VESSEL

Description: A lead vat or tank of early-medieval date, now in several fragments. Most of the base and a little under half of the wall survive together in a single piece, and there are numerous fragments of varying size; in total it is about two-thirds complete. The tank measures in total approximately 40cm in diameter and 19cm tall. The lead is c. 3-4mm thick and the object now weighs approximately 15kg. The tank can be dated to the ninth century on the basis of its decoration.

The base is oval and measures c. 404mm by 365mm. It is incomplete and slightly crumpled, and is 4.14mm thick. The wall of the vessel is composed of one large sheet of lead, 3.41mm thick and c.190mm tall, curved around with a single vertical seam. This seam was then covered with a thick undecorated applied triangular panel which runs right down to the base. A stub of square-section iron rod is sandwiched between the seam and the panel, and represents the remains of one iron handle. A second similar iron handle survives. The base and the wall appear to have been joined by setting the wall and the base into grooves on a separate lead ring, or by setting the wall into a groove channelled into the edge of the base. All joints were presumably soldered. The rim appears to have been made simply by turning the top of the wall over into a U shape.

The exterior of the vessel is decorated with eight relief-decorated triangular panels which alternate in orientation, half with the apex at the top and half with the apex at the base. The panels appear to be integrally cast as part of the vessel wall and are not separate components. The panels measure 150.04mm long by 75.22mm wide. Panels 1-3 remain on the vessel, whereas Panels 4-8 are broken off and now survive in many separate fragments; their order and orientation is unclear. All of the panels appear to have been set evenly around the upper wall of the tank.

The panels are in excellent condition with little wear, and are of two designs which alternate around the vessel. Panels 1, 3, 5 and 7 are identical, as are Panels 2, 4, 6 and 8. Panel 1 is set to the right of the undecorated handle panel, and there was presumably a second handle panel between Panels 4 and 5, making 10 triangular panels in all. The handle panels, unlike the decorated panels, appear to have been soldered on.

Panel 1 has the apex downwards and is bordered by a relief cable. The upper portion shows a pair of profile addorsed animals, apparently bird/animal hybrids; the heads, with open jaws and oval ears, are turned back to confront each other. Below is a profile quadruped, apparently hoofed and with grooving on the head and body which might suggest harness. This animal, possible a horse, has an open mouth and lolling tongue or leaf protruding from its mouth. Another cable then divides the panel and signifies a change of decorative style; the lower portion of the panel is decorated with an interlace design composed of two serpents with inward-facing heads towards the top. The apex, outside the cabled border, ends in an animal-head terminal with large oval eyes and nostrils, similar to those on 9th-century strap-ends.

Panel 2 has the apex uppermost. The border is formed by a ridge decorated with moulded chevrons facing in alternate directions which again outlines a trapezoidal field. The terminal at the top is again a stylised animal head with large oval eyes, but the snout is not visible beneath the crushed rim. The upper portion of the panel has a full-length human figure with both arms bent at the elbow, one upwards (with hand perhaps to the right ear) and the other downwards to touch the weapon, which lies obliquely across the figure's waist, the hilt below the figure's right elbow. The weapon is probably a sword, and has a downward-curving lower guard and straight upper guard. The head is large and apparently without any helmet or headgear. The figure is wearing a short triangular skirt with ribbed decoration perhaps indicating kilt-like folds. Below the figure there is a four-legged animal with a long tail, slender snout and long curved ear, perhaps a dog, which appears to have an animal or bird dangling from its mouth.

Panel 3 is apex-downwards and has decoration apparently mould-identical to that on Panel 1. Although the panel is fractured, the decoration is in general clearer; more details of the possible horse are visible, and the bodies of the birds are clearer, each with legs, tail and a wing with upturned tip.

Panel 4 is on a broken fragment of the vessel, but its apex terminal survives on the complete vessel so allows precise reconstruction. The panel depicts the human figure and is apparently mould-identical to the decoration on Panel 2. This panel is in better condition than the photograph suggests, and offers extra information; the figure here looks as if he is moustached and helmeted.

The positions on the tank of Panels 5, 6, 7 and 8 cannot at present be reconstructed exactly and will require expert conservation work. Panels 5 and 6 both have large parts of the vessel wall still attached and have some recent crumples which have cracked the metal.

Panel 7 and 8 are detached from the vessel, and Panel 8 is now in three pieces. There is some recent damage to both.

