The Runic Sword-Pommel from Sarre, near Deal Grave 91
One of two sword-pommels from Sarre in the KAS collections, Maidstone Museum, under accession number KAS 838, carries traces of a runic inscription (Fig. 1). This pommel was published with a thumbnail illustration by John Brent in 1866, in the second instalment of his reports on the excavations in the cemetery at Sarre, where he assigned it to grave 91. The grave number, in Roman numerals, is incorrectly printed as CXI, which is unambiguously a typographical error for XCI (Brent 1866, 173).
[fg]png|Fig. 1 The sword-pommel from Sarre grave XCI. Upper: engraving published in Archaeologia Cantiana VI (1866), 173; Lower: photograph in present condition, by Andrew Ward.|Image[/fg]
The object in question was cast in copper alloy and shows signs of having been gilded. Typologically, it is a sword-pommel of form SW2-b in the scheme published for the English Heritage project on the chronology of Anglo-Saxon graves and grave goods of the 6th and 7th centuries AD. That sub-group of pommels was identified with Menghin’s Type Bifrons-Gilton, which he also numbered Type 2d. There is no very sharp line of division between Menghin’s Types 2b (‘Brighthampton- ciply’) and 2d, which are also Anglo-Saxon forms SW2-a and SW2-b (Menghin 1983, 309–15; Hines and Bayliss eds. 2013, 183–4). The former are relatively simple, the latter tend to be more elaborate. The presence of a beaded ridge around the base of the inscribed pommel from Sarre and the traces of gilding are just enough to justify its classification as Anglo-Saxon form SW2-b and Menghin Type [pg317]2d. In the Early Anglo-Saxon chronological scheme this type was current in phase AS-MB, which radiocarbon dating and Bayesian modelling assign largely to the period ad 525–565. The simpler forms SW2-a/Menghin 2b originated in the later fifth century, but Vera Evison’s firm dating of this specimen to the second half of the fifth century has to be considered mistaken (Evison 1967, 72–3).
This inscribed sword-pommel is highly abraded on both of its main faces. There is a hole in one side, close to the top, and the area around the peak of the ‘pyramidal’ shape shows even greater wear. This degree of erosion and attrition implies that the pommel could have been of considerable age when buried. With one exception, only ‘ghosts’ of the runes can now be made out. The presence of these runes seems to have remained unobserved for around a century after it was excavated but was independently noted in two publications in 1967. Hawkes and Page (1967, 1–3 and pl. IIa) published photographs taken at the Institute of Archaeology, Oxford, in 1966 showing what appear to be runes towards the right-hand sides of both main faces of the pommel. Page interpreted a sequence on the upper face shown in that published photograph montage as rdæt, possibly rd͡ jæt; the photograph shows marks that are indeed consistent with that reading, albeit only by way of informed conjecture. I shall refer to this face as Face A: it is the face to the left in the photo- montage in Fig. 1 and that illustrated by Brent in 1866. Evison (1967, 89, fig. 10b) also observed ‘traces’ of runes on both faces, as did Parsons in the 1990s (1999, 59–60). Parsons considered only some runes on one face to be legible: this must be Face A, partially transcribed by Page as just noted. On the other face (Face B: to the right in Fig. 1) he could make out only unidentifiable traces of staves and arms. He could agree with Page’s reading of the more ‘legible’ Face A only in respect of recognizing an r-rune. In the late 1990s Page and Parsons jointly re-examined the pommel in what should have been optimal observational conditions at the British Museum, but could discern nothing more: indeed it is not clear that they were then able to read anything on the piece, although it does seem they remained persuaded about the presence of an r-rune on Face A. Evison, conversely, identified only a d and a u on one face, which seems to have been Face B, although Parsons evidently inferred that she meant the face on which he and Page had identified an r-rune, i.e. Face A. On Face B, where Evison read du, Page had, in the 1960s, believed he could see the innovative Anglo-Saxon diphthong rune e͜ a and possibly a particular form of c rune further to the left. But neither of those suggested rune-forms is attested before the eighth century to our present knowledge, and their presence on a sixth-century sword-pommel may be considered most implausible.
In a nutshell, between the mid-1960s and the end of the last century an experienced and careful archaeologist and two experienced and cautious runologists had produced only confused, inconsistent, and in some respects quite unlikely identifications of runes on the pommel. This is evidently a challenging case. Only some of the faint marks indicated by the persuasive photographic image published by Hawkes and Page in 1967 are now visible, however one adjusts the lighting and whatever form of magnification is used. Even with the published photograph as a guide, I could discern nothing really be convincingly runic on Face A of the pommel, which Page, nearly sixty years ago, found to carry the ‘better preserved’ sequence, and where Parsons confirmed he could make out the rune r. On Face B of the pommel, conversely, neither Page nor Parsons recognized the u read by [pg318]Evison. At the right-hand side of Face B, however, I can see what looks like quite a wide u, with, under magnification, an appropriately rounded top where the vertical main stave and side stave meet. To the right of that is what might be the truncated fragment of another vertical stave, which could be the rune i, or conceivably part of a single rune along with what Evison and I would read as u, forming an unusual variant of h — resembling, in fact, what has also been read as h on the Sandwich stone with its crossing stave connecting the top of the first vertical stave with the bottom of the second one (Hines 1998, 190–2; cf. Parsons 1994). Preceding the putative u I cannot see a complete d-rune but can make out part of the lower right-hand corner of what Evison had read as d — the bottom of the right-hand vertical stave with the end of a diagonal cross-stave meeting it.
