BULLETIN ON SUMERIAN AGRICULTURE Volunle VI Cambridge 1992 General Editors J.N. Postgate Faculty of Oriental Studies University of Cambridge Sidgwick Avenue Cambridge CB3 9DA U.K. CONTENTS M.A. Powell Department of History Northern Illinois University De Kalb Illinois 601 25 U.S.A. Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timber and trees: ancient exploitation in the Middle East: evidence from plant remains G. Willcox Levant trees and tree products N. Liphxhitz . . . . . . . . . . . v 1-31 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33116 Ethnographic evidence for wood, boats, bitumen and reeds in Southern Iraq E. Ochsenschlager . . . . . . . . . . . . . . . . 47-78 LC bois dans l'architecture: premier essd pour une estimation des besoins dans le bassin misopotarnien . . . 79-96 J.-Cl. Margueron Dcndrochronological wood from Anatolia and environs P.I. Kuniholrn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99-122 'Timber production in Presargonic Lag& M.A. Powell . . . . . . . . . . . . . . . . . . . . . . . h e r k u n g zu den neusumerischen Texten iiber Schilfirohr W. Sallaberger 'Rohr' und dessen Verwendungsweisen anhand der neusumerischen Texte aus Urnma FI. Waetzoldt Kced in the Old Babylonian texts from Ur M, Van De Mieroop * . . . . . . . . . . . . . . . . . . . . . 155-161 ISSN 0267-0658 Orders may be placed though bookseuers or direct to the Sumerian Agriculture Group, Faculty of Oriental Studies, Sidgwick Avenue, Cambridge CB3 9DA, U.K. OSumerian Agriculture Group 1992. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without the prior permission of the publishers. Printed in EngIand by Aris & Phillips Ltd., Warminster, Wiltshire . . . . . . . . . . 163-170 . . . . . . . . . . . . . . . . . . . . 171-176 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177-192 Wood, reed and rushes: a note on Neo-Babylonian practical texts G. van Driel Notes on the reeds of Mesopotamia F.N. Hepper . . . . . . . . . . . . . 125-146 147-153 . . . . . . . . . . . . . . . . . . Trees and timber in the Assyrian texts J.N. Postgate 123-124 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wood in the Old Babylonian texts from Southern Babylonia M. Van De Mieroop Lc bois & Mari J.-R. Kupper 97-98 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193-194 . . . . . . . . . . . . . . . . . . . . . . . . . . 195-196 ............................... 197-199 A corrective note on Pistaciatrees and resin F.N. Hepper Some wood identifications from Mesopotamian sites P.R.S. Moorey & J.N. Postgate Illustrations of trees by F.N. ~ k ~ ~ e r Pinus halepensis @. iv) - Populus, Buxus, Platanus (p. vi) - Quercus spp. (p. 32) - Juniperus, Tamarix @. 154) - Abies, Cupressus, Cedrus @. 162) List of contributors and addresses . . . . . . . . . . . . . . . . . . . . . . . . iii 200 PREFACE 'Jc lulnl. h of thc Bulletin on Stmerinn Agriculture contains papers from the meeting of thc ,4gr-:culltart Group held an the sympathetic environment of the Inter~~ationales W~xscnccllallslir:um of !he Univenity of Heidelberg in July 1989. Our thxlks go to Prof. Dr. i l ; ~ ~ r n r Wactzoldt t~r for organizing the meeting, and to the Deutsche Forschungsgemeinschaft and 11ri- Ministry of Ck~ltureV J I ' I ~ C generously ~ defrayed the costs of the meeting. SIIII~ r; tan Some of those at Heidelberg were not presenting formal papax, but taking part in the cIi\cussrori, nnd some of bhe papers given were not destined for inclusion in this volume. 1';trlicipmts in h e meeting not represented here in print are: Dr. A. Cavigneaux, Prof. Dr. W. t:rcy, Dr. J.-P. GrCgoire, Dr. B. HruSka, Prof. Dr. J. Renger, Prof. M. Stol, Dr. G. Szab6, and last but not least Prot', Dr. W. van Zeist. Prof. J.-CI. Margueron, though unfortunately prevented from ntlcnding in person, was able to send his paper. In July 1990 the Group met in Barcelona to discuss sheep and goats, at the invitation of Prof. G. del Olmo Lete of the Institute Oriental. The papers from that meeting will form the next volurne(s) of the Bulletin, now in preparation. The time and place of the subsequent meeting, which will probably deal with cattle, have yet to be decided. The Editors would like to take this opportunity to stress that they are always ready to consider for publication contributions on agriculture in Mesopotamia which do not derive directly from a meeting of the Group, especially those concerned with the theme of forthcoming volumes. As in previous volumes, bold type is used for Sumerian, italics for Akkadian words. Bibliographical conventions generally follow the usages of the individual authors, but Harvard system references are used for books and articles, whereas traditional Assyriological abbreviations are preferred for text editions. These can be tracked down in one of the following works: R. Borger, Handbuch der Keilschrifliteratur, W. von Soden, Akkadisches Hanmuiirterbuch, the Chicago Assyrian Dictionary. For Ur 111 texts see the works cited by Waetzoldt on p. 146. Once again, the format and typography of this volume have benefited from the support of the Literary and Linguistic Computing Centre in Cambridge: our thanks are due equally to Beatrix Bown, John Dawson and Rosemary Rodd for their help and readiness to solve our problems. We are also grateful to Kirstie Shield (Girton College) for her assistance with the type-setting and to Trinity College, Cambridge, for financial support towards this. Pinus halepensis Nicholas Postgate Marvin Powell January 1992 TIMBER AND TREES: ANCIENT EXPLOITATION IN THE MIDDLE EAST: EVIDENCE FROM PLANT REMAINS G. Willcox (C.rV.R.S., E.RA. 17, JaEs, France) Populus euphratica Plataiius orientalis Introduction The exploitation of trees and shrubs (ligneous species) has been fundamental to man's development and dates back into the Palaeolithic to the invection of fire. This natural resource, which makes up an important part of the vegetation cover in many regions of the world, has, and is being, over-exploited by man. The study of the exploitation of ligneous species through analysis of charcoal remains from archaeological sites has relevance beyond historical reconstruction, because it can throw light on the progressive destruction of forest cover and the resulting effects on the natural environment. The history of vegetation in the Middle East, however, would be very poorly understood were it not for the survival of undisturbed stratified pollen in lake beds. Pollen analysis carried out principally by van Zeist and his team at Groningen has provided evidence revealing the sequence of vegetational changes which have occurred since the last glaciation, when severe steppe conditions existed (van Zeist & Bottema 1982, 1977). Most of the vegetation cover that we see in the Middle East today is degraded, but we may find relict associations in inaccessible and isolated regions where exploitation would be impractical (Davis 1971, Zohary 1973, Guest 1966). They are what remain of the original climax vegetation. The examination of these relict associations has greatly enhanced our knowledge of past vegetation and of past potential timber resources. The cedars of Lebanon, now almost completely destroyed, are an example. Charcoal is omnipresent on archaeological sites and is readily recovered by flotation (Williams 1973, Pearsall 1989). This charcoal (almost pure carbon, which is inert and does not decompose) preserves perfectly its anatomical structure and so may be readily identified when examined under the microscope with the help of reference material and the standard manuals (see Greguss 1955 & 1959, Schweingruber 1978). It is not uncommon on archaeological sites to find charcoal belonging to species which have long since been wiped out. This is particularly true of areas where intense settlement has occurred continuously over a long time span. With four lines of direct evidence, pollen from lakes, relict vegetation assemblages, seeds and charcoal from archaeological sites, one can begin to reconstruct the ancient vegetation and thus the environment of ancient civilisations. Charcoal from archaeological sites is of particular interest to the archaeologist because it gives direct evidence of the use of tree and shrub species in the immediate area of the site, although there is also some evidence for long distance transportation. On the other hand, for the botanist who wishes to reconstruct the vegetation, there is always the risk that ancient man was selective in his exploitation, and this must be taken into account. In this paper I have used scientific names of plants to avoid confusion, but readers who wish to refer to common names may consult the list (see below) where English common names are given. A word of warning: names in English for timber or lumber are not always the same as for Timber and Trees the tree and often include several different species. A good example is deal, which wuld br any close-grained Gymnosperm. Ebony is a very dark wood that may come from several different species. Teak and mahogany may also refer to more than one timber species. Rosewoods are named after their fragrance, and are not members of the plant family Rosaceae. Phytogeographical Regions These regions are made up of distinct associations of plants and have been described by Zohary (1973) and Guest (1966). The following is a summary with reference to some of the major tree species which could have beer. exploited in antiquity for timber. Mediterranean This region concerns the coastal k l t of Turkey and the Levant where a number of imponant timber trees occur, the most important being Cedru.~libani, Quercur s r ~ pFr~.xinus ~~ spp., Pinus spp., Platanus orientalis, Abies cilicica, Cupressus sempervirem. Irano-Turaniun This vast area stretching from Turkey to China (including nonhem Iraq) contains an extremely varied Rora and a large number of woody species. However, in general this region is characterised by low rainfall and extreme temperature variation, which docs not lead to the production of good timber in the modern sense of the term, though one could consider Juniperus spp. Quercus spp. Jugkrss regia, Ulmus spp., Pistaciaz atlantica, Pinus brutio, and Tuxus haccata, as some of the possible sources of timber. These species are f~imndin the Zagros. and in the Taurus together with some Mediterranean species. Two other areas within this phytogeographical region should be considered as potential timber producers because of their special climatic conditions. These are the Euxine region where Fagus orientalis, Abies cilicica, Picea orientalis, Coryllls spp., Castanea sativa and Carpinus orientalis are to be found, and the Caspian region, where such species as Farrotiu persicta, Mnrus nigra, Z e l b v a sp. occur. These areas have been exploited in more recent times, but at presmt there is no evidence of their being exploited in the second or third millennium B.C. Saltaro-Sindion This forms a belt of sub-tropical desert steppe vegetation m i n g from west Africa, Egypt, Ardbia (including southern Iraq) to northwest India. This region is chamcicrised by high temperatures and extreme aridity. Evaporation greatly exceeds precipitation, which rnakes it a harsh environment for plant growth. The trees mentioned below can form open forest savanna relying on ground water and deep rooting systems. The northern Ximit of these trees would appear to be defined by winter temperatures and most of the endemic trees in this region are frost-sensitive. Common trees in this region are Acacia spp. Phoenix dactylifiera, Zizjphus spina-christi, Prosopis cineraria, Salvadora persica and Tanarix spp. These trees may reach a considerable height under favourable conditions, that is to say 5-10 meters. This region includes the southern part of Iraq, but little of the original vegetation remains. Sudano-Deccaniun This region forms a belt of tropical vegcrstion running parallel but to the south of the Saharo-Sindian region; unlike the former it has a high diversity of species. The trees which most concern us here are the following. Dalbergin spp.. Bosweifia spp. Ficus spp., Avicennia marina, Acacia spp., Maerua sp., Moringa sp. r~ortl~crn elevated areas. At high altitudes where temperatures are relatively lower and mr~il;ill I ~ ~ y l ispecics ~ r , such as Betula verrucosa and Pinus brutia are to be found (Guest 1966, 85). whlle 111 111clowlands one finds Phoenix dactylifera, Zizyphus spina-christi and Acacia iraqensis. The . I ~ I I N I I~mits ~ ~ of many species define the present-day natural distributions. However, past climatic c orrirlrons, even if only marginally different, would give rise to a different zonation and different ;~lrirudclimits. IIIC ~(~c~r.c\t evolution during the holocene in the Near and Middle East Pollen evidence from 1.1he scdi~nentsindicates that during the latter part of the Pleistocene climatic conditions in the Mldtllc East were such that tree species survived only in rtfuge areas and much of the territory w .I, dry sleppe, probably rather like that which occurs on the h~gll! !iilln plateau at present. In gcrlcral, lakes at the beginning of the Holocene in the Middle East appeir to have k e n more cxtcr~siveand to 'have subsequently shrunk or dried up. Thus while the broad vegetational history II:I\ emerged, the climatic reasons for these changes in vegetation are not fully understood. Major refuge areas wherc deciduous forest species could survive the adverse climatic conditions during the late Pleistocene were probably restricted to the coastal areas of the Black Sca, the Mediterranean and the Caspian (van Zeist & Bottema 1982). In addition it is possible 11l;lt thcre were minor refuge areas where particular micro-dimates favoured tree growth. There is wrnc evidence that the vegetation zones migrated southwards but this has still to be confirmed. Some ten to fourteen thousand years ago a climatic change occured which favoured the expansion of tree species in Turkey. Iraq, Syria and Iran (van Zeist & Bottema 1982). Evidence lrom carbonised plant remains from early Neolithic archaeological sites suggest that t ~ species e were more widespread in Syria (see below). Similar changes occured in Africa and north America (Lkrine 1989). In the Middle East this amelioration apparently coincided with the transition from hunlcr gathering to farming communities. The expansion of tree species was gradual and reached a maximum some 6000-4000 B.P. It was not simultaneous in all areas because of two factors, the d~stanceof sites from refuge areas and the differing rates of co!onisation. Thus the expansion of forest species was later at Lake Van than at Lake Ghab (van Zeist & Bottema 1982). From our point of riew we should note that the birth of the city states in Mesopotamia coincides with the maximum forest cover in the region and timber resources would have been more available then than at any time since. After this period of maximum expansion of the forest species there is some evidence of decLinc, hut this is variable from area Lo area and it is not clear whether this was due to climatic factors or to deforestation by man, through over-exploitation. When one examines the vegetation of the Middle East today, there is little doubt that much is highly degraded. But it is not easy to establish when this deterioration began. Indeed it is probable that the degradation of the vegetation occured at different times in different places depending on population pressure. Evidence from charcoal at A ~ v a nin eastern Anatolia and Bosra in southern Syria would seem to indicate that deforestation was a late phenomenon. Some man-made factors affecting forest deterioration Forest zonation in Western and Central Asia Within eat.h phytogcognplural i~gi(.r)w r may observe different pliurt formatiorls resulting from differen[ c ~ iif~t1u.11 * : ! i d i , ~ l i t f i i i i ~ ~ri >i ,i i ~ d ixing , B the mcxt important. 11.1 Iraq for caample, there is a markc,! '3 7 : ? : :. hc 7q,?