The Amherst Brooch

THE AMHERST BROOCH SONIA CHADWICK HAWKES Exactly fifty years ago, one of the finest pieces of seventh-century Anglo-Saxon gold jewellery, a Kentish jewelled composite disc brooch of a quality to compare with the larger and more famous brooch from Kingston Down grave 205,' came up for public auction at Sotheby's.2 Lot 117A, the Property of the Rt. Hon. the Earl Amherst, appears to have been a late entry in a sale dominated by antiquities of other places and periods, but it did not pass unnoticed. Christopher Hawkes, then an Assistant Keeper in the Department of British and Medieval Antiquities in the British Museum, has a vivid recollection of 'a young chap from Sotheby's, who had not the least notion what it was' bringing the brooch round to see if the Museum would be interested in purchasing it, and of 'being pleased to be able to recognise it as an important Anglo-Saxon brooch.' He duly reported it to Reginald Smith and thence to T.D. Kendrick, his seniors in the Department and both experts in the archaeology of the Anglo-Saxon period. Shortly after, less than three weeks before the sale on 18th July, 1934, we find Kendrick writing to E.T. Leeds, then Keeper of the Ashmolean Museum and its Department of Antiquities, and the other great authority on the subject: 'Last week I was rejoiced to see for the first time the Sarre brooch illustrated in the Gloucester Volume of the Arch. Congress. Its a beauty, but has, I think, a faked back. It is coming up at Sotheby's and we shall try our best to get it.'3 Whether Leeds would have known about the sale 1 Bryan Faussett (Ed. C. Roach Smith), Inventorium Sepulchrale (1856), 77 ff., PI. 1,1; and innumerable appearances since, down to Richard Avent, Anglo-Saxon Disc and Composite Brooches (BAR 11, 1975) no. 179, 47 ff., PI. 68. 2 Sotheby and Co., Catalogue of the Egyptian, Babylonian, Greek, Roman, Byzantine and South American Antiquities, etc., Day of Sale Wednesday, the 18th of July, 1934. f 3 Letter dated 29th June, 1934. Ashmolean Museum archives. 129 S.C. HAWKES otherwise is not known, but this tip seems certainly to have put him on his mettle. In the event, the bidding for the Amherst Brooch has been described as 'spirited'4 as the two great Anglo-Saxon archaeologists, friends and rivals, competed to secure it for their respective museums. From the copy of the catalogue used by Leeds, it is evident that he arrived early at the sale-room, for he noted the prices fetched from lot 6 onwards. Only one or two items went for more than modest sums, but the Amherst Brooch was bidden up highest of everything. Beside it Leeds wrote £260, his only sign of emotion being to ring the figure firmly, and thereafter to lose interest in the few remaining lots. And with good reason: by £10 he had outbid Kendrick and secured a great addition to the Anglo-Saxon collections in the Ashmolean Museum. It must have been with feelings of great satisfaction that he put his new treasure (protected by the leather case mentioned in the catalogue) in his waistcoat pocket and took the next train back to Oxford. How he had raised what in 1934 must have seemed an astronomical sum of money to pay for a piece of Anglo-Saxon jewellery is not recorded. Hawkes suggests that Sir Arthur Evans, always generous to the museum of which he himself had been Keeper, may have helped Leeds with an anonymous gift of money. Hawkes also recalls that Kendrick took his defeat in good part and, on being questioned eagerly the next morning, said: 'Well, there it is, Leeds has got it. It's no use grumbling, we hadn't the money. And at least it has gone to a good home.' The Amherst Brooch has been a prime exhibit in the 'gold case' amongst the Anglo-Saxon antiquities in the Ashmolean Museum at Oxford ever since.5 The curious thing is that Leeds made so little archaeological capital out of his splendid purchase. His Report to the Visitors is very spare: 'The Kentish brooch (PI. lie) already mentioned ranks with the finest of its class, a small group of large circular brooches of elaborate fabric belonging to the seventh century. It was found many years ago at Sarre, in the Isle of Thanet, Kent, and it is mentioned in Transactions of the British Archaeological Association (Gloucester), 1846, p. 340, and illustrated in colour on the frontispiece. It is of composite form, with gold rim and cloisons set on a silver back, and is in excellent preservation except for the loss of some of the settings. The pattern of the cloisons, while conforming to the general type of these brooches, presents some unusual details; the combination of garnet, yellow glass (the latter largely missing) and 4 Ronald Jessup, Anglo-Saxon Jewellery (1950), 117. 5 Registration number 1934.202. 130 THE AMHERST BROOCH green glass paste is not met with on any other known example.'6 This is good observation, so far as it goes, but it was strictly for home consumption. He and others have made the Amherst Brooch familiar to a wider audience since,7 but not in any particular way. It seems strange that Leeds, at least, did not make it the subject of a detailed study for its own sake. Technically, it is extremely interesting, and the question of its provenance constitutes a minor mystery which deserves to be resolved. I have long been suspicious of the attribution to Sarre,8 and at first thought to confine this paper to just that aspect. But on handling the brooch itself, thoroughly for the first time, and with the aid of a powerful binocular microscope, it seemed that more might profitably be said about the jewel itself, which has generally been much underrated. Unfortunately, there has not been time or opportunity to get all the desirable technical analyses done, but perhaps readers of this volume would not welcome a report stuffed with wads of scientific data. In due course, one hopes, they may appear elsewhere. Meanwhile, this article should be regarded primarily as a preliminary essay towards an appreciation of the Amherst Brooch and the various problems connected with it. THE BROOCH ITSELF A major reason why the Amherst Brooch roused such interest in 1934 seems to be that, though it was a known find dating from the early 1840s, it had been in private collection and lost to view for something like seventy-five years. In his letter to Leeds (above p. 129) Kendrick, who had just the previous year published a paper on Kentish jewellery,9 admits that he had now seen it for the first time. It is also noticeably absent from the works of older scholars who would have published it had it been accessible.10 One is forced to conclude that the last time it had been on show in archaeological circles was probably on the occasion of the Second Annual General Meeting of 6 Ashmolean Museum. Report of the Visitors (1934), 15. 7 E.T. Leeds, Early Anglo-Saxon Art and Archaeology (1936), 119, PI. XXXIII, 5; Jessup, op. cit (1950), 117, PI. XXVI, 1; Avent, op. cit. (1975), no. 178, 47, PI. 67. 8 Sonia Chadwick Hawkes, 'The Monkton Brooch', Antiq. Journ., liv (1975), 252; Sonia Chadwick Hawkes and A.C. Hogarth, 'The Anglo-Saxon Cemetery at Monkton, Thanet', Arch. Cant., Ixxxix (1974), 81 ff., note 65. 9 T.D. Kendrick, 'Polychrome Jewellery in Kent', Antiquity, vii (1933), 429-52. '" One looks in vain for it in The Baron J. de Baye, The Industrial Arts of the Anglo-Saxons (1893); G. Baldwin Brown, The Arts in Early England, III-IV (1915); and Nils Aberg, The Anglo-Saxons in England (1926), to mention just a few obvious reference books of pre-1934 vintage. 