Two six week programmes of excavation took place in July and August 1974 and 1975 at Sidi Khrebish on the site of part of the Hellenistic and Roman city of Berenice. Large scale excavation ceased in September 1973 and the work of the last two years was directed towards the completion of areas unfinished in 1973 together with a certain amount of new excavation.

The major discovery of the last two seasons has been a section of a two-period circuit of defensive walls, including a tower (Figs 1 and 2). The walls are Hellenistic in construction and from the associated dating material appear to post-date Queen Berenice's initial fortification of the city which was founded in the middle of the third century B.C. They are possibly as late as the first century B.C.

Many interesting architectural fragments came to light during the excavation, but in few cases are they sufficient to give an overall picture of any one building. Some suggest stylistic links with Tolmeita, although the Berenice versions are generally coarser and smaller in scale. For example Berenice has produced several variants of the doorway in the East portico of the Roman Villa at Tolmeita, but the forms are simpler and the scale smaller. An arched doorway which led from the street into the ‘L’ peristyle is closely similar to ones at Tolmeita.

Only the peristyle of ‘L’ can be confidently restored to its full height (Fig.3), but this building alone tells us a great deal about the character and influences of Berenice architecture. Built about 75 A.D. its order is Doric, for long a favourite in the province, and the columns are composed of two unfluted shaft drums and a capital cut from a separate block. No column is complete, but the fragments suggest a height of around 2.50 m. This only gives a lower diameter to height ratio of 1:6.5, but other elements too suggest that the building was rather squat. Above the columns is an architrave bearing a monumental inscription, and a frieze with triglyphs offset so that those over the columns could be carved from a single block instead of the normal two. To judge by the slots at the back of the frieze blocks the peristyle was roofed with wooden beams on which were laid poles or planking sealed with plaster or mud, a style which Wright terms “universal in the Near East during all ages”. After the building was abandoned the cornice blocks were dislodged so that the roof beams could be extracted, a demonstration of the scarcity of wood in the province. Good building stone was scarce too, as is shown by the widespread reuse of stone on the site and the extensive use of mud-brick. The back wall of the ‘L’ peristyle is stone-built to a height of about 1.30 m. and above that in courses of thin unbaked bricks whose dimensions conform to the wall thickness.

During the emergency excavations of 1971, a large quantity of painted wall plaster was recovered from the peristyled house ‘H’.

It is thought that this house was built in the late first century A.D. and destroyed in the third century A.D. The plaster was found in the destruction levels and lay in a dump or spread to the south of the courtyard area of the house. It was mixed with, and probably to some extent sealed by, mud-brick debris. This may account for its good state of preservation.

From its context, it appears that the plaster decorated two or three of the main rooms of the house, probably including the room on the south-west which has a fine mosaic floor. Traces of red painted plaster can still be seen near floor-level in the room on the south-east corner. The plaster was probably in situ at the time of destruction.

Unfortunately, much of the plaster was lost during the necessarily hurried emergency excavations, and what remains is often broken into small fragments. However, on the whole, both the plaster backing and the painted surface are in good condition. Work on the reconstruction was begun in 1971; it has continued during the season of 1972, 1974, 1975.

Preliminary examination of the plaster itself shows that it is of good quality. The colours appear to have been applied using both the fresco and tempera methods; the main pigments are haematite (dark red), yellow ochre, green earth, lime white and Egyptian frit (bright-blue). A full technical study of the plaster and pigments is still to be undertaken.

Fragments of a number of different painted motifs remain, including figures, foliage, inhabited scroll and varied border designs. In one room at least, the decorative scheme appears to consist of a number of panels outlined with multiple borders. Designs within the panels show architectural perspective and figures. It has been possible to reassemble the fragments of four similar architectural panels and these appear to be linked (by multiple borders) to the panels containing figures.

In two rooms of the courtyard house (no. H) on the Sidi Krebish site excavated by this society, the wall plaster carried a large number of graffiti. Obviously this must mean that the rooms were open to public use in some sense at the time of writing and, as will be seen, there are grounds in the texts for suggesting that the house was then used as a restaurant or more probably a hotel.

1. In 1936 Professor Gaspare Oliverio published an inscription cut on a sandstone stele found re-used in one of the Churches of ancient Taucheira. The apparently mundane text has attracted no attention. In 1959 R. G. Goodchild initiated serious excavation of the building concerned by the Libyan Department of Antiquities and in the course of work supervised by Mr. Breyek Atiyah another inscribed sandstone stele was found very close to Oliverio's. It became apparent that the two adjoined and that the resulting text is of some interest.

Two rectangular sandstone blocks forming part of a composite monument (together width 0.72 m × height 1.17 × depth 0.40), a damaged at the upper righthand corner, b along all edges; there is extensive pitting on the face, some of which had already occurred before the stone was inscribed.

Letters, II–III cent. A.D.: 1.1, 0.06; 11.2–5, 0.05–0.06; ε, ω; Y cut within 0 in 11.3, 4; the cutter has spaced so as to avoid holes in the face.

