Gloucester Museum Accession A2656; Bryant 1999, no. 33. On display in Museum gallery.

Found in St Oswald's precinct wall facing Pateshall Alley in 1890.

St Oswald

Quite good

The surviving edge moulding on all four faces is square. Marking-out lines survive on face D where a strand of interlace falls over the left end of the moulding bar that divides the panels, and also where the moulding bar touches the edge moulding on the right side.

A (broad): A great Mercian beast, presumably in this case a lion, stands enmeshed in knots of interlace. The beast turns to bite one of the fleshy fruits or leaves that grow from the ends of the strands of interlace. The body is outlined and hatched, with curls of fur falling down the neck and across the back. The beast's eye is set beneath a heavy brow-ridge which terminates in a tight curl. The tail swings down across the back legs. The paws and genitals are well defined, as are the teeth, especially the two canines. It has been suggested that the beast's ear is drawn out into a strand of interlace, but it is also possible that the interlace simply ends in a hollow, swelled terminal. There is a very similar terminal to the left of the beast's rump. Beneath the beast's front feet is a second enmeshed beast, with a very similar head, hatching on its upper forelegs and a spiral shoulder-joint. This second beast also bites one of the fleshy fruits or leaves.

B (narrow): Divided into two panels. (i) The upper panel contains a bird caught in a chain of interlace. The head is missing. The feathers are carved with hollow centres. The curving line of the shoulder joint continues into the fore-edge of the wing, which is drawn out into the interlace. (ii) The lower panel contains a pair of salamander-like creatures with outlined and hatched bodies and spiral hip-joints. These creatures are also enmeshed in interlace and are busy biting each other's tails or feet. The creatures are seen from above, and each has its eyes set, side-by-side, in the top of its head.

C (broad): This face is covered with a bifurcated tree-scroll, the branches of which become a tangle of interlace. 'Eye-shaped' fruits or seeds grow from the point at which the main stem splits, and from a secondary bifurcation near the top of the surviving portion of the design.

D (narrow): Like face B, this face is divided into two panels. (i) The upper panel contains a spiral-hipped, bipedal creature, with a long tail that splits in two. One strand of the tail terminates in a triangular tip, while the other is drawn out into the interlace that enmeshes the creature. The head is missing. The body outline is filled with a simple meander. (ii) The lower panel contains a pair of facing beasts, the bodies of which are treated in a naturalistic manner, well rounded and muscled. The beasts have forelegs only and their bodies are drawn out into interlace. A separate strand of interlace is tied around the creatures' necks and ends in two fleshy fruit or leaves that the creatures are biting. The beasts have small round ears, naturalistic eyes and wrinkled snouts.

Gloucester St Oswald 3 is the best known of the Gloucester crosses. Three faces of this shaft are covered with creatures that are generally related to those on a group of west Mercian sculptures from Cropthorne (Worcestershire), Acton Beauchamp (Herefordshire) and Wroxeter (Shropshire). Elsewhere in this volume it is suggested that all four carvings are products of a single centre or group of carvers, probably based in Worcester or Gloucester, and they have been described as the 'Cropthorne' group (see Chapters III and VI, pp. 25, 67). At Cropthorne the birds and beasts are differently treated; the birds are shown in simple outline with wing and tail feathers, whilst the beasts are heavily cross-hatched (Ills. 621, 625, 629–33). Even so, there is nothing like the variety shown on the Gloucester cross. In this respect the Gloucester shaft is more similar to that at Acton Beauchamp where every creature is different (Ills. 496, 498–501). On face A of the Gloucester shaft the 'lion' has a hatched body but also a mane of fur and curly swags of fur across his back (Ills. 279, 283). The creature has strongly accentuated facial features with a heavy brow ridge ending in a tight curl. On face B the bird is entirely covered with hollow centred feathers (Ills. 280, 284), similar to the bird on a cross fragment from Escomb (Co. Durham) (Cramp 1984, 77, pl. 53.253), while the sprawling 'salamanders' are hatched and have spiral hip joints (Ill. 286). On face D there is a creature with a spiral hip joint (Ills. 278, 285), but the hatching of the body is treated as a meander, while the pair of beasts below are entirely naturalistic in treatment, if this can be said of these wonderfully unnatural creatures (Ill. 282).

