The Cleveland Apollo: Apollo Sauroktonos (Lizard-Slayer) or Apollo the Python-Slayer

A life-size bronze of a young Apollo, the only bronze version of a sculptural type that survives about twenty times in marble, bought by the Cleveland Museum of Art in 2004 from a Geneva dealer for a price the museum has never published. Its owner before 1994 was a German lawyer who says he found it broken in the grounds of a family estate in what had been East Germany, and there is no document of any kind for it before that date.

Greece said publicly that it had been fished out of the sea between Italy and Greece in the 1990s and got the Louvre to leave it out of a Praxiteles exhibition by threatening to withhold every loan it had promised. Fourteen years of metallurgy, published in full by the Getty in 2017, established that the figure had been in a fire, had never been in salt water, and had been soldered onto its display plate some time between the seventeenth and the nineteenth centuries. The same analysis found that the lead in the bronze does not come from the Aegean.

In 2013 the museum concluded that the creature at the god's feet is not the lizard Pliny describes but the Python of Delphi, and renamed the object accordingly. On that reading the sculpture is not a copy of a famous statue at all: it is the statue, and it stood at Delphi.

The only one in bronze

Roughly twenty sculptures of this type are known. Every one of them is marble except a single reduced-scale bronze copy and this object, which is life size, 150 centimetres tall, cast in bronze with copper inlaid at the lips and nipples and stone inlaid in the eyes. The right eye is the original. The god stands with his weight on one leg, one arm raised, the other lowered toward a small creature on a tree trunk that has not survived; the figure leans, and the lean is the thing every marble copyist tried to reproduce.

The physical consequence of being the only bronze is that this object carries information no marble copy can. A marble copy records a silhouette. A bronze records an alloy, a casting method, a set of joins, a corrosion history and, through its lead, a mine. Almost everything contested about this sculpture has been argued from that evidence rather than from its face.

What Pliny wrote, and why Cleveland stopped believing it

The type is named from one sentence. Pliny the Elder, listing the bronzes of Praxiteles in the Natural History, records a youthful Apollo called the Sauroktonos because he is aiming an arrow at a lizard creeping towards him. That sentence is the whole ancient authority for the subject, and for two hundred years it has governed how every version of the type is catalogued and displayed.

In 2013 Michael Bennett, then the museum's curator of Greek and Roman art, argued that the creature on the Cleveland bronze is not a lizard. He identified it as the Python, the serpent Apollo killed at Delphi, and proposed that the sculpture originally stood in the Sanctuary of Apollo there. The museum renamed the object in its own catalogue, which is why its title now carries both readings joined by an or. Bennett's statement of the case was that the sanctuary at Delphi is the only place in the ancient world where a Python-Slayer by Praxiteles belongs, and that the subject is Apollo's victory over the Python, a triumph of order over chaos at the place the Greeks called the Omphalos, the navel of the world.

This matters far beyond iconography. If the creature is a lizard, the Cleveland bronze is a version of a famous statue and could be a Roman copy of it. If the creature is the Python and the setting is Delphi, then the object is not a copy of anything, and the marble Sauroktonoi in the Louvre and the Vatican are copies of it. The reidentification and the attribution stand or fall together, and both remain contested outside the museum.

How it was cast

The bronze was made by the indirect lost-wax process. Wax manipulation marks survive inside the openings at the arms and feet where modern resin repairs do not obscure them, and the walls are thick, about half a centimetre in many places, cast in large individual sections. Rectangular patches and square chaplet holes across the surface are evidence of original ancient manufacture rather than later repair. The front half of the right foot was cast separately and welded on.

Radiography carried out before the 2004 purchase showed that the head had been joined to the neck and the lower left leg to the calf in a recent restoration, using modern adhesives and a modern internal armature, and that plaster and resin fill the neck and shoulders. The same radiographs showed the figure was nearly complete. What is missing is parts of both arms and the tree trunk the creature climbed.

The lead is not Greek

Lead isotope analysis was run on the figure, the detached hand, the creature, the baseplate and the solder. Every bronze part of the statue proved isotopically almost indistinguishable, which settles the question of whether the reassembled pieces belong together: they were cast from the same melt.