G. Oliverio, Documenti Antichi dell'Africa Italiana II (Bergamo, 1936), no.288 and pl.LXXII, fig.29, whence SEG IX, 544 (righthand side only).

Photos, Dept. F.1078 (righthand side only); JR Tocra III. 51 (E. Alföldi-Rosenbaum). See plate 1.

It has been known since 1954 that a pottery kiln site is situated about 100 metres to the North West of the West Gate of Roman Tocra, in the area known locally as ‘Howa El-Faquar’ (Area of Potsherds).

While looking for tombs in the East face of the small quarry in this area (see Fig.l), Wright discovered in 1954 part of a wastage dump from a kiln. The general direction of the tip lines led him to consider that one of the kilns was situated on a ‘small but salient mound’ (see Pl.1) nearby. He published a selection of pottery from the dump and, after some discussion, based on lamp evidence, he placed the date of the group at around A.D.100.

As this area is now outside the area controlled by the Department of Antiquities and is being ploughed by local farmers, it was thought expedient to examine the area through limited excavation for several reasons. A date of around A.D.100 seemed too early in the light of a study of the coarse pottery from the recent excavation at Sidi Krebish in Benghazi: a late second or early third century A.D. date seemed more appropriate. It was hoped that further excavation at Tocra would shed more light on this. In addition, few kilns have been excavated anywhere in North Africa to date and the potential of the Tocra site for structural and organisational information about the pottery industry is very great. It was also considered an excellent opportunity to study in detail a large group of contemporary pottery and to quantify the relative proportions of the different types.

Permission was kindly granted by Mr. Masoud Shagluf, Controller of Antiquities for Western Cyrenaica, and a two week excavation took place in August 1974.

The purpose of the present study was to measure the trace element distribution of waster pieces of coarse ware found on the pottery kiln site associated with Roman Tocra. These would then be compared with other coarse ware sherds found in Cyrenaica.

Neutron activation has been chosen as the method of analysis because several elements can be determined simultaneously, and also many samples can be analysed routinely.

Samples from any sherd were obtained as follows:

The sherd was brushed lightly to remove any foreign matter. It was then sliced with a diamond-coated saw into pieces about 5 mm × 5 mm × 2 mm all surface material being removed in the process. Each piece was washed quickly in dilute nitric acid (ANALAR), rinsed in distilled water and finally boiled for 2 minutes in distilled water. The samples were dried in an oven at 120° C. During all procedures polythene gloved hands and clean tweezers were used for manipulation.

Each sample was wrapped individually in aluminium foil and labelled with a heat resistant indelible marker. All the samples were placed in an aluminium can together with suitable standards for irradiation.

In this investigation the elements determined were scandium and iron. The oxides of these elements are commercially available in a very pure form (SPECPURE).

A known weight of the oxide was dissolved in hot concentrated ANALAR acid, hydrochloric acid for iron and nitric acid for scandium, and diluted to 25 ml in a graduated flask. The resulting solutions were about 1.5 mg per ml for iron and 0.15 mg per ml for scandium.

Much interest has centred on the development of the fossil water resources at Kufrah in the central Sahara of Libya. The remote location and extremely arid climate presented unusual problems to irrigation engineers, agriculturists and stock specialists. The favourable estimates made by consultants indicated that the high quality of the water together with the larger quantities available would easily support 10,000 hectares of irrigated alfalfa for steep breeding and the groundwater resource would only be drawn down by 35 metres after 40 years of pumping (Tipton and Kalmbach — 1971). This optimistic phase was at its height in 1973 when the political leadership of the country was pressing the managers of the Kufrah scheme to envisage not just 10,000 hectares of irrigated cropping, but extensions to 50,000 hectares, preferably including settlement schemes.

In the event, however, faced with the decision of how far to extend the scheme, those responsible approved designs for the limited expansion of the project in the form of a 5500 hectare settlement scheme accommodating about 850 farmers. This sharp reduction in the target area for irrigated farming at Kufrah arose because of the results of the first effective year of operation of the scheme, results of which became available during late 1973 and early 1974. The most important indicator of the long term viability of the project was of course the drawdown in the level of the groundwater. Uneasy predictions were being circulated as the result of some important work completed by a Libyan hydrologist (Adib 1974) who had applied a different model from that used by Tipton and Kalmbach in their feasibility study. These new predictions proved to be very close to the measured rates of decline after a year of pumping, a year incidentally in which the scheme had worked at only about 70 per cent or less of its anticipated levels of water use for full fodder and livestock production. Instead of declines of a few metres which would in the early years have been consistent with a 35 metres decline over 40 years of pumping, it was discovered that in parts of the well field drawdowns of as much as 15 metres had been recorded after only one year. Critical conclusions were reached by Pallas and Fadel (1974) on the basis of these data, especially regarding the risks involved in investing heavily on the basis of contested assumptions concerning an aquifer, and they urged that important groundwater developments should be preceded by at least two years of pilot experimentation.

