The Contents of the Hoard

Kevin Leahy (Portable Antiquities Scheme)

“They let the ground keep that ancestral treasure, gold under gravel, gone to earth, as useless to men now as it ever was.” (Beowulf, l. 3168).

It is our challenge to make sure that this hoard is not useless.


The hoard boxed up at Birmingham MuseumAs the person fortunate enough to compile the first catalogue of the Staffordshire Hoard, I have been asked to provide a summary of its contents, with other speakers going on to look at various aspects of the find in more detail.  It must be said that, at this stage, some doubt surrounds even the most basic statistics about the find; the objects have not yet been cleaned which, in many cases, makes identification difficult.  However, the figures quoted below should provide a good résumé of what was found.

My first sight of the hoard, at Birmingham Museum on the 21st of July 2009, was awe-inspiring. I had been told that there were 240 finds, which turned out not to be the case; there were 240 bags and boxes.

The contents of a typical bag included a large number of broken and twisted objects of gold and silver, and the contents of each individual bag are important as they provide a degree of context for the objects.

Contents of a typical bag

The objects and fragments are almost all damaged. Although the reconstruction of the objects represents a huge challenge, the damage has advantages; it reveals details of construction that we cannot see on complete and perfect objects. The breakage appears to have happened prior to deposition; the scratches, scrapes and snaps characteristic of plough damage are rare, and it appears that the material has entered the ploughsoil very recently. The gold tends to be in better condition than the silver, some of which is brittle and in very small fragments.

Terry Herbert, the finder, and the team from Birmingham Archaeology recovered every humanly made artefact that they could, including modern finds, and the pattern of non-Anglo-Saxon finds from the site is very interesting. Even though we are within a couple of miles of the Roman town of Wall, Letocetum, there is no Roman material at all – not a scrap, not a sherd, not a coin, nothing.

Bagged soil recovered by Terry HerbertThere is also no high-medieval, no late medieval, no early post-medieval material. The land was enclosed in the mid 1830s, and all of the modern finds appear to post-date this, and include tin-glazed earthernware, some bits of nineteenth-century glass, cast iron and some nails; nothing else. Earlier maps, such as that of Morden (1695), show the field as heathland, a barren waste overlooking Watling Street.

In addition to the bags of finds that Terry Herbert recovered, he collected 21 soil bags – lumps of soil giving a response to the metal-detector and containing large quantities of tiny fragments of metal. Terry wisely kept this material as individual deposits. 

The recording and analysis of the Hoard

Gold and silver finds more than 300 years old need, under the Treasure Act 1996, to be reported to a Coroner. We needed to be able to analyse the assemblage speedily so that by the time of the inquest, when news of the find would become public, we would know what had actually been found. I was fortunate in that I cut my teeth on another large assemblage, the Cleatham Anglo-Saxon cremation cemetery, which with 1200 urns and 62 graves, is the third largest cremation cemetery in England. The methodology to analyse the Staffordshire Hoard could be readily adapted from the Cleatham study.

The database was very abbreviated, with simple tick-boxes for all of the different things that one might want to include. The dimensions and mass of each fragment were crucial in order to maintain control over the enormous amount of material.

Proportions of different metals in the Hoard

The initial list stated that there were more than 1,300 objects, more than 5kg of gold, and more than 1.3kg of silver. Later work, focussing on the smaller fragments from the earth lumps, has upped the count to more than 3,490 pieces, and the uncleaned weight has gone up a little, particularly for the silver. The database records 309.87g of copper alloy, but some of this is likely to turn out to be base silver and some of the gold objects are on copper-alloy cores. All the weights, absolute and relative, are likely to change substantially when cleaning and analysis are completed.

The easiest way to appreciate the range of objects in the hoard is to analyse them by mass. 46% of the pieces in the hoard weigh less than a gram. 

Graph of objects by mass

There is a significant number that weigh less than a hundredth of a gram, including 51 unmounted cut garnets. These are important from a technological point of view, because we can see tooling marks on the edges of the stones which would be invisible on a complete piece of jewellery.

StH 655 The largest object is the remains of the folded and incomplete gold cross (StH 655) which weighs, without its detached settings, 140.21g.  By comparison, the Crondall hoard, found in 1828, contained two jewelled ornaments, a chain and 101 tremisses, and weighed 131g. The gold objects in Mound 1 at Sutton Hoo weighed in total 1.66 kg.

