Collection online


  • Object type

  • Museum number


  • Description

    Two-handled vase; honey-coloured agate; enamelled gold mounts and cover; oviform body cut in cameo with vine branches and grapes; Pan-headed handles; acanthus leaves at base; convex cover, chased in relief with four opaque white satyrs with outstretched arms holding grapes; knob in form of cluster of grapes and leaves, enamelled in translucent green and ruby; flat underside engraved with trefoil design of branches enclosing birds and insects, filled with translucent enamel; mouth mount divided into three zones: upper, representing wreath of grapes and leaves with ruby-coloured band; below lip, broad band with formal scrolls in white on black ground relieved with gold; foot formed of four rams' heads, each with four horns, resting on short stem with vertical flutes in blue, lower member with egg-like bosses of white enamel; square plinth, engraved and enamelled with floral scrolls; horned mask with leafy beard at each corner; sunk panel of reclining Bacchus in relief on each side, in white enamel; on each side, rectangular fret in blue enamel enclosing a vase in ruby red. Openwork.


  • Producer name

  • Culture/period

  • Date

    • 1811-1834 (mounts)
    • 400 (c.; vase)
  • Production place

  • Materials

  • Technique

  • Dimensions

    • Height: 21.4 centimetres
    • Height: 1.5 centimetres (plinth)
    • Width: 11 centimetres
    • Diameter: 5.7 centimetres (plinth)
    • Diameter: 5.2 centimetres (cover)
    • Depth: 8.1 centimetres
    • Weight: 596 grammes
  • Curator's comments

    Text from Tait 1991a:-

    (i) Carved agate: authenticity is uncertain; since 1899 loosely described as “antique Roman” or “antique”, but recently attributed to the late Roman period, c. AD 400.
    (ii) Enamelled gold mounts and cover: previously described as “Italian, 16th century” and, subsequently, attributed to Benvenuto Cellini (1500-71) but now attributed to the hand of an early 19th-century copyist - before 1834 - perhaps working in London.

    Provenance: Formerly in the collection of the Dukes of Devonshire; recorded in a manuscript “Inventory of Curiosities and Articles of Vertu at Devonshire House, 1834”; also listed under the heading of “Vases of Stones” in the 1892 inventory of Devonshire House, London; lent by the 8th Duke of Devonshire to the Burlington Fine Arts Club Exhibition of European Enamels, 1897; acquired by Baron Ferdinand Rothschild about one year before his death in December 1898, and visible in a photograph of one of the vitrines in the Smoking-Room at Waddesdon Manor towards the end of his privately printed ‘Red Book’, the Foreword of which is dated 1 November 1897 (see also WB.69 - WB.72, all acquired at the same time from the same source).

