Bust of Flora

A painted wax bust of a young woman as Flora, about seventy centimetres high and weighing some twenty-eight kilograms, bought in London in 1909 by Wilhelm von Bode, director general of the Prussian collections, as a work by Leonardo da Vinci. Within two years more than seven hundred and thirty articles had been published about it in Germany, England, France, Italy, Austria and Denmark. The English press said it had been made in 1846 by a Southampton sculptor named Richard Cockle Lucas, and Lucas's son swore that he had helped his father build it out of candle ends.

The case ran for a century because the material fought back. The bust is made of spermaceti, the wax from the head cavity of the sperm whale, and a whale's carbon is not in equilibrium with the atmosphere. Radiocarbon measurements on it come out several centuries too old unless they are corrected for the marine reservoir effect, and no reservoir age for sperm whale material had ever been published.

Ina Reiche, Lucile Beck and Ingrid Caffy solved that in 2021 by calibrating against a wax relief Lucas made in 1850, of nearly identical composition, whose date is known. Every analysed constituent of the Flora bust dates after 1700. It cannot be by Leonardo. Who made it is still not proved.

What the object is

A bust of a young woman, bare-breasted, her head inclined and turned, holding a flower: Flora, the Roman goddess of spring and flowering. It stands about seventy centimetres high and weighs approximately twenty-eight kilograms. It is made of wax and is partly painted, and the polychromy is one of the things that persuaded people it was Renaissance work.

Internally it is hollow. X-ray radiography and tomography, neutron tomography and endoscopy establish that it was cast hollow in at least three wax layers by lost-wax moulding, the innermost layer being the most opaque. The cavity at the back can be opened, and when it was opened it was found to contain a wooden spar, newspaper and textile.

The material is the whole case, and it needs stating precisely. Gas chromatography with mass spectrometry shows that the wax is principally spermaceti, the fatty wax taken from the head cavity of the sperm whale, Physeter macrocephalus, with minor amounts of beeswax and a high content of palmitic and stearic acids indicating stearin or possibly tallow in the mixture. Infrared spectroscopy found lead stearates, probably a degradation product of lead white.

Wax was not a Renaissance material for finished sculpture

The first problem with the object was never technical. It was that no other wax sculpture of the Italian Renaissance is known.

Wax was central to Renaissance sculpture as a means, not as an end: it is the material of the model destroyed in lost-wax casting, and of the bozzetto. A finished, painted, exhibitable wax bust of this scale has no companions in the period. That absence should have been the first question asked in 1909, and it was, by other people, immediately.

Spermaceti compounds the problem in the same direction. It was rare and expensive in the Renaissance, and it was ordinary in the nineteenth century, when it was used for candles and for making sculptures from two-dimensional models. Chemists identified spermaceti in the Flora's wax in 1910, a year after the purchase, and again in the 1980s. The identification alone does not settle the date, because spermaceti did exist in the sixteenth century, but it puts the object in a nineteenth-century material world rather than a Renaissance one.

Bode buys it, 1909

Wilhelm von Bode was the most powerful museum figure in Europe, and he bought the bust in London in 1909 for the Berlin royal museums, hoping to add an outstanding Renaissance object to the collection. The price is given as 185,000 gold marks.

How he bought it is not agreed. The 2021 study states that Bode bought the bust at an auction in London. The German scholarly summary of the case says the seller was the London art dealer Murray Marks, acting through an intermediary named Willy Gretor, and notes that one source gives the price as 160,000 rather than 185,000 marks. This record does not choose between those accounts; it records that a sale of this size, in London, in 1909, is described in two incompatible ways in the modern literature, and that the difference matters because an auction and a brokered private sale carry different evidence about who knew what.

Murray Marks is a name that recurs in this field. He is the London dealer who bought Reinhold Vasters's studio drawings at the Aachen estate sale in the same year, 1909, the archive that would eventually re-date a large group of Renaissance goldsmiths' work in museums on two continents.

Bode attributed the bust to Leonardo da Vinci on the resemblance of the face to figures in well-known Leonardo paintings, and published the attribution himself in the Jahrbuch der Koeniglich Preussischen Kunstsammlungen in 1909, under a title that ends in a question mark, and again in Kunst und Kuenstler in 1910.

Seven hundred and thirty articles

Within two years of the acquisition more than seven hundred and thirty articles had appeared about the bust in the German and English press and in France, Italy, Austria and Denmark. The dispute continued into the 1930s.

It was not only an art-historical argument. It ran between Berlin and London in the years before 1914, and national feeling was part of its energy on both sides: a German museum had bought an English wax as an Italian masterpiece, or a British newspaper campaign was trying to humiliate the greatest museum director in Europe, depending on which paper the reader took.

