In September, 1947, whilst digging trenches for the laying of the foundations of a building in the field known as ‘ta Trapna iz-Zghira’, in the parish of Zebbug (Ref. 394228 on the standard 2-inch map of Malta), five early tombs were discovered. The workmen, not appreciating the importance of the discovery, cut a trench 2 feet wide through the middle of two of these burials, thus partially destroying the deposit. Further destruction was prevented by Mr. Ian Small, Civil Engineer of H.M. Dockyard, who stopped the work and reported the discovery to the Museum; and on inspection of the site it immediately became apparent that we were dealing with tombs of a type hitherto unrecorded. The excavation of these remains was conducted with the fullest care to secure adequate records and ensure the recovery of all the archaeological material.

The tombs were five in number, irregularly disposed within an area of 50 sq. yards (Fig. 1). They all exhibited more or less the same features. Each tomb consisted of a saucer-shaped cavity roughly cut in the rock-bed, which at this point is the soft, white, Middle Globigerina Limestone and which underlies about 2 ft. of field soil. The diameter of the tombs averaged about 6 ft., and they attained a depth of about 2 ft. at the centre. A single layer of flat, roughly chipped slabs, derived from the local rock-bed, was used to pave the tomb floor. These slabs, about 4 in. in thickness and 10 in. in width, varied in length from 6 in. to 2 ft. 3 in. Overlying the slabs there was a layer of marl, about 6 in. in thickness, in which were embedded human skeletal remains, mostly in a fragmentary state and in utter disorder.

The Cemetery at Sphagion was excavated during the last week of August 1951, under the auspices of the Ashmolean Museum—Sydney University expedition to Cyprus with the aid of funds contributed jointly by the University of Sydney and the Australian Institute of Archaeology, Melbourne, which now possesses the finds allotted to the expedition by the Department of Antiquities. Tomb 17 is in the Cyprus Museum. The University of Sydney has made a generous grant towards the cost of illustration.

The site at Sphagion (so called because of a nearby slaughter-house) lies on the outskirts of the village of Myrtou (Kyrenia District) in North-west Cyprus (fig. 1). It is situated immediately to the North of the main Myrtou-Nicosia road, within a few metres of the 20-mile post from Nicosia. The narrow, flat tract of land, now used for the cultivation of vines and olives, has a shallow surface soil overlying the soft white limestone into which the graves were cut. The land belongs to Eleni Hadji Sophokli of the village of Myrtou.

The epigraphy of Roman North Africa is well known to be richer than that of any other Western province. Not only is the total number of inscriptions large, but their wide distribution makes possible a detailed study of some municipal institutions withoutmuch danger of arguing from evidence from one city to a generalisation covering many. The structure of municipal life in the more populous parts of North Africa was characterised by the large number of communities which, though small, developed genuine civic institutions, as compared with the semi-tribal agglomerations with a few very large cities that we find elsewhere in the Western Empire. Further, many of the sites have been desolate since the end of Roman rule, and no events during that rule brought destruction widespread enough to cause significant gaps in the epigraphy taken as a whole.

The two well-preserved Roman fortresses to be described in this paper have been known for many years. They were first brought to European notice by the British-sponsored geographical expeditions of the nineteenth century, when Tripoli was the spring-board for repeated attempts to find a route into Central Africa. Although important discoveries have been made in one of these forts (Bu Ngem) in more recent years, no detailed ground-plans have previously been published.

The following notes and illustrations are primarily intended to fill this lacuna in the documentation of the African limes; but it is hoped that they may also serve to increase our knowledge of early third-century trends in Roman military architecture. The European frontiers of the Roman Empire have yielded, and are still yielding, numerous examples of first- and second-century forts, and equally numerous examples of the forts erected during the later-third and early-fourth centuries, when barbarian invasion threatened the whole Roman world.

Few archaeological problems have been longer and less conclusively debated than that of the origins of the Christian basilica. Ever since the great Renaissance architect, Alberti, noted the similarity of name and of architectural form between the Early Christian basilicas and the forum basilicas of Imperial Roman practice, students of classical architecture have been trying to establish the derivation of the Christian basilica from this, that, or the other type of pagan monument. The measure of their failure to secure general agreement is the very large and still increasing literature that has grown up around the problem. There have, of course, been certain clear advances. Some of the older theories, such as that which derived the basilica from the Pompeian type of Roman house, need no longer be taken seriously; and there is a vastly greater body of reliable evidence available now than there was even a quarter of a century ago. Despite these advances, however, we are still very far from being able to give an agreed answer to the fundamental question, why it was that the fourth-century Church adopted the apsed basilical hall as the standard form of building for the celebration of the Eucharist, and so established an architectural and liturgical pattern that is still effective to the present day. The primary purpose of this article is to review the problem in the light of recent discoveries and of recent research, and to try to define, rather more precisely than is often done, the terms within which it can usefully be discussed.

