Potpourri Vase (Vase potpourri a vaisseau)

A ship modeled entirely in soft-paste porcelain, complete with rigging, portholes and a pierced pennant, made at Sevres in 1764 to a form Jean-Claude Duplessis le Pere had designed seven years earlier. Only about twelve examples were ever produced and only ten are known to survive; this is the one now in Baltimore, and the only one of the ten that the curator who has personally examined the other nine has not yet seen in the porcelain.

Its documented history runs from an English Rothschild collection through two Paris and New York dealers to Henry Walters, who bought it in 1928 and left it, with the rest of his museum, to the City of Baltimore on his death in 1931. The Walters Art Museum's own provenance note leaves the Rothschild acquisition itself undated and unexplained, which this record preserves as a gap rather than closing it with a guess.

The form is documented, priced by its own difficulty, and imitated by name since the nineteenth century, yet this specific vase has not been the subject of a single further recorded transaction in ninety-eight years. That is itself a market history, and it is recorded here as one.

What it is

The vase takes the form of a stylized sailing ship set on a wave-form base, rigged in porcelain rather than rope. At each side a lion-like beast mask, crowned with reeds, grips a tapered spar in its mouth in place of a bowsprit. The hull carries two painted reserves: on the front, within a cartouche, sailors work on a dock with a ship behind them; on the back, a marine trophy of a shell, an anchor, strands of seaweed, two crossed oars, a flag and a trident, bound together with a painted ribbon. Both scenes are the work of Jean-Louis Morin, one of the leading figure and marine painters on the Sevres payroll in the 1760s. Above the hull, a pierced cover in the shape of furled rigging is topped by a long, wind-blown pennant, also pierced, so that the whole superstructure reads as sail and cordage rendered solid.

The body is soft-paste porcelain, the eighteenth-century Sevres paste, built up with polychrome enamel grounds and extensive gilding. The vase now stands on a later, non-original stand of carved and gilt wood; the Walters Art Museum's own cataloguing separates this replacement mount from the porcelain vase itself in its materials description, which is the correct way to read the piece, since a plinth is not evidence about the object it carries.

How a ship became a perfume box

The pierced cover is not only ornamental. This is a pot-pourri vase, a container for scented dried petals, herbs and spices, and the openwork rigging is engineered to let the scent escape into a room while keeping the contents out of sight. Eighteenth-century women commonly compounded their own pot-pourri mixtures at home, sometimes ageing a blend for years before it was considered ready, and a vase of this kind was made to be filled, refreshed and used, not only displayed.

It was also designed to be seen as part of a set. Vases of differing shapes, this one among them, were sold by Sevres to form a garniture across a mantelpiece or side table, arranged for visual rhythm rather than function alone. The documented case of the type is the example now in the Louvre, delivered on 30 May 1760 to Madame de Pompadour as the centerpiece of a five-piece garniture for her bedroom at the Hotel d'Evreux, the building now the Elysee Palace. There is no equivalent documentation placing the Walters vase in a specific eighteenth-century garniture; this is stated as a fact about the type, not assumed for this individual vase.

Duplessis's ship, and why a boat

The vaisseau a mat form was designed in 1757, and the design is generally credited to Jean-Claude Duplessis le Pere, a goldsmith by training who held the title Goldsmith to the King before becoming artistic director and chief modeler at Sevres. The factory's own institutional history of the form describes it as evoking the precious-metal ship-form table ornaments, the nef, that had adorned royal and noble tables since the Renaissance; the Frick Collection's catalogue entry for its own example uses the same word, stating plainly that the shape partly derives from the nef.

A second reading, reported by more than one secondary source but stated with consistent hedging in each, connects the ship to the coat of arms of the City of Paris, which bears a ship, and proposes that Duplessis's design gestures at this civic emblem, or alternatively at France's naval engagement in the Seven Years' War, which ran from 1756 to 1763 and so overlaps the form's entire production run. No source consulted in this research states a documented, contemporary written justification from Duplessis or the factory for the choice of a ship; the arms-of-Paris and Seven Years' War readings are both argued by later cataloguers rather than demonstrated from an eighteenth-century document, and this record states them as arguments, not facts. What is documented without qualification is the nef association and the vase's court and aristocratic clientele.

An engineering problem disguised as a centerpiece

The vaisseau a mat was one of the most technically difficult forms Sevres ever put into production. The multiple open piercings through the hull, rigging and pennant weakened the unfired body at exactly the points where the kiln's heat could cause it to sag or collapse before firing was complete. Losses in the kiln were common, and the difficulty and cost of a successful firing made the finished vase one of the most expensive models the factory sold in this period, despite carrying gilded fleurs-de-lis that might suggest a piece reserved for the crown; it was not, in fact, sold exclusively to the royal family.