Gloucester St Oswald 3 is a masterpiece of eclecticism. Kendrick (1938, 187) called it 'mature "Baroque"' and placed it in the mid ninth century, and Cramp (1978, 9–10) suggested that it belonged at the end of her Phase V or in Phase VI (the last quarter of the ninth century). However, late eighth- to early ninth-century manuscripts offer parallels for most of the Gloucester animals. The pair of creatures, and a creature in the exact pose of the spiral-hipped creature, on face D (Ill. 278) can be found on folio 4a of the Canterbury Gospels. This was dated by Kendrick to the first third of the ninth century (Kendrick 1938, 162–3). The feathered bird on face B with the hooked 'shoulder' joint (Ill. 284) can be seen in the symbol for St John the Evangelist on folio 27v of the Book of Kells, as can a maned lion with zoned body-colouring, studs on the hip, curling-tipped fur and an outlined jaw (Henry 1974, 20). St Mark's winged-lion symbol on fol. 53v in the early ninth-century Book of Armagh has curling-tipped swags of fur down the neck and across the back (Alexander 1978, 76–7, cat. 53, ill. 229). Creatures with split tails ending in leaves, like the upper creature of face D (Ill. 285), occur in the Book of Cerne (Henry 1974, 165), while the shape of the salamander's heads on face B (Ill. 286) is a very common form (to be found for example throughout the Book of Kells). Tweddle (1983, 19–20) pointed out that sprawling creatures, similar to the salamanders, can be found in the arcades framing the Canon tables in the St Petersburg Gospels (Alexander 1978, 64, cat. 39, ills. 188–91). The creatures can also be found on contemporary metalwork — for example on the disc brooches from the Pentney hoard (Webster and Backhouse 1991, 229–30), and bone and ivory objects such as the Gandersheim casket (Webster and Backhouse 1991, 177). They are, in fact, all items that one might expect in the repertoire of a carver working in the early years of the ninth century, and arguments adduced for the body hatching at Cropthorne and Acton Beauchamp should be equally relevant for this Gloucester cross (Cramp 1977, 225–30; Kendrick 1938, 187–8). Recently Rosemary Cramp has provided a valuable summary of the debate concerning the development of these creatures that has exercised art-historians since the end of the nineteenth century. In this summary she supports a late eighth- to early ninth-century date for the Colerne creatures and the beginning of the tradition on the related group of carvings from Wessex (Cramp 2006, 42–7). In the Gloucester context a date between ad 800–825 for no. 3 still seems most likely.

It is interesting to note that the parallels for Gloucester St Oswald 3 at Cropthorne and Acton Beauchamp, are both, like Gloucester, places which feature in the relatively sparse documentary record from the eighth and ninth centuries (King Berthwulf of Mercia issued a charter at Cropthorne in 840, while Acton Beauchamp was a minster mentioned in the early eighth century: see the relevant catalogue entries). This tends to confirm that these pieces were carved under high-status patronage. Wroxeter, the other site, lay in an area from which virtually no documentation survives for this period.

First quarter ninth century

Medland 1888, 196; Medland 1888–9; Browne 1903, 161, fig. 8; Smith 1913–14a, 71; Brøndsted 1924, 59, 142, 217–19, 229–30, figs. 116, 163a–b; Collingwood 1927, 128; Clapham 1930, 72; Dobson 1933, 267–8; Kendrick 1938, 187–8, 190, 208, pl. LXXXII; Cramp 1975, 189; Rice 1952a, 130; Cramp 1977, 225–30, figs. 62f, 62i, 63g; Cramp 1978, 9, 10, 13, 14, figs. 1.1(t–u), 1.2(l); Heighway 1978, 107; Heighway 1980, 212; Plunkett 1984, I, 69, 87, 89, 93, 181, 182, 185–6, 188, 189, 190, 191, 193, 198, 210, 212–13, 247–8, II, 298, 360, pl. 63 (Gloucester I); Heighway 1987, 116, fig.; Tweddle 1991a, 239, fig. 25; Tweddle 1992, 1144, 1155, 1157, 1161, 1162, figs. 574c, 574f, 578c, 581d; Tweddle et al. 1995, 38, fig. 11e; Bailey 1996a, 18, 20, figs. 8e, 9a; Bryant 1999, 154–5, 159–62, no. 33, fig. 4.10