From my own personal observation, then, only a final u on Face B might be proposed as a relatively plausible identification. The r recognized and attested to by both Page and Parsons on Face A, with the support of a photographic image, must, however, be afforded the same level of credence. In light of all of the evidence summarized here, no readings of runes on this sword-pommel can be regarded as certain. When the present observations, and indeed those of Page and Parsons from the later 1990s, are compared with the photographs taken in 1966, it would appear that the object has deteriorated considerably in the course of the last sixty years. The dark green staining around the peak of the pommel suggests chloride corrosion in over-humid conditions at some stage. Evison (1967) listed this pommel amongst a group of ‘ring-swords’, and evidently believed that the hole on one side of the pommel was where a symbolic ring had been attached. But there is no parallel for such an attachment in this position. The hole looks like further damage; it is shown in Brent’s nineteenth-century illustration (see above) and so was already there when the object was unearthed, but would appear to have become larger. The shading on the line-drawing might only show wear around the peak, and is ambiguous as to the presence and extent of any ‘copper disease’ already then present in this area; indeed, it is reasonably clear that the representation of the pommel in this engraving has been tidied up and idealized to a degree.
All the same, there are further observations of significance to be made, and if the object is decaying it is of particular value to record these now. On Face B, there is a very clear, incised, horizontal, upper framing line for the inscription. This has not previously been registered. Under x3.5 magnification, and with the lighting from the appropriate angle, I was also able to see a clear, double incised line cutting through where the inscription would have been on this face, angled at a slight slope across the face, being higher to the left-hand side than to the right. This gave the impression of a cut-mark intended to cross out or cancel the inscription. All of these horizontal/near-horizontal lines on Face B are clearly attested in the photographs taken in the 1960s. Towards the left-hand side of the same face, at the level of magnification just noted, I also saw a clearly incised, short line descending from the upper border line, at about 45 degrees towards the right, just like the upper by-stave of the rune a [in the later Old English fuþorc: æ] or l. But there were no other visible incisions connected to or associable with that to produce an identifiable rune.
More recent finds have revealed that runic inscriptions on late fifth- and sixth-century sword-pommels are a feature both of south-east England, in effect Kent, [pg319]and a region of north-east France around the Rivers Rhine and Saône (Fischer and Graf 2016). The Sarre grave 91 sword-pommel is unique amongst these in apparently having carried runes on both faces of the pommel. It is a reasonable question to ask, whether the differences in legibility various experts have found between the two faces, and the newly observed sign of a possible cancellation, could be due to the runes on each face having been added at different times and in different styles, with whatever was inscribed on one face crossed through and replaced by whatever was inscribed on the other.
It is a matter of importance to check that the present conditions of curation, particularly in respect of humidity, are optimal. It would also appear that it may not be too late for technically sophisticated surface scanning and mapping to record important information, which might not retrieve any more of what are truly long-lost inscriptions, but could reveal informative details of the pommel’s object-biography. We are very grateful that the KAS agreed to loan the sword-pommel for detailed examination and conservation assessment at Cardiff University, where it arrived in December 2024. We shall report on the results of these investigations in due course.
Examined Friday 13 September 2024; interim report updated February 2025.
John Hines
References
Brent, J. 1866. ‘Account of the Society’s researches in the Anglo-Saxon cemetery at Sarr’. Archaeologia Cantiana 6, 157–85.
Evison, V. I. 1967. ‘The Dover ring-sword and other sword-rings and beads’. Archaeo- logia101, 63–118.
Fischer, S. and Graf, M. H. 2016. ; ‘‘Runes and rivers’: considerations on the typology and distribution patters of Merovingian Period runic inscriptions along the North Sea, the Rhine and the Danube’. In É. Petremann (ed.), Des fleuves et des hommes à l’époque mérovingienne (ARTEHIS). https://doi.org/10.4000/books.artehis.24289.
Hawkes, S. C. and Page, R. I. 1967. ‘Swords and runes in south-east England’, The Antiquaries Journal 47, 1–26.
Hines, J. 1998. ‘Grave finds with runic inscriptions from Great Britain’. In K. Düwel (ed.), Runeninschriften als Quellen interdisziplinärer Forschung (Berlin: De Gruyter), 186 96.
Hines, J. and Bayliss, A. (eds) 2013. Anglo-Saxon Graves and Grave Goods of the 6th and 7th Centuries A.D.: A Chronological Framework (Society for Medieval Archaeology Monograph 33: London).
Menghin, W. 1983. Das Schwert im frühen Mittelalter (Stuttgart, Theiss).
Parsons, D. N. 1994. ‘Sandwich: the oldest Scandinavian rune-stone in England?’. In B Ambrosiani and H. Clarke (eds.), Developments around the Baltic and the North Sea in the Viking Age (Birka Studies 3. Stockholm), 310–20.
Parsons, D. N. 1999. Recasting the Runes: The Reform of the Anglo-Saxon Futhorc (Runrön 14: Uppsala, Institutionen för nordiska språk).