The archaeological remains of Libya, prehistoric as well as Classical, have long served as guides to the age of the landscapes and deposits with which they are associated. To give a familiar example, the elevation of the ancient harbours of Apollonia and Lepcis Magna relative to present-day sea level is used in determining the extent of submergence or emergence since Antiquity. Other instances of archaeological dating may be found in the studies devoted to the Quaternary geology and geomorphology of Libya which were briefly reviewed in the Society's Fourth Annual Report (1972–3, pp.9–10). Hitherto, archaeological dating has been applied to substantial portions of the landscape in order to establish when they were formed, destroyed, deformed or displaced. It is now beginning to prove fruitful in investigating the history of individual sand grains, although still with the aim of evaluating gross environmental changes.

The items whose age is here at issue are identified with the help of the Scanning Electron Microscope (SEM), whose application to the study of quartz grains is primarily associated with the name of D. H. Krinsley of the City University of New York. The SEM allows grains to be examined at a wide range of magnifications with all the advantages to be gained from a great depth of focus. Plates 1–4 illustrate the kind of detail revealed by the SEM on grains from deposits in Wadi Ganima, Tripolitania.

Speaking of Tripoli, the fourteenth-century writer al-Tijāni says: ‘when the Zughba killed Sacīd ibn Khazrūn in the year 429’. Sacīd was the Zanāta lord of the city, and al-Tijānī is explaining that the venerable faqīh Ibn Munammar was driven into exile because, when Sacīd was killed, he opened the gates to a cousin of the dead man who was promptly expelled by a brother. The reference to the Zughba is thus contained within an account of the dynasty of the Banū Khazrūn which is itself contained within a biographical notice of the jurist Ibn Munammar. There is no further explanation of the killing, and no mention of it by any other writer apart from Ibn Khaldūn, who has borrowed the story from Al-Tijānī, and repeats it twice. This presents a problem. The Zughba were Arabs, a branch of the Banū Hilāl. In that case they should not have been at Tripoli at all, as Ibn Khaldūn observes, because the Banū Hilāl did not enter Ifrīqiya until they were sent from Egypt to punish the Zīrīd sultan al-Mucizz ibn Bādīs for breaking his allegiance to the Fāṭimid caliph in Cairo, and recognising the cAbbāsid caliph in Baghdad. That would have been some time after 1048, more specifically, after 1050, when the man who is usually held responsible for the invasion, al-Yāzūrī, became the Fāṭimid wazīr. The story of the invasion at the behest of al-Yāzūrī is accepted by the principal modern authority, Hady Roger Idris, who tries to explain the reference to the Zughba at Tripoli in 1037–8 by suggesting that a fraction of the tribe may have accompanied their relatives the Banū Qurra, a people settled to the west of Alexandria, when these were sent by Cairo to support an attack upon the Zīrīds of Ifrīqiya by an Ifrīqiyan pretender, Yaḥyā ibn cAlī ibn Hamdūn, in 1001–2. The Banū Qurra had helped Ya…yā to attack Tripoli, but had then returned to Egypt. Idris wonders if this hypothetical fraction of the Zughba had not stayed behind in the neighbourhood of the city.

Misurata is the second city of Western Libya (Tripolitania) and, with its surrounding farms and villages, had some 83,000 inhabitants in the early nineteen-sixties, but has grown rapidly since then. West of the city is a narrow but fertile plain coming from Zliten between the dune-encumbered coastal ridge and a line of low hills which are the last eastward vestiges of the Tripolitanian Jebel and which come to an end shortly before Misurata. Twelve kilometres to the east, at Gasr Ahmed, the coast turns south-eastwards to form the western side of the Greater Syrtis, or Gulf of Sidra. Between Misurata and Gasr Ahmed the cultivated area narrows to become a kind of peninsula between the coast and the marshes of the Sebkha Taworga which are extensively flooded during the winter. South of Misurata, along the main road to Fezzan and Cyrenaica, the country becomes increasingly arid. East of the road cultivation soon gives place to mud flats and salt marshes, the successive sebkhas Taworga, Hishah and Al Awenat, which together cover an area some hundred kilometres long and twenty broad.

The name Misurata (Misratah) is Berber and did not come into the district until the tribal movements in late antiquity, when the Hawara spread into Eastern Tripolitania. According to tradition the Berber hero Aurigh had four sons, Calden, Meld, Hawar and Maggher. All their descendants came to be designated collectively as the children of Hawar, and among these were Wurfel and Misratah, sons of Meld. The sons of Wurfel have dwelt ever since in and to the south of Beni Ulid, and the Misratah have occupied the coastlands to the northeast. Ibn Batuta mentions passing through the country of the Misratah in 1326. Ibn Khaldun says that in his days (14th century) the Misratah were very powerful and only paid a very small tribute to the Arabs, ‘tribute which they have the air of handing over by condescension.’ He says ‘they occupy themselves by commerce, and travel frequently to Egypt and Alexandria, to southern Tunisia and to Fezzan.’ The inhabitants of the Misurata area to-day include groups of Berber origin mixed with the descendants of the first Arabs who came into Tripolitania; others, descended from Turkish immigrants; and yet others claiming Sherifian origin, that is, descent from the Prophet Mohamed himself.