Neolithic long mound some 546 metres in length, comprising a bank of earth between two parallel ditches. The ditches are 19.5 metres apart. It follows a slightly curving east-west course across the interior of Maiden Castle. Richard Bradley had suggested that the section of the bank barrow immediately west of the earliest Iron Age rampart (and thus just outside the causewayed enclosure (SY 68 NE 7)) had originally been a more conventional long barrow before being considerably extended in both directions. The RCHME survey appears to support this: this section of the barrow, 65 metres long and 15 metres wide, is considerably higher than the remainder of the monument. However, only excavation could demonstrate chronological precedence for any given section of the monument. The three elements of the monument also show differing alignments, and are separated by sections without a bank. The contour survey shows that for the whole of its length the barrow is set on a false crest, varying from some 10 to 20 metres north of the summit of the ridge, suggesting that it was deliberately. It appears not to have performed a funerary function, but may have acted rather as a symbolic barrier or territorial divide. It has been excavated on at least three occasions - by Mortimer Wheeler in the mid-1930s, by RJC Atkinson in 1951, and by N Sharples in 1985-6. It was surveyed by RCHME in 1984-5 by RCHME. Wheeler encountered some post holes, which he felt were Neolithic and therefore associated with the barrow structure. Two intrusive Saxon inhumations were also found just below the surface at the east end of the barrow. One was a supine burial with its head to the west. Grave goods including a seax and knife were recovered and indicate a seventh or eigth-century date for this burial. The second burial had been mutilated and dismembered, and did not have any grave goods. Radiocarbon dating suggests a 7th-9th century date for this burial.