On window sill in church, currently (2007) set upside down

First published in 1850 ((—) 1850, 115). Excavated from the churchyard in 1843 (so Browne 1887a, 11–12 and Taylor, H. 1906, 205) or 1830 (so (—) 1909, 169) and then placed close to the external east wall of the chancel. It was still in the churchyard in 1920 ((—) 1920) but was moved inside at some date after this. The stone had clearly been available in c. 1793 since one arm was cut back for use as a memorial inscription at that date.

St Oswald

The upper and lower arms are lost and the end-panels are somewhat worn; the stumps of the former ring are visible on the (original) underside. The right arm of face C was re-used in the eighteenth century as a grave-marker; the inscription reads: Roger Lowe/ Houghton/ 1721. / Thomas & Alice/ Holcroft/1793.

A (broad): The entire decoration on this face is framed by a roll-moulding border. Further vertical mouldings separate the decoration in the arms from that surrounding the encircled central boss. In general the decoration is ill-balanced. (i) In the left arm there are two areas of ornament. Along the top, and to the right, is a run of Stafford knots (simple pattern E) whose linking strands, with one exception, cross and re-cross; there is a markedly angular bend at one point. Below are four and a half blocks of fret pattern, based on a square unit with two crossing diagonal elements plus angled terminals, creating four opposed arrow-head triangles (type U3: Edwards, N. 2007a, fig. 7.14). (ii) At the centre of the head is a flat boss decorated with four interlinked Stafford knots in a cruciform pattern. The decoration around this is divided horizontally into two distinct areas. At the top are six blocks of fret pattern (type U3 as above), irregularly laid out to fit the curve of the arm-pits, whilst below is a run of linked Stafford knots; the connecting strands cross and re-cross at the sides but are organised in two adjacent runs below the boss, the lower run following the curved borders of the missing lower arm. (iii) In the right arm the decoration is again divided into two areas: at the top and to the right is knotwork (like that in the left arm) consisting of two crossing strands which, with one exception, form a Stafford knot between crossings. Along the bottom and to the left are three rectangular blocks of fret pattern, again of type U3 but stretched to fit the broader panels.

B (narrow): The figural scene on the end of the arm is flanked by a roll-moulding border on three sides and a flat undecorated area at the base; this lower area carries a long narrow slot which may represent later damage or originally have formed the setting for a metallic addition. In the centre of the panel is a naked figure, with large head and hollow ears, who is suspended upside down from a strand emerging from the upper border; the feet are set in profile, the rest of the body is seen en face. His arms curve down alongside his head, one ending in a hand, the other apparently terminating as a stump. He is flanked by two figures, seen in profile, who also appear to be naked. They have pigtails, marked ears, long noses, round eyes and open mouths — the left figure is depicted as though smiling, the other has a down-turned mouth. With one hand they each grasp the suspended figure's legs whilst one of their feet apparently passes behind or onto the victim's head. With their other hand they grasp a thin curved moulding which runs across their bodies to meet the lateral frame, disappearing behind (or into) the suspended man. A bulbous protrusion emerges from the lower part of their stomachs; this does not appear to be connected to the curved moulding.

C (broad): This face is now inaccessible, and the description is based on Dr R. Trench-Jellicoe's photographs, taken in 1990. (i) In the left arm is a boss which has been roughly cut flat, surrounded by three quadrupeds. The uppermost one is facing left, with two ears and a ?tail passing across the body. A second beast is set at the end of the arm, facing downwards with ear, tail and open jaw. The third is shown upside down in relation to the onlooker and closely resembles the upper beast in ears and enveloping tail. (ii) At the centre of the head is a similar boss, also cut flat, surrounded by pelta ornament, now worn but including a variety of interlocking C-spirals. (iii) The ornament in the right arm has been removed for an eighteenth-century inscription.

D (narrow): The end of the arm is occupied by a framed figural scene. At the centre is a forward-facing human figure, with a flat head, hollowed ears and fringed beard; the eyes, nose and mouth are lightly modelled. He is dressed in a full long-sleeved garment (alb); over this are faint remains of a pointed chasuble. There are traces of decorative strips representing embroidery around the edges and neckline of the chasuble and there is a central bar of ornament up the front of the vestment; further decoration can be seen on the lower hem of the alb. The figure, whose booted feet face to his left, grasps the handles of two sub-rectangular objects which narrow from base to top; below each of these is a triquetra. To the right of his head is a long-stemmed cross (probably of type B6) and there is a similar cross alongside his right arm. In the upper left corner is a building (?church or reliquary) with an off-centre door, inward-turning gable finials and a central trefoil or cruciform feature on the roof.

