Archaeologia Aeliana Series 5

Hadrian's Wall must have had a coordinated signalling system at all stages, or lose its effectiveness. Field tests indicate that, despite very difficult terrain, seventy-two wall installations from mile 35 onwards are in direct visual contact with a Stanegate site, with only one needing an intermediate relay. This explains the 'irregular' placing of milecastles. There would be no need for specialist signallers: beacons and messengers would suffice. The theory's predictive force was tested in the Newbrough-Grindon Hill area, and the revised signalling system once the Wall forts were in place is also worked out.

New information is presented for four parts of the wall which were freshly disengaged, excavated, and consolidated. No refinements to dating emerged, but differences in character were seen from place to place. A section of ditch between Heber Tower and Morden Tower revealed a quantity of pottery and small finds, botanical remains, animal bones, and soil particle evidence. A watching brief was held for a stretch of wall on the quayside.

Exploration of the quayside revealed successive tipping and new layouts, from later 13th century onwards. Industrial installations were in use beside Burn Bank in 15th and 16th centuries before being cleared for another use. Finds included textiles and cordage, caulking, leatherwork, metal and wood, and biological material including much fishbone.

The sixteen ballista bolts from Vindolanda, dating AD 85-125, support the views expressed earlier (see 89/552). There is discussion of ballistarium as magazine and workshop for tormenta.