131 S.C. HAWKES the Kent Archaeological Society, held in the Deanery at Rochester, on the 3rd of August 1859," exactly one hundred and twenty-five years ago. According to the report of the Honorary Secretary at the time: 'Lord Amherst exhibited his splendid gold Saxon brooch, found at Sarre, in Thanet, in 1843.' After giving a brief description of the brooch, he recorded what must have been the sensation of the day. 'Lord Amherst having accidentally dropped this fibula, the jar of the fall displaced the hinder part, when it was discovered that the interior was filled with a substance resembling, if not identical with, plaster of Paris.' This titbit of information can scarcely have compensated Lord Amherst for the embarrassment of such an accident and the damage to his brooch. Luckily, F.W. Fairholt had published a description and a coloured engraving of both front and back of the brooch some thirteen years before its mishap and not long after its discovery.12 Until 1934, this was the fullest source of reference. Though neither the description nor the illustration is as accurate as one could wish, they are useful in that they indicate the condition of the brooch before it was dropped and enable one to evaluate the damage done in 1859. Fairholt's engraving is reproduced here in monochrome (Plate II) to compare with the brooch as it is today (Plate I). It will be seen at once that by leaving white all cells not containing garnets or green glass, he was greatly simplifying, indeed falsifying the complexity of the polychrome inlays on the Amherst Brooch, and from his description he betrays that his eye for detail, despite his renown for draughtsmanship, was less good than Leeds's. Here, for the record, however, is Fairholt's description of the brooch in 1846. 'By the aid of Mr. Rolfe, of Sandwich, I am enabled to illustrate this paper more fully; that gentleman having obtained for my inspection a magnificent specimen discovered in 1843, at Sarr, in the Isle of Thanet, and which I have engraved the full size of the original, and coloured in facsimile as a frontispiece to this volume. The shell is of gold; the face of the fibula being divided into four compartments by concentric circles, which are subdivided into various and differently formed cells. Some of these cells are filled with a triple range of ornaments, constructed of gold wire twisted like cord, or arranged circularly; others contain a chalky substance which appears to be decomposed pearl, while other are fitted with thin slices of garnet, beneath which is placed gold foil, granulated with intersecting lines to "Arch. Cant., ii (1859), p. xiii. 12 Trans. Brit. Arch. Assoc., Third Annual Congress (Gloucester, 1846), 1847, 87-8, PI. I. 132 PLATE I (Photo: R.L. Wilkins, F.S.A.) Colour Photograph of the Amherst Brooch in 1984. Scale 1:1. THE AMHERST BROOCH PLATE II - * W . » " :?. Reproduction of F.W. Fairholt's Engraving of the Amherst Brooch in 1846 (After Transactions of the British Archaeological Association, Gloucester, 1846 (1847), frontispiece.) 133 S.C. HAWKES give brilliancy to the setting; pale blue turquoise is fitted into the cross. The outer edge of the fibula resembles that described by Douglas as a vermicular gold chain delicately milled in notches. The general arrangement of this ornament is cruciform, in accordance with the ordinary style of these relics of our Saxon forefathers; but it is not so strikingly visible as many other specimens. The reverse exhibits no trace of ornament except upon the head of the pin, where a single piece of garnet is placed. The stem of the pin has been broken and lost; it has been gilt, of which traces remain. The plate, of discoloured silver, which forms the base of the fibula is here seen, and the narrow rim of the gold shell which is secured on it and forms its face. The finished excellence of the workmanship of this jewel could not be surpassed by any artificer of the present times.' Thus Fairholt. From his engraving, it is clear that the present backplate is not a fake as Kendrick supposed, but the original 'discoloured' metal base on- which the twin lugs and broken remains of the catch-plate, which secured the pin, still survive. The pin itself, with its garnet-set back is missing: no doubt lost in the commotion after Lord Amherst dropped the brooch and people crowded round being helpful or curious. Probably, it was trampled underfoot. The central boss on the front certainly looks squashed, as if someone had stepped on it: it now overlaps the flanking quartet of little steppattern garnets which appeared less obscured in Fairholt's drawing. The whole central setting must have come adrift from the brooch and been replaced, because originally it was aligned with great care as part of what Fairhold interpreted as the 'cruciform' arrangement of cells, and certainly contributed to the overall symmetry of the design which is such a very marked characteristic of this brooch. Now the whole central setting is askew and loose in its socket. The other major damage, if Fairholt's accuracy is to be trusted, is the loss of many garnets in the outer zone of cloisonne work and also in the next zone with the panels of gold filigree. One side of the brooch is particularly affected, and here, too, the rim is dented as if the brooch landed sideways on the Deanery floor at Rochester, from which, doubtless, it would have been difficult to recover the loosened and broken garnets. It says much for the strong construction of this composite brooch that it did not disintegrate altogether. That it looks as good as it still does must be due to the skill of the craftsman employed by Lord Amherst to repair it; and to his discretion, for he did not attempt to make good with modern additions. Damaged though it is, the Amherst Brooch has been left with its integrity intact. It measures 6 cm. in diameter and its gold rim is between 4 and 5 mm. deep. The back-plate is unusual in not being made of silver but of some lead-rich alloy resembling pewter. The pin-fastenings are the 134 THE AMHERST BROOCH originals, and between the twin lugs the metal is gouged away where the grooved back of the original pin must have rubbed against it each time the brooch was fastened or unfastened. It is a heavy ornament, and it is just possible to make out a rivet-hole, to the right of the pin, which will have been made by the loop, normal on such composite brooches,13 which served to secure a safety chain. The gold rim, which binds the back-plate to the decorated gold front plate, and holds the whole construction together, is not straight like that on the Monkton Brooch but curved as on the Dover Brooch:14 around the top, it is decorated with an exquisitely wrought double-plait of twisted filigree wires between borders of simple beaded wires, while below it bears simple triple grooving. There is no sign that any of the principal cells on the front plate had been riveted to the back plate, so one must assume that on this brooch, as on other early gold composite brooches, the cell-work was simply soldered to the gold front plate.15 This front plate is normally hidden, but on the Amherst Brooch it may be glimpsed at the base of some of the empty cells. The white plaster exposed to view when the back fell off in 1859 will have been the normal packing used to cement front and back plates together, and to give weight and rigidity to the whole elaborate construction. So far as I know, only one composite brooch has been properly analysed in a laboratory, the one found at Sarre in I860,16 which was damaged enough to permit searching examination.17 In that, the cement packing was found to consist of calcite reinforced by tin ( dioxide around the rim, but obviously the composition must have varied according to the custom and resources of individual craftsmen. In describing the front of the Amherst Brooch it must always be borne in mind that the craftsman who designed it had a very precise and orderly sense of lay-out and colour distribution, and that his technical skill in space-control, placing of cloisons, and gem-cutting was first-class. The Kingston Brooch may be regarded by some as the great masterpiece of Kentish jewellery in the early seventh century, but though it is more elaborate than the Amherst Brooch, and its 13 See the Kingston Brooch, op. cit. in note 1, and the Monkton Brooch, op. cit. in note 8. 