were by innumerable pits, made the interpretation of the evidence a challenging task and the outcome none too rewarding. By the end of the season, however, when the workmen were all put to this task, the clearance of a considerable area of the main building level on the citadel had just been achieved, in spite of the sudden onset of rain and cold weather (Figs. Io-I I and Pls. Vc-VIIIb, VIIId). To shorten and clarify the following description of the plan (Fig. io), locusnumbers have been superimposed and are indicated in brackets after each related feature.
2 strings of beads>gi head-band P1 Burial 5

o0
O

stems loops anddiscs

.pins

J"pendant strings of carnelian and blue frit beads

iron bracelets

bowl

remains of small sheep

anklet scale -5 -10 15 cm bowl

Fig. 9. PI, Grave 5: the burial in situ.

As discussed below, there seems no doubt that this building level can be ascribed to the Urartian period; and there is evidence for a more precise date than simply within the whole time-span (c. 850B.C.) which could legitimately be so termed. The fact that such a long period can be called 600oo Urartian must be seen in the context of the firm political and military control of the northern half of the Urmia basin by the kings of Urartu, with only the briefest interruptions caused by the campaigns of Tiglath-Pileser III and Sargon II of Assyria.18 This level will henceforth be described as the Urartian citadel.
17 18

Iran VIII, pl. VIa-c; fig. 4. D. D. Luckenbill, Ancient Records of Assyria and Babylonia

(Chicago 1927), vol. I, paras. 795, 81 1-13 (Tiglath-Pileser III); and vol. II, paras. 139-78 (Sargon II's eighth campaign).

I
HAFTAVAN TEPE 1969

citadel(1968-1r969excavations). Fig. to. Plan of the Urartian

EXCAVATIONS

AT

HAFTAVAN

TEPE,

1969

139

It seems reasonable to interpret this building level as the residence of an Urartian provincial governor, administering a fertile district, the Salmas plain, from the settlement situated therein, whose modern equivalent is the town of Shahpur. If this is the earliest Urartian level at Haftavdn Tepe, it seems as if Urartian rule may not have been established without violence, for there is an underlying layer of burning. The extent of the Urartian citadel (Fig. Io) has revealed a feature not apparent at an earlier stage of the excavations, the uniform N.N.W-S.S.E. alignment of the buildings. Therefore it seems reasonable to reconstruct this as essentially one extensive building, laid out to a plan, with separate rooms and open areas, rather than a number of contemporary but not directly related buildings. This realization made the significance of the column-bases (P1. VIb and VIId) at once apparent, when taken together with the finding of Urartian red polished pottery, some of it of the finest quality, in the fill above the floor (P1. VIIa). These column-bases did not support a colonnade belonging to the building to the north (4, 5) and strangely masked by the blank wall of another building (8), nor for that matter did they support a colonnade belonging to the latter. The columns which they upheld supported the ceiling of a large hall (6, 7, 9) facing on to the open court (I2), with rooms in an upper storey above. The occurrence of Urartian red polished pottery, mostly bowls, both in the debris of this upper storey and in a similar context in the fill overlying the floor of a corner tower (2) seems to indicate domestic quarters, perhaps for high-ranking persons. In contrast, kitchen wares were found along one side of the largest structure (5), on the floor. The civilian character of this citadel is discernible by comparison with Urartian fortresses of varying size in the surrounding districts and at Bastam.19 Moreover, the very plan excavated at Haftavan Tepe shows that the citadel was probably never very seriously fortified, the reasons presumably being the contours and situation of the mound and the lack of stone in the immediate vicinity. Urartian fortressesalways required unlimited stone, though mud brick also was, of course, much used. Part of the Urartian perimeter wall is shown in the section of the east side of HY (Fig. I2), where four courses of mud bricks were found in situ above two courses of stone footings, with a small stone in the shallow foundation trench. The HY4A surface in Fig. I2 belongs to a terrace (3) associated with a secondary entrance to the Urartian citadel. The narrowness of this entrance suggests limited significance; but it was nevertheless guarded by two rooms (1, 2) apparently serving at least as porter's lodges. Later occupation and subsequent erosion along the steep north slope of the mound make it improbable that any trace of the approach to the terrace (3) will ever be recovered. Among the material recovered from the Urartian citadel are objects of undoubtedly Urartian appearance and others which are identical with or similar to grave-goods and other finds from the lower parts of the site outside the citadel. In the burnt debris marking the destruction of this level were scattered the remains of at least one burial, suggesting the disturbance of the level, probably in the processof stone robbing, some time after the destruction. A site such as Haftavdn Tepe, not immediately next to a source of stone, is inevitably vulnerable to quarrying of ruined buildings and the consequent problems in the elucidation of their plans. Near the column-bases, for example, only limited areas of contemporary stone paving were found in situ (Pls. VIb and VIId). The clearest proof of the Urartian date of this level is provided by the range of red polished and comparable pottery, mostly bowls of simple form and repaired from many sherds. In addition to these simple bowls (P1.VIIa) there were forms which can be paralleled at other Urartian sites: an unusual divided dish, perhaps for salad, with
strainer holes through the partition, is like one from Altintepe (P1. VIIb);20 and a fragmentary goblet is of the form found at Patnos (Giriktepe) and at Kayalidere.21 Among the finds from the burning of the citadel near the entrance rooms (I, 2), shown in the west section of HX as HX4B (Fig. I3), was a bronze fragment possibly from a quiver,22 bronze nails, studs and tassels, as well as a fragmentary example of a form of personal ornament characteristic of this period, a bronze ring threaded through a
19Wolfram Kleiss, (a) " Bericht fiber zwei Erkundungsfahrten in Nordwest-Iran AMI New Seriesvol. II(1969), especially ", pp. 2o-6; (b) " Ausgrabungenin der urartaischenFestung Bastam (Rusahinili) 1969 ", AMI III (x970o), pp. 7-65. 20 Kutlu Emre, "The Urartian Pottery from Altintepe",
21

Belleten XXXIII (1969), pp. 291-301, fig. 15. (a) Ibid., pl. III (2, 3), for cruder variants of this form; (b) Burney, AS XVI, pl. XV(b) and fig. I5 (Kayalidere). 22 Cf. ibid., fig. x8 (6, 6b) and pl. XVIII (c).

140

JOURNAL

OF

PERSIAN

STUDIES

SECTION AA

SECTION BBBe.-l

HAFTAVAN1969 , TEPE
2

'

on sectionsacrossthe Urartiancitadel: the lines of thesesections as indicated fig. io. are Fig. zr. Architectural

carnelian bead (cf. P1. VIIIa, also with a striped glass bead). Beads included many of carnelian, though the commonest were small, almost spherical yellow glass beads, a type hardly found in the burials below the citadel and distinctive of this level; a few beads of yellow glass striped with black were also found. In this same level (HX4B) was found an iron spearhead with carbonized fragments of its wooden shaft. In this same period there was evidence of a thriving industry producing ground stone vessels, best exemplified by a tripod and a bowl of basalt. Seven iron arrowheads, found together, were among other contemporary objects (P1. VIIIb). The history of this Urartian governmental residence and of the subsequent levels overlying its remains is by no means straightforward. In one part (6) there is very clear proof of replanning in a secondary phase, with the construction either of an entirely new room or of such a room created by a projection added to two screen walls flanking an entrance. One thing is obvious, that this room as
r.->ID

FfIT

WALL [W 12A

LE

?"r$ AVPPoy, .TE

LEVEL

IAFTAVAN EAST AREA 14" SECTION.

4 Fig. rI. Sectionof east side of HT.

TEPE

1969

)-

wow

OFIVMW!!
tik., I**-

WALF To? OFMoJMP

M" MOAE
RU.

UPI at.. .1o.

9

.
me

ma

0
L7

RAmeW
flLW

ae

ssus. 2o

MW
FILL

POFWALPA4QbUD

PARp

41WSiH

toccDranoMri
4
e

oFSOFT
an
-

OFfir

'"

wr

FILL NY-lie PWmumcsMAREA IMEx

14s WErT
/to
aWPts
wITP

III
Olbw

CD4seVm SEC
a

an

wrAs caAIZ

FUcIM*OF

4.w

u.O,.

fxeoez"

0]
opnacrnw0

C&/

* ss
2.

1 5
4

4.

TEPE 14AFTAVAN TEE W-AFTAVAN

AREA
HX

WEIT

Fig. z3. Section westsideof HX. of

142

JOURNAL

OF

PERSIAN

STUDIES

excavated must have been built later than the colonnade, against three of whose column-bases it abuts. It is an attractive if insufficiently documented theory that there may have been an approach by means of a staircaseto the upper storey, which would thus have extended over both the columned hall (6, 7, 9) and the kitchen to the north (5). The stairs would have had to be steep, but the height of the kitchen need have been no more than two metres at most. Only fuller exposure of the citadel, the main objective of the 1971 season, can help to confirm or refute this necessarily tentative reconstruction. The date of the destruction of the Urartian citadel, within whose lifetime the secondary phase (room 6) must be included, can be estimated from the pottery, presumably dating not much beforehand. The bowls with folded rim, found at such Urartian sites as Bastam nearby and Altintepe far away in the north-west corner of Urartu, seem from their context to date to some stage in the seventh century B.c.23 The forms of bowl from the citadel at Haftavan Tepe, however, are quite different, with simple rims and rounded profile, so that it seems virtually certain that they antedate the pottery from Bastam and Altintepe. The late eighth century B.C. consequently seems the likeliest date for the sack of this citadel; and it is tempting to suggest it was the work of Sargon II in his famous eighth campaign (714 B.c.). If so, perhaps this was one of the citadels of the land of Sangibutu which Sargon destroyed. Against this suggestion is the possibility that after the destruction of Ulhu, perhaps to be located at Marand, he marched through the district wherein Khoi is now situated, leaving the Salmas plain, with Haftavan Tepe, unscathed on his left flank.24 Immediately overlying the Urartian citadel level were fragmentary remains of well-constructed stone foundations, but at this height the depredations of stone-robbers,combined with the effects of the many pits, have left little more than an indication that there had been a renewal of prosperity, possibly in the seventh century B.C. and thus tentatively attributable to Rusa II (c. 685-645 B.C.),in whose reign Urartu enjoyed a political and cultural renaissance. Interesting though it would be to discover a Median level here, there is no evidence yet of any such remains of importance. It would of course be rather far north to expect this. In one area (G4) seven surfaces were traced before the level of the Urartian citadel was reached. This was typical of the succession of surfaces, often with little intelligible left of building remains. The one area where a reasonably clear story was obtained was along the perimeter of the north side of the citadel, where the east section of HY (Fig. 12) shows the sequence of the Urartian wall followed by a mud-packed wall which had only intermittent stone foundations; this in turn was followed by the latest perimeter wall (HY2A), an altogether more impressive structure, already partly uncovered in the 1968 season.25 This wall was found to have a semi-circular bastion (P1. VIIIc), a design which may suggest comparison with the defences of Bishdpir, built in the twenty-fifth year of Shahpir I (266 A.D.) rather than any date in the Parthian period.26 A later date than the early Sasanian period is scarcely plausible, however, in the light of the burials of Sasanian date discovered on the citadel in the 1968 season, and seemingly restricted largely to the north-west part. Whatever the precise interpretation of the discovery of these inhumations, which are termed Sasanian only as an abbreviation for " contemporary with the Sasanian empire ", two such burials were found in the 1969 season overlying the citadel wall here attributed to the late third century A.D. Moreover, a typical miniature glass bottle was found in the fill of a pit later than the early Sasanian wall, contemporary with the semi-circular bastion, against which it was cut (Fig. 13). These burials are likely to be no later than the fifth century A.D., implying a short duration for the early Sasanian citadel of Haftavan Tepe.

23

(a) Bastam: Stephan Kroll, " Die Keramik aus der Ausgrabungen Bastam 1969 ", AMIIII (1970), pp. 67-92

25

24

(b) Altintepe: Kutlu Emre, Belleten XXXIII, de (a) F. Thureau-Dangin, La HuitiemeCampagne Sargon (Paris 1912); (b) Luckenbill, AncientRecordsII, paras. 162-4.

(figs. I, 3, 4); figs. 8, 1o, I2.

Iran VIII, pl. VIIa. 26 For a reference to the inscription giving the date of the foundation of Bishapfir, see the note by A. A. Sarfaraz. Iran VIII. p. 178.

SHORTER NOTICES A SMALL BRONZE TRIPOD-STAND FROM WESTERN IRAN By P. R. S. Moorey

Although Elamite influences on the cultural development of her northern neighbours have received increasing attention in recent years,1it is not often that there appears among the numerous cast bronze objects reported from the Zagros region an example in which the influence of Elamite iconography is immediately apparent. It is this, as much as its intrinsic charm, which makes a small bronze tripod recently acquired by the Ashmolean Museum, Oxford, particularly interesting and instructive (P1. Ia).2 In form the stand is simple. Three vertical struts cast as bull-men support a broad ring of metal. Just above the feet, cast as bulls' hooves, a strip of metal runs round the struts to form a triangular brace. Only the torsos of the bull-men are rendered in detail; their bodies and legs are stylized in a plain, straight column terminating in a bull's hoof. From the top of the head in each case rise three feathers carefully chased on the surface of the upper ring, below them a pair of horns curve upwards to a point above the centre of the brow. A single hair-braid passes across the brow and falls in two long curls which frame both face and beard. Human facial features, including human ears, are boldly modelled. The beards are rendered with parallel rows of curls on the lower cheeks and wavy vertical curls falling from the chin onto the breast. Incised lines and dots on the upper arms suggest the short sleeves of a garment. Both arms are bent at the elbow with hands held clasped in front of the stomach. Surface examination suggests that each strut is a separate " lost-wax " casting to which the upper ring and lower brace were subsequentlyjoined. A very close parallel to this stand was acquired by the Museum fUr Vor- und Frihgeschichte in Berlin some ten years ago3 (P1.Ib). Another such stand, but with more crudely cast, stylized anthropomorphic struts in the form of men with raised arms, said to be from Khurvin, passed through the London Art Market in 19634 (Fig. i). Also in Berlin is an almost identical tripod-stand which has three feathers chased on the upper ring just above its junction with each strut, though the struts themselves are plain save for a bull's hoof as each foot5 (P1.Ic). Both the Berlin stands were said to come from the Kakavand region of Luristdn. The only archaeological evidence for the date of such stands as these comes from Hasanlil, where two comparable bronze tripod-stands were found in Level IV. In both cases the struts are plain, terminating in shod human feet, in one case wearing boots with turned-up toes.6 A more substantial tetrapod bronze stand from the same level has a figured design cast in low relief on the struts.7 Baked clay tripod-stands were a regular feature of contemporary pottery production in the region.8 Further to the south vessel-stands are shown on a number of the tenth-century decorated situlae regularly reported from Luristan.9 Whether these stands were of metal or baked clay is not of course clear, but
I am most grateful to Dr. P. Amiet for a photograph of the Louvre tripod; to Dr. W. Nagel for full details and photographs of the two tripods in Berlin and for permission to publish them here; and to Miss M. de Schauensee for information on the Hasanlfi tripods. E. Porada, AncientIran, pp. 70-2. 1970.402; 19.4cm. high, 8cm. diameter; Sotheby Sale Catalogue,May 4th I970, no. 37 (plate). 3 XIc J. H. Potratz, Die Weltkunst (1960), p. 35, 3546/1961; 30 fig. Io; 9-5 cm. high, 6.9 cm. diameter. 4 SothebySale Catalogue, March I ith 1963, no. io (plate); c. 12-7 cm. high; "Khorvin ". SXIc 4095/1966; 15- I cm. high, 7.8 cm. diameter.
2

HAS 58-216: R. Dyson, Jnr., Expedition(Spring I959), p. I6; cm. ArchaeologiaViva I (1968), p. 9go figure, 7 cm. high, 7"-4 diameter; HAS 62-755: 1o cm. high, 7 cm. diameter; now both in the Iran Bastan Museum, Tehran. 7 R. Dyson, Jnr., SurveyXIV (1967), pl. 1486A. 8 R. Dyson, Jnr., ArchaeologiaViva I, pp. 96-7, figs. 118, 121. 9 P. Calmeyer, BerlinerJahrbuch 5 (1965), pp. io-i i, fig. A5; pp. I2-I3, fig. A6; pp. 16-I7, fig. Aio.

143

144

JOURNAL

OF

PERSIAN

STUDIES

they seem to be slightly more substantial than the examples considered here and may be represented by a plain bronze tripod-stand reported from Luristan soon after the earliest wave of clandestine excavations there.10 Only one stand has so far been reported in the exotic cast bronze style characteristic of Luristan in Early Iron II.11 It is, however, further to the south still that recourse must be made to match the style of the bull-men on the Ashmolean stand, for it is in the " Middle Elamite " levels at Susa that their closest parallels are to be found. Both on the Untash-Nap(i)risha stela (Fig. 2) and on the glazed brick frieze from the Shilhak-Inshushinak temple (Fig. 3)12 demons, half men, half beast, wear the same beards, the same framing curls and comparable horns or horned crowns; but significantly not, so far as present evidence goes, the triple head-feathers. At Susa, moreover, the bull-men have prominent bulls' ears. Recently Barnett and Calmeyer13 have assembled and discussed most of " in the known representations of the " feather-crown the art of Iraq and Iran. The wide chronological and geographical scatter of the available evidence indicates

Fig. i.

After Sotheby

I963, lotno. o.

Catalogue,

March rith

Fig. 2. After Encylop6die Photographique de l'Art I (Louvre1935), pl. 269B.

o10A.

that no single hypothesis will explain its use. Sometimes it may be taken to identify certain specific peoples, perhaps Persians, at others it seems to have a ritual significance devoid of any particular ethnic affiliation. In the present context three representations of feather headdresses are especially relevant: the late third millennium rock relief at Sar-i Pul where in a triumphal scene of King Anubanini, King of the Lullubi, the leader of his conquered foes wears one;14 a disc pinhead of the eighth or earlier

A. xx
12

Godard, Les Bronzes du Luristdn(1931), pl. LIX.2 x18. aus Moortgat, Bronzegerdt Luristdn(1932), pl. XII.4o; now in the Pergamon Museum, East Berlin. P. Amiet, Elam (x966), pls. 284, 299.

13 R. D. Barnett, SurveyXIV (1967), PP- 300x ff.; P. Calmeyer, Actes de la XVII Ren. Assyr. Int. (Brussels 1969), pp. 184 if. 14 E. Porada, Ancient Iran, fig. 15.

SHORTER

NOTICES

145

seventh century on which a double-winged genius holding inverted lions (?) has a feather-crown;x5 and finally a ninth-century fragmentary local ivory carving from Level IV at Hasanli which shows a spearman wearing a feather " crown ".e These three occurrences suggest that, although the style of the bull-men on the Ashmolean tripod is "Middle Elamite ", what we have here is not the product of an Elamite metalsmith, but rather the work of a craftsman in the central Zagros region imitating such products in his own idiom. The feathers here denote one of the mountain tribesmen's own age-old nature-demons, not necessarilyknown at Susa, whose presence is also indicated on the Berlin tripods, once by feathers and hooves alone. The date of these tripods, following the indications from Hasanli IV rather than the more insecure basis of " Middle Elamite " iconographic parallels, may be set in
a century or so on either side of
900 B.C.;

total obscurity still in the Elamite homeland.

a period of

Fig. 3. After Encylop6die Photographique de

l'Art I (Louvre r935),pl. 276.

Fig. 4. After E. Pottier, Catalogue des

Antiquit6s Assyriennes (Paris 1924),

pl. XXX.153.

Finally, it may be possible on the basis of information presented here to claim for a ninth-century west Iranian workshop a bronze tripod-stand of unknown origin which has been in the Louvre for over a century.1 This stand (Fig. 4) is structurally very like those considered in this note, with plain struts and bull's feet. Its decoration is distinctive. On the narrow upper ring four protruding male heads are set at regular intervals, whilst identical heads appear on the lower ring at its junctions with the struts. These heads are so remarkably like a male head wearing a horned cap which formed the knob of a glazed wall-tile found in Hasanlil IV18 that some relationship may be assumed. As the glazed tiles at Hasanli are unlikely to have been made far outside the Solduz valley, if not actually on the site itself, the Louvre tripod too may well have come originally from a workshop in Azerbaijan
17 E.

W. Speiser, Vorderasiatische Kunst (1952), fig. 152 (lower). "5 Museumof Art (Novem6 V. Crawford, Bulletin of the Metropolitan ber I966), p. 128, fig. 14.

18 R. Dyson, Jnr., Survey XIV (x967), pl. 1484B; E. Porada, AncientIran, pl. 30 (colour).

pl. XXXI.

Pottier, Cataloguedes Antiquites Assyriennes(1924), no. 153,

146

JOURNAL

OF PERSIAN

STUDIES

or north Kurdistan. It has long been thought to derive from a site in Iraq, largely no doubt on account of the date it entered the Louvre.19 As recorded Assyrian military and diplomatic activity in the Urmia region begins in the ninth century the presence of a contemporary Iranian bronze tripod in Iraq need occasion no surprise, whether it went there as booty or tribute, particularly in view of the many western objects in Level IV at IHasanlfi.