    Commentary: The preceding account of the technical construction of the vase has revealed many features that are inconsistent with the known practice of sixteenth-century goldsmiths' workshops, and so the former dating of the goldsmiths' work on this vase to the mid-sixteenth century is no longer sustainable (see below). Most importantly, detailed analysis of the technical construction has revealed that the neck-mounts could not have been riveted into place until after the metal disc had been inserted into the vase and, therefore, the stem and foot elements can only be of the same period of manufacture as the neck-mounts and cover.
    When the vase was first published in 1897, it was described as “Italian, 16th century”, and one of the first suggestions that the carved 'chalcedony' body of the vase had a different origin - namely, “Graeco-Roman” - appeared in Davenport 1900. However, in Read 1902 the 'agate' vase was said to be “antique Roman” and the mounts were, again, described as “Italian, 16th century”; this distinction was repeated in Mrs Eugene Strong's entry in the Chatsworth ‘Catalogue of the Gems’ in 1909. Not until Dalton 1927 was the attribution cautiously altered to read: “The vase, antique; the mounts, Italian 16th century, perhaps the work of Cellini.” This highly significant addition was apparently based on a lengthy comparative study with the famous and fully documented Cellini salt-cellar of 1543 in the Hapsburg Collection in Vienna. Read's manuscript notes, following his visit to Vienna in 1912, are the earliest record of this new line of research, which was to receive the wholehearted approval of Ernst Kris during his visit to the British Museum in 1926, as Kris himself later confirmed in his full-scale publication of the Cellini salt-cellar in 1932, when he restated that the mounts of the Waddesdon Bequest agate vase were the work of Cellini. This attribution, accepted as “very plausible” by Pope-Hennessy in 1949, was quoted by Camesasca in 1955 and by Hayward in 1962; it was not challenged until October 1971, when the author was invited by The Society for Renaissance Studies (Warburg Institute) to lecture on 'Cellini the Goldsmith and his Trattato dell' Oreficeria' (at the Victoria and Albert Museum) and was able to contrast not only the technical features with those of the Vienna salt-cellar but also the character of the enamelled ornament with that on several documented examples, in particular the famous mounted onyx jug which the French King, Charles IX, presented to Archduke Ferdinand of Tyrol in 1570 and which has since remained unaltered in the Hapsburg Collection. Ten years later, in 1981, the author was able to publish for the first time two coloured drawings of the Waddesdon Bequest 'Cellini' vase, both of which recorded several crucial details on the enamelled gold mounts that differed from the object as it exists today, thereby reinforcing the arguments advanced in 1971 for not only disputing this attribution to Cellini but also for questioning the reputed sixteenth-century origin of the gold mounts and proposing an early nineteenth-century origin as the more convincing alternative solution. In the recent monumental monograph on Cellini (see J. Pope-Hennessy, ‘Cellini’, London, 1985) there is no longer any reference to the Waddesdon Bequest agate vase.
    The reason why the enamelled gold mounts on the agate vase in the Waddesdon Bequest had been so highly regarded by scholars like Ernst Kris is that their external appearance is closely based on a genuine Renaissance original, which is now lost but is, fortunately, recorded in the earlier of the two drawings first published in Tait 1981 (figs 37-8) - the pen and watercolour drawing on paper preserved in the collections of the Dukes of Devonshire at Chatsworth since the mid-eighteenth century. The large album, in which it appears (on folio 24, mounted along with four other eighteenth-century drawings) is labelled 'Lord and Lady Burlington's Drawings'. The album, which is quarter-leather with marbled-paper covered boards dating from about 1810 to 1830, contains a large group of drawings by William Kent as well as a miscellaneous collection of drawings, all of which date from the first half of the eighteenth century, and it seems very probable that they were brought together during the lifetime of Richard, 3rd Earl of Burlington (1695-1753), the creator of Burlington House, Piccadilly, and of Chiswick House, and patron of the architect William Kent. Lord Burlington had brought him back from Italy in 1719 and installed him in Burlington House, which became his home for the remainder of his life; he died there in 1748. When Lord Burlington died in 1753 his spectacular collection of architectural drawings (together with the rest of his magnificent collections, mostly kept at Chiswick House) joined those at Chatsworth, because five years earlier his daughter and heiress, Charlotte, had been married to the 4th Duke of Devonshire.
    At Chatsworth today this drawing of the mounted agate vase is without parallel; the lack of comparable material in that vast historic collection is a curious fact. The drawing itself has been slightly damaged by folding at some time (before it was placed in the album), and, on the reverse, in the top left-hand corner there remains (upside down) an inscribed capital A with a cut-off flourish - similar to that in the inscription on the obverse. The drawing has been trimmed so close that, on the obverse, at the top the tip of the finial has just escaped but, at the bottom, part of the flourish at the beginning of the French inscription has been lost although the two words are themselves perfectly preserved: Agathe Oriental. Although the caption seems, therefore, to indicate a French origin for the drawing, perhaps in the late seventeenth or early eighteenth century, the evidence remains far from conclusive. There is, however, no reason to doubt, firstly, that the drawing pre-dates the death of Lord Burlington in 1753 and, secondly, that it is not a sketch or design for a proposed commission but is a factual drawing of an existing object, recording as accurately as possible all the salient details, like an illustration in a pictorial inventory. Its value as a document is, without doubt, immeasurably greater than its merit as a work of art - and for that reason it is perhaps puzzling that it should have been acquired for the collection of 'Lord and Lady Burlington', unless it had, indeed, been sent to Lord Burlington or William Kent as a visual record of an object that could be purchased through some dealer and, years later, the drawing had been incorporated with William Kent's drawings and all the other miscellanea in Chiswick House. Significantly, William Kent, who was buried in Lord Burlington's vault in Chiswick Church, had bequeathed to Lord Burlington “two yellow Marble Vases with Vine Leaves” (see E. Croft-Murray, ‘Decorative Painting in England, 1537-1837, II, The Eighteenth and Early Nineteenth Centuries’, London, 1970, p. 232, where the will is stated to have been made on 13 October 1743, and in a codicil two similar vases were also bequeathed to Lord Burlington). It seems, therefore, that William Kent may, indeed, have been the owner of this drawing.
    Strangely, the connection between the drawing and the agate vase had not been made before 1981. Although the vase had been lent by the 8th Duke of Devonshire to the Burlington Fine Arts Club Exhibition in 1897, neither the drawing nor any reference to it appeared in the full-scale Catalogue of the exhibition. Similarly, it seems not to have been mentioned to Baron Ferdinand at the time of the sale of the vase immediately following the closure of the exhibition. However, earlier in the century the records of the Devonshire Collection are so worded that it may be deduced that the existence of the drawing was well known and its relationship to the vase fully appreciated, especially by the 6th Duke (born 1790, inheriting title in 1811, died 1856). It is, for example, more than a coincidence that in the 1834 'Inventory of Curiosities and Articles of Vertu at Devonshire House' this vase was described - quite unbelievably - as: “A fine Oriental Vase, engraved and raised Vine borders with handles Satyrs' heads, and richly mounted in Gold and Enamel.” The unique use within this Inventory of the word 'Oriental' is most unexpected because although there are many other agates listed none is similarly described. Furthermore, none of the entries describe the origin or age of any of the 196 hardstone objects. Instead, the emphasis was consistently on recording the type of mineral and the form of the object, with perhaps a brief mention of additions, such as the surface decoration or metal mounts. This pattern was established with the first entry of the Inventory: “A Chrystal Boat-shaped Cup; Engraved a Fight of Sea Monsters, Gilt Silver foot”. The 6th Duke, as a young man in his twenties, was noted for his enthusiasm for hardstones and mineralogical specimens, as indeed the numerous entries of the 1834 Inventory fully confirm; they even include three items (“a Spar Pillar, Blue John” and “2 White Spar Pillars”), which according to the annotation (in a different hand) were “taken by His Grace to Paris 2 Jany, 1836”. Blue John is the finest-quality fluorspar, being deep purple in colour, and was found in its greatest perfection in Derbyshire (not far from Chatsworth), but the supply was completely exhausted before the end of the nineteenth century. Although many of the items listed in 1834 were finished objects, highly polished and richly mounted, others were little more than mineralogical specimens that received the briefest of descriptions: “polished pebble”, “7 agate polished specimens”, “1 Topaz Pebble (or Nova Mina)”, “A Lump of polished Red Agate” and “A piece of black and white speckled Porphyry”. In such a context the entry describing the Waddesdon Bequest vase is conspicuous for being the only one that omits the name of the material from which it had been made - just as it is unique in appearing to give its origin. The curious form of words – “A fine Oriental vase . . .” - seems, therefore, to reflect the compiler's knowledge of the drawing inscribed ‘Agathe Oriental’ and perhaps, in a moment of carelessness or inattention during dictation, the crucial word ‘Agathe’ (agate) was omitted by the clerk. In the absence of any mention of this piece in the 1811 Inventory of Devonshire House or in any of the pre-1834 records at Chiswick House or Chatsworth, it may be assumed that the 6th Duke was responsible for the acquisition of the vase and for recognising its close relationship with the drawing entitled ‘Agathe Oriental’.
