Archaeologia Aeliana Series 4

[NY 9864]. An account of successive structural periods is given, not dependent on unpublished evidence; a table giving dates derived from the whole site is also printed. The striking features are the re-arrangements of the structures in the central block apart from the headquarters building. In Period I there is a hospital and large granary, in II a commander's house, in III two granaries, in IVA a timber commander's house and a stone ?granary, in IVB a stone commander's house and ?granary. A curious long E-W subdivided building along the via principalis, found in several periods, resembles in two of them small-scale barracks. Dates and suggested garrisons are: IA, 73-c 87, ala quing, plus legionaries; IB, c 87-c 105, same; II, c 105-c 122, coh mill eq; III, c 122-125, coh mill ped; IVa, 139-c 158, ala quing; IVB, c 158-162/3, coh mill eq; VA, not a fort but ?depot, 162/3-c 180; VB, c 180-?; VC, ?-c 207, ?abortive town. B D

NY 829748; NT 774009. The two circles at Goatstones and Three Kings near Hexham are outliers of the Scottish series of "four-poster" circles, whose most characteristic features (eg. rectangular setting, grading of stone heights, SWINE axis, sepulchral function) link them with the Clava-Recumbent Stone Circle series of NE Scotland. Such associations as they have suggest a date for four-posters in mid-2nd millennium BC. They seem to have begun in Aberdeenshire, spreading to Perthshire and further afield. In four of the five excavated examples cremations were found, sometimes with urns. These are humble, local burial monuments, not observatories; thirty-four examples from Banff to Radnorshire are catalogued.

NT 998164; NU 011158. Excavation on these sites, which lie on the same hill about 1 km apart, was directed mainly towards obtaining material for 14C assay. Brough Law hillfort, bivallate and entirely stone-built, has a double outer face to the inner rampart. No earlier palisade was found. Material from beneath the rampart, possibly consistent with its construction, yielded a date in mid-3rd century BC. The settlement on Ingram Hill, a palisaded site followed by a slight ditch and bank with palisade, was suspected of being later than the normal floruit of palisaded sites. Material from the later bank gave a date in later 3rd century BC. Au

[NZ 2750]. A new plan of the fort is given, with an account of recent work including part of a building identifiable as the commander's house and evidence for the new lines proposed for the ramparts. BD

Further flights have revealed many more sites on the coastal plain and cultivated valleys, particularly the Till valley. These are roughly classifiable as: rectangular single-ditched enclosures, which may contain one or more round houses; rectangular double-ditched enclosures; curvilinear enclosures, including possible henges; possible barrows; and field systems. Additional structures at Yeavering Bell and a Roman fort complex at Washing Well, Gateshead, are new finds of particular interest.

Although many difficulties surround the identification of Roman provincial gem-cutters' workshops, four intaglios in red jasper from S Shields and Corbridge have enough technical similarities to suggest they are products of one such workshop. The S Shields gem may represent Silvanus-Cocidius.