14 S.E. Rigold and Leslie Webster, 'Three Anglo-Saxon Disc Brooches', Arch. Cant., lxxxv (1970), fig. 5. 15/6id., 13; Hawkes, op. cit. (1975), 253. 16 C. Roach Smith, 'On Anglo-Saxon Remains discovered recently in various Places in Kent,' Arch. Cant., iii (1860), 41, Pi. Ill; Jessup, op. cit. (1950), PI. XXVII; Avent, op. cit. (1975), no. 177, p. 47, PI. 66. " A.E. Werner and Margaret Sax, 'Report on scientific Examination of an Anglo-Saxon Brooch from Sarre,' British Museum Research Laboratory, 14th October, 1967. Their results are summarized in Avent, op. cit. (1975), 19-20. 135 S.C. HAWKES filigree-work is better and more imaginative, the doisonne-work on the Amherst Brooch is decidedly crisper and technically better. It seems very likely that on both brooches several hands were at work. The man who put the filigree around the rim of the Amherst Brooch could surely have made a better job of the filigree panels on the front plate than the decidedly sloppy performance which confronts us today. Maybe Jessup was right in detecting an 'apprentice hand' at work here. As to the cloisonne-work it was virtually flawless: even the present damaged state of the Amherst Brooch allows one to say that with confidence. Proceeding from the central setting outwards to the rim, Fairholt's engraving enables us to see how precisely everything was aligned. It is to be hoped that the Ashmolean Museum's conservation laboratory may one day reposition the elements in the central roundel to restore to the brooch at least some of its original glory. As to the good fit the jeweller achieved between cloisons and garnets, that is still visible despite the severe jolting it underwent when it hit the Rochester Deanery floor all those years ago. In describing the concentric rings of ornament it will be simplest to begin with the roundel, 21 mm. in diameter, at the centre of the brooch. Its contents are themselves composite. Right in the middle, and set high, probably supported on its own tubular cell and soldered in place, is a separately constructed element in the form of a domed boss with plaited-filigree trim around the base of a plain collar designed to retain curved settings. All its cells are now completely empty, but it seems reasonable to assume that, like the others on this brooch, the quatrefoil cell at its centre was filled with white calcite paste, probably with a little gold ringlet at centre. The four curved settings, separated by straight cloisons radiating from the lobes of the quatrefoil, are most likely to have been of garnet.18 Encircling the boss is a ring divided by four straight cloisons terminating in tiny garnet-set stepped cells which abut the boss. The big curved spaces between were set with a white substance, which appears under the microscope to have been shell. Without analysis, it is impossible to say whether this was merely cuttle-fish bone or the exotic shell imported from the Indian Ocean via Byzantium, which was used in 18 Cloisonne^ bosses set with garnets are relatively rare, but do occur on the finer pieces of contemporary Kentish jewellery, on a good plated brooch from Faversham (Avent, op. cit., 1975, PI. 50); on the Kingston Brooch itself; and on the newly discovered pendant from Canterbury (Paul Bennett, 'Old Westgate Court Farm site, London Road', Arch. Cant., xcviii (1982), 221 ff., fig, 2; T.W.T. Tatton-Brown, 'A major Find in Canterbury', Kent Arch. Newsletter, 2 (Autumn 1982), p. 1 with plate; 'The Canterbury Pendant', Current Archaeology, 83 (1982), 377 with plate.). 136 THE AMHERST BROOCH exactly the same manner on the Kingston Brooch.19 But in view of the high quality of our brooch, and the general similarity of the central setting to that on the Kingston Brooch, the likelihood is that it was the more prestigious imported shell. The next zone of ornament is better preserved. Eight acute-angled triangles radiate outwards, and their fillings alternate; those pointing to the round settings in the next zone being filled with translucent pale olive glass, apparently without gold foil backing, and those aligned with the 'flask-shaped' settings being filled flush with opaque jade green glass. Between them, base outwards, are an equal number of ambitious triple-step-pattern cells, two of which retain a smooth fill of white calcite paste which seems to have been the intended final inlay. The remaining stepped cells are either empty, in one case, or contain a grey paste which, under the microscope, looks very lumpy. Very probably, it is a backing paste, of one of the very varied groups discussed by Arrhenius.20 It would appear that samples have been taken from some of these cells, but the Ashmolean has no record of it, and the Amherst Brooch does not figure in Arrhenius's lists. There is no trace of any gold foil in the stepped cells in this zone, so, in default of evidence to the contrary, it seems likely that all were once filled with smooth white calcite. The cells intervening between triangles and steps are filled with flat garnets laid over gold foils, all of which on this brooch appear to be boxed foils.21 The next zone contains four roundels enclosing quatrefoil cells, alternating with 'flask-shaped' cells enclosing stepped cells with a long cloison projecting from the upper of the three steps. In one of the quatrefoil cells there is well-preserved smooth white calcite, with the impression of what must have been a gold ring in the centre. At 19 CF. Battiscombe, The Relics of Saint Cuthbert (1956), 310, 542 ff. 20 Birgit Arrhenius, Granatschmuck und Gemmen aus Nordische Funden desfruhen Mittelalters (Acta Universitatis Stockholmensis: Studies in North-European Archaeology, Series B. 1971), 78 ff., 216-221. Her results are summarized by Avent, op. cit. (1975), 18. The technique used by Kentish jewellers seems to have been to seal the garnets and their backing foils in place by a hot-wax process, incorporating calcite, quartz and other substances, but whether this would work with other types of setting remains to be discovered. Not enough analysis has been made of this, as of other, aspects of Anglo-Saxon jewellery. 21 Richard Avent and David Leigh, 'A Study of cross-hatched gold Foils in Anglo-Saxon Jewellery', Med. Arch., xxi (1977), 1-46. The Amherst Brooch was one of the pieces of jewellery examined, no, 137, see Table I, and the foils have 3 x 3 little squares in each box, PI. IC. The British Museum's Research Laboratory's research into the manufacture of such foils, on brass dies produced with the aid of a template, will be published in Anglo-Saxon Studies in Archaeology and History, 4 (1984), forthcoming. 