This paper provides a sequel to the account previously given of Roman settlement in the eastern Gebel of Tripolitania, and is designed to complete the summary report of exploration carried out in this area during the years 1949–51. I must again express my gratitude for the assistance of friends and colleagues, in Tripolitania and elsewhere, to whom detailed acknowledgement was made in my earlier paper.

In dealing with the decline of the prosperous agricultural society of the first three centuries A.D., two sites were described on which the open olive farms were replaced, in the one case by a ditched blockhouse, and in the other by a herdsman's hut. These examples were chosen to illustrate the nature of the transition rather than to give a complete picture of the later economy. We there confined ourselves to the suggestion that the destruction of the olive farms was the work of marauding pre-desert tribes such as the Austuriani, whose incursion into the territory of Lepcis is recorded by Ammianus Marcellinus; and that the invaders' subsequent settlement left its material traces in the huts and fortified farms, and its historical mark in Procopius' description of the Libyan tribes and Ibn Abd al-Hakam's account of their traditional origin. It is our purpose here to give a fuller account of the later society, as it can be pieced together from the evidence of the sites, and to elucidate it where possible by a more detailed citation of the literary references.

A recently published history of naval power and trade in the Mediterranean in the early medieval period provides a reminder of that vital but little-explored change in the balance of maritime power which preceded the Crusades and of the movement which won back for the Christians of the West, Sardinia, Sicily, and Malta, threatened the shores of Africa and Albania, and was both a rehearsal and an essential preliminary for the great Drang nach Osten which was to follow. Little is known of the shipping used in these operations or of the tactics employed, but an attempt has been made to investigate the scanty sources for the Norman naval operations in the early stages of their conquest of southern Italy and Sicily, not merely because some of these operations were important in their immediate results, but also as an example of the naval technique of the Latin Christian military commanders in southern Europe. The subject derives additional interest from the spectacle it affords of Vikings finding their sea-legs again after many decades as farmers and cavalrymen in Normandy. It also offers one explanation for the conquest of the considerable area of southern Italy and Sicily by a comparatively small force, and once again demonstrates the eclecticism and adaptability of the Normans.

It is impossible for us, even in these days of land-rovers and jeeps, to imagine a North African landscape without a camel in it, yet the animal is nowhere to be found on the numerous Roman mosaics with scenes of country life in the museums of Algeria, Tunisia, and Tripolitania. Only two ancient writers before Ammianus Marcellinus refer to the camel in North Africa, one specifically, the other by implication. The fact that the camel spread over this region—Africa Proconsular is, Numidia and Mauretania—some time during the Roman period is, however, generally accepted. What is not clear is when and how this happened. The purpose of this note is to publish the photograph of a relief from a Tripolitanian tomb (Pl. XVIII), in the hope that it may be a small contribution to the discussion of the problem, at least as far as Tripolitania is concerned.

The evidence has often been studied. The earliest mention of the camel in the Maghreb is the note in the Bellum Africanum to the effect that twenty-two camels belonging to King Juba were captured during the skirmishing in the campaign which led up to the Battle of Thapsus (45 B.C.). It is not known how Juba obtained his twenty two camels, or whether other rich Numidians owned any, and there is no hint of the presence of the beast in Punic times. The second notice comes in the work of the late third-century African Christian, Arnobius of Sicca, who writes of the camel kneeling down when it is being loaded or unloaded.

Dr. Bernabò-brea has kindly given me a new photograph of this monument. It shows the red paint at the outer corner of the gorgon's eye, which makes her eye look bigger and more sinister. It is not, however, a leer such as many early gorgons wear: the pupil is still in the middle of the eye. A lesion in this region of the eye may be due to strain. Our artist is depicting a gorgon under pressure. We must think away two round, dark shadows behind the top of the plaque: the real top is level with the upper end of the gorgon's ears. Dr. Bernabò-Brea was good enough to discuss the plaque with me, and he allowed me to study it outside its glass case. Here are a few observations:

1. What Should Not Be There—the bud-like mark above the nose.

The new photographs show that there is no evidence for it in the original. Payne omits it from his drawing and thereby makes the picture stronger and more archaic. No other full-length gorgon seems to have a ‘bud’. Many Corinthian gorgoneia have ‘buds’, the earliest is on the Timonidas vase in Middle Corinthian times. This difference in treatment may be due to the feeling that a gorgon was a living creature, while a gorgoneion was more like a space to be filled. The ‘bud’ has come to our gorgon from a gorgoneion found in Gela, probably of a later date and from a non-Corinthian colony. The same fate of unsponsored restoration overtook a gorgon on a gravestone in Athens, but, as far as my memory serves, the restoration has since been removed.