E (top): The upper faces of both arms are plain, with traces of a ring.

F (bottom): The underside of each arm-end is decorated with a single framed panel of fret pattern. (i) The unit to the left is bisected by both right-angled and diagonal elements, creating four squares each containing four triangles with central angled depressions. The borders of these appear to interlace and form wedge-shaped crosses (Allen and Anderson 1903, ii, no. 724). (ii) The unit to the right has two crossing diagonal elements plus angled terminals, type U3 (Edwards, N. 2007a, fig. 7.14) but with double outline, creating four opposed arrow-head triangles interlocked with the angled terminals. (iii) Between these panels and the curve of the cross's armpit is the stump of a ring.

This is one of the largest pre-Norman cross-heads to survive from England; only a fragmentary cross-head from Lastingham in Yorkshire approaches its dimensions (Lang 1991, 168, ills. 587–91). Clearly it is an ambitious monument, in scale comparable to such massive crosses as those from Monasterboice, Co. Louth; Ray, Co. Donegal; and Iona (Fisher 2001, 170–1). The date of the carving is firmly indicated by the survival of the stub of the ring which once connected the arms of the cross, since all the evidence we have shows that the ring-headed form was not exploited in pre-Viking period Northumbria, but was introduced to northern England in the tenth century from Ireland or western Scotland where the type existed at an earlier date (Collingwood 1926a; Bailey 1978, 178–9).

Some of the ornament on its main faces is not unexpected in Viking-age Northumbria. Thus the organisation of three free-style animals set around a boss is repeated in the Tees valley at Winston in Co. Durham and Forcett in Yorkshire (Cramp 1984, pl. 147.774; Lang 2001, ill. 250). Similarly the run of Stafford knots also reflects a local taste, for this is a form of knotwork which had a restricted popularity to the west of the Pennines in tenth- and eleventh-century England (Bailey and Cramp 1988, 34–40).

But, alongside these regional forms, there are important elements in the shape of the head and its decoration which seem to reflect tastes transplanted from Ireland or Scotland. The cross-head, which combines a block-ended arm, taller than it is wide, with small circular or near-circular armpits, and a narrow connecting ring set close to the armpit, is a type which is familiar on slabs in Scotland and, in free-armed form, in both Ireland and (once) in western Scotland (Allen and Anderson 1903, iii, figs. 210, 227A, 231A, 309A, 322A, 393, 397; Harbison 1992, ii, figs. 38, 101, 110, 221, 405, 427, 446). It is not, however, a shape which can be readily paralleled in England (the closest is type A11). Similarly, decoration built up from blocks of fret pattern is not widely used in pre-Norman Northumbria, and certainly never on a cross-head, but it is very popular — often in that position — on sculptures in Ireland and Scotland (Harbison 1992, ii, figs. 105, 152, 364, 446; Allen and Anderson 1903, iii, figs. 32, 51, 151A, 227, 231A, 235A, 305A, 306A, 309A, 314A, 322; Cramp 1983, fig. 115). The same can be said about the extremely competent version of spiral ornament at Winwick; this cannot be matched elsewhere among Northumbrian or northern Mercian carvings, but it is a persistent element in the decorative repertoire of carvings in Scotland and Ireland (Allen and Anderson 1903, iii, figs. 66B, 139, 235A, 393; Harbison 1992, ii, figs. 15, 155, 398, 516).

One final element can be added to this catalogue of non-Northumbrian forms. The central boss on face C is part of a five-boss arrangement whose disposition and numerical significance is discussed under Cheadle 1 (p. 61). At Winwick the central boss on face A carries additional ornament. This is a highly unusual feature among northern English carvings, the only real parallel being provided by a very elaborate cross-head from St Mary Castlegate, York (Lang 1991, ills. 305, 308). It is, however, a familiar feature of carvings from Ireland and Scotland.

The evidence of the two main faces of the cross thus points to the Celtic world as a source for much of its inspiration. That conclusion forms a useful background to examination of the figural panels on the ends of the arms and their relationship to each other.