~~-t ftnC ++:d,;~!,, ar;J :b5 The role of fire m i l e wme forests depend upon fire for germination and healthy gmwth is ft;r ~ x ~ i m in p [ Austr,dia) ~ then. i l no cvidirncr lor i b ~ cin ih; Zvliddle East. On the contrary, one i t as h i n r datriwi.nt$ As man': population inazai.rd so must l l risk ~ of forest flrc. it:! o,ljv Timber and Ttew Timber and Trees In many areas today this risk has been reduced by drastic thinning of the forests so that the self perpetuating forest fire could not develop. One can imagine that in the past the frequency of forest fires would have been greater than at present. The role of grazers and browsers Man's herds, particularly sheep and goats, prevent forest regeneration because they browse on the young shoots of newly germinated seedlings. While this was also the case for wild ungulates, the fact that man became sedentary put increased pressure on particular areas, lessening the survival of seedlings, Soil erosion Removal of the ground cover by fire or grazing increases soil erosion. Added to this, the felling of trees would also leave the soil more susceptible to erosion. Types of forest exploitation Dry (dead) wood gathered as fuel Freshly cut wood from a live tree or shrub is difficult to bum because it contains high levels of moisture. The gathering of dead wood is prefened if no stocks have been made to allow wood to season. This method requires a large catchment area and is therefore best suited to hunterlgatheree and nomadic peoples. It has the great advantage of not destroying the vegetation. Timber for any purpose is best cut during the dormant period before the sap has risen (dry season or winter depending on the region). Rough timber for construction is not necessarily seasoned but when cut green it will undergo considerable shrinkage. Seasoning is essential for cabinet making. Pollizrding This is the cutting of higher branches which are then allowed to regenerate for a period of several years before cutting again This method is not often seen in the Middle East; it is however common in Europe and nonhwest India where Prosopis cineraria and Acacia spp. trees are cut in this way, the leaves being used as fodder while branches are kept for fire wood. Salk spp. may also be cut in this way to obtain young shoots for basket-making. Coppicing Trees are cut at ground level and the stocks are encouraged to mruub usually a feu trees are left as standards to encourage the upward growth of the young saplings. This method is less suitable in semi-arid environments and is only effective for fiawood production from Corylus, Carpinus and Betulus. Uprooting More common than one might expect, especially where tools are elementary and trees are scarce. Uprooting is also carried out when there is clearing for agricultural land. Felling The felling of large trees for timber could be accomplished using polished stone axes and does not require sophisticated tools. The division of large trunks was probably carried out using a splitting technique with wedges. Seasoning of timber With the development of settled village life it is probable that the systematic felling of timber and its subsequent storage and seasoning was adopted. Wood is usable as fuel after one year of drying; when green it smokes excessively and bums poorly. Where high temperatures were required this period may have been extended. Specialised use of timber Architecture The principal timbers in buildings made from mudbrick are the roof beams which support the flat moves (Aurenche 1981). The span is limited by the dimensions of the timber available and by its strength. For example poplar and oak beams cannot satisfactorily span much more than 5 meters without some kind of other support either from a truss or from a vertical pillar. Populus beams are the most common in areas where this tree is available. They frequently occur preserved in the carbonised state after a conflagration, where burning timbers are covered by earth from the roof structure. In other cases Quercus. Pistacia, and Celtis, have also been identified in this context (see Willcox 1990a, 1991akb; van Zeist 1984). In southern Arabia it is not uncommon to find date palm trunks used as roof beams. At Pella in Jordan during the Byzantine period date was used in roof construction, (Willcox 1983, unpublished report). Furniture Fine-grained woods which are resistant to insect attack are the most suitable and include a large number of species. Boat building Modem boat builders are extremely selective about their timber and rely mainly on imported tropical woods. Coniferous woods for example are not considered good enough. However this may not have been the case in the past. In considering wood from the Gulf, Theophrastes in Hist. IV, VII 7-8 mentions boat-building wood from this area, interpreted by Bretzel (1903), as being wood from the mangrove tree which grew on the coast of Bahrain. This would seem be to be an unlikely timber tree. Could it be possible that he was refering to Acacia? Trees and shrubs as sources of fuel Wood was undoubtedly the major fuel in the Middle East until the twentieth century, and the vast majority of charcoal recovered from archaeological sites represents wood used as fuel. Most woody species can be used for fuel. Some are preferable because they burn slowly giving out a constant heat. These are the dense woods such as Quercus spp. Others bum quickly giving out an intense heat over a short period of time such as Populus spp., Fraxinus spp. and Acer spp. For domestic use such as cooking and heating, the most readily available wood is used and this probably means the nearest to the habitation, regardless of how it burns or its thorniness. While timber for construction may have been the object of trade over relatively long distances, this was probably not the case for fuel. Small branches are frequently used for fuel whereas large sections are used for timber. This difference is often visible among the charcoal remains. The presence of small twigs would appear to indicate that a species is local and is proposed as a criterion for evidence that the t m g a w in the neighbourhood of the site since it would be impratical to transport anythmg except large-section branches or trunks. Charcoal manufacture has not to my knowledge been identified from the Near East. Large quantities of wood are necessary, and the finished product is easily transported over long distances, being extremely light. Today one sees it used in the urban situation or for smelting where high temperatures are required. Possible alternative combustible materials include the following-: Timber and Trcca Timber and Trees Dungcake fuel The use of dungcake fuel has been described by Miller (1984). In general the use of this material can be seen today in traditional societies from Turkey to Mia. Its utilisation may occur in areas where there is abundant timber. To make dungcake fuel cow dung is mixed with chaff or chopped straw and left to dry, often by sticking the cakes onto a vertical south-facing wall. The use of dung cake has been tentatively identified at Bosra (Willcox in prep) and at Malyan (Miller 1985). Straw Straw is sometimes seen to be used for ephemeral fires. It can be used for starting a fire or even for cooking flat Bedouin bread. Its continual use should theoretically leave a residue of silica phytoliths. Chaff This consists of the glumes, awns, rachis fragments, and chopped straw, left after threshing and winnowing ( H i b a n 1984). As mentioned above it is frequently mixed with mw dung to make dung-cake fuel. As a combustible by itself it tends to smoke excessively. Annuals When nothing else is available the dry stems of wild annuals are sometimes collected in the desert regions of the Middle East and used in domestic heanhs when the fire is required for a short period such as for heating water or mi&. Stems of cultivars such as cotton or sesame may also be used. Reeds Phragmites australis and Arundo dona both rorm stands of dense reeds in moist habitats. They have a multitude of uses, fuel being only one; I have observed the latter collected for firing brick kilns in southern Afghanistan where timber is extremely scanx. (* possible or potential timber trees) Long distance transportation of timber High quality timber could have been imported over large distances and there is indeed both archaeobotanical and epigraphid evidence for the importation of timber by the historical period and possible evidence for the prehistoric period (see below). Below is a list of potential timber trees from western Asia, including some species fmm the hdus and northern Oman, but not including shrubs which would have been used for fuel and would have been gathered on a more local scale (see above). Taxacaceae Taxus baccata List of trees and shrubs native to Iraq which could have been used for fuel Mountain forest (Zagros foothills) Quercus iMectoria* Q. libani* Q. aegilops* Acer cinerascens Crataegus azarolus* C . monogyna Pyrus syriaca* Prunus mahaleb P. amygdalus* Cotoneaster numrnularia Amelanchier integrifolia Cornus australis* Rhamnus spp. Paliurus spina-christi Ulmus sp.* Celtis tournefortii Ceris siliquastrurn* Betula verrucosa Pinus halepensis* Juniperus oxycedrus J. polycarpos* GaUery forest (TigrisEuphrates) Salix spp.* Populus euphratica* Fraxinus syriaca* Platanus orientalis* Juglans regia* Vitex agnus-castus Tamarix aphylla* T. pentandra Desert steppe Acacia gerrardii* Hakoxylon spp. Zizyphus nurnmularia 2.spina-christi* Prosopis farcta Nitraria retusa Calligonum comosum Steppe Pistacia utlanrica* P. khinjuk Prunus micmiarpa Y.arabica Ephedra spp. Cypressaceae Cupressus sempervirens Juniperus oxycedrus J . excelsa J . druuacea J . polycarpos Pinaceae Cedrus libani Picea orientalis Abies cilicica Yew. Asia Minor. This is a compact, durable wood, sought after for bows and for cabinet-making. Present-day distribution may be much reduced. Cypress. Eastern Mediterranean. Can grow very tall, specific timber uses not known. Juniper. Western Asia. Possible use as timber. Juniper. Asia Minor and the Levant. This tree can grow to a considerable size, and some authorities suggest it could have been the biblical cedar. Juniper. Asia Minor and north Syria. Has edible fruits. Juniper. Oman, Iraq (rare), southern Afghanistan, Pakistan. Similar to J. excelsa but with a more easterly and southerly distribution. Cedar of Lebanon. Southern Asia Minor and the Levant. Normally thought of as the main timber tree of Western Asia, though rarely confirmed by identification. Now drastically reduced in terms of its former distribution in Lebanon. Spruce. N.E. Asia Minor. A conifer more usually associated with cooler climates. Produces high-quality timber. Fir. Taurus. Identified from Byblos and Bosra. Suggests that timber was being exploited in Asia Minor at an early date. Timber and Trees Azadirachta indiea Neem. hnjab/Indus. Large tree, used for timber. Tamaricaceae Tamarix aphylln, T. articulata Tamarisk. Bolh these trees grow in the more southerly part of Western Asia and grow to considerable height. They are frequently cut for timber, though the quality is poor. A vicenniuceae Avicennia marina Salvadoraceae Salvudora persica Palmaceae Phoenix dactylifera Mangrove. Gulf, Bahrein and Oman. A tree occuring in muddy salt marshes and tidal creeks. Bark used for tanning. Other uses not known. Tooth brush tree. Oman, Baluchistan and southern Iran. Used for the fabrication of tooth brushes. Identified from Mundigak outside its present-day distribution (Casal 1961). Date. Southern Iraq (see Guest 1966). Trunks of date palm are frequently used for roofing beams. Identified from Pella in the Jordan valley. It is probable that southern Iraq lacked quality timber during the Sumerian period and for this reason timber had to be either cultivated or imported from elsewhere. Timber from the Taurus and Lebanon is the most obvious source but other areas should be considered. For example the Zagros mountains or the Caspian area could have supplied large quantities of high-quality timber. Timber trees available in the Zagros mountain zone have already been listed and there is little doubt that this area was the nearest region possessing timber of reasonable quality Timber from the south, that is to say, southern Arabia, is a possibility. For example the Omani mountains may have been a source of timber. Here Acacia spp., Maerua crassijliu, Prosopis cineraria, Zizyphus spinashristi, Tamarix aphylla, Juniperus polycarpos, Olea sp., and Teucella sp. would be candidates. In southern Oman and the Yemen we move into another vegetational zone with a great variety of small trees and shrubs, the majority of which do not produce good timber; the only species worthy of mention is Euclea schimperi, which has dad; hard red wood much prized for making small objects (Miller & Moms 1988). In Sind, that is the Indus region in the drier desert areas, the same group of trees are available which we find in northern Oman. In the areas further north we meet yet another zone of vegetation leading up into the Himalayan foothills which is beyond the scope of this survey. However, in the northern part of Sind there are several timber species mentioned in the list above. Methods of charcoal analysis Presence of charcoal. Charcoal occurs in archaeological sediments either associated with hearths, conflagrations or in rubbish pits. It results from incomplete combustion where the supply of oxygen has been cut off during burning. The most common reasons for this are: when a fire is stifled by its own ash, when a fire is intentionally extinguished, or in the case of a burnt building, when burning beams are covered by earth falling from the roof which suffocates the fire. Hearths are by far the most common source of charcoal and are readily sampled. It is here that we find concentrations of charcoal. Frequently charcoal also occurs as dispelsed fragments in floor deposits, fill, destruction and even in mudbrick deposits. When charcoal is found thinly diffused throughout a sediment such as mudbrick there is a possibility of it being residual and it may be of earlier date than the layer in which it occurs. It is therefore of less value. Methods of recovery Flotation is the most suitable method for the recovery of all carbonised remains. Where there is evidence of in situ burning the archaeological unit should be subjected to flotation. It should perhaps be pointed out that sediments in the Middle East react well to simple flotation techniques because of the lack of humic and clay colloids, which in Europe tend to cause the sediment to cling to the carbon thus preventing it from floating. This is not the case in semi-arid areas. Flotation is preferable to hand sampling for the following reasons: 1) seed remains invisible to the naked eye often occur in the same deposits and so should be sampled at the same time. 2) some species of wood tend to break up more easily and with hand sampling excavators tend to introduce a bias by only collecting the larger fragments. 