Anglo-Saxon Remains

Pages 373-381

A History of the County of Somerset: Volume 1. Originally published by Victoria County History, London, 1906.

This free content was digitised by double rekeying. All rights reserved.



If reliance can be placed on an entry in the Anglo-Saxon Chronicle under the year 577, and on the judgment of successive editors who agree in identifying a place there mentioned, archæology has a fixed starting-point for the treatment of post-Roman Somerset. The battle of Deorham, the event referred to, is the first recorded in the Anglo-Saxon history of the west, but though the site is generally acknowledged to be Dyrham, there is no clear indication of the route by which the invaders reached the fertile valley of the Severn. It is natural to imagine the advance on Bath to have been along the upper Thames valley, but it occurred to Professor Freeman, (fn. 1) and possibly to others, that the Saxons worked their way gradually west from the region of Southampton Water, by way of Salisbury, to Exeter. If this route were ever adopted for operations against the Britons of the west it was of a much later date than the advance against the Roman inhabitants of the Cotswolds; and it is more probable that Ceawlin led his West Saxons from the upper valley of the Thames, where evidence of their early occupation is abundant, southwards into what is now the county of Somerset.

The year 577 saw not only the reduction of the three RomanoBritish townships and the death of their chieftains, Conmail, Condidan and Farinmail, but also the complete isolation of the West Britons, who could no longer act in concert with the North Welsh of our modern Wales, and were step by step compelled to retire into Cornwall. Through Somerset their retreat is marked by sundry entries in the Anglo-Saxon Chronicle, but for two generations after Deorham there seems to have been no decided activity on the part of the English in this region. It is probable that their energies were fully occupied elsewhere, for the soil of Somerset has yielded next to nothing characteristic of the pagan Saxon. The national collection contains only three objects of the kind from Somerset, and two of these (fig. 4) are brooches of a type best represented in the important series from Long Wittenham, Berks; but specimens have been found in many parts of southern England, and they cannot be considered characteristic of any particular branch of the Anglo-Saxon stock. The third (fig. 3) is of more importance, and was evidently made in imitation of a type common in the Isle of Wight and apparently confined originally to the Jutish area. A specimen in the British Museum from Chessell Down, Isle of Wight, not only corresponds in size and outline, but explains the incised lines which have little meaning as they stand, but are survivals of the garnet settings that enriched the earlier examples. By itself this brooch can tell us little of the Sumorsætan, but it is interesting to note that Jutish forms are occasionally met with in the burials of Wiltshire and Berkshire; and its discovery with others at Ilchester, as well as the derived character of its ornamentation, justify its attribution to the second half of the seventh century, when the Saxon conquerors had reached the Parret.

Another discovery was made not far from this station on the Fosse Way. In the Castle Museum at Taunton there is an iron shield-boss of the usual pattern, that was found on Ham Hill (Hamdon) between Yeovil and South Petherton; and a Saxon dagger and spear-butt are said to have been found in one of the hut-circles on Worle Hill. (fn. 2) This site has been excavated with success, and proved to be a stronghold of an earlier population; so that the Saxon relic is here but slender evidence of occupation to any extent in the period preceding the conversion of Wessex. That was not accomplished till the middle of the seventh century, and till that date, perhaps for some time longer, it was customary to bury the warrior fully armed and the housewife with her ornaments and domestic utensils.

So far then there are some slight indications that the Saxons did not occupy in force the territory they had conquered, at any rate beyond the Bristol Avon, though the land immediately to the south of Bath no doubt passed into their hands with the Roman city; and archæological researches lend support to the theory that for three parts of a century the West Saxon territory was here bounded to the south by the earthwork known as Wansdyke.

There exist few earthworks to which a definite date can be assigned, and fewer still that can be ascribed with certainty to the Anglo-Saxon period; but what may prove a notable exception can be traced to this day in Somerset and Wilts. The Wansdyke has a course of about sixty miles, and runs from the marsh-land near the Severn at Portishead, passing by Bath to the north of Devizes, through Savernake Forest to Chisbury Camp, where it turns southward in the direction of Andover. As far as Marlborough its continuity is clear except in places where, to judge from the nature of the soil, dense woodland already formed a natural defence.