When Ellacombe (1872, 526) produced the first published sketch of face D in 1872 he captioned it: campanarius antiquissimus in camisia vestitutus. In favourable lighting there is, however, no doubt that Ellacombe's camisia is in fact an alb overlaid by a chasuble (Ills. 708, 715). This figure is therefore a priest who is shown wearing full mass vestments, which carry elaborate detail carved in low (and now almost totally obliterated) relief. Below the alb his feet are set in high-ankled shoes of the same type as are worn by the figure on the Anderton 1 shaft, by an angel on a carving from Slaidburn in Yorkshire (now Lancashire) and by clerics and saints on various Pictish sculptures (Ills. 398, 401; Bailey 1980, fig. 68; Henderson, I. 1998, 157–8; Henderson and Henderson 2004, ills. 204, 221). Similar priestly figures are found on Viking-age carvings elsewhere in Northumbria at Brompton in Yorkshire and, closer to hand and with an equally elaborate chasuble, on Neston 1 in Cheshire (Ill. 197; Lang 2001, ill. 40). Both of these latter figures carry maniples whilst the Neston man also holds a chalice aloft. They are not, however, accompanied by the plethora of supporting images which are found in the Winwick carving. These warrant closer examination.

In the upper left corner of the panel is a tower-like structure, with two finials flanking a taller form which may be cruciform in shape. There is a narrow arched door in the lower left. It is possible that this feature was intended to represent a reliquary of some kind — forms in metalwork with end finials and/or suspension loops are known, and some of these have raised ornament at the centre of the ridge (Blindheim 1984, 1–53; Youngs 1989, 129–30; Webster and Backhouse 1991, 175–6). But the presence of a seeming door and the tall vertical shape suggest that this is a towered church, and it may not be entirely coincidental that coinage from Chester of c. 920 has remarkably similar forms among its decorative novelties (Dolley 1970, 42–3, fig. 9.).

The priest is carrying a pair of bells or buckets. Browne (1887a, 17) favoured the bucket interpretation. In his support we could cite the numerous buckets carried by the diminutive figures on the ninth-century Drogo Sacramentary ivory cover, to argue that this type of object was used liturgically and could thus be an appropriate attribute of a priest (Hubert et al. 1970, pls. 214–15). As supplementary evidence we could call on the eighth-century bronze bucket from Hexham, to demonstrate that these containers were perfectly familiar in Northumbrian contexts (Bailey 1974b, 141–50). Close examination of the carving convinces me, however, that both of these objects are actually hand-bells. These are of course a familiar survival, often in reliquary form, from the Celtic world, the majority of examples coming from Ireland where they were central to the ecclesiastical tradition. In his definitive studies, Cormac Bourke was able to distinguish two types (Bourke 1980; 1983; 1997b). The form represented here at Winwick, which tapers from mouth to crown and has its handle fixed part-way in along the top, has all the characteristics of his Class 2, whose distribution is markedly to the north and east of Ireland. Whilst we can assume that the Anglo-Saxon church was familiar with such objects — acolytes carrying a pair of bells are shown accompanying the funeral of Edward the Confessor on the Bayeux Tapestry (Wilson, D. M. 1985, pl. 29) — it is nevertheless significant that in terms both of survival of, and documentary emphasis on, this type of object, the Winwick bells are once more leading us towards Celtic areas.

This comment becomes even more significant when we assemble the evidence for representations of figures with bells elsewhere in Insular sculpture before c. 1200. There are only five such examples; on all, the figure also carries a crosier, and they all come from Ireland; to these can be added the priest depicted on the shrine of the Stowe Missal (Henry, F. 1965, pl. 73; id. 1967, pls. 9, 14; id. 1970, 183, pl. 30; Harbison 1992, ii, fig. 87). It is clear that, on these Irish examples, the bell is being used as an ecclesiastical marker; this must also be at least part of the function here at Winwick.

The presence of two bells is slightly unexpected since these objects are usually made, conferred, bequeathed and indeed rung singly. But the Bayeux Tapestry suggests that such pairing might have funeral associations. More significantly perhaps, Cormac Bourke has drawn my attention to a passage in the Life of Maedhog in which the saint leaves two bells to Drumlane church to be used for excommunication (Plummer 1922, i, 270, ii, 262). The Winwick bells are thus not only symbols of a priest, but probably also indicative of his authority.

