A contour map of part of the basin of the glacial Lake Pickering was constructed from surface and sub-surface levels. This map (figs 2, 3 and 4) gives a pictorial representation of the microtopography of the area and is amplified by an additional map (fig. 13) in which the margin of the organic deposits is given for the early Mesolithic period, based on evidence outlined below. A stratigraphical survey of the basin deposits shows that the infilling of the lake in the Devensian Late-glacial began largely with the deposition of calcareous muds. In the early Flandrian there followed a natural hydroseral development from open water to reedswamp and fen carr. Towards the end of the Boreal period and during the early Atlantic, a rise in water level was responsible for the replacement of fen carr by reedswamp. The rate of hydroseral development was affected by a progressive rise of water level. This has been determined (estimated) by radio-carbon dating the basal organic muds on the 23, 24 and 25 m OD sub-surface contours at eight sites within the study area. During the early Mesolithic (Godwin's pollen Zone IV, c.9600 BP)the water level in the basin was likely to have been in the range of 23–24.5 m OD. By reference to the dating evidence, including published pollen diagrams (Walker and Godwin 1954) and new data presented in Parts 2 and 3, the stratigraphical evidence is interpreted to show that there was considerable variation in the vegetation of the lake margin during the early Mesolithic period (pollen Zone IV). At Seamer Carr there was a large area of fen carr woodland with reedswamp fringing the open water. Star Carr had a much larger area of reedswamp, with open water close to the dry land in places, but with a narrow fringe of carr. The bed of the ancient River Hertford, which drained the basin, probably in the late Mesolithic, has been identified in two of the sections which traverse the basin.

A three-dimensional study of the vegetational history of Seamer Carr from the Devensian Late-glacial to the Atlantic period has been made with the aid of six sites in the main basin. Five of these are close to the margins of the organic deposits and the sixth is from deep basin deposits. The Godwin zonation scheme has been used to allow a direct comparison with the survey of Walker and Godwin (1954). A composite section through the basin deposits, based on the detailed contour plan described in Part 1, shows that, pre-Zone IV, the open water margin abutted the dry land, and that during Zone IV reedswamp and carr gradually filled in the edges of the lake. Large expanses of fen carr occupied Seamer Carr during Zone V, but a rise in water level during Zone VI was responsible for the area reverting to reedswamp. At the Boreal/Atlantic transition, coincident with the rise in the alder pollen curve, there is a thick band of highly humified, acid peat, full of microscopic pieces of charcoal. Analysis through this band at one of the specially selected sites on East Island produced evidence of local disturbance, seemingly anthropogenic, at the wetland margin. There were two other special sites in West Embayment that were sampled for pollen analysis. The first was located over a radiocarbon dated Zone II organic deposit, sealed beneath a Zone III calcareous sand layer. The second was located in a late Mesolithic organic mud radiocarbon dated to c. 8000 BP and containing an assemblage of microliths of late Mesolithic type, identified tentatively as a composite tool (A. David pers. comm.).

Detailed boring records and surface levels, combined with knowledge of the rate of extension of organic deposits in the lake basin (Part 1), allow the production of a computer-generated contour map and isometric drawings of the microtopography of the area around the Star Carr site of Clark (1954) (figs 1 and 2). The site was at the mouth of a shallow gully on the southern face of a hillock of glacial till which could be reached from the higher ground to the N over dry ground. The hillock extended into the lake basin as a low peninsula ending in a spit, at the entrance to its relatively narrow outflow channel. The occupation area was at a point where the open water most closely approached the shore and extended from the marginal wetland on to the dry slopes of the hillock.

Pollen and lithological analysis, combined with extensive C14 dating, is described at four closely spaced points along a section through the basin deposits close to the Star Carr site. Dated sequences of local vegetation development are established. The results are used to reconstruct pictorially the environmental changes through a period of occupation contemporaneous with, and presumably an extension of, the occupation of the Star Carr site (fig. 9). The dry ground supported birch wood throughout, but natural successional changes in the wetland, on a nominal timescale, were as follows:

(a) At 9800 BP the open water of the lake lapped the shore save for a narrow zone of reeds, sedges and water plants.

(b) By 9650 BP immediately before the occupation, reedswamp dominated by the great reed (Phragmites) had extended well into the basin. There was a narrow drier zone at the landward side of the reedswamp with abundant ferns enclosing a damper patch with fen plants.

(c) At the time of the occupation, around 9600 BP, the inshore reedswamp had largely been replaced by fen dominated by the saw sedge (Cladium). The marginal environment, where artefacts were found, was still dominated by ferns but a wood layer occupied the former damp fen area, presumably having been deliberately emplaced to consolidate the ground. Cladium fen and fringing reedswamp formed an extensive tract to the E of the occupation area behind the spit mentioned above, and probably also to the W.

