An Anglo-Saxon Cemetery at Holborough, Kent

AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT By VERA I. EVISON THE SITE THE earliest indication of the presence of Saxon graves on Holborough HiU was given by a small earthenware pot and hone which were brought to Maidstone Museum in 1943, with the record that they were found 200 yards west of the Knob. In March, 1950, chalk excavations being carried out on Holborough HU1 by the Associated Portland Cement Manufacturing Company led to the discovery of human skeletons in a number of graves. The matter was reported to the coroner and the curator of a local museum, but no archseological excavation was undertaken. Further skeletons were disinterred in July, 1952; the Curator of Maidstone Museum, Mr. L. R. A. Grove, was informed, and he contacted the Inspectorate of Ancient Monuments, Ministry of Works. An excavation was undertaken under the auspices of the Inspectorate from July to September, and was continued during part of June and July of the following year. Mr. A. Warhurst, then Assistant Curator of Maidstone Museum, gave much valued assistance and the finds from the site are now in the possession of that Museum. The A.P.C.M. Co. throughout provided every possible kind of practical help such as labour and surveying assistance, and, in 1953, the use of a buUdozer for clearing topsoU. Work was facUitated by the friendly help and interest of the Works Manager, Mr. B. Buxton, the Assistant Manager, Mr. Morgan, and the Surveyor, Mr. R. Cook.1 About half way between Maidstone and Rochester, where the North Downs are cut by the Medway vaUey, Holborough Hill stood on the west bank of the river, nearly clear of the main ridge of the DoAvns.la Oval in shape, it sloped graduaUy on all sides except to the north, where it fell steeply, covered with trees and shrubs. In the time of the Saxons one must imagine the summit, 200 feet high, crowned by a Bronze Age barrow. About two hundred yards doAvn the slope towards the river was the conical Roman tumulus known as Holborough Knob,2 and between these two earlier monuments lay the Anglo-Saxon cemetery (Figs. 1 and 2). An ancient ridgeway passed over the hill, and along the foot of the Downs to the north ran the Pilgrims' Way. Continued chalk-quarrying operations have now almost completely removed the hill. 84 HOLBOROUGH MARSHES ROHAN TUMULUS JNGL0-&4X0N CEMETE1& P/gMSTORIC BAgRCM CEMENT r*W-'-h WOPKS HOLBOROUGH QUARRY SCALE IN FEET 0 500 1000 FIG. 1. Holborough and district. (Based on the 25-in. O.S. map.) [ face p . 84 HOLBOROUGH KENT ROMAN BARROW W A &<\U) SEARCHED 2 l IS1 PREHISTORIC BARROW.-- M\W CONTOURING IN FEET O.D. 50 0 SCALE 50 i 100 I 250 = 1 OLD QUARRY FIG. 2. Holborough Hill in 1952, showing contours. face p. 84 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT By June, 1952, excavations for chalk had proceeded untU the hUl had been cut almost exactly in half verticaUy in a line roughly southwest to north-east, with the exception of one deeper cut made just north-east of the crest and stopping short of the Roman tumulus (Figs. 1 and 2; PI. la). On the top of the south-west slope was the overgroAvn dip of a more ancient quarry. This is represented by the group of shrubs to the left of the photograph (PI. la). At this time there was still a working platform left on the quarry face some yards below ground level (PI. 16), and from here it was possible to see on the vertical face a section of two V-shaped ditches about 80 feet apart, and of one grave (No. 1). The eastern ditch was nearly on the highest point of the hill. METHOD OF EXCAVATION In the first season, investigation was by means of trenches cut through the 9 inches of topsoil to the level of the natural chalk, 2 feet Avide and 4 feet apart, laid out at right-angles to the one grave visible at the time. The area covered was to the north-east of the ditches. In this way, the other graves were encountered at right angles and were easUy recognizable by means of the darker filling in contrast to the Avhiteness of the solid chalk. The edges could then be traced, although the occurrence of clay-with-flint pockets sometimes made definition more difficult. Total excavation by hand is, of course, the only way to be sure of not missing any burials. At that time this was impracticable, and the method of paraUel-trenching was judged to be a fahly fast and safe way of discovering the limits of the cemetery and detaUs of the incidence and alignment of the graves. As the trenches progressed, all the burials encountered were lying in a Avest to east dhection, and it was therefore reasonably certain that no grave of normal length would be missed. This system, of course, is not workable if there are differently orientated graves, cremations or other features visible in the soil, but as here the graves were lying predominantly or exclusively west to east, and the site consisted of a thin sheet of topsoil rendered sterUe by incessant ploughing and a subsoU of pure white rock chalk in which any disturbance was immediately Adsible, the margin of error seemed so smaU as to be worth the risk. However, it appears that two fragmentary graves were missed, for when the area was cut away by the chalk digger in June, 1953, two lots of bones were discovered Avhich apparently belonged to graves aheady mostly destroyed. This means there Avas probably only a foot or tAvo of them left in the sharp chff edge overhanging the quarry. They Avere Grave 34 (of which parts of the lower half of the skeleton were retrieved) and Grave 39 (of which the upper part remained). In the second season the complete clearance of the topsoil by bulldozer obAdated the 85 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT uncertainty of graves being overlooked. The main risk with this means is that of damage to the contents of any grave not very deeply cut, but the risk in this case was justified, as only plough soU Avas removed. The excavation of the individual graves was carried out by methods which were afterwards used by Mr. Warhurst in the excavation of the cemetery at Lyminge and described by him in Archceologia Cantiana, LXIX, pp. 4-5. At Holborough a measured sketch was made of each grave, the position of objects being noted in relation to each other and to the skeleton. A vertical photograph was also taken of each grave, and the drawings (Figs. 21 and 22) were made by combining the information contained in both the sketch and the photograph. In an emergency excavation this method saves time which might otherArise be expended on a detahed and painstaking drawing in the field, and yet ensures that no detaUs are lost regarding the appearance and position of the objects. Close-range vertical photographs of groups of objects were also of great use. The cleaning of objects in the grave for the purposes of photography was carried only as far as was possible Avithout disturbing the objects in any way. Wherever practicable, finds were lifted still embedded in the earth immediately surrounding them. This is particularly necessary in the case of a complex of small objects such as may be found in a purse. The group can then be more conveniently excavated on a table in the shed than in the bottom of a grave, and a detahed drawing can then be made. With regard to the ditches, a trench Avas laid out parallel to the quarry edge and as near to it as was practicable, to cut across each ditch. The western ditch was sectioned at this time. Towards the eastern ditch there was a smaU chcular pit c. 3 feet in diameter and 2 feet deep. It was sterile and filled with loose, clean, sharply-angular chalk lumps with no seepage or infiltration of the topsoU at all, and the sides were scored with the vertical marks of the broad end of a pick. I t was therefore evidently a modern pit, for an older one would have shown discoloration and sinkage of the fill. Where the trench crossed the eastern ditch, a grave (No. 18) was seen to be lying partly in the ditch filling. Proof that the ditches were anterior to the graves was therefore conveniently forthcoming at a very early stage. Further trenches estabhshed the chcular nature of the ditch, the fact that nearly half of it had been cut away, and that the fhst trench laid down had cut through its central point. No other disturbance of the chalk was detected in the cleared areas. In the second season the rest of the topsoil was cleared off by bulldozer and the chant surface cleaned by hand, so that the whole chalk surface in the vicinity of the chcle was exposed to view. The few clay and flint pockets were investigated. 86 PLATS 1 ( "" •"r=*—• '" ' ' '" BRONZE AGE I HON AGE ® SECONDARY NEOLITHIC SHERDS 1 CHALK HOCK SCALE OF FEET TOPSOIL LOOSE EARTH COMPACT CLAY EARTH. CHALK, CHARCOAL SILT A«O CHALK LOOSE CHALK Eio. 4. Section of ditch of prehistoric barrow. Primary silting is represented by the lowest layer 6, consisting of large, sharp-edged pieces of chalk loosely packed in the centre of the ditch. A second stage of this process is evident in layer 5 of tightly-packed yeUow sht with small, weathered lumps of chalk towards the sides of the ditch. In both of these layers there were fragmentary sherds of smooth, flaky ware, and the tightly-packed silting contained a distinctive piece of rusticated pottery with double rows of widely-spaced finger-pinching (Fig. 5a). On top of this silting which half-filled the ditch, was a layer 4 mostly confined to the central part between the side silting, and containing a mixture of ash, fragments of charcoal, animal bones and a larger sherd (Fig. 56) of much the same type of rusticated ware mentioned above. Above this a narrow band of brown clay layer 3, 88 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT stretched almost from one side of the ditch to the other, and this produced part of a rim with a small perforation some way below (Fig. 5c). To this level the sherds were consistently of a Secondary Neolithic culture. A small fragment of thick, soft, less gritty ware of the native Bronze Age also appeared in layer 3, together with a coarse piece of Late Bronze-Early Iron Age ware. Above this, and immediately below the 9 inches of topsoh which uniformly covered the barrow, ditch and surrounding chalk, was a reddish-brown soh layer 2 containing various types of sherds ranging from the Late Bronze Age to Roman times. Coarse cooking-pot fragments including a base (Fig. 5d) and some thinner ware in the shape of a considerable part of a grey bowl with an angular shoulder (Fig. 5/) belong to types which begin in the Late Bronze Age and continue into the Early Iron Age. To the Iron Age also belongs the slashed rim (Fig. 5e) and to the later part of that period some thick, soapy-textured sherds. Early Romano-British pottery was represented by a few small fragments. The organic material, which all came from layer 4, was examined by Dr. I. W. CornwaU. Animal bones were fragmentary and were probably bovine. There were also snails, one fahly complete Cepaea sp., and tiny fragments of charcoal, possibly oak. PREHISTORIC POTTERY (Fig. 5) (a) Smooth, flaky ware, inner surface dark grey, red outside; rusticated decoration, apparently in a double row of pinched indentations. From 5, primary silting. (b) Similar to above, but the double rows of pinching more clearly marked.5 From 4. (c) Similar ware to above, but with greater flint grit content; slightly thickened rim with bevel sloping outwards; perforation below rim made before firing. From 3. These three sherds belong to the Peterborough Beaker ware group in general, strong affinities with the Ebbsfleet variety being evident from the finger-nail rustication, flattened rim, perforation below the rim, and the thinness and gritty content of the pottery. Professor Grimes has shown me an Ebbsfleet type of pot from Heath Row, Middlesex, which has one simhar perforation in place of one of the usual finger-tip pits in the hollow of the neck.6 An example of a straightsided pot with a row of perforations below the rim but with a rim slightly turned over, is known from Grovehurst, Kent.7 The sherds of rusticated ware are also straight in profile, and Professor Grimes suggests this indicates a beaker shape rather than the usual round-based Ebbsfleet types. (d) Flat base fragment of coarse cooking pot; rough ware with 89 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT chalk grit, dark grey inside, red outside. From 2. Late Bronze Ageearly Iron Age. (e) Flattened rim fragment with diagonal slashes on outer edge; smooth buff ware with flint grits. From 2. Iron Age; cf. various examples from Fengate, Peterborough,8 and The Caburn, Sussex.9 * £ D t t t g * » ^ ^ f Fio. 5. Prehistoric pottery from ditch. J. (/) Part of bowl with angular shoulder, grey ware with reddish shades on outside. From 2. Late Bronze-early Iron Age; cf. Fengate, Peterborough,10 Worth, Kent11 and Minnis Bay, Bhchington, Kent.12 THE ANGLO-SAXON CEMETERY (Fig. 3) AU the graves, except one, were lying more or less west to east, with the head to the west, but Grave 31, so small that it must have been 90 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT intended for an infant, was almost at right-angles to the rest. There is no attempt at ahgnment sideways, but an occasional head-to-foot ahgnment may be noticed, e.g. Graves 9-10, 22-33, 35-37, etc. ' The topsoh, which had been under the plough until very recently, was everywhere about 9 inches deep and sterile. The graves were cut into the solid chalk, and were roughly rectangular with rounded corners. Grave 8 was the largest, 9 ft. 3 in. long, 3 ft. 9 in. wide at the head and 1 ft. 3 in. deep from the top of the chalk. The deepest cut was Grave 28, with a maximum depth of 2 ft. 6 in. The original extent of the cemetery is not known. We do know that graves were disturbed in 1950 when the top of the strip about 60 feet wide which included the first part of the barrow was taken away. The operations in 1943 which produced the pot and hone sent to Maidstone Museum in that year took place in the deep bite made to the west of the Roman tumulus, and must have disposed of the northeastern part of the cemetery. Apart from the orientation, there was no rigid system of spacing, and judging from the uneven northern outline there seems to be no definite boundary. The graves as excavated are in two groups: one group clustered round the barrow includes only one inhumation with goods, No. 18; the other group lies on the eastern slope and includes all the other furnished graves, most of these being on the south-eastern side, those on the north-western fringe being without goods. This grouping into two parts may be quite artificial, for the graves may have been continuous in the destroyed