The scale of the coverage is itself a fact about the object. Very few disputed attributions have generated a literature of that size, and almost none in the popular press. It is the reason the case never quietly went away, and the reason a laboratory in Paris was still working on it in 2020.

The son's declaration

The counter-attribution was specific from the start. The Times published an article in 1910 stating that the bust had been made by Richard Cockle Lucas, an English sculptor, working from a painting.

Lucas's son Albert then made a statutory declaration, on oath, that his father had made the Flora bust in 1846, that he had helped him, and that the wax had been built up in layers from candle ends. A watercolour of the bust painted by Lucas's son also survives.

Sworn testimony from a participant is strong evidence and it is not conclusive evidence. It was given by a man with an interest in his father's reputation, about work done more than sixty years earlier, in the middle of an international controversy. What can be said is that the account is specific, was made under legal penalty, and named a date, a method and a source image.

Gustav Pauli, the museum director in Hamburg, put the attribution to Lucas on a scholarly footing: a British nineteenth-century sculptor known to have made a large number of wax sculptures after ancient models.

What was inside the bust, and what that does and does not show

When the back of the bust was opened it was found to contain a wooden spar, newspaper and other materials of the nineteenth century. Albert Lucas had described stuffing of exactly that kind.

The 2021 study dated those materials. The wood gave 110 plus or minus 30 years before present, the textile 195 plus or minus 30, the paper 170 plus or minus 30, and the three combine to 155 plus or minus 20. Calibrated against the atmospheric curve they give 1682 to 1938, 1646 to 1950 and 1656 to 1950 respectively, combining to 1667 to 1950; the ranges are broad because the calibration curve is flat across that period. All three are definitively after 1650.

The authors then add the caveat that matters, and they add it themselves: it is possible that these objects were introduced later, when modifications were carried out. A filling is not a signature. If the packing inside the cavity had been the only evidence, the case would still be open, because packing is exactly the kind of thing a restorer replaces.

That is why the wax itself had to be dated, and the wax is where the problem lay.

Why a wax made of whale cannot be dated in the ordinary way

Radiocarbon dating assumes that the carbon in the material was in equilibrium with atmospheric carbon while the organism was alive. Wood, linen, paper and beeswax satisfy that assumption. A sperm whale does not.

Carbon dissolved in the ocean is older than carbon in the air, because it has been out of contact with the atmosphere for a long time. Anything that eats in the sea therefore carries an apparent radiocarbon age that is too great, an offset called the marine reservoir effect. The global average for surface water is around five hundred years; real values range from about four hundred years in subtropical oceans to over a thousand at the poles. A sperm whale travels across all of it, integrating the reservoir ages of every water mass it feeds in.

The consequence for this object is stark. Uncalibrated measurements on the Flora's wax come out between 340 and 395 years before present, which read naively would put the bust in the sixteenth or seventeenth century, which is to say in the Renaissance. Those numbers are a large part of why the case survived so long. They are also wrong, and they are wrong for a reason that has nothing to do with the sculpture.

No marine reservoir age had ever been published for sperm whale bone or for spermaceti. The nearest available values were for other cetaceans: an average of 370 plus or minus 30 years for various North Atlantic whales, from a study of twenty-one animals caught off Norway in the nineteenth century; about 200 years for Canadian Arctic bowheads; 320 plus or minus 35 for marine mammals living near Sweden; about 350 for a seventeenth-century finback whale from Spitsbergen; and 570 plus or minus 95 for bowheads in a compilation from the Canadian Arctic archipelago. The only estimate for spermaceti was Juergen Freundlich's, of 300 plus or minus 200 years.

Freundlich's date in the 1980s, and why one measurement was not enough

The first radiocarbon date on the bust was made by Juergen Freundlich in the 1980s. He took 3.2 grams of wax from below the left arm and obtained a conventional age of 290 plus or minus 40 years before present, laboratory code KN-3224. Applying a marine reservoir correction, he gave a calendar date within the nineteenth century.

That result excluded the Renaissance, and it is now understood to have been right. It did not close the case, and it is worth being clear about why not, because the reason is methodological rather than political.

Freundlich had one sample, from one place, and a reservoir correction he had estimated himself with an uncertainty of plus or minus two hundred years. Anyone who wished to could contest the correction, and contesting the correction changed the answer by centuries. The measurement was sound; the parameter that converted it into a date was a judgement. Assessments of the casting method pointing away from the Renaissance, and the historical documents, sat alongside it in the same unresolved condition.

