Thom has proposed that megalithic monuments were planned using geometry and a unit of length — the megalithic yard. All data in Thom's data-sets not deriving from well-preserved stone circles are removed, so that these proposals can be tested on this one category of monument. The statistical evidence for the existence of the megalithic yard is found to be equivocal. The shapes of stone rings, a part from the true circle, can be most easily explained as arising from layout by eye, without geometric planning. Exceptionally, two groups of stone circles which have distinctive architectural traits appear to have been planned to be truly circular, presumably using peg and rope construction. These groups are in southern England and north-east Scotland, although the latter group is not identified with certainty.

A total of 190 axe-hammers and 82 battle-axes from all over Scotland were studied. Attention was focused on the key question of whether the implements were made from cobbles or from scree-outcrop blocks. In glaciated regions such as Scotland, the nature of the source exploited is critical, since the area from which a rock-type may be obtained is greatly enlarged if cobble-blanks were used. Exhaustive petrological and typological studies suggest that a rather haphazard exploitation of suitable-looking locally available rock-types took place. The manufacturing experiments described here were designed to discover how the implements were made, and the essential properties required of the blanks. They showed that a blank should be (a) free of flaws, (b) of a suitable rock-type, and (c) close in form to the implement required. Thirty-one timed searches were made of cobble-deposits and scree/outcrop sites within the areas of distribution of the artefacts. Cobbles, because they are more often free of flaws and better-shaped, proved much superior to scree/outcrop blocks as blanks. The main conclusion of this paper, therefore, is that the great majority of Scottish battle-axes and axe-hammers were made from cobbles. There is evidence to suggest that this conclusion may be extended to the rest of Britain, and to other types of implement made by pecking and grinding, including some unperforated stone axes.

This paper deals with a unique glass bead from the second millennium BC in Wessex. Overlooked for more than 150 years, it has now been recognized for its intrinsic interest and general importance and is here presented for its wide significance in ancient Europe and beyond (pls 8 and 9).

In Ancient Wiltshire (1812, 210) Richard Colt Hoare recorded the excavation of a barrow in a group of Bronze Age date at Wilsford: ‘No. 7 is a large bell-shaped barrow’ (now regarded as a bowl barrow) ‘composed entirely of vegetable earth. It contained within a cist a little pile of burned bones with which had been deposited a very fine brass pin, a large stone bead which had been stained red, a bead of ivory and a lance head of brass’. This account is based on the records of William Cunnington (1807, 5–6), which include a transcription of a letter from the original excavator, a Mr Owen. The dimensions of the barrow are there given as 80 ft in diameter, 9 ft high, with a circular cist 18 in deep. The barrow is described as ‘No. 6 of Mr Duke's barrows’; there is thus a discrepancy in the numbering of the barrow, since Colt Hoare referred to it as Lake No. 7, while Cunnington kept to No. 6. The barrow, though recently ploughed, still stands to a height of over 2 m, and is today known as G.42 (Grinsell 1957, 211).

The excavation of eighteen round barrows was undertaken by the late Charles Green during summer seasons from 1958 to 1960, in advance of their destruction by ploughing. The excavated barrows are members of two linear groups which occupy adjacent spurs to the east of the village of Shrewton in the modern parish of that name (fig. 1). One of the barrows investigated lies in Winterbourne Stoke parish. This western part of Salisbury Plain is termed the ‘Lower Plain’ comprising those areas having their ‘upper limit… between 400 and 450 feet OD and their lower limit where they overlook the valley trenches … between 250 and 300 feet O.D.’ (Gifford 1957, 6). In such a lower valley lie the modern villages of Shrewton and Rollestone, and through it flows the river Till, the nearest modern open water supply to the barrows. This connects the area to the River Avon via the Wylye and Nadder.

An assemblage of cairns, distinguished by the incorporation of orthostatic elements suggestive of ring structure, is described from an area of central Wales. The morphology and distribution of the sites are discussed in terms of geology, erosional processes and land-use history. It is concluded that the form taken by individual cairns is largely a function of local environmental constraints, their distribution limited by factors of survival and discovery. The susceptibility of the monuments to superficial modification is stressed and caution is urged in the application of classificatory schemes.

The Assendelver Polders, 20 km west-north-west of Amsterdam (fig. 1), consist of some 2000 ha of typically flat Dutch polder landscape, treeless and cut by hundreds of drainage ditches, primarily in use for grazing. Observations by amateur archaeologists and a few small excavations have shown that the area was relatively densely populated during the period 2150 bp to 1750 bp. There were also indications for occupation around 2500 bp. The site discussed in this paper dates to this earlier period, the Early Iron Age. Until recently, circumstances were good for the preservation of organic remains owing to the high water table of the peat which covers large parts of the polders. Wooden posts, representing structural elements, were regularly retrieved by amateurs, as well as occasional lengths of wattle-work. These favourable conditions were, however, brought to an end in connexion with changes in the allotment scheme, implemented between 1970 and 1978. The organic remains became acutely threatened by desiccation due to the lowering of the water table by c. 1 m. Because of this, the Albert Egges van Giffen Institute for Pre- and Protohistory, of the University of Amsterdam, initiated the Assendelver Polders Project. Fieldwork took place in the polders during the summer months of 1978–82.