in the relative and absolute dating of the building Ll and Rampart E of Phase 11 (pp. 111-15). The report is frustrating, for the reader has to work hard to understand it. Anyone unfamiliar with the castle has repeatedly to refer back to Illustration 1.11 to locate cuttings and to orientate detailed plans and even then information is often lacking. For the location of H. St. George Gray's trenches (p. 62) the reader should search Illus. 1.6, not 1.5. It is unhelpful to be told that the distribution of amphorae sherds may have been thrown out of building L2 through its south-western door (and one has to struggle to believe this when building L2 overlaps with Ll and not all of building L2 is shown on the artefact distribution plans 2.19) when to orientate the plan of the building (Illus. 2.20) the reader must refer back to the general site plan 1.11 (which also does not have a north point); in the next sentence you discover this was all a waste of time as building L2 is prehistoric. Reference is made on p. 24 to 'Cutting K (Illus. 2.12, sections B-B' and D-D')' whereas Illus. 2.12 refers to Site K (actually comprising ten cuttings) and D-D' is not marked. Two important artefacts of Period 11, a decorated bronze ring pendant and an iron axe-hammer, were found on, according to Illus. 2.17, a 'final' road surface in the south-west gate, although there was a subsequent repair (later, p. 29, 'a second surface') to the road, K 423, which is not shown on the section drawing (Illus. 2.13). On p. 29 the 'ring brooch came from K 554, the make-up of the repair or secondary surface, K 423'; K 554 is also not labelled on the section drawing. The section is a composite drawing of two sections 8-10 metres ('medial baulk') apart on different alignments; the caption refers to layer '659', presumably an error for 656, and it is by no means clear from the information provided that the gate structure is contemporary with the road surface '603', which is critical for the phasing of Period 11. When invited to consider the nature of the gate in relation to Banks 2 and 3 the reader is referred to Illus. 2.12; however, they are labelled neither there nor on Illus. 1.11. The reader will also find the discussion of the principal early medieval building L2 very difficult when in the caption to Illus. 2.21 we are informed (repeated in text) that the building was crossed by a medieval field ditch L 405 while on Illus. 2.20 the wall slot belonging to building L2, L 068, is drawn as though it cuts L 405 and is therefore later; Illus. 2.22 taken in 1969 and 2.23 (undated) give the impression that the wall-slot L 068 was excavated before the field ditch and yet 2.21 photographed in 1968 shows both fully excavated. The reader can also be sent chasing when dealing with the artefacts: 'Early Saxon glazed' pottery (p. 54) made this reviewer's ears prick up but, having failed to find this in the index, I concluded that for 'Early' you should read 'Late', as indeed it is later catalogued (p. 97).