La Main aux algues et aux coquillages

A human hand, 33.4 centimetres high, modelled hot in crystal, rising out of a wave of blue-green seaweed with shells applied to its back in high relief. Emile Galle made it in 1904 and died on 23 September of that year. It was shown at Nancy five weeks later, in a case of his own design called The Depths of the Sea, at an exhibition whose opening had been put back a fortnight because he had died. The Musee d'Orsay, which has had it since 1990, calls it doubtless his last work in glass.

There are four of them. Two are in the Musee de l'Ecole de Nancy under two different titles and at two different heights, one is in Paris, and a trial piece was sold at Versailles in December 1970. Anyone citing a photograph of a Galle hand is citing one of four objects, and the only reliable way to tell them apart is the measurement.

The signature on this one reads Galle, engraved in relief at the base, and it has no star. From 1905 the Galle works put a star beside the name on its industrial production to mark it as made after the founder's death, and kept it there until May 1908, when the firm removed it because customers had begun saying that anything with a star was no longer Galle. That three-year mark is the cleanest authorship line in nineteenth and twentieth-century decorative art, and it is the reason a vase signed Galle can be by Galle, by his workshop in his lifetime, by his widow's factory in 1906, or by a joint-stock company in 1930.

There are four hands, and the difference between them is millimetres

Start here, because everything else about this object is downstream of it. Four examples of the model are recorded.

The Musee de l'Ecole de Nancy holds two. One is catalogued as La Main aux algues, inventory 955.1.2, 38.5 by 12.5 centimetres, acquired in 1955. The other is catalogued as La Main aux algues et aux coquillages, inventory RD1, 32.5 by 13.2 centimetres, acquired in 1971. The Musee d'Orsay holds the third, inventory OAO 1207, 33.4 by 13.4 centimetres, acquired in 1990. A fourth was sold at public auction at Versailles on 13 December 1970 as lot 67 and is in a private collection.

The two titles are not a cataloguing muddle. They describe a real difference in the objects. On 955.1.2 the engraving of the seaweed on the back of the hand was begun and never finished, and no shell was ever applied to it: it is a hand with seaweed, and the title is accurate. The other three carry the shells. Anyone reconciling the labels should start from that and not from an assumption that curators were careless.

What the four are to each other is disputed by the two scholars who have worked on them. Francois Le Tacon holds that the two Main aux algues are first versions of the two Main aux algues et aux coquillages. Philippe Thiebaut rejects that, on the ground that reading a linear development into Galle is a misreading of how he worked, and argues instead that Galle had two fully finished pieces in hand. Thiebaut's supporting point is the more useful one for anybody cataloguing this maker: unique works do not exist in Galle's output. When an exceptional piece was undertaken, commissioned or not, four or six examples were generally made at the same time as a precaution.

So a reader who finds an image captioned La Main aux algues et aux coquillages is looking at one of three objects, and the caption will usually not say which. The measurement is the identifier. Use it.

What it is

A right hand, upright, palm forward, fingers together and slightly parted at the tips, rising out of a swirling base of blue-green glass worked to read as seaweed and water. The body of the hand is a smoky, marbled crystal shot through with amber and brown inclusions that run like veins or like weed suspended in water. Applied to the back of the hand and to the fingers, in high relief and in contrasting opaque colours, are shells: a pink whelk at the centre, a pale disc, a small dark spiral, a red form at the wrist.

The Orsay describes the technique as crystal partly blown and modelled hot, with inclusions and applications, then wheel-engraved. Every one of those operations is a different trade and a different temperature. The hand was formed while the glass was soft; the inclusions were introduced into the body of the metal; the shells were applied hot and left standing proud; the detail was cut cold with a lapidary wheel afterwards. The finished object is 33.4 centimetres high and 13.4 wide.

It is not a vase, although it is sometimes catalogued as one. The Orsay classifies it as a decorative object. It holds nothing and it does not function.

Made in 1904; shown five weeks after he died

Emile Galle died at Nancy on 23 September 1904, aged fifty-seven. The Exposition d'art decoratif lorrain, organised by the Societe lorraine des Amis des Arts at the galeries Poirel in Nancy, was due to open on 15 October. The opening was put back to 29 October because of his death.

The hand was in that exhibition. It was shown inside a display case Galle had himself conceived, catalogued as number 88 and titled Les Fonds de la mer, the depths of the sea, and it was reproduced as plate 34 of the exhibition catalogue. So the first public appearance of the object was in a room that its maker had planned and did not live to see, in a vitrine of his own devising, at a show that had been delayed to bury him.

That sequence is the reason the object is called his artistic testament, a phrase the Musee d'Orsay used for the title of a monographic exhibition built around this single work in 2004. It should be treated as a description of the circumstances rather than a statement about his intentions. Nothing consulted here records Galle saying that this was his last word.