part of the cemetery. There is no case of disturbance of an earher burial by a second one, and no two graves are very close together. Theh location must therefore have been indicated on the surface in some way, and in view of the fah distance between each, it is most probable that the marking was by means of smaU mounds. There is no obvious grouping with respect to age or sex, for the graves of men, women and children are intermingled. The graves furnished with goods are also of each type. The skirting of the perimeter of the barrow by the Anglo-Saxon graves and the sherd-bearing fiU of Grave 18 have aheady been adduced in support of the more complete state of the barrow in Anglo-Saxon times. A Bronze Age barrow was a site often chosen by the Saxons for a cemetery. This was noticed by Baldwin Brown, who listed a number of cases,13 and it is a practice also followed on the continent, in Schleswig-. Holstein, Denmark and elsewhere,14 the burials often being confined to the eastern and south-eastern slopes of the barrow. The cemetery must have belonged to the settlement of Holborough in the vaUey below, from where it would have been possible to see the smaU Saxon mounds on the south-eastern slopes of the hhl behind the Roman tumulus and encroaching on the older barrow at the top. 91 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT Some of the bodies, at least, appear to have been contained in coffins. An hon nah which was found near the top of the fill, at the foot of Grave 18, is of a type often used for this purpose. In other cases, such as Graves 8, 16, 25, 27, 30 and 32, a rectangle of black dust c. 8 feet by 2 feet appeared in the fiU some inches above the floor and continued vertically down to the floor and under the bones. In Grave 17 the vertical walls15 of the coffin at the head and along the right-hand side of the skeleton were discovered in situ. Fragments of simhar pieces of wood were found in Graves 6 and 7. In no case was there evidence of wood above the skeleton, and it seems, therefore, that the bodies were usually buried in a hdless coffin. DISCUSSION OF FINDS In addition to the grave goods obtained by excavation, three other objects are appended to the hst. The smaU pot and whetstone (Fig. 20a and b) are in the possession of the Maidstone Museum, and are recorded as having been found 2 feet under the surface 200 yards from the " burial knob," Snodland, in 1943. A spearhead is also mentioned, but is not in the museum. This is to be related to the chalk excavations going on in that year which resulted in the semi-chcular cut to the west of the Roman tumulus shown on the plan (Figs. 1-3) and is fahly reliable evidence that they came from part of the same cemetery. The provenance of the necklace of beads (Fig. 20c) is not so certain. In the possession of the British Museum (Reg. No. 1947 5-2 358), it was acquhed from Major J. P. T. Burchell who obtained it from the widow of Mr. W. H. Cook, and it is recorded that he found it on the Roman villa site at Snodland, Kent. . I t is known that Mr. Cook did once discover a skeleton on Holborough Knob. In 1952 an employee of the cement company said that during the excavations of 1943 skeletons were found with two spearheads and a necklace. Two spearheads were disposed of to a scrap-hon merchant, but the fate of the necklace was unknown. The necklace in the British Museum was purchased by Major Burchell in 1945, and as it appears to be Anglo-Saxon rather than Roman, it seemed probable that this was the missing necklace and worthy of inclusion in this report. During the excavation, the shed was broken into and equipment stolen. Unfortunately the pommel of the sword from Grave 7 also disappeared on that occasion. BUCKLES The buckle in Grave 11 (Fig. 18, 1 and PI. Ila) belongs to a wellknown series of which there are few substantial variations in the basic form, although no two are identical.10 Tho loop is always cast in one 92 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT piece with the plate, the tongue alone being movable; the end of the plate finishes in a pah of rudimentary birds' heads or animal heads, facing either inwards or outwards, the rest is perforated with shapes in the form of a T, a cross or a flat-hon, and the reserved metal decorated by eirele-and-dot motifs. The whole object is flimsy in construction, and suitable only for light straps no more than about half an inch wide. The animal heads are most clearly seen on the piece from Kingston Down, Grave 300, where the lower jaw is looped and projects above the upper jaw, giving a tusk-like effect.17 I t could be argued that the cruciform designs on the buckles from Broadstahs, Breach Down, Uncleby and Sibertswold Down Grave 131 are motifs of no particular significance, but the intention of the craftsman to delineate a Christian cross on the Holborough plate is evident as the shape of the reserved metal is that of the equal-armed curvilinear cross in religious use at the time. Another reason for supposing the Christian period for graves containing such buckles is the prevailing scantiness of accompanying objects—perhaps a knife or a spearhead only. There is sufficient evidence, however, to show that the buckle type could suitably be worn by men, women or chhdren, for essentially female belongings occur in the Uncleby grave, there were spearheads in Sibertswold Graves 150 and 176, and Grave 131 in the same cemetery was under 4 feet long and apparently contained a child. I t has aheady been noticed elsewhere18 that a somewhat similar buckle is a copy of a more complicated jeweUed type, the perforated shapes in the bronze type taking the place of inset garnets. The effect was probably enhanced by the use of a brightly-coloured leather or cloth backing the perforation. Some of the outlines used, a cross or T-shape for instance, correspond to the shapes of garnet cells in the late seventh-century gold jewellery, and the cross type formed by the reserved metal on the Holborough plate is that used for the contemporary pendants from Ixworth, and Wilton, and St. Cuthberts cross,19 as well as the heads of pins from London and Breach Down.20 On styhstic grounds, then, they belong to the later part of the period during which the deposition of grave goods was in fashion. The case is further strengthened by the associated finds of the example in Grave L at Valetta House, Broadstahs, for apart from a knife and two " girdle-irons or keys," this contained eight sceattas. This does not, of course, date the grave accurately at the present state of numismatic knowledge, but indicates a time in the seventh or the early eighth century. Decoration a jour on bronze was not widely used by the Anglo- Saxons during the sixth century, although there are examples such as swastika brooches. The openwork discs which functioned as girdlehangers appear in undoubtedly later contexts such as Grave 121 at 93 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT BurweU,21 where the same T-shaped perforation is apparent. A simUar series may be noticed on the continent, and examples occur in Werner's Groups IV and V (i.e. between A.D. 600 and 700)22 but not in the earlier groups. In many cases the perforations are of cruciform or step patterns, and the decoration is sometimes by chcle-and-dot stamping. Often, too, the arrangement is cruciform. Even closer to the object of study is a Frankish buckle with rectangular loop,23 but there are vital differences in the shapes of the loop and the tongue base, and in the use of dome-headed rivets. The date of the foreign girdle-hangers is established as the last part of the seventh century, but the openwork technique went on into the eighth century without much change in the form of decoration on the handles of bronze keys.24 The long, rectangular buckle-plate with geometrical perforation seems to be pecuhar to Anglo-Saxon England, and along with the openwork girdle-hanger discs constitutes an insular series contemporary with the continental set of objects in the same technique. Like the openwork buckle, the bhd buckle (Fig. 16, 8 and PI. 116) also belongs to the dehcate type in which the plate is cast in one with the loop. The Style I I bird's head motif is extremely common in other countries as weU, and frequently occurs in pahs, on pendants, for example from Nordendorf25 and Soest.26 Some ornate buckles include such birds' heads as shoulder decoration, and the large gold buckle from Sutton Hoo27 is a good illustration of this, for it indicates how a craftsman who set out to make a tiny buckle in the same style might find it convenient to use the two shoulder heads as sole decoration for the plate. A recent find from Holland is a bronze buckle with openwork triangular plate consisting mainly of two birds' heads.28 Three examples in which a mask is enclosed between the birds come from Skane, Sweden,29 Zealand, Denmark,30 and northern France.31 Where the position of the mask is occupied by a fish instead, it is possible that a Christian symbol is intended, as on the pendants from Kingston, Grave 161,3 2 or a smaU bronze buckle with immovable loop from Faversham.33 Also from Faversham are two magnificent gold filigree strap buckles with this motif doubled, so that there are two pahs of birds' heads, one at each end of the plate.34 A close version of the pattern in filigree work adorns the back of the Kingston brooch, and another in chipcarving, much nearer in style to the Holborough buckle, surrounds the catch of the composite brooch from Sarre.36 The nearest parallels of all, however, are two smah buckles, a fragmentary gilt bronze buckle from Faversham (Fig. 6) in the British Museum,30 and the other from Mitchell's HiU, Icklingham, in Moyses Hall, Bury St. Edmunds. Both are cast in one, with a chip-carving bhd design, but the Icklingham one is slightly smaUer, being only 0 -8 in. long. It seems likely that these, 94 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT together with the Holborough buckle, emanated from the same workshop as those from Faversham mentioned above. The remaining buckles are not so distinctive, but nevertheless, almost exact paraUels can be produced from the same milieu as that of the openwork buckles. The smaU bronze buckle cast in one piece with the loop (Fig. 16, 6) occurs at Kingston, Grave 40, and Barfriston Down, Grave 27,3 7 A smaU bronze buckle with a folded plate is common, but from Beakesbourne, Grave 32, comes one which has an edge nicked in exactly the same way as Fig. 19, Gr. 18, 3.38 PIG. 6. Bronze buckle from Faversham, Kent. J. (By courtesy of the Trustees of the British Museum.) SMALL BRONZE AND IRON OBJECTS The type of bronze catch or hasp (Fig. 18, 2c) has been found elsewhere, e.g. at Holywell Row39 and BurweU,40 Shudy Camps41 and Kingston Down.4 2 A simhar fixture with a rivet at each end was found on a piece of perforated leather at Beakesbourne and Sibertswold. 43 It is presumed, therefore, to be a bag or belt fitting. The smaU bronze loop (Fig. 18, 5) and hon spoon (Fig. 18, 6) from Grave 11 are known from other places, and occur in the same cemeteries aheady mentioned, e.g. HolyweU Row,44 BurweU,45 Shudy Camps40 and Sibertswold.47 Also common to these cemeteries are the purse-mounts of triangular shape with curling ends48 (Fig. 16, 9) as opposed to the straight type with curling ends and a smaU buckle attached to the centre of one side, which seems to have been in use earlier.49 The bronze spatula-shaped object with the broken shaft, however (Fig. 16, 7) might very well be a relic from Roman times.50 SHIELDS The four shield bosses (Figs. 14, 15 and 17) have a number of points in common: they are aU rather smaU, the greatest diameter being 5-6 in.; the dome, although shghtly convex, tends to be conical in shape with only a trace of a carination; the knob at the apex is extremely smaU, and the flange very narrow and slanting. The two remaining 95 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT grips are strap-shaped. AU this contrasts strongly with the bosses of the early pagan graves which tend to be between 6 in. and 7 in. wide with a sharp carination, wide flange, button top and a grip with bentover sides. The Holborough kind does not seem to appear in early graves, but it is by no means confined to Kent. In the cemetery of HolyweU Row, which belongs to the pagan period, there are nevertheless a few graves which have clear affinities to the later Christian group such as Nos. 11, 31, 70 and 85. In this cemetery conical bosses occur in Graves 4 and 29, 56 and 91.51 The boss from Grave 38 is described as carinated, but is actually very simhar to Holborough Grave 7, and its diameter is only 5-3 in. The associated finds were not very distinctive, but in both Grave 4 and Grave 91 there was a smaU bronze buckle with doubled-back rectangular plate such as often accompanies later burials.52 Other bosses come from PrittleweU, Essex,63 Ipswich, Suffolk,64 and Eastry, Kent,55 Barrow Hhl, Ebbesbourne Wake and Sahsbury Race Course, WUts.,58 Melton Mowbray, Leics.,57 Mitcham, Surrey,68 and the two included in the chieftain's burial at Taplow, Bucks., are also of this form. In connection with the Eastry example Baldwin Brown gives a list of similar bosses but, in doing so, he includes the sugar-loaf variety. There are, indeed, points of simharity in the lack of distinct carination and narrowness both of the flange and the total diameter, but on the whole the height of the sugar-loaf compared with its width marks it out as a distinct type. There are, however, examples which seem to fall between the two, such as that from Baginton,69 and this suggests a line of development from the Holborough type of cone to the sugar-loaf type of the late seventh and early eighth centuries. A related type which may be purely insular is the taU, straight-sided cone as from St. Margarets near Dover,60 Melbourn, Oambs.,61 Middle Down, Alvediston 03 and West Knoyle Farm, Stourton, Wilts.03 The sugar-loaf and related types are to be found also in Germany and HoUand, where they run on weU into the eighth century.64 Rivets of the disc-headed type found in Grave 7 (Fig. 15, 3c) are fahly common and have been noted in Kent,65 Yorkshhe,00 Cambridgeshire 67 and WUtshhe.68 The bronze-plated ones of diamond shape in Grave 8 are exceptional. On both shields these ornamental rivets were placed two together or singly on opposite sides of the boss and at a distance from it, as shown on the plates of the helmet from Vendel XIV.69 The angle of the boss flanges indicates a considerable curvature to the surface of the shield. Thickness of wood used is indicated by the lengths of rivet shanks as from 0 • 4 in. to 0 • 55 in. in Grave 7 and 0 • 6 in. in Grave 8. In Grave 7 the minimum width suggested by the appliques is 1 ft. 8 in., and in Grave 8 1 ft. 7 in., but here, as the shield had been 96 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT stood on its edge, it is possible by doubling the distance of the boss from the floor of the grave to calculate a minimum diameter of 2 ft. 2 in. H FIG. 7. Drawing of the radiograph of the spearhead from Grave 7. \. Cf. Fig. 15 and PI. Ula. INLAID SPEARHEAD It has not been possible to date to have this spear (Fig. 15, 1) cleaned, but the radiograph (PI. I l i a and Fig. 7) leaves no doubt that the marks are caused by an inlay of whes of some other metal than iron, for its higher density shows up well against the rusty hon blade in which it is inset. If there were any more inlaid whes, they would certainly have been apparent in the radiograph. If, on the other hand, whes had been hammered hi on other parts of the blade and have faUen out, it is possible, but by no means certain, that the extra thinness of the hon in that place caused by the empty channel might be revealed. 