(d) By 9300 BP (on the nominal timescale used for the reconstruction) the occupation was over and the wood layer and an associated ‘brown layer’ were covered by Cladium deposits on which willows were now growing.

Attention is drawn to the need for detailed work on the taphonomy of artefacts as a component of any further excavation in the organic deposits of the Vale of Pickering.

The inception of the Neolithic has always been one of the more vexed questions of British prehistory. As an issue, it has been obscured by a number of conceptual difficulties. Not least amongst these is that the transition from the Mesolithic to the Neolithic coincides with the point at which two different and opposed approaches to prehistory and its teaching meet. The first of these, which dominates the study of the Palaeolithic and Mesolithic periods, is concerned with human behaviour in terms of adaptive responses to environmental pressures. The latter, more common in the Neolithic and later periods, is more likely to consider human beings as purposive subjects, acting in pursuit of socially-defined goals. As a result, the Mesolithic-Neolithic transition represents not merely a change of lifestyle, but also a point at which our perception of the past changes. It is the boundary between two models of man.

Material culture meanings are transferred between objects on the basis of similarities and differences. In order to fix the potential ambiguity of meanings, experience is bracketed within bounded contexts. New acts always refer back to existing organized experience or texts. It can be argued that text comes before action. This idea is used in a discussion of the appearance of agglomerated and bounded settlements in the European Neolithic. It is shown that occupied settlements are often preceded by non-domestic enclosures which sometimes have a ritual nature. Ritual text seems to prefigure practical event. A similar relationship between formalized ritual texts and social action is shown in a wide range of other examples. Finally some of the potential reasons for the observed relationship between text and action are discussed.

Aerial survey in Essex has revealed a number of elongated enclosures interpreted as either long barrows or mortuary enclosures of Neolithic date. Excavation of one of these sites at Rivenhall in 1986 produced finds of flintwork and pottery which help to substantiate this hypothesis. A surface collection survey of the field containing the enclosure produced Mesolithic and Neolithic flintwork. A short discussion considers the Essex sites in their wider context.

The environmental setting of two late-prehistoric enclosures in upland Ardudwy, Gwynedd was investigated by pollen analysis of nearby valley mire deposits and of old ground surfaces revealed during excavation of the sites. Radiocarbon-dated pollen data indicate human influence on the local environment since Mesolithic times, with the major period of woodland clearance in the early Bronze Age. The two enclosures appear to have been constructed in an environment already substantially deforested. Remanent upland timber resources appear to have been substantially depleted during occupation of the sites, whilst soil acidification and/or climatic change may also have contributed to their eventual abandonment.

Two late prehistoric circular enclosure settlements near Harlech, Gwynedd were excavated by the Gwynedd Archaeological Trust in 1980–81 in advance of upland pasture improvements. Both produced evidence for the change from timber to mainly stone as a building material, commonplace in upland Britain at this time. At Moel y Gerddi there was some evidence for earlier activity in the Neolithic period, but the principal occupation occurred in the mid-first millennium BC, and comprised a roughly circular timber palisade enclosing a fairly substantial timber roundhouse. As these structures deteriorated, they were consolidated and embanked with stone, but despite this the occupation did not last very long and the site was abandoned. Erw-wen a short distance away produced similar evidence but there was no Neolithic phase and the site was re-used in the Medieval period. The principal late prehistoric settlement there comprised a single large roundhouse standing within an imposing enclosure wall but whether or not this had been preceded by a timber palisade is not known. It was clear, however, that the stone roundhouse was a replacement for an earlier timber counterpart and, quite unlike Moel y Gerddi, the occupation lasted a considerable period of time, possibly two or even three centuries. Dating evidence from either site was poor, and acid soil conditions mitigated against the preservation of most finds. The sites were apparently aceramic and probably relied on pastoralism rather than arable farming. The environmental setting and background has been considered in some detail in the preceding paper with particular reference to soil pollen studies and a bog core dug near Moel y Gerddi. Typologically the sites belong to the Gwynedd stone-built hut-circle settlements and it is argued that they could mark the first appearance of these monuments in the landscape. But in any case, the use of timber as witnessed here, was far more commonplace than has previously been supposed on these sites.