Between 1757 and 1764, the span this vase's own 1764 date mark places it at the very end of, roughly twelve examples of the form are recorded as having been produced in total. Ten are known to survive today. That is an unusually high loss rate even by the standards of eighteenth-century porcelain, and it is a direct consequence of the design's structural ambition rather than of later accident or war.

Ten ships, eight collections

The ten surviving vaisseaux a mat, cross-checked here against each holding institution's own published catalogue record rather than a single secondary list, are: three at Waddesdon Manor, Buckinghamshire, in bleu celeste, dark blue and petit vert grounds, of about 1761 to 1763; one at the Louvre, Paris, described by the museum as third size and dated about 1760, delivered to Madame de Pompadour in that year; one at the Wallace Collection, London, described in the museum's own title as of the first size, about 1761; one in the Royal Collection at Buckingham Palace, in a bleu lapis and green ground with gilded oeil-de-perdrix diaper, of about 1758 or 1759, part of George IV's Sevres purchases; one at the Metropolitan Museum of Art, New York, a pink-ground example of about 1757 to 1758 given by the Samuel H. Kress Foundation in 1958; one at the Frick Collection, New York, in green and blue grounds highlighted in gold, about 1760, bought by Henry Clay Frick in 1916; one at the J. Paul Getty Museum, Los Angeles, catalogued by the Getty itself as vaisseau a mat, deuxieme grandeur, about 1760, with front-panel painting attributed to Charles-Nicolas Dodin; and this vase at the Walters Art Museum, Baltimore, dated 1764 by its own factory mark, the latest-dated of the group in the sources consulted.

One tertiary source used in this research states that by the mid-nineteenth century all ten known examples had passed through English collections, and that today one is in France, five are in the United Kingdom and four are in the United States. Counting the institutions above independently against that claim gives exactly that split, which is a useful internal check even though the claim itself is not independently sourced beyond that one tertiary reference.

A closely related but separate object should not be confused with the ten: in 1895 the English firm Minton produced a facsimile of the Royal Collection's vaisseau a mat, made with Queen Victoria's permission and with access to the original for Minton's own artists, whose preparatory drawings survive in the Minton archive. That facsimile, documented at auction by Bonhams, is a nineteenth-century copy openly made and openly sold as such. It is not one of the ten eighteenth-century survivors and is not counted as one here.

Before Henry Walters: the Rothschild question

The Walters Art Museum's own published provenance for this vase opens with a Rothschild owner, and names him precisely: Sir Anthony Nathan de Rothschild, 1st Baronet, 1810 to 1876, of the English banking branch, a director of N M Rothschild and Sons and, from the early 1850s, owner of Aston Clinton House in Buckinghamshire, where he and his wife Louise built a substantial collection of paintings, faience and other decorative arts alongside a career that also included the High Shrievalty of Buckinghamshire and a stable of successful racehorses.

The Walters record itself, however, brackets this link with an explicit caveat: the date and mode of Sir Anthony's acquisition of the vase are unknown. This record follows that caveat rather than smoothing over it. The ownership by Sir Anthony is treated here as documented, because it is stated without qualification by the object's own holding institution; the manner and date by which he acquired it is treated as unknown, because the same institution says so.

This is the point at which version confusion becomes a real risk for this object, and it is worth stating plainly. The Louvre's own vaisseau a mat, a different vase entirely, of a different size and a different decade of production, passed at one point through the collection of Baron Alphonse de Rothschild, a member of the French branch of the family, before reaching the Louvre in 1984. Sir Anthony Nathan de Rothschild and Baron Alphonse de Rothschild are different men from different branches of a large family that collected Sevres porcelain broadly and in depth across generations. Nothing in the sources consulted here connects Sir Anthony to the Louvre vase or Baron Alphonse to the Walters vase, and no such connection should be inferred from the shared surname.

The dealers: Hodgkins and Seligmann Rey

From the Rothschild collection, the Walters Art Museum's provenance places the vase next with Edwin Marriott Hodgkins, in Paris, by 1927, again with the date and mode of acquisition unrecorded. Hodgkins is not an invented or vague name; he was a known specialist. Trading from 43 Old Bond Street from about 1900 and then 158b New Bond Street from 1904 to 1920, he built his reputation specifically in Old Sevres Porcelain, publishing a catalogue of his own Sevres collection in 1910. He relocated his business to Paris around 1920, and by 1928 was operating from 3 Rue de Berri. His personal Sevres holdings were substantially absorbed, over time, into the Walters Art Museum's own collection, which places this transaction inside a documented, ongoing relationship between Hodgkins and Henry Walters rather than a one-off sale.