2021: using a known object to calibrate an unknown one

The solution Ina Reiche, Lucile Beck and Ingrid Caffy published in April 2021 is the reason this record exists, because the method is more interesting than the verdict.

They took two series of samples: loose fragments and surface samples from the bust, then further samples from inside it, so that a claim that the surface was not original could not be used to dismiss the result, plus the wooden spar, the textile and the paper from the filling. They also sampled two wax reliefs by Richard Cockle Lucas whose dates are known: Leda and the Swan, made in 1850, in the Alte Nationalgalerie in Berlin, and Woman and Winged Woman, dated 1848.

Infrared spectroscopy and gas chromatography showed that the wax of the Flora bust and the wax of the Leda relief are almost identical: principally spermaceti of the sperm whale, with minor beeswax, in both. That similarity is the hinge. Because the Leda relief was made in 1850 and its composition matches, it can be used to determine the calibration itself.

They calibrated the Leda measurements with combinations of the atmospheric IntCal20 curve and the marine Marine20 curve until the resulting distribution sat correctly on either side of 1850, and the mixture that achieved it was fifteen per cent atmospheric to eighty-five per cent marine, with an uncertainty of ten per cent. Calibrated on the atmospheric curve alone, the Leda dates fall outside the year it is known to have been made, which independently confirms that marine carbon is present in the wax.

The same mixture applied to six wax samples from the surface and interior of the Flora bust puts every one of them after 1704, with a statistical combination given in the text as 1712 to 1950. The wax dates now agree with the wood, paper and textile dates, which the authors describe as confirming the strength and validity of the approach.

What the result proves, and what it explicitly does not

The finding is negative and the authors state it as a negative: all of the analysed constituents of the Flora bust are dated after 1700, precluding the bust from being created in the Renaissance period, and therefore precluding an attribution to Leonardo da Vinci.

What it does not do is name a maker or a decade. The authors say plainly that uncertainty in the calibration curves leads to a broad interval for the wax, with a precision of about two centuries. An object dated to somewhere between 1704 and 1950 is not thereby dated to 1846. The radiocarbon evidence is compatible with Albert Lucas's sworn account and does not confirm it; it is equally compatible with a wax made in 1780 by somebody nobody has ever named.

What narrows it is the chemistry rather than the dating: the wax of this bust and the wax of a documented Lucas relief of 1850 are, on infrared and chromatographic evidence, almost the same material. That is a strong argument, and it is an argument about a workshop's supplies rather than about a hand. Spermaceti was a commodity.

There is one small internal inconsistency in the paper worth flagging for anyone citing it. The body text gives the statistical combination of the six Flora wax dates as 1712 to 1950; the caption to the figure showing those dates gives 1707 to 1950. The difference is immaterial to the conclusion and it should not be quoted from memory.

The sentence about the people who got it wrong

The 2021 paper contains a judgement on a century of previous work that is unusual to find in a scientific journal, and it is the most useful thing in it.

Reviewing the analyses done in the 1900s and 1910s, which were of very high quality, the authors observe that the art-historical interpretation of the bust was likely influenced by the attribution that the researchers wished to obtain, an attribution to Leonardo.

That is not an accusation of dishonesty. The chemists in 1910 found spermaceti and reported it accurately. An expert who examined the surface observed cracks and read them as evidence of significant ageing. Others argued that the polychromy used Renaissance techniques, and that the face closely resembles Leonardo's figures, which it does. Every one of those observations is defensible on its own. What determined the conclusion was the order in which the questions were asked and which findings were treated as decisive, and that was set by what the institution had already paid for and published.

The Berlin museums have carried two objects of this kind, both bought on Bode's authority and both later given up. The other is the painting known as The Man with the Golden Helmet, acquired in 1897 as a Rembrandt over the objections of Cornelis Hofstede de Groot and Abraham Bredius, and reclassified as circle of Rembrandt in 1986.

Where it stands

The bust remains in Berlin, in the Skulpturensammlung und Museum fuer Byzantinische Kunst, under inventory number 5951, and it is shown in the Bode-Museum. The English-language summary of the museum's labelling gives England, nineteenth century, with a question mark. The German account of the case states that the authorship of the sculpture is still not entirely clarified.

That is the correct position, and it is worth defending against tidier versions. The 2021 study closed one question absolutely: this is not a Renaissance object and it is not by Leonardo. It left the other open. A museum that answered the first question and declined to invent an answer to the second has done its job.

The case is also the cleanest demonstration available of what radiocarbon is for. It excluded a period; it named nobody. Anyone reaching for a scientific test to confirm an attribution should read this paper and notice that its authors, having done everything correctly with three laboratories and a dozen samples, were able to say only that the thing is not five hundred years old.