137 S.C. HAWKES least one other, and probably all these cells, had the same kind of white filling. They are each surrounded by garnets. The outer settings in the 'flask-shaped' cells are also garnets, but the stepped cells within them are either empty (two cases) or just have grey backing paste. One wonders whether the jeweller had used white again. Already we have seen him using more white than is at all usual on jewellery of this period: clearly, he liked the effect of contrasting brilliant matt white with the rich red gleam of garnet and the bright yellow glitter of gold. The outermost zone of cloisonne-work has sixteen of his steppattern cells, alternately base in and base outwards, separated by long diagonal cloisons which make truncated obtuse-angled triangles. Everything is set with garnets except the stepped cells, and here again we face the problem that the contents of the stepped cells have barely survived. Of the eight such cells placed base inwards, three still retain gold foils which had evidently backed inlays of yellow glass. They alternate with cells without foils, mostly just filled with grey backing paste, but one of these alternate cells, next to one with a yellow setting, contains the remains of what appears to be semi-decomposed Prussian blue opaque glass. Assuming that his design here will have been as regular as elsewhere on this brooch, we can be safe in suggesting that the inner ring of stepped cells in this outer zone will have been set alternately with bright foil-enhanced yellow and deep blue glass. Only one of the stepped cells with its base outward has any trace of anything except backing paste, and this one, opposite one of the green triangles in the inner zone, again has the remains of a deep blue inlay. To have alternated the colours in this ring of cells would have created visual chaos, so I suggest that the jeweller may have used the blue all around the outer rim of the brooch, to finish it off with a dark border that would contrast with the light effect created by all the use of white radiating out from the centre. Finally, to return to the panels of filigree-work, between the circular and 'flask-shaped' cloisonne settings in the central zone of ornament, they consist of sheet gold inlays laid on a layer of backing paste, each with a border of beaded gold wire and three rows of filigree ornament. The rows are separated by beaded wires, and on each plate each is different. The central one normally consists of four (once three) granules enclosed by beaded wire ringlets; the others tend to comprise a row of running spirals and another of peltate motifs, all in beaded wire. No two panels are exactly alike, and there is a fair amount of botching to make up for bad spacing. Altogether the filigree panels on the Amherst Brooch present a poor contrast, not only with the masterly Style II ornament on the Kingston Brooch but also with the minutely accurate filigree-work on the composite 138 THE AMHERST BROOCH brooch from Priory Hill, Dover.22 Actually, there is a very close parallel to the filigree on the Amherst Brooch on another composite brooch from Gilton grave 42,23 which could be by the same hand, though the brooch itself is very different. That such a very precise and beautifully executed piece of cloisonne-work as the Amherst Brooch could be finished off with second-rate as well as first-rate filigree suggests that maybe each of these techniques was a specialism, and the man who constructed the brooch and set the garnet, glass and shell inlays, did not necessarily see each piece of jewellery he made through to completion. We know all too little of workshop organisation in this period. But certainly a workshop specialising in luxury jewellery such as composite brooches, and which disposed of precious imports such as gold, garnets and exotic shell, will have been rigidly controlled and highly organised, probably under royal patronage. Such a workshop may have employed master craftsmen in all the required techniques, each with their apprentices, and not all need have been equally skilled. Though the filigree on the Amherst Brooch can be paralleled on another such brooch, perhaps emanating from the same workshop, the cloisonne'-work is quite another matter. It differs from that on the rest of the known corpus in many respects, not just in technical superiority but in the use of special cell-forms and combinations not known elsewhere. For example, though the step-pattern cell is a commonplace of Kentish cloisonne jewellery, almost invariably it is simpler, with just two steps aside, even on the Kingston Brooch, the exception being on that composite brooch from Gilton grave 42, already mentioned because it has similar filigree. But there the similarities end, for though the maker of the Gilton brooch was adventurous in his use of cloisonne, the effect was messy and indecisive, not adjectives one would apply to the Amherst Brooch. Everything about it is original and different: the stepped cells are combined with triangles, not the usual semicircles; the 'flask-shaped' setting is a novel adaptation of a much looser arrangement seen on another brooch, a plated brooch, from Gilton grave XV, which also has a quatrefoil cell at its centre.24 The quatrefoil cell is rare on jewellery after the early sixth century, especially when inlaid with white, and the device of the inlaid gold ringlet is another archaism.25 Either the maker of the Amherst Brooch was a very old man, in which case he had a remarkably steady hand, or he had access to 22 Avent, op. cit. (1975), no. 174, p. 46, PI. 64. 23 Faussett, op. cit. (1856), PI. II, 4; Avent, op. cit. (1975), no. 175, p. 46, PI. 65. 2Ubid., no. 153, p. 38 ff., PI. 52. 25 Ibid., Pis. 3-5, on brooches dating back to the second quarter of the sixth century. 139 S.C. HAWKES antique pieces of jewellery and was experimenting with old and to him unusual ornamental devices. He was certainly experimenting with colour. As Leeds remarked, no other piece of jewellery of its period was quite so adventurous in that regard. For posterity's sake, it has been sheer bad luck that the substances he used did not always survive the rigours of its subsequent history. The surfaces of the surviving settings of white calcite make it clear that the Amherst Brooch has been subjected to a rigorous scrubbing, certainly before its acquisition by the Ashmolean, and probably before Fairholt saw it in 1846, to clean it up for Lord Amherst's cabinet. What damage that may have done to settings more fragile than garnet can well be imagined. We are lucky the brooch has come through all its vicissitudes as well as it has. ASSOCIATIONS AND DATING According to Fairholt, the Amherst Brooch was found with a bronze bowl, which was evidently extant in 1846, presumably in the same collection as the brooch, because Rolfe secured it at the same time and Fairholt both described and drew it. What has happened to it since is not known: it was not sold with the Amherst Brooch in 1934, and Peter Richards kindly tells me he has not been able to trace its present whereabouts.26 However, Fairholt's description of it; 'about twelve inches in diameter, the ornaments of the foot are perforated; a double raised circle is upon the bottom within; and it has two handles which move freely on sockets inserted on each side.'; his comparing it with a previously published parallel from Wingham; and, above all, his own published engraving, reproduced here at an enlarged scale (Fig. 1), make it clear that the vessel in question was one of the now familiar Coptic, or, if we follow Richards, Byzantine bowls of Group Bl.27 These cast bowls, of East Mediterranean origin, have been found in a number of rich early seventh-century graves in Kent, for example that containing the composite brooch found at Sarre in I860.28 They were imported as just one of the larger, tangible ingredients in a long-distance trade from Byzantium to the western 26 Peter Richards has made a study of Byzantine bronze vessels in England as part of his Cambridge doctoral thesis. 