97 10 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT The spear-blade thickens considerably towards its central rib, and any inlay in this position might not stand out on a radiograph. However, as the existing mark is near the edge, it is a reasonable expectation that if there were others, some at least would be placed in relatively much the same position. On the present evidence, therefore, the visible traces must be regarded as constituting the enthe original design. The first glance suggests a runic inscription, but as the strokes are combined in an entity, it might be a symbol of some other kind. On looking at other examples of Germanic ornamented spearheads, we find that not only are many of them decorated in precisely the same inlaid technique, but also that often the design consists of a runic inscription or of symbols, or a combination of both. The main series of spearheads with runic inscriptions belong to the thhd or fourth centuries A.D. and are in the Gothic or North Germanic language. They have been pubhshed together by W. Krause,70 so that it is not necessary to do more than refer to them here. There are differing opinions on the reading of the runes, but it is generaUy agreed that in each ease the word represents the name of the spear as a kenning, e.g. " raunijaR " incised on the blade from 0vre Stabu, Norway, which is the North Germanic form of 0. Icel. " reynir," the noun connected with the verb " reyna " which means " to make trial of," therefore " the prover." Some of these inscriptions begin with an unintelligible symbol, and, as Marstrander has pointed out, this foUows the custom of Greek and Latin inscriptions which began or ended with some symbol like a swastika, cross or palm leaf.71 According to his reading of the inscription on the spear from Kovel, VoUiynia, the fhst symbol is apotropeic and not a rune,72 e.g. J Simharly, the runes on the Dahmsdorf spear begin with a chcle,73 and the Stabu inscription starts with two lines now partiaUy obscured.74 The other symbols used include the customary ones among the Germanic tribes: chcle-and-dot, swastika, triskele, half-moon, but in addition there is on the Dahmsdorf spear a double fork motif (and there are others somewhat simUar on the Kovel and Mos blades) which Marstrander interprets as a lightning motif with origins in the Bosphorus. 76 A number of inlaid spears were mentioned by Kossinna,70 but apart from one more ambitious sun-symbol,77 the marks appear to be confined to chcle-and-dot motifs.78 An unusual combination of straight lines elaborates the blade from Stenstugu, Gotland, where four straight inlaid whes meet at a point marked by a chcular inlay of red enamel.79 Although the examples considered so far belong to a period much earlier than that encompassed by the Holborough cemetery, it will be 98 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT reahzed how strong a bearing they have on the subject when more contemporary weapons are examined. Extreme conservatism through centuries is apparent from the spears found at Great Chesterford, Essex,80 and Eprave, Prov. de Namur,81 which belong to a time about A.D. 500, but which wear the same traditional inlaid chcle motifs. A sthl later Alemannic spear from Wurmlingen82 of the late sixth or early seventh century carries on the tradition of an inlaid runic inscription along one side of the blade, this time a man's name, but beginning as usual with a symbol, according to Krause's reading. Other Alemannic blades bear a Christian cross.83 These other examples are of great assistance in attempting to decipher the meaning or purpose of the mark on the Holborough spear. The knowledge that runic inscriptions are quite common in such a context makes it necessary for this possibility to be considered here. The arrow-headed part could certainly be the rune f =t, but the other part does not represent any rune of the Anglo-Saxon futhark in its usual form. |~) does occur in Scandinavian inscriptions in the earlier futhark for M = e as on the diadem from Strarup.84 A bind-rune of two individual runes 'f and the early f~| is unlikely in Anglo-Saxon England unless its origin belongs to a much earher period and its appearance here is as a fossilized symbol.85 If it represented a combination of f and h > i-e - * and u, or -f and h, i.e. t and c, the second element would be in a form not known elsewhere in Anglo-Saxon inscriptions. An interpretation " Tu" must be considered, however, for certain reasons. It could represent one form of the name of the god of war known in Anglo-Saxon as Tiw, for a spelling " Tuu " does occur in one manuscript.80 Moreover, there are two amulets on which Stephens professed to read an invocation to Tiw. Of these, however, the petrified spicule of an echinite from Glostrup, Zealand, inscribed f ft is considered by Moltke to be medieval, 87 and he is unable to see any runes on the stone from Jyderup, Zealand, which Stephens transcribed as f T fAr* —" Tj™ a1-"88 The t-rune itself was known by the name of the god, and, used alone, would bear just that magical value. And what symbol is more hkely to occur on a weapon in view of the instruction in the Eddie poem " Sigrdrffumal," which says that victory runes should be inscribed on a sword, and that Tyr should be named twice?89 Perhaps this is what the rune-writer of the Kylfver stone had hi mind when he rounded off his inscription with something like a multiple t-rune which has six strokes on one side and eight on the other.90 The inscription in Etruscan characters, but in archaic Germanic language, on one of the helmets from Negau, includes the dative of the word for god—" teiwa," the word which later became used specificaUy for the name of the Anglo-Saxon and Scandinavian god of war.91 At a 99 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT much later stage, but sthl carrying on the tradition of invocation to a god, the symbol of the Christian cross appears on Alemannic spears. If the mark is to be regarded as a symbol, but not necessarily a runic one, there are unlimited possibilities, from which two or three likely ones may be selected. Simple geometrical signs such as X T Y are often used as marks of property, and where such a symbol occurs on its own, this may indeed be its only purpose. Whatever theh function, examples of this are to be found on the spear shafts from Vimose, and varieties of arrow-shaped signs occur on other objects.82 Then again, the Holborough mark has definite affinities to the lightning symbols mentioned earlier, some of which also appear on spear blades.92a Lastly, there is a possible connection with the symbol 6*. For many centuries this has been used to represent hon, the male sex and the planet Mars. Jacob Grimm remarked upon the simharity between this symbol and the rune 1" which was caUed by the name of the corresponding Germanic war god.93 It is evident that there was already by the eighth or ninth century a recognized connection between the rune, the war god, and a heavenly body, for in the Rune poem of that time there is a verse which may be translated: " T is a special symbol: it keeps pledges well with noblemen: it is always on a journey across the darkness of night; it never ceases."94 After the rune is the name Tir, possibly showing confusion with the Old Icelandic god Tyrr.95 No definite conclusion can be reached concerning the meaning of the mark on the Holborough spear, but its smallness in comparison with the size of the object (length 0-2 in. on a spearhead 18-5 in. long) seems to show that its purpose was symbolic rather than a visual aid to identification of ownership. SWORDS Both of the swords have pattern-welded blades, and the design of the one in Grave 2 is quite clear (Fig. 8).90 Remains of wood and wool (possibly sheep) on the blades suggest wooden scabbards lined with sheepskin.903 POT-HANGER There seems to be only one other pot-hanger of this kind (Fig. 13, Gr. 2, 2) in England, and this is the one from Chartham, Grave 34.97 I t is about 26 in. long, whereas the Holborough one is 34 in. long. They both consist of two main rods of twisted hon, but the connecting links are of different shapes and differently disposed. The two rods on the Holborough hanger are connected by a single loop, and the chain is attached to the extremity of one of these. The rods from Chartham, however, are each attached to a chcular ring and these rings are 100 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT connected by a chain of s-shaped links. The smaU hook at the end of one rod is suitable for attachment to a suspension ring in the roof, whhe the wide hook at the lower end makes for the easy accommodation of pot handles. In addition, the shorter hook aUowed for the simultaneous heating of a smaller dish above the larger brew or for providing a FIG. 8. Schematic drawing of the welded pattern revealed by radiograph on the sword from Grave 2. 1. stage farther from the fire. The only other object in this grave was a knife with an unusual curved blade, according to Faussett " in shape not much unlike a pruning knife." Iron pot-hangers were aheady in use in the La Tene period, and continued without any radical change of design until the end of the fourth century.98 Characteristics of the earhest types are figure-ofeight loops with a trefoil-shaped loop at the junction of the two hooks 101 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT (there are often two hooks at the lower end to hold opposite sides of the pot) as at Bigbury Camp99 and Newstead.100 Others, however, are practically indistinguishable from the Holborough example, except for a medial fixture of a double hook with a perforated, flattened central part, as at La Tene,101 and by the end of the Roman period there are precisely the same combinations of rods and ring chain with a folded loop at the end, from Saalburg,102 Vichy (AUier),103 and Mainz.104 Yet another belongs to the Migration period of Gotland,105 and is assigned by Nerman to his Period VI, 2, i.e. between about A.D. 475-500 and about 550-600. This, and the late Roman chains, show that the Holborough hanger represents only the detachable part of the equipment, and it also explains the function of the doubled loop at the top. Each side of this loop is slipped over a separate hook, so making a firmer and steadier connection than is possible by means of a single hook and loop (Fig. 9). The pot-hangers from Vendel Graves 1 and 8 and Valsgarde Grave 8100 are different; they are more ornate with double rods ending in spirals, the loops are oval, and the double hook with perforated centre is a descendant of the La Tene type. These, however, are stUl of the same recognizable pattern of twisted rods and loops.107 By the Viking period the pattern is completely altered to elongated double loops, a trefoil dividing loop and very short rods.108 I t is therefore certain that hangers identical to the Holborough one were being made by the end of the Roman period, and there is a possibility that the pattern was continued until at least about A.D. 600. Definite changes are discernible in the northern Vendel and Viking periods, however, and it may be that the earher type was discontinued. One twisted hon bar, 20 cm. long, with one end hooked and the other looped, was found at Dorestad, and this suggests the style had not died by the ninth century.109 For the purpose of suspending a cooking pot over a fire, an iron chahi of adjustable length is the obvious solution, and such a chain should not have been beyond the capabilities of local smiths in the Germanic world. Considerable conservatism of design might be expected so that products of different areas might be simhar. But as chains of identical design are forthcoming from places as distant from each other as Saalburg, Mainz, Vichy, Gotland and Holborough, the conclusion suggests itself that possibly they are the exports of one manufacturing centre. This likelihood is strengthened by the fact that the Gotland chain was found with a bronze bowl of a type which was probably manufactured in Belgium and exported to France, Germany, England and Scandinavia.110 It is feasible that the chains were made in the same centre as the bowls, and exported with them as a complete kitchen unit. 102 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT In the case of an object like this which shows httle change of form over a stretch of centuries, it is the associated finds only which can in Q & FIG. 9. Pothanger and cauldron from Gotland (after Nerman). £a. indicate a date. The sohtary comparable find in England cannot be closely dated by its accompanying grave goods, but, on the other hand, the cemetery in which it occurs appears to be a late one, for it includes 103 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT a number of unfurnished graves, and the few objects it produced have obvious contacts with the Christian period, a shver cross, for instance.111 A twisted bar which may originaUy have been part of a pot-hanger comes from a presumably Christian Saxon grave at Winkelbury Hhl, where it may have functioned as a coffin fitting.112 By the group of cauldrons, etc., in the Sutton Hoo ship burial there was an hon chain complex and it is likely that the fuU pubhcation of the contents of the ship whl reveal this complex as including one or more pot-hangers.113 POTTERY Bottles of this kind (Fig. 13, Gr. 4, PI. He) are common to both sides of the English Channel. The particular rouletting pattern is of less frequent occurrence, but the general scheme of a double undulating line of impressions with other stamps in the curves of