    The term ‘Agathe Oriental’, or ‘Agathe d'Orient’, seems to have been frequently used by the compilers of French seventeenth-century inventories, probably to denote the exceptional quality of the stone. The most fully documented example to have survived from Louis XIV's collection - albeit without its gemstones, gold foot and central mount - is now in the Galerie d'Apollon, Musée du Louvre (inv. no. OA. 8; Marquet de Vasselot, Paris, 1914, no. 1015; also Marquet de Vasselot, A propos de quelques gemmes des anciennes collections de la Couronne, in ‘Congrès d'Histoire de I'Art organisé par la Societe de I'Histoire de I'Art français, Paris 1921: Compte-rendu analytique’, Paris, 1922, pp. 192-4; more recently, the hardstone of this curious vessel, reclassified as sardonyx, has been compared with a Byzantine example dating from the ninth to the eleventh centuries AD in the Treasury of San Marco, Venice, although it is acknowledged that the surface carving on the exterior of the bowl was executed in France in the mid-sixteenth century when the gold mounts and foot were added - see D. Alcouffe, ‘The Treasury of San Marco, Venice’, ed. D. Buckton, exh. cat., Milan, 1984, p. 157, fig. 15a). This most extraordinary vessel was purchased by the King in 1685 - twenty-four years after the death of its former illustrious owner, Cardinal Mazarin, in whose posthumous inventory of 1661 it is described as: “Un grand vaze dont le corps est d'une seulle agathe d'Orient travaillée à godrons . . .” That lengthy description (quoted in extenso in D. Alcouffe, The Collection of Cardinal Mazarin’s Gems, ‘Burlington Magazine’, CXVI, 1974, p. 521, Mazarin inv. no. 324, fig. 20) continues with a detailed record of the enamelled gold mounts that were set with numerous precious stones (rubies, diamonds, emeralds, etc.) and concludes with the weight - all of which corresponds exactly with the annotated water-colour sketch that had been prepared and sent to Louis XIV from Germany at the time of the negotiations for its sale. The sketch is numbered '17' and inscribed in ink: ‘Vaze d'Agathe orientalle’ (at the top) and ‘Envoyé de Francfort le 28 Avril 1685’ (at the bottom of the sheet), together with numerous smaller annotations relating to the composition of various details on the vase (preserved in the Cabinet des Estampes, Bibliothèque Nationale, no. Lf3, pl. 17; H.O, 394, L.O, 203; see J. Adhémar, in ‘Collections de Louis XIV: dessins, albums, manuscrits’, exh. cat., Paris, 1977, p. 326, no. 350, with illus.). The sketch, which is stated to have belonged to the “agence royale Robert de Cotte” (no. 1919 of the MS inventory of the drawings, etc., in Monsieur de Cotte's possession) and to have been acquired in 1811, is catalogued as an anonymous French work of the seventeenth century and provides additional support for the tentative attribution of the Chatsworth drawing, also inscribed ‘Agathe Oriental’, to a French hand in the second half of the seventeenth century.
    Significantly, the Royal Accounts (‘les Comptes’) of Louis XIV mention not only a payment on 27 March 1685 of 4,800 livres to Sr. Alvary for this “grand vase d’agate onix garni d'or . . . qu'il a acheté pour le service de S.M.”, but also several similar agate vessels in the same year - all coming from Germany. It is, therefore, probable that these vessels would have been similarly sketched and that, for some reason, another of these drawings, also made about the same time, was later brought to England, where it eventually came to rest in the album of 'Lord and Lady Burlington's Drawings' at Chatsworth.
    Without doubt, the 6th Duke would have discovered the drawing within a few years of inheriting the title at the age of twenty-one in 1811. He lost no time in exploring the vast hidden areas of his inheritance, shunning the world of politics and even that of the Court, despite being a close friend of the Prince Regent, later George IV. The King made him Knight of the Garter and chose him as his representative at the Coronation of Tsar Nicholas in 1826, but three years later he resigned the office of Lord Chamberlain within ten months of being appointed. He was a lifelong bachelor and devoted much of his time during some forty-seven years to enriching the library and collections and to expanding the gardens and buildings, for he clearly had developed a deeply personal attachment for Chatsworth - a love that he strove to share with as many as possible, keeping the house and grounds open to the public from 10 a.m. to 5 p.m. (see F. W. Thompson, ‘A History of Chatsworth’, London, 1949). His early re-organisation of the collections of drawings is well documented and, indeed, he removed “nearly one thousand original sketches” from their albums and had them hung for all to see in “the Gallery leading to the Chapel” (see Stephen Glover, ‘The Peak Guide’, Derby, 1830, p. 27). By 1835 he had improved the display of the drawings for the unexpectedly large numbers of visitors within the newly created 'Sketch Galleries', where (in his own words) “I have classed them according to the several schools of painting; but I am sure that the arrangement must be very imperfect . . . Few things at Chatsworth are more admired . . . They hardly ever saw the light in my Father's time, nor in mine often, till I rescued them from portfolios, and placed them framed in the South Gallery below” (‘Handbook to Chatsworth and Hardwick’, written (and privately printed) by William George Spencer Cavendish, 6th Duke of Devonshire, 1845, pp. 34-6). Undoubtedly, this coloured drawing inscribed ‘Agathe Oriental’ would have attracted his attention not only because, as a collector, he was noted for his connoisseurship of hardstones, engraved gems and, later, of marble sculpture, but also because it is so conspicuously unlike any of the other drawings in the collection.
    Searches in the archives at Chatsworth and in London have not discovered papers that chronicle how or when the 6th Duke acquired the matching vase - presumably after 1811 and certainly before 1834. Since few, if any, persons outside the Duke's personal staff and circle of close friends would have known of the existence of the drawing inscribed Agathe Oriental, it seems incredible that the vase itself should almost simultaneously become available and be offered to the Duke, the one person most likely to recognise it and appreciate its connection with the coloured drawing at Chatsworth. If he did record the details of the acquisition of this vase and, perhaps, some details of its previous history, then that information was neither incorporated in the 1834 Inventory - nor in the 1892 Inventory of Devonshire House, where the vase was valued at £650. This later Inventory was made after the death of the 7th Duke and lists the contents of the rooms in the House before itemising the Plate, the Gold Boxes and, then, the 'Vases of Stones'. In both Inventories the five vases that were later to be lent by the 8th Duke to the public exhibition of 1897 had been listed in an unbroken sequence, and, interestingly, all five were kept together when they were sold to Baron Ferdinand Rothschild. The other four (WB.69 – WB.72) are, in the author's opinion, each demonstrably equipped with new mounts of gem-set and enamelled gold - all apparently dating from between 1811 and 1834. The same appears to be true of the so-called 'Cellini' Waddesdon vase.
    A detailed study of the Chatsworth drawing and of the mounts on the vase itself reveals the following major inconsistencies:
    (i) On the square foot (as depicted in the drawing) the two 'classical' vases on either side to the oval niche do not have handles - unlike those on the gold mount itself.
    (ii) On the square foot (as depicted in the drawing) the two 'half-spaces' at either end are left empty - unlike those on the gold version, where the 'half-space' is slightly increased and accommodates one half of a 'classical' vase, again complete with handle. Because each half is adjacent - but at right angles - to an identical half on the adjoining side, the visual impression has been created of four 'classical' vases, each with two handles at the neck, being bent through ninety degrees at each of the four corners of the foot. No such eccentricity of design would be correct in the Renaissance, and neither the Chatsworth drawing nor any sixteenth-century design by a court artist is known to include such a strange detail.
