Proceedings of the Prehistoric Society 33

Attention is drawn to the dangers of uncritical use of some simple statistical methods, particularly percentage frequency graphs, under five main headings: sample, percentage, ordering, typological and perceptive errors. Of these, the authors devote most attention to ordering errors arising from the arbitrary arrangement of a sequence of artefact types along the horizontal axis of the frequency graph, and they underline the elementary point that rearrangement of the artefact sequence leads to a quite different shape of graph. They next discuss typological errors, in which the basic uncertainty is inherent in the subjective assignment of objects to types which are not themselves rigorously defined, and conclude with some remarks on perceptive errors in comparing lines or curves. The pitfalls are all of the kind which would be avoided by the use of ordinary critical common sense; the message is summed up in the sentence: "Archaeologists are warned that although certain statistical techniques will provide useful tools .... these techniques will not automatically do their thinking for them". KEP

A study of animal populations, pollen and technological evidence from Neo and BA settlements in the Alpine Foreland suggests in some detail how the different societies shaped their natural environment. The application of statistical methods to prehistoric faunal samples provides some means of assessing herd compositions and breeders aims. For example, the absence of working oxen in Cortaillod and Michelsberg cultures implies hoe agriculture; and the pollen evidence bears this out. Horgen cultural achievement may be greater than its crude craftwork suggests, and the "nomadic" Corded Ware peoples in fact practised settled agriculture with the ox-drawn ard, relying little on hunting. Increased overwintering of stock - difficult at this high altitude - was achieved in EBA by careful pasture management. Later still, the Swiss Urnfield settlements achieved an efficient integration of crop and animal management which may partly account for their wide distribution in Central Europe. In general, the reduction of damp forest by increasingly efficient means gradually removed the main limitation on farming development. See also Man, 3, 1968, 64-75.

SP 526348. Several cuttings were made through the defences, the W entrance was fully excavated and small portions (chosen after magnetic survey) of the six-acre interior were examined; there were four EIA and three later phases. Phase 1, ?6th/5th cents BC, was pre-rampart occupation. Phase 2, ?5th cent BC, saw the erection of an impressive inner rampart of three-tiered wall construction and an outer, dump-built rampart, each with a V-ditch. The inner entrance, fully inturned, had a pair of C-shaped guardrooms. After continuing occupation the fort was strongly attacked and fired in ?early 4th cent BC. Disuse was followed by Phase 4, a renewal of the defences in quite different form in ?late 2nd cent BC. Abandonment after this occupation was followed by two RB phases, and 18th cent landscaping further altered the site. Appendices treat the magnetic survey, land mollusca, archeomagnetic results from hearths, human and animal bones, and edible snails.

SU 184419. A large bowl barrow standing some 8ft high before excavation was found to be the final phase of a monument with at least two earlier phases: a primary circular ditched enclosure was succeeded by an unaccompanied inhumation in a deep oval grave pit attributed to the Beaker period and covered by a low bell barrow under which were concentric stake circles. This mound was subsequently truncated in the Food Vessel phase and four inhumations and several cremations, one in an Enlarged Food Vessel, were inserted. Remains of a fire and a concentration of flints were found over these burials. A final mound of turf and chalk was built and further cremations and inhumations continued to be inserted into it and the surrounding ditch. On the evidence of the pottery and 14C dates for two phases, the site was used as a cemetery from c 2000 BC for at least 1000 years and probably longer. Au