the waves is also to be found in both England and France. In Maidstone Museum there are at least three examples of such rouletting, aU on bottles; two of them from Sarre. Some of these are presumably represented by the sketches published in the original account of the excavation.114 It is not possible now to be certain of the grave groups, but the distinctive finds of a garnet-inlaid pyramid in Grave CCXI and amethyst beads in Grave XX are late features. Faussett found a similar bottle in Grave 10, Barfriston Down, together with a small silver buckle and spearhead, 116 and another came to light at Harrietsham.116 Across the Channel, the rouletting occurs on a bottle at Clery, Somme.117 A wide-mouthed jar from Canterbury118 which seems to be a rather odd form, is merely the lower part of an unfinished bottle vase. Sthl another form of pot in Kent bears this type of decoration, as may be seen from the biconical bowls from Wickhambreux and from Faversham.119 These occur also in France, at Mazinghem.120 At Broadstahs121 the rouletting again appears, but this time on a curious wheel-turned jar with shoulder bosses which can be paralleled by a jar from a seventh- or eighth-century context in the late Merovingian cemetery of Villers-devant-Orval, Belgium.122 The Broadstahs jar, together with another from Faversham, were included by Myres in his types of Romano-Saxon pottery, but with the proviso that they might be Frankish imports from the Rhineland.123 In a region so closely in contact with the continent as Kent, it is difficult to disentangle the influence on native pottery of pre-Saxon Roman wares or contemporary Frankish types. In view of the late connections mentioned above, however, and the absence of known Connections with the migration period, it seems that they must represent the fusion of wheel-turned and hand-made traditions on the continent. As to the smaU, hand-made pot (Fig 20, 1, PI. II, d), Faussett Ulustrates one generally similar from Chartham Down where it was found in a 104 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT chUd's grave.124 The only particulars given is that it is smaU, black, and would hold a quarter of a pint. Another is listed by Mortimer in Grave No. 8 in the late cemetery of Garton Slack, Yorks.126 These pots have an essential difference, however, in that the constriction at the neck is much less definite. Moreover, the criss-cross tooling to produce burnishing in lines does not seem to occur in the kind of pottery forthcoming from Saxon graves. This tooling, either in vertical lines, or in a criss-cross pattern, and either on a rough surface or a fahly smooth one so that burnished streaks are produced, is a characteristic of pottery of quite other provenances. The relationship of the Holborough pot to these other sources is further accentuated by the fact that the tooling often occurs on pots of globular form with sharply constricted necks. Without exception, hand-made pots like this came from late Saxon occupation sites in Kent. Theh dates have not yet been precisely determined, but they seem to begin in the eighth century and go on possibly unth Norman times. Canterbury is one of the sites producing such ceramic material, and a coarse domestic bowl with overall treatment (smoothing) of the surface has aheady been published,126 but there are also others from Canterbury and Dover. From a settlement site at Sandtun, West Hythe, Kent, there is a slightly larger black burnished pot of comparable shape, but decorated with groups of vertical grooves in panels, obviously a descendant of the early Anglian types. This site is unpubhshed, but Mr. G. C. Dunning informs me that the coarse ware occurring in the same level as this pot indicates the late seventh or eighth century. A well-known example with vertical smoothing is the spouted pitcher found at Richborough in loose association with the coins of Offa.127 The Holborough pot, then, has affinities with the latest of the pagan pottery and with the earhest examples of the Christian period. WHETSTONES The smaU hone is a mud-stone foreign to south-east England, and must therefore be an object of trade. Although Anglo-Saxon whetstones have not yet been subjected to particular scrutiny, some from medieval sites have received notice on several occasions, and especially by G. C. Dunning and E. M. Jope,128 with the result that various kinds of rocks have been estabhshed as being in use for this purpose. Sedimentary rocks, probably obtained from the Midlands, are in evidence from the eleventh to the early fourteenth century, and these sources had already been tapped by the Romans.129 Hones of schist have also been listed, showing a denser distribution in south-eastern England, and a possible source of supply in France. On the other hand, a coUection of twenty-eight Yorkshhe hones (some of which belong to the Viking period) has recently been examined,130 and included only 105 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT one example of schist. For this and the fifteen granulite rocks examined, together with a metamorphosed sUtstone, sources in Scotland were suggested; the sedimentary types could have been found in the glacial drift of Yorkshhe, and one sandstone is of Kentish Rag. The importance of accurate petrographical study of each specimen is hereby made very clear. It is to be regretted that there is as yet no such reliable data avahable for the early Saxon period, but a collection of examples is now being made with this end in view. For although whetstones have been regarded as of rare occurrence in Anglo-Saxon graves, there is an appreciable number. The following is a preliminary list: 1. Guildown, Surrey, Grave 136. Surrey Arch. Coll., XXXIX, 26. 2. Horton Kirby, Kent. K.A.S. Catalogue, p. 32, No. 507. 3. Bifrons, Kent. Idem, p. 33, No. 520 (two). 4. Uncleby, Yorks. B.B., IV, p. 414, PL XCVII, 5 (a dozen). 5. Camerton, Somerset. Som. Arch. Soc, LXXIX, Grave 95, p. 60. 6. Galley Lowe tumulus, Derbyshhe. T. Bateman, Vestiges of the Antiquities of Derbyshire (1848), p. 37. 7. Soham, Cambs. Proc. Comb. Ant. Soc, XXXIII, Grave No. 1, p. 156, Fig. 1, 2. 8. Sutton Hoo, Suffolk. B.M. Sutton Hoo Ship Burial, PI. 6. 9. Caistor-by-Norwich, A.S. Cem., Grave 10. Norwich Museum. 10. Croydon, Surrey. Surrey Arch. Coll., XIII, 18. 11. Faversham, Kent. British Museum. 12. Harrold, Beds. M.O.W. excavation. 13. Dover, Kent, Grave 162 (two). M.O.W. excavation. 14. Great Chesterford, Essex, Grave 115. M.O.W. excavation. 15. Sutton Courtenay, Oxon., Houses IV and XIV. Arch., 73, 157, and Arch. 76, 69. 16. Wykeham, Yorks. Habitation site. 17. Linford, Essex. Habitation site. M.O.W. excavation. 18. Elmswell, E. Yorks. ? Roman or Saxon. Hull Mus. Pub. No. 198, PI. IK. 19. Caistor-by-Yarmouth. M.O.W. excavation. 20. Grimstone End, Pakenham, Suffolk. Proc Suffolk Inst. Arch., XXVI, Part 3, p. 206, Fig. 30a. 21. Minninglow, Derbyshhe. Sheffield Museum. There are also the following whetstones of the late Saxon period: 1. Guildown, Surrey, Grave 183. Surrey Arch. Coll., XXXIX, Fig. 11 and p. 32. (124.) 2. Whitby, Yorks. Arch., LXXXIX, Fig. 18, 6. 3. Thetford, Norfolk. M.O.W. excavation. 106 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 4. WaUingford, Berks. Proc. Soc. Ant., VIII, 468; B.M. Anglo- Saxon Guide (1923), 112, Berks. Arch. Journ., LI, 70. 5. Guildford, Mount Street. Proc. L.O.W. Nat. Hist, and Arch. Soc, II, 684, Fig. 6, 4. 6. Northampton Castle. Idem, 686, Fig. 7, 5 and 6. 7. Clifford Street, York. Proc. Yorks. Geol. Soc, 141-8. 8. Mawgan Porth, Cornwall. Ed. R. L. S. Bruce-Mitford, Becent Archceological Excavations in Britain, Fig. 59e. In addition, whetstones of the same period are known from the Celtic areas of the British Isles, e.g. Garranes, Ballycatteen and Lagore in Ireland and Bantham in Devon,131 and from these areas, too, come the carved whetstones of the Sutton Hoo type.132 With only one or two exceptions, the pagan period hones are of fah size and ghth, and are made of some rough-textured rock such as sandstone. These are known also hi the later period, but then there is an increasing number of smooth-textured stones small enough to be easily portable, and often pierced at one end for suspension. This type may also be noted abroad, at Dorestad, Hedeby, Bhka and Vendel and in Viking settlements in Scotland.133 It is this type to which the Holborough hone belongs, and although it is impossible to make useful observations without the advantage of petrographical tests, it is at least possible to see that the small portable hone was in more general use in the late Anglo-Saxon period. CONCLUSIONS The number of finds from this cemetery is very smaU, and yet practically every fragment retrieved makes its own important contribution towards an assessment. In the discussion of types above, it has been seen that almost without exception, very simUar objects and sometimes practically identical ones are to be found in the group of Kentish cemeteries excavated in the eighteenth century by the Rev. B. Faussett, and the graveyards of the Cambridge area excavated by T. C. Lethbridge in more recent times. Mr. Lethbridge came to the conclusion that BurweU134 was a Christian cemetery, for the reasons which he gives in his report, and this theory he further elaborated when publishing Shudy Camps136 a few years later. Since then a cemetery at Melbourn has been partly excavated and may be added to this group.130 The archeology of this period was subjected to an examination on a more far-reaching scale by E. T. Leeds,137 who, whUe not agreeing that this type of burial was Christian, at least conceded that it was semi-pagan. Besides the Kentish and Cambridgeshire sites aheady 107 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT mentioned, Leeds quoted a number of finds from various places spreading from Yorkshhe to Somerset. As a result of his exposition of the data, it was apparent that fashions at that time did not tend to develop on local lines, but that a certain homogeneity pervaded the style of trinkets and other furniture deposited in the graves. The distinctive openwork buckle, for instance, was as likely to turn up hi Yorkshhe as in Kent, and shver bullae were common both to Cambridgeshire and ' Somerset. This is to cite but two of the sahent points regarding the objects of the late group, however; other features are the consistent orientation of the graves from west to east, the large proportion of graves containing no objects at all, the almost complete absence of brooches from the women's graves and the lack of weapons with the men. In most of these particulars, Holborough is comparable. The relationship of the type of objects found there with the Cambridgeshhe and other late groups, has aheady been adequately commented upon. The graves are all lying in the dhection west-east. Out of thirty-nine graves actually excavated or definitely identified, only nine contained any objects at all. Of the graves excavated of women who were provided with possessions, not one could boast of brooch or beads; a knife, buckle, pin or keys represent theh richest possessions. The few glass beads apparently from a disturbed grave are small and simple, as usual in late graves. On the other hand, the men with grave goods were not poorly provided with weapons, for there were two swords, apart from spearheads and shield bosses. None of the types are early, however, and we know that chieftains were provided with their weapons after death at least as late as Sutton Hoo and Taplow. Apart from the resemblance of certain objects to finds in the Cambridgeshhe group, there are Christian connections evident from two of the buckles. The openwork buckle from Grave 11 has a cross of undoubted Christian form in current use; the bhd buckle in Grave 7 is very simUar to a buckle from Faversham which may have emanated from the same workshop as one with a fish symbol. Most intriguing of aU are the smaU pot and hone which were found in 1943 in association with other objects 200 yards from the Roman tumulus. According to the records of the Cement Company, the area excavated in that year was the concave incursion to the west of the tumulus. The farthest part of this is rather less than the 200 yards from the tumulus recorded as the find spot, but presumably this figure only represents an approximate guess on behalf of the finder. The fact that the spot was so near the site of known graves, and the fact that other Saxon objects, including a spearhead, were forthcoming at the same time, are a firm indication that it was part of the same cemetery 108 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT which was disturbed, and not some other complex such as a habitation site. This being so, it is extraordinary that the two objects, the hone and the pot, appear to be almost unknown types in pagan burials, even late ones, but are typical of finds from habitation sites as yet insecurely dated, but probably starting at least in the eighth century. From the character of the other finds from this cemetery, there seems no doubt that the site was in use in the latter part of the seventh century, and it must be considered whether the local inhabitants continued to use it into the eighth century, and later, when the practice of providing grave furniture was abandoned. The question of how far into the seventh or eighth centuries the custom of depositing grave finds continued in England is a vexed one. There are occasional finds which have been assigned to the eighth century,138 e.g. the openwork zoomorphic plate from Gilton. At least two Saxon graves have contained a smaU glass stud with a moulded pattern of sunken channels. This is the type of jewel which forms part of the embeUishments of the Ardagh chahce, the Ekero crozier139 and other such objects generaUy aUocated to the eighth century. They appear to be of Irish origin, and some in the process of manufacture have been found at Lagore140 where the excavator allocated some to the seventh century. One forms the central roundel in the pin chain at Roundway Down, Wilts.,141 and the other appeared without any mounting in Grave 5 at Camerton, Somerset.142 In both graves there were also pendants of a type belonging to the end of the pagan burial series. Then, too, there are a number of graves in England containing weapons and a sugar-loaf shield boss. In Germany some of these may be dated to the eighth century.143 In addition to the English examples enumerated by Tischler144 there is one from Rodmead Down, Wilts., in Devizes Museum, and one from Colchester in Colchester Museum. Another was recently retrieved from a grave at Bury St. Edmunds in association with a spear and two bronze buckles by