This study is concerned with the exploitation of resources by human groups in the Carmel area over a period of about 50,000 years. To this end an attempt is made to evaluate the changing economic potential of the ‘catchments’ of individual sites, for this enables us to make a comparative analysis of hunting-gathering and agricultural economies.

Although the study of prehistoric sites is now usually complemented by some treatment of their setting, the situation in their immediate vicinity—the principal concern of the inhabitants—tends to be neglected or at any rate overshadowed by generalized statements regarding the physiographic, vegetational, climatic or kindred zones of which they form a (not necessarily typical) part. The difficulty that is encountered by attempts to harmonize the findings of the various specialists who nowadays contribute to archaeological studies may, in fact, be due to the limitations of a zonal approach, for the populations or phenomena with which each of them is concerned do not necessarily refer to a single ‘catchment’ area and may have little connection with the picture of the exploited territory that is imprinted on the site record.

Lunulae are commonly considered to be Early Bronze Age gold sheet collars produced solely in Ireland in imitation of jet and amber spacer-plate necklaces found elsewhere in the British Isles. Decorated with a distinctive style of incised geometric designs, they are more carefully and intricately ornamented than Beaker gold sheet-work, but less sophisticated than the superior Wessex gold objects. The term lunula has come to distinguish the Early Bronze Age gold crescent collar from the term Halskragen, a probably distantly-related type of copper crescent collar found in north-west Germany.

It is now nearly 20 years since the late W. F. Rankine first drew attention to the use of Portland Chert as prehistoric material (Rankine, 1951). He mapped the distribution of such artifacts in southern England, as known to him. No attempt was made to differentiate between artifacts of different cultural affinities or periods. Also, when he wrote, hardly anything was known of the prehistory of the Isle of Portland, Dorset, and of the immediate neighbourhood where the chert occurs in natural form.

During the past few years rich stone age habitation sites and workfloors have been found on the Island and along the landward shore of the Fleet nearby (Palmer, 1963–1968). Research is continuing in the area. This lends far greater importance to the question of the distribution of Portland Chert artifacts.

The present paper will assess the significance and scope of the distribution of Portland Chert artifacts in relationship to the industries in the Portland area. Investigations in the field and in numerous museums and private collections have greatly lengthened Rankine's original list of Portland Chert artifacts. It is likely that even now numerous specimens in other collections have not been traced, and that the distribution of these artifacts is even wider than suggested by the present list (Appendix A).

In 1934 Professor V. Gordon Childe wrote an important paper entitled ‘Neolithic Settlement in the West of Scotland’ (Childe, 1934), in which he sought to correlate the distribution of what are now known as Clyde chambered cairns (Scott, 1969a) with the incidence of the comparatively scarce light and well drained soils of the predominantly mountainous region in which the cairns are found. He visualized such soils as restricted to raised beach platforms, apparently having the so-called 25-foot beach in mind, and to alluvial gravels (1934, 22). He considered that such soils would have been more easily cultivated by primitive methods than heavier soils, and would indeed for that reason have been deliberately sought by incoming Neolithic settlers (1934, 23). He assumed that tombs would have been erected on slopes immediately adjacent to settlements (1934, 23), and consequently that such tombs could today be taken as indicators of settlement in the absence of other evidence.

The Somerset Levels are the largest area of low-lying ground in south-west England, covering an extensive region between the highlands of Exmoor, the Brendon Hills and the Quantock Hills to the west, and the Cotswold and Mendip Hills to the east (Pl. XXIII, inset). The Quantock Hills and the Mendip Hills directly border the Levels themselves, and reach heights of over 250 metres above sea level. The valley between extends to 27 metres below sea level, but is filled to approximately the height of the present sea by a blue-grey clay. The Levels are bisected by the limestone hills of the Poldens, and both parts have other smaller areas of limestone and sand projecting above the peat deposits that cap the blue-grey clay filling. In this paper we are concerned with the northern part of the Levels, an area at present drained by the River Brue.

The flat, peat-covered floor of the Brue Valley is some six kilometres wide and is flanked on the north by the Wedmore Ridge, and on the south by the Polden Hills (Pl. XXIII). In the centre of the valley, surrounded by the peat, is a group of islands of higher ground, Meare, Westhay, and Burtle. These islands, which would always have provided relatively dry ground in the Levels, are linked together by Neolithic trackways of the third millennium B.C. Several of these trackways formed the basis of a paper in these Proceedings in 1968 (Coles and Hibbert, 1968), which continued the work of Godwin and others (Godwin, 1960; Dewar and Godwin, 1963).