Throughout its length the dyke or rampart was strengthened by a ditch on the northern side, and was evidently intended to meet a hostile advance from that quarter. The scientific and elaborate excavations of the late General Pitt-Rivers on the line of this and the kindred work called Bokerly have for ever disposed of the theory that both were erected by the Belgæ before Cæsar's time, as step by step they forced a passage into the heart of the country from the southern coast. References to archæological literature on the subject are furnished by the General, (fn. 3) who was careful to distinguish between the tangible results of his excavations and the comments he thought fit to make upon them. It was in this judicious spirit that he pronounced the Wansdyke to be of Roman or post-Roman origin; and though the evidence of date is more decisive in the case of Bokerly Dyke, he was disposed to regard them as almost contemporary. The more southern earthwork is proved by coins to have been thrown up at or after the time when the Roman forces and officials withdrew from Britain; and a partial examination of the larger work afforded grounds for the belief that the Roman road from Marlborough to Bath was of earlier date and for some distance ran beneath the rampart. (fn. 4)

There is little choice then but to assign the Wansdyke to the period between the Roman and Saxon dominations. On what occasion this barrier was erected it is difficult to surmise, but it may well have served to arrest the southern advance of the Teutonic occupants who settled near the upper Thames and left but few traces of their presence during the pagan period in the counties further to the south.

It is true that Wiltshire is comparatively rich in Saxon grave relics; but the cemetery at Harnham Hill near Salisbury has been assigned to the Christian period on independent evidence, (fn. 5) and it is possible that other Saxon graves in the county were not much earlier. In reliance therefore on data afforded by excavation, the suggestion may be hazarded that the Wansdyke marked the first, and Bokerly a subsequent, line of defence against West Saxon encroachment from the Thames.

So far as Somerset is concerned the written records point to the same conclusion, and the tale of the Anglo-Saxon Chronicle may now be resumed. (fn. 6) In 652 a forward step was taken by Cenwalh, third in descent from Ceawlin, and an English victory is recorded at Bradford-onAvon. Six years later the Saxon king advanced his frontier to the Parret, thus including the fastnesses of the Polden Hills. From the Parret the next advance was to the coast, for in 682, after an interval of a quarter of a century, the Britons were driven to the sea by Centwine. In the meantime Wessex had been involved in domestic struggles, and at least the first ten years of the succeeding interval were occupied with the Kentish campaigns of Ceadwalla and Ine. The victory of the latter king in 710 comes immediately after the abdication of the Mercian king Cœnred, (fn. 7) and Ine no doubt availed himself of such an opportunity of extending his dominion in the west, where the interests of Mercia and Wessex were beginning to clash.

The entry under 710 in the Anglo-Saxon Chronicle is of special interest as giving the name of the British king against whom this campaign of the West Saxons was directed; and a contemporary letter of Aldhelm bears independent testimony to the historical character and political importance of King Geraint. It is open to conjecture that the defeat of the Britons compelled them to retire to the hills of Exmoor, and thus leave what is now the western portion of the county in English hands. The extent of this new accession to the West Saxon realm may indeed be marked by the county border as it exists to-day, in which case the Sumorsætan would have been in full possession early in the eighth century. The further doings of King Ine, the foundation of Taunton and the civil war that necessitated the razing of that fortress by his queen in 722 belong rather to the political history of the county, and are not essential to a treatment of the AngloSaxon remains recovered from the soil. For our present purpose the important points are the successive advances of the English, as recorded in their own chronicles (rather than in those of the conquered Britons), and the collateral evidence of British survivals under Anglo-Saxon rule. In this connection some ethnological evidence may be adduced. 'It is a fact that the people of the eastern half of the county have, on the whole, broader heads, lighter hair and darker eyes than those of the western half. In all these respects the eastern men approach more to the ordinary English, the western to the Irish standard. These are the clearest and most important differences between them, and are very much what we might have expected to discover.' (fn. 8) As to the line of separation, a local observer has stated his opinion that there is a notable difference in physique, as well as in dialect, between the people to the east of the Parret and those to the west of it—the eastern men being larger and having more of the Saxon type. (fn. 9) There are obvious reasons therefore why the actual relics of the conquerors are scanty in these parts; for, to leave its mark on the speech and stature of the inhabitants, the frontier must have been maintained for a considerable period, and the conquering Saxon held at bay.