3) flotation separates the sediment from the charcoal with a minimum of mechanical s a s s . The most suitable technique for flotation and one which I have used on many sites in the Middle East is decribed by Williams (1973). Charcoal which does not float is caught in the wet sieve (3mm diameter). After flotation, the charcoal and other carbonised material such as grain should be slowly dried. Rapid drying causes the charcoal to fracture. Identification Identification is carried out using a reflected light microscope with dark field. The charcoal is fractured in the following planes, cross section, tangential section and longitudinal section, in order to expose the diagnostic features. In the Middle East identification is normally made to genus level only; however in Europe specific identifications are possible because of the reduced number of species (Vernet & ThiCbault 1987). While several publications may aid identification there is no substitute for modem reference material. This is collected during field excunions and is then prepared by heating to three to four hundred degrees centigrade and cutting off the oxygen supply by wrapping in aluminium foil leaving a small hole for the volatile gases to escape. After three or four hours one is left with nearly pure charcoal. Quantitative analysis On Near Eastern sites recovery is often only partial and the quantity of charcoal which survives is the result of the fortuitous circumstances of the carbonisation process. Neither the volume, weight, or number of fragments necessarily relate to the quantities burnt in the original fire, or the available timber. Thus to avoid bias, presencelabsence analyses are well adapted to the data when the percentage of archaeological units as a total of all units sampled with a given species is used as a measure of abundance. With this method it is necessary to have a minimum number of samples from each period or site before a quantitative estimate can be made on the availability of timber by comparing one site (or phase) with another. When the minimum number of samples is not available, phases or periods must be grouped together as in the case of Can Hasan (Willcox 1991b) where several phases were treated as a single period. A Willwx Timber and Trees Timber ad lLrr Willwx sample is equal to the sum total of charcoal from any one archaeological unit. This may vary from a few small fragments to several litres. To compare different periods or sites there must be a minimum of ten, preferably more, units sampled from each period. In European cave sites where preservation, recovery techniques and excavation offer more complete data, researchers have used volume, number of fragments, and weight in order to make a quantitative assessment (Chabal 1990, Vernet 1990, Heinz 1990). Their results show that these methods are valid for these particular sites and recovery methods. Similar methods were also used at Malyan in Iran with convincing results (Miller 1985). Charcoal data from selected sites in Western and Central Asia In this section I have attempted to amass as much information as possible on charcoal identification from the Near and Middle East including much of my own unpublished material. In some cases a list of identifications has been given, in other cases the bibliographical reference. Each site is listed by country with a note on location and archaeological context. Presence analysis was used as a measure of frequency from each phase or site. Thus presence within a sample (= archaeological unit), regardless of quantity, is the criterion. Where sample numbers are less than 5, percentages have not been calculated. + ++ +++ ++++ * = = = = = present in 1-25% of samples present ** in 25-50% of samples " 50-75%" " '* w 75-100 w v presence where the sample size is <5. With small sample size the quantitative estimates become dubious. In the tables Prunus refers to Prunus and/or Amygdalus TURKEY Afvan Central eastern Anatolia 40 kms N.W. of Elazlg. Multi-period, Chalcolithic to Islamic. Samples were taken from four sites within a radius of five kilometers. Excavated by D. French, British Institute of Archaeology at Ankara (Willcox 1974). Site location map with shaded area giving approximate distribution of Mediterranean and Irano-Turanian forest associations, excluding Saharo-Sindian species such as Zizyphus spina-christi, Acacia spp, and Prosopis cineraria. (see next page) Wiwx Timber and Trees Acer Alnus Crataegus Elaeagnus Chalco E.B.A. + + + + + Ficus Fraxinus Helmom Ear Med Late Med + + Timk Willwx The majority of samples came from features which were not identified owing to the small size of the trench. But it must be presumed that most came from hearths and thus represent the remains of fuel. The more steppic species such as Pistacia and Amygdalus are common, while Quercus is + ++ + + + Juglans Junipems less well represented. One possible interpretation is that Quercus had only just begun to colonise the area. C d e r Htiyzlk Central eastern Anatolia 40 kms east of Malatya. Excavated by J. Cauvin, C.N.R.S. Aceramic Neolithic. (Willcox 1991b). Celtis Fraxinus Pistacia Lycium Paliurus Pinus . Quercus Salicaceae Pistacia Platanus N= Prunus This long sequence from nearly three hundred samples (many of which had to be sub-sampled because of the large quantity of material) illustrates how systematic sampling techniques can reveal a vegetational history. Here there is good evidence of progressive deforestation in the area of the four sites. Exploitation of forest climax species in the early phases gives way to progressively more use of spiny and riverine species in the later phases when cultivated species such as Pinus also appear. Today the area is almost totally devoid of Quercus. Can Hasan Southern Anatolia 20 kms N.E. of Kararnan. Aceramic Neolithic. Excavated by D.French, British Institute of Archaeology at Ankara (Widcox 1991b). hygdalus Celtis +++ Crataegus Juniperus Ulmus + ++ + ++ + + ++ + N= 97 Quercus Rosa Salicaceae +++ 84 Caytinu Eastern Anatolia 75 kilometers northwest of Diyarbekir. Aceramic Neolithic, 8th millennium B.C. (van Zeist 1972). gdalus Fraxinus Pistacia + + + ++ The high levels of Salicaceae from this site may be due to burnt building timbers. However large quantities were also found in hearths, where this wood appears to have been the preferred fuel. The relatively lower levels of Quercus do not necessarily suggest that this tree was less common in the area. The important point is that the environment of the site contained the climax forest association. Quera Rhamnus Rubus Salicaceae Tamarix Ulmaceae indeterminate Pinus nd' h N +-I Pistacia Quercus Tamarix N= +++ +++ +++ ++++ +++ 23 (% based on three phases) Here we see confirmation that early farmers in Anatolia settled in areas which were already forested though these forests were open. The three early Neolithic sites, Can Hasan, Cafer H(iyuk, and Caycinu corroborate this. Other finds from Turkey 1st and 2nd millenium sites such as Gordion, Kiiltepe and Acemhuyuk produced large quantities of carbonised timber, now undergoing dendrochronological analysis by P.Kuniholm. The impressive quantities of timber used in the architecture indicate that there was no lack of timber for this period in Anatolia (Kuniholm, pers.comm.). Buxus sempervirens * Cedrus sp * Juniperus spp P. sylvestris * * Pinus nigra * Taxus baccata * Taxus baccata and Buxus sempervirens may have come from furniture. Timber and Trees Timber and Trees This site is situated in the wooded hills of the Jebel Druze. Again we see evidence of deciduous SYRIA Bosra Southern Syria. Roman to Islamic. J-M. Dentzer, Mission ArchBologique en Syrie du Sud. CNRS (Dentzer 1985, Willcox, in preparation). Qwrcur, now extinct in the area. The presence of Pinus occurring in the context of a humble dwclling suggests that it was endemic. Bouqras Islamic Byzantine Roman + Capparis cf. Cedrus Crataegus + Olea Pinus Pistacia + ++ Quercus (evergreen) Rhus Salicaceae ++++ +++ ++ + N= 19 by J. Cauvin. (van Zeist 1984, Willcox unpub.). +tt + Quercus Tamarix ++ + ++++ + + 4-1- N= 25 (hand samples collected during excavation) + + + Vitis vinifera Notthern Syria, Euphrates valley. Aceramic Neolithic 9th millennium B.C. Excavated Early Neolithic Pmnus Pyrus Quercus (deciduous) see van Zeist 1989. Sabi Abyad Mureybet ++ + + + see van Zeist 1985. Chenopod cf. Haloxylon Fraxinus cf. syriaca Populus (euphratica) The presence of Quercus is of interest but the low frequency could represent an exceptional hpon or even wood which drifted down the Euphrates. On the other hand the seed remains (Plstacia and Ficus) unless also imported, suggest a vegetation much richer than that of today. 6 16 Abu Hureyra Deciduous Quercus does not grow in the region today and thus must have become extinct, perhaps through over-exploitation. The presence of Pinus is difficult to interpret since at present it is not possible to establish whether it was part of the natural vegetation which became extinct or whether it represents an introduction. The poor number of samples from the Byzantine levels make a true evaluation difficult. Quercus may have been gathered from the hills some 20 kms to the northeast which to this day are still wooded. One the other hand the presence of scattered Crataegus, Pistacia and Pyrus trees on the plain around Bosra suggest that the area did indeed support more trees in the past. Sia Southern Syria, Jebel al-Arab. Roman. Excavated by F. Villeneuve (Institut Fmsais de 1'ArchBologie du Proche Orient). (Willcox, in preparation) Roman Vitis vinifera + + + + + + ++++ + + N= 47 Crataegus Olea Pinus Pistacia Pyrus Quercus (deciduous) Quercus (evergreen) Rhamnus Northern Syria, Euphrates Valley. Epipalaeolithic and Aceramic Neolithic. @illman et a1.1989). Epipalaeolithic Acer Populus Salk Tamarix * * * * This site produced a large assemblage of plants with extensive sampling and the meticulous work glrrled out by G. Hillman and his team at the Institute of Archaeology (London). Only a pvisiond analysis of the charcoal remains has been carried out, but in light of the finds from Mumybet, further analysis would seem justified. All those timber species so far identified come h e riverine habitat, which in this case would be the Euphrates. Seed remains suggest a Wdcr range of tree and shrub species (see below). r#ll Seh Hamad see Frey et al. 1991. Timber and Trees Willcox Pella 1983). JORDAN Jordan Valley. Multi-period. Excavated by tne University of Sydney, Australia (Willcox Kebaran Natuf Chalco Bronze * * Amygdalus Celtis * * + Crataegus Ficus Nerium Olea * + Iron Hell/Rom Byzan + + * + ++ ++ * + + + + ++ + * * * Quercus Rhamnus +++ + * As with a number of other sites, both the gallery forest and the climax forest, which today exists only in relict form, were exploited at the same time. This indicates that the climax species were more available in the past, than at present, but lack of continuity does not permit an exact date for the deforestation. Jawa 200 km N.E. of Amman. Chalcolithic. Excavated by S. Helms (Willcox 1981). Acer + a + Phoenix Phragmites Pinus Pistacia Umayd Mamluk Timber nd Tnu Willcox Amygdalus Chenopod + + +++ + + Crataegus Pistacia Quercus decid Quercus evergreen N= * * * Salicaceae Tamarix Thymelea Zizyphus Quercus is present in the form of both deciduous and evergreen (Q. calliprinus) species, the latter being present only in small quantities. It is possible this this wood was imported from the Jebel Druze, which is 75 kilometers due west and where these species are found today (see under Sia). It would be difficult to imagine these species growing in the vicinity of Jawa without proposing a substantial climatic change. In the absence of further evidence it would appear best to reserve judgement on the origin of these finds. Dharih Southern Jordan, Roman. Excavated by F. l'Arch&logie du Proche Orient). (Willcox unpublished). The site of Pella, situated on the eastern slopes of the Jordan valley, is even today within reach of timber from higher elevations where there is a mixture of Mediterranean and Irano-Turanian species, while in the valley bottom there are species typical of the Sudano-Sindian vegetation. Juniperus Olea The samples from the Kebaran and Natufian levels indicate that full climax vegetation was established at the site or at least within the catchment area at this relatively early date. The Lisan lake, of which the Dead Sea is now the much reduced remainder of this large body of water, may have locally affected the climate. It could have favoured forest species during the last Ice Age when large areas of the Middle East were steppe (see above). Populus Tamarix N= The Natufian levels produced only Quercus. During the Bronze Age again we see the forest climax dominants being exploited, but in the latter periods less desirable species were used. Phoenix was used in roof construction during the Byzantine period. The Byzantine and Umayad periods show a dearth of forest species in favour of riverine species. Tell Deir Alla Villeneuve (Institut Fran~ais de Pistacia Samples of charcoal were taken by hand from the area of a sanctuary. Today the species identified do not occur in the area around the site. However these species are common in the hills rising on the east side of the Jordan valley, for example near Petra. Of particular interest were the stones of Celtis found with the charcoal because this site is south of its present habitat. Raum Fenan (Wadi Araba) see Baierle et al. 1989. Jordan valley, 1200-500 BC. (van Zeist 1985). LEVANT Fraxinus Olea Platanus Populus Quercus decid Quercus evergreen Tamarix N= + + + 48 Jericho Jordan valley, west bank. Aceramic Neolithic and E.B.A.: see C. Western 1971. Extensive analyses of charcoal from a wide range of sites has been carried out in Israel. The reader is referred to Liphschitz (1986 & 1989, and see this volume). Willcox Timber and Trees Willcox IRAN Abdul Hosein S.E. Iran, Lorestan. Aceramic Neolithic, 7th millennium B.C. Excavated by I. Pullar (Willcox 1990a). Tamarix + + ++++ +++ ++ + N= 20 Chenopod Crataegus Pistacia b u s Salicaceae Identifications from this site indicate an open PistacialArnygduIus formation as described by Zohary (1973) It is clear that by the early Neolithic in this region the Quercus-dominated forests had not colonised the area. Pollen evidence (van Zeist 1982) indicates that this occurred at a later date. Similar evidence was forthcoming from Ganj Dareh, a site of similar date occuring in the same general region (see below). Large fragments of charcoal which apparently came from roof beams are Pistacia. Jarmo Foothills of the Zagros. (Braidwood et al. 1983). Quercus * Populus Tarnarix * Malyan Bronze Age site in the Zagros mountains 46 kms northwest of Shiraz. Altitude 1100 meters (Miller 1985). Acer Amygdalus Capparis Daphne Fraxinus Juniperus Piitacia + + + +++ + + Prosopis Quercus Ulmaceae Vitex + + + + 97 89 Vitis N= * The --- results - Ganj Dareh S.E.Iran. Aceramic Neolithic, 7th/8th millennium B.C. (van Zeist et.al. 1984). Flotation samples indicate that the immediate surroundings of the site were deforested during the early phase when Populur and Juniperus were the most common species. During the later phase periods the inhabitants of the site went further afield for their fuel into the surrounding hills where they exploited Amygdalus, Quercus, and Pistacia (Miller 1985). Species which may have cnlonised - - -- - the deforested zone such as Prosopis and Capparis were only found in the later period. There is also evidence for the use of dung-cake as fuel during this period. -- IRAQ Hand samples Celtis Pistacia Salix/Populus N= Charcoal analyses carried out at Groningen indicate that as for Tepe Abdul Hosein the Quercus forests were not present during the Neolithic. However Celtis is present at Ganj Dareh but absent at Tepe Abdul Hosein both in the charcoal remains and in the seed remains. Published identifications from Iraq come largely from individual finds of beams, objects or furniture (see Moorey and Postgate, this volume). Below is a summary of the results. Taxon Bunus Cedrus libani Cupressus Fagus Juglans Morus Pinus Platanus Quercus Site Nuzi Nuzi Ur Tell Seh Hamad Nimrud Nimrud Brak, Nirnrud,Ur, Tell ed-Der Brak Brak, Nimrud Timber and Trees Willcox Timber and Trees The table above gives a list of identifications of imported timber from sites in Mesopotamia. Other finds include Populus and Fraxinus which probably grew in the Tigris Euphrates valleys. Date beams have also been recovered from several sites. CENTRAL ASIA Kanduhar Southern Afghanistan. Multi-period. British Institute for Afghan Studies. (Report lolthcoming). 