The presence of an archaeological site on Kosipe Sacre Coeur Mission was first noted in 1960, when axes and waisted blades were found by Father L. Willem during excavations for church foundations. Word of the site was sent to Mr W. Tomasetti, then Assistant District Officer, Department of Native Affairs, Tapini, and he informed White of it. Excavations were made there in June 1964 (White, 1965, 41–3; 1967). In 1966 the site was visited by Crook who collected further carbon and soil samples and in August 1967 White and Ruxton carried out further archaeological and geomorphological investigations. This report covers the entire history of excavations at the site.

Kosipe lies about 135 kilometres north of Port Moresby and 20 kilometres north northwest of the Woitape Sub-District office, Central District, Papua, at 147° 16′ E, 8° 21′ S (fig. 1). The area around Kosipe has not yet been the object of a reconnaissance geological survey. Observations by Crook along the Woitape-Kosipe road, south of Kosipe and on the Kosipe-Tanipai track northwest of Kosipe along the Ivane River indicate that the oldest rocks are slates and other low grade metamorphics of unknown age with steeply dipping foliation. These are overlain, probably sub-horizontally, by several hundred feet of basic volcanics, probably largely basalt. The volcanics are probably of late Tertiary or early Quaternary age; they are strongly dissected and are not obviously related to any centres of extrusion.

The practice of Archery in the Bronze Age and early Iron Age as it survives to us in the archaeological record is difficult of interpretation owing to the low survival value of most archery equipment—the one relic of fairly high survival value being the arrow-head, of flint, bronze or iron. This paper deals with a topic little touched upon previously, and, setting aside the tradition of stone arrow-head manufacture (which persisted well into the Bronze Age in various parts of Europe) essays a treatment of the evidence for the manufacture and use of metal arrow-heads. These will be divided into three types, tanged, socketed and spurred—a threefold division which will appear to be significant not only from the purely typological point of view, but also in the more general context of prehistory.

A catalogue of all finds known to the writer in 1967 will be found at the end of the paper. While this catalogue makes no claim to be complete, it is representative enough to serve as a firm basis, and although further research may bring more examples to light, it will probably not reveal any major new concentrations of metal arrow-heads within Europe (at any rate within the framework of existing knowledge).

Parallelopiped bone dice form a small but important group in the bonework of the Scottish Iron Age. The overwhelming mass of this material has been produced by excavations in which the stratigraphy has not, save in exceptional circumstances, been recorded. Moreover, the large assemblages are from sites which give every indication of being multi-period, notably the brochs and wheelhouses of the Atlantic Province (the Province is defined in Piggott, 1966, 5, 4, fig. 1). This lack of meaningful cultural or chronological groups has been an open invitation to select those objects, with parallels in the supposedly better defined and dated groups to the south, in order to draw wide-ranging inferences concerning the pre-history of Scotland. Such a procedure enables the bulk of the material to be ignored as sui generis. It is not the writer's intention to produce another study of the might-have-beens of the Scottish Iron Age, but rather to present a corpus of the dice and to consider the previous interpretations in which these dice have formed part of the evidence.

There is no need to stress that at the present day it is becoming increasingly accepted that the further expansion of prehistory depends not only on the development and ever greater precision of the actual methods of archaeology, but also to an equal degree on the application of natural scientific research methods. This has been well exemplified in the series of systematic studies on petrological research on stone artifacts published in these Proceedings.

A similar tendency can be noted of late in many other European countries, France (Cogné and Giot, 1957; Giot, 1964), Austria (Zirkl, 1956–9), Germany (Frechen, 1965; Wiegers, 1950; Scholz, 1968), Hungary (Végh and Viczian, 1964), Poland (Kowalski and Kozlowski, 1959), the Soviet Union (Gajduk, 1963; Kovnurko, 1963; Petruň, 1967) and Switzerland (Hugi, 1948).

In Czechoslovakia, too, from 1960 onwards, regular petrological-archaeological co-operation has been carried on between the Department of Prehistory at the Philosophical Faculty of the Purkyně University of Brno and the Department of Mineralogy and Petrology of the same University. To begin with, the stone building material of Great Moravian architecture in the region of the lower Dyje valley and the lower Morava valley (Štelcl and Tejkal, 1961–7) was petrologically examined and latterly the investigation has been extended to cover the chipping and polishing industry of the Stone Age (Štelcl, 1967; Malina, forthcoming; Štelcl and Malina, 1966, and forthcoming).