Mr. A. R. Edwardson, Hon. Curator of Moyses HaU. A Saxon cemetery of the eighth to eleventh centuries is no doubt usually unrecognizable as such, as it simply consists of unfurnished graves. However, even these thoroughly Christianized cemeteries are liable to produce some chance find, possibly through a hurried burial of the person in his clothes, or through the individual whim of a mourner. Such late burials with an occasional possession are attested at Guildown 145 (a hone and coin) and perhaps, too, at Stockbridge.146 A section of Holborough cemetery, of unascertainable extent, had aheady been cut away before investigations began. Apparently the spears, etc., retrieved from these were near the cluster of find-bearing 109 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT graves to the south-east, bordering the quarry face. It may be that the graves containing possessions in the section partiaUy destroyed were the earher ones, so that the cemetery would have been extending westwards and northwards, and the unfurnished graves on these sides might represent the graves of the later and more strict Christians. But with the vital evidence of the other part of the graveyard missing, this is only conjecture and the empty graves may merely have belonged to slaves. On the other hand, if, as seems probable, the prehistoric barrow was standing proud at the time, the cemetery would have been begun on the barrow itself and would have spread to the south and east down the slope facing the settlement. It does not seem to have been of any considerable extent, if one may judge from the fact that graves were not noticed in earher commercial excavations, and that the finds recorded may aU be dated to a span of about a half-century. One may imagine, therefore, that it is a cemetery opened for the first time in the seventh century to meet the needs of two or three famUies who had settled at the foot of the hhl by the stream running into the Medway, where a charter of A.D. 838 records a mhl, and where an old mhl still stands.147 Further importance is added to the site by the fact that although the ancient terraceway known as the PUgrims' Way turns northwards along the river at this point, it does approach near to the river on both banks. A ridgeway also came down to the river here148 over the crest of Holborough Hill (Fig. I)149 and it seems that a crossing at Holborough would have been convenient at certain times of the year.160 Theh immediate neighbours do not seem to have been quite as numerous as one might expect. There are a number of pagan Saxon remains at the mouth of the Medway, but the main concentration is to be found in and near the Roman town of Durobrivae (Rochester) (Fig. 10). Some of the finds here are noticeably early, e.g. at Strood and Chatham Lines, and contiguity of the Roman and Saxon cemeteries at Strood and Higham probably indicates overlapping of the two cultures. Penetration farther up the Medway vaUey, however, seems on present evidence to have been fahly limited. Settlements along the smaU tributaries to the west and the scarp followed by the PUgrims' Way are indicated by the graves found at Holborough, Lad's Farm and Wrotham. A gravel pit at Aylesford produced the glasses and other objects in Maidstone Museum, but there are no traces of the early Saxons on the east bank of the Medway until as far up as Maidstone, where there was a cemetery of considerable size. This is surprising, for Roman remains on this side of the river are as plentiful as those on the west. The upper reaches of the Medway towards Tonbridge which must have been covered by trees on the fringes of Andredsweald are then completely devoid of any signs of the Saxons, but burials ocour 110 THE MEDWAY VALLEY - PAGAN SAXON SITES O SINGLE BURIAL • TWO Oil HORE GRAVES • HABITATION SITE O UNASSOCIATED OBJECTS SCALE IN MILES 1 0 1 l-W,- in—-J.... 2 3 4 5 i i i i - " TW »3 MEDUM STROOD ESTER. o •/<7 ROUGH VRPTHAtt© 2 AYLESFORD ^AIDSTO FIG. 10. The Medway Valley. (Based on the 1-in. O.S. map.) [face p. 110 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT along the tributary of the Len at HoUingbourne and Leeds in the valley, and at Thurnham. Boxley, Harrietsham and Lenham near the PUgrims' Way. At Harrietsham continuity in the use of pagan and Christian burial sites is shown by the discovery of a crystal ball, pottery, etc., in the churchyard. It is noticeable that the objects discovered at all the sites south of the cluster round Rochester are, with few exceptions, to be dated to the late sixth or seventh century. AU the finds are not avahable for study, and it may be that the relevant evidence is not to hand, but so far it appears that in the fifth century the Saxons were content with theh settlements at the mouth of the river on Watling Street, and that they spread southwards in relatively smaU groups during the two foUowing centuries. This is quite a different picture from that of the more eastern part of Kent where the rich cemeteries are clear evidence both of early settlement and of dense population. If the banks of the river were not occupied in the years immediately following the end of Roman rule, the areas cleared no doubt became overgrown and were so avoided for some time. Holborough cemetery, therefore, which represents a smaU and nominally Christian community of the seventh century, is not out of line with the general character of " pagan " Saxon settlements in the upper Medway valley. KEY TO SITES (Fig. 10) Location Type 1. Cliffe at Hoo Cemetery 2. Cliffe (West Court) Settlement 3. Higham Cemetery 4. Otterham Creek, near Cemetery Upchurch 5. Otterham Creek (west 1 grave side) 6. Strood 2 graves 7. Temple Farm,-Strood 2 graves 8. Strood—Cuxton 1 grave 9. Watts Ave., Rochester Cemetery 111 Reference Arch. Cant., XIII, 562. V.C.H. Kent, I, 377. Roch. Nat., No. 130, Vol. VI, 53. Arch. Cant., XXVIII, xci. V.C.H. Kent, I, 385. B.B., IV, 629. J.B.A.A., II, 347. Coll. Ant., II, 161 and VI, 260. V.C.H. Kent, I, 375. B.B., IV, 738. Maidstone Museum. Coll. Ant., II, PI. XXXVI, 158. B.B., III, 115, 138. V.C.H. Kent, I, 376-7. Arch. Cant., XXXIV, 158. Arch. Cant., XXIII, 7. J.B.A.A., II, 192. V.C.H. Kent, I, 376. B.B., IV, 741. Coll. Ant., II, 158. Arch. Cant., II, xii. Coll. Ant., V, 129. V.O.H. Kent, I, 377. Coll. Cant., 121. Arch. Cant., XXI, lv; XXII, liv; XXIV, liv. V.O.H. Kent, I, 376. B.B., IV, 741. P.S.A., XVIII, 78. AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 10. Star Hill, Roohester 11. Rochester 12. Chatham Lines 13. Gillingham 14. Lower Halstow 15. Holborough, Parrington Lane 16. Holborough, Lads Farm 17. Holborough Hill 18. Aylesford 19. Maidstone 20. Wrotham 21. Wrotham 22. Hollingbourne 23. Harrietsham Church 24. Harrietsham, Pilgrims' Way 25. Harrietsham, Court Lodge Farm 26. Leeds 27. Lenham 28. Lenham 29. Thurnham 30. Chart Sutton 31. Boxley Hills I t has not been possible to (Arch. Cant., XXXIV, 158). Cemetery 1 grave Cemetery 1 grave 1 grave Cemetery Cemetery Cemetery Cemetery Cemetery Cemetery 1 grave Cemetery J.B.A.A., IX, 408. Coll. Ant., Ill, PI. XXXIV, 1-7, 209. B.B., IV, 741. Arch. Cant., XXIII, 3. Arch. Cant., LI, 205. Nen. Brit., Tumuli, I-IV, VI, IX-XII, XVI-XX. B.B., IV, 738-40. V.O.H. Kent, I, 375-6. Arch. Cant., XLVI, 201. Coll. Cant., 60. Coll. Cant., 137. Maidstone Museum. Roch. Nat., No. 130, Vol. VI , 53, and No. 131, Vol. VI, 107. Arch. Cant., 1, 166; XV, 72 and XXX, 535. V.O.H. Kent, I, 385. B.B., IV, 741. Arch. Cant., XXXIV, 158. Maidstone Museum. J.B.A.A., III, 248 and 323. V.C.H. Kent, I, 387. B.B., IV, 741. Arch. Cant., LXV, 160-6. K.A.8. Cat., Nos. 236, 474 and 513 B.B., IV, PI. CXXXIX, 4. Roch. Nat., 131, Vol. VI , 107. Arch. Cant., XLV, xliv. Maidstone Museum. Maidstone Museum. Maidstone Museum. Ant. Journ., XX, 380-2. Maidstone Museum. Maidstone Museum. Cemetery 1 grave 1 grave 1 grave 2 graves 1 grave Cemetery 1 pot Pot and spears locate the position of the finds at Bradford Piatt THE HUMAN REMAINS By R. B. DENSTON From the Duckworth Laboratory, Faculty of Archceology and Ethnology, Cambridge The human remains from Holborough, including those of at least four individuals extensively damaged by the operations of a mechanical excavator, represent some three dozen persons, five of whom were infants, chUdren, or adolescents of indeterminate sex. Of the adults, most appeared to be fahly young, in the thhd and fourth decades of life. Sixteen of these were adjudged to be males and fifteen females, but the fragmentary nature of many specimens did not make for ease 112 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT in sexing. For the same reason very few were sufficiently complete for the customary measurements to be taken on the skulls and principal long bones. For example, only forty-five post-cranial bones were in a measurable condition, and even then no more than ten statures could be reconstructed from the lengths, five of male and five of female subjects. As frequently occurs, the skuhs were in a better state of preservation, though many had to be reinforced with plaster of paris preparatory to theh study. Tables based on the multiple regression formulae for whites of Trotter and Gleser161 have been used to reconstruct the Holborough statures. These range from 168 cm. (5 ft. 6 in.) to 179-5 cm. (5 ft. 11 in.) in the case of the males and from 159 cm. (5 ft. 3 in.) to 169 cm. (5 ft. 6 | in.) in that of the females. The respective male and female means are 175 • 1 cm. (5 ft. 9 in.) and 164 • 6 (5 ft. 5 in.), a difference of 10 • 5 cm. or 4 in. between the sexes. Using the Trotter-Gleser formulae, Dr. J. C. Trevor has kindly recalculated for this report the stature of a large series of male Anglo-Saxons from 1,066 long bones, measured by Dr. A. H. Miinter,152 which he gives as 173-4 cm. (5 ft. 8£ in.), a figure 5 cm. or 2 in. higher than that obtained by Dr. Munter from the Pearson formulas, the best avaUable at the time. It will be seen that the five Holborough Anglo-Saxon males are 1-7 cm. or two-thhds of an inch taUer than those studied by Munter, though little weight can be attached to such a result owing to the smah size of the sample considered here. The same caution must be apphed when the cranial measurements and indices of the Holborough series are examined. The maximum glabeUo-occipital length of the males, 194-9 mm., is far larger than any comparable mean as yet recorded in this country for Anglo-Saxons, but theh maximum biparietal breadth, 138-2 mm., is correspondingly smaUer. Such an increased length and diminished breadth results in a very low mean cranial index of 70-3, that of most Anglo-Saxon series so far studied lying in the region of 75. Likewise the mean value of the basi-bregmatic height of the Holborough males, 142-8 mm., is exceptionaUy large for Anglo-Saxons. But this is affected by the measurement of one very large and powerful specimen. AU the results just quoted, however, depend on no more than six individuals. Simhar measurements on the Holborough female crania do not reveal any striking departures from observations previously made on Anglo-Saxon skulls of that sex. DENTAL REPORT By H. W. NOBLE INTRODUCTORY An examination of the dental condition of the skeletons was carried out to discover if the dentitions exhibited any unusual features. 113 i i AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT Remains of the dentition were present in connection with 31 of the skeletons discovered. In some cases only a few isolated teeth were present, whhe in others there were fragments of one or both jaws and more complete sets of teeth. In 13 cases it was possible to assemble the remains satisfactorUy enough to obtain a reasonably complete picture of the dentition. Two conditions, not commonly encountered in the mouths of present-day patients, were thought to be worthy of mention because they illustrate the effect of the more primitive environment upon the dentitions of these people. Signs of a seasonal disturbance of the development of the teeth were noted, and the degree of attrition or loss of tooth subs'tance by wear was greatly in excess of that seen today. \ n> V" I l!l • I Fia. 11. The effect of the rhythmic seasonal variation in nutrition upon the surface appearance of a tooth. TOOTH CALCIITOATION The formation and calcification of the dental tissues of the permanent teeth commences at birth and continues throughout chhdhood and adolescence. This process is very dependent upon the continual presence of an adequate supply of the various food factors and vitamins which ensure the proper formation of the dental tissues. If this supply is deficient for a period of time, then, as far as the dental tissues are concerned, it cannot be made good later and evidence of this disturbance of the normal diet remains throughout life and can be seen in the erupted teeth as poorly-formed layers of enamel and dentine (see Fig. 11). Evidence of such a disturbance is frequently seen in the teeth of present-day patients who have suffered from one or other of the chUdhood fevers. 114 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT The teeth of these Anglo-Saxon people showed shght signs of such a disturbance intermittently throughout theh length, as if the essential food factors had been deficient for a certain period each year. This seasonal variation probably represents a difference between the enamel and dentine formed during the summer months and that formed during the whiter months when the vitamin D formed by sunlight and the other necessary foodstuffs were not so freely avahable. Whhe a certain degree of seasonal difference in tooth calcification is to be expected in dweUers in a temperate zone with its seasonal variation in sunlight, this Saxon farming community showed a greater degree of seasonal difference in tooth calcification than is seen in theh descendants today. It therefore seems probable that, although they were reputed to have been good tillers of the soil, theh standard of husbanding theh crops did not ensure an adequate supply of the food factors necessary for proper tooth formation and calcification during the winter months. A seasonal variation in tooth calcification has also been seen recently at Glasgow Dental Hospital in a patient who was brought up on a croft in an isolated island in the Outer Hebrides. The condition, which is most readily detected in the incisors and canines, was noted in nineteen of the specimens examined. It was not present in two specimens and could not be ascertained in the others owing to the absence of incisors or canines or to the degree of chalk erosion. ATTRITION The degree of occlusal and interproximal attrition present was most marked in the majority of the dentitions exammed. The first permanent molars were the teeth most severely affected. It was apparently common for the dentine to be exposed on the cusps of these teeth by about the fifteenth year, and by about the twenty-fifth year one or more of the original pulp horns, now filled in with secondary dentine, would appear. Simhar changes occurred at shghtly later intervals in the lower incisors, upper incisors, canines, premolars, second and thhd molars, in that order. The pattern of occlusal attrition can be seen from the diagrams of the typical cases selected in Fig. 12. The blacked-out area within the tooth outline indicates an area of exposed dentine on the occlusal surface or incisal edge. Broca163 defined four degrees of attrition as foUows (PI. III6): Fhst degree: enamel abraded without obhteration of the cusps or exposure of the dentine. Second degree: disappearance of cusps and partial exposure of dentine. 