600-3OOBC 300-0BC GULF Failaka Island in the Gulf of Kuwait. Hellenistic and Bronze Age. Excavated by J-F. Salles, C.N.R.S. (Willcox 1990b). Bronze Age Hellenistic Chenopod Phoenix Pinus Tamarix Zizyphus N= Acacia, Zizyphus and Tarnarix all occur on the island today. Ho~czvertheir status as indigenous components of the natural flora is dubious. The absence of Acacia indicates that it may have been introduced at a later date while Tamarix and Zizyphus were present, at least, since the Bronze Age. Pinus which occurs only in the Hellenistic levels is almost certainly an importation associated with the port. Northern Oman. Bronze Age. K. Frifelt, Danish archaeological mission (Carlsberg Foundation). Bat Acacia Prosopis Zizyphus These results are based on a preliminary analysis only. They reflect the contemporary vegetation occumng in the vicinity of the site. Despite the extreme desed conditions in this tropical latitude these trees are able to survive because of their deep rooting system in areas where there is abundant ground water. Thus trees occur in the plains or wadi bottoms which lie adjacent to large mountain areas which supply the ground water. The hills around Bat are bare but this is not due to deforestation as it is improbable that they could have supported tree species during the last 10,000 years. However at high altitude, above 2000 metres, the climate is less harsh and there is another vegetation belt with Juniperus polycarpos and Ceratonia siliquia. OAD-600 60&12OO 1200-1738 Celtis Chenopod Ficus Fraxinus Juniperus Lycium Moms Pinus Pistacia Platanus F'Nnus Salicaceae Tamarix Vitex agnus-castus Over one hundred samples from several different periods were taken from this site. As at Ai Khanoum there is very little available timber in the immediate area today. 80 kms. to the north we located forest remnants which were still being exploited and in some cases the timber was brought by truck to the modem town of Kandahar. At present timber sold for fuel comes only by truck because of the relatively long distances involved. Other alternative resources which are more local, such as reeds, are brought by camel. The three species (Juniperus, PLtacia and Prunus) typical of the supposed climax vegetation of the region are more common in the early levels than in the later ones. This would appear to indicate that these trees became less available. Pinur, which must be an introduced exotic species, was present during all phases. In addition to the charcoal evidence, Celtis was present as seed remains. This find of Celfis is well outside the known natural distribution and is further evidence for a reduction in this species (see below). Shorarghaf Nonhem Afghanistan, valley of the Oxus. Bronze Age. Excavated by H.-P. Francfort C.N.R.S. (Willcox 1989, 1991). (see next page) Willcox Timber and Trees 250-1600 B.C. Amygdalus Elaeagnus Lycium Pistacia Populus Salk Tamarix N= Sixty-six flotation samples were recovered and analysed indicating that the inhabitants used wood from the gallery forest which was approximately 2 km from the site. There is also evidence of open forest steppe species such as Pistacia and Amygdalus but they are not common. At the present time in the region the two species are extremely rare, indicating that the forest species were more extensive. Sarazm Soviet Tajikistan 50 km E. of Samarkand. Bronze Age. Excavated by R. Besenval and H.-P. Francfort, C.N.R.S. (Besenval 1987). 26W2100 B.C. Amygdalus Celtis Elaeagnus Fraxinus Phragrnites Pistacia Salicaceae Tamarix N= Flotation samples were examined from this site. The results indicate a similar situation to that at Shortughai, where the gallery forest which follows the rivers and is very characteristic of the region supplied the majority of timber identified from the charcoal remains. Ai Khanoum Northern Afghanistan, valley of the Oxus. 2nd century BC. Excavated by P. Bernard, CNRS (Willcox unpub). Tamarix sp. ++ + + ++++ ++ + + + N= 33 Celtis sp. Crataegus sp. Elaeagnus sp. Junipe~ssp Pistacia sp. Platanus sp Salicaceae Thirty-three samples recovered from the treasury were processed to separate the charcoal and Timber and Trees Willcox other plant remains. The presence of charcoal apparently resulted from the residues of fuel from smelting activities which took place immediately after the Hellenistic period in the second century B.C. Today there is very little available fuel in the region with the exception of the riverine gallery forest where Elaeagnus sp. is dominant, associated with Salix sp. and Tamarh sp. These forests, which are extensive, are the only source of timber for fuel today unless one travels 40 to 50 kms to higher altitudes. It is therefore significant that among the samples the most widely represented tree is that of Juniperus sp. and not the riverine species. Thus Juniperus and Pistacia would appear to have had a much more extensive distribution during the second century B.C. than at present. Mundigak Southern Afghanistan. Bronze Age. Excavated by J. Casal (1961). Analysis R. Porkres, M u s k National d'Histoire Naturelle, Paris. cf. Salvadora persica Zizyphus vulgaris Unfortunately this site was not extensively sampled and only two determinations were made, the second from the fruit. Both species are surprising since they do not grow at this altitude in the region today, winter temperatures being too low. However in view of the limited sampling it would be unwise to draw any conclusions since these samples may not be representative and could therefore be misleading. Introduction and cultivation of exotic species At present there is little evidence from charcoal for the cultivation of fruit trees. This results from the difficulty of separating closely related species (an exception is Morus from Nimrud; however the reliability of this identification is in question). From charcoal remains it is difficult to establish whether wood for timber has been cultivated or collected from its natural habitat. There are two possible approaches. 1) Growth ring width in the case where a species was grown under irrigation in an arid environment. 2) Tracing the spread of certain species outside their natural area of distribution. The one taxon whose cultivation probably dates back the furthest is Populus spp. As charcoal this been found on several Neolithic sites. Populus spp are widely cultivated today and are still used for roof beams. They occur commonly on archaeological sites; for example for the Neolithic period it was found at Cafer Htiyiik, Can Hasan, Tepe Abdul Hosein, Ganj Dareh Tepe, Mureybet, and Abu Hureyra. Populus is well suited to irrigated conditions and is traditionally cultivated in association with irrigation and can be easily grown from cuttings. Pinus charcoal has been frequently found outside its natural habitat, however it is difficult to establish whether this represents an introduction or an importation of timber. In many parts of the Middle East Pinus halepensis is cultivated more often as an ornamental than for timber. This species could have been cultivated in southern Iraq under irrigated conditions. Gymnosperms are propagated from seed rather than from cuttings and therefore require more care. Records of identifications of Pinus outside its natural habitat are common, especially for the historical periods, for example Bosra, Apan, Failaka, Pella, ~ a n d a h a rand the Mesopotamian finds. Willcox Timber and Trees Elaeagnus angustifolia occurs naturally as the dominant in the gallery forests of Central Asia. It appears to have been introduced into the Near East, but the exact date of its introduction is difficult to establish. Correlation with pollen and seed evidence Pollen evidence comes from undisturbed sediments where pollen accumulates under natural conditions. Pollen from archaeological sites is problematic because of the nature of the deposition. The pollen may be residual and there is no constant environment for the steady accumulation of pollen-rich sediments derived from the atmosphere. In contrast charcoal occurs in situ in archaeological deposits so the problem of residuality is eliminated. On the other hand man would have been selective in his gathering of wood and also in his gathering of fruit. It has already been demonstrated that there is a good correlation between pollen obtained from Zeribar and charcoal from Ganj Dareh and Tepe Abdul Hosein (Willcox 1990a). Further west in southern Anatolia the presence of Quercus on aceramic Neolithic sites also correlates with the pollen evidence. In northern Syria the presence of Quercus at Mureybet is surprising and one is tempted to consider the possiblity of importation of timber. On the other hand seed remains from Abu Hureyra of Pktacia, Celtis, Prunus, Pyrus, and Mespilus gerrnanica suggest that during the Epi-Palaeolithic these species extended further east (and south) than we might have expected from the pollen record. Further seed evidence to corroborate this comes from early Neolithic sites in the Damascus basin (van Zeist 1984) where Pistacia, Arnygdalus, Crataegus and Pynrs were identified. Thus in Syria it appears that during the Neolithic, forest species occurred further south and east than at present. Evidence of the use of timber in antiquity from charcoal analysis In the vast majority of cases charcoal from flotation samples represents the remnants of fuel used in domestic or industrial hearths. But, where a building has burned the roof timbers invariably survive. To take just a few examples, there is evidence for Quercus at Mureybet, Populus at Asvan, Pistacia at Tepe Abdul Hosein, Pinus at Bosra and Juniperus at Gorciion being used for roof timbers. In general the species used reflect the species most available. Evidence for other uses is extremely rare because burnt objects rarely retain their identifiable form. Evidence for deforestation andlor vegetation changes ftom charcoal analysis For many sites in the Near and Middle East, the comparison of the species range obtained from charcoal analysis of prehistoric and historic levels with the species available in the immediate environment of the site today (up to 10 kilometres) indicates a drastic reduction in forest cover. However this assumes that ancient man gathered his everyday fuel in the vicinity of the sites and did not travel to obtain his fuel. The use of alternatives to wood for fuel no doubt developed in areas which lacked timber or became deforested; today the use of dungcake fuel is very widespread, and may occur in wooded areas where authorities forbid the cutting of timber, as in Jordan, Syria and Turkey. Perhaps the best evidence for real decline is the replacement of preferred species such as Quercus or Pistacia by smaller spiny shrubs such as Paliurus, Zizyphus, Lyciurn, Crataegus, and the Chenopods. This appeared to be the case at Asvan, Bosra, Pella and Kandahar. However there is a need to examine more multi-period sites where there is a long sequence, including more recent levels. If one excludes the possiblity of long-distance transportation, there is evidence for lower altitude limits for woody plant formations in the Neolithic and Bronze Age periods, from sites Willcox Timber and Trees such as Abu Hureyra, Mureybet, Jawa, Jericho, Kandahar and Tepe Abdul Hosein. A specific example is Celtis which has been identified from the following sites: Ganj Dareh, I'clla, and Kandahar. These sites are well outside the modem distribution of this species, showing (I considerable reduction in its distribution. Quercus, Pinus and Cedrus are trees much in demand for their timber, and these three spccies would appear to have been reduced by over-exploitation, while other species, particularly fruit-bearing trees of the Rosaceae family, were preserved. Today in parts of northern Syria and 'I'urkey we see natural orchards where the only remnants of the original forest are dispersed fruit rrces which may be used as grafting stock. IDossibleevidence of trade and long distance transport of timber Here one must make the distinction between imported products and trees which are introduced for cultivation. One of the ~)roblemsis that it is not always easy to distinguish these two categories. For example finds of I'inus, Morus, and Juglans from Mesopotamia could be either imports or introductions. Here the tcxts may be of some help. On the other hand Cedrus and Buxus represent imported timber. Outside Mesopotamia, Quercus from Neolithic levels at Mureybet may be an import (one cxplanation is driftwood coming down the Euphrates). Pinus from Hellenistic levels at Failaka is assumed to be an import, however its presence at Kandahar is more ambiguous. Clearly more samples need to be analysed before firm conclusions can be made about the origin of these woods. Finds of forest species such as Quercus and Acer in hearths from Jawa in the Black Desert some 100 kilometers from the present day habitat are curious, particularly since many samples represented branches of only a few years growth. One is inclined to reserve judgement until complementary evidence is available. Conclusion Evidence from charcoal analysis complements pollen data in the study of the vegetational history of the Middle East and it adds to our knowledge of the exploitation and use o f woody species. This in turn provides evidence for trade and the development of arboriculture. Recause man-made deposits are involved, some results tend to be ambiguous. Complementary data should in the future resolve certain questions such as whether firewood was local or imported. This problem is particularly true of sites situated away from forested areas but near enough to allow the transport of fuel. There is unequivocal evidence for deforestation in the Near East, which may have led to dcsertification in some regions. The evidence comes from both charcoal and studies of the modem vegetation cover. While over-exploitation is undoubtedly responsible, climatic change may also have played a r61e, though at present it is difficult to distinguish between man's influence and climatic change. Altitude limits of woody formations appear to have been lower than at present during the Neolithic and Bronze Age periods, and more detailed studies could possibly reveal the vegetational sequence which led to deforestation. During the Neolithic in Syria, Iran and Turkey the vegetation cover appears to have been more favorable to man than at present. For the Bronze Age all the evidence points to a richer environment in terms of tree and shrub species when compared with the present-day situation, although the evidence is still scarce. For the later historical periods we continue to see the same forest species being exploited. Examination of more recent Islamic levels at sites such as Asvan, Willcox Timber and Trees Bosra and Kandahar indicate that the destruction of the forests and the local extinction of some species was a relatively late phenomenon. BIBLIOGRAPHY Frifelt, K. 1985 Oreguss, P. 1955 1959 Aurenche, 0. "Un exemple d'architecture domestique en Syrie au VIII millhaire: la maison XLVII de 1980 Mureybet", in J.CI. Margueron (ed.), Le Moyen Euphrate, zone de contact et d'kchanges (Actes du Colloque de Strasbourg, 10-12 mars 1977), pp. 35-53. Baierle et al. H.U. Baierle, W. Frey, C. Jagiella & H. Kiirschner, "Die Brennstoffressourcen im Raum 1989 Fenan (Wadi Araba, Jordanien) und die bei der Kupferenverhiittung verwendeten Brennstoffe", Der Anschnitt (Bochum, Selbstverlag des Deutschen Bergbau-Museums), Beiheft 7, 213-222. Bretzel, B. 1903 Botanische Forschungen des Alexunderzuges (Teubner: Leipzig). Besenval, R. 1987 "Dkouvertes rkentes h Sarazm (R.S.S. du Tadjikistan)", Acadkmie des Inscriptions et Belles-Lettres, avril-juin 1987, 441456. Casal, J-M. 1961 Chabal, L. 1990 Davis, P.H. 1971 Frey et al. 1991 Guest, E. 1966 Heinz, C. 1990 "Further Evidence of the Third Millennium BC Town at Bat in Oman", Journal of Oman Studies, Vol. 7, 89-104. Identification of L,iving Gymnosperms on the Basis of their Xylotomy (Budapest). Holzanatomie der Europaischen Laubhiilzer und Straucher (Budapest). Flora of Iraq, Vol.1 (Glasgow) "M6thodologie et PaltWxologie du gisement prkhistorique de 1'Abeurador (Hkrault) au passage M6solithiquW6olithique d'aprhs l'analyse des charbons de bois: premiers r&ultats", in T. Hackens, A.V. Munaut & C. Till (eds.), Wood and archaeology I Bois et archt!ologie (Pact, 22; Belgique) 219-229. Millman, G.C. 19848~5 "Traditional husbandry and processing of archaic cereals in recent times: the operations, products and equipment that might feature in Surnerian texts. Part I: The glumed wheats", Bulletin on Sumerian Agriculture 1, 114-152 & 'Tart 11: The free threshing cereals", Bulletin on Sumerian Agriculture 2, 1-31. Fouilles de Mundigak, Vol. I. (M6moires de la Mlbgation Archhlogique Frangaise en Afghanistan, Tome XVII). Hillman, G.C., Colledge, S.M. & Harris, D.R. "Plant food economy during the Epipalaeolithic period at Tell Abu Hureyra. Syria: dietary 1989 diversity, seasonality and modes of exploitation", in D.R. Harris & G.C. Hillman (eds.), Foraging and Farming: Evolution of Plant Exploitation (Landon: Unwin & Hyman), 240-268. '‘L'Btude pal&-kologique de sites pmton~storiques partir des charbons de bois: la question de l'uNr.6 de mesure. Mnombrement de fragments ou pesks?", in T. Hackens, A.V. Munaut & C. Till (eds.), Wood and archaeology I Bois et archt!ologie (Pact, 22; Belgique), 189-205. Liphschitz, N. "Plant economy and diet in the early Bronze Age in Israel: A summary of present research", 1989 in P. de Miroschedji, L'urbanisation de la Palestine l'age du Bronze ancien (BAR International Series 527/ii), 269-277. Plant Life of South-West Asia (Edinburgh). Dentzer, J.