27 Joachim Werner, 'Zwei gegossene koptische Bronzeflaschen aus Salona', Antidown Michael Abramic, i (Split 1954-7), 115-28; 'Fernhandel und Naturalwirtschaft im ostlichen Merowingerreich', 42. Bericht der Romisch-Germanischen Kommission, 1961, 310-12. ™Arch. Cant., vi (1860), PI. IV. 140 THE AMHERST BROOCH Fig. 1. Reproduction of the Amherst Bowl, enlarged from F.W. Fairholt's engraving. Scale c. 1:4. (After Transactions of the British Archaeological Association, Gloucester, 1846 (1847), fig. on p. 88.) Germanic kingdoms, whose minor ingredients included amethyst beads, porphyri, cowrie shells from the Indian Ocean, and spices. They were shipped from Constantinople, or Alexandria (or wherever), to Italy; thence taken, presumably on pack animals, over the Alps to southern Germany, where they are found amongst the Alamanni; thence shipped again, down the Rhine through Austrasian Frankish territories, and so on to the utmost ports of call, amongst the Frisians and the English. Kent, the dominant trading power amongst the English kingdoms at the time, retained the majority of those it acquired, notably the Bl bowls.29 The date, of all but a few rare sixth-century imports,30 appears to have been a short period from the very end of the sixth century through at most the first three decades of the seventh. This dating is provided not so much by the find-combinations in graves, because some bowls were kept a long while before burial, but by the Imperial coins which formed part of this trade. Light-weight gold solidi were minted specially for trade with the barbarians, and in the West the last issues to be represented in any significant number were those of Heraclius and Heraclius Constantine (613-29).31 Thus the import of the Bl bowls had probably ceased by c. 625. Fairholt's drawing suggests that the bowl buried with the Amherst Brooch was in better condition than that 29 Sonia Chadwick Hawkes, 'Anglo-Saxon Kent c. 425-725', in (Ed.) Peter E. Leach Archaeology in Kent to AD 1500, 1982, 76, fig. 33. 30 C.J. Arnold, The Anglo-Saxon Cemeteries of the Isle of Wight (1982), 59-60, figs. 7, 26, iii, 45, i. 31 Howard L. Adelson, Light Weight Solidi and Byzantine Trade during the sixth and seventh Centuries (Numismatic Notes and Monographs, New York, 1957); Sonia Chadwick Hawkes et. al., 'X-Ray fluorescent Analysis of some Dark Age Coins and Jewellery', Archaeometry, 9 (1966), 99. 141 S.C. HAWKES found with the Sarre Brooch in 1860, and this accords well with the internal evidence for the dates of these two brooches. Only very few of the Kentish composite brooches have been found with useful datable associations, so such external evidence has to be considered together with information afforded by the brooches themselves. Since in number they are so few, stylistic considerations are perhaps less informative than technical matters such as the quality of the gold employed and the size of the garnets. It has long been established that the gold supply in England depended on imported coin, and as the supply of high quality Byzantine gold dried up, and Frankish and, eventually, English moneyers had to eke out their dwindling resources, by the use of alloys increasingly rich in silver, so the jewellers, too, had to use increasingly poor-quality gold alloys as the seventh century progressed.321 discussed this in some detail when publishing the composite brooch from Monkton, with its pale alloys only 55 per cent and 41 per cent fine, which is likely to have been made as late as the 640s.33 The present subject, the Amherst Brooch is quite another matter, however. Its gold has been analysed34 and found to be 83 per cent fine, indicating that it was made before the gold famine set in. According to the system of calculation proposed by David Brown, based on the gold standard of Merovingian coinage produced outside the Provencal mints, the maker of the Amherst Brooch will have been using a gold-melt composed of coins of the period 596-613.35 The brooch from Gilton grave 42, discussed above as in some respects similar, has a similar gold content of 85 per cent,36 the Kingston Brooch 83.5 per cent and the Dover Brooch 81 per cent.37 These have all been considered early on stylistic grounds, and the results of the gold analyses confirm this very nicely. Probably, all were made during the first two decades of the seventh century, and the Amherst Brooch, with features in common with the late sixthcentury plated brooches, may well have been the earliest of them all. We now have an analysis result for the composite brooch found at Sarre in I860,58 whose gold at only 69 per cent fine places it in the next generation, composed of coin-gold of the period 613-29. It was found K Hawkes et. al., op. cit. (1966); P.D.C. Brown and F. Schweizer, 'X-Ray fluorescent Analysis of Anglo-Saxon jewellery'. Archaeometry, 15 (1973), 175-92. "Hawkes, op. cit. (1975). " Brown and Schweizer. op. cit. (1973), 180, AM 27. »Ibid., 185. v- Hawkes et. al., op. cit. (1966). 106-7, L12. •" M.J. Hughes et. al., 'Report on the Analysis of the Gold from the Sutton Hoo Jewellery and some comparative Material', in Rupert Bruce-Mitford, The Sutton Hoo Ship Burial, ii (1978), 618-25. •* Idem. 142 THE AMHERST BROOCH with a necklace of coins from Provencal mints, which cannot have been assembled before c. 615,39 and were then equipped with loops and worn long enough to become abraded. I have previously suggested that the Sarre composite brooch, which shows little sign of wear, may not have been made before c. 625,40 and the analysis results certainly bear me out in thinking it a piece transitional between the early all-gold masterpieces and the later bronze-cell composite brooches made towards the middle of the seventh century.41 The relevance of this to the Amherst Brooch will become clear during the discussion of provenance (p. 147). Meanwhile, there is just one further pointer to the early date of the Amherst Brooch which deserves a brief mention. Its maker disposed of large slabs of garnet, relatively speaking: he was not, like the makers of later composite brooches, such as that from Monkton, having to re-use old settings, or cut down second-hand materials into tiny rectangular settings as on the latest composite brooches.42 In her forthcoming paper on flat-cut garnets, Mavis Bimson suggests that such garnets were exported ready polished from the Byzantine Empire as part of the trade-goods for the western Germans, and that at some point in the earlier seventh century the supply failed, so that our Anglo-Saxon jewellers were reduced to cannibalizing and adapting existing materials.43 It seems virtually certain that the garnets travelled with the light-weight Imperial gold solidi, Byzantine bowls, and the other trade-goods mentioned above, and that new supplies of garnets, like bowls, failed to reach Kent after c. 625. This would certainly fit with what we see on the now well-dated composite disc brooches. The Amherst Brooch was made when flat garnets were still plentiful and economies not yet necessary. PROVENANCE All that Fairholt recorded about the finding of the brooch, except that it was found at Sarre in 1843, was: 'The grave in which this fibula was discovered was broken into accidentally by some workmen in digging for gravel, and no record of its contents kept. The only article •" Arch. Cant., iii (1860), PI. II; Jessup (1950), PI. XXVII; Rigold and Webster, op. cit. (1970), 7, note 19. 