Attentive readers of Layard's Discoveries in the Ruins of Nineveh and Babylon will have observed that four of the lithographic plates and a number of small woodcuts in the text are taken from sketches by S. C. Malan, and that the Preface conveys ‘grateful thanks and acknowledgments’, among others, ‘to the Rev. S. C. Malan, who has kindly allowed me the use of his masterly sketches’. A fortunate chance has lately brought the writer of this article into communication with the artist's family, in whose possession his works are still happily preserved. Mr. Walter Malan, his grandson, has very generously allowed not only examination of the drawings contained in the bound volume entitled Armenia and Nineveh, but the reproduction of those which accompany this article; the first care of which must be that of renewing to him, for a like privilege, the thanks which Layard expressed to his grandfather.

The graffiti from the tilery and pottery which the Twentieth Legion maintained at Holt, Denbighshire, included one which Professor A. Guillaume acutely recognised as written in neo-Punic script. He rendered the text as M‘QR YTN (or M‘QR YNA) that is “Ma‘qar, (son of) Yathanbaal (or Yathanmilk)”. But a close study of the graffito, undertaken by the first-named writer of this note in conjunction with Professor H. F. D. Sparks, has shown that Professor Guillaume's alternative reading is to be preferred, and may be expressed as M‘QRYN’. The second and seventh letters, here transcribed as rough and smooth breathing respectively, represent the vowels, in this case, of a Latin word. The third letter will represent hard c, as Q (qoph) is often used to transcribe this sound in non-Semitic words. The fifth letter (yodh) would stand for long i. From the photograph the exact form of the final letter (aleph) is not clear, but can be seen by close inspection. The evidence, however, of the drawing (here given as Figure 1) was confirmed by a squeeze and shows how the straight stroke not only touches but actually intersects the other stroke and thus forms a cross with terminal hook, which was a common neo-Punic form of aleph.

Animal remains from two seasons of excavation at the Tell of Abu Salabikh (1975–1976) were examined during a month's visit to the site in October 1977. A proportion of this material was then selected for further study and was kindly sent on loan to the British Museum (Natural History) by the Department of Antiquities, Baghdad. Work is continuing on these specimens and on the equid skeleton described by Postgate and Moorey which had been loaned to the Museum a year earlier. The photographs, Plates XIV–XVI, were taken on site by Mr. David Nicolson and our thanks are due to him for his patience in the face of dust and sandflies.

The collection of animal remains was large, numbering perhaps more than 10,000 elements which were mixed with an equal or greater number of human bones. The animal bones were in a very fragmentary condition as is usual with the remains of food. They had suffered further breakage by the action of the compacted soil of the floors under which they were buried as well as by disturbance and re-burial following human interments. Furthermore the very high concentration of salts in the soil contributed to the fragmentation process. Shortage of time prevented analysis of this comminuted bone but the dimensions of every measurable element were recorded.

The Near Eastern seal collection of the Gulbenkian Museum of Oriental Art, University of Durham, had a single immediate origin. It was part of the collection of antiquities belonging to the Dukes of Northumberland, formerly housed in Alnwick Castle. Most of the Egyptian antiquities were acquired by the 4th Duke during the first half of the 19th century, and in 1880, at the expense of the 6th Duke, a printed catalogue of the whole collection as it then existed was issued, compiled by Samuel Birch, Keeper of Egyptian, Assyrian and Oriental Antiquities in the British Museum. This describes four Near Eastern cylinder seals, nos. 2016–2018 and 2020, a very modest start. The greater number of these was collected after this time by the eldest son of the 7th Duke, who first came to public attention as Lord Warkworth, but was later styled Henry Algernon George, Earl Percy. He was born in 1871, travelled extensively in the Near East in 1895, 1897 and 1899, and died prematurely in 1909. As noted in the Dictionary of National Biography, Supplement 1901–1911, 904, “true to the traditions of his family, he began to collect antiques, particularly cylinder seals”, and significantly an Akkadian cylinder seal design adorns the front board of his book Notes from a Diary in Asiatic Turkey (London, 1898). This material joined the family collection, and it appears that he was the last of his line to pursue this interest, so that nothing was added after 1910. The collection remained in Alnwick Castle until the 9th Duke deposited it on loan in the British Museum in 1939. The easier access which this move was intended to create was, however, prevented by the outbreak of war. In 1950, when conditions had returned to normal after the war, thanks to the generosity of the present Duke and the initiative of Professor T. W. Thacker of the University of Durham, the whole Northumberland collection, with a few exceptions, was acquired for that University, where it is now housed in the Gulbenkian Museum of Oriental Art, close to its ancestral home.