    (iii) On the upper surface of the square foot (as depicted in the drawing) two of the white enamelled masks with beards are clearly depicted, but the rest of the upper surface of the square foot is left plain - unlike the gold version, where the space between the applied masks is filled with a translucent enamelled vine scroll along each of the four sides.
    (iv) On the circular stem (as depicted in the drawing) the white egg-like bosses are each decorated with four separate dots of gold (arranged in a diamond pattern) - unlike those on the stem of the Waddesdon Bequest vase, which are each ornamented with a four-lobed gold rosette.
    (v) On the neck (as depicted in the drawing) the four roundels (with white rivet heads) are flesh-coloured (not a bright red as on the vase) and the swags of drapery are white (not the vivid deep blue of those on the vase); above, the bold arabesque design is painted in blue on a light ground (not white on a black ground, as on the Waddesdon Bequest vase).
    (vi) Finally, the rim-mount (as depicted in the drawing) projects outwards even further and its upper convex surface is less rounded than on the vase itself. Certainly, the two profiles are distinctly not the same.
    The above list of very precise differences has been compiled because none of them can be convincingly dismissed as the simple errors of an incompetent seventeenth-century artist of inferior calibre. Furthermore, all these differences - apart from the colours on the neck - are to be found repeated in a second drawing of the vase. Unfortunately, this second drawing had no history before 1885 when it was purchased, along with a collection of some fifty miscellaneous drawings, from an unnamed dealer by the Victoria and Albert Museum (inv. no. D. 64-’85; see P. Ward-Jackson, ‘Victoria and Albert Museum Catalogues: Italian Drawings Vol. I: 14th—16th Centuries’, London, 1979, no. 495, where it was not recognised as relating to the drawing at Chatsworth nor to the vase in the Waddesdon Bequest; it is, however, stated to be “a design for a vase and cover decorated with grapes and vine leaves. Pen and ink with gold paint and blue watercolour on the base and cover”). Its identification as 'a design' and its attribution to a sixteenth-century Italian workshop are wholly unconvincing, especially as the draughtsmanship is greatly inferior to that of the Chatsworth 'inventory-type' drawing. For example, on the London drawing the vine branch on the right-hand side of the body, the rams' heads around the cagework element and the relief on the cover are particularly incompetent. Inexplicably, the oval niche on the square foot has been carefully drawn in a less elongated form - with a corresponding adjustment to the Bacchus figure, which is less reclining - than on the Chatsworth drawing. However, all the other details on the foot are exactly repeated, but, curiously, the neat effect has been marred by a later hand crudely drawing two additions on either side of the square foot, as if experimenting with the idea of enlarging the foot by 'half a space' on either side.
    The age, origin and purpose of this drawing, therefore, remains uncertain, although it certainly possesses all the characteristics of a drawing of a finished object, aiming to record the most prominent features in detail - albeit with less skill than in the drawing inscribed ‘Agathe Oriental’. Consequently, there is no new information that does not corroborate the evidence of the Chatsworth drawing and reinforce the conclusion that the enamelled gold mounts on the Waddesdon Bequest vase are not those depicted and, therefore, must be of more recent origin - an interpretation that agrees with the results of the recent physical examination of the dismembered mounts.
    If the mounts are not original, what evidence is there for attributing the carved agate two-handled body to a workshop in the late Roman period? In 1897 only one of the five Devonshire pieces lent to the Burlington Fine Arts Club Exhibition was thought to be 'ancient' - the lapis-lazuli vase (WB.70), which was described as “probably late Roman of about the second century”, although the lapis-lazuli cover was stated to be “Italian, 16th century, surmounted by a small figure . . .”. The identical ascription, “Italian, 16th century”, was bestowed in 1897 on this two-handled agate vase but was decisively abandoned by Charles Hercules Read within a year of Baron Ferdinand's death. His new attribution can be found in a little-known paperback handlist of the Waddesdon Bequest, which is very prominently marked ‘UNDER REVISION’, both on the cover and on the title-page, and has an Introduction (pp. v-xii) bearing his signature and the date “December 1899”; in it, he states “. . . one piece stands pre-eminent - the mottled agate vase (no. 68, pl. IV), of which the body is an example of ancient Roman cameo work ... it is somewhat rare to find an ancient vase of hardstone so symmetrical and graceful in outline and it is probably this quality that led the goldsmith of the Renaissance to bestow his skill upon it” (see ‘The Waddesdon Bequest, The Collection of Jewels, Plate and other Works of Art bequeathed to the British Museum by Baron Ferdinand Rothschild, M.P.’, London, 1899, p. vi). This passage was, indeed, revised to read: “The only other relic of classical times is the agate cameo vase (no. 68), which may be assigned to imperial Rome . . .” (Read 1902, p. x). Apart from three similarly worded versions (Davenport 1900, Strong 1909 and Dalton 1927), Read's belief in the classical origins of the carving on the agate body appears not to have received careful consideration until 1943, when Marvin Ross included the vase in a list of extant “antique agate vases, either whole or in a fragmentary state” (Ross 1943, p. 30, fn. 118). Furthermore, Ross stated that “of these, only one, the Waddesdon Vase in the British Museum, seems to have any relation to the Rubens Vase . . .” and he illustrated it on page 33 in fig. 17. Ross proceeded to relate it to a vase in Vienna (see A. Furtwangler, ‘Die Antiken Gemmen’, III, Berlin, 1900, p. 341, figs 192-3) and to a silver repoussé case enclosing a glass phial in Baltimore (Walters Art Gallery, inv. no. 57,932 ; H. 9.5 cm); he tentatively concluded that these “comparisons seem to indicate that this vase belongs with late Roman gem carvings, in which case it precedes the Rubens Vase in date and represents a type from which the latter derived”.
    For the Rubens Vase Ross succeeded, where previously there had been disagreement among scholars, in marshalling “convincing arguments against an early- or late-Renaissance dating of the piece and in favour of placing it in the fourth or fifth century of our era” (Harden and Toynbee 1959, p. 202, pl. LXXIV). These two classical scholars not only illustrated the Rubens Vase but also included a photograph of the Waddesdon Vase, about which they stated “the design is tamer and the undercutting of the ornament less daring. It may well be somewhat earlier than the Rubens Vase and is technically less close than is the latter to the figured glass cage-cups” (Harden and Toynbee 1959, p. 202, pl. LXXV, a, b). The other recent major publication, also by a classical scholar, lists it as the 'Waddesdon Vase' and places it immediately before the Rubens Vase, which is dated to the end of the fourth or beginning of the fifth century AD (see Bühler 1973, pp. 76-7, nos 108-9, pls 36-7). Furthermore, Bühler compares the carving of the acanthus leaves on the Rubens Vase with those on the Waddesdon Vase, and for the form of the latter he refers to a piece in Rome (discussed in G. Lippold, ‘Die Sculpturen des Vaticanischen Museums’, III, 2, Berlin, 1956, p. 236, no. 84, pl. 115); however, the latter is a more typical classical oviform vase (without handles) and lacks the unusual elongated shape of the Waddesdon Vase. For certain details of the carving, especially the drilling of small dots, he cited the unprovenanced chalcedony fragment with a leaping panther amidst foliage from the Townley Collection (British Museum, Department of Greek and Roman Antiquities; see Bühler 1973, p. 77, no. no, where the fragment (H. 4 cm) is dated “4-5 Jahrhundert n. Chr.”; also, see H. B. Walters, ‘Catalogue of the engraved gems and cameos, Greek, Etruscan and Roman, in the British Museum’, London, 1926, p. 375, no. 4052, fig. 115 (a line-drawing), but there is no reference in this entry -nor, indeed, elsewhere in the volume - to the Waddesdon Bequest's “antique Roman” two-handled agate vase).