115 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT Thhd degree: complete exposure of dentine with further reduction of tooth height. Fourth degree: wear extending to neck with disappearance of the crown. From examination of these specimens it was seen that second degree attition in every tooth in the dentition normally occurred after approx- O O 8 o O O . «®. O 0 o 0 <9, G©, ®. >&& S-17 Vr» Crave I I . ° 20-25Yrs Grove 23 A^, S> o 0 ®<; m o €1 0 © Q © '•a©©^ ® * % ®. ^><3><»iS> O* 25- 3 0 Yrs Grove 21. d 30 -35 Yrs, Grove 9- Fia. 12. The gradual spread of 2nd and 3rd degree attrition over the ooolusal aspect of the dentition accompanying increase in age. 116 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT imately 30 years of age. Infrequently thhd degree attrition was encountered in the fhst permanent molar teeth on one side of a dentition, but this excessive unhateral wear could usuahy be attributed to accidental damage to the dentition on the opposite side of the mouth. The extent of attrition present in the individual teeth of a dentition corresponded roughly to the length of time that the tooth had been in use, and this appeared to be the factor responsible for the varying degrees of attrition present in the individual teeth of any one dentition. The dentition as a whole, therefore, appeared to undergo attrition evenly. The cause of such a pronounced degree of attrition is usuahy to be found in the nature of the diet and in the manner of preparation of the food. The grinding of corn in sand-stone querns with the resultant admixture of gritty particles has been cited as an important factor. There is also no doubt that the teeth and muscles of the jaws were employed more strenuously in the mastication of meat and vegetables than is the case with present-day standards of cooking. While it is generaUy beheved that nature has endowed the dental tissues with certain defence mechanisms to counteract the effects of loss of tooth substance as a result of attrition, it would appear that in these specimens the ability of the dental tissues to tolerate attrition was being fully utilized. Such a degree of attrition is rarely seen in present-day mouths, but it is interesting to note that the defence mechanisms which were evolved to counteract attrition stUl operate in a feeble attempt to hinder the progress of dental decay which appears to have supplanted attrition as the curse of the 'human dentition. GRAVE INVENTORY In these descriptions right and left refer to the right and left sides of the skeletons themselves. Grave measurements are given in the order: length, width, depth. The depth measurements are exclusive of the topsoil (9 in.) and are taken from the natural chalk level. Unless described otherwise, the bodies were extended on theh backs with legs straight and arms straight down by theh sides. The bones disturbed by the mechanical excavator consisted of at least four individuals: adult male 30-35, adult male 35-40, young adult male 20-30, young adult female 20-25. Graves disturbed by excavations for chalk: GRAVE No. 1 Part only of this grave remained at the edge of the quarry, and had been excavated. Its remaining width was 2 ft. 10 in., remaining length 3 ft. 9 in., and its depth 1 ft. 3 in. in the chalk. None of the bones or finds may be definitely attributed to this grave. 117 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE No. 2, skeleton 1. Sword (Fig. 13, 1), length 2 ft. 6 in., width c. 1 -5 in., tip missing; remains of smaU hon pommel. The radiograph shows that the blade is pattern-welded with three parallel zones. The pattern is an alternation of straight and diagonal lines as Fig. 8. Judging from the clarity of the radiograph, it seems likely that there is only a single plane of patterned zones. 2. Iron pot-hanger (Fig. 13, 2), total length 2 ft. 10 in. A hooked rod with twisted shank is linked to a second similar rod by means of a ring; to the farther end of the second rod is joined a row of five interlinked rings with a terminal doubled-over loop. 3. Bronze ring (Fig. 13,3), diam. 2-4 in.; almost round in section, but flattened on two sides. GRAVE No. 3, skeleton 1. Spearhead, length 13-5 in., slightly angular blade, split socket (Fig. 14, Gr. 3, 1). 2. Spear, socket fragment only, length 2-3 in. 3. Shield boss; one of the foUowing was found in this grave, it is not known which. The associations of the other are unknown. (Fig. 14, Gr. 3, 3), diam. 5-1 in., ht. 3 in. Small knob, convex dome, sloping waist, narrow flange with knob-headed rivets. Space between flange and end of rivet is 0 -45 in. 4. (Fig. 14, Gr. 3, 4), diam. 4-7 in., ht. 3-3 in., rudimentary knob, convex dome, carination only shght giving conical appearance, narrow flange. GRAVE NO. 4, skeleton 1. Black pottery bottle (Fig. 13, PI. lie), burnished on upper part, grey ware, rim missing, ht. 9-1 in., max. width 7-4 in., dents in walls and base. Four rows of rouletting in one continuous line; the pattern consists of a double wavy line of small oblong impressions, each of the spaces in the upper curves is occupied by a chevron and triangle impression, and the lower curves alternately by a double chevron and grid of smah squares. Graves excavated GRAVE NO. 5 8 ft. 9 in. X3 ft. Xl ft. 10 in. Young adult male, age 20-30 years. Skull to right; remains only of arm and leg bones. No finds. GRAVE No. 6 8 ft. x2 ft. 9 in. Xl ft. 4 in. Adult female? age 30-40 years. Skull slightly right; remains of leg bones and lower vertebrae. Fragments of wood above skull (? coffin). 1. Knife (Fig. 14, Gr. 6, 1) at right hip; length 4-9 in. 118 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 2. Bronze pin (Fig. 14, Gr. 6, 2) below chin; length 2 in., head missing. 3. Iron pin (Fig. 14, Gr. 6, 3) near bronze pin, length 1-7 in., point missing, head spherical. GRAVE No. 7 (Fig. 21) 9 ft. 3 in. X3 ft. at head, narrowing to 2 ft. 6 in. at foot Xl ft. 3 in. deep. Male, middle age. Remains of skull, right humerus, pelvis and both legs. 1. Spearhead (Fig. 15, 1), length 18-5 in., parallel-sided blade, spht socket. The radiograph (PI. Ilia) shows a line of lower density runnmg along each side of the blade about 0 • 3 in. from the edge. This probably means that the cutting edges were made separately and welded on to the central core. Near the angle of the blades to one side there are a few thin lines of higher density in the form p^, which must represent a pattern of inlaid whes (see p. 97 and Fig. 7). At a distance of 7 feet from the tip and in line with the blade were some shapeless fragments of rust, probably remains of the ferrule. Part of the remains of the shaft may be indicated by some black dust extending nearby along the waU of the grave for about a foot. By the spear blade was a patch of blackened earth and rust about 3 in. X4-5 in., together with some tiny fragments of bronze—perhaps a small box. To left of skuU. 2. Sword (Fig. 15, 2), length 3 ft., width c. 2 in. The hon pommel (see p. 92) 2 • 5 in. wide, bore traces on the under side of the lengthwise grain of the upper wood guard. The radiograph of this sword is not sufficiently clear to indicate the number of zones of pattern-welding, but there is evidence of both long straight lines and short diagonal lines. Lying on right femur. 3. Shield boss (Fig. 15, 3a), width 5-6 in., ht. 3-2 in., conical, shghtly convex, small knob and narrow flange; the rivets on the flange have small, knob heads, and the width of the wooden shield pierced by the rivets was 0-4 in. The grip (Fig. 15, 36) is a flat strip of hon, length 6-1 in., widening shghtly at each end. Three iron rivets (Fig. 15, 3c), domed disc heads, width 1-8 in., the shank indicating a depth of wood of 0 • 55 in. The boss was lying near the right edge of the middle of the grave at some distance from floor level, so that the point was just projecting above the grave edge. Two discs were lying together, 6 in. from the floor, at one side of the boss, and the other 3 • 5 in. from the floor at the other side. The minimum width of the shield indicated by the discs would be 1 • 7 feet. It would appear that the shield was broken before being placed in the grave, for had it been laid down flat, as the angle of the boss indicates, the wooden part would have projected over the edge of the grave. The appliques have sunk to a lower level than the 119 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT boss, perhaps because, being smaUer, they could slip easily between the lumps of chalk. 4. Knife (Fig. 16, 4). In fragments; total length about 6 in. It was lying under the lower part of the spear socket. Near the knife, approximately the position of left waist, was a group of objects amongst traces of wood: 5. Iron buckle-loop (Fig. 16, 5), width 1-4 in., round in section; remains of an hon plate folded over one side of the loop. 6. Small bronze buckle (Fig. 16, 6), length 1 • 1 in. The loop is cast in one with the triangular plate, at each corner of which is a rivet attaching a separate back plate. 7. Bronze spatula (Fig. 16, 7), length 2 • 4 in. The rod is incomplete and square in section, and widens to a flattened oval at the end. 8. Small bronze buckle (Fig. 16, 8; PI. 116), length 1 in. The loop is in one piece with the plate which consists of two bhds' heads in chip-carving with beaks touching at the end farthest from the loop. The buckle is attached to a flat fragment of wood, probably ash, as ihustrated. There are apparently three projections at the back; one is broken off, but the visible one is 0 • 3 in. long with a perforation near the end. The other is embedded in the flat piece of wood, the grain of which runs transversely across the buckle, and on top of this is another piece of material which looks like leather. The buckle was therefore obviously backed by a thin piece of wood, and it is difficult to see how the strap could have been threaded through the rigid loop. Possibly the buckle was merely decorative, like the cloisonne' buckle at Sutton Hoo. 9. Two shapeless fragments of iron rust, with various accretions including a thin hon rod, wood, etc. The two pieces fit together and an X-ray revealed the shape of a purse mount (Fig. 16, 9), length 3-6 in., one side straight, the other widening in the centre, the ends curving inwards. GRAVE No. 8 (Fig. 21) 9 ft. 3 in. X 3 ft. 9 in. at head, narrowing to 3 ft. 3 in. near foot X 1 ft. 3 in. Young adult male, age 20-30 years. Skull shghtly left, remains of arms and legs. At a depth of about 4 in. from the chalk level, black dust (? carbonized wood) began to appear in the shape of a rectangle about 7 ft. 3 in. X 1 ft. 9 in. This continued down to the floor of the grave and under the bones, and presumably represents the remains of a lidless coffin or bier. 1. Shield boss (Fig. 17, la), diam. 4-7 in., ht. 2-8 in. Convex conical shape with only a slight trace of a carination, a short point broken off, and narrow flange. The grip (Fig. 17, 16) is a strip of hon with shghtly incurving sides, 4 • 8 in. XO • 5 in. Four iron rivets (Fig. 17, 120 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT lc) each with a flat lozenge-shaped head covered with bronze, 2 in. X 1-1 in. The boss was placed on edge near the middle of the left side of the grave, pointing towards the grave waU and with its top rim on a line with the chalk surface. The lozenge-shaped rivets lay in pahs on either side with a maximum span of 1 ft. 7 in. The shield must have been placed on its edge outside the coffin and leaning against it. As the distance from the point of the boss to the floor was about 13 in., the actual width of the shield must have been at least 26 in. and more if sinkage has occurred. 2. Spear (Fig. 17, 2a). Leaf-shaped blade and spht socket, length 13-8 in. Ferrule (Fig. 17, 26), a flat piece of hon folded into a conical shape, length 4 in., max. diam. 0-6 in. The spearhead was to the left of the skull, 7 in. above the floor, and the ferrule near the foot of the grave, 2 in. from the floor, the total length of the spear being 7 ft. 6 in. 3. Knife (Fig. 17, 3), length 8-3 in. To left of position of pelvis, pointing towards the head of the grave. 4. Iron buckle loop (Fig. 17, 4), with tongue and plate. The loop is a flattened oval, round in section, width 1 • 6 in. A double rectangular hon plate, 1 -3 in. X0-9 in. folds over the loop. A radiograph showed traces of transverse whes inlaid on the loop, and a possible row of inlaid chcle-and-dot ornament along the longer side of the plate. GRAVE No. 9 7 ft. 4 in. X2 ft. 3 in. Xl ft. 5 in. Male, age 35-45 years. Skull facing left; remains of arms and legs with left forearm bent in. No finds. GRAVE No. 10 4 ft. 7 in. x l ft. 10 in. X4 in. Child, sex uncertain, age 3-6 years. Remains of skull, left humerus and femur. No finds. GRAVE No. 11 (Fig. 22) Irregularly cut, 7 ft. 7 in. X3 ft. Xl ft. 4 in. ChUd, sex uncertain, age 12-15 years. Skull tUted forward; remains of right humerus and both legs. 1. Bronze buckle (Fig. 18, 1 and PI. Ila), length 2 in., loop width 0 • 9 in. The loop, decorated by a pah of scored transverse lines each side of the tongue, is cast in one with the plate. This is decorated by perforations, the reserved metal forming a cross with curving arms near the tongue, and two rudimentary bhds' heads facing each other at the farther end. Chcle-and-dot stamps decorate the reserved bronze and form the eyes of the bhds' heads. Three rivets secure a thin back plate. At right waist. At the position of the left pelvis there was a complex of iron and bronze objects, presumably contained in a box. This is deduced from 121 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT the position of the objects which were lying in a close group, the outline of which formed a right angle. Also, amongst the hon fragments there were the foUowing, apparently parts of a handle suitable for a box: 2a and 6. Iron handle, fragments. An hon rivet with ring head (Fig. 18, 2a), length 1-6 in., retains the looped end of a handle. There is also the ring of the other rivet with the loop of the other end of the handle (Fig. 18, 26). 