A POSSIBLE SOURCE FOR THE SCULPTURES OF THE AUDIENCE HALL, PASARGADAE20 By Trudy S. Kawami
The Achaemenian ruins at Pasargadae, although long known and often described, have usually taken second place to the later and more elaborate remains at Persepolis and Susa.21 Because of the fragmentary state of the ornamentation of the Pasargadae buildings, emphasis has been placed on the architecture, especially as it indicates the early artistic influences on the Achaemenians.22 Enough of the sculpture remains, however, to provide a fruitful area for study. Achaemenian sculpture decorate the south-eastern and north-western doorways. The two jambs of each door bear the remains of almost life-sized pairs of composite beings, each pair mirrored on the opposite jamb. The images are carved in the black stone so that they appear to be standing on a shallow ledge about 75 cm. above the threshold. The north-westerndoorway is the more damaged of the two and all that can be seen on the jambs are a pair of bare human feet and a pair of clawed feet, both pairs pointing in the same direction. On the better preservedsouth-eastern doorway the figures exist from the knees down and give a clearer indication of their identity (P1. IIa and b). First comes a pair of bare human feet with the edge of a short skirt seen at the knees and a long fish-skin hanging down covering one leg almost to the ankle. This is followed by a pair of bovine legs separated from the preceding legs by a narrow rod or staff. The pleated end of a belt-tie can be seen on P1.IIb. These two figures can be reconstructed as a fish-garbed man followed by a standard-bearing bull-man. This occurrence of the fish-garbed figure is unique in Achaemenian art.24 the particular type of the figure popular since the time of Sargon II (721-704 B.C.). This type is characterized by the costume comprised of a long fish-cloak covering most of the head, body, and one
19 G. Perrot and C. II Chipiez, Histoire de I'Art dans l'Antiquite'

In the so-called Audience Hall, dated c. 546-30 B.c.,23 several provocative

examples of early

The image of the fish-garbed priest, known in Mesopotamia for close to a thousand years,25 depicts

(1884), fig. 393. 20 I would like to express my gratitude to the Kress Foundation whose grant through the School of Art, University of Iowa, facilitated the research resulting in this paper; to the Graduate College, University of Iowa, for aid in acquiring photographs; and to Dr. R. D. Barnett, Keeper of Western Asiatic Antiquities, the British Museum, and his staff for much appreciated assistance. A version of this paper was presented to the 181st meeting of the American Oriental Society, in Cambridge, Massachusetts, on April 6th 197 I. 21 For a summary of early European visitors to Pasargadae see the bibliography in the study by A. Gabriel, Die Erforschung der KenntnisderGeographie Persiens,die Entwicklung abendldndischen Persiens (Vienna 1952). For excavation reports see Ernst Herzfeld, AMI I, pp. 4-16, and the continuing reports by David Stronach, Iran I (1963) to III (1965). 2 Ernst Herzfeld, Archaeological History of Iran (London '935), pp. 1-43.

de L. Vanden Berghe, Arche'ologie l'Iran Ancien (Leiden 1959), pp. 20-3. C. Nylander, lonians in Pasargadae(1970), p. 128. 24 First published by Ali Sami in Pasargadae(Shiraz 1956), p. 49, who dismissed it as an aberration in Achaemenian art. See also Vanden Berghe, op. cit., p. 22. 25 The earliest known example is a Kassite cylinder seal of about the fifteenth century B.c. in the Pierpont Morgan Library: E. Porada, The Collection the PierpontMorgan Library,Corpus of of AncientNear EasternSeals I (Washington, D.C. 1948), p. 65 and pl. LXXX. I wish to express my thanks to Dr. Porada for calling this seal to my attention and for helpful comments regarding the motif. Whether the image represents a genius or a costumed priest is a separate question, and as such is outside the scope of this article. Two references to the problem are A. L. Oppenheim, " Akkadian pul(u)-h(t)u and melammu ", JAOS (1943), PP- 31-4; and Edith Porada, " Seal Impressions of Nuzi ", Annual of the AmericanSchools of Oriental Research XXIV (1944-45), p. I21.
23

Pl. Ia. Bronze tripod--Ashmolean Museum 1970.402.

Photo: Ashmolean Museum.

Pl. Ib. Bronzetripod-Berlin XIc 3546/1961.

Photo: Berlin.

Pl. Ic. Bronzetripod--Berlin XIc o495/1966.

Photo: Berlin.

SHORTER

NOTICES

147

leg, a knee-length kilt, and bare arms and feet. Customarily the priest carries a pail and sprinkler (Pl. IIIc). A comparison of the Pasargadae remains with a Sargonid seal suggests strongly that the Achaemenian figure was of this specific type. As examples of the Sargonid type in glyptic, stone, and metal are numerous, and as these examples come from Babylonia as well as Assyria, the particular type of the fish-garbed priest provides little indication of the source of the motif used at Pasargadae. It does, however, suggest that the source was of a late Assyrian date, Sargonid at the earliest. More revealing is the use of the image as a pair of life-sized guardians flanking a door (Pl. IIIb). The fish-garbed priest appeared in this manner, so far as is known, only in Assyria. Assurnasirpal II (885-860 B.c.) and his son Shalmaneser III (859-824 B.C.) employed the image at Nimrud in the Temple of Ninurta,26 and in Fort Shalmaneser;27 and later at Nineveh Sennacherib (704-681I B.C.) decorated seven doorways in his palace with the figure twice life-size.28 At Nimrud the fish-garbed priests wore earlier versions of the costume including short fish-cloaksand long, fleecey undermantles, or skin-tight scaley suits. The examples from Nineveh, on the other hand, are of the same short-kilted, long-cloaked type as those at Pasargadae (P1.IIIa). This parallel suggests that the Ninevite sculpture, specifically that in the Palace of Sennacherib, is the source of the fish-garbed priest motif found at Pasargadae. The figure associated with the fish-garbed priest at Pasargadae also demonstrates a connection with Nineveh and the Palace of Sennacherib. The bull-man with the standard as seen in the remains of the south-eastern doorway is a fairly common Mesopotamian motif, found often in glyptic and on clay but plaqueS29 rarely in monumental sculpture. Similar Assyrian use of a standard-bearing figure as a door guardian in stone occurs in the Palace of Sennacherib at Nineveh.30 Here the being is a curlyhaired human rather than a bull-man, but the function of supporting a standard and guarding a doorway is the same in both cases. Furthermore, the practice of grouping the fish-garbed priest with other composite beings has parallels only in the Nineveh palace where in at least three instances the fish-garbed priest was found in the company of other colossal beings.31 These accompanying figures also give an indication of the identity of the beings found on the north-western doorway at Pasargadae. Among the Ninevite companions of the fish-garbedpriest is a figure with a lion's head and talons instead of feet,32with which the badly damaged Pasargadae examples can be identified. The bare-footed partner was probably a short-skirted warrior with a raised mace. These figures occur individually as Assyrian apotropaic figurines, but rarely as a group. The only examples of this same combination in monumental sculpture are two large reliefs from the Palace of Sennacherib at Nineveh,33 providing yet another connection with the sculpture of the Audience Hall at Pasargadae. Some of the unique motifs found in the sculpture at Pasargadae have been explained as the result of Babylonian influence on the Iranian plateau.34 Yet it can be shown that the closest parallels for the fish-garbed priest and the associated motifs in the Audience Hall are not Babylonian but Assyrian. These motifs, their combinations, and the manner of use occur, furthermore, in a particular palace at Nineveh, that of Sennacherib. This derivation and its careful adaptation suggests that one of the early sources of Achaemenian art is Assyria.

26

27

C. J. Gadd, The Stonesof Assyria (London n.d.), p. 249. David Oates, " Excavations at Nimrud (Kalhu) 1962 ", Iraq XXV (1963), pp. 29-30. These figures are not sculptures, however, but frescoes. 28 The image was found in pairs in Rooms XXV, entrances b and d; XXXVIII, entrances g and i; LX, entrance b; and LXV, entrances b and h. For a description see A. H. Layard, Nineveh and Babylon (London 1853), pp. 343-4, 442 and 460; and A. Paterson, The Palace of Sinacherib(The Hague 1915), pp. 8, io and 13. 29 M. E. L. Mallowan, NimrudandIts Remains (New York I966), II p. 388, n. 28. 30 At the entrance to Court VI. See S. Smith, AssyrianSculpture in the British Museum: From ShalmaneserIII to Sennacherib

(London 1939), p. XXXV; and Paterson, op. cit., pl. 22. The only other known example of this motif is on a slab from Nimrud dated to the time of Esarhaddon and illustrated in R. D. Barnett and M. Faulkner, The Sculptures Tiglath-Pileser of III (London 1962), p. 23 and pl. CXII. 31 Rooms XXV, entrances b and d; and LX, entrance b. See Layard, op. cit., pp. 442 and 460; and Paterson, op. cit., p. 8. 32 Room LX, entrance b. See Layard, op. cit., p. 460. 33Ibid., p. 462. There is also an unpublished drawing in the Malan sketchbook in the British Museum Library, Add. Ms. 45360, f. 24, showing the upper portion of a lion-headed figure at Nineveh. 34 Sami, op. cit., pp. I18-19.

148

JOURNAL

OF

PERSIAN

STUDIES

Independent Persian contact with the Assyrians existed at least as early as the reign of Assurbanipal (668-626 B.c.) when, according to contemporaneous inscriptions, Cyrus I sent his eldest son Arukku to Nineveh with gifts of tribute,35 undoubtedly including the excellent Persian horses. It is possible that he and his entourage returned with not only portable examples of Assyrian craftsmanship, but also with memories of the impressive royal palace. As a member of the princely Persian houses, Arukku would have been in a position to effect the spread of Assyrian motifs and styles by his patronage. Later many Persians were in Nineveh aiding in the Median sack of the city in 612 B.c. Furthermore, it is unlikely that the great buildings were immediately obliterated. The successive resettlements of Nimrud after its fall suggest that the architecture was habitable for some time.36 Similar resettlement was probable at Nineveh,3sa providing further opportunity for Persians to view the captured Assyrian capital with its royal buildings. At Nineveh the fish-garbed priest was connected with the person of Sennacherib by reason of its extensive use in his palace. Here it functioned as a royal protector, an important but by no means usual task of the image. At Pasargadae the motif also appears in the doorways of a royal rather than religious building, as the magical guardian of the ruler. The similarity in the use of the motif at Nineveh and its use at Pasargadae further reinforces the possibility of a Ninevite source for the Pasargadae sculpture. The form and use of the motif suggest that the Persians were aware of at least some of the significance of Assyrian motifs and utilized them to establish their own position and power. As a new political power it was imperative for the Persians to be as important, both militarily and culturally, as the Assyrians before them.3" One way of achieving this importance was to awe visitors to the capital with great buildings and brilliant decoration. To this end the Persians employed in the Audience Hall, where official visitors were most likely to go, the motifs and figures which were the magical companions of the Assyrian king. This association emphasized the idea that the Achaemenians were protected by the same powerful genii as were the other great kings before them. In other words, the fish-garbed priest was now used in a political as well as religious manner to underscore the fact that the Achaemenian ruler was the legitimate successor to the Assyrians. The intention was to establish the young Achaemenian dynasty as heir to Assyrian power. The use of the motif from the time of Sennacherib for political reasons sets the early sculpture of Pasargadae apart from the work that was to follow. It marks the art of Pasargadae as a distinct and separate stage in the development of Achaemenian sculpture.

NEUTRON ACTIVATION ANALYSIS OF SOME OBSIDIAN SAMPLES FROM GEOLOGICAL AND ARCHAEOLOGICAL SITES By A. Mahdavi and C. Bovington*
Obsidian is a volcanic glass which occurs in restricted sites. It fractures concoidally and gives very sharp cutting edges: cutting tools and other objects made from this material were used by prehistoric
* Seconded, under Technical Aid agreement by British Government Overseas Development Authority, to Tehran University Nuclear Centre, P.O. Box 2989, Tehran, Iran. K6nighaus. Kyros I, ein Zeitgenosse Assurbanaplis ", AfO VII (1931-32), PP- 2-4; and R. Campbell Thompson and M. E. L. Mallowan, " The British Museum Excavations at Nineveh, 1931-32 ", AAA XX (1933), pp. 86 and 9536 The fact that Nimrud was not completely abandoned is suggested by three post-Assyrian occupations of the citadel
35 Ernst F. Weidner, " Die alteste Nachricht uiber das persische

producing pottery identical with late Assyrianware: David in Oates, Studies the Ancient Iraq (London Historyof Northern 1968), pp. 58-60. 36aIbid., pp. 60 f. 37 Cyrus was not yet king of Babylon. Thus he would resort to Assyrian rather than Babylonian imagery to emphasize his legitimacy as a ruler.

Audience south-eastern Pl. Ila. Sculptured doorway, Hall, Pasargadae.(FromVanden Berghe, Archeologie de 1'IranAncien, pl. 22a.) doorjamb,

Pl. IIb. Opposing Audience Hall, Pasargadae.Photocourtesy John G. Corbett. of jamb, south-eastern doorway,

CaesarMalan at Ninevehon June r5th on Caesar Malanat Nineveh June Pl. IIIb. Drawing made by Solomon Pl. IIa. Drawingmade Solomon by in to ioth i850. The HebrewwordDagon found next to the tail of the i85o. A noteon thereverse Malan's handgivesLayard permission use this in book that the drawingas an illustration a forthcoming on thecondition thedrawing imagewas addedby the artistand reflects nineteenth-century upper Add. Ms. 45360, f. 30o. to priest with the biblicalDagon. The be returned Malan. The British MuseumLibrary, identification thefish-garbed of Photocourtesy the Trustees theBritishMuseum. BritishMuseumLibrary,Add. Ms. 45360, f. 31. Photocourtesy of of of the Trustees theBritishMuseum. of

datableto the late eighthor earlyseventh centuries Pl. IIIc. Cylinder impression, seal B.C., thePierpont MorganLibrary,acc. no. 773. Photocourtesy thePierpont of MorganLibrary.

SHORTER

NOTICES

149

man and are found in many prehistoric archaeological sites throughout the world. These sites are frequently far away from the geological sources of obsidian, and therefore suggest the existence of considerable trade over large distances. In the Middle East, obsidian artifacts appear in sites from c. 7500 B.c. and the end of obsidian trade coincides with the development of copper and bronze. It is possible to characterize and relate archaeological samples and geological sources of obsidian in terms of their trace element composition. In the analysis of these trace elements neutron activation analysis has proved most satisfactory. The classification of samples is mainly based on Na and Mn content and Na/Mn ratio. A comprehensive study of geological sources of obsidian in the U.S.A., Greece and Turkey has been carried out by G. A. Wright,38 C. Renfrew39 and A. Gordus,4? but not all the possible sources in the Middle East have been sampled. In the present work we have analysed a number of excavated samples and some source samples for their Na and Mn content. The map (Fig. 5) shows the geographical location of the samples. The excavated samples are from three different archaeological sites in Iran, from the Hasanlti area, from Susa and Jaffarabad, and surface finds from sites in the Marvdasht plain. Samples 13-17 are from different geological sources in Central Anatolia and the Lake Van area in Turkey, see Table I. Table I Sample Number I 2 3 4 5 6 7 8 9
o10

Archaeological or Source Site Mohammair, IHasanli, Western Azerbaijan Pisdeli, Ijasanlil, Western Azerbaijan Balukchi, .Hasanl, Western Azerbaijan Balukchi, Ijasanlti, Western Azerbaijan Balukchi, I.asanlfl, Western Azerbaijan Balukchi, IHasanli, Western Azerbaijan Balukchi, Ijasanlfi, Western Azerbaijan Jaffarabad, south-west Iran (745 1) Acropolis, Susa (940) Jaffarabad (872) Apadana Palace, Susa Apadana Palace, Susa (1035) Nemrut Dagh, Turkey (Lake Van area) Suphan Dagh, Turkey (Lake Van area) Acigol Crater, Turkey (Central Anatolia) Gollo Tepe, Turkey (Central Anatolia) Hassan Dagh, Turkey (Central Anatolia) Marvdasht, Firs Marvdasht, Fars Marvdasht, Firs Marvdasht, Firs Marvdasht, Fars

Probable Period Iron Age 4000-3000 B.C. 4000-3000 B.C. 4000-3000 B.C. 4000-3000 B.C. 4000-3000 B.C. 4000-3000 B.c. c. 4000 B.c. c. 4000 B.C. c. 4000 B.C. c. 4000 B.c. c. 4000 B.C. Source sample Source sample Source sample Source sample Source sample B.C. c. 4500-3500 4500-3500 B.C. c. c. 4500-3500 B.C. c. 4500-3500 B.C. c. 4500-3500 B.C.

II
12

13
14*

15 16 17 18
19
20

21 22

* Sample 14 is a very poor quality obsidian.

The analyses were carried out by neutron activation. The cleaned samples and standards of Na and Mn were activated in a flux of Io10on/cm.2/sec., in the Tehran University Nuclear Reactor. The induced activities were measured using a 40 cc Ge(Li) detector model L-612, Nuclear Diodes, and an
Near-EasternTrade: Wright, ObsidianAnalysis and Prehistoric 75oo to 35oo B.C. (Museum of Anthropology, University of Michigan 1969), no. 37. 39 C. Renfrew, J. R. Cann and J. E. Dixon, B.S.A. 6o (1965), 38 G.
o40 Gordus, G. Wright and J. Griffin, " Obsidian Sources A.

p. 255.

Characterized by Neutron Activation Analysis ", Science I61 (1968), pp. 382-4.

150

JOURNAL

OF

PERSIAN

STUDIES

MeV of 8oo channel analyser, Intertechnique model DIDAC 8oo, the peak area of y-energy 56Mn and y-energy I-37 MeV of 24Na were evaluated according to Covell.41 The sampleso'-84 classified are in five groups according to their Na and Mn content, see Table 2. Samples 14, 15 and I6 represent the lowest Na content but differ in their Mn content and are therefore considered in two separate groups A and B. The remaining two source samples and the excavated samples are placed in three groups, C, D, E. Group C contains only excavated samples and they do not compare with any available source samples. The closest source comparable to group D is sample No. 17. Source sample I3 can be fitted very well with group D.
BLACK SEA

LAKE
*is

TUZ

TUR

KEY

13io

-,
QTEHRAN

?N
',IRAQ

6.12 SUSA

IRAN

.... 0 SCALE IN 200 MILES 400

8-22 ,SHIRAZ

location samples. Fig. 5. Mapshowing of It is evident that the use of Na/Mn ratios to identify samples can be misleading, for example samples 5 and 14 have identical ratios but consideration of their absolute composition would clearly lead one to separate them. The near identical results obtained for samples 15 and I6 are, at first sight, disturbing: however, according to the donor, R. A. Watson, these two flows are within a few kilometres of each other and may represent one geological source. It is hoped to obtain samples from various parts of these flows for further study. Neither source, however, appears to have been a major supplier of obsidian. The similar composition of the excavated samples comprising group C, Susa, Jaffarabad and the Marvdasht plain is not surprising since on typological evidence the areas from which these samples were obtained are contemporary and show strong evidence of considerable cultural contact. Hassan Dagh (Central Anatolia) appears to have been the source of obsidian of the Hasanlfi samples comprising group D but does not appear to have participated in trade over longer distances. The appearance of obsidian in an Iron Age context at Mohammair (sample i) may at first appear
41 D. Covell, Anal. Chem.31 (1959), p. 785.

SHORTER

NOTICES

151

surprising. It should, however, be remembered that in the vicinity of its sources obsidian is still used to this day for crude agriculturalimplements and presumably trade in obsidian and its use persistednearer the sources long after the larger scale, long distance trade had finished. Nemrut Dagh (Lake Van) seems to have been a major source, having supplied the samples comprising group E. This shows that trade existed between Nemrut Dagh and Susa and Jaffarabad (more than 2000ooo km.), going by way of Hasanlfi. Further studies will be undertaken to resolve the problem of the Acigol Crater and Gollo Tepe source and more samples, both excavated and source, will be analysed to build up a better picture of prehistoric trade in these areas.
Table 2

Group
A B

Sample
14*
15* I6*

Na%
5-00oo 3-o6 3"o7

Mn% 102o
48 6-7 6-5

Na/Mn
62

45 47

II

3"45
3 36 3.41 3"39

6-3
6-5 6-35 6-43 6 I9

55
52

18
19
22

55
52

3 E
6 17*
2

3"43 3.65 3"76

3.70 4-02

6*4 6-8
6-5 7"2 5-8 5"4

55 57 55 68
68
62

54 56

3"95 3-68 3-85 3-68

4-02

5 7 9
10
12

3"7o

5"9 5.7 5-6 5-8

5"9

13*
20 21 * Samples from geological sites.

3.66
3"92 4-OI

67 66 68 63
66

3"99

5"9 5"79
5"72

69
70

THE JAMSHIDI OF KHURASAN: AN HISTORICAL NOTE By Andre Singer

Within that area of Khurasan between Mashhad and Birjand, and up to the Afghanistan-Russian border in the east, live a number of tribal groups that have found themselves bounded by the extensive Turkoman population to the north, the Baluch to the south, the Dasht-i Kavir to the west, and the political boundaries to the east. Most prominent among these are Hazara, Timuri, Kurdish and Barbari peoples. A smaller group, now only to be found in five villages of Persian Khurasan,42is the Sunni Afghan tribe of the Jamshidi.
42 Meaning that only in these five villages (MahmfidTabid Bla,

Amghun, Qal'ah-i Mfiri, Tavakkuli and Kalat-i Sfifi) do the people regard themselves as a separate ethnic group with the tribal name of Jamshidi, as opposed to being the Jamshidi

of the Timuri tribe. In Afghanistan there are consider.ta'ifa more ably Jamshidi than in Iran, but it is impossible to estimate their numbers there.