-M. (ed.) 1985 Hauran I: Recherches archkologiques sur la Syrie du Sud h l'kpoque hellknistique et romaine (IFAPO Bibliothhue archhlogique et historique, Tome CXXIV). Freitag, H. 1977 Timber md l'rma Willcox "The ple~glacial, late glacial and early postglacial vegetations of Zeribar and their present-day counterparts", Palaeohistoria XIX, 87-95. W. Frey, C. Jagiella & H. Kiirschner, "Holzkohlefunde in DW-Katlimmu/Tall fhmad und ihre Interpretation", in H. Kiihne, A. Mahmoud & W. RUllig (eds.), Berichte der HamadlDiir-Katlimmu (Verlag Reimer Berlin), Band 1. Ausgrabung Tall Mande~lle,J. "Plants", in Journal of Oman Studies: Special Report: The Scientijic Results of The Oman 1977 Flora Survey 1975, pp 229-267. 1978 Wild flowers of northern Oman (Aylesbury). Miller, A. & Moms, M. Plants of Dhofar. The southern region of Oman, traditional and economic uses (Edinburgh: 1988 McDougall). Miller, N. 1984 1985 "The interpretation of some carbonised ct:real remains as remnants of dung cake fuel", Bulletin on Sumerian Agriculture 1, 4547. "Paleoethnobotanical evidence for deforestation in ancient Iran: A case Study of Urban Malyan", Journal of Ethnobiology 5 (I), 1-19. Timber and Trees Willcox Timber and Trees Willcox Mouterde, SJ. La Flore du Djebel Druze (Paris). 1957 1991a "Carbonised Plant Remains from Shortughai Afghanistan", in J. Renfrew (ed.), New Light on Early Farming, Recent Developments in Palaeoethnobotany (Edinburgh University Press). Pearsall, D. 1989 1991b "Cafer H6yiik (Turquie): Les charbons de bois nbolithique", Cahiers de I'Euphrare (Editions Recherche sur les Civilisations; Paris) 139-150. Paleoethnobotany: A hand book of procedures (New Yo*). Salles, J.-F., & Calvet, Y. Failaka: Fouilles Fran~aises1984-198.5 (Travaux de la Maison de I'Orient, No 12). 1987 Schweingruber, F. Mikroskopische holzanatomie (Zurich). 1978 Thiebault, S. "Palaeoenvironment and ancient vegetation of Baluchistan based on charcoal analysis of 1988 archaeological sites", Proc. Indian. Nut. Sci. Acad., 54 A, No. 3, 501-509. Vernet, J.-L. and Thiebault, S. "An approach to northwestern Mediterranean recent prehistoric vegetation and ecological 1987 implications", Journal of Biogeography 14, 117-127. Vernet, J.-L. "The bearing of phyto-archaeological evidence on discussions of climatic change over recent 1990 millennia", Phil. Trans. Royal Soc. Lond., A, 330, 671-677. in prep. a "Les macro-restes vbgbtaux et l'konomie vbgbtale de Bosra et Sia, Syrie du Sud: r&ultats pour les @rides romaine, byzantines et islamiques", Hauran 11: Recherches Archologiques sur la Syrie du Sud, I'Epoque Hellknistique et Romaine. in prep. b "Charcoal from Epipaeolithic and Neolithic levels at Mureybet (Syria)". Williams, D. "Flotation at Siraf", Antiquity 47, 188. 1973 van Zeist, W. 1972 "Palaeobotanical results of the 1970 season at (hy6nii". Helinium 12, 13-19. 1978 "A postglacial pollen diagram from Lake Van in eastern Anatolia", Review of Palaeobotany and Palynology 26, 249-276. 1980 "Holocene vegetation and climate of northwestern Syria", Palaeohistoria 22, 11-125. van Zeist, W. & Bottema, S. 1977 "Palynological investigations in Western Iran", Palaeohistoria 19, 19-85. Villeneuve, F. "L'6conomie male et la vie des campagnes", in J.-M. Dentzer (ed.), Hauran I: Recherches 1985 archkologiques sur la Syrie du Sud h 1'6poque hellknistiqw et romaine (IFAPO Bibliothhue archblogique et historique, Tome CXXIV). 1982 "Vegetational history of the Eastern Mediterranean and the Near East during the last 20,000 years. in J.L Bintliff, W. van Zeist & H. Woldring (eds.), Palaeoclimates Palaeoenvironments and Hwnan Communities in the Eastern Mediterranean Region in Later Prehistory (B.A.R. International series 133), 277-321. Western, C. "The ecological interpretation of ancient charcoals from Jericho", Levant, 111, 31. 1971 1985 "Past and present environments in the Jordan Valley", in A. Hadidi (ed.), Studies in the history and archaeology of Jordan I1 (Amman). Willcox, G.H. "A history of deforestation as indicated by charcoal analysis of four sites in eastern 1974 Anatolia", Anatolian Studies 24, 117-133. van Zeist, W. & Bakker-Heeres, J.A. "Archaeobotanical studies in the Levant 3. Late Palaeolithic Mureybet", Palaeohistoria 26, 1984a 171-199. 1979 "Preliminary report: Analysis of charcoal from Can Hasan III", Bulletin of the British Institute of Archaeology at Ankara, 1979. 1981 "Appendix F", in S.W. Helms, Jawa, the Lost City of the Black Desert (London). 1983 Unpublished report on archaeobotanical finds from Pella, Jordan. van Zeist, W., van Smith, R.M., Palfenier, M., Suwijn, M. & Casparie W.A. 1984 "An Archaeobotanical study of Ganj Dareh Tepe", Palaeohistoria 26, 201-224. 1987 "List of trees and shrubs of economic importance in Iraq", Bulletin on Sumerian Agriculture 3, 101-106. van Zeist, W. & Waterbolk-van Rooijen, W. 1985 "The Palaeobotany of Tell Bouqras, eastern Syria", Palkorient 11/2, 131-147. 1989 "Etude ArchCobotanique", in Francfort H.-P. (ed.), Fouilles de Shortughai; Tome I1 (Mbmoires de la Mission Archblogique Fran~aiseen Asie Centrale), 175-185. 1989 1990a "Charcoal remains from Tepe Abdul Hosein", in J. Pullar, Tepe Abdul Hosein: A Neolithic site in Western Iran. Excavations I978 (BAR International Series 563), 223-227. Zohary, M. 1973 1990b "The Plant remains from Hellenistic and Bronze Age levels at Failaka, Kuwait: A preliminary report", in F. Salles, Failaka Fouilles Fran~aises19864988, 43-50. 1984b "Archaeobotanical studies in the Levant 1. Neolithic Sites in the Damascus Basin: Aswad, Ghoreife, Ramad", Palaeohistoria 24,165-256. "Plant remains from Tell Sabi Abyad", in M.M. Akkermans, Excavations at Sabi Abyad (B.A.R. International Series 468), pp.325-335. Geobotanicalfoundations of the Middle East, 2 vols (Stuttgart-Amsterdam). LEVANT TREES AND TREE PRODUCTS Nili Liphschitz (Tel A viv University) Introduction Two main sources of information contribute to the reconstruction of past vegetation: written documents and botanical remains gathered in archaeological excavations. The historical documents include the Bible, the Mishna and the Talmud, books of classical writers, as well as memoirs of pilgrims and travellers. The botanical evidence is based on pollen grain analysis, seeds and fruits identification, and mainly on analysis of construction timber and wood used for everyday tools, objects and fuel. Careful examination of each source and combination of all data enables us to reconstruct the past landscape. Quercus rnacrolepsis (=aegilops) Written Sources Q. calliprinos Quercus iMectoria Information on the trees of the Levant before the dendroarchaeological research started, both in Israel and the Near East, was based on written documents, i.e. the Bible, the Mishna and Talmud, books of Greek and Roman writers such as Theophrastus and Pliny the elder, and memoirs of pilgrims and elders who visited the Holy Land through the ages. Names of plants described in the written sources differ from the modern scientific identification. The names of the plants mentioned in the Hebrew Bible are translated to different genera according to the language. For example, the Hebrew modern name for cypress (Cupressus sempervirens) is "berosh", but the identification of "berosh" is controversial and uncertain. In an English translation of the Bible "berosh" is translated as fir. In a German translation it is referred to as 'Tannenbaum" - fir while in a French translation "berosh" is referred to as cypress. Several botanists suggest the "berosh" is the Abies cilicica while others prefer the Juniperus excelsa. Today the "berosh" is referred to as Cupressus sempervirens (Liphschitz & Biger, 1989a). The same holds true for one of the common trees of today - the pine (Pinus). The Hebrew modern name for pine - "oren" - is mentined only once in the Bible (Isaiah 4414). In an English translation of the Bible (Friedlander, 1881), one can find the mention of pine in Isaiah 60:13 "The glory of Lebanon shall come unto thee, the fir tree, the pine tree and the box together, to beautify the place of my sanctuary". The Hebrew original text there speaks about "tidhar" which is not recognised, yet one authority believes it might be a fir (Anderson, 1979, p.134). There is some dispute among the authorities as to the reference to pine in Nehemiah 8:15 "Go forth unto the mount and fetch olive branches and pine branches and myrtle branches and palm of thick trees". The Hebrew text refers to "etz shemen" which means "oil tree". Some authorities accept "etz shemen" as Eleagnus angustifolia. The same holds true for Isaiah 41:19 where the Hebrew name again is "etz shemen". Some refer to it as pine while others refer to a plane m e (Feldman, 1956; Felix, 1957; Anderson, 1979). According to the Septuagint, the first Greek translation of the Bible, "oren" means "Pitus", i.e. Pinus in Latin. In the Assyrian language there is a tree named "urnu" and the Aramaic translation of the Bible by Yonatan Ben Uziel explains the "oren" in Liphschitz Levant Trees and Tree Products Isaiah as being "urnu". According to Krupnik (in the Talmudic dictionary) the "umu" means "Lorbeerbaum" in German, which is Laurus nobilis, as it is identified with the Talmudic mention (Rosh Hashana section 23, p.1) of Aronim-An which means "Ar" i.e. the Hebrew name for k u r u s nobilis. In his dictionary, which is the first in modem Hebrew, Ben Yehuda suggests that the Biblical "oren" is Quercus i.e. "Eiche" in German (Ben Yehuda, 1950). According to Yona Ebn Jenach "tirza" is Snubar - an Arabic word meaning Pinus. There are many other translations of the word "oren" being other trees mentioned in the Bible and ancient literature. It is therefore impossible to find the exact name for the Biblical tree (Liphschitz et al., 1988-9). Numerous trees mentioned in the Bible are not recognised today, and their botanical identification is not certain. Such trees are "teashur", "gopher", "etz shemen", "tina", "tidhar", "bacha", "sneh", "armon", "almugim", "hovnim", "zeelim", "aviyona", etc. A long list of names of trees mentioned in the Hebrew Bible are not known today. Among them one can find names like "afarsemon", which today refer to the genus Persirnun, which is foreign and did not grow in Israel in the past. Trees mentioned in the Bible lack botanical description. The identification of the Biblical Israeli local trees by translators who lived in Europe, Africa or the United States, with foreign trees which were familiar to them, resulted therefore in numerous errors. Levant Trssr ad Tree Prodwt, Liphschitz time of Justinian in the 7th century AD, and the original constructions still exist today. In St. Catherine's Monastery hundreds of logs used in the construction of roofs, bdconier, walls and passages (figs. 1 4 ) enabled a careful examination of the whole complex (LiphscNtz & Waisel, 1973). Restorations of the roof of the El Aqsa Mosque and replacement of the original mof logs enabled an examination of the original timber of which the original m f was made (Lev-Yadun et al., 1984). Evidence fkom Plant Remains Pollen Grain Analysis One of the accepted methods for the reconstruction of past vegetation is pollen grain analysis. Usually pollen diagrams are aimed to represent the relationships between the AP (Arboreal Pollen) and NAP Won-Arboreal Pollen). Changes in the relationship between the AP and the NAP in various periods point to changes in the climate i.e. a wetter or cooler climate or a drier and warmer one. However, the quantity of pollen deposited per unit area depends upon a number of factors such as the frequency of the species in the area, its absolute pollen production which varies individually and according to the growth conditions of the specimen, and the dispersal mechanism of the pollen i.e. wind pollination versus insect pollination. Some other sources of errors might be local over-representation proximal to the site, long distance transport mechanism, differential destruction of pollen grains, vertical displacement of pollen after primary deposition, and the most important of all - the inability to distinguish in numerous cases between various species of the same genus, or even between various genera of the same family (Faegri & Iversen, 1975). Results of pollen analysis should therefore be considered very carefully, especially in historic excavations. Usually pollen analysis is conducted in prehistoric excavations or from cores taken from geological strata. However, information gathered from pollen analysis can add to the information gathered from other sources of research. fig. 1. St. Catherine's monastery in South Sinai: general view. The Basilica (the tiled roof) dates to the 7th century AD, the mosque to the 11-12th century AD and the guest wings to the 17th-18th century AD. Dendroarchaeology Direct evidence for the existence of trees in the past can therefore be obtained via dendroarchaeology. The dendroarchaeological research is based on the identification of wood remains gathered in archaeological excavatins or sampled in ancient buildings. i. Sources of information The existence of well-preservea ancient buildings, where wooden logs and beams are used in construction, is very rare in Israel. Such buildings are usually sacred places such as churches and mosques, e.g. St Catherine's monastery in South Sinai, El Aqsa Mosque in Jerusalem and the Church of the Nativity at Beith Lehem. They were founded at the fig. 2. The mosque with the logs in the construction. 35 Liphschitz Levant Trees and Tree Products Liphschitz Concerning archaeological remains, only a few sites were well preserved. Usually, d&# are located in the arid region. Secondly, only in very few cases were they not destmysd by bn following wars. One of the sites preserved is the Roman castle in En Boqeq (Dead Sea Redon) where logs were used especially as part of the guard tower roofs (fig. 5:l-2) (Liphschitz & Waisel, 1973, 1991). Another, well famous, site, partially preserved, is Masada, including the rampart and parts of the northern and western palaces (Liphschitz et al., 1981; Liphschitz & Lev-Yadun, 1989). fig. 5. The Roman castle in En-Boqeq. 1. One of the guard towers with the original logs in its roof. 2. A burned log of date palm in the yard. Liphschitz Levant Trees and Tree Products fig. 6. Burned wood remains in the excavation at Tel-Sheba dated to the Iron Age: a burned log of olive. Most of the archaeological sites were burned during wars, and most of the botanical remains, either logs or small pieces of wood, are c h m d (fig. 6). Very often small pieces of charcoal are dispersed all over the dig. These samples must be accurately numbered before being sampled (fig. 7). ii Basic assump2iuns Dendroarchaeology is based on several basic assurn~tions: 1. Timber for everyday use for construction of simple structures and for fuel was taken from the close environment and therefore represents the local natural vegetation. Only rarely was precious wood imported from far distances, i.e. for the construction of sanctuaries, palaces or governmental buildings, as in the case of cedar of Lebanon (Cedrus libano. Liphschitz fig. 7. Burned wood remains in the excavations at Tel Sheba dated to the Iron Age: charred small pieces of wood dispersed in the area. Two logs (I and H) are seen. 2. The basic ecological requirements of plant species in the past are the same as today; therefore changes in the composition of plants in the past indicate macroclimatic change. Climatic changes are especially prominent in transition regions between two climatic zones, and therefore vegetational changes will be obvious there. 