The signature is the argument

The Musee d'Orsay's inscription line is nine words long in the original and it decides the authorship question: a signature engraved in relief at the base, reading Galle.

Take care with what the cut of that signature does and does not prove, because it is easy to get backwards. A signature engraved in relief is a cameo signature: the ground is taken away and the letters are left standing. Samuel Provost, who has done the archival work on the marks, holds that on the best pieces the signature is worked in intaglio, cut into the glass in an elaborate hand-script, and that a plain cameo signature on a piece claiming to be top-flight is a warning rather than a reassurance. On his test the relief signature on this object argues nothing in its favour. The authorship here does not rest on the mark at all. It rests on the date of 1904 and on the object's presence in the Nancy exhibition of that year, in a case Galle designed, catalogued and photographed at the time.

What the mark does settle is the other half of the question, and it settles it decisively: there is no star. The Galle works marked its posthumous industrial production with a star beside the name. If this object carried one it would not be Emile Galle's work in the sense the label claims, and the label would be wrong. It does not, and it is not.

The rule that follows is simple enough to write on a card and is broken constantly in the trade and occasionally in print: check the date of any Galle object against 23 September 1904 before calling it his own work, and then check the mark. A piece made in 1906 and signed Galle is a Galle-works piece, not a Galle.

The star: what it was, and exactly when it ran

After Galle's death his widow Henriette Galle took over the direction of the factory, which she ran until her own death in 1914. The industrial series produced under her direction were marked with a star beside the founder's name to identify them as posthumous.

The dates given for that mark in most reference books are 1904 to 1906, and they are wrong. That chronology descends from a single piece of testimony given by Rene Dezavelle in 1974. Samuel Provost re-dated the mark from the firm's own correspondence, principally Henriette Galle's letters to Albert Daigueperce, who managed the Paris depot. The decision to add the star was not taken before January 1905, and the mark was introduced in the first half of that year. It was removed in May 1908.

The reason for the removal is recorded and it is commercial. The factory's director Emile Lang reported customers who were claiming that everything bearing the star was no longer Galle. The mark that had been added out of respect had turned into a discount. Three years after introducing an honest disclosure of authorship, the firm withdrew it because the market punished it.

That is a fact worth holding on to well beyond this object. A maker's mark that truthfully distinguishes posthumous production from lifetime work is exactly what a provenance archive wants, and the reason so few of them exist is that they cost the estate money.

The full sequence of marks, and how to read one

Provost's chronology of the marks after 1904 runs as follows. A first mark, standardised after 1905, derives from the prototype used for the industrial series before Galle's death. The second, from 1905 to 1908, is the underlined crossed-G with the star, with variants in the position of the star and in whether the final e is a Latin acute or a Greek epsilon. The third, from 1908 to about 1920, is the same underlined crossed-G with the star removed; it is the longest-lived mark and it carried the firm through the First World War.

Then the marks multiply with each reorganisation of the company. A simple unaccented-underscore Galle appears from about September 1920, after the departure of the sales director Daigueperce and of the designers Louis Hestaux and Paul Nicolas, and is concentrated on small pieces: bonbonnieres, atomisers, cups, averaging under fifteen centimetres. A short-lived mark with an inverted G-tail runs from about 1920 to 1924 and is associated with the designer Jean Rouppert, who left for Lyon in that year; one grapevine vase carrying it is incised with the date 28 November 1922. From 1925, when the firm became a joint-stock company and recruited George Dethorey and Theodore Ehrhart, an underlined tangent-G mark appears on the larger and costlier pieces. From about 1927 a vertical, Japanese-styled signature reading top to bottom appears, overwhelmingly on the crackled vase series. Two ornamental marks, one horizontal and one vertical, run from about 1927 or 1928 to the end on the prestige series: the polar bears, the elephants, the large floral pieces.

The end date matters too and it has also been corrected. The main furnace was shut down in 1931 and the design team had been dismissed in late 1930, but production continued until 1936; Francois Le Tacon established that in 1993 against a literature that had closed the factory in 1931. Provost notes that there is almost no hard evidence at all for the decade from about 1928 to 1936, because the firm stopped exhibiting.

The practical upshot for a cataloguer is that a Galle mark is a dating instrument. Read it properly and it gives a date range of a few years. Read it as a name and it gives you nothing but a brand.

What the museums call the later pieces

The best way to see how this works is to look at what careful institutions actually write on the label, because they do not write Emile Galle.