152

JOURNAL

OF

PERSIAN

STUDIES

The following brief account deals with the presence and the demise in the nineteenth century of what was once a powerful tribe in Afghanistan and a potentially powerful one in Iran.43 The vicissitudes of nineteenth-century politics and military activities are the major factor contributing to the decline of this tribal group. Although processessuch as the ease with which the Jamshidi were assimilated into neighbouring groups are of importance, it is the political events that precipitated the tribal upheavals that will be detailed here. References to the Jamshidi before the nineteenth century are rare. V. Barthold (translated in Schurmann)44claimed not to have found any medieval references to the Jamshidi, and Schurmann's al-Mulik of I1725-29.45 Following the successful own researchesreveal nothing earlier than the Tadhkirat campaign by Shah 'Abbas I against the Uzbeks in 1599 the Jamshidi Sultans, with those of other tribes in the vicinity of Herat-Qipchaq, Timuri, Taheri and Hazara-came to Herat to pay Shah 'Abbas homage.46 Ten years later the Jamshidi are again reported around Herat, this time fighting against the 5000 Almanchi Uzbeks brought via Kakh from Jam, Khwaf and Bakharz.4 During the reign of Nadir Shah there is further reference to the Jamshidi fighting in the Herat district.48 They must at this time (early eighteenth century) have been a formidable fighting force, for Qazem records them as having participated in two major conflicts, on both occasions being prominent in the battles.49 The first of these was the defence of Herat against Nadir Shah in 1731. Although the Jamshidi, Taimani and Hazara were independent of the Vazir of Herat, Zaman Khan Abdali, they did support the defence of the city and together with the Abdalis and Ghilza'is were defeated for their troubles. Then again, of 6000ooo the Jamshidi under the leadership of Shadi Khan joined with Dervish 'Ali Khan in his rebellion against Nadir Shah. They therefore participated in the attack on and capture of Balkh, after which they returned to Herat.50 At the beginning of the nineteenth century, the Jamshidi were still a belligerent tribe,51fighting with the Kala Nau Hazara52and selling slave captives to merchants (probably Turkoman traders) of Khiva and Bukhara. The origins of the Jamshidi are unclear.53 Schurmann's evidence is confusing; he claims that a Persian origin for these people is unlikely, but seems to base this observation on the affinities drawn between their yurt-type tent culture with that of Central Asia.54 Lambton classes them as one of a group of Turkish tribes but gives no reason for this classification except to point out that her division was one of convenience, " leaving aside ethnological, anthropological and sociological criteria ", and was thus not attempting to show origins.55 The most recent authoritative attempt at classifying the Jamshidi is by Ferdinand who, using the evidences ofJanata and Ligeti, denies that, as a part of the ChahdrAimaq, they are of Turkish or Mongol origin as has been generally held.56 He points out that the Aimaqs57
has the existence of the Jamshidi as a tribal group in Iran been acknowledged. None of the other census materials seem to mention them. For more detailed ethnographic material on the Jamshidi in Afghanistan see H. W. Bellew, An Inquiry into the Ethnographyof Afghanistan (London I891); H. C. von der Gabelentz, " Ueber die Sprache der Hazaras und Aimaks " in ZDMG XX (1866); H. F. Schurmann, The Mongols of " EthnoAfghanistan (The Hague 1961); K. Ferdinand, graphical Notes on Chahir Aimaq, Hazara, and Mogh61 ", Acta OrientaliaXXVIII (1964). 44 Op. cit. 4 Translated by V. Minorsky (London I943). 46 Iskandar Beg Tfirkoman, Tarfkh-i 'Alam Ardyi 'AbbdsiI. 41 Ibid., II. 48 MahmfiadQazem, Ndma-yi 'Alam ArdyiNddiri; M. Elphinstone, An Accountof the Kingdomof Caubul(London 1879), p. 336 n. 49 Ibid. 50Ibid. 51 J. B. Fraser, Narrativeof a Journey into Khorassan... (London " Report on the Tribes ", 1825), p. 256; P. J. Maitland, IV Commission, (1891i), etc. Afghan Boundary 52 In Iran the Shi'a Hazdrajtt Hazdra are known as Barbari and the name Hazdra used only for the Sunni Kala Nau Hazdra. 53 The etymology of the name Jamshidi is unknown to the tribesmen themselves, although vague references were made to
43 Ironically, only in the I97I (tenth) edition of the Iran Almanac

Jamshid in folk mythology. I have found no decisive literary referenceson this point. "5Op. cit., pp. 53 if. of 55A. K. S. Lambton, art. " Ilt " in Encyclopedia Islam'. Admittedly the yurt as mentioned in connectionwith Schurmann, above, can be seen to have strong affinitieswith those of the Turkoman, rather than with those from Mongolia (Ferdinand,op. cit., pp. 197-8), but this in itselfis no evidence for ethnic origin. C. M. MacGregor, "A Contribution Towards a Better Knowledge of the Topography,Ethnology, Resources and History of Afghanistan", Gazeteer Central of Asia,partII, p. 134,vaguelylumpsall Aimdqsinto an" Afghan origin" group, but gives no evidence and can be disregarded here. 56 In op. cit. See also A. von Janata, Die Bevolkerung Ghor (Vienna 1963), and L. Ligeti, " Recherches sur les dialectes mongols XIV et turcsde l'Afghanistan ActaOrientalia ", Hungarica (1954). ' Referringto the ChaharAimiq group, comprising combinaa tion of some or all of the following: Timuri,Jamshidi,Hazdra (Kala Nau), Firfiz-Kiii,Taimani and Zfiri. Both Ferdinand and Schurmannagree on the vaguenessof the term Chahar Aimqq (lit. Four Tribal Groups or Settlements) as a useful classificatoryterm. Without new evidence there is no point in further speculation here. The earliestreferenceI have so far found to the Jamshidi as a part of the Aimaq is that in Turkoman,op. cit., vol. II.

SHORTER

NOTICES

153

have been exposed to Central Asian (Turkish) influences and to extensive inter-marriage " especially in the case of the Jamshedi, Firuzkui ", etc. There seems to be little to contradict the view held by Ferdinand and others,58 and by the Jamshidi themselves, that they are of Persian origin. During the reign of Zaman Shah in Afghanistan (i1793-18oo), the Jamshidi appear to have been mainly living in the region around Badghis with pockets in Karriikh, Mary, and possibly as far north as Urganj.59 That they were reported so far north is not surprisingwhen it is remembered that they were nomadic pastoralists and thus prepared to move wherever there was sufficient pastureland for their herds. They could, on the other hand, only have survived in Urganj and Mary by the agreement and sufferance of the Tekke Turkoman tribes. It is unlikely that any remained there after the Russian occupation of those areas (see below).60 It is highly probable that small groups of Jamshidi moved to the Jam and Khwaf region under the protection of Fath 'All Shah Qajar'sadministration there, following Qilich Khan Timuri's abortive attempt at supporting Zaman Shah's designs on Herat in 1802; but the first major migration from Afghanistan to Persia came during and immediately following the hostilities between Persia and Herat after the siege of 1856. At the beginning of the nineteenth century, the Jamshidi and the Kala Nau Hazara were weakening each other by continual fighting in Badghis, leaving the leadership in the Herat district of both the Timuri and Jamshidi to Qilich Khan. They were thus used against Mahmlid Shah Durrani and doubtlessjoined the Timuri in their flight to Khwaf. With the ascendency of Shuja' al-Mulk in Afghanistan (1803) and the return of the blinded Zaman, many Jamshidi would in any case have returned from Persia to Afghanistan.61 During the brief occupation of the Herat region by the army of Nasir al-Din Shah Qajar, an enforced migration of the Jamshidi in two waves took place. Husam al-Saltana*62had the greater part of them moved from Herat to Pul-i Khattin and Shfiriya. Then approximately 200ooo63 families moved with Allah Yar Khan to Sar-i Jam and thence to Kaneh Giisheh, Karbukkeh, Mashhad and Qfichan. Meanwhile, in Afghanistan, those Jamshidi who had moved to Andkhui returned to Herat through Badghis.64 From this time onwards, the tribe grew weaker and its history became one succession of movements after another as they found themselves either fighting the Turkoman or Hazara in Khurasan, or else being shifted from area to area by Persian and Afghan rulers. A further blow was dealt to the tribe by the 1871 famine which almost completely killed off those Jamshidi still surviving on the Persian side of the border.65 The leadership of the Jamshidi during the disturbances of the eighteen-eighties was in the hands of the three sons of Khan Agha (Mahdi Quli Khdn),66and it was under these men that a further and irreversibledecline in Jamshidi fortunes came about. By 1885 a British news-writer, Mirza Muhammad Taqi Khan, was able to report to Sir West Ridgeway (who was leading the British Boundary Commission in Afghanistan) that the Jamshidi " have recently lost their status as recognized tribes under their own chiefs and are reduced to the position of ordinary peasants ",. This was an exaggeration, but nonetheless a reflection of the straits into which the Jamshidi appear to have fallen. A great deal of their trouble stemmed not so much from local antagonisms with other tribal, peasant or religious groups in Persia and Afghanistan, but rather from their unfortunate geographical location in the heart of the region for which there were British, Russians, Persians and Afghans struggling for
58 Minorsky, op. cit., p. i88, J. Stewart and Maula Bakhsh in vol. I, p. 131, of Gazeteer Persia (1910o): " they were originally of Persians but through continual contact with the Turkoman have lost nearly all trace of their Persian origin ". 59 Stewart, op. cit., p. 131. 60 Isolated tribal groups were likely to be found along the border with Russia even after the border establishment in the eighteeneighties. This tribal transplantation was done by Amir 'Abd al-RahmAn in Afghanistan in much the same way as Shdh 'Abb3s I had done it in Persia (see G. Curzon, Russia in CentralAsia (London 1889), p. 365). 61When, however, the Aimdqs formed part of the army of Sfifi Islam and "]zbak Mulld in I8o6 and were defeated, it is unlikely that any Persian Aimaqs would have wished to return under the protection of Firfiz al-Din in Herat and leave Persian Khurasan. I.e. Sultan Murdd Mirzi, born in 18i17 the son of 'Abbas MirzA Na'ib al-Saltana. 63 Stewart, op. cit., p. 131. 64 This was done under the leadership of Mahdi Quli Khan, Khan Aghd, who replaced Mir Ahmad Khan Jamshidi as chief of the northern Jamshidi in 1857 at the latter's death, and who himself was later murdered by Sardar Ayylib Khan in Herat (i88o), despite the fact that Ayyfib Khan was his son-in-law. 65 According to C. E. Yate (Khorassan Sistan (Edinburgh 1900), and p. 32) their numbers fell to a mere I50ofamilies at this time. Famine struck again in 1875, but there are no figures on how this affected the Jamshidi. c Sc. Aminfillah Khan, Yalangtfish Khan and IHaidar Quli Khan. 67 Simla Records,Foreign Dept., Secret F, no. 246.
2

154

JOURNAL

OF

PERSIAN

STUDIES

control. The ambitions of Alexander II and the Russians culminating in their Central Asian annexa",69 and the vacillations of the Persians and the between these policies, meant that the Russian pretence that they were not expanding their Afghans border, but merely punishing the depredations of the Turkoman, was generally swallowed.70 It meant however that the Turkoman tribes, particularly after the slaughter at G6k Tappa,7 gradually directed their raids southwards, and it was upon the Aimaqs and the Khurasanian peasants that the brunt now fell. Although the Jamshidi were noted fighting men, they were no match for the numerically greater and more mobile Turkoman moving southwards.72 When therefore in 188o Yalangtish Khan took over his murdered father's position as paramount Khan of the Jamshidi, the tribe had not surprisingly been alienated from Ayyfib Khan73 and was distrustful and frightened of the Russians' approach. If the Jamshidi had been aware of the vacillations in England over whether Herat and the surrounding area should be ceded to Persia,74 they might have been equally distrustful of the British. As the change of government in Britain in I88o led to the abandonment of this policy of cession, the Jamshidi retained the hope that the British would keep back both the marauding Turkoman and the expansionist Russians. Yalangtfish Khan showed his proBritish attitude by, as governor of Pandjih, helping Yate (see above, n. 24) and his party withdraw from the area in I1885 at the Russian approach. Some Jamshidi, however, were making approaches to the Russians and suspicion was therefore thrown on the whole tribe. The British (through Ridgeway) wanted to maintain the status quo of the Jamshidi and Hazara and so appealed to the Amir 'Abd alRahman for a policy of conciliation. He agreed to this, but only reluctantly, as his letter to the Foreign " Savageness and rudeness are Secretary of the Government of India shows: ingrained in them, and the clay with which their bodies are formed is prepared with the water of infidelity . . . they are nomads, and [therefore] ungrateful."'75 Mirzd Taqi Khan warned the British of Jamshidi him, he was able to do little to prevent Amir 'Abd unrest,76 and although Ridgeway agreed withMuh.ammad al-Rahmdn from crushing them in I886, firstly by moving them out of Bald Murghab and replacing them with Ghilzd'is and Ishlqzd'is,77 and secondly by imprisoning and killing both Yalangtfish Khan and Aminullah Khan and replacing them with their younger, weaker and less popular brother Haidar Quli Khan. Meanwhile, those Jamshidi who had attempted previously to settle in Sar-i Jam on the Persian side of the frontier found the Turkoman too hostile, and therefore divided up into two groups. The larger group was sent back to Herat by Allah Ydr Khan; the other group settled around Khaneh Gisheh and Mashhad. Allah Ydr Khan found that Sarddr Ya'qib Khan did not want him in Afghanistan, and so returned to Persia to replace Isma'il Khan78 as chief of the Jamshidi there, whose numbers had however dwindled to approximately 300 families. By I888 all the Jamshidi in Persia were living in Pas Kamar
tions,68 the British diplomacy of" masterly inactivity
Khiva, I839; Aq-Masjid, I853; Tashkent, I865; Samarkand, 1869; Qizil-Sii (Krasnovodsk), I869; Khokand, I875-76; G6k Tappa and Akhdl, 1881; Mary, I884; and Panjdih, I885. 6 See F. Kazemzadeh, Russia and Britain in Persia, 1864-1914 (1968) for a description of this diplomatic struggle. The expression " masterly inactivity " was not coined for this particular border, but it remains an adequate description nonetheless. 70 Eventually even by Ndsir al-Din Shah who, after I869, found that he could not count on British support; also his Foreign Khan had quarrelled with the British Minister Mirz I.Husain Charg6 d'Affaires in Tehran, Thomson, over British inactivity. 71 An estimated 14,000-15,000 Turkoman were reported slain in the battle and only fifty-nine Russians (Kazemzadeh, op. cit., p. 76). 72 However, Maitland (op. cit., vol. I, p. xx) points out that there was an agreement and friendship between the Jamshidi and the Sariq Turkoman, especially over the division of Panjdih lands. This was possibly a result of the Sariq also being weakened by the Tekke and thus needing an ally. Although the cavalry escort for the British Afghan Boundary Commissioners was
68

made up of Jamshidi irregulars (1884-85), the Turkoman had a tradition of horsemanship and horse-breeding that had been instrumental over the decades for maintaining their independence. 73 Up till his defeat in 1881 by 'Abd al-RAli.mn, Ayyfab Khan still controlled Herat and its environs. 14 Eastwick's communication to the F.O. in I871; Rawlinson to the F.O. also in 1871; Thomson to Salisbury, 1879, no. 253; etc. 75Simla Records,op. cit., no. 244. 76 Ibid., no. 246. Most of them being moved to Ubeh, with a smaller group "77 going to Kushk (Maitland, op. cit., vol. II, p. 602). 7 Ismd'il Khan was the paramount Khan of the Rauti tribe, a group which moved to Herat from Kandahar and became divided into two parts, one linking up with the Taimani, the other with the Jamshidi. During Allah Ydr Khdn's absence, Ismd'il Khan, who had taken his group of Rauti to Persia with the Jamshidi, was regarded as the strongest leader. Allah Ydr Khan had left Sar-i Jam for Herat after the Persians had realised he had sent many of his tribe without permission back to Afghanistan.

SHORTER

NOTICES

155

under the leadership of Dhulfiqar Khan, who had replaced his father as paramount Khan. The Jamshidi of Khurasan were moved to their present location north-east of Turbat-i Jam by the government of Nasir al-Din Shah, following the flight of Dhulfiqdr Khan to Russia in 1889, when he was suspected by the Persians of negotiating with the Russians.7" The number of Jamshidi in Persia steadily increased and was swollen by those families escaping the pressures of the Amir 'Abd al-Rahman, but they still found themselves too weak to maintain themselves as an effective fighting unit against the Turkoman. Their continually unsettled state, and their inability to surmount the various political, social and ecological pressures during the nineteenth century, led them to find security outside their own tribal structure. Through marriage and voluntary association they thus joined up with the stronger units of the Timuri tribe in Khurasan, into which they have today been largely incorporated. This is a continuation of what Ferdinand calls " a consistent tribal feudal system, a sideline to the tribal genealogical system ",80 i.e. a complex form of re-organization under a strong leadership, which in this instance has come from outside the Jamshidi themselves.81 Succession of Jamshidi Khans in the second half of the nineteenth century Mahmfid Khin
I

Qara Khan

Jabbar Khan

Zaman Khan - - - - - - - -

Mir Ahmad Khan

31

'Abdullah Khan

14 Mahdi Quli Khan IHaidarQul Khan

A
Allah Ydr Khan

Aminullah Khan Yalangtash Khan

5i

17

Dhulfiqar Khan

BI

8
Muhammad 'Azim Khan Rita Quli Khan I-8--Paramount Khan in Afghanistan A-B-Paramount Khan in Khurasan (Persia) Mahmfid Khan

STAIRCASE MINARETS ON THE PERSIAN GULF By David Whitehouse

In 1938 Schacht drew attention to the existence in certain mosques in Egypt and parts of central and western Anatolia of a distinctive architectural feature: the staircase minaret.82 Instead of a tower, the minaret comprises a small kiosk on the roof of the mosque, approached from the ground by an external flight of steps. In 1954 the same writer discussed the presence of the staircase minaret in two other regions, Tunisia and northern Nigeria.83 In Tunisia he noted minarets at Djerba and Sousse which consist of massive bastions ascended by an external stair, while in Nigeria he reported that
Dr. Brian Street, Mr. Alexander Morton and Mr. Iraj 79Duringhis absencein Russia,his youngerbrotherMuhammad for 'Azim Khan took over the leadershipof the tribe. Haqiqat-KhAn kindly reading the draft of this article and 80Op. cit., p. 183. offeringconstructivecomments. 82J. Schacht, " Ein archaischerMinaret-typ in Agypten und 81 This note was written while the writer was a Fellow of the V Anatolien", ArsIslamica (1938), pp. 52-54. British Institute of Persian Studies and was working on a Social ScienceResearchCouncilProjectunder the guidanceof 83J. Schacht, "Sur la diffusion des formes d'architecture ProfessorSir Edward Evans-Pritchard.To the organizations religieuse musulmane A travers le Sahara", Travaux de de II GavinHambly, I'Institut Richerches and guide are offeredthanks; also to Professor Sahariennes (1954), pp. 11-27.

156

JOURNAL

OF

PERSIAN

STUDIES

staircase minarets are universal in the mosques of the Ffilani, who probably acquired the type from the Maghrib. More recently still, Schacht returned to the subject and described an even wider distribution, including Omdurman in Sudan and the coast of East Africa, at Kaole, Shamiani, Ishikani and Gedi.84 Finally, further examples have been recognised in North Africa, where it seems likely that the staircase minaret was a standard feature of early F timid mosques,85 and on the Persian Gulf. My purpose here is to describe the minarets of the Gulf coast and to suggest a possible link with staircase minarets in East Africa. My starting point is Bandar-i Taheri, the village which occupies the site of medieval Sirdf, 240 km. south-east of Bushire. Taheri is among the larger villages of the coast, with a population of nearly two thousand, evenly divided between the Sunnite and Shi'ite sects. It has two mosques: a Sunnite mosque " built, according to local informants, more than a hundred years ago " and a Shi'ite mosque completed in 1377/1957-58 (P1. IVa and b). The former building (Fig. 6, lower) is typical of the mosques built on this part of the coast today. It is a rectangular structure, 7-3 m. deep and II -4 m. wide, with a courtyard of similar dimensions in front of the fa?ade and an enclosed ablution area at one side. The mosque rests on a low platform and is built of mortared rubble, with a flat roof of poles, matting and mud. The outer walls are pierced at regular intervals by openings containing doors or, as in the qibla wall, windows with bars and shutters. The interior is divided along the major axis by an arcade which supports the roof. The mihrdb (demolished since we planned the mosque in 1966) was contained in a rectangular salient and, as usual in this region, had a small window facing Mecca.86 The minbarwas a small structure of plaster and stone, two steps high and o-6 m. square. The minaret is a flight of eighteen steps, m. 5"5 long, leading to the north angle of the mosque; there is no kiosk and the muezzin simply stands on the roof. In the nearby Shi'ite mosque, the staircase leads to a small kiosk and a more elaborate kiosk surmounts the Sunnite mosque at Akhtdr, 8 km. north of Tdheri, which was built only a few years ago (P1. IVc). In this area, the earliest datable mosque with a staircase minaret was found at Siraf in 1966 and It stood in the ruins of the Great Mosque and was dated to the excavated completely in I967-68.8 fifteenth century by pottery with underglaze ornament (including SirdfI, P1. VId) and unglazed painted ware (comparable with SirafII, P1. VIf). The mosque measured 6 -4 m. by II - 8 m. internally and had a porch in the centre of the faqade (Fig. 6, upper). As in the Sunnite mosque at TAheri, the interior was two bays deep and three bays across, divided by an arcade which supported the roof. The mihrdb occupied a rectangular salient, with a plastered stone minbar immediately to the right. The mosque itself was built of mortared rubble, with a levelling course o044 m. above the floor. Elsewhere at Siraf, similar levelling courses were used in the square building at Site E, also of the fifteenth century.88 The walls were 6 m. thick and rested on the floor of the Great Mosque. The similarity between the fifteenth o. century mosque and those of the present day extends to the minaret, which in the former consisted of a solid L-shaped structure abutting on to the east angle of the building. It survived to a maximum height of I 4 m. and the first three steps remained. Although the ancestry of the Taheri type of mosque may be traced back to the ninth century89, the fifteenth century structure is the earliest staircase minaret so far discovered on the Iranian coast of the Persian Gulf.90 In East Africa, too, the earliest examples belong to the fifteenth century, or slightly earlier. Garlake records the occurrence of staircase minarets in the north and south courts of the Great Mosque at Kilwa, rebuilt by Abil' 1-Mawdhib or his descendents,91 in the fifteenth century ablution court of the Malindi mosque, also at Kilwa, and in the medieval mosque at Kaole, where the minaret consisted of a

" J. Schacht, Further notes on the Staircase Minaret ", Ars OrientalisIV (x961), pp. 137-41. " 85 Hugh Blake, Antony Hutt and David Whitehouse, Ajdabiyah and the earliest FAtimid architecture ", Libya Antiqua (in the press). 86 A similar window occurs in the post-medieval imdmzddehat Sirdf, described on p. 82 of this volume. 7 Sfrdf II, pp. 46-48.
81

88 Strdf II, p. 5489 See, for example, the mosque in the bazaar at Sirdf: Strdf IV, fig. 5 and pl. IVb; SfrdfV, fig. 13 and pl. IXb, in this volume. 9o Note, however, the existence of a short flight of steps attached to the minaret of the Great Mosque at Sirdf in its latest phase: Sirdf II, pl. Ic. 91 Neville Chittick, " Kilwa: a Preliminary Report ", Azania I (1966), pp. 1-36.