3. The disappearance of plants from the area does not necessarily always mean a climatic change; man's impact on the composition of vegetation, i.e. clearance of forest areas for agriculture and introduction of new cultivated species and varieties, must always be taken into consideration. Liphschitz Levant Trees and Tree Products I up"-"i. Levant 7hu d Itw CN# The dendroarchaeological research provides important data on the distribution of tlrsdl limber supply, as well as indication for various uses of local native versus imported timber d u w antiquity. The proportional amounts of the various tree species found in the excavations repmmnt their availability in the region, as well as the intensiveness of exploitation in the examined site. Preservation of timber depends both on the climatic conditions as well as on the history of the ~ g i o nand the degree of preservation of the site. Very often sites were burned and destroyed during wars and the residue includes only small remnants of carbonised wood. Lack of appropriate techniques for handling charred wood hindered the collection and preservation of wood remains, resulting in the loss of most valuable data. Only in the last 20 years new methods were developed which enable identification of the charred wood up to species level. 111. Methods of analysis Dendroarchaeorogical analysis is based on the identification of the wood remains up to the species level, based on the anatomical three-dimensional structure of the wood (fig. 8:l-4). The exact identification is of special importance in Israel, where in spite of the small area, different species of the same genus represent different climatic conditions. For instance in the genus Quercus, Q. boissieri (Cypress oak) represents the very humid, high elevation species, Q. calliprinos (Kermes oak) is a typical Mediterranean species and Q. ithaburensis (Mt. Tabor oak) occurs in the dry regions of the Mediterranean territory, bordering the semi-arid districts. Concerning the genus Pistacia, P. palaestina (Terebinth) is a typical Mediterranean species occurring together with Q. calliprinos, P. lentiscus (Lentisk) occupies the low elevations of this territory, and P. atlantica (Atlantic pistachio) is an Irano-Touranian species. Reconstruction of the past: Levant trees and their products during antiquity fig. 8. Anatomy of burned wood originating from an Iron Age Layer at Tel Sheba: three dimensional structure of Acacia rdiana. 1-2. Cross section. 3. Tangential longitudinal section. 4. Radial longitudinal section. 40 The arboreal landscape In spite of its small area Israel is a meeting-point of three phytogeographical territories: the Mediterranean humid, the Irano-Touranian - semi-arid - and the Saharo-Arabian - arid - thus representing three different climatic regimes. A tropical penetration exists along the Jordan VaLley. The typical vegetation ranges accordingly from humid Mediterranean forest and maquis to a desertian one. The natural Mediterranean maquis and forest vegetation of Israel was until very recently commonly considered to be composed mainly of four components: Pinus halepensis - Aleppo pine, deciduous oak - Quercw ithaburenst, evergreen oak - Quercus calliprinos (Kermes oak) Pistacia palaestina (Terebinth) association, and carob - pistacia association - Ceratonia siliqua Pistacia lentiscus (Lentisk). The Irano-Touranian territory is characterised by Tarnarix aphylla (Tamarisk) and Retama roetam (White broom) and the Saharo-Arabian territory is characterised by the existence of various acacia species: Acacia raddiana, Acacia tortilis and Acacia negevensis in the wadis. The area is sparsely covered by small bushes or completely bare (Zohary, 1959, 1962, 1973). Accumulating data obtained from dendroarchaeological investigations all over Israel, in numerous sites, dating to the last 10000 years, show clearly that timber for everyday use for construction of simple structures was taken from the vicinity of the sites and therefore represents the local natural vegetation. Combined research in palynology, dendroarchaeology and written historical sources showed that the arboreal landscape of the Mediterranean territory of Israel in the past is quite different from that of more recent times. Analysis of wood remains and pollen from archaeological excavations suggest that Pinus halepensis, Quercus ithaburensis and Ceratonia siliqua (carob) did ' r Liphschitz Levant Trees and Tree Products Liphschitz Levant Trees and Tree Products not have a dominant position in the forest and maquis of Mediterranean Israel. Pinus halepensis and Cupressus sempervirens were rare (Liphschitz & Biger, 1989; Liphschitz et al., 1988-9). Ceratonia siliqua was very rare in Israel and its present dispersal probably dates from the last few centuries (Liphschitz, 1987a). The view that Quercus ithaburensis was dominant in the forest which once covered the Central Coastal Plain, as suggested by Eig (1933) and accepted by the geobotanists (Zohary, 1962, 1973) was also found to be erroneous (Liphschitz et al., 1987). Olea europaea (olive), which is a typical Mediterranean element, was part of the landscape, but was cultivated since the Early Bronze Age (Liphschitz et d., 1991), and its share in the arboreal landscape is therefore not the primary native one. The ancient vegetation of the Mediterranean territory was dominated by the Quercus calliprinos - Pistacia palaestina association (Liphschitz & Biger, 1990). On the other hand, the typical vegetation which characterises the semi-arid and the arid areas today did also characterise it in the last 10000 years. The wood remains were mainly made of Tamarix, mostly T. aphylla (tamarisk) and of various species of acacia, mainly Acacia raddiana (acacia), but also of Acacia tortilis and Acacia negevensis. Phoenix dactylifera (date palm) was extensively used in oases, but as in the case of the olive, this species was cultivated since ancient times (Liphschitz, 1989a). Timber trade and imported wood Commercial activities concerning timber trade in ancient times were known from all the Mediterranean region. The northern and humid mountain areas supplied wood for construction and shipbuilding for the southern arid countries like Egypt and South Israel. The local trees which grew in Israel were not suitable for large roofing needs in the prosperous temples and palaces of urban Israel of the Middle Bronze age onwards. Analysis of the timber used in the simple as compared to the monumental constructions provides information on timber trade and the degree of prosperity of the relevant period. The wood of Cedrus libani (cedar of Lebanon) was highly esteemed during antiquity and was used for special construction purposes. Its wood was highly regarded in Egypt by the builders of sarcophagi, coffins and other appurtenances of burial. King Cheop's Royal ship was made predominantly of cedar and the coffin of Queen Merit Amon was made of this timber. Cedar of Lebanon never grew in Israel. Its timber was used in the MB palace in Lachish, in the northern palace at Masada, in the LB temple at Lachish and in the Philistine temple at Tel Qasile, in residential buildings at Herodion and Kipros, in El Aqsa Mosque, and in the Byzantine church at Tel Masos in the Negev (Liphschitz 1987b). Apart from Lebanon cedar Juniperus phoenica (red juniper) was also imported to various monumental sites in ancient Israel. The red juniper is absent from Israel, Lebanon and Syria. It occurs in North Central Sinai and in the Highland of Edom. All the remains, except two pieces, were found in the Hellenistic-Roman periods in three sites only - Masada, En Rachel and Moa (fig. 9:l-3). Most of the wood was collected at Masada where about half of the timber was made of this species (Liphschitz & Lev-Yadun, 1989). The other two trees which can be seen as imported trees are the cypress and the Aleppo pine. These two trees were probably part of the arboreal vegetation of Israel but their occurrence was very rare. It seems that they may represent trees that in some periods were planted in Israel while in other periods they were imported from nearby areas (Liphschitz & Biger, 1991). fig. 9. Remains of Juniperus phoenica wood originating fmm Moa. 1. Cross cut of a log. 2. Worked object. 3. A branch with the bark. Liphschitz Levant Trees and Tree Products Everyday life: Furniture, tools and objects Timber was used for furniture, coffins, statues, tools and objects. Investigation of these wooden objects can shed light on the techniques and preferred types of timber during antiquity. Coffins discovered in Jericho and Ein Gedi were made mainly of Ficus sycomorus (sycamore) (Liphschitz 1986, 1988a). Identification of the woody species which were used for manufacturing certain objects and tools from Masada (Liphschitz & Lev-Yadun, 1989) and from Ein Gedi (Liphschitz, 1988b) show that combs, bowls, wooden boxes and other objects were made of Myrtus comrnunis (myrtle), Nerium oleander (oleander), Juniperus phoenica (juniper), Tarnarix (tamarisk), Acacia raddiana and A. tortilis (acacia) and Ficus sycomorus (sycamore). The selection of specific wood species for manufacturing certain objects seems to reflect an old tradition. Shipbuilding Timber was very important particularly in shipbuilding. Sunk shipwrecks near the coast, very often well preserved, enable reconstruction of the entire ship with the original timber. Botanical remains of submarine excavations include parts of ships and boats, parts of constructions of the harbour and remnants of the cargo. Research on these remains can provide information on suitable water-resistant timber and shed light on the foreign commerce and navigation lines during ancient days. Analyses of sunk shipwrecks along the coasts of Israel reveal that most of the timber was European conifers, but specific parts like nails were made of hardwoods (Liphschitz, 1981, 4, 5, 6). Human impact: deforestation, cultivation and introduction of fruit trees Dendroarchaeological analyses not only provide information on past landscape and environmental ecological conditions but also reflect human impact on the environrnment following deforestation and introduction of cultivated fruit trees and plants. As was previously mentioned, before the increasing population invaded the woodlands of the plains and the slopes of the hills, and later on of the mountains, the Mediterranean region was covered by a climax vegetation of Quercus calliprinos - Pistacia palaestina forests and maquis (Lipschitz & Biger, 1990). Cutting of wood served for timber and fuel supply and also for clearing areas for agriculture. Massive deforestation which was followed by soil erosion resulted in a change in the climax vegetation and replacement of the forests by lower plant communities. Large regions became devastated. At the same time cultivated trees replaced natural forests. This change is well represented by the very high proportions of olive remains, both timber and stones, in the archaeological excavations, from the great number of oil-presses and from olive orchards which still cover today's large Mediterranean districts (Liphschitz, 1987~).The role of the date palm in the economy of ancient times is of special importance due to its occurrence in the desert regions. It was one of the first fruit trees which were taken into cultivation in the Levant, and each part of the tree was exploited for the benefit of man (Liphschitz 1989a). Local cultivated trees as well as foreign tree species like peaches (Liphschitz, 1989b) were introduced rather lately, replacing the native local vegetation of the region. Liphschitz BIBLIOGRAPHY Anderlind, L. 1890 "Mittheilungen iiber Starke Baume in Syrien", Zeitschrift des Deutschen Palastina-Vereins 13: 220-227. Anderson, D. A. 1979 AN the Trees and Woody Plants of the Bible. (Word Books, U.S.A.) Ben Yehuda, E. 1950 A complete dictionary of ancient and modern Hebrew. (Poplar Edition, La-am, Tel-Aviv), Vol. 1, p. 397. Faegri, K. & Iversen, J. Textbook of Pollen Analysis. (Oxford). 1975 Feldman, U. 1956 Plants of the Bible. Plants of the Mishna @vir, Tel-Aviv). Felix, J. 1957 Plant World of the Bible (Masada Rarnat Gan). Lev-Yadun, S., Liphschitz. N., & Waisel, Y. 1984 "Dendrochronology of Cedrus libani logs from El Aqsa Mosque roof', Yearbook of the Israel Soc. Antiquity 17, 92-96 (Hebrew with Eng. Sum.) Liphschitz, N. 1981,4,5,6 Dendroarchaeological investigations in Israel. (Mimeographed reports no. 90, 1 13, 129, 130. Institute of Archaeology, Tel-Aviv University (Hebrew)). 1986 "Overview of the dendrochronology and dendroarchaeology in Israel", Dendrochronologia 4:37-58. 1987a "Ceratonia siliqua - an ancient element or a newcomer", Israel Jour. Bot. 36, 191-197. 1987b "Cedrus libani in ancient Israel", in Vilnay II (ed. Schiller), pp. 291-294 (Hebrew). 1987c "Olives in ancient Israel in view of dendroarchaeological excavations", in eds. Helzer & Eitam Olive oils in antiquity - Israel and neighbouring countries, (Univ. Haifa) pp. 139-145. 1989a 'The Date Palm (Phoenix dactylifera) in Israel during antiquity dendroarchaeological research", Hassadeh 69:2022-2025 (Hebrew). in view of 1989b "The Almond (Amygdalus communis) in Israel during dendroarchaeological research", Hassadeh 70:86437 (Hebrew). in of antiquity view Liphschitz, N. & Waisel, Y. 1973 "The effect of human activity on composition of the natural vegetation during historic periods" Proc. 4th Sci. Conf. Israel Ecol. Soc. F, 1-18. 1991 "Dendroarchaeological investigations in Israel: En Boqeq - Dead Sea Region" in En Boqeq (ed. Gichon) (in press). Liphschitz Levant Trees and Tree Products Liphschitz, N., & Lev-Yadun, S. 1989 "The botanical remains from Masada: identification of the plant species and the possible origin of the remnants'' BASOR 274, 27 -32. Liphschitz, N., & Biger, G. - 1989 "Cypress 35-45. 1990 "Ancient dominance of Quercus calliprinos (Kermes oak) - Pistacia palaestina (Terebinth) association in the Mediterranean" Israel Jour. Vegetation Sci. 1, 67-70. 