The first two seasons of excavation by the Iraq Directorate-General of Antiquities at Eridu (Tell Abu Shahrein) lasted from December to March in 1946–47 and 1947–48. They were directed by Professor Fuad Safar, with myself as Technical Adviser. A third season was already in progress when I left Iraq early in 1949, to take up a new post as Director of the newly founded British Institute of Archaeology at Ankara. Preliminary reports on the excavations were published by Professor Safar and myself in the journal Sumer, as will be seen from the short bibliography below, and they were followed by various specialist articles and commentaries. The final publication has raised some more complicated problems, and, to judge by the many enquiries on the subject which Professor Safar and I myself have received, it seems probable that archaeological colleagues and others may be interested in the steps more recently taken to make this important material generally available.

In 1965 Roman Ghirshman drew attention to a particular element of Elamite architecture which he called the “salle à quatre saillants”: this is a rectangular room with four pilasters placed near the short walls and a large doorway in the middle of its long side, opening on to a courtyard. It must have served as the reception, audience and throne hall. Ghirshman identified this architectural form in successive Middle Elamite levels on the Royal City mound at Susa in houses belonging to officials of the Sukkalmahhu, and so it may reflect Royal Palace designs, as do the Neo-Assyrian Residences at Khorsabad. He has dated these levels from c. 1800 to c. 1300 B.C. He also noted its reappearance at Susa in the Palace of Darius. Here it is slightly modified: there are two “salles à quatre saillants” placed side by side with doorways piercing through them both (see also fig. 11). It would be possible to maintain that the double and single “salles à quatre saillants” are distinct but this has neither chronological nor geographical significance.

The eight centuries between these buildings cannot yet be filled with more Elamite examples of the “salle à quatre saillants” because very little Elamite architecture has been discovered. It does not appear in the funerary palaces at Choga Zanbil and has not been found on the Iranian plateau at this period. The Kassite palace at Dur-Kurigalzu shows no trace of it, and Geoffrey Turner in his recent study of Late Assyrian palace plans did not draw attention to this feature.

Ashmolean Museum inv. 1937. 993. Photo: P. R. S. Moorey, Ancient Iraq (Assyria and Babylonia) (Oxford, 1976) 47, plate XXX (reverse only). Cf. Plate IX (enlarged). An originally unbaked clay tablet: W., 8·2 cm, h., 6 cm, th., 1·4 cm (maximum).

The tablet was in the possession of Stephen Langdon and came to the Ashmolean as part of Langdon's bequest. No Kish excavation inventory number is scratched on the edges as on material from Kish. There is no reason therefore to assume Kish as the provenance. No exact provenance is in fact available for the tablet. The tablet is inscribed in Greek script on both sides. It was referred to but not published by Sollberger in his important collection of the fourteen known fragments of Babylonian texts in the Greek alphabet, the “Graeco-Babyloniaca” texts. To this collection can now be added the Harvard tablet, which contains part of an incantation inscribed in cuneiform on the obverse and Greek script on the reverse.

The Ashmolean tablet differs physically from these in two respects, (a) Size. It is considerably smaller in all dimensions than the other tablets in the British Museum with Greek script which Pinches estimated to have been about 5 in wide by 9 in high when complete. Their thickness is also greater at 2–3 cm. (b) Oblong shape. Apart from tiny chips the other fragments (including the Harvard text) are from tablets which had the usual shape of Babylonian tablets, height greater than breadth. It is like the other tablets in being unbaked. The size and shape of the Ashmolean tablet are, however, typical for Seleucid period cuneiform business documents.