II II
II

I
LJ

I-- .
L _-J

Ii

?opening

-i
7 7 If

cc

minaret

porch porc

r-

--Ir

I-I

rI,
I.J1

graves

soak-aaway

approx N

ablution area

opening

10

I 1
BP GS RPW & DW

I
metres

minarets: upper, at at Fig. 6. Mosqueswith staircase fifteenthcentury Sirdf; lower,nineteenth century Tdheri. (Scale i : 2oo).
14A

158

JOURNAL

OF

PERSIAN

STUDIES

flight of eight steps abutting on to the fagade.92 According to Garlake, tower minarets were unknown in the region until the nineteenth century.93 Regular contact between the Persian Gulf and East Africa already existed in the tenth century, when Mas'fidi recorded the presence of Sirdfi merchants at Qanbalu and at this date merchandise from the Gulf was reaching such sites as Kilwa,94 Unguja Ukuu95 and Manda in the Lamu archipelago.96 Until recently it was believed that the so-called " Shirdzi " migration to East Africa also occurred in the tenth century. The traditional account, based on the Kilwa Chronicle, holds that c.957 one 'Ali b. al-IHusain (or Hasan), of the ruling family of Shirdz, migrated to Kilwa, where he founded a local dynasty. Recently, however, Chittick has pointed to inconsistencies between the Arabic and Portuguese versions of the Chronicleand argued that the migration took place in the second half of the twelfth century, if not later.97 In the late fourteenth or fifteenth century, pottery with underglaze ornament was exported from Persia, and at Kilwa the House of the Mosque was embellished with about three hundred Persian bowls, set in the vaulted ceilings.98 I suggest that the staircase minaret was introduced to East Africa from the Persian Gulf, or perhaps to both areas from a common source elsewhere in the Indian Ocean, as a result of this persistent traffic.

A RECENTLY IDENTIFIED

FRAGMENT OF THE CYRUS CYLINDER By C. B. F. Walker

A small fragment of a Babylonian cylinder inscription in the Babylonian Collection of Yale University has recently been identified by Dr. P.-R. Berger of the University of Munster as a part of the British Museum's Cyrus Cylinder (BM 90920). The fragment, numbered NBC 2504, was first published

plaster cast of the fragment has confirmed that it joins the broken end of the text of the Cyrus Cylinder supplying parts of lines 36-45. It is expected that arrangements will be made for the fragment to be rejoined to the cylinder in the near future. Dr. Berger is at present preparing a new edition of the text of the Cyrus Cylinder as part of a larger work on royal inscriptions of the Neo-Babylonian period; but in view of the interest in the Cyrus Cylinder aroused by the recent celebrations of Cyrus's 2500th anniversary the Department of Western Asiatic Antiquities has felt it appropriate to make this preliminary announcement of his discovery. A brief allusion to the new fragment was made by Dr. Berger in Ugarit ForschungenII (i970), p. 337The cylinder was found at Babylon in the course of Hormuzd Rassam's excavations on behalf of the British Museum. The excavations, begun in February 1879, were authorised by a firman from the Sultan which permitted Rassam " to pack and dispatch to England any antiquities he found " (H. Rassam, Asshurand the land of Nimrod, p. 259). It appears from his correspondence that the cylinder was found in March 1879. While Rawlinson states in his first publication of the cylinder (J.R.A.S. I880, p. 83) that it was understood to come from the excavations at Birs Nimrud (Borsippa) and Rassam (op. cit., p. 267) says that it was discovered in the ruins ofJimjima (a part of Babylon), in a letter from Rassam to Birch, Keeper of Oriental Antiquities in the British Museum, dated 2oth November 1879, he states, " The Cylinder of Cyrus was found at Omran with about six hundred pieces of inscribed terracottas before I left Bagdad." This is as one would expect, as Omran is the site of the Temple of Marduk at Babylon. The Yale fragment was presumably removed from Rassam's excavations, or found on his dumps, and later purchased by Nies in Baghdad or Europe.
92

in Inscriptions theCollection by J. B. Nies and C. E. Keiser in Babylonian ofJ. B. .Nies,Vol. II, no. 32. A

Peter S. Garlake, The Early IslamicArchitecture theEast African of Coast (Oxford I966), p. 84. 93 Note, however, an inscription with a date equivalent to 1269, recording the construction of a minaret in the Great Mosque of Mogadishu: Neville Chittick, " The 'Shirazi' Colonization of East Africa ", Journal of African History, VI 3 (1965), PP- 275294, quoting E. Cerulli in Somalia I (1957), pp. 2-1o.

94 Chittick (1966), pp. 5-1o. 95 Neville Chittick, " Unguja Ukuu; the Earliest Imported Pottery and an Abbasid Dinar ", Azania I (1966), pp. 161-63. 96 Neville Chittick, " Discoveries in the Lamu Archipelago ", Azania II (1967), pp. 1-31. 97 Chittick (1965). 98 Chittick (1966), pp. 23-24-

builtin 1957-58. in minaret theShi'ite mosque, Pl. IVa. Taher~.Staircase

built in thenineteenth in minaret theSunnitemosque, century. Pl. IVb. Tdheri. Staircase

in Pl. IVc. Akhtdr. Staircase minaret t

SHORTER

NOTICES

159

The text of the new fragment refers to additional offerings (presumably in the temple of Marduk) instituted by Cyrus and to his restoration of the fortifications of Babylon. Among other works (the passage is still incomplete) he lists the inner city-wall, named Imgur-Enlil, the brickworkof the bank of the moat, perhaps also the outer city-wall, Nimit-Enlil, and gates with decorative ornament, threshold and pivots in bronze and copper. Most interestingly he adds, " In it (i.e. in the gateway?) I saw inscribed the name of my predecessor King Ashurbanipal." This must be an allusion to the discovery of an earlier building inscription, almost certainly the cylinder L6 (translated in D. D. Luckenbill, A.R.A.B. II ??963-4) of Ashurbanipal, King of Assyria 668-627 B.C.,which commemorates his restoration of Imgur-Enlil, Nimit-Enlil and the gates of Nimit-Enlil. It is interesting to see Cyrus thus continuing the antiquarian interests of his enemy and predecessor Nabonidus, if the allusion is not motivated simply by a desire to be seen to respect local traditions. The reference to Ashurbanipal is of particular significance as I understand that ProfessorHarmatta is to publish in a forthcoming volume of a Acta Orientalia study of the Cyrus Cylinder in which he demonstrates that in literary form its closest are the inscriptions of Ashurbanipal rather than Neo-Babylonian royal inscriptions. parallels The new addition serves to emphasize the essential character of the Cyrus Cylinder as not a general declaration of human rights or religious toleration but simply a building inscription, in the Babylonian and Assyrian tradition, commemorating Cyrus's restoration of the city of Babylon and the worship of Marduk previously neglected by Nabonidus.

UN OBJET CULTUEL (?) D'ORIGINE IRANIENNE PROVENANT DE NIPPUR By P. de Miroschedji

' Dans Iran IX (1971), p. 155, B. Brentjes a publi6 un objet en stdatite en forme de "sac main" originaire de Sokh, dans le Ferghana (Uzbekistan). Nous avons pu 6tudier r6cemment au Musde du Louvre99le moulage d'un objet de la meme catigorie dont l'original, conserv6 au Musde d'Istanbul, ' proviendrait des fouilles amdricainesde la fin du siecle dernier Nippur.100 L'objet mutil6 mesure aujourd'hui 26 cms de longueur et 13,3 cms de hauteur; son dpaisseurest de 4,2 cms. L'anse a disparu, mais on peut sans doute lui restituer une forme semi-circulaire. Le d6cor, tres mal conserve, est grave plut6t que sculpt6, ce qui donne aux figures un tres faible relief. Le recto (Fig. 7a et P1. Va) reprdsentetrois palmiers stylists dont les branches supdrieuresportent des r6gimes de dattes, tandis que les branches infirieures retombent mollement vers le sol; quant au verso (Fig. 7b et P1. Vb), il figure simplement deux torsades. Cet objet est le quatrieme d'une categorie dont malheureusement aucun exemplaire n'a &t trouv6 au cours de fouilles rdgulieresou scientifiques: deux d'entre eux furent publids par Mme Y. Godard, et avec pour provenances d6clardesrespectivement l'Azerbaidjan et la region de Palmyre en Syrie,o10 d'Asie Centrale, est celui que B. Brentjes a publi6. Notre exemplaire est une le troisieme, originaire rdpliquepresque parfaite de celui de Palmyre, qui prdsenteaussi trois palmiers sur le recto et une torsade (au lieu de deux) sur le verso. D'autres objets de formes comparables, mais gindralement en albatre, ' relkvent encore de cette cat6gorie. Un fragment sans d6cor trouv6 Sialk par R. Ghirshman remonte probablement au debut du IIIme mill6naire av. J.-C.102 Un exemplaire du British Museum, dont la forme est celle d'une boite surmontded'une anse imitant la vannerie, et dont le d6cor, compos6 d'yeux et de rosaces, est dispos6 sur une seule face, date peut- tre du milieu du IIIme mill6naire av. J.-C. ;103
99Nous remercions tres vivement M. Pierre Amiet, Con- 101(Philadelphia, 1917). III (1938), pp. 336-7 et 3o10-I I; voir aussi F. A. Athdr-6-Iran servateurdu D6partementdes Antiquit6sOrientales,qui nous a aimablementcommuniqu6ce moulage pour 6tude et qui Durrani, " Stone Vases as evidence of connection between PakistanI (1964), 6 nous a autoris Ale publier. Mesopotamia and the Indus Valley", Ancient 100 H. V. Hilprecht, Die Ausgrabungen Universitdt von der Cf. pp. 88ss et pls X-XI. 102Fouilles de Sialk I (Paris, 1938), p. 55 et pl. LXXXV: S.223. im zu Pennsylvania Bel Tempel Nippur(Leipzig, 1903) et J.-P. on and Peters, Nippuror Excavations Adventures the Euphrates 103BMQ IX, pl. VII: 43 = F. A. Durrani, op. cit., p. 89.

160

JOURNAL

OF

PERSIAN

STUDIES

de Fig. 7. Un objeten ste'atite d'origineIranienne, provenant Nippur.

SHORTER

NOTICES

161

(dans le cas de l'exemplaire de Sialk) ou des imitations (dans le cas de ceux d'Ur et de Suse) des equivalents en st6atite. La distribution gdographique de ces derniers est assez d6concertante et n'aide guere a pr6ciser leur sur origine. Toutefois, le d6cor de torsades s'observe sur divers fragments de vases en st6atite,105 des ainsi que sur la plaque jarres de Suse D106et sur des vases mdtalliques a long bec "du Luristan",107 perforCeen bitume inscrite au nom du pretre Dudu et trouv6e a Tello;l08 tous ces objets, dat6s du milieu du IIIme mill6naire av. J.-C. (Early Dynastic IIIA-B), sont apparemment de fabrication l1amite. Quant aux trois palmiers stylis6s, ils possedent des paralleles rigoureux sur l'exemplaire de Palmyre, mais aussi sur trois vases en st6atite, dont deux au moins furent import6s en M6sopotamie.109 Ce genre de repr6sentationsdu palmier figure encore sur un cylindre-sceau de Tepe Yahya IVB110et est Son sur plusieurs cylindres-sceaux susiens de l'6poque des sukkalmahhu.11l origine l1amite d'autant plus vraisemblable que la stylisation du palmier est diff6rente dans le domaine m6sopotamien aux 6poques contemporaines et post6rieures.112 C'est donc probablement dans le sud et le sud-ouest de l'Iran, et plus particulierement dans les regions soumises t l'influence de la civilisation 61amite, qu'il faut chercher I'origine de cette cat6gorie ' d'objets anse, comme d'ailleurs celle des vases en st6atite sculpt6e sur lesquels apparaissent les memes themes dacoratifs. Cependant, les objets a anse pr6sententcertaines particularit6sstylistiquescommunes telles qu'un d6cor en tres faible relief obtenu principalement par gravure, et une repr6sentation des surfaces rugueuses a' l'aide de courtes entailles en forme de virgule;x13 ces caract6ristiquessuggerent qu'ils 6taient fabriqu6spar un petit nombre d'ateliers, dans un but, cultuel ou autre, qu'il est difficile de priciser.

" des objets identiques, mais sans d6cor, ont &t6 trouv6s Suse dans une couche que R. de Mecquenem attribuait "au xxximime sicle".o04 On pourrait consid6rerces objets en albAtrecomme des prototypes

AN INSCRIBED BRONZE VESSEL FROM LURISTAN

The Vessel,by P. R. S. Moorey Our attention was recently called to a sheet bronze vessel in the possession of Charles Ede, Ltd., of London. We are most grateful to the owner for allowing us to publish this vessel here and for providing a photograph (P1.VIa). It had been much corroded when received, but in the course of chemical cleaning a short inscription in cuneiform script was revealed on the neck. The circumstances of discovery leave no doubt that the inscription was engraved in antiquity and is not a recent addition intended to increase the object's value. This is confirmed by close examination of the signs, which show no trace of fresh cutting with a steel tool.
MDP XXV (1934), p. 182, fig. 6: 2; le Mus6e du Louvre 109 Un vase de Khafadje conserv6 au British Museum: cf. F. A. conserve plusieurs objets identiques, tous en albaitre. Le Durrani, op. cit., pl. V; un vase du Mus6e du Louvre donn6 " XXXIIIeme siecle " de R. de Mecquenem semble correspar le vendeur comme provenant d'Uruk: cf. G. Contenau, Manuel d'Archdologie OrientaleII (Paris, 1932), p. 644, fig. 448; pondre A " Suse C" dans la classification de L. Le Breton et un vase d'origine inconnue du Metropolitan Museum: (cf. Iraq XIX (1957), PP. 94ss), c'est-A-dire environ aux niveaux 18 16 des fouilles entreprises actuellement par J. cf. E. Herzfeld, Iran in the Ancient East (London, 1941), pl. Perrot sur l'Acropole de Suse. XXIV. 105 Cf. A. Parrot, Sumer (London, 1960), p. 140, fig. i68: A; 110 C. C. Lamberg-Karlovsky, Excavations at Tepe Tahya, Iran, Mission Arche'ologique Mari I, Le Templed'Ishtar (Paris, 1954), de 1967-i969 (Asia Institute Monograph I, Shiraz, 1970), pl. 21. pls XLVI:268 et XLVIII (ohi la torsade se compose de 111 MDP XII (x912), fig. 132 et XXV (1934), fig. 82: 4; RA XLIV (1950), pl. V : 44l'enroulement des corps de deux serpents). 106 MDP XIII (1912), pl. XXVII: 112 Cf. C. C. Lamberg-Karlovsky, op. cit., p. 66, n. 6. I I. Pour des 107 P. Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah exemples illustr6s sur des cylindres-sceaux m6sopotamiens du IIIeme mill6naire, voir P. Amiet, La glyptique misopotamienne (Berlin, 1969), p. 15, fig. 12 (gruppe 5). 108 A. Parrot, Sumer, p. 138, fig. 167: A. L'inscription pr6cise archaique(Paris, i961), nos 387, 1139, 1163, 1346 et 1482. que Dudu fit venir cette plaque d'URU.Akt, un district 11 Cf. le dessin des r6gimes de dattes sur notre exemplaire et sur oriental situ6 peut-&tre aux confins de l'Elam: cf. C. J. Gadd, celui de Palmyre, et le corps de l'aigle sur celui d'Azerbaidjan; " The Cities of Babylonia ", CAH I, chap. XIII, p. 41. ce proc6d6 s'observe tres rarement sur les vases en st6atite: Notons que des torsades, mais de type diff6rent, ornent aussi voir par exemple le corps des serpents sur le vase du British Museum (supra,n. Io9) et sur un fragment de Tepe Yahya IVB plusieurs objets susiens en bitume: cf. P. Amiet, Elam (Auverssur-Oise, 1966), pp. 166 et 170-71. (Iran IX (1971), p. 93, fig. 3: U).
104

Pl. Va. Un objeten ste'atite de d'origineIranienne, provenant Nippur: recto.

PI. Vb. Verso.

162

JOURNAL

OF PERSIAN

STUDIES

The vessel is a simple, hammered sheet-bronze jar with a slightly everted rim, tall neck, and broad, inclined shoulder. There is a constriction running round the body about two-thirds of the way down cms. high and has a maximum diameter the side and the vessel has a slightly raised base. It stands 8.-2 cms. at the lip). Such vessels have long been associated with the cemeteries of of 10 - i cms. 7 (7" Luristan clandestinely excavated in the past fifty years. Examples have recently been found in controlled excavations at War Kabud in western Luristan by the Belgian Luristdn Expedition. The cemetery in which they were found has been dated by the excavator to the later eighth or earlier seventh century B.C.114 The form had a wide distribution about this time since comparable vessels have been excavated at Susa in Khuzistan15 and at Uruk in Iraq.116eThis is the first example with an inscription to be fully published, though there are others about (see note 114) with inscriptions of uncertain authenticity. Vessels of this simplicity are likely to have been produced by any number of workshops satisfying local demands for standard items of household equipment, for such they seem to be. They are a metal variant of an ancient ceramic form and there is no reason to suppose them exclusively funerary in function.

The Inscription (Pl. VIb), by W. G. Lambert

The inscription is clearly cuneiform, and the most conspicuous feature of it is that the signs are reversed on the bronze, so that it must be read from right to left. A few of the horizontal and diagonal wedges towards the end of the row of signs have a slightly thicker left-hand end, but where there is a bold clear head, it is always at the right-hand end. Except for the first group of six wedges, which are more widely spaced than the rest, they run together and are not clearly divided into signs. The one " clearly separated sign, the first, is the Sumerian KISIB seal ", probably, as will be seen, used here as an for the Babylonian kunukki" seal of". Despite the running together of the remaining wedges, ideogram they do in fact yield the Babylonian personal name (with determinative) mla-ba-li. The name is adequately rendered except that two small inner wedges of the LA are missing, but that is no objection when the craftsman responsible clearly did not read and understand cuneiform. The sign-forms could be either Assyrian or Babylonian, except for the BA, which is Babylonian. Thus the whole line reads:

kunukki(KI?IB)
Seal of Ldbasi.

mla-ba-Ji

The personal name Ldbasi, originally ld-abdf, " may I not be ashamed ", seems to occur first in a Middle Assyrian document from the time of Ninurta-tukulti-Ashur (written mla-ba-dJ, see C. Saporetti Onomastica I Medio-Assira (Rome 1970), p. 297), 1133 B.C. The present writer has not noted any further occurrences before two in the Late Assyrian Harper letters of the 8th and 7th centuries B.C.: documents from the end of 222:I and 472 rev. 5, written both times mla-ba-.Ji. In the Late Babylonian the 8th century to Seleucid times it is a common name written variously mla-ba-.i, mla-ba-19, mla-a-ba-Ji, and mla-ba-a-li. The rare Assyrian occurrences prove little, and lack of evidence prevents one from learning how much before c. 70oo the name became common in Babylonia. It seems not to occur in Cassite-period and other documents which extend downwards to about I o100 B.C. Thus the inscription identifies the vessel as " the seal of Mr LdbAi ", which patently it is not. Not only the rendering of the signs but also the content confirms the illiteracy of the engraver. That the signs are reversed shows that they were copied from a cylinder seal, since these were the only objects on which cuneiform inscriptions were put in reverse, so that they would appear the right way round in an impression. Seal inscriptions of the type " seal of (name of owner) " come into fashion only with the Cassite and Middle Assyrian styles, though they remain common until the end of cylinder seals in Mesopotamian styles in the Late Assyrian and Late Babylonian times.
114L. Vanden Berghe, Opgravingen Pusht-i Kuh, I: Kalwali en of the new inscription this might merit re-investigation. in de War Kabud (r965-I966) (Brussels, 1968), pl. 34a. Compare x'15 Mecquenem, MDP XXIX (1943), pp. 49-50, fig. 42: 6, 8. P. Calmeyer, Datierbare Bronzen aus Luristan und Kirmanshah (Berlin, 1969), Group 53; fig. I 9g on p. xi6 is described as 116 E. Strommenger, Gefdsse aus Uruk (Berlin, 1967), pls. 48: 4, 49: 5. having a "gefdlschteInschriftauf dem Gefdssbauch". In the light

SHORTER

NOTICES

163

All the evidence relating to the inscription fits the archaeological dating of the vessel to the first quarter of the first millennium B.C., and the inescapable conclusion about the inscription is that it was copied from a Babylonian seal by a craftsman who had no idea of its meaning whatsoever. Since the vessel belongs to an Iranian type it is most likely that the inscription was engraved in Iran, where use of cuneiform was very restricted at this time. This confirms a conclusion of the present writer about certain other inscriptions on Luristan and other Persian bronzes (see AfO XXII 9 iff.), that when of Babylonian origin, they were copied in Iran from stray Babylonian objects, and do not prove the Babylonian origin of the object or the engraver. Why an illiterate craftsman should want to adorn his products with writing is a question which cannot in this case be answered with certainty, but parallels can be quoted. Some Sasanian seals have bungled inscriptions, and, more recently, some Oriental carpets have woven into their patterns phrases from the Qur'dn which cannot be deciphered in their garbled state.

cm. Luristan,Iron Age II. (Photograph courtesy Pl. VIa. Sheetmetal with a cuneiform on height by jar inscription the shoulder, 8.2 Ede, Ltd., of London.) of Charles

inscription. Pl. VIb. The cuneiform

SURVEY OF EXCAVATIONS

IN IRAN DURING

1970-71

Borazjan A summary of work at this site has been published in Bastan Chenassiva Honar-eIran, Revued'Archiologie et d'Art Iranien VII. Ganj Darek Tepe

Archaeological Service of Iran, and was financed by the Canada Council. Mr. M. H. Khushabi served as Representative of the Archaeological Service. Our previous excavations (see IranV, VI, VIII) had shown that apart from a few recent burials the mound contained only prehistoric materials representing the earlier phases of the Neolithic. At least five main occupation and/or building levels (A, B, C, D and E) are represented in a vertical thickness

ended on 19 August, 1971. The work was a collaboration of the Universite de Montreal and the

The third full season of excavation at Ganj Dareh, Kermanshah District, began on 29 May and

S/A
,..,CA" "* HakFlovn -'*" .. " *TABRIZ
SHasa ,.""T."

U.S.S.R.

CASPIAN SEA S A

do

0..0. "*

0'*.
MESHED

Takhb-i- 5ulaiman (3 TEHRAN eHAMADAN Gcdin Tepe-.

Tepe ?*Tureng
Shahr-i Qumis

,* to .,.

"

i%I

0
-- --.

Tepe Ganj DorehIP

*lSFAHAN
Shahr-i Sokhbe SKERMAN
0,

(.

Qabr Sheykheyn Su'v** sjid-i-Sulaiman HaFI TepeN Cl-i

0'

t.

