1a, in south chancel wall (exterior) 1m above ground level and c. 4m from south-east corner. 1b, in the course above 1a and one stone to the west. 1c, same course as 1a, but 2m to the east.

'During the restoration some Sax. stones, sculpted with knotwork, came to light; they are built into the walling of the chancel' (Cox 1924, 254). We presume, however, that Cox was suggesting that these stones were first noticed during this restoration work (presumably that of 1903–4) rather than discovered here during that time and reset in these positions. The account of the restoration work does not mention them at all ((—) 1903–4a, xi–xii). It is therefore probable that the stones were originally reset in these positions when the chancel wall was built in the early thirteenth century.

St Clement

All moderate, weathered

Stone 1a. A fragment from a much larger stone. The panel represented here has a cable-moulded border around two sides, and is decorated within by interlace in low relief. The strands have a double incised medial line. The interlace itself is a simple pattern E knot. This run develops in a most unusual way: the two loose ends are twisted around each other to form a vertical division in the interlace pattern, resembling a vertical cable moulding, before they develop further (off the present stone) into what was probably a long run of four-strand plait.

Stone 1b. A fragment which retains only its cable-moulded border and, next to this, the crests of a run of interlace of unknown type. The interlace was clearly quite tightly organised but little more can be said of it.

Stone 1c. A fragment which seems to incorporate a whole panel of interlace in low relief framed by a cable-moulded border at one end, beyond which is a second, unmoulded border, and borders of unknown type on the remaining three sides. The interlace itself is motif type vi (Fig. 10) and has a single incised medial line.

Stone 1c appears to be a panel from a mid-Kesteven cover (Chapter V), and probably represents a vertical panel from one end of the flank. Stones 1a and b are only connected with 1c because of their identical stone type and because they were all reused in conjunction. Like 1c, however, stone 1a has interlace of very much the same technique and can be seen as part of a panel from the flank of a mid-Kesteven cover. But if 1a does come from the flank of such a monument, it has the unusual feature (both within the mid-Kesteven group and within Anglo-Scandinavian interlace in the county) of the interlace developing continuously into a vertical run of cable-like two-strand plait, which itself forms an internal boundary within the pattern but allows the interlace to be continuous through the border. Such a device would be quite appropriate for the division between the transverse terminal and the long horizontal panels, taking the place (as it were) of the usual cable-moulded border between these two panels. So this attractive and unusual device is probably not an objection to an interpretation of the piece as the flank of a mid-Kesteven cover. The only other example of two-strand plait in the county, Syston 3a (Ill. 365), is on a panel of uncertain type but probably not from this group of covers. Even this example, however, does not provide a good parallel for Rowston, as the two-strand plait is not integrated with the interlace to either side. Stone 1b can also be reconstructed as a part either of the flank or the lid of a mid-Kesteven cover, but so little survives that this is not the only possible reconstruction. However, given its reuse alongside 1a and 1c such a reconstruction is made more likely.

Stone 1c, then, is certainly from a mid-Kesteven cover whilst 1a and 1b can be reconstructed as such (Fig. 9). As all three were reused in association, an origin within a single monument of this type is a strong possibility and consequently a date between the mid tenth and the early eleventh century is probably indicated.

Mid tenth to early eleventh century

Cox 1924, 254; Davies 1926, 18 (stone 1a only); Pevsner et al. 1989, 616