    In conclusion, Read's initial description of the carved agate body as a Roman antiquity - unsupported by sound art-historical argument - has gradually been endorsed by classical scholars, despite the lack of any mention of it by Walters in 1926. However, in the absence of a well-researched case, comparable with that advanced in respect of the Rubens Vase (see Ross 1943, pp. 9-39; D. Alcouffe, Gemmes Anciennes dans les Collections de Charles V et des ses Frères, ‘Bulletin Monumental’, vol. 131, pl. 1, Paris, 1973, pp. 41-6, fig. 1; also Alcouffe 1984, p. 85, fig. 2e), the Roman origin of the Waddesdon Vase will remain less securely established than that of the Rubens Vase, which, for all the superficial similarity of its decorative motifs, is otherwise very different. As expressed in Marvin Ross's commentary on the Waddesdon Vase (Ross 1943, p. 31), “not only its proportions, but the whole spirit of the thing is quite different . . . the restrained design shows none of the almost eccentric vigour and the daring undercutting of the Rubens Vase”. One might add that the truly sculptural quality of the Pan-head handles on the Rubens Vase is immeasurably superior to the carving of the heads on the Waddesdon Vase, where the lack of modelling and absence of detail (such as the row of teeth or the locks of hair) is strikingly apparent. Furthermore, as present colleagues in the Department of Greek and Roman Antiquities are quick to point out, little comparable material has survived and so, without adequate data, the necessary criteria for judging the age of the carved agate body of the Waddesdon Vase is still lacking. An understandable reluctance to confirm the vase's reputed origin in the late Roman period will persist unless fresh evidence comes to light.
    In a sense new evidence has emerged - the inscribed Chatsworth drawing, the 1834 Inventory description of the vase, the special enthusiasms of the 6th Duke and, to a less quantifiable extent, the London drawing. Between
    1811, when the Duke inherited the title, and 1834, when the Waddesdon Vase was first recorded, the long-lost Rubens Vase had reappeared - after nearly 200 years - and, not surprisingly, became one of the three highly prized objects in William Beckford's collection chosen to adorn the engraved title-page of ‘Illustrations Graphic and Literary of Fonthill Abbey, Wiltshire’, by J. Britton, F.S.A., which was published in London in 1823 - probably soon after Beckford had acquired it, perhaps through Christies. Britton's text extols the quality of the carving, concluding that it “affords reason for believing that this rare and very curious article must have been executed by a Greek artist of Asia Minor (‘Christie's Catalogue’, p. 41)”. That Beckford was aware that it had belonged to Rubens is established beyond doubt; indeed, he possessed an engraving after Rubens's drawing of two views of the vase. Whereas Beckford thought it had been pawned in Rubens's time for a large sum, it had in fact been sold by Rubens to Daniel Fourment between 1626 and 1628, less than ten years after he had purchased it in Paris. The vase was later being sent to the East Indies when the boat was captured by the Dutch and, despite Rubens's efforts (through the office of the Dutch East India Company) to locate it, it remained 'out of sight' not only in his lifetime but until 1823. It is always difficult to gauge the excitement that such a major rediscovery might provoke, especially in the cultivated circles of London society during that extravagant reign of George IV, but for a collector and connoisseur of carved hardstones and mineralogical specimens, such as the wealthy young 6th Duke of Devonshire, the appearance of the Rubens Vase in England could not be without interest. Significantly, Beckford did not part with it in his lifetime and it entered the collection of his son-in-law, the Duke of Hamilton, where it remained until 1882. Did the advent of the Rubens Vase stimulate craftsmen and dealers in London to seek out and 'rehabilitate' damaged specimens - even, perhaps, create de novo clever pastiches in the late Roman style? There was no shortage of skilled gem-engravers at this period, particularly in Rome, Paris and London, for gem-cutting as an art form was extremely fashionable throughout the last quarter of the eighteenth and the first three decades of the nineteenth century. The skill was not lacking, but, in the case of the Waddesdon Vase, the design would also have had to be known and, unfortunately, there is as yet no evidence that the London drawing even existed - let alone that it was available for copying - at that time probably midway between 1823 and 1834. However, it is not impossible that it was made in the 1820s for that very purpose because the Chatsworth drawing was in Derbyshire - probably being mounted, along with other miscellaneous drawings, to be placed in the new large quarter-leather album labelled 'Lord and Lady Burlington's Drawings'.
    Before leaving the vexed question of the origin of the carved agate body of the Waddesdon Vase, attention must be drawn to its perfect state of preservation despite the clean break that has resulted in the loss of the stem and foot; there appears also to have been considerable damage to the former neck of the vessel since the massive gold mount is attached with rivets passing through the walls of the body below the neck. For the agate vase to have been broken at both ends and yet not to have sustained any damage to the delicate handles, which have been thinly carved from the same block of agate, is truly remarkable.
    Unlike the handles on the Rubens Vase, which do not project and have only a minimal amount of undercutting beneath both pairs of sturdy horns, the handles of the Waddesdon Vase are particularly vulnerable. Each pair of horns is not only carved completely in the round but projects a long way from the sides of the vessel. Furthermore, the two horns of each pair separate and are carved as individual thin strips of curving agate for about half their length. These four narrow independent strips commence before the handles reach their highest point and begin to curve inwards. Consequently, the handles are far more fragile than those carved on the Rubens Vase - and yet they did not suffer when the Waddesdon Vase was broken. Other hardstone vessels from Classical Antiquity are carved with handles of even greater delicacy and complexity but survive intact only when the rest of the vessel has not been badly damaged (see Alcouffe 1984, pp. 129-35, no. 10, col. pl., and figs 10 a-j); for an example of comparable high quality in the British Museum, see the fragment of a sardonyx handle terminating in a leaf (in the Townley Collection, Department of Greek and Roman Antiquities - see Walters 1926, no. 4021).
    The method by which the goldsmith attached the new stem and foot to the broken end of the agate vessel is ingenious but wholly out of character with Renaissance practice. With one small exception, gold has been extravagantly used throughout, even for the totally hidden elements (such as the long square bolt and screws). Furthermore, given the nature of the break at the lower end of the body, the goldsmith could scarcely have devised a more effective method of attaching a stem and square foot of enamelled gold that would resemble - at least superficially - the design recorded in the two drawings. There is, however, nothing in either drawing to suggest that the frieze of white enamelled rams' heads is an openwork 'sleeve' made to fit over a hollow gold 'cap' of roughly hemispherical form, but the goldsmith has chosen to make a double-walled gold 'calyx' - altering the design of the palmette in the process - because it enabled him to employ a late eighteenth- and early nineteenth-century practice common among restorers of both cameos and intaglios. Instead of trimming and refashioning the end of the vase so that the gold mount of the stem could fold over and 'grip' it in the Renaissance manner, the goldsmith has, in effect, extended the vase downwards in gold. He has moulded the gold to the irregular and undulating shape of the fracture so that the hollow - but internally strengthened - gold 'cap' fits perfectly against the broken surface of the agate, leaving no perceptible gap at any point. This distinctive method of restoration, using only a soft gold, became especially popular among gem-collectors, where so many cameos and intaglios of the ancient world had understandably been found damaged or chipped. Despite their imperfect state, they were in great demand among rich collectors and became even more saleable when the missing areas had been skilfully filled with gold, smoothly moulded to the fractured edge and the newly restored gem set in a handsome gold frame or a plain gold finger-ring (see Dalton 1915, nos 74, 108 and 111, pl. VIII; also, no. 165, which is a large - and very undistinguished - chalcedony cameo of the head of Ceres dating from the second quarter of the nineteenth century and yet has been framed in enamelled gold to simulate damage as if it were a gem of classical origin that had been badly chipped in two places: for an illustration see Hugh Tait (ed. And contrib.), ‘The Art of the Jeweller, A Catalogue of the Hull Grundy Gift to the British Museum: Jewellery, Engraved Gems and Goldsmiths’ Work’, London, 1976, p. 238, no. 398. For examples in the Department of Greek and Roman Antiquities see Walters 1926, nos 3498, pl. XXXV; 3500, pl. XXXV; 3501, pl. XXXIV; 3631). This method of restoration - so reminiscent of the technique of dental gold fillings - was especially favoured in England, as can be seen, for example, in the gem collection of Richard Payne Knight, bequeathed to the British Museum in 1824, and indicates that the goldsmith responsible for the mounts on the Waddesdon Vase was more familiar with this practice than with those of the Renaissance. The latter are excellently documented, for example, in Oeding's drawing of the gold mounts that were added to the famous Roman onyx vase from the Grotta of Isabella d'Este (inventoried in 1542) and since 1666 kept in the Kunstkammer of the Dukes of Brunswick (see G. Bruns and A. Fink, Das Mantuanische Onyxgefäss, ‘Kunsthefte des Herzog Anton Ulrich-Museums’, Brunswick, V, 1950, pp. 1-50, figs 1, 3).