.,'
0 -1

Maliyan.o

Persepolis

ZAHEOAN

-o0

?
*

SHIRAZ*;

ruzjam

" "
STraf SAUDI

?*TepeYakya
Tepe Doshl-i Del

ARfABIAo
500 lo kilometres r

165

166

JOURNAL

OF PERSIAN

STUDIES

of nearly eight metres. Among the interesting features of the site are evidence of early and simple pottery, of early goat domestication, and of a complex level (D) of structures made in part of planoconvex bricks apparently representing a small village destroyed by fire. The principal aims of the 1971 season were (I) to expose a larger horizontal area of this burned village; (2) to clarify certain stratigraphic problems posed by the succeeding levels C, B and A; (3) to uncover as much as possible of Level E which had been exposed earlier, in 1965 and 1967, only in sondages of a few square metres; and (4) to find out, by trenching from the central area to the flanks of the mound, the extent of each occupation. We were also anxious to recover more palaeobotanical materials than had been found in previous seasons. With the help of seven student assistants and up to forty local workmen the area excavated during the 1971 season was nearly doubled for a total (exluding the soundings on the edges of the site) of over 150 square metres. The " central excavation area " which extends southward from the centre and summit of the mound (Pl. Ia) was enlarged by excavating along the east, south and west sides, from the present surface to Level D. In the northwest corner the excavations of 1969 were also continued into Level D. Two lateral trenches were excavated east and west of the central excavation area. These operations exposed a further 75 square metres of Level D so that a total of I44 square metres of this interesting level are now revealed. With this exposure it is now possible to have a better idea of the layout of structures, the construction sequence, the range of techniques used and perhaps something of the functions of certain of the structures. Thus there is some evidence that at least in the centre of the site some of the structures may originally have been two-storied buildings, with a living surface supported by wooden beams overlying in some cases the small alcoves or cubicles that had so puzzled us in previous seasons. Very few of the occupation floors had survived the fire since the contents tended to drop into the " basement " when the beams collapsed. There is also evidence for the roof arrangements, through the many impressions of beams and canes on the baked clay of the rubble filling the structures. The functions of the subfloor cubicles are not yet clear, though the presence in some of small and large clay pots suggest some kind of storage activity. We do not yet know what might have been stored. Among the interesting finds this season was a number of very small compartments built inside one of the cubicles, made of thin vertical slabs of clay with bevelled edges that had apparently been prefabricated and sunbaked before being placed in position and plastered. However, perhaps the single most revealing discovery of the season in Level D was a small cubicle, situated almost precisely under the present summit of the mound, which contained two skulls of what are probably wild sheep attached one above the other to the plastered interior of a small niche built against one wall of the cubicle (Pl. Ib). The skulls are complete except for the mandibles and the horn cores are intact. This kind of arrangement seems unique for the the Zagros Neolithic up to now and is reminiscent of (though considerably older than) some of the " shrines " with skulls of wild cattle at Qatal Huiyiik in Anatolia. Another interesting feature of this level, which probably would not have survived in the absence of an intense burning, are several large boulder mortars around whose upper edges clay rims had been carefully built. These baked rims, if found out of their context, can easily be mistaken for the rims of large, lightly baked bowls. The bases of the mortars were at times set in mud or clay foundations packed with pebbles, presumably for stability. The area now exposed of Level D shows a series of mainly rectilinear structures built of long planoconvex brick, of strip-chineh and sometimes of ordinary chineh. A few walls were simply constructed of old rubble plastered on both faces. In a few cases we found that flat stones were also incorporated into the walls. The structures are tightly clustered with no signs of streets, courtyards or large open spaces
some walls have circular or oval " portholes ", (Pl. Ic). There are no clear indications of doorways; often sealed with clay discs or cones. Mud plaster was used for both floors and walls. The structures were in some cases repaired or rebuilt following collapses or weakening, and sometimes modified in It is possible that at least two fires were responsible for the destruction of this level. rebuilding. A trench, 6 x 2 m., was excavated from the central area at the southeast corner, eastward nearly to the present edge of the mound. It showed that Level D probably extends (or extended originally) under cut through a shallow depression which had been the present field surface. This trench, incidentally, and noted earlier in this part of the site; the depression yielded sherds which seem to be Achaemenid the first time such materials have been found here. Islamic,

SURVEY

OF

EXCAVATIONS

167

Another trench, 8 X 2 m., was excavated westward from the northwestern corner of the central area to the " West Cut " created by clandestine digging on the flank of the mound some years ago. This was an attempt to follow the various levels distinguished in the centre of the site to the outer edge, and thus gain an idea of the area of each occupation. Unfortunately this was less successful than expected due to the very hard deposits and.the unclear stratigraphy on this side of the mound. Just why the two sides of the site differ so much in compactness of soil and preservation of architectural detail is still uncertain. It had been hoped that most of Level D in the central excavation area could be cleared and removed during 1971 so that a large section of the underlying Level E might be exposed. Unfortunately less was accomplished than planned because of the complex nature of Level D, and this project must await another season. However, some twelve metres of Level E were uncovered in the northwest corner and in the west trench. This level seems to consist of a series of lenses and deposits about one metre thick, with no solid architecture, grinding stones, human figurines or pottery but with much the same types of flint tools and animal figurines as in the succeeding levels. The only structures so far noted are circular or oval shallow depressions dug into the virgin soil and containing in most cases ashes and burned stones. One is partially surrounded by an arc of small stone slabs placed on edge (Pl. Id). Possibly they represent fire-pits like those reported at Mureybit in Syria at about the same time, though whether they were used for heating or for roasting or parching food is uncertain. The present sampling of Level E suggests a series of encampments by groups using impermanent shelters and possessing only some of the traits that characterize their successors at this site. The excavations in Levels A, B and C resulted in larger and better samples of materials but few unexpected new details. It now appears that Level C is restricted, at least insofar as it is preserved, to the northwest corner of the site, and Level B, although covering a larger area, is essentially a continuation of it. Both are composed of quadrangular structures of chinehand mud brick, with several occupation floors in each case. The final occupation of Ganj Dareh is Level A. Although much disturbed by rodents, roots, leaching and recent burials, it has revealed two interesting features: a domed oven or kiln with several layers of powdery grey substance (probably burnt limestone); and an empty cist or box made of stone slabs resting on their edges. It now seems possible to subdivide Level A in spite of its poor state of preservation. Alarge number of prehistoric human skeletal material representing adults, adolescents and infants was found in I971, bringing the total to twenty-six individuals at least. Most came from Level D, including several infant burials in subfloor cubicles. Both flexed and extended positions are represented, and some burials may be secondary. In one case three extended skeletons (an adult, an adolescent and a " sarcophagus " made of mud bricks and covered child) were found together inside a curious elongated with a kind of mud roof. The bodies rested on a thick layer of greyish powdery material, perhaps burned limestone, and one had a finely polished stone pendant in its rib cage, the only evidence of grave goods so far found at Ganj Dareh. In cleaning a section of the West Cut we discovered a human mandible resting on a former surface facing an intact pair of hands which had apparently been severed from the arms and were lying palms upward nearby. Altogether, while the state of preservation of the human materials is variable (only one reasonably complete skull could be reconstructed), the remains should throw some much needed light on the physical types present and the burial customs at this period. Among the clay objects were several human figurines, the usual quantities of cones, discs, spheres and animal figurines (including a bearded goat) and some types not hitherto found here such as small seal-like pieces. There are also several new types of worked bone with double perforations, and a bone fragment with grooves that may have held flint blades. Flint artifacts are abundant in all levels and show little change from top to bottom of the site. Blades with edge polish (" sickle sheen ") are fairly common except perhaps in Level E. No obsidian has yet been found. Fragments of polished stone bowls occur in small quantities. Some fragments and impressions of wood were found this season, including what looks like interwoven withies (for wattle-and-daub walls?), and a notched cane fragment may represent the butt of an arrowshaft. As reported earlier, preliminary study of the faunal remains by Dr. Dexter Perkins of the Columbia University Faunal Research Centre suggests that goats were being domesticated here-a conclusion

168

JOURNAL

OF PERSIAN

STUDIES

that fits well with recent studies of the Asiab fauna. Large new samples of faunal materials were collected in 1971, including sheep, cattle, deer and pig. More animal hoofprints were observed in the plano-convex bricksof Level D. The situation for floral remains is less conclusive in spite of the flotation practices followed, although there is much indirect evidence for plant use in the form of mortars, pestles, clay bins and containers. Several more of the very large clay pots were found built against the walls and floors of cubicles in Level D, while in Level B were some smaller but very simple clay vases. It is now consideredpossible that some of the pottery at Ganj Dareh (probably the earliest yet reported in Western Asia) may have been fired at a low temperature at the time of fabrication, and that their preservationhere was not entirely due to accidental hardening in later fires. In addition we recovered in Level D several miniature clay vessels about five centimetres high which may represent toys or models of larger vessels. The three seasonsof excavation at Ganj Dareh reveal a detailed sequence of early village life in Levels D to A when presumably the bases for food-production and sedentary communities were being established in this region. The available radiocarbon determinationsfor these levels suggest that they belong to the late 8th and perhaps early 7th millennia B.C. The age of the basal deposits (Level E) is still uncertain, and so is its real nature. One further season of excavation is planned to complete the investigation of Levels D and E.
PHILIP

E. L.

SMITH

Ghubayrd The first season of excavations at the early and medieval Islamic site of Ghubayra were conducted between late March and late May 1971. The excavations were supported by the School of Oriental and African Studies, the Percival David Foundation of Chinese Art, the British Academy from the Stein-Arnold Fund, the Royal Ontario Museum, Toronto, the Corpus Inscriptionum Iranicarum and by private sources. Ghubayrd lies some 8o km. south of Kirmdn in the eastern part of the Bardsirvalley at the junction of the Chiri and Ghubayri rivers. The important remains consist, from north to south, of two prehistoric mounds, a stone tower called locally Xsiydb,the ruins of numerous mud-brick buildings, the ruins of a large building known as Keshmeshkhdneh raisin-house), two octagonal imdmzddehs the and (the remains of an extensive fortress called locally Qal'a-yi Khdn. The purpose of the excavations was to investigate the occupation sequence of the town, the nature of the structures that are still visible above ground and the possible cause and date of abandonment of the site. The first trench, opened on the side of one of the prehistoric mounds (Site A), was a sondage to establish the succession of layers in this part of the site. This mound is situated near the northeastern angle of the site close to the junction of the Chari and Ghubayrd rivers. After passing through two layers which produced respectively Seljfiq and Sdmdnid pottery, the trench reached a level containing Tall-i Iblis II, III and IV pottery types of the second and third millenia B.C., together with a considerable number of flint implements. The second trench was opened some Ioo m. west of Site A, where the remnants of a kiln had been observed (Site B). The kiln was uncovered and found to be a brick kiln. The pottery inside and around the kiln was identical with that of the two upper layers of Site A.
The focal point of interest of Ghubayri is the Citadel area. On a level plateau, to the north of the Citadel, an extensive industrial area was revealed with six metal-workers' furnaces (P1. IIa), henna pits and storage jars. Dating evidence was provided by finds of a copper coin belonging to the 8th/I4th century and by underglaze-painted pottery of the same period. One of the most conspicuous features of this area was the Central Mound of the Citadel. A number of caves and tunnels were uncovered here. At the southeastern part of this mound Room I a was uncovered, where a complete bronze bowl with engraved decorations, many pieces of fine glass (P1. IIIc) and pottery vessels, a bronze ladle that was hung up on the wall, and three coins from the Muzaffarid period were found.

SURVEY

OF EXCAVATIONS

169

A further room beyond (Room 5) produced rich pottery finds and fragments of an ivory chessman. The room next to it on the eastern side (Room 3) produced seventeen large decorated unglazed jars (Pl. IIb), two underglaze-painted bowls of the Sultiandbd type (Pl. IlIa), glazed oil lamps and a number of iron and bronze utensils and jewellery. Room 4 on the southeasternside of the mound contained an underglaze-painted inkpot, two large bronze ladles and a coin of the Muzaffarid Shah Shujd' (765/1364-786/1384) minted at Shiraz (P1. IIIb). Apart from the rooms and caves a number of vertical shafts were found each of them leading into an underground chamber. These chambers were not connected with each other and they all contained slip-painted pottery of the Samanid type, lustre-painted and underglaze-painted vessels and numerous bronze and glass fragments. It seems that virtually every one of the rooms uncovered here was not a built structurebut originally an excavation in the subsoil. At the same time in some cases fallen roofs had been repaired with mudbricks and floors were plastered. From this first season of excavations the following conclusions can be drawn: (i) that the northern part of the town of Ghubayra had settlements already in prehistoric times, and that the pottery from there shows contacts with Tall-i Iblis; (ii) that there was an active industrial quarter north of the Citadel in Muzaffarid times; and (iii) that the Central Mound in the Citadel area, which was excavated during this season, had caves and tunnels which were used as service- and store-roomsin early and medieval Islamic times.
A. D. H. BIVAR and G.
FEHERVARI

HaftavdnTepe The third season of excavations at Haftavan Tepe, near Shahpur, lasted nearly two months, from the start of excavations on July 27 until the close of the camp on September 22. This has proved to be the most interesting season so far, both for its own results and for the promise of future success. The discovery of mud brick buildings in three quite different periods surviving to a height of from two to four metres has been in marked contrast to the poor preservation of the buildings found in the two previous seasons. The scale of these structuresis alone enough to show that Haftavan Tepe was for centuries a major town, perhaps the largest of all around Lake Urmia, especially in the late third and the early second millennia B.C. It has now been possible to distinguish seven main cultural periods at Haftavan Tepe, though there must be several more in the unexcavated levels deep in the citadel. The periods distinguished are as follows: Haftavdn I-Sasanian; Haftavdn II-Median and Achaemenian; Haftavan III-Urartian This season considerable work was carried out in the first, second, fifth, sixth and seventh periods. In Area C4, on the citadel, mud brick walls were found standing over three metres high, after excavation of a deep fill of fallen bricks. This building level was not burnt, and few objects were found in situ. There was, however, enough pottery to show beyond any doubt that this level must date to the Early Bronze III period. One room had on its north side the impression of a massive vertical wooden post, with that of a horizontal beam directly above it, thus indicating the full original height of about 2*5 metres of the ground floor room, from floor to ceiling. Some of the highest surviving remains of this level m. thick. When the building clearly belong to the upper storey, supported on walls not less than I.5 became dilapidated, the ground floor rooms were filled in. What survived of the upper storey was re-used in the second main phase of this level. It is to this second phase that the plan uncovered in Area CI, immediately west of C4, belongs: here was found a winepress, built of slabs forming the sides and floor and set against the corner of a wall; holes allowed the grape juice to flow out. Nearby was a II (c. 100ooo-800oo Haftavan V-Iron I (c. (c. 850/800-714 (?) B.C.); Haftavan IV-Iron B.c.); Bronze III (c. 2300-1900 Haftavan VI (c. I900-1200 Haftavan VII-Early B.C.); Iooo B.c.);
1200B.C.).

170

JOURNAL

OF PERSIAN

STUDIES

kitchen area with a large storage jar and a pit filled with querns. The whole appearance of the walls of this level suggests that they had stood abandoned and exposed to weathering for some time. The pottery includes a small proportion of matt painted ware related to the Painted Orange Ware of the same time at Hasanlti. In the overlying level were a few sherds of black ware with incised pendant triangles, paralleled at Trialeti in Georgia, one of many indications of cultural contacts with TransCaucasia. The discoveries in the 1971 season suggest the existence of a major public building in Period VII. In two new areas of excavations (X and Y) building levels attributable to the next period (VI) were uncovered. These levels seem to date to c. I800-1500 B.C., antedating the sequence excavated in the JX sondage in the 1969 season, completed this season with the excavation ofJX/8, beneath which water was reached, showing a rise in the water table. In X was found a kitchen complex in a courtyard (X/4), with a burnt layer and several storage pots in situ, one being full of carbonized wheat. In Y was found a level (Y/4), burnt and yielding much painted pottery, burnished and in some examples decorated with birds, a style similar to that of Kizil Vank to the north. The origins of this painted pottery are probably, however, in the Urmia region, not further north. Beneath this level was one (Y/5b) with impressive mud brick walls, one being strongly battered on the outside face and cutting across an earlier phase (Y/5d), one of whose walls stands two metres high. Characteristic of this period was the use of a narrow wooden beam laid horizontally along the base of the wall. These discoveries in Y point to a clear choice for future work, showing as they do the existence of a public building. The culture of early Haftavan VI was highly distinctive, with no evidence of the connections with upper Mesopotamia shown at Dinkha Tepe by the occurrence of Khabur Ware. In Haftavan V there seems to have been a decline both in material culture and in population. The buildings were constructed of stones set in mud, and the centre of the settlement was probably on the west side of the mound, where the large trench PQis situated. Here the work of the two previous seasons was continued, a second house contemporary with that found in 1968 being uncovered. Between and beyond were animal enclosures with several phases. The pottery includes a minority of the well known Iron I grey wares, side by side with the red wares persisting from Haftavan VI; among shapes is the tumbler found at Godin Tepe, where its date of c. 1200oo B.C. fits well with the evidence from Haftavin Tepe. Only limited excavations were carried out in levels attributable to Haftavan IV and III, largely because the discoveries in the post-Urartian levels on the citadel prevented the exposure of more of the plan of the Urartian building level. In W, a new area near the south fringe of the mound, Level 2 seems to be of Period IV. Urartian pottery was found in the overlying level. The most unexpected discovery of the 1971 season was that of a massive mud brick structure, at first thought to be a gateway but later seen to be a tower, with a slot through the greater part of the thickness of the outer wall facing north: this could have served for an archer. This feature, together with the manner in which the tower was built into the earlier mound, with a deep square shaft cut down inside it, suggest comparisons with the Median fortifications of Godin Tepe; and it is tempting to date this tower to the Median period, though it may be Achaemenian. Whatever its precise date, it seems reasonable to associate it with the large defensive ditch found in the 1968 season on the west side of the mound. The entrance to the tower seems to have been from the south-east corner and associated with stone cobbling which suggests a paved open area within the citadel of Haftavin II. The basement shaft of the tower must have served for storage or else have remained empty. Belonging to the Sasanian period (Haftavin I) is the latest perimeter wall round the top of the citadel, through which a small doorway with wooden threshold was found. To the west was a horseshoeshaped tower, of the same plan as that found in the 1969 season. It should be possible in due course to trace the whole surviving circuit of the Sasanian wall. A few burials of later Sasanian date, resembling those found previously, were excavated. Prospects for further work at Haftavin Tepe must now be reckoned with certainty as being very promising, especially for Periods VII, VI, III and II. The fourth season is planned for 1973.
CHARLES BURNEY

SURVEY

OF EXCAVATIONS

171

Masjid-i Solaiman La vue de la Terrasse sacree de Masjid-i Solaiman (P1. IVa) permet de distinguer quatre sanctuaires que nous avons identifies: No. No. No. No. I2
-

34-

A Un podium de l'6poque acheminide remani6 (les deux murs visibles) l'apoque parthe. Le " Grand Temple " qui semble avoir &t6celui de la deesse Athena (peut-etre Hippia?). Le Temple d'Heraclks. Le Temple de l'Ouest. La divinite d6dicataire reste inconnue.

La reprise de la fouille autour du podium a permis le degagement entier de sa face primitive SudOuest et de son angle Ouest (P1. Va). Sa face Nord-Ouest avec un escalier avait ete remaniee; quanta sa face Nord-Est, tres prouvie du fait d'un 6boulement, elle avit requ des murs de soutenement qui devaient consolider ce qui restait du podium (P1. IVc). En ce qui concerne la face Sud-Est du podium, l'boulement la fit disparaitre entierement avec la majeure partie du mur qu'elle supportait, puisque le monument primitif avait et6 biti sur le bord d'une terrasse qui surplombait presque a pic d'une vingtaine de metres, un profond ravin. Plusieurs pi6cettes 6lymaides des premiers siecles de notre ere proviennent de cette fouille, confirmant l'activite de ce sanctuaire du feu jusqu'a l' poque sassanide. Elles voisinaient avec de nombreux petits objets votifs appartenant, en grande partie a des parures fdminines, tels que de modestes bagues, bracelets, epingles, colliers de perles. Le seul objet de l'Fpoque acheminide qui ait 6te mis au jour lors de cette fouille est un scarabee en pierre grave d'un sujet classique de cette epoque: un roi terrasant un monstre (P1. IVb). Les travaux furent repris egalement sur le " Grand Temple " oi', at l'extreme fin de la campagne precedente et sous le sol, dans un endroit que je croyais tre une favissa, furent mis un jour quelques objets votifs (des cavaliers en particulier). La fouille de cette annie, faite en profondeur, a d~montr6 que ces objets provenaient d'un temple anterieur qui preceda celui qui fut entierement digage par nous. A une profondeur de pres de trois metres, fut atteint le sol de ce temple (P1. Vb et d), qui semblait avoir fit sauvagement d~truit en recouvrant nombre d'objets qui y avaient 6td d~pos6s par les fiddles. On y a trouvi des figurines de la deesse nue (P1. Vc); une tate de femme en terre cuite (P1. VIa); un personnage (cavalier ?) i coiffre occidentale. La poterie 6tait nombreuse: un rhyton ~ protomb de taureau (P1 VIe); un vase theriomorphe (P1. VId); plusieurs petits flacons 6maillis i deux anses (P1. VIIIb); des flacons de forme etiree (P1. VIIIa) ou d'autres, piriformes, 6maill6s de vert et incis6s (P1. VIc). Parmi les objets en m6tal, signalons un vase en bronze (P1. VIIIc). Nous rencontrions fr~quemment des figurines qui repr6sentaient des cavaliers mac6doniens reconnaissables a leur coiffure nationale la causia, armes d'un bouclier. D'autres cavaliers montaient, chacun deux chevaux entre lesquels 6tait placfe une figurine de dfesse nue (P1. VIb). De 1i provenait 6galement une petite tate de jeune homme, en pierre blanche, de l'6poque achfmfnide (P1. VIIb), trbs 6tg la proche par sa facture et sa coiffure de teu te du jeune prince en pte de lapis-lazuli qui avait t Persepolis.x trouvie L'intrpt de cet ensemble d'objets reside dans la possibilith de leur attribuer une date relativement exacte. En effet, le debut de ce premier temple qui fut aussi le plus ancien, doit remonter ~ l'6poque qui suivit la conqu&e de l'Iran par Alexandre le Grand. Quant Asa fin marquee par sa violente destruction, elle doit dater du milieu du IIe sifcle avant notre are, lors de la conqute du royaume d'Elymaide par historiques.p Mithridate I, ce que rapportent les sources d'Huraclks o0 d'autres fragments devant le parvis du temple egalement Les travaux furent poursuivis de sculpture ont 6ti mis au jour. Ce sont: une petite tate d'H6raclks (P1. VIId); un boeuf bossu; un (P1. VIIc) et une belle tratede reine lymaide, probablement du I-IIe sitcle de sphinx ou griffon ailU notre *re (P1. VIIa). R. GHIRSHMAN

xR.

