Collingwood's drawing (1907, fig. a on 342) is misleading since the two pieces are arranged one above the other and made to match an upside-down cross-head. The following description rests on a new reassemblage.

A (broad): A flat edge moulding continues across the top of the panel. At the right, at the base of the fragment, is a full face head with incised eyes and nose in a single line and a pointed hat or helm. The area alongside this figure is lost. Above, occupying the whole width, is a crude human figure with legs astride and elbows raised in a crab-like manner. The eyes are drilled and the nipples are incised circles. At each side is a broken angular frame.

B (narrow): This is the longer surviving side. The flat edge moulding, which extends across the top of the panel, contains a slightly slimmer inner flat moulding. Within the panel is a continuous run of bungled interlace, apparently a form of three-strand plain plait with irregular breaks resembling Carrick Bends. There is an irregular looped terminal at the top. The strands are lightly median-incised.

C (broad): The flat edge moulding extends along the top of the panel. Within is the top and pendant head of a profile beast with double outline and fetter bands. The fore leg is crammed against the top of the panel, the foot having three frond-like toes. The head has an incised elliptical eye, a spiral, and looped jaws, threaded by a bar which is crossed by the tongue. A fetter crosses the chest. The surface is damaged in front of the snout.

D (narrow): The edge moulding is flat and flanks an inner frame, a flat moulding, which extends across the top of the panel. It contains a three-strand plain plait using fairly well-modelled, median-incised strands.

This shaft very closely resembles the ring-headed crosses at Middleton (nos. 2 and 7). The portrait head with pointed hat on one face with the Ryedale bound dragon on the reverse, together with the interlace on the sides, matches the ornamental scheme of the Middleton group. The beast's jaws derive from those at Sinnington and tie in with both Levisham and Middleton. The present animal, in the disposition of its leg and fetter, and in the treatment of the jaws, is almost identical with Middleton 2 (Ill. 680). The bungled interlace is another indication of that sculptor's work.

The strange stance of the upper figure has Viking parallels, notably the toggle from the Thames (Wilson and Klindt-Jensen 1966, pl. XLVf), though an equally primitive, phallic version occurs on an unpublished carving at Thornton Steward in the North Riding. There are also Manx parallels for the akimbo stance at Bride (Kermode 1907, no. 97B) and the circular nipples at Michael and Maughold: (ibid., nos. 129 and 72A).