The excavation produced a total of 23 metal artefacts; five of copper alloy and eighteen of iron. This material was recovered from two trenches only (B and E). 7 ne relationship of types to trenches is given in table 2.

Copper alloy (fig, 9).

1. Terret ring. Oval in form and of circular cross-section, with a low ridge-moulding around the outer edge of the loop. The loop tapers gradually from 7.2 mm diameter at the terminals to 6.2 mm opposite them. The terminals are circular (12.8 mm diameter) and have a shallow groove incised around them close to their inner faces. The attachment bar is parallel-sided and of rectangular section (6.6 mm x 4.5 mm). External dimensions: 54 mm x 47.2 mm; weight 41.6 gm.

Excavation of trenches B and E produced a small quantity of metalworking residues which was subjected to a simple visual examination in order to determine the nature of the metalworking activity. Four basic types of ironworking slag were identified in addition to a small quantity of fired clay, ammounting to 1,44 kg of debris. Further metallographic analysis would be required to continue the following divisions.

To the north of Milfield, Northumberland the A697 cuts through a cluster of ring-ditches centred at NT933347. Discovered during aerial survey by Professor N. McCord, these were included in Dr A. Harding's survey of the ritual henges and related monuments of the Milfield Basin (Harding 1981, 101). Subsequent aerial reconnaisance by Mr T. Gates has increased the number of visible ring-ditches to at least eleven and, of these, three types may be tentatively isolated (fig. 1). The smallest appears to have a diameter no greater than 3.5 m, but with a ditch some 2 m in width (3 examples); in the second type the ditch appears less than a metre in width, but enclosing a larger area estimated at between 9 and 20 m (3 examples). A single example (Site 1 below) lies outside these dimensions and it was this, together with an example from the second group, that was chosen for excavation in 1982 and 1983.

Eskbank Nurseries, a market garden allotment some 200 m long and 30 to 40 m wide, lies at approximately 70 m OD on a relatively level terrace overlooking the south bank of the river North Esk (fig. 1). The excavated area is centred on NT320660 and lies within the two superimposed Roman camps first noted as crop marks by St Joseph in 1965 and subsequently excavated by Maxfield in 1972. Her excavation areas I and II were sited to investigate a pit alignment, visible on the aerial photographs and plotted by her (Maxfield 1974, 142, fig. 1). In the event no pits were discovered by excavation, a fact attributable to an unfavourable combination of weather and soil conditions (Maxfield 1974, 150). The pit alignment did not, of course, appear as a crop mark in the deep market-garden soils of the Nurseries but, by extrapolation of Maxfield's plot, it was clear that it must cross the northern half of the allotment, possibly intersecting the north ditch of the earlier of the Roman camps (Maxfield's Camp A) on the west edge of the site.

This study attempts to define patterns in the artefact distributions from subsoil features at Winklebury Camp, an early Iron Age hillfort in lowland Britain. Two aims are identified: the first is to elucidate the nature and organization of past human activities within early Iron Age hillforts; the second is to determine whether meaningful patterns can be identified within the artefact distributions preserved in subsoil features. A variety of hypotheses drawn from the literature are presented and these are compared with the results of a number of statistical techniques. It is possible to distinguish patterns within the artefact distributions from Winklebury and some interesting results are produced. Specifically, it can be suggested that Winklebury was not participating in specialized craft production, that round-houses had a variety of functions, and that four-posters did not have a domestic function.

Round houses and an enclosure, surviving parts of a nucleated settlement on boulder clay terrain at Roxby (near Staithes, north-east Yorkshire), discovered from the air in 1972, were excavated 1973–81. Most were dated to the immediately pre-Roman Iron Age, but one round house, standing in an area of marks of former cross-ploughing, had native Romano-British pottery, and in the last phase of ditch silting, sherds of sixth century AD Anglo-Saxon stamped ware. The economy was based on mixed farming, but two of the Iron Age houses also contained iron working comprising both smelting and smithing. These houses also yielded fragments of jet and glass and were interpreted as a repair workshop, rather than a production unit. Great structural detail had been preserved and was recorded. The houses were architecturally different and represent a significant addition to the prehistoric round house data. They lie in that part of the township of Roxby which escaped medieval ploughing, and probably represent a fraction of the total original settlement. This and other data in north-east Yorkshire show that an economy based on settled mixed farming, not on semi-nomadic pastoralism, was widespread across the boulder clay encircling the North York Moors.

Two farmsteads, one of late Iron Age (second-first centuries BC) date and the other dating to the early Romano-British period (first-second centuries AD), were excavated at Copse Farm, Oving. The site is situated within the Chichester dykes on the Sussex/Hampshire Coastal Plain. The Iron Age farmstead produced pottery spanning ‘saucepan’ and ‘Aylesford-Swarling’ traditions, a transition in ceramic production which is poorly understood in Sussex. Information on the agricultural economy and small-scale industries (principally metalworking) practised at this site give an insight into the way the Coastal Plain was settled and exploited at the end of the first millennium BC.

Strap-unions of the types discussed in this study were made during the later part of the pre-Roman Iron Age and during the earliest part of the Roman occupation of Britain. Their distribution (fig. 1) and details of their chronology and cultural associations and also the evidence for their precise use are considered in the concluding sections of this paper.

Observations made soon after recent major earthquakes in Algeria, California and Japan have shown that repeated motions on buried fault planes in seismically active areas cause incremental folding of the overlying rocks and sediments. The deformation causes characteristic changes to river profiles and is therefore a factor which must be considered when applying techniques of site catchment analysis in areas of tectonic uplift. Here we examine the relationship between tectonic uplift and the palaeoenvironments of palaeolithic sites in North-west Greece, which is one of the most seismically active areas of Europe. Uplift can substantially alter local topography and sediment distributions and therefore undermine the economic viability of human settlements. Paradoxically uplift can also create stable conditions highly favourable to long-term habitation in two ways: by maintaining well-watered sediment traps which provide a climatically insensitive environment; by accentuating enclosed topography, which facilitates the control of mobile prey species.