From Hodgkins, the vase passed to A. Seligmann, Rey and Co. in New York, which had it in stock by 1928, the year Henry Walters bought it. This firm was the New York branch, established in 1921, of Arnold Seligmann and Cie. at 23 Place Vendome in Paris. Arnold Seligmann was the younger brother and former partner of Jacques Seligmann, who was, until his death in 1923, the leading Paris dealer in French eighteenth-century decorative arts and a dealer with whom Henry Walters did an enormous volume of business. A 1912 family dispute split the brothers' firms; Arnold continued independently at the Place Vendome address and, through the New York branch, continued to place major French decorative arts with American collectors, Walters among them, through the 1920s, 30s and into the 1940s.

Henry Walters, 1928

Henry Walters (1848-1931) did not start this collection; he inherited its foundation from his father, the Baltimore shipping and railroad magnate William T. Walters (1820-1894), and spent much of his own life extending it into a deliberate, comprehensive survey of art history rather than a personal taste collection alone. In September 1900 he bought adjoining houses in the Mount Vernon neighborhood of Baltimore and had them adapted into a purpose-built, palazzo-style gallery, which opened to the public in 1909, so that by the time the vaisseau a mat entered the collection in 1928 the building meant to hold it already existed and was already operating as a public museum.

By the 1920s Walters was a major, repeat buyer of French eighteenth-century decorative arts from the Seligmann dealer network, a pattern this vase's route through Arnold Seligmann's successor house fits. The same year, 1928, Walters also bought a later Sevres eared vase from the same firm, A. Seligmann, Rey and Co., a second, independently documented purchase from the same dealer in the same twelve months that corroborates rather than merely illustrates the buying pattern.

No price for this specific vase was found in any source consulted for this record, whether from Hodgkins, from Seligmann Rey, or as a total paid by Walters, and none is asserted here. This is recorded as an open question rather than estimated. Walters died on 30 November 1931, bequeathing his entire collection, by then numbering more than 22,000 objects, along with the museum building and an endowment, to the City of Baltimore; the Maryland General Assembly incorporated the Trustees of the Walters Art Gallery in 1933, and the institution opened in its present public form in 1934, with the vaisseau a mat among the objects that made that transfer.

Faking a masterpiece: Samson, and the Minton exception

Sevres soft-paste porcelain of this quality has been extensively imitated since the middle of the nineteenth century, most prominently by the Paris firm founded by Edme Samson, active from the 1840s onward, whose output is now understood to make up a large share of the material sold on the open market as Sevres. Samson's wares are technically distinguishable: Samson used hard-paste porcelain rather than the soft paste of the genuine eighteenth-century factory, its glazes tend to be glassier and more glossy than the original, its modeling can be stiffer or run to the wrong scale, and its color and gilding are frequently heavier or less accurate than the period original, even where a Samson piece has had its own marks removed or altered to pass as authentic.

No source consulted in this research raises any authenticity concern specific to Walters 48.559. Its factory marks, the interlaced Ls, the date letter and the painter's monogram, are recorded by the museum without qualification, and no dealer, cataloguer or scholarly source in this research questions the vase's status as an eighteenth-century Sevres original. The one documented copy connected to the vaisseau a mat form as a whole is the 1895 Minton facsimile of the Royal Collection's example discussed above, made openly and with the Royal Collection's cooperation; it is a legitimate historical reproduction, not a case of intended deception, and it is unconnected to the Walters vase specifically.

A market that essentially does not exist

No auction sale of this vase, and no auction sale of any of the other nine surviving vaisseaux a mat, was found in the sources consulted for this record. Since Henry Walters's purchase in 1928, this specific vase has not changed hands again; it passed directly from his private collection into a public museum by bequest three years later and has remained there for ninety-five years since. Of the other nine, all are now held by museums or by the National Trust at Waddesdon Manor, none of which are active sellers of core collection pieces of this rarity.

That absence is itself informative rather than simply a gap. A form of which only ten examples were ever made and only ten survive, all now inside institutional walls, is a form that has effectively been removed from the market for the better part of a century. Where a modern comparable price might otherwise anchor a valuation, none exists in the record, and this record does not manufacture one.