Though earlier traces of the Sumorsætan are for the most part wanting, the county has yet produced a relic of Anglo-Saxon times that for intrinsic value and historical interest far surpasses anything of the period. It was in 1693 that what has long been known as the 'Alfred jewel,' was discovered during some excavations about three miles from a site that is intimately connected with the heroic figure whose millenary has recently been celebrated.

The Isle of Athelney lies very near the meeting-place of Tone and Parret, and the mound now known as King Alfred's Fort commands the junction of the streams. To the north-west, about half-way to Bridgwater, is an estate known either as Newton or Petherton Park, the villages of North Petherton and North Newton being little over a mile distant. The discovery took place in the park, which was then the property of Sir Thomas Wrothe, and five years later the jewel was in the hands of his uncle, Colonel Nathaniel Palmer of Fairfield. In 1717 it passed to the latter's son Thomas, who carried out his father's wishes by presenting it to the Bodleian Library. It has thus been the property of the University of Oxford for nearly two centuries, and is to-day perhaps the greatest treasure of the Ashmolean Museum.

The first description of the jewel was contributed to the Philosophical Transactions (No. 247, 1698) by Dr. William Musgrave of New College, and published by Dr. Hans Sloane, who had been elected secretary of the Royal Society in the very year of the discovery. It would be tedious to record the numerous notices that have since appeared, but special mention must be made of the elaborate account in the first volume of Hickes' Thesaurus (1705), pp. 142–4.

The full-size drawings (figs. A, B, C, D) will render a lengthy technical description unnecessary, and its general features can be presented in a few lines. An oval framework of gold encloses an enamelled plaque, which is protected by a thick slab of crystal. The figure is in cellwork, the coloured enamels being separated and the outlines traced by means of vertical partitions of gold attached to the plate, which serves as the base and is engraved with a floral design on the back. The openwork lettering passes all round the sloping edge of the crystal, except where a monster's head projects from the point of the oval and provides a socket for a thin rod that was originally fastened by a rivet still in position.

To account for the existence and character of such a product of the goldsmith's art is a task of greater difficulty, and what internal evidence there is must be reviewed. The Anglo-Saxon inscription is clear enough : + AELFRED MEC HEHT GEWYRCAN, or, in modern English, 'Alfred ordered me to be worked.' The Roman forms of the letters are here of little evidential value as they were in common use on the coinage of the time ; but the peculiar character of the language is unmistakable. HEHT is really a Mercian form, and Professor Earle said that both MEC and HEHT were archaic in the ninth century, the one being never, and the other rarely, found in the prose of the tenth century. MEC had been superseded by ME, while HEHT had given place to HÊT; but the older forms were still at the service of the poet, who sometimes retained them for metrical reasons. Both MEC and ME are Anglian forms, and the occurrence of the former in West Saxon poetry is due to its retention when the poems were altered into the dominant dialect ; nor is it impossible that Alfred employed an Anglian goldsmith. (fn. 10) The old form of the personal pronoun occurs twice in similar phrases on a gold ring of about King Alfred's time, now in the national collection, and favours the common attribution of the Alfred jewel.