4,1 Hawkes et al., op. cit. (1966), 107. 41 Hawkes op. cit. (1975), 252-3. 42 Avent, op. cit. (1975), no. 171, PL 61; nos. 181-5, Pis. 71-75. 43 Mavis Bimson, 'Garnet Cutting in the Dark Ages', Anglo-Saxon Studies, 4, forthcoming. 143 S.C. HAWKES known to have been found with it, was the bronze bowl here figured. . .'. The reference to gravel in connection with the Anglo- Saxon burial at Sarre, where the cemetery is on chalk, sounds a trifle suspect, but is understandable if the information came at secondhand from William Rolfe of Sandwich, who seems to have been acting as intermediary between Fairholt and the owner. It is somewhat curious, however, that the owner is never once mentioned by Fairholt. Even more curious is the fact that no writer about the Amherst Brooch until now has paid attention to a couple of entries in The Archaeological Journal, iv (1847). On p. 164 we learn that, on 7th May, 1847, 'Lord Holmesdale exhibited a large metal dish of Roman work, cast and then finished on the lathe; and a most beautiful gold fibula . . . found in the Isle of Thanet in 1841. It was set with pieces of coloured glass, tastefully arranged.' On pp. 253-4, the archaeological correspondent, J.Y. Akerman,44 who is writing about Anglo-Saxon composite brooches, continues: 'With the exception of that already referred to from the collection of Mr. Faussett, the next largest specimen is in the cabinet of the Right Hon. Lord Holmesdale. It was purchased together with a bronze vessel, in 1841, by Mr. Rowland Freeman, a medical man, at Minster, Thanet, from a labourer who had discovered them both a few days before, about four feet deep in the chalk. The spot where they were found is described in Lewis's History of Thanet (p. 48) as an ancient burying ground about three rods east of the town, and there are many barrows near it hitherto unexplored. In turning up the soil human bones are invariably found for some distance, and a few years back, a stone coffin was dug up in the old burying ground, and is now used as a water trough in a farm-yard close by. The person who found them stated that they were in the same grave, and not many inches apart, but the fibula was not in the brass vessel: there was something attached to the ornament having the appearance of a small chain of some material into which gold had been interwoven, but as soon as it was touched it pulverized.' This has all the authority of a first-hand account, which 44 In the Archaeological Journal, the author of the very full and competent sections devoted to archaeological finds is not named. Audrey Meaney, A Gazetteer of early Anglo-Saxon Burial Sites (1964), 129-30, under Minster, refers to the anonymous correspondent as Akerman. Doubtless, she checked with the records of the Royal Archaeological Institute to verify her facts. It would make sense, however. Until his appointment in 1848 to the post of Resident Secretary and Librarian to the Society of Antiquaries of London, Akerman had been dependent for his living on other secretarial employments (Joan Evans, A History of the Society of Antiquaries (1956), 255-6), and Albert Way, Director of the Society of Antiquaries and chief sponsor of the new Archaeological Institute, must have been a useful patron and source of fees. 144 THE AMHERST BROOCH Fairholt's does not, with Akerman quoting directly from a reliable, archaeologically educated informant, who must surely have been Rowland Freeman himself. He, or rather the labourer who dug the grave, contributes the little extra detail that there had been what reads very like a piece of gold-brocaded braid, perhaps a safety-chain or a dress-border, which would be unusual in a woman's grave of the early seventh century.45 So, why has this admirable account received virtually no attention in the literature about the Amherst Brooch? The explanation, probably, is that the Amherst Brooch became to some extent a casualty of the rivalry between the two societies, The British Archaeological Association (whose adherents included Charles Roach Smith, William Rolfe and F.W. Fairholt, the engraver and their protege) and the Archaeological Institute (whose adherents included Albert Way and his protege J.Y. Akerman).46 The timing of Lord Holmesdale's exhibition of the brooch and bowl to the Archaeological Institute in 1847, and the publication date of Akerman's account of it, probably precluded any updating of Fairholt's text and any cross-referencing by Akerman, even had either wished any contact with the other in the matter. As we have noted above (p. 131), the brooch itself disappeared from view between 1859 and 1934, and one suspects that most people who have written about it in the last fifty years (if they knew of Akerman's account at all) must have assumed that there had been two sets of brooch and bowl; Lord Holmesdale's found at Minster in 1841 and since lost; and Lord Amherst's found at Sarre in 1843. Audrey Meaney certainly adopted this view when compiling her Gazetteer, and actually has a map reference in Minster parish for the 1841 find. But, of course, there can only have been one brooch and one bowl. It takes only a few moments with Burke's Peerage to find out that Holmesdale is a subsidiary title of the Earl of Amherst, and that the Second Earl, who exhibited the Amherst Brooch to the Kent Archaeological Society in 1859, had succeeded to the title only two years before. Prior to that he had been Viscount Holmesdale, and it was as such that he exhibited the same brooch to the Archaeological 45 For gold braids, see Elisabeth Crowfoot and Sonia Chadwick Hawkes, 'Early Anglo-Saxon gold Braids', Med. Arch., xi (1967), 42-86; Elisabeth Crowfoot, 'Early Anglo-Saxon gold Braids: Addenda and Corrigenda', Med. Arch., xiii (1969), 209-10; D.R.J. Perkins and Sonia Chadwick Hawkes, forthcoming. Most, from women's graves, date from the first half of the sixth century. 46 For the account of the turbulent history of the birth of the Royal Archaeological Institute and its separation from the British Archaeological Association, see Joan Evans, 'The Royal Archaeological Institute: a Retrospect', Arch. Journ., cvi (1949), 1-11. 145 S.C. HAWKES Institute in May 1847. Fairholt's silence on the question of ownership in 1846, unthinkable in those days where a title was concerned, suggests that Lord Holmesdale may not have acquired brooch and bowl until after all the publicity they were given at the Gloucester meeting of the British Archaeological Association in that year, and that they were new treasures to him when he exhibited them to the rival society early in the following year. The person from whom William Rolfe had obtained them for Fairholt to engrave was, presumably, Rowland Freeman. Conceivably, Rolfe had bought them himself, perhaps in 1843. Certainly, there must have been some early complexity in the history of the disposal of the finds which led Fairholt to get the date wrong. These simple facts established, we come now to the question whether the Amherst Brooch was found at Sarre or Minster, and for that it is necessary to examine most carefully the information given in volume 4 of the Archaeological Journal. Here I could not have coped without risk of grave error but for generous help given by Mrs. Patricia Hopkins, who is working on a history of Minster, and Mr. D.R.J. Perkins, who is chief archaeologist for the Isle of Thanet Archaeological Unit. According to Mrs. Hopkins,47 'Rowland Freeman was appointed Medical Officer for the Isle of Thanet Union at the first meeting of the Guardians in 1835 (KCC records G/Thl),' and was thus in charge of the hospital and workhouse at Minster. 