1991 Cupressus sempivirens in Eretz Israel during antiquity", Israel. Jour. Bot. 38, Timber trade in ancient Israel (manuscript). Liphschitz, N., Lev-Yadun, S., & Waisel, Y. 1981 "Dendroarchaeological investigations in Israel: Masada" IEJ 31, 230-234. ETHNOGRAPHIC EVIDENCE FOR WOOD, BOATS, BITUMEN AND REEDS IN SOUTHERN IRAQ Ethnoarchaeology at al-Hiba Edward Ochsenschlager (Brooklyn College, City University of New York) A project to determine the possible bearing of modern ethnographic information on some of the archaeological data was undertaken during the 1970-1971 and 1972-1973 seasons of the excavations at Tell al-Hiba in southern Iraq. The information reported in this paper was collected at that time with the support of the Research Foundation of the City University of New York and the National Endowment for the Humanities. Liphschitz, N., Lev-Yadun, Y., & Gophna, R. 1987 "Dominance of Quercus calliprinos (Kermes oak) in the Central Coastal Plain in Israel during antiquity", IEJ 37,43-50. Liphschitz, N., Biger, G., & Mendel, Z. 1988-90 "Did Aleppo pine (Pinus halepensis) cover the mountains of Eretz Israel during antiquity", Israel - People and Land 5-6, 141-150 (Hebrew with Eng. Sum.) Liphschitz, N., Gophna, R., Hartman, M., & Biger, G. 1991 Zohary, M. 1959 1962 1973 "Beginning of olive (Olea europaea) cultivation in the Old World: a reassessment", Jour. Archaeol. Sci. (in press). Geobotany. (Sifriat Poalirn (Hebrew)). Plant Life of Palestine (Ronald Press, New York). Geobotanical Foundations of the Middle East. (Gustav Fischer Verlag, Stuttgart). Al-Hiba is located on the edge of the permanent marsh which lies below Shatra on the Shatt al Gharraf. The Gharraf is a river that flows south-west from the Tigris at Kut in the direction of Nasiriya. Around the ancient mound are contiguous areas of seasonal and temporary marsh. Melting snows in the mountains to the north causes the Gharraf to overflow its banks, flooding these areas annually. The inundation reaches its height in May and begins to recede in June. By August the temporary marsh is covered with a growth of sedges and grass ready to welcome the Hadij, nomadic Bedouin who arrive with their herds of goats, sheep and camels to take advantage of the pasturage afforded. The waters reach their lowest point in September and October. In November there is usually a slight rise in the water level and, with h e rainy season, beginning in late December or early January, sudden short floods may occur. Many of the Beni Hassan grow winter crops of barley and wheat on small plots of land. Crops are largely dependent on the winter rains, nowadays supplemented by mechanical pumps, for water. During the spring both the Beni Hassan and the Mi'dan sow rice in the seasonal marshland flooded by the annual inundation. Most also have small herds of sheep and goats and flocks of chickens and turkeys, and some keep small herds of cattle. Mi'dan keep herds of water buffalo which depend on the reed shoots, rushes and sedges of the marshes for their fodder. Each of these three peoples occupies an important ecological niche in the area. WOOD, BOATS AND BITUMEN Very few trees of any size exist in the villages around al-Hiba. Most of them are palm trees or willows. They are usually the property of individual families and if another wishes to make use of one of them, he must purchase it for barter or cash. To buy a whole tree would make the process of small scale woodworking in the village prohibitively expensive. Luckily there are other ways that one can acquire wood. Dead branches from trees are sometimes available and the local wood-worker can re-use old and broken pieces of wood discarded by householders because they can no longer fulfill the use for which they were designed. For instance, the wood salvaged from a broken wooden plow can be used for many purposes. Roof supports and beams are often salvaged in their entirety for use in new structures. For smaller projects such as mallcts, tool handles, grain pounders, loom parts, etc. pollarding is regularly practiced. A branch is harvested Ochsenschlager wood, boats. bitumen and reeds Ochsenschlager from a tree by carefully breaking or sawing it off so that the tree continues to survive. Other kinds of wood, usually already sawn into more or less standard lengths and widths, are available in market towns such as Shatra, with a greater variety of size and type available in Nasiriyah. These, I am told by merchants in both places, are almost entirely imported and they are very expensive. Today most carpenters and wood craftsmen live and work near their source of supply in the market centers. The majority of wooden articles and tools used by the villagers are bought ready-made in these towns. There is still, however, a certain amount of wood use and manufacture in the villages. Mud Brick Forms Mud bricks are molded in forms made by the brickmaker from hardwood planking bought in the market towns. The forms are usually nailed, but sometimes are put together with screws. All bricks manufactured today are strictly rectangular and their size is dependent on the forms owned by the master brickmaker. As the forms are made of wood and as wood is scarce and comparatively expensive, the size of each master's bricks tend to be exact for a fairly long period of time. With only three brick makers active in the general area, it is possible to distinguish the products of each by minor variations in size. Wooden Gates, Doors and Windows Mud brick structures are most often seen in the old fortified complexes of major sheikhs. These fortifications are usually surrounded by high walls made of pist or more often of mud bricks. They are as wide as 2.5 meters on the bottom and as high as 5 meters. Entrance is gained through a strong, solid wood gate sometimes studded with large, flat-headed nails. In the two cases still extant in the area, the gates are made of imported hardwood planks joined together, perpendicular to the ground, by nailed horizontal planks with a diagonal plank between them. The hinges are mounted on the horizontal planks and attached to heavy wooden posts sunk into the ground on each side of the gateway, capped with a wooden lintel and incorporated into the mud walls. Within one finds the sheikh's dwelling apartments, a large reception room, storage facilities and dwelling structures for guards and servants. Some parts of this complex can be built of either reeds or pist, but the family dwelling apartments and storage facilities are almost inevitably built of mud brick or baked brick and have wooden doors which are often smaller variations of the outer gates. Some rooms at least have modern framed and glazed windows. These gates, doors and windows are inevitably made and installed by carpenters hired from the market towns. Merchants' sara'if and village mud brick or pist houses may have similar doors or doors made of tin nailed to a wooden frame. The mud houses may also have framed openings of wood for windows and sometimes wooden shutters. These are mostly made by unskilled local workmen who tend to use wood from local trees (preferably willow) if it is available. Columns and Roof Supports The reception hall in a sheikh's compound can be either a large and splendid mudhif or an especially grand mud brick building. If the latter, it is usually of such size as to require column-like roof supports made of smoothed palm-trunks down the center of the floor space. Occasionally a larger mud building in a village will require one or two roof supports of the same type and material. Roof Beams Roofs of both mud brick and pist buildings can be either flat or pitched and the Wood, boats, binunen and reeds beams supporting them of either wood or reed depending on the buildings' size. If of wood, the (nmks of native palm trees are generally employed. Sometimes these are squared with adzes before use, but usually they are merely smoothed. Flat roofs are surfaced with layers of loosely laid reeds, saplings and tamped mud, or layers of reed mats and tamped mud. Pitched roofs are covered with reed mats. Mallets Wooden mallets can be used for pounding rice but are most often used for softening reeds for use in making mats, baskets, and occasionally rope. The wooden handle is thrust through a hole drilled in the side of a cut-off section of a larger branch, split, and splayed open with a small wedge of wood. Handles Tool handles are usually made from a branch of the appropriate diameter. In most cases the handle is more or less shaped with an adze. The handle can be attached either by splaying one end, or by embedding the end of the handle and the particular tool in a matrix of bitumen. Grain Pounders and Mortars Among the Mi'dan a ca. 60 cm. hollowed out section of palm tree is used as a grain pounder for husking rice. Two or more women usually participate in the pounding process. Each strikes downward into the hollowed palm section mortar with her wooden or reed pounder in asynchronous rhythm. As her pounder lands, she utters a grunting sound in a different key from that used by her companion or companions. The palm tree section is hollowed out with an adze and chisel. The pounder can be made in several ways. One way is from a smaller section of palm fastened to a wooden (usually willow) handle by drilling a hole of appropriate size through the palm section, pushing the handle through and splaying it with a wedge of wood. A second way is to slightly round the thicker end of a willow branch. A third way is to shape a bitumen head on the end of a length of wood or reed. Mortars are made in similar fashion from smaller sections of palm tree. A piece of tree branch of correct length can be used as a pestle. The branch is cleaned of its bark and sometimes (but not always) rounded at the bottom with an adze. The pestle can also be made by building up a bitumen head on the end of a stick or reed. Milking Pails Pails made from a hollowed out section of palm trunk with an adze and chisel are used most often for milking both water buffalo and cows. The pail is shaped on the outside so that it diminishes in diameter from top to bottom. The conical bottom ends in a knob. Loom Parts The parts of a loom can be made from a wide variety of wood or even from reeds depending on the size of the object to be woven. The looms of the village carpet weavers are made of hard woods, but there is usually someone in every household who can make woven belts, small bags, slings, etc. Looms for these smaller objects tend to be temporary and can be made from whatever wood is available in the village. Wooden Boats and Bitumen Use of boats is rather unequal among the three groups of peoples living in the vicinity of al-Hiba. The Bedouin, migratory Arabs who spend but a few months in Ochsenschlager Wood, boats, bitumen and reeds the area at the end of the summer, the fall and sometimes early winter, never own boats. They swim their livestock across the canals when moving from place to place or if absolutely necessary rent boats from others. Their favourite transportation, the majestic and sometimes bitterly complaining camel, usually conveys them safely and dryly from canal bank to canal bank or across the shallows of the marshes during the driest season of the year. The average village household, members of the Beni Hassan tribe, may or may not own a boat of some kind. Those who have not, borrow from those who have in case of need. Those who own boats are usually fishermen, proprietors of small village stores or men otherwise engaged in trade. Bitumen covered boats are most frequently found amongst the Mi'dan where every family owns at least one. They use them constantly for transporting fodder for their water buffalo, for fishing, and for transportation. Every morning the men set out in their boats to harvest enough fodder to feed their buffalo that evening. During the spring when high water prevents the buffalo from grazing the marshes, they must redouble their efforts and if short handed are joined by their entire families. In the villages near al-Hiba, two categories of bitumen covered boats are regularly used. The mashuf is a long, narrow boat of graceful line and is made in many sizes. Today all the larger ones are usually called tarada, a name formerly applied only to a very long example of this category with rows of interior studding if it was owned by a sheikh. In the front and back it curves upward from the water line terminating in a long, tapered bow and stern. Decked areas at front and back serve as platforms on which men may stand to pull the boat through shallow water by reaching forward with a long reed or bamboo, fixing it on the marsh or canal bed, pulling the boat toward it, and giving a strong push on the pole as the taraa'a glides by. In deeper water they are paddled with spade shaped boards nailed to lengths of reed or bamboo. If two people are paddling in the same boat they both paddle on the same side. The smallest example of the mashuf category is called a mataur and is of the same general design. It lies much lower in the water, however, and is designed to carry only one man. These boats are used primarily for hunting and fishing in the marshes. They can be poled through the water in a similar fashion to the tarada by their single occupant standing in mid ship. Frequently the punter uses the reed or bamboo shaft of his fish spear for a pole. This allows him to bring the spear into play more quickly. These spears, with ten to twelve foot shafts, have five pronged, barbed, iron heads and are used for spearing fish or retrieving birds which have been wounded by gun shot from the punters gun When stalking birds feeding in the marshes, a hunter usually uses his arms for paddles while lying on his stomach. The belam is wider and has higher sides than the taraa'a. Its bow and stern terminate in fairly heavy, carved pieces of wood, sometimes of a rather fanciful design. These serve a utilitarian function and can be used for binding two boats together, side by side, for towing one boat behind another, or for pulling a boat upstream from the tow-path along canal banks. The belam is usually moved forward in open water by setting the pole in the marsh bed at the bow and moving along the side to the stern. The punter then removes his pole and moves back along the other side to the bow where the process is repeated. Constructing the Frame The center for boat building for the area is located at Chabaish and nearby Huwair where a whole village specializes in making boats. Some craftsmen are very well-known and the boats they build are recognized on sight and highly prized. The price for a new tarada about ten forearms in length is from 14 to 15 dinars. Although boats of any kind are Ochsenschlager Wood, boats, bitumen and reeds seldom made in the al-Hiba area, they are frequently so thoroughly repaired and patched as to be almost entirely rebuilt. These boats are built or repaired with salvaged parts by itinerant craftsmen whose work, it is generally acknowledged, is not up to the standards of those in the famous boat building centers further to the east. For replacement ribs they use acacia or mulberry wood as did most of the craftsmen who originally built these boats. They use any of a number of hardwoods, all of which are imported from abroad for the keel. For the rest of the boat the carpenter tends to use any size or kind of wood or wood fragments available, including the wood salvaged from empty packing cases. The keel of the boat consists of a large plank which will become a part of the outer wood surface in the finished product. To this are attached the bottom sections of the ribs which vary from 4 to 7 cm. in width, are about 3 to 4 cm. thick, and are placed 12 to 14 cm. apart. Such measurements are misleading, however, for I have never seen a boat maker measure anything. They work by eye alone. Planks are fitted on the outside of these ribs to form the bottom of the boat. If boards to stretch the entire length are not available, several pieces can be fitted between ribs. Major holes in the boards or gaps between planks can be repaired with pieces cut to fit with an adze. Even a section of local palm wood is regarded as suitable for this purpose. The nails which are made by blacksmiths out of iron brought from Basra are inevitably driven through the planks and into the ribs. The heads of the nails are slightly blunted and they are sometimes dipped in oil to help keep them from splitting the wood. Longer nails are either cut off by hammering or clenched into the wood of the rib. Side sections of rib are added when the bottom is completed and nailed precariously to the bottom section of the rib. The ground on which the boat is built, upside down, will serve to hold the ribs in place until they are secured by the side planks nailed to them. When these middle-section ribs are completely covered, the top section of ribs is added and covered in similar fashion. Although the outer boards are now nailed to the central part of the boat, they gape open at prow and stern without support of any kind. In the smaller tarada these ends are nailed to shaped prow and stem pieces of wood which project a considerable distance above the side of the boat. In the larger and wider belam, fairly large, but lightweight pieces of wood, shaped at the top to accommodate the ropes by which they are towed upstream in the canals, or behind fishermen setting out nets in the marshes, are used for prow and stern pieces. When wood of appropriate size is not available, they can be manufactured by