(Paris I963), fig. 294. Mddes, Acheminides Ghirshman, Perse: Proto-iraniens,

172

JOURNAL

OF PERSIAN

STUDIES

QabrSheykheyn Excavations at Qabr Sheykheyn, funded by the National Science Foundation and the University Museum of the University of Pennsylvania, were carried out between October 28, 1970 and February 18, 1971. The assistance provided by Dr. Henry T. Wright of the University of Michigan, Mr. M. H. Khoshabi of the Archaeological Service of Iran, and the expedition's eight field staff members is gratefully acknowledged. The site lies about 30 km. south-east of Dizful and 20 km. north-west of Shushtar,just south of the Dizful-Shushtar road. Qabr Sheykheyn is small, approximately 8o m. long and 40 m. wide, its highest point rising 6 m. above present plain level. The mound was selected for excavation for quite specific reasons. Surveys of the prehistoric mounds of central Khuzistan by Robert McC. Adams in 1961 and Frank Hole in 1969 were analyzed by Hole in I969 for evidence of settlement change in the Susiana-Susa A periods. Hole's preliminary analysis suggested a hypothesis of change in settlement type during the Susiana d-Susa A period: a shift away from low settlements of small, uniform domestic structures in the early-middle Susiana periods to higher settlements, crested by a single, more substantial domestic structure, in the Susiana d-Susa A period. Survey results showed Qabr Sheykheyn to be suitable for testing this hypothesis. The mound is one of several late Susiana-Susa A settlements in the region clearly defined by the eastern drainage of the Barikh River, an area previously untested. Being a terminal Susa A settlement less than four-tenthsof a hectare in size, Qabr Sheykheyn offered an opportunity for horizontal clearance and the exposure of most of a Susa A settlement. Large scale excavation revealed four distinct, but closely successive periods of occupation. Preliminary ceramic analysis shows all of these occupations to be of late Susiana-Susa A times. The pottery assemblage includes fine paste " Susa A " beakers, shaved cups, ring base and concave base bowls; buff, grit paste, high neck jars and deep bowls; and deep, flat base, red ware bowls. Burnished black ware bowls and plates, previously known in Khuzistan only from Tepe Sabz and Tepe Farukhabad, occurred infrequently. The slight qualitative differences between the Qabr Sheykheyn assemblage and those of other excavated mounds of this period may reflect temporal and/or proximal factors. Quantitative data on the Qabr Sheykheyn pottery is now being assembled. Horizontal clearance across the top of the mound uncovered major portions of the architectural remains of periods II, III and IV. Period I, the mound's latest occupation, was only represented by a poorly preserved plaster floor with a quantity of in situ Susa A sherds. Period IV: 150 square metres of Period IV occupation were excavated, revealing the greater part, but not the complete plan, of a large domestic structure of at least four rooms. The walls of this house were constructed to thicknessesof 54 and 8o cm. employing bricksmeasuring 52 X 25 x 8 cm. A rectangular four-chamberedstorage bin was built into the corner of one room while a rectangular one-chamber oven was constructed against one of the compass oriented walls of another room. Red and black burnt floor areas and smoke blackened in situ sherds indicated cooking areas, but no hearths were found. Period III: Almost 500oo square metres of the Period III settlement were uncovered, disclosing, in some areas, three phases of rebuilding. In phase III C two house areas and several interesting architectural features were recovered. House I, to the south, consisted of three rectangular rooms defined by unbonded walls one brick (22 cms.) thick. Flecks of green-blue plaster still clung to these walls even
2 by though they were disturbed by later constructions. The central and largest room of this house, 4" m. in size, abutted the eastern-most of two brick platforms which lay between House I and House 2. 2.2 The platforms, of unbaked mudbricks measuring 56 x 10 x 8 cm., follow the pattern of all the mudbrick architecture at Qabr Sheykheyn, the sides being aligned with the cardinal points of the compass. On the eastern platform, which covered an area of more than 25 square metres, an infant with bracelets of minute paste beads on its ankles and wrists was located. The western platform, more than 20 square metres in area, was constructed to form a rectangular basin, 3 x 4 m., two brick courses deep. A few Susa A painted fine paste sherds and a broken grir ding stone were situated in the basin, but no other indications of the basin's function were found. To the north of these platforms lay House 2, also built of unbonded one-brick-thick walls forming

SURVEY

OF EXCAVATIONS

173

rectangular rooms. One room, with a red burnt hearth area against its western wall, contained an adult burial just below the surface of its plastered floor. The body was oriented east-west, extended on its back; a small red ware pot and a plain buff shaved cup were deposited just east of the head and one unbaked clay "sling missile" was laid in the crook of the right elbow. An adjacent room had a well " plastered floor with in situ sling missiles ", several black ware bowls, and one of the more than fifty " hoes " located during the excavation. stone In Periods III B and A the almost complete plans of three houses were recovered. These structures were built of mudbricks measuring 38 x 23 X 10 cm., and 26 x 22 x io cm., with unbonded walls usually o one brick thick, thinly plastered inside and out. House I, to the south, was of particular interest for its well preserved doorway with an outside step, a baked clay doorsocket and a I m. wide pit filled with potsherds. This rectangular, free-standing, one room house was destroyed at its eastern end by a later pit, but probably measured 3 x 5-25 m. House 2, a complex of walls interrupted at points by intrusive burials, was cut off by the later Period II house construction. House 3, to the north, was also partially destroyed by intrusive burials, but the information recovered from its excavation conforms to the architectural record for Houses I and 2: one or two brick thick, unbonded walls enclosing small rectangular rooms. An open area south of the entrance-way to House 3 was littered with smoke blackened and cracked stones, a large flint blade core, and a stone " hoe ". Period II: When the Period III occupation was abandoned and subsequently fell into ruin, the southern crest of Qabr Sheykheyn was sloping gently downward toward the east. On top of, and up against, the ruins of Period III a platform of unbaked mudbricks, each brick measuring 56 x 21 x o10 cm., was laid over an area of about 30 square metres. The purpose of this platform was probably twofold: to level this area and provide a strong foundation for new construction. The eastern end of the platform consisted of eight courses of brick whereas the western consisted of only three. The building erected upon this platform, the only building constructed in Period II, was more substantial than any built in Period III. It consisted of three rooms constructed from unbaked mudbricks measuring 60 x 22 X 1O cm., and 40 x 22 X IO cm. Most walls of this structure, though still unbonded, were three bricks thick. The house itself was oriented north-south following the long axis of its central room. This room, measuring 5 - 6 x 24 m., had a small rectangular bin and a dividing wall built up against its western wall, and a ledge or supporting wall against its eastern wall. Two doorways gave access to the central room. One door swung open into the central room and gave passage to the eastern room. Two baked clay door sockets and a wall depression for a doorpost attest to this. The second doorway led out of the central room to an open " court " on the north. An area littered with potsherds lay in front of this exit, recalling the sherd-filled pit by the doorway of House I, Period III B. In spite of its proximity to the surface of the mound, and the thirteen intrusive burials which were sunk into its confines, a considerable number of artifacts were discovered in situ in this house. Red conglomerate grinding stones and ground stone pounding tools were situated on the floor of the central room, along with a buff, grit-tempered vessel which is similar to objects at Susa called " wall-pegs ". Two large buff, grit-tempered jars, smashed in antiquity, lay on a black burnt area in the southern half of this room. On the floor of the eastern room the skeleton of an infant, probably abandoned with the building, was recovered. In the corners of the western room ground stone tools, large pieces of bitumen, a painted pot lid and sherds from burnished black ware bowls were found. The pattern of settlement unveiled at Qabr Sheykheyn records significant changes from Periods IV through II: the Period II occupation being but one impressive three-room house built upon a brick platform, the Period IV and Period III occupations being an extensive series of domestic units running the length of the mound. Analyses of flotation samples and faunal material, as well as distribution pattern studies of pottery types, lithic materials, and other artifacts, may provide more information about the different social units occupying these different settlements. Further excavation of small Susiana and Susa A settlements may confirm the original hypothesis that the settlement pattern at Qabr Sheykheyn was typical of Susa A times and reflects the division of Susiana-Susa A period society into new units of social organization.
HARVEY WEISS
15A

174

JOURNAL

OF PERSIAN

STUDIES

Shahr-iQfmis Following the soundings carried out in 1967, the second season at Shahr-i Qimis lasted for six weeks from late August to early October, 1971. The work was sponsored by the National Geographic Society with further help being received from the British Museum, the Ashmolean Museum and the British Institute of Persian Studies. The excavations were concentrated on four of Shahr-i Qtmis' more prominent structuresof Parthian date. At Buildings IV, VII and XIII it soon became clear that we were dealing with somewhat similar mud-brick constructions,each of which came to serve a probably secondary, funerary purpose. Each building exhibits a cruciform plan with a narrow staircase and four or more rooms on the ground floor. Traces of a first floor plan also survive at Sites VII and XIII. In Building VII mixed deposits of human and animal bones were found high in the filling of certain ground floor rooms. In the first campaign this same structure produced a hoard of Parthian drachmae of the early ISt century B.C.1 This year one room yielded a cache of fine pottery vessels including a tall jar with high loop handles and two opposed spouts in the base. m. Of greatest interest this past season, however, were the results obtained at Site V, some 5oo00 to the north of Site VII. Here we were able to uncover the greater part of a three-storied Parthian building over 23 X 23 m. in size with a protruding bastion at the mid-point of each external wall. Entered by a single" maze entrance " akin to others known from Parthian sites in Turkmenistan, this once formidable building sufferedfrom several local conflagrations before being severely burnt and then abandoned as a place of habitation. But before it was left wholly alone this structure was also used as a repository for human and animal bones, most of which had been exposed before they were buried. The deep, formerly vaulted rooms of Building V contain pottery that can be linked with that of Parthian and perhaps pre-Parthian settlements in the Gorgan plain. Moreover, while these same rooms have not provided any coins, they have produced a fine series of bullae,that is to say, clay impressions from bezel rings and stamp seals, not to mention one (or possibly two) impressionsfrom cylinder seals. These various impressionsshow geometric or abstract motifs, scenes with human figures-in worship, in battle or engaged in other activities-and animal designs. The latter usually depict individual animals, both horses and gazelles being represented with special care.
JOHN HANSMAN and DAVID STRONACH

Shahr-iSokhta The cultural sequence of Shahr-i Sokhta is a continuous development, subdivided in four periods, which covers more or less the whole of the third millennium B.C. This was first determined in 1969, but we had still to document this sequence from a single section. In 1970 the major trench was therefore laid out along the southern side of the Burnt Palace, the large building of Period IV whose destruction is probably contemporary with the end of the settlement. On a surface of 40 X 20 m., we have brought to light a group of houses and minor structures, built up on each other in successive phases. The analysis of them, together with that of the materials found, confirmed the periodization first recognized. The Palace was built above at least two adjoining buildings of Period III and the line of the foundations had been adjusted to the several irregularitiesof the earlier structures. The evidence does not show that the walls of Period III had been raised to be used as the foundations of the large building superimposed. While the cultural deposit of Period III is on average one metre thick, that of the preceding Period II is always more than three metres in this, as well as in all the other trenches dug in Shahr-i Sokhta. Period II seems to have lasted for quite a long time. In the present season Period I has only been excavated in a small pit (2 x 2 m.) on the western side of m. from the top of the Palace. the major trench. In this area, virgin soil has been reached, exactly 8.3 This deposit consists of a natural terrace madeff silted marls stratified with layers of sand.
1 See J. Hansman and D. Stronach, " Excavations at Shar-i Qiimis, 1967 ", JRAS (1970), p. 47 and pls. IXb and IXc.

to of Pl. Ia. GanjDareh Tepeat the beginning the 1971 season,showingthe P1. lb. Two sheepskullsattached thewall of a smallnichein a cubicle central area. excavotion parallelwalls of LevelD, GanjDareh. The nicheis formedby two short orpilasters.

Pl. Ic. Several thesmallcubicles LevelD, GanjDareh, constructed of of of to bricks.In theright-hand cubicle a largeclaypot attached is plano-convex wall. theplastered

out or P1. Id. A shallowdepression fire-pitin LevelE, GanjDareh,scooped and burned cobbles stone fragments. of virginsoil andcontaining

Pl. Ha. Ghubayra z971:

citadelindustrial furnace. area, metalworkers'

Pl. IIb. Ghubayrdcitadel, : Room Storage inprocess excavation. of 3. jars

A.D. type,in darkblue,light blue,and white. Abouti3th century P1. IIIa. Ghubayrd:bowlfrom Room3. 'Sultandbdd'

from Room4. ShdhShujd'(765/1364 P1. IIIb. Ghubayri: Muzaffaridbrasscoin to 786/1384), Shiraz mint. Scale 2:1.

Pl. IIIc. Ghubayrd: citadel,Roomia. Glassperfume-bottle A.D. 12th to zi6thcentury

de entierement P1. IVa. Vuede la terrasse Masjid-i Solaiman, ddgagde.

Pl. IVb. Masjid-i Solaiman: Podium. Scarablee achiminide.

Pl. IVc. Masjid-i Solaiman: Podium,c6tdNord-Est.

Sud-Oest. Pl. Va. Masjid-i Solaiman: Podium,cdtd

Solaiman: GrandTemple. Mursdestrois Pl. Vb. -Masjid-i superposes. bdtiments

Pl. Vc. Masjid-iSolaiman: GrandTemple. Temple primitif, cuite. figurine,terre

Pl. Vd. Masjid-i Solaiman: Grand Temple. Sol du temple primitif.

primitif,tatede P1. VIa. Masjid-i Solaiman: GrandTemple. Temple cuite. femme, terre

P1. VIb. Masjid-iSolaiman:Grand Temple. Templeprimitif, cavalier sur macidonien deuxchevaux, portant aussi unediessenue.

Pl. VIc. Masjid-i Solaiman: Grand Temple primitif, poterie Temple. emaillie et gravie.

terre primitif, vasethdriomorphe, cuite. P1. VId. Masjid-i Solaiman: GrandTemple. Temple

Pl. VIe. Masjid-i Solaiman: GrandTemple. en Temple primitif,rhyton terrecuite.

de Pl. VIla. Masjid-i Solaiman: Terrasse. Surface, tWte reine dlymaide,pierre.

primitif,tte P1. VIIb. Masjid-i Solaiman: GrandTemple. Temple en achgminide pierreblanche.

Pl. VIIc. Masjid-i Solaiman: Terrasse. Surface, sphinx ou griffon ailli, pierre.

Pl. VIId. Masjid-i Solaiman: Terrasse.Surface,tdted'Heracles, enpierre.

Pl. VIIIa. Masjid-i Solaiman: GrandTemple. Temple ordinaire. primitif,poterie

Pl. VIIIb. Masjid-i Solaiman: GrandTemple. Temple emaillie. primitif,poterie

Pl. VIIIc. Masjid-iSolaiman: GrandTemple. Temple primitif,vaseen bronze.

Seal. Presentation Pl. IXa. Tall-i Malydn: Cylinder scene,tableand vessels. Height

2"5 Delaporte, Catalogue des Cylindres, Musde du Louvre, Paris 1920o, P1. 34:5)-

cm. Mi59:

Iran Bastan, Tehran. (Cf.

snake. Height 2-35 cm. MO3: Pl. IXb. Tall-i Malyan: Cylinder Seal. Presentation scene,man-headed
(Cf. ibid, P1. 34:12).

Iran Bastan, Tehran.

Seal. Lion chasinganimal. Height Pl. IXc. Tall-i Malyin: Cylinder

i"9

cm. M088:

Iran.Bastan,Tehran.(Cf ibid, Pl. 34:zz).

view the Sokhta: Houseof Pit. General from south. In theforeground long,narrow Pl. Xa. Shahr-i whichhasproduced strawand timber store-room manywell-preserved objects.

Pl. Xb. Shahr-iSokhta: PeriodH early. Smallstylized and steatopygou female figurinewith marked pregnancy character.The style is close to the South Turkmenistan from Namazga IL figurines

de vue en Pl. Xla. TurengTip6, terrasse achMme'nide, gdnreale la partie explore'e 1971.

et mur Pl. XIb. TurengTipde, exterieur tou

Pl. XIIa. TepeDasht-i Deh: bonegamingpieces. Height5 cms.

bowl. Diameter cms. Celadon 20 P1. XIIb. TepeDasht-i Deh: imitation

SURVEY

OF

EXCAVATIONS

175

Among the most interesting finds are a clay female figurine of Namazga III type (Pl. X b) in the late layers of Period II; a bronze pin with winding ends (Pl. X c) and polychrome ware in Period I. This last has been quite a surprise, as we had thought on the basis of earlier campaigns that this class of pottery was a typical product of Periods II and III. Quite remarkablehas also been the discovery of a pot buried in the NE corner of room CLXVI and containing camel-dung, most likely stored as fuel. on the disputed problem of the camel's domestication. Several scholars, among them Zeuner, put the beginning of it to the second half of the second millennium B.C. The extensive excavations in area X have been continued on a very restricted surface in order to bring completely to light another building of Period I-IIII. Again this was a residentialunit of rectangular plan divided into seven rooms, facing the " House of Stairs ". The main feature is a brick-lined square pit, excavated on the southern side and by that it has been currently called the " House of Pit " (P1.X a). Some of the rooms are connected by means of small staircases, due to the variations in level, so common in third millennium towns. On the eastern side, the house originally ended with a long rectangular store-room. In a second phase, an external staircase was set up along the NE corner which totally covered the small store-room. The bricksof the staircase sealed all the deposit of the small rooms, thus preserving most of the perishable materials too. We could therefore recover a complete basket and fragments of another, textiles, timber objects and an intact V-shaped sling, made out of a tamarisk twig. During the same phase the plan of the house, originally consisting of five rooms was enlarged on the western side by adjoining two small square rooms, which further reduced the width of the lane dividing the " House of the Pit " from the " House of Stairs ". The phase of enlargement " House of Pit IIb " corresponds, therefore, with " House of Stairs IIc ". Shahr-iSokhta Project: TepeRud-iBiyaban2 This is a small mound, recognized since i 968 as a workingcentre for pottery. On its surface are spread over 50 kilns and the whole of it is thickly covered with slugs and wasters. A small test was done in x969 mound. We have now cleared 9 kilns and the vertical section shows three major phases of occupation, the natural soil being reached at a depth of 4-8 m. The kilns are rectangular in plan, with a single or double firing chamber at one end, a long atriowith sloping floor, a small entry and two chimneys. They are all of the downdraught type and they could easily reach a temperature of over 15oo0 C. On the outer sides of the entrances were actually piled thousands of overbaked sherds and vitrified slugs. The highest temperature was quickly reached inside the kilns and the degree of speed was hardly controllable, as is witnessed by the huge amount of wasters and the destruction of at least two of the excavated kilns. In fact over a certain temperature the walls of the kilns themselves smelted and vitrified, reducing the amount of circulating air so that the furnace had to be put out of use. The kiln was then filled with the surroundingwasters, as in kiln 35, or partly reused for a new one, as were kilns 5 and 32. The 4 -8 m. of the total archaeological deposit are datable to Shahr-i Sokhta Period III, whose layers are nevertheless rarely thicker than I - 5 m. This was caused by the different use of the two sites: the deposit-quickly piled wasters on the former and slow everyday accumulation on the latter. Twenty-nine km. separate Shahr-i Sokhta from Tepe Rud-i Biyaban 2; it is probable that the
pottery produced here was distributed to the near-by small villages, which are concentrated in large numbers on the Rud-i Biyaban dried delta. Other pottery working centres of this period presently recognized are Tepe Raikes, 3 kms. east of Shahr-i Sokhta, and at Shahri-i Sokhta itself, on an area c. I hectare wide, localized on the northern side of the mound. Single, isolated kilns are anyway present on several mounds. Unfortunately in I971, due to the drought and to personal trouble, it was impossible for the Italian Archaeological Mission to undertake the usual campaign. Anyway for 1972 we are planning an intense and longer season.
MAURIZIO TosI

As the floor is datable to the middle of Period II (2600

B.c.),

the find is important for the light it sheds

to clear some of the exposed kilns. In I970 a 30 X o10m. trench was finally set up at the centre of the

176 Tall-i Malydn

JOURNAL

OF PERSIAN

STUDIES

Tall-i Malydn is a large mound (c. 3oo00 hectares) located in the Kur River basin, Fars province, some 46 kms. north of Shirdz. The site was visited by the author in 1968 during a survey of the Baizd district. Soundings were first undertaken some years ago by the Archaeological Department of Fars and the first season of excavation under the joint sponsorship of the University Museum, University of Pennsylvania, and the American Institute of Iranian Studies was conducted in June and July, 1971. This season's objectives included a preliminary chronology and an initial exploration of the cultural topography of the site. The chronological results may be summarized in tabular form:

Period
Late period Hiatus Middle Elamite

Source Evidence of
Operation D and surface Inscribed bricks from surface Operations A, B, D, and surface Operations B, F, and surface