    However, his workshop must also have had to supplement the information provided by the London drawing with first-hand knowledge of certain rare pieces of French court goldsmiths' work of the middle of the sixteenth century in the sophisticated Fontainebleau Mannerist style. Because of the great losses in France, the opportunity to study French Renaissance gold and enamelled mounts of this type today is as good in Vienna as in Paris because in 1570 Charles IX of France had presented to Archduke Ferdinand of Tyrol not only Cellini's famous gold Saliera ('salt-cellar') of 1543 but also two other objects from the Royal Treasury, the gold St Michael's cup and the beautiful onyx jug mounted in enamelled gold. These three masterpieces entered the Archduke's Schatzkammer at the Schloss Ambras, near Innsbruck, where they safely remained, along with other pieces of similar character, until 1806, when to escape the threat of Napoleonic armies the Ambras Collection was brought to the Lower Belvedere Palace in Vienna, where it became the 'Imperial and Royal Ambras Collection'.
    Whether these objects were known to the goldsmith who made the mounts of the Waddesdon Vase or whether it was the examples in the French royal collection (now in the Galerie d'Apollon, Musée du Louvre) that served as his source is more difficult to determine. Certainly, the relief ornament on the foot and the stem of the Waddesdon Vase seems a rather obvious, but faint, echo of the achievements of Benvenuto Cellini on his great 1543 salt-cellar, although, surprisingly, the nineteenth-century goldsmith has feebly repeated the identical design on all four sides of the foot by the mechanical process of using a mould -not a characteristic of Cellini's salt-cellar! Furthermore, he has introduced a note of inconsistency by decorating the upper surface of the foot with a translucent enamelled vine-scroll decoration that is stylistically at variance with the rest - although he also repeats this error on the disc applied to the underside of the cover. The rest of the cover, together with the neck mount, is more accomplished both in design and in execution. The curving surface beneath the rim is decorated with the kind of arabesque pattern (in white and black enamel with the finer details executed in gold) that the mid-sixteenth-century French Renaissance goldsmiths excelled in producing on Court objects, such as the rim-mount for Cardinal Mazarin's agathe d'Orient cup that Louis XIV purchased in 1685 and which is now in the Louvre (discussed above). It can also be seen in a more elaborate and better-preserved form on the mounts on the equally well-known gold and onyx jug that Charles IX of France gave to Archduke Ferdinand of Tyrol in 1570 and which entered the Ambras Collection; most recently it has been published as a work of about 1560-70 by the Paris goldsmith Richard Toutain the Younger, who became a master goldsmith in 1559 and died in Paris in 1579 (see Helmut Trnek in ‘Führer durch die Sammlungen’, Kunsthistorisches Museum, Vienna, 1988, p. 190; also, E. Kris, ‘Meister und Meisterwerke de Steinschneidekunst in der Italienischen Renaissance’, Vienna, 1929, pp. 102-4, no. 469, pl. 117). However, the strikingly high quality that is found on these two large spectacular examples of mounted hardstone vessels can also be seen on more intimate little items, such as a gold case for reading-glasses (preserved among the Hapsburg treasures in the Kunsthistorisches Museum in Vienna, inv. no. 2080/81). If the arabesque zone on the underside of the rim-mount of the Waddesdon Vase is compared with these three documented examples, its lack of crisp definition and subtle complexity of interlace becomes immediately apparent.
    The source for the enamelled bunches of grapes in high relief on the finial of the cover of the Waddesdon Vase seems likewise to have been either the 1543 Saliera by Cellini or a lesser piece like the gold enamelled mount of the hinged cover on the unique Renaissance glass ewer in the Ambras Collection (Kunsthistorisches Museum, Vienna, inv. no. 1393; illustrated in ‘Meisterwerke’, Vienna, 1955, where the mounts are attributed by the Museum to a French workshop; also E. V. Strohmer, ‘Prunkgefässe aus Bergkristall’, Vienna, 1947, p. 29, pl. 20; Hayward 1976, p. 372, fig. 376-8; it is also discussed as the probable source used by the faker Reinhold Vasters before 1872 when making WB.77). Unexpectedly, the finial on the Waddesdon Vase is made separately and fixed through a hole in the centre of the cover with four small gold clips; however, superficially the actual workmanship of the enamelled relief of the finial faithfully creates a convincing effect. Nevertheless, neither the enamelling of the four kneeling satyrs nor that of the background of vines around them is of a comparable quality, whilst on the interior of the cover the translucent enamelled vine scroll (with birds and insects) on the gold disc conflicts with the mid-sixteenth-century French character of the cover. Furthermore, the method of attaching the gold disc bearing this feeble design is exceptional, for it relies entirely on friction caused by pushing the thin enamelled gold fillet into place around the circumference.
    Therefore, although the goldsmith seems to have drawn upon a direct knowledge of certain Fontainebleau court works of art, he seems to have tackled the more technical problems concerning the mounting of the Waddesdon Vase with the methods of an early nineteenth-century skilled craftsman - as, for example, when he devised the internal construction of the foot, the method of bolting the various parts together or the riveting of the neck-mount through the walls of the agate. However, it has to be remembered that it was not until 1819 that Cellini's saltcellar was published as the work he had made for Francois I in 1542-3 and had so fully described in his Vita and in the Trattati (see Hugh Tait, ‘Catalogue of the Waddesdon Bequest in the British Museum: Vol. II. The Silver Plate’, London, 1988, pp. 15-16). It, too, had survived almost unnoticed in the Schloss Ambras until 1806 when it was brought to Vienna. What a profound impact the Ambras Collection must have had on those who were able to see it in the Lower Belvedere Palace, especially after 1819 when the richness of its contents was made available in print (see Alois Primisser, ‘Die Kaiserlich-Königlich Ambraser-Sammlung’, Vienna, 1819). After the upheavals of the Napoleonic Wars and the Congress of Vienna in 1815, there was time and money for such pursuits, and both the craftsmen and their patrons were unlikely to have remained indifferent to these rediscovered Renaissance masterpieces from the 'Tyrolese line' of the Hapsburg dynasty.
    Indeed, one other major work created in this taste was sold to William Beckford in 1819 by a London dealer, Edward Baldock, as the work of Benvenuto Cellini and has become known as the Fonthill ewer. Fortunately, it is now in public ownership, and after prolonged study it has been established that although the carved smoky crystal-quartz bowl is a genuine hardstone bowl that was carved around 1680 in the Prague workshop of Ferdinand Miseroni, it could not have been mounted in its present elaborate gem-set and enamelled gold mounts because they are in the 'Celliniesque' style of more than a century earlier (see Clare Vincent in ‘The Jack and Belle Linsky Collection in the Metropolitan Museum of Art’, New York, 1984, p. 179, no. 95, illus. in colour; Tait 1988, pp. 16-18, fig. 2). Not only are there contemporary written accounts of the Fonthill ewer both in William Beckford's correspondence and in the publications of 1822 (Christie's sale catalogue) and of 1823 (J. Rutter, ‘An Illustrated History and Description of Fonthill Abbey’, Shaftesbury, 1823, p. 7), but there are also illustrations which establish that it has not changed its appearance since Beckford purchased it. It is still a work of outstanding technical brilliance but, alas, the identity of the goldsmith who made the 'cinque-cento' dragon-handle and openwork foot still eludes detection. His workshop could be suspected of having made the gem-set and enamelled gold mounts on a pseudo-Renaissance rock-crystal ewer, now preserved in the Abegg-Stiftung, Berne, Switzerland. Neither the shape of the vessel, with its almost hemispherical domed foot and elongated neck, nor the style of the wheel-engraved foliate decoration is correct for the sixteenth century, although it is characteristic of the early nineteenth-century collectors' perception of the Renaissance style. While the mounts are less encrusted with gemstones than those on the Fonthill ewer, the quality of the enamelling and of the modelling of the dragon-handle and the masks is strikingly similar, especially the distinctive technique of leaving the faces in burnished gold without any enamel, except for the white teeth and the coloured eyes, with their dark pupils contrasting with the surrounding brilliant white enamel. On these two ewers neither the actual forms of the dragons nor the patterns of their scaly skins have been repeated, and yet the dramatically effective sense of springing movement and tension is the same - perhaps, indicative of a common origin and of the high quality of this goldsmith's versatile use of particular motifs. His workshop, copying from the Chatsworth drawing - or, more probably, from the modern copy of it that is now in the Victoria and Albert Museum - could, likewise, have been responsible for the creation of the 'Cellini-esque' mounts on the Waddesdon Vase.

