Search

This paper explores the cross-cultural portrayals of an unusual and striking musical instrument, the carnyx, on the coinages of the Romans and the inhabitants of Iron Age Britain and Gaul. Fashioned as a snarling boar, the carnyx was a war horn used by the Gauls and Britons that not only captivated the minds of their artists, but also those of the Romans. This paper studies the cross-cultural phenomenon of its appearance in the coin iconography of the late second to late first centuriesbc. This simultaneous analysis of Roman, Gallic and British coinage reveals that while each culture had a shared belief in the carnyx’s military role, each culture also had its own interpretation of the object’s significance. To the Romans, it was a symbol of the barbarian, to be cherished as a war trophy after a Roman victory, but to those northern Europeans, it was a sign of pride and spiritual significance. An image’s meaning is, therefore, seen to transform as it crosses into a new cultural context.

In 1998 the author published ‘Genetics, linguistics and prehistory: thinking big and thinking straight’, a critique of late twentieth-century attempts to synthesize the disciplines of genetics, linguistics and archaeology. This paper assesses subsequent progress, using examples from various parts of the world, including Britain, Denmark, Finland, France, Frisia, Germany, Hungary, Ireland, Italy, Micronesia, Portugal, Spain and the Canary Islands. The growing importance of mitochondrial DNA and the Y chromosome, rather than classical population genetics, is emphasized. The author argues that ancient DNA and early linguistic data should be used more. Languages mentioned include Aquitanian, Basque, Celtiberian, Etruscan, Finnish, Hungarian, Iberian, Lepontic, Lusitanian, Pictish, Raetic, ‘Tartessian’, Thracian and the Ladin dialect of the Italian Alps. Aspects of the ancient linguistic geography of Scotland and the Iberian peninsula are discussed, as is the difficulty of deciding the direction of spread of Indo-European and non-Indo-European languages. The potential of ancient place and personal names is illustrated from Celtic.

The decorative motifs of palmettes and interlocking S-patterns on the Battersea Shield have a clear place in a well-understood tradition of Celtic art, while the plan of the shield – its schema – has not had such a place. Its essential form is a deliberately constructed configuration crafted specifically by a constructional geometry, within a tradition which persisted and later pervaded Insular art.

The Welshman Edward Lhuyd (?1659/60–1709), Keeper of the Ashmolean Museum, was a naturalist, philologist and antiquarian. He wrote the Welsh additions to Camden's Britannia (1695) and undertook extensive research for an Archaeologia Britannica. He was part of the scientific revolution centred on the Royal Society and was influenced by the flowering of Anglo-Saxon studies in late seventeenth-century Oxford. Although many of his papers were destroyed, sufficient evidence survives to assess his methodology for recording early medieval antiquities – particularly inscribed stones and stone sculpture in Wales and other Celtic areas – as well as his analysis of them. His legacy is of considerable importance and he may be regarded as the founding father of early medieval Celtic archaeology.

Mamucium, the Roman name of Manchester, is often explained as ‘place on the breast-shaped hill’ from the hypothetical British mamma ’breast; breast-shaped hill’. But the name of Mansfield, Nottinghamshire, suggests this is baseless. A likelier etymology is ‘place on the river called Mamma, “mother”’, apparently the old name of the River Medlock, perhaps revered as a Celtic goddess.

One of the objects on display in the ‘Celtic Iron Age’ gallery of the British Museum is so small that in fact it may not be immediately appreciated for what it is – a cast bronze knife with exquisite curvilinear decoration which measures a mere 11omm in length (figs Ia−b2a−c).To date only the briefest of descriptions of the knife have appeared in print accompanied by illustrations which, it must be said, hardly do it justice.2 Thus, some further notes may be added to draw further attention to one of the miniature insular masterpieces of what, despite the best efforts of many, is still widely termed ‘early Celtic art’.

In 1960 and 1961 Ole Klindt-Jensen published two short notes about a golden statuette of a Celtic warrior, soldered to a fine brooch. He was convinced that the warrior did not belong to the brooch, and thought that they had been combined fairly recently. J. M. de Navarro added a comment to the 1961 note, concluding: ‘My impression (from photographs only) is that the brooch might date from the fourth century BC and the figure not before the latter half of the third or the second century BC, i.e. that it was added later.’ Klindt-Jensen's notes were accompanied by plates, and at the same time another photograph was published on the front cover ofCelticum, volume I. A decade later the brooch was shown at the Royal Scottish Museum, Edinburgh, and at the Hayward Gallery, London, as item no. 35 in the Early Celtic Art exhibition held in 1970. The catalogue entry was based on Klindt-Jensen's note, but no photograph was published. In the mid 1970s R. M. Rowlett prepared a paper on the authenticity of the brooch, including metal analyses and a comparison of its proportions with those of some La Tène II brooches: Rowlett considered that the figure of the warrior was contemporary with the rest of the brooch, which he accepted as an authentic antiquity. His paper was eventually published about twenty years later.