The other elements set around the priest are equally symbolic of sanctity and power. The two crosses clearly fulfil this function but so also do the triquetra at the base of the panel. This form of ornament is, of course, frequently employed in Insular ornament, often in the arms of crosses or to fill triangular spaces. In such contexts it is difficult to know how far to invest the motif with symbolic meaning. Despite this it has often been claimed that the triquetra is a Trinitarian symbol (e.g. Blindheim 1985, 52). Sadly, there are no accompanying inscriptions in early medieval art which would help confirm this identity, such as occur with the triangles and interlinked circles of a later medieval period (Magerřy 1975; see also Elbern 1955). It is therefore helpful to draw on evidence which is well distanced from the Insular art world: the sculpture of eighth- and ninth-century Italy. Here triquetra frequently occur on ciboria and on closure- and altar-slabs where they are associated with crosses, grape clusters and other symbols of the Church and the Eucharist (Rotili 1966, pl. XIII; Serra 1974, pls. CXXI, CXXV, CCIV, CCXLII; Cecchelli 1976, pl. LXXIII; Bertelli 2002, pl. CI; Ermini and Peroni 2003, pl. XXXVII; Bailey 2007, 460–1). In such contexts it is difficult to resist the conclusion that they are charged with Christian symbolism. Similarly in Britain we find triquetra, at a variety of dates, placed in highlighted positions on sculptures: alongside a cross on the slab from Killaghtee, Co. Donegal; above clerics at St Vigean's, Angus; on top of a disc-headed cross from Iona; above the cross at St Madoes, Perthshire; surrounded by a series of symbols at Meigle, Perthshire; at the centre of a slab from Kilduncan, Fife; on the top of the frithstol at Hexham, Northumberland; accompanying an angel at Slaidburn, Yorkshire; and on the end of a hogback from Plumbland in Cumbria (Henry, F. 1965, pl. IV; Henderson and Henderson 2004, ill. 220; Allen and Anderson 1903, iii, figs. 310, 342; Fisher 2001, fig. 17; Trench-Jellicoe 2005, ill. 3; Cramp 1984, pl. 186.1028; Coatsworth 2008, ill. 696; Bailey and Cramp 1988, ill. 535). Given such emphasis, they must here also be invested with symbolic significance. Admittedly, St Augustine may only have lent his written authority to the use of the triangle as an appropriate symbol for the Trinity, but his continual recourse to the value of the numeral three in his struggles to explain the essential unity of the threefold godhead, does offer further support to the argument that the triquetra provides a very appropriate symbolic expression of the Trinity (Migne 1865a, col. 455).

In summary, this panel is a visual statement of the power and authority of the priest and the Church, based upon the Trinity and the cross, using a symbolism which would be readily understood in late Anglo-Saxon England even though its representation owes much to Irish antecedents.

The scene on face B provides major interpretative challenges (Ills. 709, 714). In earlier publications I identified it as the death of Isaiah, sawn in half at the hands of Manesse (Bailey 1980, 159–61; id. 1990, 7; id. 1996a, 80–1; Bernheimer 1952). Though the story of Isaiah's murder was known to Bede and figured in texts known to have circulated in Anglo-Saxon England, this identification must now be rejected for it was based upon an inaccurate drawing by Romilly Allen which appeared to show the central figure being cut in half using a bow saw (Bailey 2007, 461–6, fig. 36). The 'blade' of the saw does not exist — as is evident from the rubbings made by Allen which are now preserved in BL Add. MS 37551, item 85. The flanking figures are actually depicted as either binding the central victim or poking at him with narrow instruments; the protrusions which were misinterpreted as parts of a saw blade may represent deformities on the flanking figures or, less likely, be (marginally misplaced) genitalia. In summary, Anglo-Saxon England certainly knew the story of Isaiah's martyrdom but this is not a representation of it.

What then is portrayed here? A long-standing local tradition recognised in this panel the dismemberment of the Northumbrian King Oswald after the battle of Maserfelth in ad 642 — an act which was to scatter a multiplying set of his bodily relics across Britain and Europe (Browne 1887a, 16–17). This interpretation depended in part upon the identification of the local place-name Makerfield with Bede's Maserfelth — an identity which is philologically impossible and which has recently been effectively discredited (Stancliffe 1995). It also, however, depended upon the church's dedication and the proximity of a St Oswald's well (Taylor, H. 1906, 210–11). We cannot be certain about the antiquity of the well, though the saint is regularly associated with water sources. Nor can we be sure about the period in which the church received its present name, though it was clearly dedicated to St Oswald by the date of the Doomsday Book compilation. Despite these uncertainties a depiction of Oswald's dismemberment would still remain a possibility. But that explanation can be firmly rejected on two grounds: first, there is no saw, and thus no dismemberment; secondly, the nakedness of all the participants would pose major problems for any identification with an Oswald depiction.