Ring cairns in Wales have been viewed as a discrete form of structured cairn, a stony ring enclosing an open central area, dating to the second millennium be, whose distinctive character may reflect a particular role not primarily connected with the interment of the dead. Fifteen likely ring cairns were noted in cairn groups on five ridges in SE Dyfed and on the Gower Peninsula. Although the position of ring cairns on these ridges to some extent overlaps that of other cairns, ring cairns tend towards locations particularly dominated by higher ground which are not so favoured by the other cairns. A range of structural elaborations was present amongst the ring cairns including kerbing, in-filled interiors, orthostats in the bank, and breaks in the ring possibly representing entrances.

Two ring cairns adjacent to each other, GCRC I and II, on Cefn Bryn, Gower were excavated. GCRC I was a stony ring bank, originally with two entrances, one of which had been blocked in antiquity. It enclosed an open area at the centre of which was a pit containing a very little burnt bone. GCRC II comprised a ring bank of stone with inner kerb, against which was a pit containing earth and charcoal flecks. It certainly had one entrance, which was deliberately blocked perhaps at the same time as the interior was filled with stone to the level of the top of the kerb. The pit contents at the rings each gave mid-ind millennium be radiocarbon dates. It is possible that the pit fill at GCRC I came from an earlier burial elsewhere so the determination may not reflect the time of deposition at GCRC I.

A case can be argued that ring cairns in the area were built for the public performance of ceremonial on the basis of the apparent absence of formal burial at GCRC I and II, and the structure and siting of the monuments suggesting their suitability as arenas. Ceremonial may have been connected with mortuary ritual, although undated evidence for early agricultural activity in the vicinity of the rings indicates a potentially wider social and economic context for their use. While aspects of the evidence for activity at the two excavated ring cairns may be paralleled at structurally similar monuments in N Wales and S W England, some variations in form and activity are apparent which could indicate differences of emphasis during the time they were in use across the region. Finally, ring cairns are not the only form of broadly contemporary monument to which a ceremonial role can be assigned, and the differences in monumental form could well reflect different ritual requirements, or even social distinctions though too little work has been undertaken as yet on possible social models for the region for this idea to be developed.

As part of a continuing programme of assessment and investigation of prehistoric activity in the coastal area of E Cleveland, Cleveland County Archaeology Section excavated an insubstantial and enigmatic monument which had a ritual use continuing from the late Neolithic into the early Bronze Age.

Ever since Martin's definition of Breton tumulus burials in 1900 the discrete distribution of biconical handled vases or vases à anses has been well known (Martin 1900). In 1928 details of similar vessels in the Channel Isles became readily available through Kendrick's survey of the Bailiwick of Guernsey (1928). In Southern Britain, however, the nineteenth century finds at Portland G2 and Winterbourne Stoke G5 have reained unconfirmed and consequently discussions on Armorico-Wessex connections in the Early Bronze Age have treated the question of cross-Channel ceramic exchange with notable circumspection.

This paper discusses the complex societies which flourished on the central plateau of southern Africa between the Zambezi and Limpopo rivers between c. AD 800 and 1500, and the models which can be proposed for how they functioned and why they developed. The principal archaeological monuments left by these societies are their regional political centres, the stone enclosures or zimbabwes (fig. 1), of which Great Zimbabwe is the best known and most elaborate (fig. 2). (The traditional spelling zimbabwe(s) is used in this paper rather than the correct but lesser known spelling dzimbahwe singular and madzimbahwe plural.) They varied considerably in size, but the largest probably housed populations numbering several thousands — Great Zimbabwe itself has been estimated to have had a population of some 30,000 people (Huffman 1984) — and their construction implies organized labour on a substantial scale. The main population lived in densely clustered huts outside the stone enclosure. Artefacts suggest that they were commoners, with the servants of the king and minor officials living close to the central hill. Beyond was an outer ring of prestige residences with their own housing units (fig. 2). ‘Great Zimbabwe was the product of a highly stratified society: the stone walls are essentially demonstrations of the prestige of a ruling class, a symbol of political authority that spread over the whole plateau’ (Garlake 1973, 14).

The cultural period immediately prior to the Talayotic phase in Minorca, the Pretalayotic, belongs to the Chalcolithic—Early Bronze Age and is represented by a series of megalithic monuments. When of a funerary nature, these consist of megalithic chamber-tombs in the eastern part of the island and of rock-cut graves of elongated plan in the western. So far, apart from the newly-discovered site of Morellet, no habitations are known in eastern Minorca and the new site appears to indicate that they were either of a seasonal nature or merely rock-shelters. In the western half of the island dwelling-sites are clearly evidenced by villages of habitation — navetas (navetiforms) such as Son Mercer de Baix which continued in use during the Early Talayotic period. It should be noted that although their method of construction would incline one to class them all as belonging to an identical Mediterranean cultural group, the finds indicate an autochthonous evolution which in time gave rise to the insular peculiarities, including some of a constructional nature, typical of Minorca.

