Shields and parade armour

One of the finest examples of the repoussé technique combined with engraved ornament is represented by the bronze scabbard mount from the River Witham (Figure 7.2). This is a pre-eminent example of the incised ornament integrating closely with and reinforcing the repoussé component. The broadly diagonal layout of the ornamented panel has been frequented remarked with reference to Hungarian Sword Style analogies, but the Witham example is singular. Its design, unfolding from a sub-circular element slightly offset below the hilt-guard, and tapering progressively in a double loop, has evoked images of a stylized bird in flight, partly prompted by broad similarities to the more explicit engraved designs on the Wandsworth shield roundel, though this kind of perspective interpretation is alien to the mainstream of La Tène art. The component motifs nevertheless are familiar enough, including hair-spring spirals, peltae-within-peltae, split palmette, together with a more distinctive 'bat-wing' motif with vessica-shaped ends. Linear hatching, not remotely basketry, echoes the simple form from Newnham Croft. The form of the scabbard's fragmentary chape, which survives only in a nineteenth-century drawing, does not suggest an early type, though like Standlake it could have been a composite piece. The ornament is quite worn, suggesting that it was in circulation for several generations, so it is quite possible that the sword and scabbard's original date of manufacture was as early as the beginning of the third century bc.

Longevity of use is also attested by the Witham shield, one of the archetypes with the Wandsworth shield-boss of the developed insular style. The Witham shield (Figure 7.3) traditionally stands at the head of the insular series, modelled on Gaulish

Scabbard Boss
Figure 7.2 The Witham sword and scabbard mount. Drawing by D. W. Harding from original in Alnwick Castle, Northumberland, by courtesy of the Duke of Northumberland.

Figure 7.3 The Witham shield. Adapted from Brailsford (1975a) and Jope (2000).

and ultimately Italic types, and sometimes reflecting in consequence southern ornamental styles. Its beaten bronze facing, doubtless originally mounted on a wooden backing, covered a sub-rectangular shield with rounded corners and nearly parallel sides. It is certainly a composite piece that was substantially modified at least once in its life. Behind its surviving central boss, spine and terminal roundels can be seen the ghostly outline, preserved by differential patination, of a slender-bodied and spindly-legged boar. This might be less than convincing, were it not for many small rivet-holes that follow the outline, and must have served for holding the emblem in place.

In its modified form the dominant element is the spine, consisting of central umbo to protect the hand-grip and two terminal roundels, the umbo being located slightly above the mid-point on the long axis of the shield. Despite the complexity of infilling and spandrels, the initial impression of the design of the central umbo is its symmetry. On either side are paired, circular voids enclosed by peltate elements; along the axis of the shield are comma-leaves, pointing down the spine. The central roundel has a setting of three coral studs, suggesting a date of manufacture not later than middle La Tène, after which coral went out of fashion north of the Alps. Crucially different from Torrs, however, this design is based upon rotational not fold-over symmetry.

In terms of their repoussé ornament the terminal roundels are a matching pair. They are bordered by a wavy relief ribbon, leading to a pair of comma-leaves with spiral terminals that seem to simulate a pair of staring eyes. The central cupped boss, missing from the upper shield roundel, contains a domed seven-petalled flower. Both terminal roundels are supported by a pair of exaggerated sub-equine beasts that, like their counterparts on the Battersea shield, may well have had apotropaic qualities. Their features include multiple-leaved palmettes, unlike the standard La Tène attenuated form, below their eyes, over their flaring nostrils and sprouting like ears or antlers from the sides of their heads. From the snout of these beasts depends a low-relief S-scroll, leading to a tongue pointing down the spine, which Jope (2000) regarded as crucial to the dating of the shield. Dot-ended lines, not unlike Cerrig-y-Drudion, infill lobes within the design, suggesting a date perhaps earlier in the third century than Jope would admit to.

On the flat surfaces the terminal roundels are engraved with figure-of-eight scrolls that abut each other, sometimes with spandrels, rather than flowing as a continuous tendril. The use of hair-spring spirals and split-palmettes to create the illusion of bird's heads or wings resonates with both insular and Continental engraved styles. A more recent find from Chiswell Green, Hertfordshire (Jope, 2000, Pl. 89d), includes similar engraved motifs on a small bronze knife, the handle of which terminates in a cast bird's head and beak. But the sense of formality of composition of the Witham engraving, belying its lack of symmetry in detail, contrasts with the erratic freestyle of the Torrs horns or Witham scabbard, and suggested to Jope (1971b; 1978) Hellenistic or southern models.

The surviving fragment from the Thames at Wandsworth of shield boss and spine (Figure 6.9) must originally have been a sub-rectangular shield of the Witham type, but significantly smaller, though sharing the same asymmetric proportions above and below the spine, as Jope's reconstruction effectively showed (Jope, 1976, Fig. 1). In its ornamental affinities, as we have seen, it draws not on southern sources but on the Central and Eastern European Plastic Style, both in the modelling of the birds' heads of its central roundel and in the features of the face-mask that supported the missing terminal roundels. The design of the central roundel is dominated by two repoussé birds' heads, again in rotational symmetry. They are linked and divided by a raised wavy line comparable to those enclosing the terminal roundels of the Witham shield, and like Witham, too, the central design of the Wandsworth piece has leaf-like fingers pointing down the spine. The appearance of sullen severity of the humanoid face is achieved by a combination of techniques. The spiral eyes, as reconstituted, simulate 'bags' under the eyes (a device that may be compared to the 'bags' under the owl's eyes of the Bra cauldron), while the straight angularity of the mouth conveys a mood of gloomy resentment. Similar spiral-boss eyes can be seen on the rein-ring 'from Paris', or on the terret from Mezek, Bulgaria, underlining the Plastic Style affinities of the insular product. Finally, we should remark the small areas of engraved infilling, notably of hair-spring spirals and dog-tooth edging, the former already seen at Torrs and Witham, but both notable features of Irish scabbard engraving (Figure 5.9). As for dating, it seems unnecessary to defer its construction much after any of those already assigned to the earlier third century.

