All four stones carry relief ornament on [A] and [B] but decoration only fills a small part of face [B] immediately adjacent to the angle of the arch. Bruera 3 differs from the rest in that it is not a through stone passing through the thickness of the opening and its ornament is continuous from [A] to [B] around the angle. The other voussoirs taper towards the inside of the wall. No border moulding exists between the ornamental panels on the two faces of Bruera 4, 5 and 6 but the decoration is not continuous across the angle.

Bruera 3 (Ills. 66–7): The decoration is formed by thick fleshy strands and can be read as an angular central-stemmed stripped spiral-scroll, set horizontally, which springs from an angular U-shaped container on [B].

Bruera 4 (Ills. 62–3): On [A] are a set of spiralling forms emerging from a horizontal stem. On [B] two inward-spiralling mouldings drop from a broad flat horizontal moulding.

Bruera 5 (Ills. 64–5): On [A], to the left, are a series of three chevrons set on their side. Flanking their apex and to the right is somewhat disconnected spiral-scroll ornament. On [B] are two parallel mouldings which terminate in outward spirals.

Bruera 6 (Ills. 68–9): On [A] is a central-stemmed stripped scroll with outward-spiralling branches. On [B] a central moulding splits into outward spirals.

All of these stones are clearly re-used in their present position and most have been re-cut. In their current setting they post-date the Norman doorway since none of the arch stones are bonded to that feature. Given: (a) that they are set irregularly around the arch; (b) that Bruera 3 differs from the others in its ornamental organisation and in not being a through-stone; and (c) that there is no balance of ornament within the group, it follows that the arch has been formed from an assemblage of stones which originally fulfilled some other function. In addition, the [B] face of nos. 4, 5 and 6 has clearly been cut away to accommodate the various decorations surviving on [A].

The following hypothetical phase sequence would explain these observations:

(i) The carvings now represented by nos. 4, 5 and 6 were originally vertical shafts, left undecorated at the base but with sides ornamented towards the top with varieties of vertical mouldings terminating in spirals represented now by the [B] faces.

(ii) The tops of these shafts were then cut off and the decoration on faces [A] added. This stage must pre-date any arch usage since the decoration on Bruera 5 [A] bears no relationship to that on face [A] of nos. 4 or 6 — and the decoration of that stone does not read logically in its present position.

(iii) Re-use in present arch with, probably, Bruera 3 added to the set.

All of the decoration on [A] and [B] can be paralleled within the late pre-Norman period. Irregular stripped and spiralling scroll-forms recur within the region on Cheadle 1 and Whalley 3 (Ills. 73, 675, 677) — and the latter cross has Bruera's combination of chevrons and spiralling scroll. Pendant mouldings with spiralling terminations can be found at Ellerburn, Levisham and Sinnington in Yorkshire, and Sockburn in Co. Durham, as well as Kilfenora in Ireland (Lang 1991, ills. 435–6, 631–4, 810; Cramp 1984, pl. 137.737; Harbison 1992, ii, fig. 371).

On this reading of the arch-stones sequence, Bruera 1 and 2 would fit best with phase 2 of the arch development.