An ornament of such costliness and splendour must have been executed for a person of no mean station, and the occurrence of the name of Aelfred in such a context would alone justify a belief that the relic once belonged to the founder of England's greatness. Other pieces of jewellery connected with the royal family have, by a fortunate accident, come down to us from that time, and in the British Museum are the gold finger-rings of Alfred's father, Ethelwulf, and of his sister, Ethelswitha. The belief, thus further encouraged, is rendered almost a certainty by the fact that the Alfred jewel was dug up in the immediate neighbourhood of Athelney, the scene of the king's humiliation before his enemies the Danes. It was in 878 that the fortunes of England were at their lowest ebb, and Athelney the only English ground that was not given over to the heathen raiders. There we know that Alfred was, devising means for his country's deliverance ; so that, accepting the connection of the jewel with the great Alfred as more than probable, the temptation becomes stronger to fix its date and place of origin with more precision.

The absence of the regal title from the inscription is an argument in favour of dating its manufacture before 871 ; but if the attractive theory, put forward a quarter of a century ago, is the correct one, it is more likely that the jewel was made some time after 878 and presented to the abbey which the king devoutly founded on the Isle of Athelney, when peace was at length secured at Wedmore.

Anglo-Saxon Remains from Somerset

With all deference to the lately deceased scholar from whose charming monograph on The Alfred Jewel many of the above details have been derived, it cannot be conceded that the latest theory as to its use is preferable to that of Bishop Clifford. At the Bridgwater meeting of the Somersetshire Archæological Society in 1877 the Bishop of Clifton put forward the view that the Alfred jewel was the head of an ' æstel of (the value of) 50 mancuses' such as the king sent with a copy of his translation of Pope Gregory's Pastoral to every bishop's see in the kingdom. Now in the preface to that translation, written some time after 890, Alfred mentions as one of his instructors in Latin the mass-priest John, whom he made Abbot of Athelney; and it is therefore by no means improbable that a copy with its æstel came in due course to the monastery. What the æstel was is uncertain, (fn. 11) but from its Latin equivalent, indicatorium, it was clearly a small rod or pointer to be used with the volume, and perhaps inserted in the binding. That the Alfred jewel had originally a rod of some stout but perishable material inserted in the socket is obvious; and it may be that, of the royal gift to Athelney, the æstel-head alone survived. It is true that none of the existing copies (those belonging to the sees of Worcester, Canterbury, and Sherborne) retains any pointer of this description ; and the form of the jewel is not adapted, as Bishop Clifford himself supposed, to the hand. Much less is the remarkably similar jewel (fn. 12) from Minster Lovel, Oxon, likewise preserved in the Ashmolean Museum. This was to all appearance used for the same purpose, belonging to the same period; and, while it is easily conceivable that Oxford possesses two æstel-heads dating from the revival of letters under Alfred, it is surely improbable to a degree that either jewel was originally the central ornament of a crown or coronet. The suggestion that they severally graced the helmets of Anglo-Saxon war-lords is even less plausible, and other conjectures as to their use are as numerous as the interpretations of the central figure of the Alfred jewel.

The halo which can perhaps be detected on the front, would testify to the religious character of the subject, but opinions are divided as to the identity of the figure. Whether the representation is of the glorified Saviour or the Pope of Rome, even (as some have suggested) of St. Cuthbert or St. Neot, the two ' sceptres ' are clearly the essential attribute, though not sufficiently distinctive for purposes of identification. They may be seen as attributes of Our Lord in the Gospels of St. Chad, of the late eighth century, and in the Book of Kells. (fn. 13)

A dissertation on the technique of the enamels would here be out of place, but comparison with a golden ouche or brooch formerly in the Roach Smith collection and found in London, shows that portraiture in the same style and material was not unknown on jewellery at that time both in England and abroad. (fn. 14) Whether all or any examples found in this country were of home manufacture or imported from the continent can only be decided by the production of more evidence than is at present available. Meanwhile the fact that the enamelling is the least satisfactory part of the Alfred jewel will weigh, even with the most sceptical, in favour of its insular origin. The Anglo-Saxon goldsmiths were famous, and a craft that could produce the wonderful jewellery of the Kentish graves in the seventh century would be quite equal to producing the Alfred jewel at the end of the ninth, when Christianity had brought civilization in its train. Moreover our museums show that enamelling, as well as other elaborate methods of decoration, were practised both in this country and in Ireland with eminent success before the year 900. The famous Ardagh chalice, which is enamelled by three distinct methods, belongs to that period.