The second result is more awkward. The lead isotope ratios of both the statue and the solder exclude a geological source in the Aegean, and therefore exclude Laurium, the largest source of Aegean lead in antiquity. The published candidates are the Cevennes in France and the Balkans. This does not prove where the statue was made or where it stood, because metal travelled, and a fourth-century Athenian foundry could buy lead from anywhere in the Mediterranean. It does mean that the one piece of hard geological evidence in the object points away from Greece rather than towards it, in a case where a Greek ministry has asserted Greek origin.

The alloy itself is a low to medium tin bronze. Published figures give 6.7 to 7.6 per cent tin and 8.6 to 16.3 per cent lead for the creature, with means of 7.08 per cent tin and 4.70 per cent lead in the right arm. The baseplate is a different metal: a mean of 3.44 per cent tin against 19.52 per cent lead.

It burned, and it was never in the sea

The corrosion on the figure is consistent with an object buried for a long period and then exposed again to weather. Beneath the surface, metallography found a band of cuprite particles formed by oxygen diffusing inward at high temperature, and lumps on the back of the left thigh containing cassiterite crystals that solidified from an oxygen-rich melt at around a thousand degrees. The figure was in a fire, early in its history. The baseplate was not.

The corrosion products also excluded the possibility that the sculpture had been buried in a marine environment. That is a direct answer to the specific allegation Greece made, and it is the reason the museum was able to state in 2013 that laboratory tests conclusively prove the sculpture was not a recent discovery and was not recovered from the sea. A reader should note that this is a negative finding about salt water, not a positive finding about a German garden. It rules out one account of where the object came from. It does not establish another.

The baseplate is the clock

The most useful object in this record is the thin bronze plate the statue stands on, catalogued separately as 2004.30.d. It could never have supported the sculpture on its own, so it is a display fitting rather than an ancient base; ancient bronze statue bases are stone. It is barely corroded compared with the figure, and what corrosion it has is different in character. The soft solder holding it on has a composition inappropriate for an ancient solder, and its eta phase is still preserved where the eta phase has vanished from every other solder on the object. The published conclusion is that the plate was attached in the eighteenth or nineteenth century.

That single finding is the strongest argument the museum has. A statue mounted for display two hundred years ago was above ground two hundred years ago. It cannot have been dug up or fished up in the 1990s.

Two further details in the same study are rarely quoted and are worth more than the headline. Substantial epsilon-phase corrosion survives on the feet of the creature, consistent with its having been reattached to the tree when the statue was first displayed after excavation, which means there was an earlier restoration campaign nobody has dated. And under the little toe of the left foot the solder is ancient: the toe was once soldered to a piece of metal in antiquity, so how the statue was displayed in its own time has to be reconsidered.

One test said the opposite

The alpha spectrometry is the part of the technical record that has been reported as simpler than it is. The method tests for the radionuclide lead-210, which has a half-life of 22.3 years and cannot be renewed once smelting has broken the decay chain, so it is only measurable in metal younger than about 110 years. No measurable activity was detected in the samples taken for the later study, and the conclusion drawn was that the lead in the solder, in the bronze of the baseplate and in the cast-on front half of the right foot is all older than about 110 years.

In the earlier phase of work the result went the other way. Two solder samples were tested; one gave no usable result, and the other showed measurable radioactivity, which would indicate that the lead or the tin had been produced within the last hundred years. That is in the published paper, and it is the finding that lay behind the museum's 2008 statement to the Cleveland press that the solder was in the vicinity of a hundred years old. The two rounds of testing do not say the same thing. A record that quotes only the later one is quoting the half of the evidence that suits the object.

The garden in eastern Germany

The museum's own provenance line has three entries: Ernst-Ulrich Walter, Germany; Phoenix Ancient Art, 2004; the Cleveland Museum of Art, 2004 to the present. Everything before Walter is narrative rather than record.

The account the museum gave on acquisition is that the statue had been on a private estate in the eastern part of Germany, later East Germany, well before the Second World War, installed in the garden and taken for a late eighteenth or nineteenth century cast. After reunification in 1990 Walter reclaimed the family estate and found the sculpture in pieces. In 1994 Lucia Marinescu, formerly director of the National History Museum of Romania, saw the fragments while touring the estate; in the same year the sculpture was sold, and it was then reassembled and restored. Marinescu presented a paper on it at the sixteenth International Congress of Antique Bronzes in May 2003. A Cleveland representative saw it at the Geneva gallery of Phoenix Ancient Art in 2002, and the museum bought it in 2004, announcing the purchase on 22 June that year.