This article is based mainly on archaeological surveys of the Iraqi Directorate General of Antiquities between the years 1938–1965. In preparing this study, the writer examined more than 7,000 files in the Directorate's archives. The reliability of the report on each site has been carefully assessed. The numbers following the site names indicate serial numbers of relevant files in the Directorate's archives. The name of the district (Qadha or Nahiya) in which a site is located usually follows the serial number.

Examination of the available surface material from all over Iraq, suggests that Uruk pottery is more popular in the north than the south. It is not possible at present to tell whether this fact is significant, as the existing southern surveys are far from complete. Moreover, in the south there is the added problem of silting and subsidence which would have tended to obliterate sites, in particular small ones, of Uruk and earlier dates. However, where extensive surface surveys have been carried out, as, for example, in the Diyala and Nasiriya Liwas, and now in the Eridu and Warka regions, a number of Uruk sites has been found. In spite of the smaller number of recorded sites, one is inclined to treat the south as the focal area at this time, owing to the social and urban development which is known to have taken place there, and for which we have no parallels as yet in the north. However, only further excavations in the north and further survevy in the south can satisfactorily resolve these problems.

Despite scholarly efforts that now extend over more than a century, the governing principles of Akkadian verse remain elusive. It is obviously not based, like Greek or Latin verse, on the counting and measuring of syllables. The idea that it is based on the counting of accentual peaks has a much greater immediate appeal. There are long stretches of poetic texts that seem amenable to analysis on these lines. Most scholars would be willing to recognize the existence of a “standard” Vierheber, a verse with four apparent accentual peaks, giving the sense of a balance of two against two, as in Enūma eliš I 47–50:

īpulma Mummu / Apsû imallik,

sukkallum lā māgiru / milik mummīšu:

“Hulliqamma, abī, / alkata ešīta;

urriš lū šupšuhāt, / mūšiš lū ṣallāt.”

This measure may be called “standard” because it occurs at all periods, and in many texts it predominates. But everywhere we find shorter lines interspersed, on no discernible principle, and often longer ones too. The shorter lines generally have three apparent accents, but on occasion only two, while the longer ones may have five or six. According to A. E. Housman,

To think that two and two are four

And neither five nor three

The heart of man has long been sore

And long 'tis like to be.

But what vexes the Akkadian metrician's heart, if he has one, is

To think that two and two are four

but sometimes five or three.

In a recent paper published in the journal Climatic Change, we put forward the hypothesis that drought and overpopulation played an important, if indirect, role in shaping the sudden decline of the Assyrian Empire during the mid-to-late seventh century b.c. This argument was partly predicated on five paleoclimatic proxy records for conditions in different parts of the northern Near East during the first millennium b.c., each of which indicates that relatively arid conditions affected much of the region during the seventh century b.c., especially during its middle decades. Here, we revisit the textual and paleoclimatic proxy evidence for a period of drought in more depth to examine whether this evidence does in fact support the climatic component of our hypothesis. In this paper, we show that the available proxy evidence supports the notion that there was some kind of regional climatic perturbation that affected much of the Near East during the latter half of the seventh century b.c., which caused conditions in many parts of the region to become more arid. The strongest signal for this short-term episode of aridification, which we have termed the “Late Assyrian Dry Phase,” is observed at approximately 650–600 b.c. These proxies thus corroborate and provide the background for the Neo-Assyrian textual evidence for drought during the mid-seventh century b.c.

Alkaline glazes were first used on clay-based ceramics in Mesopotamia around 1500 B.C., at the same time as the appearance of glass vessels. The Roman Empire used lead-based glazes, with alkaline natron glass being used only to produce objects of glass. Chemical analysis has had some success determining compositional groups for Roman/Byzantine/early Islamic glasses because of the discovery of major production sites. Parthian and Sasanian glass and glazed wares, however, have been found only in consumption assemblages, which have failed to inform on how they were made. Here we reanalyse compositional data for Parthian and Sasanian glazes and present new analyses for Parthian glazed pottery excavated at the early third century A.D. Roman military outpost of Ain Sinu in northern Iraq. We show that some Parthian glazes are from a different tradition to typical Mesopotamian glazes and have compositions similar to Roman glass. We propose that Roman glass was recycled by Parthian potters, thereby suggesting that as yet undiscovered Mesopotamian glass production centres ordinarily supplied glass for indigenous glazed pottery. Furthermore, if recycling glass to make glazed pottery was extended to indigenous glassware, this may provide an explanation for the paucity of Parthian and Sasanian glass in the archaeological record.