The influence of an art best represented by discoveries on Irish soil may without difficulty be traced on the Alfred jewel itself. The so-called boar's head that holds the socket at the base of the jewel is nothing but an example in the round of what is more often seen in the illuminated manuscripts of the Irish school. (fn. 15) Parallels in metal are less frequently found in this country than in Scandinavia, but are even there attributed to Irish influence.

The artistic influence of the Irish school, spread through northern Europe by the zeal of missionaries, culminated about the year 900, (fn. 16) and is perhaps again exemplified in Somerset by a remarkable open-work brooch of gilt-bronze (fig. 5) discovered in Pitney churchyard some years ago. The front is slightly convex and is ornamented in a most pleasing manner with an animal form that can be distinguished among the interlacing curves. In style it differs altogether from the early Anglo-Saxon treatment of animal forms, and bears a very close resemblance to the design of a book-binding (fn. 17) preserved in the museum of the Royal Irish Academy and no doubt found in Ireland. (fn. 18) Its date is not determined, but the binding is if anything a little later than the Pitney brooch.

Another relic of more than local interest may perhaps be dated from a consideration of the site of its discovery. A gold finger-ring (fig. 2), now exhibited in the Pump Room at Bath, was found during excavations for the new Guildhall above some graves, and about three feet higher than the present basement floor level, at a spot near the back entrance and opposite the Central Police Station. It is formed of two sets of four wires, the bezel or front being composed of two figures of 8 interlaced, the strands consisting of two twisted wires with a single wire on either side of them. At the back the tapering wires are hammered together into a diamond-shaped joint that is itself a characteristic feature of gold and silver work in the Viking period (a.d. 700–1000). There is evidence that for a long time after the capture of Bath by the Saxons the Roman buildings were suffered to decay; but the city was no longer waste when Edgar was there crowned king in the year 973. Hence there is nothing improbable in the view that the ring was manufactured in the tenth century, and was subsequently lost on the Roman site by some important personage, perhaps at the very time of the coronation.

Leaden Coffin-plate found at Bath

Leaden Coffin-plate found at Bath.

In the same collection is a leaden obituary cross that was fully described (fn. 19) by the late Major Davis, who was the hon. curator of the Corporation Museum. The accompanying illustrations (fn. 20) will sufficiently indicate its size and character, but the inscriptions on both sides, which were engraved on a wax model and transferred to the lead by casting, may here be set out at length, mainly in accordance with a reading by Mr. Blakiston included in the pamphlet already referred to. The front has the names of the four evangelists round the border, and four titles of the Deity, viz. ELOE, or ELOI, ADONAI, SABAI and ΘΕΟΙ, between the arms of the cross. These titles are of special interest as being of Gnostic origin and generally used in the middle ages as words of power.

Down the centre is a Latin invocation:—

Qui in virtute crucis mundum (purgavit)
Tartara disrupit claustra celestia aperuit
Et omnibus dedit pacem fidelibus Salutem
the words and letters in italics being supplied. This may be translated. 'He who by the power of the Cross redeemed the world, burst asunder the gates of Hell, opened those of Heaven, gave peace to all and salvation to the faithful.'

On the arms the inscription is apparently as follows, between the symbolic letters Alpha and Omega:—

Christe omnium hominum cunabula cuncta disponens
Purifica me squalore sorde volutatam
Supplex tibi Domine deposco miserere mei
which may be rendered: 'O Christ, who orderest the birth of all, purify me who am polluted by the stain of sin, I suppliantly beseech Thee, O Lord, have mercy on me.'

The inscription on the back is of a more personal character and professes to give the date:—

Anno ab incarnatione Domini nostri die
xv Kal. Octobris . . . Eadgyvu . . .
congregationis soror

This may mean that Eadgyvu, a sister of the community, died in the year . . . after the Incarnation of Our Lord, on the 17th September.

