Proceedings of the Prehistoric Society 39

TM 157134. An interim report of the 1969-70 excavations. The cultural, ecological and stratigraphical evidence recovered is compared with the results of previous investigations of the Pleistocene channel deposits as a basis for a general interpretation of the site and a reassessment of the Clactonian industry. The recovery of sharp artefacts and faunal remains associated with buried surfaces was an important feature of the excavations. Preliminary geomorphological and pollen studies suggest a series of occupations relating to the Anglian and early Hoxnian stages. The authors discuss the significance of their findings for the dating of the channel deposits and an interpretation of the relationship between the Clactonian and Acheulian industries in SE England. C B

[ST 422403]. The excavation of the Sweet track at two sites in the Somerset Levels has yielded unique material of late 4th millennium BC. The track was built across a reed marsh, and the builders used a variety of timber obtained from both nearby fen-wood and from the dry land of the Burtle sand and the Poldens. Pollen analyses indicate that certain areas of the Levels were particularly favourable for forest clearance and cultivation, and the debris from the track shows a range of woodwork so far unique to the British Neolithic. Au

The broad distribution of highland fields is indicated and some examples are described in detail from Ordnance Survey records. SW England, Wales, and England-Scotland north of the Trent are subdivided for consideration. Three principal forms of field remains are apparent, roughly datable respectively to 3rd-2nd millennium, late 2nd to mid-1st millennium, and late 1st millennium BC into RB times. Towards the end of the development came a significant decline in quality of highland as compared with lowland land-use, possibly a result of worsening climate enforcing stock-raising in the highlands.

A new analytical procedure for blade assemblages involves the construction of dispersion diagrams enclosed by three axes, allowing easy reference to the relative values of length, width and thickness. Triangular dispersion diagrams have obvious advantages over conventional two-axis plots, and any number of diagrams can conveniently be prepared using a single card master. If strict definitions for blade, length, width and thickness are adhered to, this is a reproducible technique which will allow accurate comparisons between blade assemblages analysed by independent workers. C B