A double-ditched sub-rectangular enclosure measuring c.42 by 48 m was excavated. It was shown to have been a Late Bronze Age settlement, with a single central roundhouse opposite the only entrance, and with a rectangular structure in one corner. The site produced a typical range of Late Bronze Age artefacts. The site is low-lying, and environmental data from the waterlogged fills of a well indicate an open landscape of damp grassland. These factors together with the absence of waste from the earliest stages of crop-cleaning among carbonized plant remains from the enclosure, point to a primarily pastoral economy.

A few Neolithic features and artefacts were also found but these are not considered to represent permanent settlement.

The standard recent authorities on the history of archaeology date the invention of a specific word for prehistory to 1833, saying that Paul Tournal of Narbonne used the adjective préhistorique (‘prehistoric’ in the English translation in Heizer 1969, 91; and in Daniel 1967, 25, following Heizer 1962) or the noun préhistoire (Daniel 1981,48) in an article about French bone-caves.

This is not true. The word Tournal used was antéhistorique (Tournal 1833, 175), and the mistake has arisen from working with an idiomatic translation into English, which rendered ‘anté-historique’ as ‘prehistoric’ (Tournal [1959]) instead of the original French. (Grayson 1983, 102., however, quotes Tournal's original French correctly.) The earliest use of ‘prehistoric’ seems to be Daniel Wilson's of 1851 in The Archaeology and Prehistoric Annals of Scotland (1851), as the older histories of archaeology say (eg Daniel 1950, 86 (reprinted in Daniel 1975, 86); Daniel 1962, 9), before the error about Tournal began to circulate.

In the summer of 1978 pottery and flintwork were noticed in the sections to the south of Cliffe Village during the laying of a pipeline by British Gas (TQ 734744) (fig. 1). This led to the excavation of a series of small trial trenches by Mr David Thomson with the help of local volunteers in the same year. The retrieval of a Beaker and Collared Urn suggested an early Bronze Age site, and excavations by Dr Ian Kinnes for the British Museum were done in September 1979. Although the excavated features contained mainly Iron Age pottery and metalwork, both seasons' work also produced a large quantity of flint artefacts ranging from Mesolithic to Bronze Age in date. The following report is an analysis of the Mesolithic tranchet axe manufacturing debitage which could be distinguished as a discrete group from the other flintwork. It is not intended to present a comprehensive flint report for Cliffe, but to provide a framework for analysis at other sites where tranchet axe production has been shown to take place (Wymer 1962; Parfitt and Halliwell 1982).

Well-defined late Neolithic structures in lowland Britain are rare. The contents of pits with Grooved Ware have, therefore, been used to interpret both settlement patterns and economic activity in this period. Pits occur both individually and in groups, are normally circular or slightly oval in plan and rarely exceed 1 m in diam or 2 m in depth (Wainwright and Longworth 1971, 250). Most were probably for grain storage (Field et al. 1964, 367) which were later filled with refuse.

The Amesbury area, 3 km E of Stonehenge, has produced several groups of pits of this type. They include five at Ratfyn (SU 15954205) N of Amesbury (Stone 1935; Warren et al. 1936, 196), four at Woodlands (SU 151431) 250 m S of Woodhenge (Stone and Young 1948; Stone 1949), a group within a henge on Coneybury Hill (SU 13434161) (Richards 1984, 183) and others on King Barrow Ridge (SU 135425) (Richards 1984).

The study of grave goods plays a central role in two very different kinds of archaeology. As groups of artefacts that were deposited together, grave finds occupy a key position in chronological studies. At the same time, it is commonly supposed that the selection of artefacts for deposition with the dead may be some reflection of the social position that they had enjoyed in life; the contents of different graves may be studied for evidence of wealth and status. Although chronological studies have the longer history, these two types of analysis ought to be most informative where social variation can be traced over a lengthy sequence. As we shall see, this raises special problems.

In September 1986 a small excavation was done by S. T. to investigate an area of primitive-style mine-workings on Copa Hill, Cwmystwyth, Dyfed (SN816756). A small opencast and several overgrown tips associated with pebble hammers occur where the copper-rich Comet Lode outcrops on the brow of the hill. Copa Hill is within an area of seventeenth–twentieth-century lead-mine workings which extend for 1.2 km along the N side of the Ystwyth Valley (fig. 1). Radiocarbon dating of charcoal from within one of the tips suggests that mining commenced in the middle Bronze Age.