Other Anglo-Saxon lead tanks: There are several lead vats or tanks known from the middle or late Anglo-Saxon period. Jane Cowgill published a list of twelve examples similar to the Cottingham tank in her study of the Flixborough tanks (in Evans and Loveluck 2009, 267-77, esp. 269):

- two nested lead tanks containing a hoard of 14 iron tools and a bell from Flixborough, Lincolnshire

- a lead tank containing a hoard of iron tools from Westerley Waterless, Cambridgeshire, now in the Cambridge Museum of Archaeology and Anthropology (accession number Z 16465; Hughes 1880; Fox 1923, 300 and pl. XXXV)

- a lead tank containing an iron cauldron with a hoard of iron tools inside from St Saviourgate, York (Ottaway in Evans and Loveluck 2009, 259)

- two nested tanks from Willingham, Cambridgeshire (Bartlett 1984)

- a folded vessel at Riby Crossroads, Lincolnshire (Cowgill in Steedman 1994, 267-71)

- fragments of three vessels buried as scrap at Bottesford, Lincs

a single fragment metal-detected from Roxby, Lincs

a tank published as Roman, from Kenilworth (Guy 1991)

Of these, four are decorated; the larger tanks from Flixborough and Willingham, and the single vessels from Westley Waterless and Kenilworth. Both Cambridgeshire examples have relief ornament withiin pendent triangles, and the Willingham tank alternates these with standing triangles in similar manner to the Cottingham lead tank. All the motifs appear to have been made by impressing the sand mould with a model before using it to cast the sheets of lead. The Cottingham tank, however, is more elaborately decorated and better finished than any other Anglo-Saxon lead tank known (Cowgill in Evans and Loveluck 2009, 269 and 276 n. 1).

There are also now three further examples of similar lead tanks recorded on the PAS database; LANCUM-DA3E25 (Grassington, North Yorkshire), LANCUM-227D54 (Cumwhitton, Cumbria; now in the Tullie House Museum) and WMID-7FF438 (Parwich, Derbyshire). They are all relatively small and undecorated; the Derbyshire example is particularly well preserved.

The date-range of these tanks is still fairly imprecise, running in general from the 8th to 10th centuries. The Flixborough tanks were not found in the main excavation, but during later quarrying, and the find is dated by the tools to the 8th to 10th centuries (Ottaway in Evans and Loveluck 2009, 258). The tank from Riby was found in a ditch which also contained Middle Saxon pottery (Cowgill in Steedman 1994, 267). The Westley Waterless tank can perhaps be dated to the 8th to 10th century by its interlace ornament. The York, Willingham and Bottesford tanks remain unpublished and the Roxby and Kenilworth tanks were stray finds with no contextual dating evidence.

Other forms of lead tank of Anglo-Saxon date are known; tall cylindrical examples from Maidstone, Felixstowe and Folkestone; smaller, later examples from Garton-on-the-Wolds (three) and Stidriggs, Dumfries-shire (Cowgill in Evans and Loveluck 2009, 269); and a unique decorated rectangular example from Willingdon, Sussex, now in Lewes Museum, with pendent triangles filled with 11th-century interlace and vertical lines of mock cabling (Blair 2010, 173, figs. 13c, 14 and 16c-d).

Tanks dated to the Roman period tend to be larger, and generally have vertical lines of mock cabling interspersed with circular motifs which often have Christian significance (Watts 1988, 210). The Kenilworth tank is decorated with a zig-zag line of mock cabling forming pendent triangles and it is difficult to know whether it should be assigned to the Roman or the Anglo-Saxon group. The findspots of tanks thought to be Roman are concentrated in the Cambridge region (Guy 1991, 108).

Function: There have been many suggestions of the possible purpose of these vats or tanks. Although some have been deposited with iron tools inside, it is generally felt that they would have made impractical toolboxes (Leahy 2003, 165-6). Study of the larger, usually better-made 4th-century Romano-British tanks, generally decorated with applied vertical and horizontal strips as well as circles, crosses and other Christian motifs, has led to the suggestion that they were used in Christian rituals, perhaps for foot-washing or as baptismal fonts (Guy 1981; 1989; 1991; Watts 1988). Cowgill argues against this function for the Anglo-Saxon tanks, however, emphasising the small size and poor finish of the tanks; she suggests an alternative use as standard measures (in Evans and Loveluck 2009, 269 and 271-3). Blair puts forward a case for lead tanks being involved in solemn rituals involving water, including baptism (2010, 160-1).

Although the art on the Anglo-Saxon tanks tends not to have obvious Christian affinities, some at least of the Anglo-Saxon tanks might have had a liturgical function. The larger Flixborough tank carried what appears to be a stylised chi-rho or six-pointed star, and the bell contained therein was decorated with a Greek cross over an X; simple copying from the late Roman examples seems unlikely, and a liturgical function is possible for both objects. Ottaway has also suggested that the 12 carpenters' tools within the hoard may symbolise the carpenter Christ or the twelve apostles, and that the context for the deposition may be the end of construction of the church on the site (in Evans and Loveluck 2009, 261).