    Bibliography: ‘Catalogue of European Enamels from the earliest date to the end of 17th century’, Burlington Fine Arts Club, London, 1897, p. XX, p. 75, no. 245; C. Davenport, ‘Cameos’, London, 1900, p. 43, col. pl. IV.; Charles Hercules Read, ‘The Waddesdon Bequest: Catalogue of the Works of Art bequeathed to the British Museum by Baron Ferdinand Rothschild, M.P., 1898’, London, 1902, no. 68, pl. XVII; Eugene Strong, ‘Catalogue of the Gems at Chatsworth’, London, 1909; O.M. Dalton, ‘The Waddesdon Bequest’, 2nd edn (rev), British Museum, London, 1927, no. 68, pl. X; E. Kris, ‘Golschmeidearbeiten des Mittelalters, der Renaissance und des Barock. I Teil: Arbeiten in Gold und Silber’, Publikationen aus den Kunsthistorischen Summlungen in Wien, Band 5, Vienna, 1932, p. 27; M. C. Ross, The Rubens Vase: its History and Date, ‘Journal of the Walters Art Gallery’, VI, Baltimore, 1943, pp. 31 ff, fig. 17; J. Pope-Hennessy (ed.), ‘The Life of Benvenuto Cellini’, London, 1949, p. 484, pl. IX; E. Camesasca, ‘Tutta l’opera del Cellini’, Milan, 1955 (2nd edn 1962), pp. 67-8, pl. 80; D. Harden and J. M. C. Toynbee, The Rothschild Lycurgus Cup, ‘Archaeologia’, xcvii, 1959, p. 202, p;. LXXV, a-b; J. F. Hayward, The Mannerist Goldsmiths - I (Italian Sources), ‘The Connoisseur’, March 1962, p. 157, fig. 2; Hans-Peter Bühler, ‘Antike Gefässe aus Edelsteinen’, Mainz, 1973, p. 76, no. 108, pl. 36; Hugh Tait, ‘The Waddesdon Bequest: The Legacy of Baron Ferdinand Rothschild to the British Museum’, London, 1981, pp. 57-60, figs 36-8; Hugh Tait, 'Catalogue of the Waddesdon Bequest in the British Museum. II. The Curiosities', British Museum, London, 1991, no.15, pl. VIA, figs. 189-211; Hugh Tait, ''La collection de Fonthill: la première collection de faux Renaissance', in 'Les vases en pierres dures', Musée du Louvre, Paris, 2001, pp.222-223.