2c. Bronze catch (Fig. 18, 2c), possibly on the box, or on a bag contained therein; length 0-6 in.; flat bronze cut to a lozenge shape in the centre, with perforated disc ends, one of which is slotted to slide over a catch pin; scored line decoration in the centre. The box apparently contained the foUowing objects: 3. Knife (Fig. 18, 3), length 4 • 8 in. 4. Bronze needle fragment (Fig. 18, 4), length 0-9 in. Point and head are missing, but one end begins to flatten as if for the eye. 5. Bronze loop (Fig. 18, 5), length 0-6 in. The two flattened ends are fastened together by two rivets, probably originaUy on to leather; an hon ring passed through the looped end. 6. Bowl of iron spoon (Fig. 18, 6), length 1 in., chcular, with remains only of shaft. 7. There are numerous fragments of hon shafts, probably belonging to keys, for the wards also of two of these remain (Fig. 18, 7a and 6). GRAVE No. 12 Roughly cut, 7 ft. 6 in. X 2 ft. 1 in. X 7 in. Female, age 55-65 years. Skull on left side; remains of both legs with knees flexed. No finds. GRAVE No. 13 7 ft. 4 in. x 2 ft. 4 in. at head and 2 ft. 1 in. at foot Xl4 in. Male, age 20-30 years. Skull to left; skeleton fahly complete except for left ribs apparently destroyed by roots. No finds. GRAVE No. 14 This grave had aheady been excavated and part of it, 4 ft. 6 in. X 2 ft. 8 in. Xl ft. deep, was left on the chalk edge. Reported to contain nothing but bones. GRAVE No. 15 7 ft. 6 in. x 2 ft. 6 in. x20 in. Female, age 18-22 years. Skull thted backwards and lower jaw fallen forwards; remains of left and right humerus, backbone, ribs, pelvis and both legs. 1. Knife (Fig. 19, Gr. 15, 1), length 4-6 in., point missing; traces of wooden handle. At left hip, point upwards. 2. Bronze pin (Fig. 19, Gr. 15, 2), length 2-5 in.; spherical knob head with one ring moulding. Mid chest. GRAVE No. 16 8 ft. 6 in. x2 ft. 6 in. x l ft. 7 in. Male, age 20-30 years. Skull 122 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT tUted down and half left; remahis of arms and legs; left forearm bent shghtly inwards. No finds. GRAVE No. 17 7 ft. 3 in. x2 ft. 6 in. Xl ft. 10 in. Male, age 40-45 years. Well preserved skeleton, skuU to right; both arms folded across body with the right forearm on top of the left hand; right ankle crossed over left. No finds. GRAVE No. 18 (Fig. 22, PI. IV) 7 ft. 6 in. x 2 ft. 7 in. at head and 2 ft. 3 in. at foot Xl ft. 5 in. deep. This grave was considerably undercut at the foot, and the head was cut into the fiU of the eastern side of the barrow ditch. Female, age 30-40 years. ChUd, sex uncertain, age 3-4 years. The adult skeleton was lying with skull to the right. The child was on top of the right side of this skeleton, its skull covering the lower jaw and right shoulder, and its feet just outside the right knee. 1. In the topsoU above this grave was a bronze pin (Fig. 19, Gr. 18, 1). Length 1-2 in., a flat, spatulate head (broken); the shaft widens and flattens towards the centre. The fill of the grave contained a number of small pieces of Romano- British pottery and a piece of tile. 2. Near the top of the fiU at the foot was a bent nail (Fig. 19, Gr. 18, 2). Length 2-4 in.; flat, disc head and four-sided shaft. 3. Small bronze buckle (Fig. 19, Gr. 18, 3). Flattened oval loop, width 0-7 in., decorated by scored transverse lines. The plate (broken) is a rectangular sheet doubled over the loop, fixed with small rivets, and decorated by triangular nicks along the edge farthest from the loop. On the lowest vertebrae of ithe adult. 4. Bronze sheet fragments (Fig. 19, Gr. 18, 4). The largest piece is 0-8 in. X0-6 in., two of the sides being original, the others broken; two punched holes. On the left femur. 5. Iron object (Fig. 19, Gr. 18, 5). Length 4-1 in., rectangular in section at one end, shapeless (? partly missing) at the other. Under top of chUd's skull. 6. Iron pin fragments (Fig. 19, Gr. 18, 6), total length 2-3 in.; head missing. Middle of adult's chest. GRAVE NO. 19 6 ft. 9 in. x 2 ft. Xl6 in. Female, age 18-25 years. Skull right; body placed partly on right side; right hand on top of right femur; left hand in pelvis. No finds. GRAVE NO. 20 6 ft. 9 in. Xl ft. 10 in. Xl ft. 3 in. Female, age 20-30 years. SkuU thted forward and set on lower jaw; remains only of arms and legs, right forearm bent shghtly in. No finds. 123 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE No. 21 8 ft. 3 in. x2 ft. 6 in. Xl ft. 5 in. Male, age 20-30 years. WeU preserved; skull right; right forearm bent across with hand inside left femur. No finds. GRAVE No. 22 7 ft. 9 in. X2 ft. 6 in. X1 ft. 5 in. Male, age 30-40 years. Fragmentary remains of skull, arms and legs. No finds. GRAVE No. 23 6 ft. 6 in. X2 ft. x 8 in. Male, age 20-30 years. Bones fahly weh preserved; skuU bent to right and down; both hands in pelvis. A tiny fragment of colourless glass was near the right femur. GRAVE No. 24 7 ft. x 2 ft. 4 in. X1 ft. 2 in. Male, age 20-30 years. Well preserved; skull inclined shghtly down and left; left hand in pelvis. No finds. GRAVE No. 25 7 ft. 3 in. X2 ft. 9 in. at head and 2 ft. 4 in. at foot X2 ft. 1 in. deep. Female, age 20-30 years. Well preserved. Skull to left; right hand top right femur. No finds. GRAVE No. 26 3 ft. 1 in. X1 ft. 9 in. X1 ft. 2 in. No bones or finds, but a concentration of brown earth in this small grave probably indicates the remahis of a child. GRAVE NO. 27 7 ft. 3 in. X2 ft. 6 in. Xl ft. 6 in. Male, age 25-30 years. Skull to left; remains of arms and legs; both hands together at top left femur. No finds. GRAVE No. 28 (PI. IV) 7 ft. 6 in. X2 ft. 9 in. at head, 2 ft. 6 in. at foot and 1 ft. 6 in. deep. Male, age 20-30 years. Complete skeleton; skull bent down to right. No finds. GRAVE No. 29 5 ft. 4 in. X2 ft. 4 in. Xl ft. 6 in. Infant, sex indeterminate, age probably under 3 years. Bones fragmentary; skull on right side, arms slightly bent. No finds. GRAVE No. 30 7 ft. X2 ft. 3 in. Xl ft. 2 in. Female, age 20-30 years. Bones well preserved, skull tilted forward; hands on top of femurs. No finds. GRAVE No. 31 2 ft. 10 in. x l ft. 4 in. X6 in. No trace of body; probably an infant's grave. No finds. GRAVE No. 32 6 ft. 9 in. X2 ft. x l ft. 4 in. Female, age 20-30 years. Well preserved; skull turned left. No finds. 124 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE No. 33 6 ft. 7 in. x2 ft. 1 in. X l l in. Female, age 20-30 years. Bones weU preserved; skull thted forward and slightly right. No finds. GRAVE No. 34 As the chalk was being excavated, a grave apparently partly left in the chalk face was disturbed. The bones recovered belong to the lower half of the body. There was no trace of skuU or ribs, and no finds. Male, age difficult to determine, but not over 45. GRAVE No. 35 6 ft. 11 in. x2 ft. 6 in. x 9 in. Adolescent of indeterminate sex, age about 12-15 years. Fragmentary remahis of skuU, arms (bent slightly in) and legs. No finds. GRAVE NO. 36 7 ft. 4 in. X2 ft. 4 in. X1 ft. Male? age 18-25 years. Skull to right and down; remains rather fragmentary. No finds. GRAVE No. 37 7 ft. 7 in. X2 ft. 7 in. Xl ft. 3 in. Male, age 35-45 years. SkuU bent down to left; fahly well preserved. No finds. GRAVE No. 38 3 ft. 4 in. x l ft. 8 in. XlO in. Infant's grave; no trace of skeleton. GRAVE NO. 39 Disturbed by digger (at edge of chalk face). Skull and two humerus bones rescued. Female, age 30-35 years. MAIDSTONE MUSEUM (see p. 92) Small, hand-made pot (Fig. 20, 1; PI. lid), globular, with round base and sharply constricted neck with everted rim, apparently incomplete. Height 3-7 in., maximum width 3-9 in. Grey ware, red surfaces covered with a dark brown shp; faint criss-cross tooling over the upper part. Small hone (Fig. 20, 2), 2-7 in. x0-4 in. X0-3 in. Smooth greygreen stone, rectangular in section, rounded at one end; the other end is broken off at a point where the stone was perforated for suspension. Dr. P. A. Sabine, Geological Survey and Museum, reports as follows: " A sUty mudstone containing smah subangular and angular grains (up to 0-05 mm. diameter) of quartz set with flakes (up to 0 • 18 mm. long) of altered biotite in a fine-grained matrix composed of smaU flakes of biotite, chlorite and clay minerals, probably mainly illite but perhaps with some kaolinite. SmaU grains of Umenite altered to leucoxene and some smaU grains of pyrites are also present. Some siltier partings occur. " The rock does not occur in any ' solid ' geological formation in South-East England, though it could possibly be a pebble in drift deposits. The rock is of too common a type to merit an 125 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT attempt to match it. It resembles mudstones from the Lower Palaeozoic and perhaps other rocks which outcrop over extensive areas of the country, but not in South-East England." BRITISH MUSEUM, Reg. No. 1947 5-2 358 Forty-seven glass paste beads; of these twenty-one are bright yeUow and twelve are of the same type but in double segments (Fig. 20, 3a). Seven are decorated with marvered trahs: Fig. 20, 36 yeUowish-white with light green traU (2), Fig. 20, 3c brown with buff traU, Fig. 20, 3d red-brown with white traU, Fig. 20, 3e brownish-red with yeUow trahs, Fig. 20, 3/ buff with brown traU, Fig. 20, dg red-brown with yeUow traU. Two egg-shaped beads are bright green with a cream end (Fig. 20, 3h) and the rest are monochrome: Fig. 20, 3j red-brown, Fig. 20, 3k and Fig. 20, 31 white, Fig. 20, 3m red-brown with darker streaks, Fig. 20, 3n dark brown. NOTES 1 Acknowledgments. I have to thank Miss E. Meikle for the drawings illustrating this article; Mr. L. Biek, for the radiographs of the spearhead and sword; Dr. L. Mullins of Messrs. Kodak Ltd., for the print of the spearhead radiograph; Professor W. F. Grimes for advice on the prehistoric pottery; Mr. G. C. Dunning for generous help; Miss M. Blumstein, formerly Assistant Curator of Maidstone Museum, for information regarding Medway sites. Much of the comparable material quoted in the discussion was seen in museums on visits financed by a grant from the Central Research Fund of the University of London. l a Grid No. 2J-in. Sheet 51/66 698627. * This tumulus was excavated by R. Jessup in the summer of 1953: " Excavation of a Roman barrow at Holborough, Snodland," Arch. Cant., LXVIII, pp. 1-61, and the reader is referred to this article for information regarding the Roman occupation of the area, and for general historical detail. 3 D. B. Harden and R. C. Treweeks: " Excavations at Stanton Harcourt, Oxon., 1940, I I ," Oxon., X, pp. 16-41. * G. Payne, " Celtic interments discovered at Shorne," Arch. Cant., XXIV, p. 86. 6 This rustication appears to be of the Holdenhurst kind: Grahame Clark, " The Timber Monument at Arminghall and its affinities," Proc. Preh. Soc, N.S. I I (1936), PI. XI . 6 Cf. S. Piggott, Neolithic Cultures of the British Isles (1954), Fig. 49, 3. 7 Ibid., p. 308. Another, from Stevenston Sands, Ardeer (Ayrshire), is mentioned here. R. Jessup, Archceology of Kent (1930), p. 48, Fig. 7, 1. 8 C. F. C. Hawkes and C. I. Fell, " The Early Iron Age Settlement at Fengate, Peterborough," Arch. Journ., C, 188-223, Figs. 3 and 4. " C. F. C. Hawkes, " The Caburn Pottery and its implications," Sussex Arch. Coll., LXXX, 217-62, Fig. C, 8a and 9, Fig. L, 26 and 27. 10 C. F. C. Hawkes and 0 . 1 . Fell, op. cit., Fig. 6, 01. 11 Ant. Journ., XX, 117, Fig. 1. " F. H. Worsfold, " A Report on the Late Bronze Age Site excavated at Minnis Bay, Birchington, Kent, 1938-40," P.P.S., N.S. IX (1943), 28-47, Fig. 6, 1, 2 and 4. 13 G. Baldwin Brown, The Arts in Early England (1926-37), III, 132-7. 126 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 14 F. Tischler, " Der Stand der Sachsenforschung, archaologisch gesehen," Bericht der Romisch-Germanischen Kommission 35, 1954 (1956), 109. 16 Identified as probably oak, by Dr. K. Wilson. 16 1. Grave L, Valetta House, Broadstairs; Howard Hurd, Some notes on recent archceological discoveries at Broadstairs (1913), pi. on p. 25; some of the seeattas found with it, Fig. 12. 2. Sittingbourne (Dover Corp. Museum); C. R. Smith, Collectanea Antiqua (1848-80), I, PI. XXXVI , 4. 3. Cherry Hinton, Cambs., Mus. of Arch, and Ethn., Cambridge. 4. Uncleby, Yorks., Proc. Soc. Ant., XXIV, pi. facing 152, 7; for grave group, see E. T. Leeds, Early Anglo-Saxon Art and Archceology (1936), PI. XXVII, 31. 5. Kingston Down, Grave 300, B. Faussett, Inventorium Sepuichrale (1856), PI. IX, 7. 6. Sibertswold Down, Grave 131, ibid., PI. IX, 12. 7. Sibertswold Down, Grave 150, ibid., PI. IX, 8. 8. Sibertswold Down, Grave 176, ibid., PI. IX, 13. 9. Sibertswold Down, Grave 144, ibid., PI. X, 1. (This consists of a bronze loop, but an ivory or bone plate.) 10. Breach Down (probably), British Museum Reg. No. 79 5-24 54. 17 E. T. Leeds, op. cit. (1936), pp. 103-4, where comparison is made to Alemannic buckles with similar animal heads. 18 Ant. Journ., XXXV, p. 33 and PI. IXa and 6. 18 R. Jessup, Anglo-Saxon Jewellery (1950), PI. XXVI I I , XXX and XXXI , 1. 50 Ibid., PI. IX, 3 and 6. These crosses also are decorated with ring-and-dot motif. 21 T. C. Lethbridge, Recent Excavations in Anglo-Saxon Cemeteries in Cambridgeshire and Suffolk (1931), Fig. 36; cf. also Fig. 29. 22 J. Werner, Mttnzdatierte Austrasische Orabfunde (1935), Taf. 11, A.6; 12, A.9; 18, 12; 23, 6; 27, B.a; 35, A.S; 36, A.2a. 83 C. Boulanger, Mobilier Fun&aire (1902-5), PI. 45, 5. 24 B. Almgren, Bronsnycklar och Djurornamentik (1955), pp. 17-19; cf. PI. 3 and Table I, E-F. 26 W. A. von Jenny, Die Kunst der Qermanen (1943), PI. 61. , 0 Germania, 14, 166. 27 British Museum, The Sutton Hoo Ship Burial (1947), PI. I. 29 J. Ypey, Kunst en Schoonheid (1955), PI. 72. 20 B. Salin, Die Altgermanische Thierornamentik (1935), Fig. 314. 30 J. Brandsted, Danmarks OUtid, III (1940), Fig. 268, p. 291. 31 W. A. von Jenny, op. cit., PI. 59. 32 E. T. Leeds, op. cit. (1936), PI. XVI I I , 6. 33 B.M. Guide to Anglo-Saxon Antiquities (1923), Fig. 42. 34 E. T. Leeds, op. cit. (1936), PL XVIII, a. 36 T. D. Kendrick, Anglo-Saxon Art (1938), PI. XXXI , 6 and 7. 30 Reg. No. 1102 70. 37 B. Faussett, op. cit., PI. X, 3. 38 Ibid., p. 153, Fig. 7. The nicks are actually more like the Holborough example than seems from this sketch. 39 T. C. Lethbridge, op. cit. (1931), Grave 83, Fig. 33, 6; Grave 85, Fig. 18B, 4; Grave 97, Fig. 35, 3. 10 Ibid., Grave 3, Fig. 22, 3; Grave 6, Fig. 22, 4. 41 T. C. Lethbridge, A Cemetery at Shudy Camps, Cambridgeshire (1936), Grave 11, Fig. 2, 6. 