The cross was probably attached to the coffin of Sister Eadgyvu, belonging to some convent in or near Bath; but her identification with the wife of King Edward the Martyr is anything but certain, and Mr. W. H. Stevenson points out that the use of v for f in Eadgyvu indicates the eleventh century. If the latter date is correct, comparison should rather be made with the inscribed leaden crosses (fn. 21) of that period found at Bury St. Edmunds and elsewhere, one of which may be seen in the national collection.


  • 1. Somersetshire Archæological and Natural History Society, Proceedings (1874), pt. ii p. 7.
  • 2. Somersetshire Archæological and Natural History Society, Proceedings (1856–7), vii. pt. ii. p. 51. The bone comb found on this site with an iron spearhead, and exhibited in 1864 to the British Archæological Association (Journal, xx. 329, see also xxiv. 61), may have been of early British date.
  • 3. Excavations in Bokerly Dyke and Wansdyke (1892), iii. 25; see also pp. xiii. 246.
  • 4. Somersetshire Archæological and Natural History Society, Proceedings, vi. pt. ii. p. 101, and vii. pt. ii. pp. 9, 50. The view expressed in volume for 1876, pt. i. p. 63 must be erroneous.
  • 5. Davis and Thurnam, Crania Britannica, pt. ii.
  • 6. T. Kerslake, 'The first West Saxon Penetration into Somersetshire,' Somersetshire Archæological and Natural History Society, Proceedings (1876), pt. ii. p. 61.
  • 7. Professor Freeman on King Ine, Somersetshire Archæological and Natural History Society, Proceedings (1872), pt. ii. p. 41.
  • 8. Dr. Beddoe, 'Ethnology of Somerset,' Somersetshire Archæological and Natural History Society, Proceedings (1873), xix. pt. ii. p. 65; pt. i. p. 37.
  • 9. Mr. Prankard of Langport, quoted in same paper, p. 70. See also Mr. F. T. Elworthy's paper, 'On the Grammar and Dialect of West Somerset' (1876), pt. i. p. 43, and pt. ii. p. 31.
  • 10. Additional notes on the grammar of the inscription have been kindly furnished by Mr. W. H. Stevenson.
  • 11. Æstel is described as a book-mark in Lye's Dictionarium, but in Toller's edition of Bosworth's Anglo-Saxon Dictionary it is said to be a tablet for memoranda, consisting of two waxed leaves joined together by a hinge and framed or covered with gold to the value of 50 mancuses. Mr. Sweet admits that his rendering ('clasp') is purely conjectural (Gregory's Pastoral Care, p. 473); and it is translated 'stylus' in Elfric's Glossary.
  • 12. Figured in colours in Professor Earle's The Alfred Jewel, p. 47.
  • 13. Ibid. p. 77.
  • 14. A very similar piece of work, probably found in north Germany, is compared by Dr. Franz Bock with Hiberno-Scottish productions of the seventh and eighth centuries and figured in Die byzantinischen Zellenschmelzen, pl. xxiv. fig. 2; see p. 366.
  • 15. Westwood, Facsimiles of Anglo-Saxon and Irish Miniatures, plates 3, 18, 20 : all attributed to the eighth century.
  • 16. Sophus Müller, Die Thier-ornamentik im Norden, p. 92.
  • 17. Figured by Miss Margaret Stokes in Early Christian Art in Ireland, fig. 46, p. 110.
  • 18. Irish pilgrims are said to have settled at Glastonbury, and educated St. Dunstan in the first half of the tenth century. See Stubbs, Life of St. Dunstan (Rolls Series), pp. lxxviii. lxxxiii.
  • 19. The Saxon Cross found in Bath, July, 1898, by Chas. E. Davis, F. S. A.
  • 20. Photographs by W. G. Lewis, Bath : full size.
  • 21. Society of Antiquaries, Proceedings, iii. 47, 165; iv. 211.