Proceedings of the Prehistoric Society 55

Continuing erosion of the axe production sites prompted a field survey which identified and recorded 566 distinct working sites within 35 groups. A fourfold categorization has been developed. Production was on a massive scale and appears to have developed towards greater sophistication and organization. Estimates of relative flake quantities for each group of sites allowed comparison of varying levels of activity at each group, and differing strategies at the two main production sites were also examined. Petrographic studies of outcrops and debitage suggest the possibility of attributing dispersed products to specific production areas. Au (abr)

A complete gazetteer of all 900+ sites is offered and their distributions, locations, and characterization discussed along with a brief examination of dating and significance. Two main types of rock art are defined, the Boyne or Passage Grave type (with a wide range of motifs) and the Galician type (with a more restricted range). These two types overlap occasionally, on natural rocks and tombs. Sixteen main rock art regions are distinguished. A few sites in the gazetteer are noted as later than AD 1.

The image of the 'footloose Celtic cowboy' in the north has finally been dissolved by the discovery of numerous sites in the Border country of very narrow ridged cultivation (cord rig), associated with prehistoric settlements of various types, and of many examples of pre-Roman cultivation terraces. Arable farming was clearly important in the Tyne-Forth area. The Northumbrian evidence is presented here, with gazetteer for it and for the Scottish material which is to be expounded later by the second contributor.

Renewed stratigraphic and pollen analytical research was conducted. At Peacock's Farm, Shippea Hill, most occupation was of Meso date, c 18000 BP uncal, with minimal Earlier Neo. At Letter F Farm, Littleport, there was some Meso but mainly Earlier Neo of c 5000 BP uncal. The 'Meso black band' at Peacock's Farm spans 1700 years, mainly 8500-6800 BP, and Meso occupation areas on the sandy ridge date to the first half of this period. Dense forest shrouding the wetland and sandy ridge received minor damage at c 8500 BP, the start of the Meso occupation; substantial clearance some 250 years later was followed by forest regeneration. On balance the evidence suggests pronounced human rather than climatic impact. Increased territoriality from 9th millennium BP is suggested by the microlith styles. The Neo evidence appears to indicate short visits only.

A bone flute from the cave at Veyreau, Aveyron, associated in limestone concretion with a bone dated to 4th millennium BP uncal, is a five-hole block and duct piece closely similar to the White Castle flute. It raises questions about the development which must have preceded the Veyreau flute (see also 90/325).

Detailed field survey is reported and a comparison made with the fragment of chevaux-de-frise excavated at Castell Henllys (Dyfed) and with other sites having this type of defence (or status indicator). The new evidence gives further support to several of Harbison's generalizations (see 72/1394), and suggests a late Iron Age date for Carn Alw.