Two recent discoveries have added greatly to our knowledge of Iron Age shields in Britain. First, the Chertsey Shield, unique in the Celtic world in being made entirely of bronze. Second, a collection of miniature shields of a shape hitherto unknown. Bindings from full-size shields of this shape had been found previously, but were not correctly identified until the miniatures provided the clue. Now it seems that these distinctive hide-shaped shields were one of the most common forms used by British warriors. At least thirty-eight examples—including twenty-three miniatures—have now been added to the corpus of Iron Age shields from Britain.

The limestone head from the Bon Marché site, Gloucester, has long been considered one of the finest examples of Romano-Celtic sculpture. A detailed description is accompanied by a discussion of its original architectural context, and the possibility of a Romanesque date is argued. The essential similarities between many simple facial representations of different dates is, however, emphasized.

SummaryA re-examination of the fragmentary remains from Cerrig-y-Drudion suggests that they belonged to a pair of similar vessels which may have been lids but were certainly not hanging bowls. The recognition of a mistake in the previous reconstruction allows some of the pieces to be rearranged to produce a more or less symmetrical reconstruction of the design on one of Britain's most important pieces of Early Celtic art.

The ‘Celtic Cabinet’, acquired from John Britton when the Wiltshire Archaeological and Natural History Society and Devizes Museum were founded in 1853, is published for the first time. The cabinet, with its watercolours and plaster models, is described. The megalithic subjects of the watercolours which decorate it are identified; two of the paintings are ascribed to John Sell Cotman; the others are related to topographic artists in John Britton's employment. The models in the cabinet of Stonehenge and Avebury are identified as standard products of Hen Browne, the first guardian of Stonehenge. The meaning and associations of ‘Celtic’ as a name for megalithic antiquities are explored. It is shown that the cabinet was made, probably about 1824, for George Watson Taylor of Erlestoke Park, Devizes; the circumstances of its making are discussed, together with the intellectual context of the piece and its relation to Britton's plan for a Druidical Antiquarian Company.

SummaryThis paper examines the two dominant cults of Sky and Fertility in the Romano-Celtic world. There are no literary allusions to a link between the two concepts, but a substantial body of archaeological data does exist which testifies to a close iconographic association between the images and symbolism reflecting beliefs in celestial and fertility powers. The most important material evidence consists of a group of small clay statuettes representing female figures decorated with what are probably solar symbols. Allied to this is evidence for an association between the Celtic Mother-Goddesses (theDeae Matres) and cosmic motifs, particularly the solar wheel. It is also possible to identify traces of a fertility aspect to the cult of a male solar divinity. Whilst the simplistic argument for a divine ‘marriage’ between sky-father and earth-mother (Hatt 1965) is not here advanced, there is, nevertheless, strong evidence from Gaul and Britain that the two major cults of the Romano-Celtic world were clearly intertwined. This association need not be such a paradox as it might seem; the sun, like the rain, was essential for crop-growth, and the iconography appears to demonstrate acknowledgement of, and gratitude for, this interaction between natural—or supernatural—phenomena.

This paper describes the context and nature of an assemblage of forty-four Palaeolithic hand-axes from a Roman religious site at Witham, Essex. The hand-axes are considered to have been derived from several sources, and it is suggested that the Romano-British occupants of the site deliberately selected them for their shape and placed them in the bottom of two large man-made depressions. In the light of stone axe finds on continental temple sites, and of classical Roman texts and traditions, the possibility arises that the Witham finds may have represented ‘thunderbolts’ in the worship of Jupiter or a local Celtic equivalent.

SummaryTwo La Tène I swords found in the River Thames more than a century ago are in the remains of iron scabbards decorated with dragon-pairs. These designs, which have only recently been revealed, are similar to those on La Tène scabbards throughout Celtic Europe from France to Romania: ninety examples can now be listed, and half of them have been added in the last decade. The discovery of dragon-pairs in the Thames emphasizes the links that must have existed between Britain and the rest of the Celtic world in the decades around 300 B.C.