Object types within the Hoard

Continuing to look at the proportions by mass, the bulk of the hoard (60.1%) consists of parts of weapons. Counting individual objects is absolutely pointless, because of the numbers of small detached items such as rivets from sword-hilts and foils from helmets, and because of the fact that many of the fragments may join.

Composition of the hoard

Swords and seaxes 

There are currently 92 pommels from swords (with two-edged blade) or seaxes (with single-edged blades). Unlike other counts, this total is likely to fall, because it includes fragments that probably join, but it is a useful total as there can only be one pommel per weapon.

The other hilt fittings total 354. There are plates from the guards at either end of the hilt, such as StH 562 (a flat plate which retains a rivet with a beaded-wire collar) and StH 557 (a plate with upturned edges). There are rings of beaded wire from each end of the grip, such as StH 470 (a larger one from the lower end) and StH 1426 (a smaller one from the upper end); the shadow of the wire ring can still be seen on StH 567, two plates which are still assembled as a lower guard.

StH 562  StH 557  StH 470  StH 1426

StH 563

Hilt-collars perform the same function as the wire rings around the grip, and can range from narrow (e.g. StH 122) to wide (such as StH 278). There are also a great many rivets, washers and caps, most of which appear to have come from sword-hilts; examples include StH 1281and StH 1536.

StH 122  StH 278  StH 1281 StH 1536

How these all fit together can be seen from the complete sword-hilt from Mound 1 at Sutton Hoo.

Sword hilt from Sutton Hoo Mound 1Exploded diagram of sword hilt from Sutton Hoo

A sword-hilt could be quite easily dismantled by first levering off the pommel, exposing the riveted end of the tang. When the rivet is cut off, most of the hilt fittings would then slide off the tang. It seems that the exception was the plates from the guards; the guards were made in the form of a sandwich of organic material between gold or silver plates. The plates tend to have been pretty badly bent and broken, presumably as they were dismantled to remove the horn, or wood, or ivory from the middle of the sandwich.

StH 672 StH 461 StH 551

Many of the pommels look worn, for example the filigree on StH 353 is rubbed almost smooth, as is the relief decoration on StH 711. The extent and the patterns of use-wear have the potential to be very informative.

StH 353 StH 711 

In addition to the certain sword-hilt components are four rings which probably came from sword pommels (StH 543, StH 531, and StH 1625, and perhaps StH 1204), and a stone bead which looks very much like a pommel ring (StH 764).

 StH 543 StH 531 StH 1625 StH 764


From scabbards, there are at least nine pyramidal mounts (e.g. StH 302, StH 451 and StH 462), and a pair of buttons (StH 675 and 1425).

StH 302 StH 451 StH 462


There are many pieces which appear to be from helmets. The most immediately recognisable is StH 453, a gilded silver cheekpiece with remarkable Style II decoration. Notice that there is only one, showing that (at least in this case) the hoard certainly contained an incomplete object.StH 453 StH 678

 Several features which can be seen on the Royal Armouries’ reconstruction of the Sutton Hoo helmet can also be found in material from the hoard. StH 678 has a curvature which would be appropriate to fit around the cap of a helmet, but we don’t appear to have any of the other heavy fittings (such as eyebrows and a nose-piece) which occur on Swedish parallels.


               Sutton Hoo helmet with Staffordshire Hoard parallels


There are several fragments of C-shaped edging strips (e.g. StH 11, StH 15, StH 23, StH 81, StH 617, StH 618, etc) totalling 109 cm in length. StH 842 has a double bend, perhaps appropriate for an edging of the type used at Sutton Hoo on both the helmet and the purse lid; StH 617 has a right angle.

StH 842  StH 617 Sutton Hoo purse lid detail

The total length of reeded strip fragments (such as StH 50, StH 74, StH 98, StH 99, StH 147, StH 1451 etc) comes to 156 cm.

 StH 776   StH 50

Both C-shaped edging and reeded strip did have other applications in the early Anglo-Saxon period, being found on items such as the Taplow drinking cups and horns but the curvature of the silver strips from the Staffordshire hoard point to them being on a helmet.

StH 1556

The Pressblach foils are paralleled on the Sutton Hoo helmet and on on several helmets from Scandinavia (see Karen Høilund Nielsen’s paper for more details). The foils on the Sutton Hoo helmet, however, are made from tinned copper alloy; the Staffordshire Hoard’s foils are of silver.