The Sippar Library tablet IM 124485 is a new source for Tablet XX of Enūma Anu Ellil (EAE), the great compilation of Babylonian celestial and meteorological omens. The twentieth tablet of the series, which deals principally with lunar eclipses on the fourteenth day of each month of the year, was edited by Francesca Rochberg in 1988 along with all the other tablets of lunar-eclipse omens in EAE (Rochberg-Halton 1988: Chapter 10). Rochberg was unable to report the whole text of her MS M = ND 4357, a Neo-Assyrian tablet from the library of the temple of Nabû at Kalaḫ; it can now be consulted as CTN IV 5 (Wiseman and Black 1996: Pls. 5–6, 145), though the copy of the reverse is inadequate. In addition a Late Babylonian exemplar of a further commentary, written in the time of Philip Arrhidaeus for the scholar Iqīša of Uruk, has come to light in W23300 (now IM 75990), published as Uruk IV 162 (von Weiher 1993: 103–5, 186). Despite these additions to knowledge, some of the text of EAE XX remained poorly enough preserved to make the discovery of a new manuscript very welcome.

The new tablet allows seven sections of the text of EAE XX to be reconstructed in full, and our understanding of the technical terminology refined as a consequence. The chief interest of this tablet of EAE emerges more clearly than before. The common denominator of the twelve lunar-eclipse omens of EAE XX is eclipses that, at least notionally, set in “above” and clear “below”, as observed in 1. 66 of the tablet published here. However, the observed phenomena that especially distinguish the protases of EAE XX from those of other calendrical lunar-eclipse tablets appear to be particular to partial eclipses. The progress of the eclipse to a point at which the disk is half eclipsed (imšul) or more (eli mašāli illik) is explicitly recorded on six occasions (§§1.2, IV, V, VII, VIII, IX). The portents relate either to the moon's “emblem” (šurinnu), a term that signifies the moon in eclipse (§§1.1, IV), to its “horns” (qarnu), i.e. the cusps of the partially eclipsed disk (§§VIII, IX, X, XI, XII), or to both (§§V, VII). It seems that what the compiler of EAE XX considered most portentous were the appearance, behaviour and other aspects of the lunar disk while the moon was half, or more than half, eclipsed.

The following texts come from the Sippar Collections of the British Museum and offer substantial new data on numerous aspects of the administration of the Ebabbara, particularly as regards crown relations and privatised functions. Among the more important points are:

(a) Text No. 1, a fragmentary text dealing with the gardeners of Šamaš, mentions a royal official (rēš šarri) in charge of the gardeners and a letter (šipirtu) from Nabonidus in Tema, the first evidence for that king exercising direct control over the Ebabbara from his Arabian capital.

(b) We also have definite indication that certain official positions connected with the collection of dues payable to the temple could be hired out. Specifically this involves the collection of tithes (No. 8) and the office of gugallu (No. 7). The occurrence of the ša muhhi ešrê and gugallu in letter orders had already prompted speculation that these two duties were rented out in the same manner as the rab sūti and this is now shown to be the case. Other examples of privatised functions are a contract to supply baked bricks to the temple (No. 6) and an undertaking by a širku (temple slave) to supply some of the material requirements of the offerings (No. 2).

(c) Hints of temple affairs outside the Sippar region are given by the appearance of the šangû and the rab ummāni in Opis (No. 4). The text concerns sheep, though whether they were there for pasture, sale or provisioning temple personnel is not clear. The sheep coming from Rusapu were part of a cycle in which temple flocks might be pastured far away with animals rotated as required (Nos. 11, 12). Sheep belonging to the temple are also found on the road in No. 10.

In his Anabasis recounting his march on Babylon with the Persian prince Cyrus in 401 B.C., Xenophon provides his readers with many descriptive details about the area that would come to be called Mesopotamia. He speaks of villages, of wheat, barley and dates, of fields under cultivation and of many other indications of settled life. He also offers a brief description of the flora and fauna of Mesopotamia, in which he mentions wormwood, aromatic shrubs, wild asses, bustards, gazelles and ostriches. Musil, who travelled extensively in this region at the beginning of the twentieth century, commented favourably on the accuracy of some aspects of the picture drawn by Xenophon.