  • Bibliography

    • Tait 1991a 15 bibliographic details
    • Read 1902 68 bibliographic details
    • Dalton 1927 68 bibliographic details
    • Meredith 2015 82 bibliographic details
  • Location

    On display: G2a/dc2

  • Exhibition history

    1998 May 11-Nov 1, Bucks, Waddesdon Manor (National Trust), 'Baron Ferdinand de Rothschild Collection'

  • Condition

    (i) The carved agate body is undamaged, except for the loss of the neck-rim, stem and foot, leaving an almost smooth horizontal break below the ring of acanthus leaves; the break rises slightly at one point (almost directly beneath one of the two handles). The full extent of the damage that occurred around the mouth of the agate vase is hidden beneath the immovable gold that now forms the neck of the vase. (ii) The enamelled gold mounts and cover are well preserved, except for minor areas of damage to the enamel, especially to the white enamel, which appears to have crazed, both on the rams' heads (on the stem) and on the reclining Bacchus figures (in the centre of each of the four sides of the foot).

  • Subjects

  • Associated names

  • Acquisition name

  • Acquisition date


  • Acquisition notes

    This collection is known as the Waddesdon Bequest under the terms of Baron Ferdinand Rothschild’s will.

  • Department

    Britain, Europe and Prehistory

  • Registration number


Two-handled vase carved from a single block of honey-coloured agate. The body of the vase is hollow and oviform, except for the two integral handles that project on either side at the widest part. Each handle is in the form of a pair of horns rising almos

Two-handled vase carved from a single block of honey-coloured agate. The body of the vase is hollow and oviform, except for the two integral handles that project on either side at the widest part. Each handle is in the form of a pair of horns rising almos

Image description



If you’ve noticed a mistake or have any further information about this object, please email: 

View open data for this object with SPARQL endpoint

Object reference number: MCN313

British Museum collection data is also available in the W3C open data standard, RDF, allowing it to join and relate to a growing body of linked data published by organisations around the world.

View this object

Support the Museum:
donate online

The Museum makes its collection database available to be used by scholars around the world. Donations will help support curatorial, documentation and digitisation projects.

About the database

The British Museum collection database is a work in progress. New records, updates and images are added every week.

More about the database 


Work on this database is supported by a range of sponsors, donors and volunteers.

More about supporters and how you
can help