The two other principal shields are more controversial in terms of dating. The circular shield from Wandsworth (Figure 7.4) was apparently found in 1849 during the same dredging operations in the Thames that recovered the face-mask shield, but obviously with no demonstrable association. Covering the central hand-grip is a hemispherical umbo, a feature that is normally assigned to the first century bc or later, while an outer circular flange is embellished in repoussé and engraved techniques. This may have formed the central component of a sub-rectangular shield comparable to those discussed earlier, or may simply have been the mount of a circular wooden or leather shield. The slightly flattened hemispherical umbo is ornamented in engraved technique, essentially with two independent tendril designs that stand in rotational relationship, though not completely symmetrical in detail. This central element is enclosed by a moulding on which wavy-line highlighting recalls earlier examples. The relief ornament of the surrounding plate also consists essentially of two independent scroll-like elements in rotational relationship, though the terminal element of each is detached from the main scroll. The focal icon of each is a bird's head, perhaps originally with coral studs for eyes. Birds also feature is the engraved ornament, more obviously as infilling of the repoussé plate but also in schematic form on the central dome. The engraved ornament makes particular use of triangle-within-triangle, pelta-within-pelta, linear and dotted infilling, 'bat-wing' and star-rosette, all familiar from earlier contexts. Behind one repoussé bird's head is an engraved design that echoes the over-and-under figure-of-eight of Hungarian scabbard ornament. The image of the bird rising from water in perspective (Jope, 2000) seems to be in the eye of the beholder, and probably should not determine the debate on dating. An important difference between the Wandsworth shield roundel and other examples of the repoussé technique in Britain, however, is the sharp-edged character of the repoussé work, in contrast to the more rounded technique and higher relief masses, for example, of Torrs or Witham. But this would not disqualify it from classification within the European Plastic Style tradition (Jope, 1976, 183).

Conventionally regarded as the latest in the shield series (though the case for an earlier date has become more compelling in recent years) is the splendid example from the Thames near Battersea (Pl. 9). Like the fragmentary face-mask shield from

Battersea Shield
Figure 7.4 The Wandsworth circular shield, with detail of central boss ornament enlarged. Adapted from Brailsford (1975a) and Jope (2000).

Wandsworth, it is small for practical use, being just 77.5 cms in overall length. Its hand-grip cover, too, assuming its association with the shield to be genuine, is small, suggesting that it was designed for a woman or juvenile if not for a man of slight physique. At a distance an undersized shield might serve to enhance the relative stature of the holder, but in any event, the Battersea shield was more probably intended for prestigious display than for military utility. It conforms to the sub-rectangular form of the Witham class, but with a much more pronounced waisting of the sides, a feature that some commentators have attributed to Roman influence. It also follows the Witham model in having three roundels, a larger central one and two equal-sized but smaller terminal roundels, though in this instance they were made in three separate pieces. All three were then mounted on a back sheet comprising four separate pieces, and the whole, with its edge-binding, was presumably once again attached to a wooden or leather backing.

The ornament of the roundel-plates is achieved by fine repoussé work, supplemented by no less than nine 'enamel' inlay settings on each. Above and below the central boss of the larger roundel are designs that Stead (1985b) likened to bespectacled faces, the goggly eyes enhanced by 'enamel' settings, below which small, glum mouths recall the sullen expression of the Wandsworth face-mask. Balancing these on either side of the roundel are two matching elements that can be read as opposed comma-leaves depending from a flattened pelta, from the arms of which tendrils reach out to join the spectacle-masks in curved triangles. Stead has shown that the detailed motifs within these peltae and triangles violate the symmetry of the overall design, though the basic design is one of strict fold-over symmetry about the spinal axis of the shield. The design of the two terminal roundels is identical in conception, and only marginally different in detail of execution. It consists of linked fleshy S-scrolls with 'enamel'-inlaid settings at their intersections. Each roundel has a flanged extension, bearing a bulbous-nosed animal with spreading antlers that acts as a support for the central roundel. This device again recalls the Witham shield, though on the latter the heads support the terminal roundels, facing along the spine towards the centre.

The 'enamel' settings are plainly a dominant feature of the Battersea shield, and for Hawkes their pseudo-cloisonné division into key-patterned compartments was one indicator of a late Iron Age 'Belgic' date. The settings are not true cloisonné, which holds cut glass inclusions; nor is the opaque red glass true enamel, but glass heated to a spongy consistency and pressed into the recesses. Hawkes further argued a date in the last quarter of the first century bc for the shield on grounds of stylistic influences in its ornament from Augustan Roman silverwork, in which view he was supported by Jacobsthal and de Navarro among others (for references see Stead, 1985b, 25-6). Against the case for a late dating, Stead pointed to various aspects, technical as well as stylistic, of the Battersea shield, for which respectable antecedents as early as the fourth century could be cited. While the weight of evidence probably favours a less radical revision of its dating, chemical analysis of the 'sealing-wax red enamel' has now indicated that it cannot be later than the second century bc (Jope, 2000, 351). We should be clear that this is strictly the date of the enamel, which in principle could have been incorporated from the craftsman's recycled stock into the assembly of a later shield. This, however, would doubtless be regarded as special pleading, so the later second century may stand as the current best assessment of the date of the Battersea shield.

0 0

Post a comment