Hiatus (?)
Kaftari

Hiatus (?)
Banesh

The identification of the late period, which is probably Parthian or Sasanian, must await further excavation. The Middle Elamite period is known solely from the inscribed material described by Dr. Erica Reiner in the accompanying note. The latest levels of the Kaftari period, as defined by Vanden Berghe de (Archedologie l'Iran Ancien, p. 42 with references), are assigned to the early second millennium on the basis of three radio carbon dates.2 The Banesh period, characterized by bevel rim bowls, low trays, heavy pedestal bases, club rims and some painted ware, may be as early as the late fourth or early third millennium.3 The architecture uncovered includes fragmentary remains of the late period, three levels dating to the Kaftari period and one level dating to the Banesh period. The Kaftari and Banesh buildings were all made of unbaked brick (44 X 22 x 8 cm.) walls with thick mud plaster. One Kaftari building level had fine red painted walls and contained fallen fragments of wall painting in red, yellow, black and white. The inscribed bricks and a stone plaque from the surface, as well as the single tablet, found in Kaftari levels are discussed below. Other finds, mostly of the Kaftari period, include six intact cylinder seals and several stamp seals. The most common motif is a presentation scene with one seated and one standing figure. This scene, which is found in several styles and on both stamp and cylinder seals, is associated with a man-headed snake in one case and with a table supporting three vessels in another (P1. IX). Close parallels for most of these seals are found among the seals of local style at Susa dating from the early half of the second millennium.4 Many female and animal figurines were discovered, as well as bronze ornaments and tools, numerous flaked stone tools, stone bowl fragments, and beads of various semi-precious stones. At the height of its occupation in the second and third millennia B.C. Malydn must have been a place of considerable economic and political importance. Its local predominance is perhaps best gauged by the fact that it covers more than twice the combined area of all 77 sites in the Kur River basin known to have been occupied during Kaftari times. The participation of Malydn in extensive trade is attested by the presence of obsidian, lapis lazuli, fragments of steatite hut pots, and Persian Gulf shell. The inscriptions of Elamite rulers as well as strong parallels with the glyptic style of Susa suggest a strong political bond between Malydn* and Elam. At the same time the distinctive local pottery style of Malyan indicates a considerable degree of cultural autonomy. We hope to learn more about this intriguing situation during future seasons of excavation. W. M. SUMNER
Unpublished dates, Tehran University Nuclear Centre: 3TUNC3i: 4815+_?9IBP. TUNC 28: 3618+65 BP., TUNC 29: 3635?63 BP., TUNC 4 I am indebted to Dr. Edith Porada and Pierre de Miroschedji for personal communications regarding the stylistic affinity and 30: 3640 ? 65 BP., 5730 half-life. I am indebted to Charles dating of the seals. Bovington and Azizeh Mahdavi of the Tehran University Nuclear Centre for these dates. * For a possible identification of this site see " Elamites, Achaemenians and Anshan " above. EDITOR.
2

SURVEY

OF EXCAVATIONS

177

Tall-i Malydn-Inscribedmaterial The inscriptional material falls in two groups: Old Babylonian (see I below) and Elamite (see II). I. Old Babylonian. M129 is an Old Babylonian tablet (written in Sumerian) enumerating offerings for various days of a month whose name is broken. Parallel tablets have been found at Susa and published in MDP XXVIII, nos. 461 and 462. II. Elamite. This group consists of (a) fragments of bricks and (b) a fragment of alabaster.
(a) Six fragments of three or more bricks (M

inscriptionsof Hutelutuv-Iniuginak,to which two duplicates, both in a private collection,

are known. Three other fragments, M oo004, and 17o are too fragmentary and badly 00oo5,

oo001, 002,

00oo3, o018, 032,

171) represent brick

preserved to permit identification. The alabaster fragment contains a curse formula, so far known only from the stela of (b) Silhak-Invuginak; there exists, however, a fragment of a stone vase, MDP XI, no. 104, which can be assigned to Hutelutul-Invulinak, the writing and context of which is very similar to the Malydn alabaster fragment.
ERICAREINER

TepeYa For a report on the 1971 season see pp. 89-Ioo above. .yd
The Yahyd Project: TepeDasht-iDek Excavations by the Harvard Expedition to Tepe Yahyd were continued at Tepe Dasht-i-Deh from

June to August 1971. Adjoining the trench of 1970 on the central mound (Iran IX (1971), pp. 182-3)

35 sq. m. was excavated to the floors of the earliest major architectural level (period 2) and a further 50 sq. m. to the floors of the succeeding period (period 3). Part of the west edge of a large fortified residence of period 2 was excavated. The building which probably dated to the I3th century A.D. was approximately 53 m. sq., Its defensive outer wall consisted of twin lines ofpise'65 cms. thick and 70 cms. apart, the space between being filled with stone and gravel. Outside a glacis sloped into a moat-like depression. Inside was a range of windowless store rooms surmounted by a second storey now collapsed which appears to have been used for domestic purposes. The innermost rooms are as yet unexcavated. Finds from the debris of the second storey included a fine inscribed mould for the manufacture of pottery together with sherds of " Saljuq " plain and lustred pots. A relatively short period of time must have elapsed before the construction of a smaller building in period 3 since the exterior wall of period 2 was reused in the succeeding period. Three complete rooms and parts of an open or semi-open courtyard were excavated. Resting on the latest floors were large numbers of objects which testify both to the domestic nature of the area excavated and to the haste of its abandonment. In a group beside a hearth were found a bronze pen case, a dice and two bone gaming pieces the remains perhaps of an abandoned game of backgammon. The more complete of the two gaming pieces stands five cms. high and is decorated with a delicately carved geometric design coloured in black and red. The other has an identical design in red (P1.XIIa). Iron knives, spoons, a bronze
lamp and other domestic objects were found together with a number of fine glazed bowls. One of these with a frit body and a hammer rim has a green glaze and fluted exterior in imitation of the Chinese late Sung " lotus bowls " imported in large quantities in south eastern Iran in the 13th century A.D., This piece was found drilled through the foot ring so that it could be hung for display (P1. XIIb). Other pots had underglaze decoration paralleled on the finest " SultinbiTd " pottery. However, bowls with the turquoise ground most characteristic of this type were much less common than those decorated in blues, purples and greens on a white ground. Two surface discoveries on the mounds around the central building are of particular interest. In one area a large concentration of" buns " of iron slag suggests extensive iron working. Beyond the iron slag and 12o m. SE. of the excavation are traces of unmortared rubble walls and column or pier bases,
16

178

JOURNAL

OF PERSIAN

STUDIES

which probably belong to a mosque, the arcaded prayer hall of which was seven bays wide and three bays deep (26 m. by 8 m.). Mounding to the east indicates the existence of a courtyard surrounded by mud brick walls or a single arcade. The plan indicated suggests the Umayyad and early 'Abbasid mosque as represented at the Great Mosque at Sirf, although the details and date can only be determined by excavation.
ANDREW WILLIAMSON

TurengTpe' ' La campagne de 1971 Tureng Tepe a donn6 des resultats d'une importance exceptionnelle, tant sur le plan historique qu'architectural. Elle a 6te entierement consacree a l'exploration du tepe principal (le " Mound A ") dont une large surface a ete l'objet de nos recherches,a la fois au sommet et sur la pente meridionale. L'essentiel du travail a porte sur les periodes IV, V et VI, mais les periodes III Cl et VII nous ont 6galement livre des vestiges d'un grand int&ret. PiriodeIV: Vers la fin de la periode IV, et plus precisement vers le debut de l'epoque achemenide, se situe l'edification d'une inorme terrassede brique, haute de quelque 14 m., qui &taitimplant6e sur les flancs d'un tepe ant6rieur dont elle avait recouvert le sommet. Une partie de la fagade meridionale a pu etre degag&ecette annie, sur une largeur de plus de 20 m. (P1. XIa). Le mode de construction de cette terrassea pu tre 6tudi6 en detail; il est du plus haut int&rft. L'edifice est constitue de murs et de blocs independants les uns des autres; les murs dessinaient un cadre exterieur et une sorte de grille interieure dont les vides etaient remplis de blocs rectangulaires eux aussi appareillis. Le mur de ceinture etait fonda sur un lit de tres grosses pierres noyees dans un mortier brfld et reposant ellesmemes sur plusieurs assises de briques egalement brulles disposees dans une tranchee qui avait 6te creusee a travers des couches beaucoup plus anciennes. Les briques de la terrasse sont de dimensions inusities: 75/45/16 cm. Elles sont tant6t crues, tant6t brfiluesplus ou moins complktement, comme l'taient les briques de fondation, ce qui suppose 6videmment une combustion volontaire. Le sommet de la terrasse n'a 6te atteint que par places; il ne semble pas qu'aucun monument s'y Elle rappelle en ceci comme a d'autres points de vue les terrasses de Pasargades ou de soit dresse. ' Masjid i-Suleiman, dont le sommet parait avoir ete l'origine depourvu de tout edifice. Il est probable que ces terrassesavaient en elles-memes une fonction cultuelle. Il est en tout cas capital qu'une construction de ce type ait ete pour la premiere fois decouverte dans cette partie de l'Iran, l'ancienne Hyrcanie qui a sans doute joud un r6le de premiere importance dans la gen~se de la religion zoroas" trienne. Il est au reste probable que la terrasse de Tureng Tep6 date du VIe siecle, car en juger d'apres la ceramique, sa construction est sirement posterieure au niveau 1 du " Mound C ", lequel remonte au VIIe siecle. Elle fut d'autre part detruite des avant la fin de la p6riode acheminide. III Pe'riode Cz.: La tranchee de fondation du mur de ceinture de la terrasse acheiminide a et6 t travers les vestiges d'un important idifice d'epoque T.T. III Cl, peut-etre un palais, dont les creusfe de bois et dont les murs de brique crue reposaient sur un lit de 6taient revetues d'un facades et de cailloux. La placage de ces fondations est analogue a celle du batiment T.T. III Cl technique pierrailles trouv6 dans la zone E; en outre une ceramique abondante, dont un " dipper ", ne laisse aucun doute bitiment. sur la date de ce
Piriode V: Aprbs la destruction et l'abandon de la terrasse achiminide, les ruines servirent de soubassement Attoute une succession d'ddifices qui couvrent la totalit6 d'une p~riode allant du IVe sidcle avant notre Areau debut de l'empire sassanide. Plusieurs niveaux datent encore de la dernibre phase de l'empire ach6minide; ils sont caractdrisis par divers 6difices en brique crue dont la destination n'est pas encore claire. La cframique sonore dicouverte ddji en 1969, t surface flamm~e brun rougeatre et noire et A bord rouge, date du plus recent de ces niveaux ach~minides et se retrouve encore au
2

Strabon,XVI,i, I8. I1s'agitdes templesd'Ath6naet d'Art6mis qui rapporterentAMithridate I, io.ooo talents. Voir aussi

Justin, XLI,6,7,8.

SURVEY

OF EXCAVATIONS

179

A niveau suivant, presileucide. Puis l'apoque sdleucide, en depit d'une evidente continuite dans l'6volution de la ceramique (continuite qui s'affirmed'ailleursjusqu'au debut de la periode islamique), des imitations locales de la ceramique hellknistiqueapparaissenten abondance. Enfin la periode parthe comporte plusieurs niveaux, dont le plus recent se signale a nouveau par d'importants edifices de brique crue, malheureusement endommages par les fondations de la premiere forteresse sassanide. Piriode VI: La fouille de la forteresse sassanide s'est poursuivie et ses refections successives se laissent maintenant distinguer plus clairement. Le niveau le plus ancien, qui date probablement du d6but de l'empire, comportait un temple du feu qui reste encore a explorer, et des tours exterieures semi-elliptiques, perches de meurtrieres et dont les briques crues dessinaient des frises au d6cor geom trique (P1. XIb). L'une d'elles, conservie sur 4,20 m, a 6t entibrement explore. Comme & Istakhr, cette tour ouvrait sur un corridor interieur dont les portes 6taient disposees en chicane. Plus tard ce couloir interieur fut entierement rempli de briques cependant qu'un glacis etait edifi6 en avant du mur d'enceinte et des tours. Enfin un dernier etat de la forteressesemble avoir a nouveau comporte un temple du feu. Piriode VII: Des apres l'abandon de la forteresse, qui date probablement de la destruction de l'empire sassanide, des fosses furent creusees dans les ruines; elles renfermaient une ce'ramiqueabonfait dante, d'epoque assurementislamique, mais ofU encore entierement defaut la ceramique a glagure. Puis fut difie le batiment stuqud qui a 6te explore en 1967 et 1969 et scrute a nouveau cette annie. Il s'averedisormais qu'il s'agit d'un temple du feu dont la construction est contemporaine de l'introduction dans cette rigion, probablement vers la fin du VIIIe siecle, de la ceramique a glaqure, dont deux fragments seulement ont 6t6 decouverts dans les fondations de l'edifice. La destruction de ce temple zoroastrienne semble pas posterieurea la fin du Xe siecle et est meme vraisemblablement plus ancienne, a en juger d'apres la ceramique trouvee dans les nombreuses fosses qui l'ont travers6 et qui n'ont livr6 aucun materiel plus recent. Aucun habitat correspondant a ces fosses n'a ete decouvert: peut-tre les populations qui detruisirent le temple du feu faisaient-elles partie de ces groupes nomades islamises, venus du Turkestan, dont on sait le r6le dans l'radication du zoroastrianisme dans le nord de l'Iran. La chronologie des periodes les plus r6centes de Tureng Tepe, telle que nous l'avons present~e pricfdemment (IranVIII, 1970, p. 208), doit donc tre modifiee comme suit: IV: V: VI: VII: VII A: VII B: VII C: Age du Fer (? VIe-Ve siecle); P6riode intermidiaire (IVe sidcle B.c.-IIIe siecle PNriodesassanide (IIIe-VIIe siecle); Islamique ancien: fosses du VIIe-VIIIe siecle); temple du feu (fin VIIIe siecle); fosses a ceramique A glagure (IXe-Xe siecle).
A.D.);

Il est evident que devant l'importance de ces r6sultatsil faudra maintenant envisager l'extension de la fouille sur une ichelle beaucoup plus large.
JEAN DESHAYES

ADDENDA
Seh Gabi An initial season of excavations was conducted at the site of Seh Gabi by the Iran Project of the Royal Ontario Museum in the months ofJune, July and August, 1971. The site is located in the saddle between the Kangavar and Assadabad valleys in the central Zagros, approximately 7 km. to the northeast of Godin Tepe. The site consists of six distinct mounds in an area 550 x 250 m. The highest of the mounds, Mound was the principal focus of excavations during this season, although the other mounds were also tested. B,

180

JOURNAL

OF PERSIAN

STUDIES

stones. A number of infant burials were also found. These are almost invariably in bowls, with the body tightly contracted. In two instances, textile impressionswere found, and in one of the two, what appear to be fragments of the textile were recovered. In yet another of the burials, copper beads were present. Of the other mounds tested, Mounds A, E and F seem to form a cluster and may originally have been a single mound. The material here is different from that on Mound B, and corresponds to that from Godin VII and VI. The lowest materials on these mounds is very like that at Godin VII. The pottery is characterized by a plain buff coarse ware in a limited repertoire of shapes. Notched horizontal lug handles and an incised or moulded rim decoration like that at Godin are the fossil indicators of this assemblage. The material that overlies this is like that at Godin VI and should probably be compared to Giyan VD as well. Painted wares appear, but they are different from those at Mound B. The pottery is rarely overfired, and the paint is matt and not usually blistered. There appears to be some overlap in the range of motifs, but in general they are quite different. The plain wares of this material at Seh Gabi are closely related to those of Godin VI and quite distinct from all of the plain wares of Mound B. No architecture was recovered from Mounds A, E and F, but two burials were found in Mound E. Both were children, buried in large pots with a large bowl covering the opening of the pot. Aside from
the vessels, no grave goods were found with these burials. The relative position of Mounds A, E and F as compared with Mound B is as yet unclear. It may tentatively be suggested that Mound B is the earlier, and that the sequence is Dalma, Giyan VC, Godin VII and Godin VI. Such a proposal presents certain problems, since the Godin VII plain wares are not encountered at Giyan, and Giyan VD (= Godin VI) seems to follow directly upon VC. There is some tentative independent confirmation of this reconstruction at Sagsabad, where Godin VII plain wares were found in context with terminal Siyalk III materials. Once again, analysis of this material in the laboratory and future planned excavation will hopefully solve this problem.
Louis D. LEVINE

and that which did appearwas of mud brickconstruction. Above the Dalma period materialswere materialsrelatingto the assemblage Giyan VC. The of transition betweenDalma and GiyanVC is as yet unclear,but appearsto have been gradual. In level of A, the uppermost the GiyanVC relatedmaterials,an area of 300 squaremetreswas opened. One house plan and partsof otherswere uncovered,all in poor statesof preservation.The concomplete structionwas of mud brickwith a white plasterovercoaton the face of the walls. At least one of the levels B and C, were also exposed,albeit in a roomswas burned. Two furtherlevels of architecture, muchmorerestricted area. LevelB is clearlypartof the sameassemblage level A. LevelC, however, as the and thoserelatedto GiyanVC. This hypobetweenthe Dalma materials may represent transition thesiswill requiretestingin the laboratorywith the analysisof the materialsexcavated. The potteryof the levelsA and B consistsof a red slippedplain ware that is highly burnished and fromthat of the Dalma period,a plain buffcoarseware,and a finger-tip ware. impressed quite different All are hand made and strawtempered. A paintedpotteryalsooccurs. It is usuallyoverfired, with the paint oftenblistered.The fabricis dense,buff,and grit tempered. Someexamplesappearto have been made on the wheel. to The othersmallfindscorrespond thosefromthe Dalmaperiod. Clayspindlewhorls,miscellaneous bone awlsand a bonepunchor handle,bone and shell beads, shapedclay pieces,clay animalfigurines, chipped stone, again mostly blades, and variousground stone mortars,pestles, querns and rubbing

In Mound B, the entire sequence was tested, from virgin soil to the summit, a total deposit of c. Io m. The lowest levels produced pottery similar to that found at Dalma Tepe, Azerbaijan. A plain red slipped ware, a variety of punctuated wares, and a painted ware were all present. There was also a variant of the painted ware that was bichrome, although this was extremely rare. Other small finds from this period included a limited chipped stone assemblage (mostly blades), miscellaneous clay pieces (spindle whorls, clay balls, etc.) and worked bone objects (awls, spatulas). Little architecture was uncovered,

SURVEY

OF

EXCAVATIONS

181

Suseet Susiane Les travaux de la Ddligation archdologique frangaise 't Suse et &Djaffarabad ont &t poursuivis en Mlle G. Dollfus, charg6 de recherche au CNRS, (voir ci-dessous). A Suse,sur le chantier stratigraphique de l'Acropole, recherches furent menses comme les anndes les sous la direction de A. Le Brun, attach6 de recherche au CNRS, assist6 par G. Sabath6.2 prdcddentes La fouille porta principalement sur la couche 17. Des 1969, cette couche avait paru constituer, avec la couche I6 susjacente, une des charni res de la sdquencestratigraphiqueobservie au flanc de la tranchde de Mecquenem. En 1970 nous avions notd que la c6ramique de la couche 16 se diffdrenciaitnettement de celle de la couche I7; la couche 16 6tait en outre caractdrisle par d'importants vestiges architecturaux (3 tats de construction) tandis que la couche 17 apparaissait en coupe constitute de terre cendreuse et de debris. Nous avions recueilli dans les trois niveaux 16 A, B, C plusieurs tablettes protodlamites. La question se posait de savoir si la couche 17 en poss6dait 6galement. En 1971 la fouille montra que la couche 17 comprenait, en arrierede la coupe, des vestiges d'habitations importants (17 B), de meme orientation nord-est sud-ouest que les constructions du niveau I6, objets se trouvaient sur le sol supdrieur. Apres l'abandon et la ruine de cette habitation, le secteur vit une accumulation progressive de d6bris (17 A) qui parait indiquer que l'occupation se poursuivit normalement a proximit6. La materiel archdologiquede 17 A et 17 B est homog ne. Ce materiel comprend des tablettes: 6 en 17 A, 3 en 17 B; 4 autres proviennent du contact I7-16. La plupart d'entre elles portent des ddroulements de cylindres que l'on retrouve encore sur des bouchons et scellements de jarres (animaux antith&tiques, araigndes, animaux et personnages, scenes de chasse, oiseaux &tate triangulaire). Plusieurs cylindres en calcaire et en stdatite furent 6galement trouvis. Les tablettes ont la forme de " coussins bombds sur les deux faces "; elles ne portent que des chiffres. Les premiers signes apparaissent sur les tablettes trouvdesau contact de 17 et de 16; ces tablettes pourraient reprdsenterune transition entre les & tablettes a notation numbrale du niveau 17 et les tablettes acriture proto dlamite des niveaux 16 et suivants. ' La cdramique est caractdrisdepar des 6cuelles et des vases grossiers hauts alvre amincie, dans la proportion de 35% du total en 17 B et de 36% en 17 A. Les vases grossiershauts sont months &la tournette; ils ont dtd ddtach6sdu support a la ficelle (string cut). Les formes caractdristiquesdu niveau I6 (fonds en moignon) n'apparaissent pas ici. Ainsi, tout en prdsentantdes traits nouveaux, le niveau 17 B se rattache sans difficultda l'ensemble Suse l'apparition des niveaux 22 &17. L'apparition d'un systme de notation numbrale precede bien & d'un systeme d'6criture. Reste a verifier si les niveaux infirieurs (18 et la suite) ne contiennent pas ddj& ces memes tablettes a notation numdrale. On notera en tout cas que cette invention se place dans un contexte archdologique caractaris6,en ce qui concerne la ciramique, par la presence des 6cuelles grossieres. Celles-ci ne se rencontrent plus dans le niveau I6 oiu apparaissent les tablettes proto 61amites. Palais du Chaour Une deuxieme campagne de fouilles, sous la direction de A. Labrousse, assist6 par R. Boucharlat,
permit d'achever le d~gagement de la grande salle hypostyle.3 Ses dimensions sont de 37,20 m sur 34,50 m. Ses abords furent pr~cis~s; si trois de ses c6tds paraissent avoir it6 bordds de portiques, le x Pour mais de plan different. Trois 6tats d'occupation furent reconnus (I7 B 1,2,3); de nombreux vases et 1971 du mois de janvier au mois d'avril.1 A Djaffarabad, une troisieme campagne s'est ddroulke du Io janvier au
25

fdvrier sous la direction de

les ann6es pr6c6dentes voir Iran VIII (1970o), p. 190-4, d'une premiere 6tude de la glyptique par P. Amiet, ibid., et IX (1971), p. 178-81. On trouvera un compte rendu p. 217-233 et par une 6tude descriptive des documents 6pigraphiques par F. Vallat, p. 235-245. d6taill6 des fouilles A Djaffarabad et A l'Acropole de Suse de I969 A 197I dans les Cahiers de la De'ligation arche'ologique 3 Le portique nord est seul conserv6 mais la pr6sence du radier autorise raisonnablement la reconstitution des portiques est et franfaise en Iran I (Paris, I971). 2 Voir le compte rendu pr61iminaire des trois sud. Un compte rendu d6taill6 des fouilles de 1970 et 1971I au premieres camChaour paraitra dans le no 2 des Cahiersde la DAFI. pagnes (1969-1971) A l'Acropole de Suse par A. Le Brun dans les Cahiersde la DAFI 1, p. 163-216. Ce compte rendu est suivi

IJSE

iVAi~AIS *'.1
0
10 20 30
40

-)A
ETATDE RECONSTRUC
Fig. r.