SummaryThe Carmarthen brooch, datable not later than c.a.d. 50 and perhaps as early as c. 25, is a direct two-piece successor to the pre-Roman, one-piece, Aylesford type, now known to be considerably later in date than used to be thought. The two-piece construction, which characterizes the entire British trumpet-brooch series, was adopted as a convenience when the bow was given a massive form in order to accommodate Late Celtic decoration in relief, of a western character and antecedents. The central moulding and foot carry petalled ornament of a ‘transitional’ or ‘supporting’ type: this ornament is reminiscent of that sometimes found between the circlet and terminals of early La Tène torcs abroad, but is probably better understood here as the carefully-designed continuation en suite of the open flower or rosette upon the headloop, and thus as half-open flower. Only the Carmarthen brooch retains such a headloop rosette.The Carmarthen brooch was soon copied in bronze, and on these copies the relief appears in a coarsened and simplified form. The copies are entirely from western findspots. In the southern part of the region in which they occur, the medial petalling is retained (as at Lydney); but in the northern part, additional simplification has led to the suppression of the medial petalling (as at Segontium). Thus the Carmarthen brooch may be seen as the parent of the more fully-developed Rii class of Collingwood's typological sequence, and also of at least one strand in the Ri class, which has been considered up to now as the original form. In the west Midlands, perhaps seizing the opportunity offered by the simplified relief, craftsmen began to apply red enamel to the hollow areas, and to emphasize the central moulding, with its petalling, so that this became the principal plastic ornamental element (as at The Lunt, by a.d. 75). In the north of England, these novel features, enamel and exaggerated moulding, were further developed. The fine northern brooches (Backworth—not enamelled, but engraved—or Risingham) do not have headloop rosettes, and it was in the presence of their exaggerated central mouldings that Collingwood devised his celebrated ‘acanthus’ theory. By the Antonine period, brooches were being produced with polychrome enamel, and with their mouldings confined, very often, to the front of the bow; many of these have been found along the Rhine. Some late derivatives of the Carmarthen type are known from Wroxeter, and form a final group in the Appendix above.

One of the most spectacular pieces of martial equipment in use among the Celtic peoples in the later stages of the La Tène culture was the animal-headed war-trumpet, the name of which, in Greek versions, has been preserved variously as karnon or karnyx. In the latter form, the name carnyx has been applied by archaeologists to the fairly plentiful representations and the very few surviving fragments of such instruments. Of the latter, the best-known is that represented today only by drawings and engravings, dredged from the river Witham at Tattershall Ferry in 1768. It is the purpose of this paper to discuss this object once again, and also to put forward the suggestion that the sheet-bronze object in the form of a boar's head, found at the beginning of the nineteenth century at Deskford in Banffshire and fortunately still surviving, is in fact the mouthpiece of another carnyx.

The Colchester Celtic mirror has been known to students of the period since a photograph of it was published by Henry Laver in the Proceedings of our Society (x·x, 1905, p. 213) as one object in an important grave-group. Laver noted that the back of that portion of the mirror-plate which is preserved is ‘ornamented with a spiral pattern’, but this pattern is not to be seen in the reproduction. Another reference in the literature is that by R. A. Smith, Archaeologia, lxi, 338, who mentions ‘traces of engraved scrollwork’ and an ‘imbricated filling, not of the usual basket-pattern’. Mr. E. T. Leeds produced another group photograph of the find; Celtic Ornament, fig. 10 and p. 30.The writers of this article collaborated for the purpose of studying the ornament, and of reconstructing the complete design, if that should prove possible. The incised work was found to be very difficult to follow owing to a pustule-like corrosion all over the mirror fragment. Thus, as one of us remarked after the first attempt, ‘it is only here and there, either in a strong cross-light, or with the lens, that one can see and copy it’.Thereafter both of us worked on the mirror fragment, and gradually the pattern emerged; it is reproduced in fig. 1. This drawing is necessarily an imitative representation of the original; it records not only actual marks on the bronze, but also direct inductions, such as the continuation of a partly preserved curve or circle, a line drawn on one side to match a line preserved on the other, and the development of the ‘matting’ or hatching from intermittent appearances. We have thought it well to indicate corrosion, but the dense obscurity of the original due to this cause cannot be reproduced.Portions of the rim of the mirror survive. This is as usual a separate strip of metal bent in a C-shape round the edge of the plate. The handle of the mirror is an attractive variant of a familiar design which can be better appreciated in the drawing than in the photographs previously published.