nailing two or more smaller pieces of wood together. In both kinds of boat, an additional pair of ribs, usually of less sturdy wood than those in the main body, are then added to both the bow and the stern. The bow and stern are then often capped with flattened tin cans. The boat is strengthened on the interior with two or three braces across the width of the boat, nailed to the interior ribs as well as the exterior planking, and a series of narrow boards cut to fit over these braces are then nailed to the interior surfaces of the ribs around the top of the boat. In the past, if the tarada was intended for a sheikh, according to an informant, this wide planking around the top of the ribs was decoratively studded with rows of large, flat-headed, round nails (sometimes of bronze). No boats with such decoration or of the size reported for a sheikh's tarada (ca. 40 feet long) now exist in the area. In the belam the boards which make up the interior part of the gunwale are especially strong to bear the weight of the pullers who move it through the water. The basic framework is finished at this point, in spite of cracks in the surface and unseaworthy joins. Covering the Boat with Bitumen The same itinerant craftsmen who repair or rebuild boats Ochsenschlager Ochsenschlager Wood. boats. bitumen and re& Wood,boats, bitumen and reeds also recover them with bitumen. New bitumen, which is said to come from Hit, can be purchased in Chabaish or Basra (about 80 kilos for one dinar), but it is not often needed in large quantities for bitumen can be used again and again and is usually salvaged from an old boat (as are most of the wooden boards) to build a new one. If a boat is leaking from small cracks and the bitumen coating is still in good condition, the cracks can sometimes be sealed by heating them with a length of burning reeds and then rolling the area with a short section of reed. Ordinarily a boat is stripped of its bitumen with a hammer and chisel every year and its hull is repaired. Small pieces of wood are thinned and shaped with an adze, then nailed over the larger holes in the framework or fitted between wider cracks. The old bitumen, with whatever new bitumen is necessary, is heated to liquid consistency and stirred occasionally with a stick Sometimes a trough is made in the mud at the side of the canal. The bitumen is liquified in this trough by a fire fed by reeds in a hole dug below it from the side of the canal bank. Most often a split and unrolled metal drum suspended on parallel mud-brick or mud walls over a fire of dung patties serves as an open-air heating oven. The top of the drum coated with ashes from the fire to keep the bitumen from sticking serves as a carrying tray for the hot liquid. A shovel full of liquid bitumen is distributed over its surface like a large pancake and camed to the repaired boats. One or two workmen spread the bitumen evenly over the surface with sobay, wooden rollers with conical ends. They frequently moisten the rollers with water and pick out lumps of extraneous material or unrnelted bitumen. The exterior of the boat is coated in sections, the bottom first, next one side, then the other. Two or more layers are usually applied until the protective coating is ca. one to three cm. thick. One man quickly spreads wet mud from the canal bank over each section as it is finished, while another throws cans of water over the mud to keep it moist and pliable. The mud, according to informants, fuses to the bitumen and gives the coating extra strength. When the bottom and sides are completed the boat is turned upright and layers of bitumen are spread over the decks, the top of the prow and stem, and the edges of the boat's sides. The final stage is packing the bitumen in the interior rib joins or wherever a large crack in the wood or a poor joint seem to require it. The boat is usually allowed to dry overnight before launching. Recently a Mi'dan family whose water buffalo had developed a taste for bitumen moved to Qaria Harran. The village forced them to locate at some distance from the village on a spit of land from which the buffalo can make their way to the marsh without coming in contact with village boats. If they come upon an unprotected boat they will eat away the bitumen from keel to top and from end to end, often destroying the wood in the process. The most interesting part of bitumen application is the semi-carnival atmosphere that prevails. Smoke from a boat builder's fire serves as a beacon drawing many spectators from the nearby village or villages, among them some who need a small amount of the substance. There are many uses for small quantities of bitumen in the village households. Although each villager could purchase separately small quantities of raw bitumen and liquify it for his own purposes, this is almost never done. The itinerant craftsman, usually with one or more helpers, provides the know-how and whatever additional bitumen is needed beyond that salvaged from the job at hand. The boat owners, who pay for the additional bitumen and the craftmen's labour render assistance to him, and also moderate the distribution of small bits of bitumen for home projects. This is for them an excellent opportunity to cement friendships and their positions in the community. More important, it serves as a public forum for regulating inter and intra-village affairs by calling to account those who have violated traditional modes of behaviour. The approach of people seeking bitumen seems to be self-regulated. No more than one or two request some of the sticky substance at any one time. Only when one villager's project is completed does another take his place. Those who have acted appropriately in the past are allowed to dip into the hot liquid or have it delivered to them on a slab of tin while spectators and boat owners recall what they have done for others or the community in the past. They have inevitably brought their projects with them and squat out of the way of the boat makers to complete them. Sometimes one of the boat craftsmen will even lend a helping hand. Throughout their stay they keep up a friendly banter, to a certain extent with one another, but more especially with the boat owners or craftsmen. Those who have not behaved appropriately are also allowed to have bitumen for their needs eventually, but the tenor of conversation is much different in such cases. The individual is openly and publicly chided for poor spirit or particular unfriendly and ungenerous acts before his or her request is granted. The banter in these cases usually continues for some time and in a rather sharp vein with the spectators joining quite freely in the discussion. Even close friends or relatives, whose lack of civic spirit the boat owner might be inclined to overlook, do not escape unscathed. Spectators feel entirely free to initiate discussions of someone's behaviour. How effective these exchanges are may best be judged by imagining oneself praised for good behaviour or chastised for bad by the majority of the population among which one lives and works. The importance of this kind of verbal exchange and material sharing in building and maintaining a sense of community should not be underrated. The role of village crafts in promoting community cohesiveness has always been a large one. Individual Projects Mace Among the most interesting objects fabricated here is the mugwar, a mace used for protection when walking abroad, most often against dogs. The end of the wood or reed handle on which the mace head is to be formed is first roughened with a knife or saw and then dipped into the hot bitumen and twirled until some of the substance adhers to it. When this is cool enough to touch, it is formed into a knob, balanced and secured by rubbing between the palms of both hands. The process is repeated until the head has attained the required size and shape. The bottom of the knob, which has spilled over on the stick, is then cut away with a knife to give an even appearance, and the knob is burnished with a wet piece of reed and a wet cloth. Sometimes the knob is decorated with bronze tacks or other insets. The mugwar can also be used on people. Pole Knobs Boat poles are made from large reeds or imported bamboo. Moving any of the boats at speed puts great stress on the pole. Even the most skillful boatman is likely to fracture the more friable reed pole. As a result, the stronger bamboo pole is more often used by those who make a living from their boats. Those who use reeds select them very carefully and always carry two or three replacement poles with them except on the shortest trips. Knobs on the end of the poles are made in the same fashion as the mugwar head. Waterproofing Baskets Both the sabat and gofa can be waterproofed or liquidproofed by coating the interior with bitumen. Stands for Water Jars The kuz is porous and water evaporates from the bottom. If it stands on the ground it tends to create a mud hole. Therefore, a recess is often dug in the ground and the hole lined with bitumen. Water leakage thus tends to be contained. Evaporation from the bottom Ochsenschlager Wood. boats, bitumen and reeds of the kuz helps keep the contents cool. Grinders Corn grinders made of mud disks will grind corn but tend to abrade rather rapidly. When covered with bitumen they last longer and grind better because of their harder surface. Mortars Mortars of various sizes and shapes are sometimes made of bitumen over a sun-dried mud core. The shape varies from round to oblong, usually with fairly low side-walls from 10-15 cm. high. Pestles consist of bitumen knobs built up over one or more reed sticks of appropriate length. Most villagers pride themselves on mortars of wood or metal. Tool Making and Repair Blades and teeth of various kinds can be attached to their wooden or reed handles by means of a daub of bitumen. In some cases they are originally constructed in this fashion. In other cases they are repaired with bitumen only when other kinds of joins are weakened or broken. REEDS The marshes of southern Iraq provide an ideal environment for reeds (gramineae), rushes Guncaceae), sedges (cyperaceae) and other grasses. It is difficult, I am told, for a botanist to describe or identify the many species of these plants that exist here because of their extraordinary variety and the similarities of their basic structures. During the excavations I became quite interested in the ways reeds and rushes were used by people in the surrounding villages. Reeds are called gasab, rushes are known as bardy and sedge is kaulan. Although the inhabitants clearly know what kind of growth each plant will produce, the difference between the words 'gasab' and 'bardy' in common usage seems to be based largely on function rather than scientific botany. Young reeds can be referred to as bardy when harvested, as can certain sedges. The term gasab is usually confined to the growth of the larger, thicker plants even if in reality they are a variety of rush and sedge. This identification according to function is further born out. The Mi'dan spend most of the day foraging for fodder to feed their water buffalo at night. During the rainy season when the marshes are too deep for the water buffalo to graze, their owners (often the entire family) must provide the animals with sufficient fodder for both day and night. They harvest grasses, sedges and the young shoots of reeds and rushes. All of these are referred to as hashish, a word usually translated as grass. One should note that certain reeds and rushes are eaten by people. The tender young shoots of reed are chewed like sugar cane, to which their taste bears a distant resemblance. Young shoots of certain rushes have a kind of licorice taste, and a hard yellow cake is made from rush pollen in the spring time. The pith from yet another rush (or sedge) is eaten raw and is also cooked in a sweet pudding. Architecture Buildings Made of Arched Reeds Because of their size and architectural splendor, the grand mudhif (pl. madhaif) built by sheikhs as guest houses many years ago still dominate the horizon Ochsenschlager Wood, boats, bitumen and reeds as one approaches a village lucky enough to preserve one. But, at least in modern usage, it is neither size nor beauty of construction alone which differentiates it from other arched reed structures such as the raba (pl. rabat) or the bayt (pl. buyut). The mudhif is used solely as a guest house and for no other purpose. The raba, on the other hand, has an entrance at both ends, a partition (bench or screen) in the middle, and while one end may be used as a guest house, the other end is a dwelling. The bayt is strictly a one room dwelling but may be used to entertain guests when cleared of household paraphanalia and furbished with reed mats and carpets. Both mudhif and raba are sleeping places at night for guests passing through a village. A mudhif is quieter than a raba either day or night for it is usually built somewhat separate from the rest of the village. Construction of an Arched Reed Dwelling We were able to watch and record the building of two madhaif, for excavation use and one small raba in one of the near-by villages. In building these reed structures, the matter of size was an important consideration in terms of labor. For anything over 5 arches, we were informed, we would need the supervision of a specialist from one of the near-by villages. Candidates for the job all had fairly sizable madhaif of their own, and the details of building them had been passed from father to son in their families. The mudhif of Sheikh Mersin, who was chosen to superintend the job, was very large. The interior floor of this mudhif covered an area seven meters wide by twenty-one meters long. The eleven arches were about one meter in diameter at their base and the centers of the arch about fifteen meters above the ground. One hundred and eighteen ribs were tied to the arches as a foundation for the reed mats. According to Mersin, it took 170 men three weeks to erect this structure. Members of the village freely contributed their labor. The Sheikh furnished the material needed, provided food for the workers during the construction and a lavish banquet once the building was finished. Nowadays, with the Sheikh's authority eroded by the central government, there was less incentive for his followers to rebuild the mudhif although the older men still used it daily as a social gathering place. Furthermore, in his financially straightened condition, Mersin could no longer afford the requisite food or material. Few sizable madhaif, we were told, had been built in the area in the last 20 years, and none of these were of the size and splendor of the major madhaif of the past. Sheikh Mersin engaged seven men who would help him build our mudhif and twelve women who- would snip the reeds of their leaves and carry them from the marshes to the building site. It took six days to build a seven arched mudhif from start to finish. The ground plan was laid out with rope and stakes outlining the interior space that would result when the building was finished. All measuring throughout the building process was based on forearm lengths (dhira', pl. adhro'a) and hand spans (shibir, pl. ishbar). Reed sticks were cut to Mersin's measurements and used throughout in the measuring process. One man, carefully supervised, was put to digging the holes for the structure's main supports. The holes, a little over two forearms deep, were dug at an angle slanting towards the interior of the building. Meanwhile the other men were building simple supports, rather similar in concept to sawhorses, on which the major architectural members would be assembled. Each of these supports was made of two small bundles of reeds sunk in the ground about 4 forearms apart with a third bundle tied between them at about chest hight. Reeds for this pair of supports came from the growing pile that the women were bringing on their heads from the edge of the marsh for the building. Men had harvested them the previous day with a sickle shaped saw (mingal, pl.