With Isaiah and Oswald thus removed from the scene, we can perhaps take a more positive approach to interpretation, and narrow down the possibilities, by drawing on the evidence of analogous figural compositions on other Insular sculptures. Significantly, in the light of what has gone before, these come from Ireland and Scotland. One is from St Vigean's, Angus, whilst the other six are from Ireland: Arboe, Co. Tyrone; Castledermot, Co. Kildare; Kells, Co. Meath (two scenes); Monasterboice, Co. Louth; and Ullard, Co. Kilkenny (Henderson and Henderson 2004, ill. 221; Harbison 1992, ii, figs. 40, 644, iii, figs. 749, 751, 955, 956). All show figures flanking or alongside a naked and reversed third person, often grasping his legs; they have recently been conveniently listed and illustrated by Peter Harbison whose commentaries and tables summarise the interpretations offered for them (Harbison 1992, i, 222–3, 246–7, 306–7, tables 1–4). Among these scenes, Arboe offers a very close parallel to Winwick in the manner in which the naked victim's arms curve downwards around his head, whilst Ullard provides an analogy for the seemingly distorted — almost animal-like — faces of the flanking figures (Harbison 1992, ii, figs. 40, 644). In Ireland and Scotland the scenes are interpreted, with varying degrees of confidence, as the Judgement of Solomon, the Massacre of the Innocents, the Fall of Simon Magus, and Saints Paul and Anthony overcoming the Devil. Yet, whilst offering comparisons for the general composition, these Irish and Scottish scenes differ in two vital respects from the Winwick panel: the flanking figures are clothed and nowhere do we find the central figure apparently suspended by one of his feet (Bailey 2007, 468–9).

The combination of a suspended naked victim, flanked by repulsive and equally naked figures who appear to be binding or prodding the man between them, does however fit with both depictions and literary descriptions of a soul in hell which survive to us from the late Anglo-Saxon period. This, I suggest, is the meaning of the Winwick scene; its iconography draws upon a complex of inter-related 'attack and suffering' scenes, which are exemplified on early medieval sculpture from Ireland and Scotland, but here invests the type with a new meaning.

The early medieval conception of hell depended upon a series of sources. Among these, the most significant within the Insular world was the Visio Sancti Pauli, a text which existed in a variety of forms and whose shorter redactions show significant connections with England and Ireland in their manuscript tradition (Silverstein 1935; Healey 1978, 41–57; Biggs et al. 1990, 66–7; Wright 1993; see also Henderson, I. 1997, 48–50). In these texts, and in the Insular Latin and vernacular literature dependent on them — as well as in Anglo-Saxon art — all of Winwick's elements are present. The vulnerable nakedness of the soul is stressed in the Visio Pauli and Christ and Satan (Silverstein 1935, 28; Finnegan 1977, i, 134); it is depicted on the early ninth-century Rothbury cross and is frequently shown in late Saxon manuscripts and ivories (Beckwith 1972, pl. 16; Cramp 1984, pl. 215.1224; Jordan 1986, figs. 17, 21; Noel 1995, pl. 57; Hawkes 1996, 88–90; Horst et al. 1996, fig. 30a). Commonplace also in art is the fact that the devils are both wingless and naked, and several commentators have noted that there is a strong English tradition of portraying them as hideously threatening, with distorted features and with genitalia on display (Tselos 1959, 139; Temple 1976, ills. 200, 265; Jordan 1986, 303–6; Harbison 1992, iii, fig. 940; Noel 1995, pls. 44, 57; Horst et al. 1996, 126). The souls are both bound and tortured in Christ III and are depicted as such in manuscripts — the devils using a variety of long thin objects (spears, flails, forks and hooked instruments) which may be what is being held by the flanking figures at Winwick (Cook 1900, ll. 1620–23; Jordan 1986, pls. 12, 14; Dodwell 1993, pls. 50, 342).

The reversed suspension of the souls is a theme which is present both in redactions of the Visio Pauli and in other apocryphal Latin texts (Silverstein 1935, 31, 42; Himmelfarb 1983, 84–92). One version of it, in which the souls are suspended from trees, was much exploited in Anglo-Saxon and Irish vernacular literature, and is a theme which has been fully explored by Charles Wright (Silverstein 1935, 69–72; Wright 1993, 113–21). But a treeless and reversed suspension is also found, among the catalogues of suffering, in other pre-Conquest texts dependent on the Visio Pauli. It occurs, for example, in the Carolingian-period Latin homiliary preserved in Cambridge Pembroke College MS 25, an eleventh-century manuscript from Bury St Edmunds, and in an anonymous Old English homily dependent on a Pembroke-type antecedent (Cross 1987, 192–3; Bailey 2007, 471). If, moreover, one or both of the hands of the victim are represented at Winwick as being amputated then this is yet another of the judicial punishments awaiting the damned in hell (Silverstein 1935, 28; Semple 2003, 237–8).