StH 1412

These fragments may be minute, but there are a great number of them. By measuring their thickness and relating this to the total mass of the foils, we can immediately work out the total area represented. Not all the thicknesses have been captured yet but a working average has been established of about 0.3 mm. We have 98.67g of foils, which gives an area that these foils would cover of 314 square centimetres. To put this into context, the cap of the Sutton Hoo helmet has an area of 906 square centimetres. Conceivably, therefore, all of the foils could have fitted on one helmet.

X-ray of soil lump from plate 02Many of the smaller fragments that are likely to have come from helmets were contained within the earth blocks. The University of Liverpool’s X-ray of one of these earth blocks shows reeded strips and Pressblech foils particularly clearly.

Possible shield fittings

It seems likely that StH 652, a magnificent gold mount depicting two birds of prey clutching a fish in their talons, came from a shield. The head of the fish (StH 1249) was found separately by Terry Herbert during the course of the excavation.

StH 652   StH 1249 

 StH 543 is another candidate for a shield mount. This double ring mount is very like the one on the shield at Sutton Hoo, although it has a curved underside, which might point to it being perhaps a drinking horn mount, like that from Grave 7 at Valsgärde (Bruce-Mitford 1978, 134 and fig. 101). This little item emphasises how little we know at present about the original uses of the objects within the hoard; there is so much work to do.

StH 543      Shield-ring from Mound 1 at Sutton Hoo

The bird and fish motifs are quite commonly found separately on Anglo-Saxon shields (Dickinson 2005). Cleatham grave 25 contained a bird and a fish mount on the same shield, and as both were made out of iron it seems that they were being used at both ends of the market (Leahy 2007, 235, figs. 18 and 86).

 Objects from Cleatham grave 25

Christian objects

StH 655 is the great cross, which is probably an altar cross or a processional cross and perhaps originally fixed to a wooden backing. This might well be linked to the military items in that it could have been carried into battle, itself a kind of weapon whether against an enemy which was pagan, or indeed one that was also Christian. This is a magnificent object, although the Style II decoration is perhaps a little rustic.  
StH 655     Reconstruction of the great cross StH655

The other crosses in the hoard include StH 303, a pendant cross with very dirty filigree and a central garnet. There are also two cross-shaped mounts, StH 820 and StH 920.

StH 303 StH 820 StH 920

StH 550 is also perhaps from a cross. It is of course the famous inscribed strip (see papers by Elisabeth Okasha, David Ganz and Michelle Brown for more detail on this). It reads ‘Rise up O Lord and may thy enemies be dispersed and those who hate thee driven from thy face’, which comes from Numbers 10.35, or from Psalm 67 in the Vulgate, which is Psalm 68 in modern Bibles and Psalters. I will not discuss this further, although as someone who works in Lincolnshire I must mention that when St Guthlac, having been a warrior as a young man, became a monk and then went to become a hermit, he sought true desolation in Lincolnshire. Here he was tormented by Welsh-speaking demons, and Guthlac, according to Felix’s life of the saint, dispersed the demons by reciting this very Psalm, Rise up O Lord (see Elisabeth Okasha’s paper for more detail on St Guthlac).

This may not, of course, be the only inscription in the hoard. It is possible that when we start to clean it we will find scratched graffiti including runic inscriptions. Although nothing has been seen yet, it is a consideration that we need to bear in mind as we work through the rest of the hoard.

Decorative techniques

Looked at by number, filigree was by far the most common technique, appearing as a primary technique on 62.4% of the objects. 24.9% of the objects in the hoard are mainly cloisonné-decorated, and 12.7% have relief decoration.

Decorative techniques in the Hoard

The decoration on the material is difficult to assess in its uncleaned state, especially on the filigree items which hold the earth very firmly. Karen Høilund Nielsen’s paper looks in detail at the Style II, which appears on about 80 to 90 pieces; one of the most stunning is the seax hilt-plate, StH 567, with animals directly comparable to those from f. 192vo of the Book of Durrow.

StH 567
Book of Durrow f.192vo, detail drawn by George Speake

There is one possible example of Style I so far, the silver relief-decorated fragment StH 298.

StH 298

60% of the sword pommels are decorated with filigree and 25% with cloisonné; a few have both techniques. Relief decoration is less common, and a few pommels are undecorated. Most of the undecorated examples are copper alloy, and it is likely that they are actually the copper-alloy cores of sword pommels which have lost their gold sheathing.

Decorative techniques found on the sword pommels

There are 22 cloisonné plates of various shapes, and 48 cloisonné strips of various forms. Of these strips, six are angled or folded into an L shape (e.g. StH 373, StH 662 and StH 773).

StH 373 StH 773

The Sutton Hoo sword pommel was made up from a series of separate components, and it is possible that some of these may be from composite objects made up in a similar way.