Yet, for all the verisimilitude of Xenophon's description, anyone who has travelled in Mesopotamia, or who is familiar with the accounts of it found in later classical authors such as Strabo or Pliny, cannot help but feel that something is missing in Xenophon's account. What is missing is any reference to pastoral life in the region, particularly the camel pastoralism long associated with Arabia and Arabs. Sheep are mentioned, in passing, only once in the sections of the book devoted to Xenophon's passage through this area, in association with the villages of Cyrus's mother, Parysatis, situated along the upper Tigris, and there is no mention whatsoever of goats or camels, or of nomadic pastoralists who might have herded flocks of such animals. This dearth of positive evidence for nomadism in Xenophon's account might tempt one to conclude that pastoral nomads—at least Arab nomads of the kind described by Strabo and Pliny—had not yet arrived in this area in 401 B.C. But Xenophon himself thwarts such a ready conclusion by calling the part of Mesopotamia lying east of the Khābūr river “Arabia”, a name which in itself implies the existence of Arab nomads.

The original meaning and the changes in the content of the term Amurru in the course of the second millennium B.C. have been one of the most intensively discussed and debated subjects of Near Eastern history. The name appears from the late third millennium on, both in phonetic spelling and in the logographic writing MAR.TU. It is now generally agreed that in the early part of the second millennium the term refers loosely to the Syrian regions situated west of Mesopotamia proper and to the inhabitants of these regions. The more difficult questions under debate are concerned with the original meaning of Amurru, whether ethnic or geographical, and with the exact nature and origin of its equation with MAR.TU. These important questions however are of little concern for the issue discussed in this paper.

The broad sense of Amurru as indicated above, often used in opposition to Akkad, is gradually restricted to a more limited region of central and southern Syria. A Mari text has the “Land of Amurru” in a sequence after the lands of Yamhad and Qatna, whereas another text mentions “the messengers of four kings of Amurru” alongside the messengers of Hazor. Texts from Alalakh VII and IV mention individuals from the “Land of MAR.TU”, without however any clues to its whereabouts.

During his second expedition to Assyria, October 1849 to April 1851, Layard excavated extensively at both Nimrud and Nineveh, as in his first campaign of November 1845 to June 1847, but now principally in Sennacherib's palace on Kuyunjik. The two London institutions, the British Museum and the British Library, house a rich corpus of primary material, in the form of drawings, sketches, plans, notebooks, diaries and journals, letters and other manuscript documents, which not only add much information and detail to our knowledge of Layard's achievements, as published in his second book, Nineveh and Babylon (1853), but also give a vivid and colourful picture and almost day-by-day account of his excavations. These sources are here described and discussed, and in a subsequent article will be used to trace the course of Layard's second campaign at Sennacherib's palace on Kuyunjik.

Since this article deals for the most part with the primary sources and only secondarily with Layard's published account, Nineveh and Babylon, the rooms and courtyards of Sennacherib's palace are here referred to as Chamber A, etc., as found in these sources and also in Nineveh and its Remains (1849). It was only in Nineveh and Babylon that Layard changed to the system of designating rooms by Roman numerals (Fig. 1). Concordances for converting the chamber or room letters to numbers and vice versa will be found on p. 213.

Between 1927 and 1930 a joint expedition by the University of Oxford and the Field Museum of Chicago to the site of Kish opened a series of soundings (ZY, Y, YW and YWN) on the mound of Tell Ingharra in order to investigate the earliest periods of the settlement. This paper presents a detailed analysis of the YWN sounding utilizing previously unpublished documentation and excavated objects, and offers a new interpretation of the evidence. In addition, a new radiocarbon date assigns the latest phase from the sounding to the ED IIIb to Akkadian transition and it can be compared usefully with existing dates from several central and north Mesopotamian sites.