Reporting since has added intermediate steps the museum does not list: a sale to a Dutch dealer in 1994 and a sale to the Aboutaam brothers in 2001 through an anonymous collector. Those are journalistic attributions, not documents, and this record grades them accordingly. No photograph of the statue in the garden, no estate inventory, no customs entry and no export licence has been published.

What Greece said, and what it did

Greece raised the object publicly in 2006. In December that year a French news agency quoted an unnamed source in the Greek Ministry of Culture saying the Apollo had not been found in a German garden at all but was probably sold illegally after being found in the 1990s by an Italian vessel in international waters between Italy and Greece.

The consequential move was not the allegation but the loan policy. When the Louvre assembled its Praxiteles exhibition in 2007, Greece made clear that if the Cleveland bronze were included it would stop lending to French museums, and it withheld the Ephebe of Marathon, itself a bronze recovered from the sea in 1925. The Cleveland Apollo was not shown. A single object was kept out of the one exhibition where it could have been compared directly with the securely provenanced bronzes that would have tested it, and the comparison has still not been made.

The polemic that followed, in the magazine Apollo, called this cultural blackmail and argued that comparison with the Marathon bronze would have revealed, as photographs cannot, whether the Cleveland figure is an early or a late copy of Praxiteles. That sentence concedes more than its author intended: the writer defending the object assumed it was a copy.

What the dealers said

Phoenix Ancient Art is the Geneva and New York gallery of the brothers Ali and Hicham Aboutaam. They have not named the person from whom they acquired the Apollo. Ali Aboutaam's public position was that the gallery stands by its provenance. Hicham Aboutaam told the New York Times in March 2007 that the Apollo was proven to have been in circulation more than a hundred years ago, that the Greeks were not saying it should not be in Cleveland or that it was stolen, and that the objection was a public relations exercise.

The museum's own position, given by a spokeswoman in 2006, was that it had acquired the Apollo after more than a year of research by an international team of specialists working from legal, art-historical and technical angles, with an emphasis on the object's history.

The report nobody outside the museum could read

The methodological complaint against this acquisition was never really about the German garden. It was about who could see the evidence. When the museum told the Cleveland press in 2008 that the solder was about a hundred years old, the report behind that statement was not in the public domain, and the reporter said so in the same article. The archaeologist David Gill made the sharper point: the statue had been in fragments when first recorded, so a date for the solder joining it to a plate cannot by itself date the statue, and the full technical report was needed before anyone outside the museum could judge.

That objection was met, nine years later and largely on its own terms, when the Getty published the full analysis by Colleen Snyder, Ernst Pernicka and Peter Northover in 2017 under a Creative Commons licence, complete with the sample tables. The delay is part of the history of the object. For thirteen years the strongest evidence for the museum's case existed only inside the museum, and during those thirteen years the argument was conducted on assertion.

The price, and what a museum purchase price is

The Cleveland Museum of Art has not published what it paid. The figure most often repeated is five million United States dollars; one account gives 5.2 million. Both are reported rather than confirmed, and this record does not treat either as established.

What is documented is the source of the money. The purchase was made from the Severance and Greta Millikin Purchase Fund, an endowment rather than an appropriation, which is why the acquisition required no public vote and no external approval. There was no auction, so there is no hammer price and no buyer's premium to separate; this was a private treaty purchase from a dealer, and the only number that could ever exist is the invoice.

What the object is now

It is on view in Cleveland, catalogued under both names, with its detached forearm, its detached creature and its baseplate accessioned as separate objects and separately photographed. All of that photography is released by the museum under CC0, which is why this record can show the parts as well as the whole.

No claim has been filed. Greece has not gone to court, no restitution agreement exists, and the museum has not deaccessioned. Cleveland has returned other antiquities in the same period, including a bronze figure sent back to Turkiye, which shows the institution is capable of acting when the evidence turns; it has not turned here. The position is a stalemate resting on an absence: there is no document before 1994, and there is no proof that a document ever existed.