'The census return of 1841 gives his age as 60', but he did not retire for some time after this. So, in 1841, he will have been living at Minster, and if one examines Akerman's phraseology and interprets it in that sense; 'purchased . . . by Mr. Rowland Freeman, a medical man, at Minster, Thanet . . .'it would dispose of the Minster provenance for the brooch and bowl. The 'labourer who had discovered them both a few days before' could have been working elsewhere. Indeed, the account of the find goes on to say that 'The spot where they were found is described in Lewis's History of Thanet (p. 48) as an ancient burying ground about three rods east of the town.' This refers to Sarre not Minster, but as Mr. Perkins rightly points out, in both the first and second editions of Lewis (the page reference quoted is to the second) what is described is not a burial-ground at all but the site of the ruined church of St. Giles, which had already fallen down by the time of Lewis's first edition in 1723. 'Upon the hill, to the eastward of the town about 30 rods, on the left of the great road leading from Sarre to Monkton . . . stood the Parish Church.' According to Perkins, this distance (misprinted in Arch. Journ., iv), about 151 m., Letter, dated 12.9.1984. 146 THE AMHERST BROOCH is 'far out', the site of the church having been established by excavation '300 m. east of the Windmill, on the edge of Perkins' chalk pit [no relation] with unexcavated seventh-century graves within 20 m.'48 The windmill was the site of the discovery of another composite brooch and bowl in 1860, cf. above (pp. 141-2), and the ground to the east up to the chalk-pit was excavated by John Brent in 1863-64 and shown to be the site of a large Anglo-Saxon cemetery.49 In 1863, Brent wrote: 'Although as will be seen, it has proved to be one of our richest and most interesting Saxon burying-places, it appears to have escaped the notice of all Kentish antiquaries. . . . Certain Saxon relics, however, were reported to have been found here some years since, such as a bronze stoup, a drinking glass, and a fine fibula.'50 He gives no references and seems unaware of any of the recent publications of the Amherst Brooch and bowl, so could have been relying on local information. He may even have been referring to objects not from the Amherst collection at all: the British Museum has objects from the Durden collection which possibly came from Sarre c. 1858.51 These include the famous quoit brooch and a bronze-bound wooden bucket, more likely to be termed a stoup than a bronze bowl would have been. As to their find-place, by Brent's own admission there were two chalk-pits in the cemetery-area before he began his own operations. In the light of all this, it seems very likely that the Amherst Brooch and Byzantine bowl came from Sarre, and that we should start referring to a 'Sarre 1841' grave to partner that of 'Sarre I860'. There is a generation's gap between the dates of the two brooches, so the burials could have been of successive women in the same rich family. The family concerned, pretty certainly, will have been that of the king's reeve, who commanded the garrison which controlled the royal port at Sarre in the earlier seventh century.52 So, how would Rowland Freeman come to acquire antiquities from Sarre in 1841? Mrs. Hopkins remarks in her letter that 'Rowland Freeman may have been visiting Sarre in connection with his work for the Workhouse. Sarre came under the Western division for which 48 Letter, dated 21.8.1984. 49 John Brent, 'Account of the Society's Researches in the Saxon Cemetery at Sarre', Arch. Cant., v (1862-3), 305-22; vi (1864-5), 157-85; vii (1868), 307-21. 50 Brent, op. cit. (1862-3), 308. 51 Proc. Soc. Ant. Lond., 2nd s., xiv (1891-3), 314-5. It is the collection reputedly from Crundale Down, which, due to confusion in Durden's notes, seems to have been wrongly attributed, and which most probably came from Sarre. 52 Sonia Chadwick Hawkes, 'Early Anglo-Saxon Kent', Arch. Journ., cxxvi (1969), 191-2; Hawkes, op. cit. (1982), 76. 147 S.C. HAWKES Freeman was responsible.' Freeman 'had been interested in antiquarian matters for many years', and had in youth intended to publish a revised edition of Lewis's History of Thanet.53 From the description of Sarre published in Arch. Journ., iv (1847), he had certainly kept his eyes open during his visits, while, presumably, the chalk-pits were being opened, and had Brent bothered to do his homework he would have found that at least one earlier Kentish antiquary had noticed the cemetery at Sarre and recognised its potential. What Akerman wrote can only have come from Rowland Freeman himself: 'there are many barrows near it hitherto unexplored. In turning up the soil human bones are invariably found for some distance, and a few years back, a stone coffin was dug up in the old burying ground, and is now being used as a water trough in a farm-yard close by'. This information could well derive from notes that Freeman had made in his own copy of Lewis's History, whether the second edition, or his own proof copy of the aborted edition of 1809. If such a thing existed, of course, the pagination would in all likelihood have changed to accommodate the enlarged amount of information, and a misunderstanding could have arisen over the site, if Rowland Freeman and Akerman were using substantially different editions, or the former was quoting from annotations unknown to the latter. Certainly, though the case for Sarre is very neat, the question remains why the Ordnance Survey gave Rowland Freeman's finds a grid reference in Minster parish, and why Charles Roach Smith consistently referred to them, even when publishing the similar finds made at Sarre in 1860, as coming from Minster. He did so at least three times.54 This cannot have arisen in error from a superficial 531 am indebted to Mr. D.R.J. Perkins for the following information about this. 'A further item of interest is the possible existence of a third edition known as The Huggett Copy by Rowland Freeman, published (reputedly) at Canterbury in 1809. The Rev. Roger Huggett, of Stone Farm, at Broadstairs, was a contemporary and friend of Lewis. Strongly attached to the study of local history, his own copy of Lewis's History was covered in hand-written notes and became a family heirloom. The volume was lost for a number of years, then discovered in possession of Mr. R. Freeman, of Canterbury, who, it would seem from an advertisement in the Kentish Gazette, was about to publish a new, third improved edition. But, by the exertion of Mr, Boys, the well-known Margate solicitor, it was recovered and returned to a descendant in 1810. So far as is known no third edition was ever published.' Extract from Memo attached to letter dated 1.9.1984. Mr. Perkins adds 'But there could have been a printer's proof.' 54 'Found in 1843, upon the property of the Marchioness of Conyngham, at Minster . . . a coloured engraving was presented by Mr. Fairholt . . . 1846', Inventorium Sepulchrale (1856), p. xxi; in referring to the composite brooch found at Sarre in 1860, 'It belongs in the same class as Lord Amherst's found at Minster', Arch. Cant., iii (1860), 41; and he uses similar terms in Collectanea Antiqua, vi (1868), 148. 