The main argument for a function as baptismal fonts comes from the apparent use of models in other materials for the few pre-Conquest stone fonts that survive. Relationships can be seen between wooden vessels and stone fonts, and also these lead tanks and stone fonts; coupled with their tendency to be found apparently deliberately buried, often in watery environments, it seems likely that there was a ritual use for at least some of these tanks, particularly a highly decorated example such as the Cottingham tank.

Parallels to the decoration: The two designs appear to have been cast from moulds made from just two models (rather than using individual models for each separate panel). The animal-head terminals on both types of panel are very similar to those found on strap-ends of Thomas's Class 1, broadly datable to the ninth century (Thomas 2003). Otherwise parallels are mainly to be found within stone sculpture and, as with sculpture, what appears coarse or clumsy at first sight may reflect the scale and technique used.

Pendent triangle ornament is found on other tanks (e.g. Westley Waterless, Willingdon) but the alternating orientation of the panels on the Cottingham tank, and the way the designs are tailored to this, is unusual. Pendent triangles and, occasionally, terminal animal heads can also be found on stone sculpture (Bailey 1980, 187-8).

Panels 1, 3, 5 and 7

The subdivisions seen in these panels is very familiar from ninth-century metalwork decorated in the Trewhiddle style. A very good parallel comes from PAS database no. LIN-BF5F52, a strap-end of Thomas's Class 1, Type A, Trewhiddle-style. Both the subject matter and the bold scale of the ornament are however more reminiscent of stone sculpture than of anything within the metalwork repertoire.

Leslie Webster has commented that all of the animals (the interlaced serpents, the addorsed bird/animal hybrids and the quadrupeds) appear to be versions of Mercian-style animals, with the quadrupeds having distinctive body markings which can be paralleled in Mercian sculpture. Webster has also noted that from the mouth of the quadrupeds comes a very distinctive leaf, of a kind which appears on late 9th- to early 10th-century sculpture from the south-west of England (e.g East Stour in Dorset, and Colyton in Devon; Cramp 2006, 51 and fig. 22).

Panels 2, 4, 6 and 8

The human figure is reminiscent of those on Viking-period sculpture from north-eastern England. A list of these runs to over a dozen, most of which are found in distinct geographical clusters and are thought to be the work of relatively few sculptors. Although they are carved on Christian liturgical items (stone crosses) they are currently interpreted as secular portraits, reflecting the aristocratic military ideals of the tenth century (Bailey 2000, 16; but note arguments for an earlier date in Wilson and Klindt-Jensen 1966, 103 and 114). Perhaps the most celebrated examples of these warriors include those on a fragment of cross from Weston (Roesdahl et al. 1981, 61 and F16) and on the cross-shaft at Middleton (both North Yorkshire; see images below). The flared, pleated skirt or kilt worn by the Cottingham figure is particularly close to that on the Weston cross.

The attitude of the warrior on the Cottingham lead tank, with elbows bent at right angles, one hand up to the ear and the other down to the sword, does not appear to be paralleled on any other warrior figure. The hand to the ear is reminiscent of the figure representing Hearing on the late 9th-century Fuller Brooch (BM 1952,0404.1).

Barry Ager has commented that the sword, with down-curved guard, is a variant of Petersen's type M (Petersen 1919). This is not a Viking type, but more likely to be Anglo-Saxon, and dates to the mid to late 9th century.

The profile animal is also distinctive and further research within the corpus of stone sculpture may well produce further insights.

Conclusion: The Cottingham lead tank, although it has many parallels within tenth-century sculpture, also has motifs which suggest a ninth-century date may be more appropriate; these include the form and structure of the panels, their animal-head terminals, and the shape of the sword-hilt. A date somewhere within the late 9th century is probably therefore appropriate. The object also stands on the border between art conventionally thought of as 'Anglo-Saxon' and 'Viking', and further research and collaboration between sculpture and metalwork specialists is needed to draw out the significance of this.

At present, Panels 2, 4, 6 and 8 appear to contain the earliest version of the standing warrior motif, and what seems to be the only example of a warrior figure in a medium other than stone. It adds considerable weight to the argument that much of the art on stone sculpture was derived from pre-Viking, English models, but this time from a different medium and object type (Bailey 1980, 206). Another point of correspondence with the stone figures is that this motif may again have been placed on a liturgical object, in this case perhaps a baptismal tank.