About two hundred and fifty yards west of Harlow railway-station, in the Essex border-parish of Latton (Ordnance Survey 6-in. map, XLI, NW.), a small gravel-capped oval hill known as Stanegrove or Standing Groves rises to a height of some 20 ft. above the marshy banks of the river Stort. In 1764 and again in 1819 ‘very strong walls’ were observed here; and other foundations ‘evidently Roman’, tesserae, many other Roman relics, and a stone coffin are recorded to have been discovered in the fields extending for a mile to the north-east, more particularly during excavations for gravel within 250 yards of the mound itself. Coffins thought to be of Roman date were found also near the station when the railway was built in 1841 and were re-buried in the station-yard. Amongst the finds from the area are specially noted ‘a great number of Roman coins chiefly of Emperors from the first Claudius to Valentinian’, including ‘several silver pieces of Sabina, Faustina the Elder, and Constantinus Junior’, and a few British, of which the following are specified: (1) one with a helmeted head, with CUNOBELINA; reverse, a hog and TASCHOVANIT (sic); (2) another ‘with a head on one side; on the other a man striking upon an anvil; (3) ‘one with a star, between the rays of which are the letters VERLAMIO; reverse, an ox’. The general character of the evidence closely resembles that obtained from similar riverside gravels in many parts of south-eastern Britain, and indicates a fairly extensive and continuous occupation, perhaps not of a very high order, from later prehistoric times to the last quarter of the fourth century.

In 1872 there came into the possession of the British Museum a small bronze fitting. In length 71 mm. and cast in a single piece, one end is in the shape of a bovine head, the other being a ring set at right-angles to the head, the shank having the appearance when viewed from the side of being bent back on itself to complete the circle; the maximum outer diameter of this ring is 27 mm. The animal's brow is defined by a double ridge which forms the lip of a question-mark-shaped socket whose body is represented by the top of the fitting. The shank has a diamond cross-section swelling gently from the moulded lips of the ring up into the head. The ridge on the rear surface sweeps out towards the top giving the end a triangular section. The casting is technically not wholly successful; the base of the ring has a stress fracture, there is a weak point at the top of the ring, and the sockets in the head (recently filled in with candle-wax) and on the return foot of the ring seem to be due in part to faulty pouring into the mould; certainly in the latter case there seems never to have been any intention to form an inlay setting although there is another possible use (see p. 28 below). Otherwise it must be said that the surface of the bronze has a fine finish with little or no sign of subsequent tooling.

The bronze brooch illustrated from photographs on pl. xi (and here also from drawings by Mr. F. Cottrill), was found a number of years ago by Mr. T. W. Armitage of Trent Lock, Long Eaton, on the high ground known as Red Hill, above the steep right bank of the river Soar just south of its confluence with the Trent in southern Nottinghamshire, between the villages of Thrumpton and Ratcliffe-upon-Soar and on the west side of the main L.M.S. railway line from Leicester to Nottingham, Long Eaton, and Derby. The site is habitually arable land, and Mr. Armitage has collected a number of antiquities from its surface, including besides this brooch several Roman coins and Romano-British bronze brooches along with some amount of mainly Romano-British pottery, to which, in the winter of 1943, he drew the attention of the Society. In subsequent correspondence he readily gave permission for the publication here of this unusual brooch, which he had himself recognized as a Celtic product of the pre-Roman Iron Age; and in lending it to the present writers for the purpose, he allowed it also to be cleaned and examined at Oxford, under Mr. Jacobsthal's direction, by Mr. V. R. Rickard in the laboratory of the Ashmolean Museum. The photographs were taken in the studio of the same museum.

The Romano-Celtic temples of Britain, unlike those of Gaul, rarely provide evidence of their dedications and the nature of the cults practised within them. At Silchester (Insula XXXV) and also at Worth, Kent, fragments of statuary found within the temple cella have been suggestive, although not conclusive. The remaining examples yet excavated in Britain (some eighteen in all) have been far less communicative: not one has yielded an inscription or votive object which would leave no doubt as to the identity of the god whose worshippers frequented it.Since our knowledge of the local cults popular in the civilian areas of Roman Britain is so limited, an object which, though found ninety years ago at Farley Heath in Surrey, has hitherto escaped publication deserves belated recognition not only for its own sake, because it displays very curious features, but also for the light that it may throw on the subject.

In 1929 the Duchy of Cornwall exercised its powers under the Ancient Monuments Acts, and placed Tintagel under the guardianship of the Commissioners of H.M. Works. During the clearance of the medieval castle it seemed desirable to investigate the earlier history of the site, in order to test the basis of the Arthurian traditions. In 1933 a fortunate discovery revealed dry-built walls on a terrace on the north-eastern cliffs. Subsequently an examination of the area surrounding the chapel disclosed indications of similar buildings, and trial excavations during 1933 and 1934 showed that these buildings belonged to a Celtic monastery.