The Rijksmuseum holds a vase with maple branches, inventory BK-1983-35, 29.5 centimetres high, in colourless glass with a double yellow and orange-red overlay, etched and cut. Its maker line reads glassblower: Emile Galle, workshop of. Its date is given as about 1904 to about 1906. Its description records, in the museum's own words, the signature Galle with a star. That is a posthumous object, correctly described, and the star is on it.

The same museum holds a small vase with honesty seed-pods, BK-1972-74, 8.8 centimetres high, again given to the workshop of Emile Galle, dated about 1906 to about 1914, with the signature Galle and no star. That is a post-1908 object, correctly described. Two vases, one museum, one signature word, ten years and one mark between them.

The Metropolitan Museum of Art catalogues an Autumn Crocus vase of about 1900 to the studio of Galle and, in its own metadata, to the Etablissements Galle. That is the third category, and the one most often lost: workshop production made while Galle was alive and running the business, which is neither his own hand nor posthumous.

One correction, offered here rather than to the museums. The Rijksmuseum's date range of about 1904 to about 1906 for the star vase is the old Dezavelle chronology. On Provost's revised dating the star cannot have been applied before the first half of 1905 and ran to May 1908, so that vase should read about 1905 to 1908. The object is not misattributed; the window is simply drawn in the wrong place, and it is drawn in the wrong place in a great many catalogues.

Who ran the works after him

Henriette Galle directed the factory from her husband's death until she died in 1914, with Victor Prouve, Galle's friend and a founder of the Ecole de Nancy, as artistic director. The First World War stopped production. Afterwards Paul Perdrizet, Galle's son-in-law, restarted it, added new designs and concentrated on the multi-layer cameo glass in floral and landscape patterns that most people now picture when they hear the name.

That is the crucial commercial fact about this maker. The overwhelming majority of objects on the market signed Galle were made by a company, after his death, to designs he did not make, using an industrial acid-etching process he had introduced but did not personally execute. They are not fakes. They are perfectly genuine products of the Etablissements Galle. They are simply not by the man whose signature is on them, and the difference in value between the two is two orders of magnitude.

The fakes, which are a separate problem from the posthumous work

Alongside genuine posthumous production there is a substantial body of outright forgery, and Provost divides it in two.

The first category is counterfeit: pieces carrying imitation Galle signatures with no connection to the factory at all. Some are newly manufactured copies, principally from Romania and China. Others are older French and Lorraine glass by other makers, Paul Nicolas, De Vez, Burgun and Schverer among them, rebaptised with a Galle signature. That second kind is the harder one, because the object is a genuine period piece of related manufacture and only the mark is a lie.

The second category is alteration: genuine Galle objects modified, to hide damage or to raise the apparent quality, to the point where Provost's view is that they cease to count as Galle originals. A cut-down rim, a re-engraved passage or a replaced element turns an authentic object into something else, and the trade does not always say so.

The signals Provost sets out for the counterfeits are art-historical rather than mechanical: decorative combinations that make no sense, such as terrestrial plants with marine elements; naive or cartoonish drawing well below the workshop's standard; visible erasure or modification of the original design; clumsy hot applications; motifs the workshop did not use, such as trails of bubbles in a marine subject; and a signature left in flat cameo on a piece that claims to be of the first quality, where the firm cut intaglio.

From the trade side, the appraiser Philip Chasen adds that on authentic pieces the decoration was wheel-polished to a semi-gloss against a matte ground, so a reproduction that is uniformly matte gives itself away; that a dark band round the rim and a straight-cut rim are bad signs; that the foot is often too large and bulbous; and that the accent on the final e is a constant on authentic signatures. Provost's chronology qualifies that last point usefully: the accent is always there, but it is sometimes a Latin acute and sometimes a Greek epsilon, and which one it is helps to date the mark.

Chasen's own figure for how much of this is in circulation is the one worth quoting. Of the appraisals he carries out for individuals and insurance companies, about half turn out to be fake.

What Galle actually costs, and why the range is so wide

The name spans four orders of magnitude, and the reason is everything above.

At the top, a marqueterie de verre vase with an iris, called Bleu fleur d'eau, made between 1901 and 1904, 32 centimetres high on a gilt bronze base and worked over entirely with the wheel, was sold at Sotheby's in Paris on 21 May 2015 against an estimate of 200,000 to 300,000 euros. It made 435,000 euros including fees. Three examples of the model are documented; one was shown at the Dresden international exhibition of 1901 and another at the Ecole de Nancy exhibition in Paris in 1903. Its provenance was as good as a Galle provenance gets: acquired directly from Emile Galle in 1904 and in the same family ever since.

The auction record for the maker is older and higher. Sotheby's sold a Les Coprins lamp in 1989 for about 1.1 million US dollars: one of six examples made in 1902, in iridescent blown glass in the form of a group of mushrooms, on a wrought-iron base, eighty centimetres high. The exact date within 1989 is not established in the sources consulted here.