148 THE AMHERST BROOCH reading of the report in Arch. Journ., iv (1847), for Roach Smith was a great authority on the archaeology of Kent and must have been familiar with Lewis's History of Thanet. He also kept up with new finds and was an active correspondent with everyone involved in archaeological work in the county. He was a particular friend both of William Rolfe of Sandwich, who had 'obtained' the brooch and bowl from Rowland Freeman, and of F.W. Fairholt, who engraved them for the British Archaeological Association. That Roach Smith, of all people, should tacitly have rejected the Sarre attribution accepted by everyone else in his day, and preferred Minster, is a matter of great seriousness, to be treated with respect. It implies that he had inside information, presumably from Rowland Freeman himself. In his last years, Roach Smith recorded in some detail his adventures on the occasion of his first visit to Sandwich and Thanet.55 He gives no date, but internal evidence suggests that it took place a year or so after the passing of the Municipal Corporations Act of 1835. He tells how, on the occasion of his first meeting with Rolfe 'early in my antiquarian youth and late in the autumn', he set out to walk from Richborough to Reculver, calling, on Rolfe's recommendation, on 'Mr. Freeman, a surgeon at Minster . . . about the middle of the afternoon'. From the context, it seems that Freeman presented Roach Smith with a copy of his poem about Reculver:56 he certainly tried to dissuade his young visitor (Smith would have been about 30) from 'attempting to cross the marshes from St. Nicholas at such a season, and so late in the day.' However, Roach Smith persisted: T passed by places rich in Saxon cemeteries then undiscovered: and arrived at St. Nicholas late in the afternoon. Here my troubles began.' He got lost, wet, muddy and very cold. This fascinating story tells us not only that Roach Smith met Rowland Freeman in person, with whom, indefatigable correspondent that he was, he will have kept in touch, but that he passed the sites of Saxon cemeteries in the plural. Between Minster and St. Nicholas, he can have passed very few places that were to prove rich in Saxon cemeteries during his lifetime. Apart from Sarre, of course, with the major finds from Monkton well in the future,57 the only likely candidate will have been Minster itself. It does seem, therefore, that Roach Smith had positive knowledge of a cemetery at Minster. 55 Charles Roach Smith, Retrospections, social and archaeological, i (1883). 'First visit to Richborough and Reculver', 1-4. 56 Regulbium, a poem, 1810. 57 Hawkes and Hogarth, op. cit. (1974); Perkins and Hawkes, Arch. Cant., ci (1984), forthcoming. 149 S.C. HAWKES Indeed, it comes into his list of places yielding the richest remains of Saxon 'persons of high position and affluence'.58 He probably knew about the burial in Minster churchyard, of a man with sword and seventh-century glass vessel,59 but that was an old find: he was writing about the future, and no doubt what he considered to be the find-place of the Amherst Brooch and bowl. A place like Minster, where a monastery had probably succeeded a royal residence, can have had more than one cemetery. The Ordnance Survey reference quoted by Audrey Meaney, TR 317644, points to a site 600 m. east of the church and north of the old track from Minster to Ebbsfleet. One does not know its authority, but D.R.J. Perkins, who knows the area well, tells me that it is an open site on brickearth over sand, so the labourer who found the Amherst Brooch and bowl 'four feet deep in the chalk' could not have been working there. However, he goes on to suggest that, if the reference had been wrongly calculated, when the conversion to the metric grid took place, or had been wrongly copied (and this is by no means unusual with grid references), and was out by just one figure, and had originally read TR 317654, this would bring the site into the area of a known Anglo-Saxon cemetery. This extends across the Down north of Minster, south of Dunstrete (as in the case of Monkton), from the hospital to beyond the Minster laundry, where inhumation burials were found in 1983, at TR 316654. Air photographs of the area, of what must be a very large cemetery, show a scattering of small penannular-ditched features which must surely have been Anglo- Saxon barrows of the seventh century. If the Amherst Brooch and bowl had been found at TR 317654, this will have been beside a minor road just 400 m. east of Minster Hospital where Rowland Freeman was in charge when the labourer brought them to light.60 Mrs. Hopkins records that 'buildings were erected at fairly frequent intervals when the workhouse first opened',61 and Mr. Perkins notes that 'they were digging chalk close by'. Given all the circumstances, what could be more likely than that the grave containing the Amherst Brooch was disturbed in the course of works connected with the Minster hospital and workhouse? In such a case, Rowland Freeman would naturally have been the first to be informed by the labourer concerned. His interest in antiquities must have been locally well known. However, at this long remove in time, with the documentation so 58 Arch. Cant., iii (1860), 43. 59 James Douglas, Nenia Britannica (1793), 52, 71, PI. XVII, 4. 60 Information from D.R.J. Perkins, by letter, 21.8.1984. 61 Letter, dated 12.9.1984. 150 THE AMHERST BROOCH conflicting, it really is impossible to decide between the rival claims of Sarre and Minster. I have tried to set out the case for both, I hope impartially, and now find myself honestly without an answer to the question of provenance. If the archaeologists at the time were at odds about such a fundamental matter, and that really is odd, only a minor miracle, such as the rediscovery of Rowland Freeman's copy of Lewis's History of Thanet, or the finding of a stone coffin in a farmyard at Sarre or Minster, could now resolve the question beyond all doubt. I admit that I shall go on privately thinking of it as the Minster Brooch, while publicly calling it the Amherst, but that arises from a form of homage to Charles Roach Smith. Perhaps others will succeed, where I have failed, in finding the definitive clues to its original find-place. CONCLUSION Meanwhile, I hope I have achieved what I set out to do: to bring what is perhaps Kent's finest piece of Anglo-Saxon jewellery back to the notice of the Kent Archaeological Society, in this special anniversary volume oi Archaeologia Cantiana. The Amherst Brooch has been in Oxford now for fifty years, and been taken too much for granted. It has now been taken out of the 'gold case' and dusted down a little, and in the process has started a great many lines of enquiry. Perhaps others may follow them further with better result. The Amherst Brooch is worth pursuing further. ACKNOWLEDGEMENTS I would like to thank David Brown and Arthur MacGregor, Assistant Keepers in the Ashmolean Museum, for allowing me the 'freedom' of the Amherst Brooch, and the Visitors for allowing me to publish it. The new photographs were taken in the Oxford Institute of Archaeology by Robert Wilkins, F.S.A. Patricia Hopkins and Dave Perkins have responded generously to demands on their time, and expertise in the history and archaeology of Thanet. As always, my husband Christopher Hawkes has taxed his memory and exercised his critical judgement whenever asked to do so, and it is my pleasure to state, that in his 80th year, both remain first-class. So, to him, the first informed person to see the Amherst Brooch in 1934, seventy-five years after its last public exhibition to the Kent Archaeological Society, I dedicate this essay. Without his recognition of its importance, the Amherst Brooch might have strayed further from Kent than Oxford. 151