 Other unidentifiable slabs of cloisonné include a group of recessed or slotted strips such as StH 131, 275, 371, 513, 677 and 696, and perhaps more. Gold plates decorated with filigree animal decoration like StH 67, StH 68, StH 69, StH 89 and StH 556, were inserted into the recesses; one at least is still in place on StH 275.

StH 89  StH 556

StH 275

Pieces with a curved and pointed end (e.g. StH 668 and StH 673), were clearly made to fit into or up against something.

StH 668  StH 673

There are triple-faced strips (e.g. StH 356 and StH 357) decorated with cloisonné on three faces; StH 356 has an attachment lug, but this has not yet helped in assigning a function to this or any of these mysterious pieces.

StH 356 StH 357

At the moment the ideas are many and various, and include saddle mounts, book fittings, and shrine fittings. These pieces are taking us into new and exciting ground.

What's not in the Hoard?

Firstly and most importantly there are no sword blades; there is no Anglo-Saxon iron at all. Every bit had been stripped, not only from the swords but also from the helmet(s). Some of the Pressblech foils have got traces of iron corrosion on the underside, and the process of stripping might explain why so much of the silver is in small fragments. 

There are no coins, and there are also no feminine dress fittings, although among early Anglo-Saxon material in general these are more common than sword fittings. The hoard has been very carefully chosen from material which is either exclusively masculine, or apparently genderless. Within this material, however, there are no recognisable harness fittings, there are no large masculine buckles similar to the well-known examples from Taplow, Crundale, and Finglesham. There are two tiny buckles in the hoard (StH 144 and StH 685), of the sort that look appropriate for a helmet strap, but the big and showy baldric fittings such as we see at Sutton Hoo are completely absent. It seems that it was not just gold they were interested in.

StH 144 StH 685

It is also noteworthy that although the findspot is in the interface between Anglo-Saxon England and the Celtic West, there is nothing that looks in any way Celtic; it is wholly Germanic.

The wider context

Since the Portable Antiquities Scheme (PAS) has been in operation, it has added significantly to Menghin’s distribution of gold and silver sword-hilt fittings in England (1983).

As of 1st October 2010, the PAS has recorded nine gold and 24 silver sword pommels of sixth- and seventh-century date, plus six gold and 13 silver pyramids. The resulting map, with the site of the Staffordshire hoard marked as a rosette, shows a concentration of finds of silver in Kent, and of gold in East Anglia, which are perhaps to be expected. The band of gold finds in Lincolnshire and central England, though, looks very like it is following the course of the river Trent and its tributaries.

Gold and silver sword-hilt fittings recorded on the PAS database  The course of the River Trent

Locally, the PAS has recorded another seventh-century find, a gold and garnet pendant, from Hammerwich, now in the Potteries Museum (Evans 2004). From the same field as the hoard comes WMID-C28605, a copper-alloy roundel decorated in Style II interlace and with a central glass setting. It was found a hundred metres away, but it shows other seventh-century activity in an otherwise barren field.

7th-century gold and garnet pendant from elsewhere in Hammerwich          7th-century harness mount from the same field as the Hoard

Lastly, a map of all sixth- and seventh-century gold and silver finds recorded by the Portable Antiquities Scheme shows how the Staffordshire hoard fits into the broader pattern of precious-metal use in the early Anglo-Saxon period. It is close to the edge of the known distribution, but not beyond it. 

Gold and silver objects of early Anglo-Saxon date recorded on the PAS database

The Staffordshire Hoard cannot be seen in isolation and must be studied in the context both of material recorded by the PAS and of finds from elsewhere in Europe.  This discovery will re-ignite interest in earlier discoveries such as Sutton Hoo and Taplow, which can now be studied in the context of a much larger data-set.  We have interesting and challenging times before us. 


  • Bruce-Mitford, R. L. S., 1978.  The Sutton Hoo Ship Burial vol. 2: arms, armour and regalia (British Museum Press, London)
  • Evans, A. C., 2004. ‘Hammerwich, Staffordshire: Anglo-Saxon gold necklace pendant’ Treasure Annual Report 2004, 82, cat. no. 115 (Department for Culture, Media and Sport, London)
  • Leahy, K., 2007. Interrupting the Pots: excavation of the Cleatham Anglo-Saxon cemetery, North Lincolnshire (CBA Research Report 155)
  • Menghin, W., 1983.Das Schwert im frühen Mittelalter (Theiss, Stuttgart)