At the other end, Sotheby's offered a small enamelled and gilt glass vase of about 1900, nineteen centimetres high, signed with the EG monogram and the Cross of Lorraine and with Emile Galle Nancy underneath, in its Important Design sale of 25 June 2020, at an estimate of 2,000 to 4,000 euros. That is an estimate and not a result: what the vase actually made is behind a login and this record has not obtained it. It is quoted here only for what the house expected, which is the honest use of an estimate and the only one.

An expectation of two thousand euros at one end and a realised million dollars at the other, under the same signature. What separates them is date, hand, technique and mark, in that order, and the hand is the one the price databases cannot see.

What it meant, and which parts of that are evidenced

The Musee d'Orsay's own reading of the hand is that its form recalls Buddhist ritual gesture, the votive offerings of antiquity and medieval reliquaries: a hand held up, presented, offered. The sea supplied Galle with a large part of his decorative vocabulary, and the museum sets the object beside Baudelaire, quoting the opening of L'Homme et la mer, that the free man will always cherish the sea and that the sea is his mirror in which he contemplates his soul.

Galle was explicit about this kind of thing in his own lifetime. He read a paper called Le Decor symbolique to the Academie de Stanislas at Nancy on 17 May 1900 and published it that year. Whatever a reader makes of the hand, the maker was a man who had set out in print, four years earlier, a theory of how ornament carries meaning.

Two further readings should be graded and separated. The first is the museum's own note that the object appears to confirm an oral family tradition, that the hand was a protest at the sinking of the Russian squadron at Port Arthur during the Russo-Japanese war of 1904 to 1905. That is family testimony reported by the holding institution, and it should be read as such: not a document, not nothing. It also carries a chronological caution, since the object was finished in 1904 and the war ran into 1905.

The second is the reading, current in the French literature on the piece, that the hand belongs with Ophelia, drowned among weeds, as Millais painted her, so that the object hangs between emergence and dissolution. That is a modern interpretation by later writers and is recorded here as reception, not as intention.

What is documented, and is enough on its own, is that a dying man made a hand coming out of the sea, put it in a case called The Depths of the Sea, and did not live to see it shown.

Where it has been since 1990

The object stayed in the family from the 1904 exhibition until 1990, and the sequence inside the family is not documented. The donor was Jean Bourgogne, Galle's grandson, an entomologist of international standing, born on 14 February 1903 and so a year old when the hand was made; he cannot have been its first owner, and who held it before him is not recorded. When it was lent to the Galle exhibition at the Musee du Luxembourg in 1985 and 1986 the catalogue described the lender only as a private collection. The gift was at first recorded as anonymous and was afterwards attributed to Galle's descendants. It went through the acquisition committee on 15 March 1990, the council on 21 March and a decree of 10 April, and the work was attributed to the museum the same year. Bourgogne died on 10 March 1999. It was shown at once, in the museum's own survey of seven years of acquisitions, in the winter of 1990 to 1991.

Since then it has been out constantly and in serious company: the 1900 exhibition at the Galeries nationales du Grand Palais in 2000; a monographic exhibition at the Orsay in 2004 built around this single work and titled after it; the Galle centenary shows at the Edo-Tokyo Museum and the National Museum of Art in Osaka in 2005; Sade. Attaquer le soleil at the Orsay in 2014 and 2015; Les origines du monde at the Orsay in 2021; and Bleu profond, on artists and the invention of the deep sea, at Les Franciscaines in Deauville in 2025.

It is also, as of 2025, one of the hundred objects chosen for a book on works that tell the story of the climate. A glass hand made in Lorraine in 1904 is now being used to talk about the sea in the twenty-first century. That is where this object currently is, in both senses, and the Orsay record was last modified on 2 July 2026.

Why there is no photograph of the object on this page

This record carries no image of the hand itself, and the reason is a rule this publication applies against its own interest.

A photograph of a three-dimensional object carries its own copyright, however old the object is. Emile Galle died in 1904 and his work is long out of copyright; the photographs of it are not. Two images of this hand are on Wikimedia Commons under a public domain tag, and both tags are applied to the sculptor's death date rather than to the photograph. One of the two states on its own file page that it was taken at the Musee d'Orsay in 2021, by a named photographer, during an exhibition this object appeared in. A 2021 photograph is not in the public domain because the man who made the glass died in 1904.

Using it would put a false licence line on this page. So the plates below are all of other objects, every one of them released as CC0 by the museum that owns it, and every one of them chosen to carry a part of the argument: what Galle's own signatures look like, what the workshop's look like, and what the star looks like when a museum records it.