I know of no other depiction of suspended souls in Insular pre-Norman art, though the theme is one which does appear in two early twelfth-century psalters (Morgan 1982, nos. 23, 51). In those later works, however, there are numerous souls who are suffering in this way. Here at Winwick the stress is on the drama of individual damnation, just as it is on the c. 1070 Gunhild ivory cross where two devils flank another naked, but upright, soul (Beckwith 1974, 44–7).

If all this can be accepted, then we must consider the relationship between the two figural panels, whilst recognising that we now only have a very small part of the total original carving. I suggest that the link is relatively straightforward: it depicts the horrors of hell which await those who reject the saving power of the priesthood and Church, founded on the Trinity. Archbishop Wulfstan's Institutes of Polity make precisely this point (Jost 1959, 104, 107–8).

The Winwick cross has always been an impressive piece of sculpture. It can now be seen in its tenth-century context, strongly influenced in its shape, abstract ornament and iconographical organisation by the Celtic west but expressing a Christian message which was well understood in Anglo-Saxon England.

Tenth century

(—) 1850, 115; Robson 1851, 206; (—) 1854a, 343; Ellacombe 1872, 526, fig.; Beamont 1875, 8, 16, 61; Rylands 1880, 62; (—) 1881, 219; Allen 1881, 91–3, fig. facing 92; Allen and Browne 1885, 355; (—) 1886a, 311; Allen 1886, 328; Allen 1887, 328, fig. 126; Browne 1887a, 14–18, pl. V, figs. 1–4, 10–11; Glynne 1893, 93–4; Allen 1894, 4, 8, 9, 15, 18, 21, 23, 26, pl. III; Allen 1895, 135, 141, 152, 154, 157–9, figs. facing 157; Cox, E. 1895, 238; Harrison 1896, 4; Taylor, H. 1898, 43; Howarth 1899, 9; Taylor, H. 1901, 219–25, pls. facing 212, fig. facing 224; (—) 1905a, 213; Allen 1906, 134–5, pl. on 134; Garstang 1906, 262–3, fig. on 263; Taylor, H. 1906, 5, 204–10, pls. facing 204, 205, 209, 476; (—) 1909, 169; Ditchfield 1909, 119–20; Fisher and Ditchfield 1909, 5; Farrer and Brownbill 1911a, 124; Wickham 1915, 172, 175; Browne 1916, 51; (—) 1920, 175–6; Collingwood 1926a, 329; Collingwood 1927a, 140, 163, fig. 155; Wallis 1932, 40; Wainwright 1945–6, 103; Tupling 1948, 8; Pevsner 1969a, 15, 432; Grealey 1976, 18; Edwards, B. 1978a, 75–6; Bailey 1980, 25, 26, 159–61, 231–2, 236, 242, 264, fig. 29, pl. 56; Plunkett 1984, I, 160, II, 309, 379, fig. 27; Fellows-Jensen 1985, 407; Freke and Thacker 1987–8, 34; Bailey 1990, 7, fig. 3; Philpott, F. 1990, 66; Kelly, D. 1991, 124; Kenyon 1991, 102; Edwards, B. 1992, 58; Cambridge 1995, 157, 162; Bailey 1996a, 80–1, fig. 39; Bailey 1996b, 29, fig. 2; Hawkes 1997b, 149; Edwards, B. 1998, 87; Bailey 2003, 222, fig. 10a; Higham, N. 2004a, 45; Blair 2005, 310, 342; Salter 2005, 4; Hadley 2006, 215; Newman, R. M. 2006, 104; Pollard and Pevsner 2006, 18, 680–1, fig.; Bailey 2007, 449–72, figs. 31–7

[1] For the documented history of the church and its pre-Norman status see: Farrer and Brownbill 1906, 262–3; Freke and Thacker 1987–8, 34–6; Cambridge 1995, 157; Blair 2005, 342. For additional discussion, rubbings and drawings, see the Romilly Allen collection, BL Add. MS 37550, items 738–58; BL Add. MS 37551, items 80–5; and Manchester Public Library, Hibbert Ware S. MSS: Msf 091 H21, vol. IX, 47, dated June 1847.