Issue no 39, November 1998

A block of fields with prehistoric boundaries containing the remains of medieval ridge and furrow has been identified in South-West Cambridgeshire - perhaps the first example of prehistoric field boundaries to be recognised in the part of England dominated by the open field `strip farming' system.

It was thought that all traces of prehistoric fields were obliterated when open fields and nucleated settlement were introduced across much of Midland England between the 9th and 12th centuries. `Celtic' fields were thought to survive only in East Anglia, the West and North, and in upland areas such as Salisbury Plain. However, an analysis of aerial photographs, early maps and field survey by Sue Oosthuizen, a landscape historian at Cambridge University, has suggested that prehistoric fields around the village of Caxton may have been incorporated into the furlongs (or blocks of strips) of the three-field system.

The village of Caxton lies in a shallow valley and is bisected by an early Roman road, Ermine Street, which runs north-south. According to a pre-enclosure 18th century map, the fields to the north and south of Caxton were mostly aligned on the road, with the strips running either parallel or at right angles to it. Immediately around Caxton, however, the fields are aligned NE-SW and lie in a block bisected by the road, indicating, according to Ms Oosthuizen, an arrangement that predates the road.

The fields, with ridges up to 260 yards long, are much larger than surviving Celtic fields elsewhere in England, suggesting that some internal boundaries at Caxton may have been `rubbed out' to create larger areas that still respected the former fields' alignment. The lanes climbing out of Caxton's valley are also aligned NE-SW, even on the higer slopes where fields with a `later' alignment predominate, and these may be survivals of prehistoric tracks.

The overall pattern of parish boundaries in the area shares the NE-SW alignment. `The alignment of parish boundaries to these small fields at Caxton shows that the underlying settlement and agricultural pattern of the area owes more to prehistoric land-use than has previously been realised,' Ms Oosthuizen said.

Before the Norman Conquest, over two-thirds of the land in Caxton was owned by `sokemen', or small freehold farmers - rather than by grander overlords. This pattern of ownership may have influenced the survival of earlier boundaries. The freemen of Caxton perhaps decided, for convenience, to transform the fields they already had into the ridge-and-furrow system rather than lay out an entirely new pattern - a process that could have been achieved in a single season between harvest and sowing.

Hillforts occupied `only in later years'

Further light on the function of Iron Age hillforts has been shed by a geophysical survey of 18 hillforts in Wessex. The survey, by English Heritage, has confirmed, broadly, that lightly-defended early enclosures contain little or no settlement evidence, while strongly-defended later enclosures were typically densely settled - a hypothesis suggested by Barry Cunliffe some years ago during work at Danebury in Hampshire.

Within this broad distinction, however, the survey showed, unsurprisingly, that each hillfort differed from the others in numerous points of detail and that there was no standard pattern of development during any period.

The hillforts were distributed across Wiltshire, Hampshire, West Berkshire and South-West Oxfordshire. Examples of lightly-defended sites containing no more than a few pits include Woolbury (Hants), Perborough Castle (Berks) and Martinsell Camp (Wilts). These enclosure - perhaps used seasonally as gathering places - are thought to be early by analogy with excavated examples of similar form elsewhere, such as Balksbury in Hampshire.

Forts with numerous lines of ramparts and intensive occupation evidence, also dated by analogy, include Barbury Castle and Castle Ditches (both Wilts). Castle Ditches contained over 50 roundhouses, streets, enclosures, and evidence of zoned activities. `It has all the characteristics of a defended town,' said Andy Payne, director of the project.

Interesting individual forts include Oldbury (Wilts), Ladle Hill Camp and Norsebury Ring (both Hants). Oldbury was found to contain a distinct inner enclosure with intensive occupation, suggesting the fort either retracted at a later date (possibly during Dark Age re-occupation) or grew, like Maiden Castle. Ladle Hill has long been interpreted as an unfinished hillfort, from what seemed to be abandoned dumps of building spoil and from the incomplete defensive circuit. The survey's discovery of no interior features whatsoever seems to confirm the interpretation.

Norsebury is almost completely ploughed out, but the survey traced its entire circuit. From one of its two entrances, a ditched avenue led to a large sub-circular enclosure some 30m across in the centre of the fort. According to Mr Payne, the enclosure may have contained a timber shrine, as is thought to have existed at Danebury.

Bronze Age `bunsen burner' discovered

One of the fundamental mysteries of Bronze Age metal-working may have been solved with the discovery that metal-workers of the 2nd and early 1st millennium BC possibly possessed a prototype of the bunsen burner.

The mystery has been to explain how early bronze-smiths produced much of the intricate, decorative metal work known from the period. Few metal-working tools are known, certainly none capable of handling filigree work or soldering.

Now, however, a perforated ceramic Bronze Age artefact with a central hole in its base, traditionally interpreted as a culinary device, has been recognised as a possible early bunsen burner. The discovery was made during experiments by Jacqui Wood, an archaeologist who specialises in reconstructing various aspects of Bronze Age life (see BA, July 1995).

Earlier this year, Mrs Wood was examining one of the perforated pots at the Lake Ledro Museum in Northern Italy, and noticed its inside had been subjected to intense heat, making it resemble crucibles used in reconstructions of tin and bronze smelting. Thinking at first that it was a kind of lantern, she lit a fat-soaked rushlight underneath and watched the effect.

`It immediately produced this tall, pencil-like flame through the hole in its base,' she said. `The air rushed in through the holes in the sides and the flame did not blow out in the wind.' The value of such a flame is that it would have allowed the metal-worker to get his or her hands close to the flame without getting burnt.

The artefact was widely distributed around Europe, and has been found in Italy, Sweden, Poland, and possibly Bavaria, Hungary and Lithuania - but not Britain. It had been thought of as a device to prevent a skin forming on heated milk in a pan, or to strain curds in cheese-making. `People just assumed that if it was ceramic, it must be domestic,' Mrs Wood said.

Peter Northover, an archaeometallurgist at Oxford University, said that to test the hypothesis you'd need to look for traces of metal on the Lake Ledro pot and estimate the temperatures implied by the scorch-marks inside. `But if she's right, it could explain why it's found on the Continent and not here. You wouldn't have needed that sort of flame to make British bronze and gold - and we don't have any Bronze Age silver - but you would have needed it to make some of the goods found the Continent,' he said.