The work of art here to be published was found in the excavations of 1938 at Elmswell, about two miles west of Driffield in the East Riding of Yorkshire (6-in. O.S. Yorks. E.R. CLXISE.). These excavations, following on those carried out by our Fellow Mr. A. L. Congreve from 1935 to 1937; have shown the site to be that of a native settlement of the Parisi, occupied—probably continuously—at least from Flavian times until the end of the Roman period, and again in pagan Saxon times. Of the pre-Roman Iron Age nothing has yet come to light beyond two British coins, one of the Iceni (c. A.D. 25) and one possibly of Cunobelin, but the site is separated by little more than the line of the Driffield—Malton railway from the Eastburn area where, in building operations in 1938, as many as seventy-five Iron Age graves were discovered even in the relatively small plots of ground excavated. These contained the remains of flexed inhumation-burials, each originally covered by a small barrow, exactly as in the well-known Danes' Graves and Arras cemeteries; and the date within the three centuries directly preceding the Roman Conquest was confirmed by several finds of grave-goods, notably an iron sword of La Tène type, bronze ornaments, an ‘involuted’ brooch, and several vessels of pottery. Further, in another part of the same site, Parisian pottery of Roman age was found, the earliest of which agrees exactly with the earliest from Elmswell. There can in fact be little doubt that this whole area was an important centre of settlement in the east Yorkshire equivalent of La Tène times, where ‘the Iron Age B’ culture of the Parisi passed directly into the romanization that was introduced at the Flavian conquest of the early seventies A.D.

In his account of the leaf-shaped bronze swords of the Hallstatt period, the late M. Déchelette wrote: ‘Doubtless one might ask whether this weapon might not have been brought to the British Isles by the first Celtic invaders, but that is purely conjectural (une conjecture fragile), for it is difficult among the British finds the same period to detect any really characteristic analogies’ (vol. ii, pt. 2, p. 724). The same paragraph points out that bronze swords of identically the same type have been found in regions as widely separated from each other as Scandinavia, Bohemia, and Ireland, to which one might add Finland and Southern France. The conclusion is irresistible that these swords were derived from a common centre of dispersal, and that they were not evolved independently in each region. Had the evolution taken place locally there might have been similarity, but not identity of type. I propose to bring forward evidence in support of the hypothesis that, towards the close of the Bronze Age, the British Isles were invaded by the first wave of Celtic-speaking peoples, bringing with them leaf-shaped bronze swords, many other entirely new types of bronze objects, and at least two types of pottery new to these islands and evolved somewhere on the Continent. I suggest that these invaders may have been Goidels, arriving about 800–700 B.C. Possibly the new types under review may not all be strictly contemporary; and there may have been more than one wave of invasion. But there can, I think, be no doubt that an invasion on a large scale took place at about this time.

During excavations carried out in 1937 by Mr. G. J. H. Neely on a Viking and pre-Viking site at Ronaldsway in the Isle of Man, a bronze balance-beam was discovered on the edge of a disturbed grave with a lead weight close by. A full account of the excavations appears in this number of the Antiquaries Journal. The balance-beam is, however, of unique interest, and it has been decided to make it the subject of a separate paper which is based on a comparison of the balance with recognized Viking, Saxon, and Romano-British types and on the results of an experiment conducted at the Science Museum, South Kensington. The balance and weight are now in the Manx Museum.

Sir Martin Conway's paper on portable reliquaries, tracing the origin of the familiar gabled type, includes a list of the Irish specimens; to this list it is possible to add another, which has twice been mentioned, but I believe neither described nor illustrated. Its existence is not generally known, for it was not alluded to by either Coffey or Romilly Allen in their works on Celtic Art. This reliquary was formerly deposited with the Royal Irish Academy by its owner, Mr. William Monsell, of Tervoe, co. Limerick, afterwards Lord Emly. It is described in one of the old Museum Registers as ‘A Shrine for holding relics 4¼ inches long, 3¾ high, and 1¾ broad. In shape of a house with roof sloping from both sides and ends. One side is deficient, but supplied with cork. Formed of wood with brass-ornamented ridge-pole, and brass bending at angles. One side and the adjoining slope of the roof is ornamented ingeniously with inlaying and enamel, the side having two, and the roof one circular ornament divided into compartments, which are subdivided by divisions radiating from their centres.’ The shrine was returned to Mr. Monsell in 1872. A plaster cast of it is, however, in the collection, from which the illustration is made. The colours in which the cast is painted show that the metal plates are bronze, not brass. The reliquary opens by means of hinges placed at the back, the upper part being a true lid. The ridge-pole terminates in animals' heads. The bronze plates covering the front of the shrine are ornamented with a species of fret pattern; they are inset with three roundels arranged like those on the Lough Erne and Copenhagen shrines. The empty centres of the roundels may have contained half-beads of amber (pl. IX, fig. 1).