In the issue of this journal for Autumn, 1956, I published under the title ‘Two medical texts from Nimrud’ a fragment from the Nābû temple library which had been excavated the previous year. The piece is numbered ND. 4358, and in the issue referred to it was described as follows: ‘This text is a catalogue, perhaps sufficiently important to be called the “Nimrud Catalogue”, of a series “sa-gig”. The series in question we have not yet learnt to call by this name, but it is otherwise well known as enūma ana bīt mar marṣi āšīpu illaku.’ As to its physical description, it was stated that the fragment was ‘the lower half of a single-column tablet, obverse and reverse thus providing a continuous text which must represent about half the original.’

It is the purpose of the present communication to announce that the missing portion of the tablet has been identified.

The new piece is ND. 4366. It came to my notice through a copy made by D. J. Wiseman at Baghdad towards the end of 1959, and I am grateful for Wiseman's generous consent that I should edit the text in this publication. The text was collated by me in the course of a personal visit to Baghdad in the Spring of 1961, and I was then able to confirm my suspicion that it in fact physically joined ND. 4358. There is some loss of material at the join and down the left edge, and the quality of the surface rapidly deteriorates as the end of the reverse is reached, but every line of the tablet is now represented in some part at least. The copy will be included in Wiseman's forthcoming volume of texts from the Nābû temple library, and thus in the present edition the text is conveyed in transliteration only.

The publication in 1955 of the fourth volume of Sir Leonard Woolley's Ur Excavations series set a coping-stone upon the remarkable edifice of recorded research which he and his collaborators in the “Combined Expedition” have achieved. This volume deals finally with the pre-Third Dynasty discoveries, other than those connected with the Royal Cemetery. But, having in the sequence of publications given precedence to Volume V (The Ziggurat and its Surroundings), its appearance has been long delayed, and, as explained in a foreword, it has now proved necessary to publish the original text as prepared for printing in 1935, almost without reference to the bearing which subsequent work in the same field has had upon its contents. Clearly its usefulness will be greatly enhanced if this omission can gradually be repaired; and though an attempt can hardly be made to do so with a standard of thoroughness comparable to that of the original work, the following notes are intended as a small contribution. They concern the discoveries connected with the Al ‘Ubaid period at Ur and the light thrown upon them by subsequent finds, particularly at the sites of Tell ‘Uqair and Tell Abu Shahrein (Eridu).

The bronzes illustrated on Plates XXXVII, XXXVIII, 1, 2, and XXXIX, 4 are all unstratified and the provenance of only one of the daggers is known. This is illustrated on Plate XXXVII, 1 and comes from a site above Lahijan in the Elburz mountains east of Resht. It was presented to the Institute of Archaeology by Mrs. Iris Ainley. The dagger is 35 cms. long and the lunate pommel measures 5·9 cms. in width. Plate XXXVIII, 4 is in the United States National Museum, Smithsonian Institution, Washington (No. 422.991), and is published here by courtesy of the Associate Curator, Mr. G. W. Van Beek. It measures 38·4 cms. in length and the pommel is 4·1 cms. in width. Both the daggers (Plates XXXVII, 1 and XXXVIII, 4) show the distinctive features which relate them to daggers from the region of the Talish, where tombs excavated in both the Russian and Persian areas by De Morgan in 1890 and 1901 have yielded many comparable examples. Plate XXXVIII, 4 can be compared to my type 43 (Iraq VIII, 1946, Pt. 1, Pl. V, 43). This type cannot be included among the common types of Talish bronze dagger, but a dagger from Hiveri, published by De Morgan, has a solid hilt with three circular ribs and the type example from Veri, with one rib round the solid hilt are both comparable. A dagger from Luristan with hilt similar to the Veri example is also published by Godard, and again the skeuomorphic moulding representing the thong which bound hilt and blade together on earlier blades is visible. But, as Schaeffer has pointed out, the cemetery at Veri must have lasted for a considerable period. Most of the tombs can be assigned to his “Talish récent II” c. 1450–1350 B.C. (dated by analogy with comparable material in tombs at Hassan Zamini and Agha Evlar which contained seals of Kirkuk style) but there is an earlier group contemporary with Khodja Daoud Köpru (Talish récent I c. 1550–1450 B.C.) in which the swords are all narrow rapiers with tangs.