42 G. Baldwin Brown, op. cit., PI. XCVII, 7. Grave 142. 43 B. Faussett, op. cit., Beakesbourne, Grave 29, p. 152; Grave 38, p. 154; Sibertswold, Grave 180, p. 133. 44 T. C. Lethbridge, op. cit. (1931), Grave 70, Fig. 18, A.3; Grave 85, Fig. 18, B.3. 46 Ibid., BurweU, Grave 6, Fig. 22, 6; Grave 61, Fig. 29, 3; Grave 83, Fig. 33, 8; Grave 97, Fig. 35, 1. Iron spoon—Burwell, Grave 90, Fig. 30, 6; Grave 83, Fig. 33, 1. 40 T. C. Lethbridge, op. cit. (1936), Shudy Camps, Grave 10. 47 B. Faussett, op. cit., Grave 60, iron spoon, PI. 12, 10, and J. R. Mortimer, Forty Years' Researches (1905), PI. LXXXV, 646. 127 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 48 T. C. Lethbridge, op. cit. (1931), Holywell, Fig. 18, A. l ; BurweU, Fig. 27, 7; op. cit. (1936), Shudy Camps, Fig. 1, A.5, Fig. 1, B.2 and Fig. 2, 4. 48 e.g. T. C. Lethbridge, op. cit. (1931), Holywell, Fig. 3, 3. 60 London Museum Catalogue: London in Roman Times (1930), PI. XXXVI I , 5. 61 T. C. Lethbridge, op. cit. (1931). The bosses from Graves 4 and 29 are not Ulustrated; Grave 56, Fig. 21, B , No. 2. The boss from Grave 91 is not illustrated; it is a taUer cone than Grave 56, having a maximum diameter of 5 in. and height 3-4 in. 62 e.g. B. Faussett, op. cit., p. 65, Fig. 3. 53 Southend-on-Sea Ant. and Hist. Soc, I, plate opposite p. 121. 64 Archceologia, LX, Fig. 3. 66 G. Baldwin Brown, op. cit., I l l , PI. XXI I , 2 A and B. 66 Devizes Museum Catalogue, p. 250, PI. LXXXIVa, 1. G. Baldwin Brown, op. cit., IV, 656. 67 Leicester Museum. 68 London Museum, recent acquisition (1956). 69 Proc. Coventry and Dist. Nat. Hist, and Scientific Soc, II, Fig. 3. 00 J. Douglas, Nenia Britannica (1793), PL XXV, II. 61 D. M. WUson, " Initial excavation of an Anglo-Saxon Cemetery a t Melbourn, Cambs.", Proc. Camb. Ant. Soc, XLIX, Fig. 2. 62 Devizes Museum Catalogue, p. 243. 63 Devizes Museum, No. 299.305. 64 H. Zeiss, " Spatmerowingisch-Fruhkarolingische SchUdbuckel von Zuckerhutform," Festschrift filr P. Reinecke (1950), p . 173; P. la Baume, " Die Wikingerzeit auf den Nordfriesischen Inseln," Jahrbuch des Nordfriesischen Verei/ns filr Heimatkunde und Heimatliebe, 29, pp. 23-9, Kar t e 3. 66 G. Baldwin Brown, op. cit., Ill, p. 198, FoUcestone; B. Faussett, op. ci'., p. 5, Grave 5 and p. 11, Grave 23, Gilton; p. 63, Grave 124, Kingston; p. 115, Grave 81, Sibertswold; Howletts, British Museum. Studs occurred in Tumulus I a t Chatham, J. Douglas, op. cit., pp. 3-4, and the shield boss was probably of the smaU conical type as the diameter was only 4 in. 06 J. R. Mortimer, op. cit., Figs. 749a and 757. 67 T. C. Lethbridge, op. cit. (1931), Holywell Row Grave 56, p . 30. 08 E. T. Leeds and H. de S. Shortt, An Anglo-Saxon Cemetery at Petersfinger, near Salisbury, Wilts. (1953), Graves LVIII and LX, Fig. 9. 09 H. Stolpe and T. J. Arne, La Nicropole de Vendel (1927), PI. XLI , 3 and 4. 79 W. Krause, Runenmschriften im alteren Futhark (1937), Chapter II. For further notes on the runes on the Mos spear see C. J . S. Marstrander, " De Nordiske Runeinnskrifter i eldre alfabet," Viking, 1952, pp. 164-5. There is, in addition, an inscription on a slate spearhead from ElgsjS, Lappland, Fornviinnen, 1907, p . 257, and C. J . S. Marstrander, op. cit., pp. 276-7, Fig. 235. 71 C. J. S. Marstrander, " De Gotiske Runeminnesmerker," Norsk Tidsskrift for Sprogvidenskap, 3 (1929), 136. 72 Ibid., 26-35, Fig. 1, and summary in French, p. 149, but cf. W. Krause, op. cit., p. 442. 73 Ibid., Abb. 14. 74 S. Bugge and M. Olsen, Norges Indskrifter med de celdre Runer (1891-1924), I , p. 415, photographs on pp. 413, 415 and drawing p. 416; also C. J. S. Marstrander, op. cit. (1929), Fig. 55. 76 Ibid., pp. 130-5. 70 Kossinna, " Uber verzierte Eisenlanzenspitzen als Kennzeichen der Ost germanen," Zeitschrift filr Ethnologic, Band 37 (1906), pp. 369-407. I have not been able to consult the Polish publications mentioned. 77 Variscia, 1829, I, Taf. I l l , 1. 78 The unusual figures on the Elgsjo spear are thought to be later additions. 78 O. Almgren and B. Nerman, Die Altere Eisenzeit Qotlands (1923), Fig. 601. 80 V. I. Evison, " Early Anglo-Saxon Inlaid Metalwork," Antiquaries Journal, XXXV, PI. IIIc. p. 41. 81 A. Dasnoy, " Les premieres damasquinures merovingiennes de la region namuroise," Annates de la Soc. Arch, de Namur, XLVII, PI. IV. 128 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 82 W. Veeck, Die Alamannen in Wurttemberg, 1931, S.82, 303, Taf. N.8 and 0.8, and W. Krause, op. cit., Abb. 129, S.652-3. 83 W. Veeck, op. cit., Taf. 71, B.7 and Taf. 78, A.3. 84 C. J . S. Marstrander, op. cit. (1929), pp. 179-80; L. Jacobsen and E. Moltke, Danmarks Rune-mdskrifter (1942), No. 18. 86 The runes TE occur in two bracteate inscriptions, where i t has been suggested they might represent " tiwaR " and " ehaR," i.e. Odin and his attribute the horse; W. Krause, op. cit., pp. 460-2; see also p. 432. 86 Cotton Cleo. A.HI: T. Wright, Vocabularies, Vol. I I (1873), p. 58. 87 L. Jacobsen and E. Moltke, op. cit., Column 304, Fig. 599; and E. Moltke, " Mediaeval rune-amulets in Denmark," Acta Ethnol., Ill, 134. 88 G. Stephens, Old Northern Runic Monuments of Scandinavia and England (1866-1901), Vol. II, pp. 859-61. 88 Finnur J6nsson, Scemundar Edda (1926), Sigrdrifumal, verse 5, p. 308. 90 There are, of course, other interpretations of this, see W. Krause, op. cit., pp. 432 and 477-8; cf. a gold ring from Austria in the Ashmolean Museum, Ant. Journ., XX, 329, but the specific instructions in the Eddie poem, combined with the fact that some runic inscriptions do in fact end with a multiple t-rune superfluous to the wording, are very good evidence. 91 F. Mosse, " L'origine de l'Ecriture Runique," Conferences de I'Institut de Linguistique de I'Universiti de Paris, X, 64-5. 92 W. Krause, op. cit., Nydam spears, Abb. 19 and 20, Skaangsten, p. 570. 92a Dr. R. Drogereit has suggested a connection with the Christian tau in runic form. 93 J. Grimm, trans. J . S. Stallybrass, Teutonic Mythology (1900), p. 200. 91 E. v. K. Dobbie, The Anglo-Saxon Minor Poems (1942), p. 29, 11.48-50. For the usage of " sum " in this passage, see C. L. Wrenn, Beowulf, p. 290. 86 As the MS. Cotton Otho B.X. was burnt in the 1731 fire, the only source is G. Hickes's printed version, Thesaurus (1705), P t . I, pp. 143-5. The names of the runes were apparently added by Hickes from another MS.—see E. v. K. Dobbie, op. cit., pp. xlvi-xlvn and p. 157. 90 For recent work on this technique, see H. Maryon, " A Sword of the Nydam Type from Ely Fields Farm, near Ely, Cambridge," Oamb. Ant. Soc. Proc, XLI, 76. A. France-Lanord, " La fabrication des epees damassees aux epoques merovingienne et carolingienne," Pays gaumais, lOe annee, Nos. 1-2-3 (1949). P. Lebel, " Provenance et diffusion des epees damass6es," Rev. Arch, de VEst et du Centre- Est, tome 4, p. 149. A. Liestal, " Blodrefill og Mil," Viking, 1951, pp. 71-98. Other blades found in England are mentioned in Arch., XCVI, pp. 167-9. 9°a I am indebted to the British Hat and Allied Feltmakers Research Association for examining the wool. 97 B. Faussett, op. cit., PI. XV, 22 and p. 172. The iron links illustrated on PI. XV, 23 were probably part of a similar chain. There are also one or two iron hooks and interlinked iron rings in the British Museum, recorded as being Anglo- Saxon iron from Kent. 98 E. Brenner, " Zur Geschichte des Kesselhakens," Maimer Zeitschrift, V, 1910, pp. 50-3. 99 W. B. Dawkins, " On Bigbury Camp and the PUgrims' Way, " Arch. Journ., LIX, 212, PI. II, Fig. 6. 100 J. Curie, The Fort at Newstead (1911), PI. LXIV, 3. 101 V. Gross, La Tene: un oppidum Helvite, PI. VIII, 3; P. Vouga, La Tine (1923), PL XXVI I , 5. See also: from Duhren, L. Lindenschmit, Die Alterthumer unserer heidnischen Vorzeit (1864-1911), V, Taf. 15, 284; l'Abbe J. PhUippe, " Le Fort-Harrouard," VAnthropologic, T.46 (1936), Fig. 40. 102 L. Jacobi, Das Romerkastel Saalburg, 1897, Taf. XTV. 103 S. Reinach, Catalogue Illustri du Music des Antiquitis Nationales (1926), T.I, Fig. 281, 25795. 104 E. Brenner, op. cit., Fig. 2. These and a more compHcated type are listed and discussed, with distribution map in S. Piggott, " Three metal work hoards of the Roman period from southern Scotland," Proc S.A. Scot., LXXXVII, pp. 13-14, 24-5, 42-4 and Fig. 3. 129 12 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 105 B. Nerman, Die Volkerwanderungszeit Gotlands (1936), Taf. 57, 611. 106 H. Stolpe and T. J. Arne, op. cit., PL X, 9, XXXI, 5, 6; G. Arwidsson, VaUgdrde 8 (1954), Taf. 31, 838. 107 Cf. also J. Petersen, Vikingetidens Redskaper, Fig. 219, pp. 410, 416. 108 Ibid., PL XXII, 7, 8 and 9; S. Grieg, Gjermundbufunnet (1947), PL XIII, 2. 109 Oudheidk. Mededeel, XI, Afb. 62, 93. There is also an iron chain of s-shaped and oval loops, Afb. 62, 92. 110 A. Bjarn, " Bronsekar og Glass begre fra FoUcevandringstiden," Det Kongelige Norske Videnskabers Selskabs Skrifter, 1929, Nr. 6, p. 24; G. Ekholm, " Bronskarlen av Ostland- och Vestlandstyp," ibid., 1933. H. Norling-Christensen, " Vestlandskedler og malede glas," Kuml, 1953, 47. G. Ekholm, " Neues iiber die Westlandkessel," Trierer Zeitschrift, 23, 224-30. 111 B. Faussett, op. cit., PL XI, 17. 112 Pitt-Rivers: Excavations in Cranbourne Chase (1888), II, PL CL, 20. 113 C. W. PhUlips, " The excavation of the Sutton Hoo Ship burial," Ant. Journ., XX, 174. 114 Arch. Cant., VII, PL IX and X. 115 B. Faussett, op. cit., p. 137, PL XX, 4 and 4a, buckle PL IX, 14. The simpler rouletting pattern on the pot from Grave 117, Sibertswold, PL XX, 5, may be compared. 116 G. Baldwin Brown, op. cit., IV, PL CXXXIX, 4. 117 Bulletin archiologique (1907), 25, PL VII, Fig. 4; also in the grave were a sword, knife, hammer-axe, tweezers, pricker, comb and buckle with shoe-shaped rivets. 118 A. Bourgeois and R. TerraUlon, " Le cimetiere merovingien de Mazinghem," Bull. Trim, de la Soc. Acad, des Antiquaires de la Morinie, XVII, PL III, 7, PL IV, 20 and PL V, 24, 25. 118 Arch. Cant., XXXIX, PL I, left and right. 120 Ibid., centre. 121 P.S.A., XXIII, p. 279, Fig. 4. 122 Baron de Loe, Belgique Ancienne, IV, PL 83, 3. 123 J. N. L. Myres, " Romano-Saxon Pottery," Ed. D. B. Harden, Dark Age Britain (1956), Fig. 2, 5 and 6. 124 B. Faussett, op. cit., PI. XX, 8. 126 J. R. Mortimer, op. cit., PL LXXXV, 654. 120 Arch. Cant., LXVIII, p. 128, Fig. 15. 127 Richborough, III, PL XLII, 362. 128 G. C. Dunning, Proc. I.O.W. Nat. Hist, and Arch. Soc, II (1937), 683-95; K. M. Kenyon, Excavations at the Jewry Wall Site, Leicester (1948), 230-2, Trans. Shrop. Arch. Soc, LIV, 142-4, Arch. Cant., LXIX, 152-5; E. M. Jope, Oxon., VIII and IX, 106, Berks. Arch. Journ., LI, 70; L. R. A. Grove, Arch. Cant., LXVIII, 157-8. 129 D. Atkinson, Report on Excavations at Wroxeter (1942), pp. 129-30, PL 44a. K. M. Kenyon, op. cit., 273, PL XXII. 130 J. E. Morey and K. C. Dunham, " A Petrographical study of medieval hones from Yorkshire," Proc Yorks. Geol. Soc, XXIX (1953), 141-8. 131 It is suggested in this publication that as there are no other goods with the skeleton the whetstone may have been displaced from another, an earlier, burial, but as it was found in an appropriate position—by the pelvis—there appears to be no reason why it should not be in its original place and so represent an eleventhcentury burial. 132 Garranes, P.R.I.A., XLVII, C.77, Fig. 12. Ballycatteen, P.R.I.A., XLIX, C.35, Fig. 10. Lagore, P.R.I.A., LIII, C.184, Fig. 92. Bantham, Ant. Journ., XXXV, Fig. 4, 5. 133 Proc. Suff. Inst. Arch., 25, 57-8. 131 Oudheidk. Mededeel., XI, 85, Afb. 66. H. Jankuhn, Die Ausgrabungen in Haithabu (1943), 130, Abb. 68. H. Arbman, Birka, Die Grdher, Taf. 186-8. H. Stolpe and T. J. Arne, op. cit., PL VII, 9, XI, 6, XV, 10, XXI, 13 and XXV, 9. These are often made of slate: J. Petersen, op. cit., pp. 620-1; A. Banteknann, Tofting, cine vorgeschichtliche Warft an der Eidermilndung, 1965, p. 74. 130 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT 135 T. C. Lethbridge, op. cit. (1931), pp. 82-4. 130 T. C. Lethbridge, op. cit. (1936), pp. 27-9. 137 Proc. Oamb. Ant. Soc, XLIX, 29-41. 138 E. T. Leeds, op. cit. (1936), pp. 96-114. 139 B. Faussett, op. cit., PL VIII, 7; see T. C. Lethbridge, op. cit. (1936), p. 29. no j?_ Henry, Early Christian Irish Art (1954), p. 40, and ibid., " Irish Enamels of the Dark Ages and their relation to the Cloisonne techniques," Ed. D. B. Harden, Dark Age Britain, pp. 85-6. Ant. Journ., XXXV, PL X, XI. 141 P.R.I.A., LIII, pp. 129-31. 142 Devizes Museum Catalogue (1934), PL LXXXII. 143 E. T. Leeds, op. cit. (1936), PL XXXI. 144 F. Tischler, op. cit., pp. 117-20. 145 Ibid., Note 59, p. 123. 140 A. W. G. Lowther, " The Saxon Cemetery at GuUdown, GuUdford, Surrey," Surrey Arch. Coll., XXXIX, pp. 1-50. 147 N. Gray HU1, " Excavations on Stockbridge Down," Proc. Hants. Field Club and Arch. Soc, XIII, pp. 247-59. 148 Birch, Cart. Saxonicum, No. 418. 149 Arch. Cant., LXVIII, Fig. 1. 160 I. D. Margary, " The North Downs main trackway and the PUgrims' Way," Arch. Journ., CIX, pp. 417-49, with map, Fig. 6. 161 MUdred Trotter and Goldine C. Gleser, " Estimation of Stature from Long Bones of American Whites and Negroes," Amer. J. Phys. Anthrop., N.S., Vol. 10 (1952), pp. 463-514. 162 A. Heinrich Munter, " A Study of the Lengths of the Long Bones of the Arms and Legs in Man, with Special Reference to Anglo-Saxon Skeletons," Biometrilca, Vol. 28 (1936), pp. 258-94. 163 Broca, P. (1879), " Instruction relatives a l'etude anthropologique du systeme dentaire," Bull. Soc. Anthrop. Paris, 2, Ser. 3, 149. 131 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE 2 «2fP On* «£&•>. °*T m X s i ) ' 5=rf =#>; - . - . X 4 * w ^ i.Uii.iliJl.i..i„ FIG. 15. Grave 7, 1 J, 2 J, 3a-c J. 134 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE 7 Ife W M FIG. 16. Grave 7, 4 and 9 J, 5-8 J. 135 GRAVE 8 i ««... *<® '-^J1"l"l"1l"lllvl"vniniH)iiP FIG. 17. Grave 8 J. 136 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT llH»»HI ®s^r\® < —f T * <~1L. ^K© © GRAVE II > H „ ' ^L, r x; i \M" , FIG. 18. Grave 11 }. 137 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT GRAVE 15 GRAVE 18 FIG. 19. Grave 16 J; Grave 18 |- except 2, 138 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT l _ „ ! I I ^Q I & i B 'C i D < E • G ' H D' ©L M N FIG. 20. Unassooiated finds, 1 J, 2 and 3 J. 139 AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT wood 9-\ ..5 « » • & <% spear shaft? •"//»////' »/' GRAVE 7 GRAVE 8 FIG. 21. Graves 7 and 8, plan fe. 140 PLATE IV (II) Vertical photograph of Grave IS. (6) Vertical photograph of Grave 28. AN ANGLO-SAXON CEMETERY AT HOLBOROUGH, KENT n 2[q+b-« \ l*fc^7 woodenif'" f box J B GRAVE II GRAVE 18 FIG. 22. Graves 11 and 18, plan fg. 141