In Taranto Archaeological Museum is a group of bronze objects, which are believed to have been found at Surbo, a village some 5 kilometres north of Lecce in Apulia. When Professor Drago was Director of the Museum, he heard of the existence of a group of bronzes, already partly dispersed: he bought the remaining bronzes but could gain no further information concerning the provenance or the dispersed objects. The Atto d'Immissione of Taranto Museum, No. 679, dated 20 March 1952, mentions seven objects under the Inventory Nos. 17140–17144 and lists the sword hilt, shaft-hole axes, hammer-heads and chisel. Under No. 17142, it describes ‘Una accetta a margini rialzati o coltello ascia. Cms. 2,3 × 0,7’. The measurement must be an error and the description might apply to the medial winged axe or to the spatulate axe. In view of this ambiguity, the eight bronze objects from Surbo, now grouped together in the Museum, are described here, noting that, if the spatulate axe could with confidence be omitted from the group, the remaining bronzes cohere well and an Apulian source is credible. It will be wise to keep these points in mind throughout the following discussion, where the bronzes will first be considered individually and then as a group.

Nearly seventy years ago Wilhelm Worringer first wrote that ‘ultimately all our definitions of art are definitions of classical art’ (Worringer, 1953, 132). Today, the study of Western European art history, old or modern, the products of peasant craft-centres or urban ‘schools’, has in the course of time developed its own methodology and, almost, mystique. In contrast, the study of many branches of prehistoric art in Europe and elsewhere is all too often seen as a mere extension of the skilled but subjective approaches of classical archaeology without considering the suitability of the latter's application. The use of the classical art-historian's intuitive methods built up not just from visual exprience but a detailed background of literary, historical and philosophical studies must in fact be almost entirely denied the student of prehistoric or primitive art. It is perhaps only natural that principles of classical art history should be applied to later European prehistory, though it is often difficult to arrive at a precise definition of these principles. It was Johann Joachim Winckelmann who made the first systematic application of categories of style to the history of art (Gombrich, 1968, 319). Sir John Beazley, the greatest of all modern classical art historians followed in this tradition basing attributions ‘on the grounds of tell-tale traits of individual mannerisms’ (Carpenter, 1963, 115 ff.) a scheme first applied to painting less than a century ago by the Italian physician Giovanni Morelli (Gombrich, 1968, 309 ff.) and followed at the turn of the nineteenth century in the study of Italian painting (Lermolieff, 1892–3). With Beazley it is, however, difficult to follow step by step his methods of work.

It is now nearly twenty years since the first radiocarbon dates for the neolithic cultures of Europe astonished and perplexed prehistoric archaeologists by falling a millennium earlier than the existing chronologies had suggested. Since that time most (but not all) archaeologists have become reconciled to radiocarbon dating, and to a chronology for Europe based, at least for the neolithic period, on the radiocarbon dates which have become so widely accepted as now to be conventional.

But a second radiocarbon revolution is now taking place. The calibration of radiocarbon by tree-ring dating indicates that the now conventional radiocarbon dates are not early enough. Dates for the neolithic, and this time for the early bronze age also, will have to be set earlier by several centuries. The consequences of this second revolution for the picture we have of prehistoric Europe are in many ways more radical than those of the first. For this reason, perhaps, there has been a marked reluctance among some archaeologists to face the consequences of the new calibration, although an emerging concensus of opinion among botanists and physicists seems to favour it. The present article sets out to identify the three principal points of difference between the conventional chronology and that suggested by the tree-ring calibration of radiocarbon. It is argued that on independent archaeological evidence the new, calibrated, chronology can be seen to yield a plausible and coherent general picture. While many problems have yet to be clarified, there seems no reason why the new dating should not be welcomed as offering a stimulatingly fresh view of the prehistory of Europe.

Belle Tout is the second of a series of headlands where the South Downs meet the English Channel between Eastbourne and Brighton. Three kilometres to its east, the chalk downland comes to an end at Beachy Head, while Birling Gap immediately to the west provides direct access to the shoreline. Belle Tout, in between, is a gorse-capped plateau of virgin downland, 40 hectares in area with chalk cliffs rising to a maximum height of 84 metres (fig. 1).

The even lines of the headland are only broken against the cliffs where a shallow valley runs down towards Birling Gap from the highest point at Belle Tout lighthouse. At a height of 40 metres this valley veers to the south and is cut by the present cliff edge. It was within the valley at this point that the Beaker site was discovered and excavated (NGR TV557956). The natural subsoil throughout is Upper Chalk save within this valley where the chalk is covered by a drift deposit of loam with patches of flint debris laid down under periglacial conditions. Under normal vegetation this surface loam would be acid, but a mildly alkaline reaction in the soil has been brought about by gorse roots penetrating to the underlying chalk.