SURVEY OF EXCAVATIONS

183

Cette salle est bloqude au nord par des salles plus petites dont quelques unes ont td d6gag6es. ' Deux inscriptions fragmentaires provenant de bases de colonnes permirent F. Vallat de lire a nouveau le nom d'un Artaxerxes; cette lecture renforce l'hypothese de la construction du palais par Artaxerxes II. Dans le mur ouest de la salle hypostyle furent d6couvertes,sur sept assises,depuis la base du mur, des briques dont la terre avait dtd mdlangde au moment de leur fabrication avec du mercure liquide. L'oxydation du metal dtait complete pour les inclusions les plus petites, mais les plus importantes contenaient encore une fraction du mdtal liquide dont plus d'un kilo au total a pu tre recueilli. L'6tude des vestiges postdrieursa l'abandon et a la destruction du palais permit de reconnaitre, antdrieurementtala pdriode islamique, cinq occupations successivescorrespondanta la piriode parthosassanide. Dans une petite cachette furent trouvies des coupes en cramique, une fiole en verre, un n6cessairede toilette et des objets de parure en m6tal et en os. La prospection dlectrique du tipd du Chaour par A. Hesse, charg6 de recherche au CNRS, donna les grandes lignes du plan gandral du palais. Apadana La restauration de l'Apadana fut poursuivie avec l'aide du Service de Protection des Monuments historiques.4 Les recherchesprdliminairesconduisirent a des observations intdressantla construction de la terrasse et le plan du palais. A l'angle sud-ouest du palais, la terrasseest constitute sur toute sa hauteur au dessus du niveau de la nappe aquifdre, soit sur plus de 18 m, par une accumulation de graviers et de galets de rivibre avec quelques debris. Jusqu'ici la plus forte 6paisseur de graviers observie l'avait t6 dans les tranchdes de fondation des murs des cours intdrieures du palais; elle n'excedait pas i I m. On rapprochera ces mesures du texte de la charte de fondation qui parle " de 20 et de 40 coud6es ". ' La masse de gravier de la terrasseest contenue l'ouest par un puissant mur de terre pisde qui forme 6galement le rempart. La face de ce mur, de profil galb6, fut ddgag6 l'angle nord-ouest du tip6 de l'Apadana, sur toute sa hauteur depuis le niveau de l'eau; elle prdsente un parement de briques crues recouvertes de torchis. D'autres observationssont venues 6clairer les preoccupations d'ordre technique des constructeursdu palais et leurs connaissancestechnologiques: leurs efforts cependant repurent empecher des tassements, compensas lors d'une restauration du palais, probablement celle d'Artaxerxes II, puis la formation au traversde la terrassed'un systeme de trois grandes failles est-ouest; l'une de ces failles est antdrieurea la pariode islamique; une autre est encore active aujourd'hui. Dans ces failles, des sols basculkrentpar
grandes plaques et des 16ments de mur s'y enfoncerent qui constituent parfois les seuls vestiges sur

c6td ouest, par contre, est occupd sur toute sa longueur (37,20 m) par une seule salle de 4,Io m de large.

lesquels appuyer aujourd'hui une reconstitution. De l'examen de l'infrastructuredu palais ressort h l'dvidence l'unit6 de sa conception et de sa construction. Tous les vestiges de construction (autres que ceux de la pdriode dlamite) trouvds au-dessous du niveau des sols du palais sont partie intigrante de sa construction. Ainsi, les huit bases dviddes digagdes par Mecquenem sur la face nord de la grande cour de l'est, a 2,50 m environ du dessous du niveau du sol de cette cour, furent-elles mises en place lors de la construction du palais; deux dans chacun des angles nord-est et nord-ouest de la cour, les six autres de part et d'autre des trois portes qui
s'ouvrent dans la fa<ade nord de cette cour; ces bases ont pu servir t maintenir en place le pied de huit grands mats. La d~couverte en place de nombreuses pierres supports de crapaudine (deux d'entre elles ont conserve leur crapaudine de bronze scellde au plomb) permit de reconnaitre des passages et des portes qui n'avaient pas 6td signalks jusqu'ici. I1 en rdsulte un plan du palais (Fig. I) assez difffrent de celui que nous connaissions; ce plan se rattache nettement t celui des constructions palatiales assyrobabyloniennes.
JEAN PERROT 'Un rapport prdliminaire sur les travaux conduits &l'Apadana depuis 1969 paraitra dans le no 2 des Cahiersde la DAFL

184

JOURNAL OF PERSIAN STUDIES

Djafarabad Une troisieme campagne de fouilles s'est daroulkesur le tpd Djaffarabad (Khuzistan) du Iojanvier L'objectif de cette campagne dtait de d6couvrir l'6tablissement aussi largement que possible, au niveau 3, mieux prdserv6que les niveaux suptrieurs I et 2. La fouille fut conduite sur 325 m2 ce qui donne une superficie totale de 425 m2 avec la zone fouillke en 1969.5 Trois 6tats de construction furent reconnus: 3A, 3B, 3C, du haut vers le bas. Seuls 3A et 3B furent ddgag6s. Les briques crues ont un module diff6rent selon qu'elles ont dt6 employdes pour l'616vationdes murs (o,40o 0,20o o,o85 m) ou rdserv6esa la construction de certaines structures domestiques (o,6o x
au 25 f6vrier 1971.

parfois plusieurs lits, plus rarement constitudspar un lit de briques. La couverture des habitations dtait composde de nattes enduites de bitume. Les pi&ces,rdparties autour d'une aire centrale d6couverte, vide de toute structure secondaire, semblent rdpondre a des fonctions diverses: chambres d'habitation pour les plus grandes tandis que d'autres, plus petites, pourraient etre des ateliers ou des magasins. Les structuresdomestiques, " silos ", foyers, fours culinaires, plate-formes sont rdpartiesa l'intirieur des ' chambres ou leur proximit6 immddiate. Un deuxieme grand four (a poterie ?) a dtd retrouv6 au cours de cette campagne. La c6ramique recueillie dans le niveau 3 est semblable a celle des niveaux I et 2; la poterie claire reprdsente6o% de l'ensemble, la poterie claire a d6cor sombre 23%, la poterie commune 2%, la poterie rouge clair a couverte rouge fonc I %. Cette derniere s'apparente a celle recueillie par C. Goff dans le Luristan.6 L'outillage lourd, houes, haches, masses, et les ustensiles domestiques, meules, pilons, molettes en pierre, sont tr s abondants. Fusaioles en terre cuite, " balles de fronde", boules et bouchons de jarre en terre crue, outils en os completent cet assemblage caractdristiquede la pdriode I de Suse (Suse A). Les 6changes avec Suse, distant seulement de 7 km, semblent attestds par la presence de la belle poterie peinte et de la c6ramique rouge brunie, etc.... Mais, comme le montre l'analyse qualitative et quantitative du matiriel archdologique,Djaffarabad offrirait plutSt un aspect rural de cette civilisation.
GENEVIVEDOLLFUS

0o,25X o, Io m). Les sols observes sont le plus souvent en terre battue ou recouverts de nattes formant

GodinTepe Full scale excavations were conducted at Godin Tepe in central western Iran from late May to mid-September 1971 sponsored by the University of Toronto, the Royal Ontario Museum and the Archaeological Service of Iran. The excavations were supported financially by the Royal Ontario Museum, the Babylonian Collection of Yale University, the Ford Foundation and the Canada Council. Efforts concentrated primarily on extending downwards the Deep Sounding, an area of approximately 700 square metres along the north edge of the site. Secondary efforts included: (I) continued clearance of the Median "manor" house of Period II, (2) test trenching on the south and southwest sides of the Citadel mound, and (3) a test trench on the western Outer Town flat in the area of the brick kiln at the foot of the Citadel mound.
Operationsin the Deep Sounding.On an average we were able to push excavations here down another three metres (in some operations more, in others less). By the end of the season we were still involved over most of the area with the very lowest levels of Period III. Some Ioo square metres, however, penetrated into Period IV remains. The Period III deposit was very much thicker along the southern side of the Deep Sounding than along the north (where our original test trench of the 1965 season was located). Therefore we were unable to reach our goal of completing the clearance of the Period III deposit over the whole of the sounding this season. In the past we have been able to speak of "building levels" within Period III. The nature of the stratigraphy encountered this season prevents continued use of such terminology. At no point in the
5 Une description d6taill6e des structures est donn6e dans les Cahiersde la De'lgation arche'ologique franfaise en Iran 1, p. 23-28.
6 Iran IX (i971), 139 et fig. 5: I-I1, I4; fig. 6: 5, 24. 'Cahiers de la DAFI I: 211I.

SURVEY

OF EXCAVATIONS

185

smaller columned hall is the outer wall of the manor house in its final configuration, (3) that the smaller columned hall had been added to the building after the original construction, and (4) that the north wall of the large columned hall had once been much narrower, more regularly constructed and had been marked by arrow slots and pilasters. Excavations at the other end of the building uncovered part of a secondary construction added to the east of the magazines (partially defined in 1969) and a large tower on the southeast corner. This tower must have originally been the dominant feature of the building, for the very thick walls (in places up to 3 metres) indicate an imposing original height. It had been added sometime after the construction of the south bank of magazines discovered in 1969. Small Soundings.Two small soundings on the Citadel mound confirmed evidence from other seasons that Period III material consistently lies immediately below the surface. A small trench at the foot of

Period III deposit cleared this year was there a break in the architectural sequence involving the whole area of the Deep Sounding. In short, we are dealing, over a large area, with a continuous occupation of considerable duration. Breaks in the sequence did appear in segments of the area under clearance, but if the results are considered as a unit one cannot speak of anything other than what we came to call "spiral stratigraphy". The term is appropriate because that is exactly what the sections show: if linked together, and if one works around the operation from section to section, a continuous stratigraphically defined line can be traced downwards from the top to the bottom of the deposit uncovered. For any given moment in time one can describe the buildings occupied at that moment. If, however, you were to try to express that moment in terms of a "building level" for purposes of description, then you would have to speak of literally a dozen or more such "building levels" for a dozen or more moments in time. Furthermore, each "building level" would be only a slight modification in terms of plan on the levels above and below it. Thus it was that much effort in the past season was devoted to understanding this phenomenon and dealing with it in the ground and on paper. As regards the architecture discovered, we continued to uncover fairly complex buildings which, on the whole, cannot be described as monumental. At the west end of the sounding, however, certain walls were defined in the earlier Period III levels which were massive and which may, on further examination, be related to a structure to the west outside the area being excavated which can be called monumental. The principal small finds of the season for Period III consisted of pottery, mostly sherds, but also sometimes whole vessels. We continued to get materials which can be related, in the main, to Giyan III and IV. In the lowest Period III strata, however, certain marked changes began to appear in the ceramic assemblage. The rather tight, precise, animal and geometric motifs heretofore characteristic of this pottery gave way to a much more free-form and looser geometric style which cannot be compared very closely to any published material from Tepe Giyan. It was also observed that certain types of plain wares characteristicof the upper half of the Period III deposit began to disappear in the lower strata of that period. Thus it has now become clear that, when the necessary final analyses are completed, there will be a major ceramic division in Godin III, with the lower phase generally incomparable to anything known from Tepe Giyan. Structures cleared in Period IV consisted of a series of brick and chinehwalls and hearths which probably defined some kind of industrial or workshop area. Certainly they did not represent houses, for the walls were much too thin and flimsy to have supported any weight. The walls may have acted simply as windbreaksfor the numerous hearths. Some distance to the west of this complex was a quite remarkable hearth: a large, carefully constructed rectangular area which had been lovingly plastered (and replastered and reconstructed several times with great care). Moulded into the plaster were basins and troughs which were apparently designed to hold liquids. Low benches lay along two sides of the structure (the north side was destroyed by erosion). The whole was burned and covered with fine ash. Large quantities of ash had accumulated to the east in thick strata which connected this hearth with the complex of flimsy walls and hearths described above. No small finds were recovered revealing the function of this feature. Reluctant as one is to reach such a conclusion, it just might have had some religious significance. Further excavation to the south in 1973 may yield useful evidence. Excavations at the northwest corner of the structure as defined in I969 The PeriodII Structure. two more fortification towers previously suspected, (2) that the western wall of the confirmed: (I)

186

JOURNAL

OF PERSIAN

STUDIES

the Citadel mound to the west in the area of the cut related to the modern brick kiln revealed Period V materials immediately below the surface and Period VI material only about one metre below that. A possible area excavation here in 1973 could permit the clearance of up to 300 square metres of both Period V and VI deposits. Ceramics and small finds in the usual quantities were found in all areas opened. Quantitatively at least the bone and vegetable material recovered was quite satisfactory and we continued to add to our considerable number of well stratified radio-carbon samples. T. CUYLER YOUNG, JR.

ABBREVIATIONS AASOR AfO AJA AJSL AJ AK AMI ANET AOr Arch Anz AS BA Besch BASOR Belleten BGA Bib Or BSA BSOAS CAH CIA El ESA IAE ILN Iranica JA JAOS JEA JHS JNES JRAI JRAS JRCAS KF LAAA MAOG MDAFA MDOG MDP MJ OIC OIP OS PZ RA REI
SAA SAOC Soy Arkh SS Survey TT WO WVDOG ZA ZDMG

Annual of American Schools of Oriental Research Archiv fir Orientforschung American Journal of Archaeology American Journal of Semitic Languages and Literatures Antiquaries' Journal Antike Kunst E. E. Herzfeld, Archaeologische Mitteilungen aus Iran Pritchard, Ancient Near Eastern Texts Archiv Orientatlny Archaologischer Anzeiger Anatolian Studies Bulletin van de Vereeniging... de Antieke Beschaving, Hague Bulletin of American Schools of Oriental Research TUrk Tarih Kurumu: Belleten Bibliotheca Geographorum Arabicorum Bibliotheca Orientalis Annual of the British School at Athens Bulletin of the School of Oriental and African Studies Cambridge Ancient History Corpus Inscriptionum Arabicorum Encyclopaedia of Islam Eurasia Septentrionalis Antiqua E. E. Herzfeld, Iran in the Ancient East (194I) Illustrated London News Iranica Antiqua Journal Asiatique Journal of American Oriental Society Journal of Egyptian Archaeology Journal of Hellenic Studies Journal of Near Eastern Studies Journal of the Royal Anthropological Institute Journal of the Royal Asiatic Society Journal of the Royal Central Asian Society Kleinasiatische Forschungen Annals of Archaeology and Anthropology, Liverpool Mitteilungen der altorientalischen Gesellschaft Memoires de la D61gation Archeologique frangaise en Afghanistan Mitteilungen der deutschen Orientgesellschaft M6moires de la D6lgation en Perse Museum Journal, Philadelphia Oriental Institute, Chicago, Communications Oriental Institute, Chicago, Publications Orientalia Suecana Praehistorische Zeitschrift Revue d'Assyriologie Revue des ltudes Islamiques

Soviet Anthropology and Archaeology Oriental Institute, Studies in Ancient Oriental Civilisation Sovetskaya Arkheologiya Schmidt, H., Heinrich Schliemanns Sammlung trojanischer Altertumer A Survey of Persian Art from Prehistoric Times to the Present, ed. A. U. Pope. Oxford, 1938 Turk Tarih, Arkeologya ve Etnografya Dergisi Die Welt des Orients Wissenschaftliche Veroffentlichungen der Deutschen Orientgesellschaft Zeitschrift ffir Assyriologie Zeitschrift der Deutschen Morgenlaindischen Gesellschaft 187

NOTES ON TRANSLITERATION FOR CONTRIBUTORS TO IRAN

I. OLD AND MIDDLE PERSIAN

It is recognized that no rigid lines can be laid down here, but it is suggested that the Old Persian Texts, Lexicon, syllabary should be transliterated according to the table in Kent, Old Persian.Grammar, p. 12; that for Manichaean Middle Persian and Parthian, the transliteration system given in Andreasvol. III, p. 66, should be used; whilst for Pahlavi, the table of Manichaica, Henning, Mitteliranische alphabets given in Nyberg, A Manual of Pahlavi,new edition, p. 129, may be used as a reference for transcription.
II. ISLAMIC AND MODERN PERSIAN

The system used for the Cambridge Historyof Islamshould be used here as far as possible. Consonants (a) Arabic yb
'
j z

th kh
d

(t'

sh

. J

k 1
m

,
dh
r

d
t

3
.
j 3

n
h Y
W

-a (in constructstate: e gh f -at) , (b) Persian additional and variant forms. The variant forms should generally be used for Iranian
,

names and for Arabic words used in Persian. # p j z

Ss3 zh

g
v

0 . Sch (c) The Persian " silent h " should be transliterated a, e.g. ndma. Vowels Arabic and Persian. Short: " a Long: I or d Doubled form: i) -.. iyy (final u au Diphthongs L; i s"ai Notes I. The izdfa should be represented by -i, or after long vowels, by -yi, e.g. umard-yijdnki. 2. The Arabic definite article should be written as al- or 1-, even before the so-called " sun letters ", e.g. 'Abd al-Malik, Abu 'l-Nasr. 3. The macrons of Abi and Dhfi (Zi) should be omitted before the definite article, e.g. Abu '1- Abbis (but Abfi 'Ubaida). It is obvious that for the rendering of linguistic and dialectical material, and possibly also for contemporary literary and spoken Persian, this rigorous system of transliteration is inappropriate; contributors should use their discretion here.

III.

GENERAL

POINTS

2. Conventional English equivalents (without macrons or diacritics) should be used for the names of countries, provinces or large towns, e.g. Khurasan, Shiraz. Otherwise, all place-names should be rigorously transliterated. Archaeologists are asked to be especially careful in representing the names of little-known places at or near sites. 3. Modern Turkish names and words should be written in the current romanized Turkish orthography. 4. Where classical Greek and Latin renderings of Old and Middle Persian names exist, these familiar forms should be used for preference. 188

I. Names of persons should be rigorously transliterated.

BRITISHINSTITUTEOF PERSIAN STUDIES APPLICATION INDIVIDUAL FOR MEMBERSHIP NAME..................................................................................................... NAME.. ......................................................................... DESCRIP ION T ........................................................................................................................................................... (Title, Decorations,Degrees) ADDRESS .......................................................................................................................................................................
.....................................................................................................................................................................

OCCUPATION

...........................................................................................................................................................

I wish to apply for Membership of the British Institute of Persian Studies as an individualsubscribing member paying*i?1/?4per annum and *enclose cheque/banker's order for this sum. *1 wish to make a covenant for a period of seven years. Signature ............................................................................................................................. D .... ate..........................................................................
* Delete whicheveris inapplicable.

Paymentmay be made in Dollarsor Rialsas follows: U.S. $10 or Rials720. Price of BackNumbers: ?5 ($12 or Rials900). For the convenienceof subscribersin Europe,payment may be made direct to the British Bank of the MiddleEast,20 AbchurchLane,London,EC4N7AY. Pleasespecify Account No. 654-674-01.

APPLICATIONFOR MEMBERSHIP A CORPORATE BODY BY NAME ............................................................................................................................................................................... ADDRESS ....................................................................................................................................................................... Application is hereby made for Membership of the British Institute of Persian Studies as a corporate subscribing member paying the sum of ?50 per annum. A *cheque/banker's order for this sum is enclosed. Signature ............................................................................................................................. Date ..........................................................................................
* Delete whicheveris inapplicable.

189

MEMBERSHIP-SUBSCRIPTIONS Under paragraph i i of the Articles of Association of the Institute, the Council has prescribed the following scale of subscription for members.
i.

Individuals

(a) Those subscribing C4 per annum or more. They will have the right to receive an invitation to the Annual General Meeting and other meetings of the Institute and will receive the Journal of the Institute, which it is hoped to publish annually. (b) Those subscribing ?i per annum. They will have rights as in (a) above, except that they will not receive the Journal.
2. Corporate

bodies

The subscription will be a minimum of ?50 per annum. They will have the right to nominate one representative to attend meetings of the Institute and vote thereat (paragraph 9 of the Articles). They will receive one copy of the Journal. 3. Unincorporated bodies The subscription will be a minimum of ?50 per annum and the privileges will be as for corporate members subject to the provision of paragraphs 7 and 8 of the Articles of Association. 4. Forms should be sent to the Hon. Secretary, J. E. F. Gueritz, Esq., M.A., 85 Queen's Road, Richmond, Surrey.

190

FORM OF COVENANT I .......................................................................................................................................................................................... of ......................................................................................................................................................................................... hereby covenant with the British Institute of PersianStudies that for a period of seven years from the date of this Deed or during my lifetime (whichever shall be the shorter period) I will pay to the said Society from my general fund of taxed income such an amount annuallyas after deduction of income tax at the Standard Rate yields the net s um of .............................................................................................................................................................................. The first paym is to be made on ent the......................................................................................................... day of ........19 .............. ................................................................. ......................................... .......... Signed, sealed and delivered by the said: In the presence of: Dated th is

NOTICE It is hopedthat as manyas possibleof those applying Membership the Institutewill undertake for of a seven-yearcovenant,thus makinga furthersubstantial contributionto the resourcesof the Institute.

BANKER'SORDER To.........BAN BANK.........BAN ..................................................................................................................................................................... ADDRESS.ADDRESS ..................................................................................................................................................................... Please pay to the British Bank of the Middle East, 20 Abchurch Lane, London EC4N 7AY, for the credit of the British Institute of Persian Studies, Account No. 654: now and a similar sum on Ist Januaryof each year 674-01, the sum of ? until further notice/for the next six years.* Signed ................................................................................................................................. Date .........................................................................................
* Delete whicheverdoes not apply.

191

I ".

*-'q250

-.

Ioo400Kilometres

. pt~:
L/SPLAX'

2t~ies

.- !i'-

"~

, , , , , , ,

SZARBAIJAN
. LfR-ezayeh

Tabriz,--

U.
-

S. """ "

S.

R. -

0..organ.,. SZAazv.

'-

"$"

.*/

+? r
<

NDARAN~eshed . *Mz Damghmlt

.-'' ?

TEHRAN
K HU
hah
*

"''*
R A S A N

SHamladL
Ke
0

rmans

SR

7-A \0 ?*"%

.Kashmur

1.Susa,

? Isfahanl4

SI

ST

AN

/ /-..aasargadae Shira
.-"S--

KermaiL
..-a .ersepohis ,Own&"Men
oRM
E

" ."

S"-

SAUDI . ARABIA
JA.. ."i ';" . . " ". " -b;:

Bandar "Abbas

-.-

"-

. :

."

.-

""-" "

"

PRINTED 109

IN

GREAT

KINGSWAY

BRITAIN LONDON

BY WC2

LTD HEADLEY BROTHERS KENT AND ASHFORD