The Picture Her Father Refused

The refusal is one sentence in one memoir

The reason La Princesse du pays de la porcelaine was available to be bought by a Liverpool shipowner, and so to end up above a fireplace in Washington, is that the sitter's father did not want it. The passage everything rests on was published by Elizabeth Robins Pennell and Joseph Pennell in 1908, in the first volume of their life of Whistler, and it reads: the two girls wanted their father to buy it, but Mr Spartali did not like it, he objected to it as a portrait of his daughter.

That is the whole of the documentary basis. It was printed forty-four years after the sittings, and the Pennells are transmitting the recollection of Marie Spartali Stillman, the sitter's elder sister, who had sat in the room throughout. The University of Glasgow's catalogue raisonne quotes the same sentence, and the scholarly literature on the painting cites the Pennells and the standard monograph on the Peacock Room, which cites the Pennells in turn. A single second-hand memory is carrying a very large building.

It is worth being exact about the objection as reported, because it is not a judgement about painting. He objected to it as a portrait of his daughter. Whatever else the picture is, it was offered to a family as a likeness and refused as one.

No price was ever named for the offer to her father

The figure that circulates in connection with this refusal belongs to a different transaction. After the family declined, Rossetti took the canvas and sold it, and it is that sale which produced the hundred pounds. The Pennells say Whistler agreed to take a hundred pounds and that Rossetti placed the canvas in his own studio. Whistler's own recollection, in a draft letter of 1892, is less firm than his biographers: he writes that the picture was bought at the instigation of Rossetti and that he received for it one hundred pounds, and then corrects himself, or fifty, I forget which.

No source consulted names any sum offered to Michael Spartali. Our own record for the painting stated that the Pennells reported the 1892 result as four times the sum Whistler had originally been paid, and that the original sum is not given in the sources consulted, so the multiple cannot be converted. That was wrong, and it has been corrected: the Pennells give the hundred pounds at volume 1 page 124. The report of the multiple that our record was citing is at volume 2 page 127, and the catalogue raisonne's own note misplaces it in volume 1, which is probably how the two became detached. The arithmetic also favours the higher figure, since 420 guineas is 441 pounds, which is nearer four times a hundred than four times fifty.

One conflict in that table is worth flagging rather than smoothing. The Pennells give the Freer purchase as five thousand guineas, against the 3,750 pounds in the catalogue raisonne. They are not reconcilable and this article does not settle them.

Refusing pictures had happened before, and that part is primary

The strongest independent evidence about Michael Spartali as a buyer is not about his daughter's portrait at all, and it sits in letters written at the time rather than in a memoir written in 1908. In 1866 he commissioned four still lifes from Henri Fantin-Latour, accepted and paid for two and declined the other two. The catalogue's editors attribute that rejection to Alphonse Legros persuading him the four pictures were all alike, so it is not a clean case of a buyer simply changing his mind. Fantin was left unpaid until another buyer took the rejected pair, and he pressed Whistler about the money in February 1867. Then, in a letter of 16 November 1868 about a different unpaid buyer, he asks Whistler a question that settles the character of the earlier episode: is it like Spartali, a collector who does not like them.

What matters is the second letter rather than the first: whatever the reason in 1866, by 1868 a contemporary painter was using this man's name, in his own hand, as shorthand for a buyer who does not like what he ordered. It does not prove the portrait story, which remains single-rooted, but it makes it a normal event rather than an inexplicable one, and nobody writing about the Peacock Room appears to have set the two beside each other.

One correction to our own record while we are on the father. He was a merchant when his daughter sat, a partner in the London Greek firm Spartali and Company, which had been Spartali and Lascaridi until 1854, and he became Greek consul-general in London only in 1867, after the picture was finished. Our record introduced him with the consulship, which reads as though the sitter were a diplomat's daughter at the time of the sittings.

What the only witness says about the sittings

Everything known about what it was like to be painted as the Princess comes from Marie Spartali Stillman through the Pennells, and the detail is domestic and physical. The sittings ran twice a week through the winter of 1863 to 1864, with Marie always accompanying her sister. The Japanese robe, the rug and the screen were ready on the first day and the pose was given then. Whistler partly closed the shutters to keep out the direct light. Christine stood at one end of the room and Whistler would look at the picture from a distance, then suddenly dash at it, give one stroke, then dash away again. They arrived about half past ten or a quarter to eleven and it was often long after two before they lunched. Raw tomatoes were a surprise to the two Greek girls, who had never eaten them except overcooked, and the canned apricots and cream they had never eaten at all; the roast pheasant belongs to one particular menu rather than to the standing lunch. It was four and later before they were released. He kept her standing while he worked on the head and never let her rest.

The costume was not hers. The Smithsonian's own label describes the series as putting western models in Asian dress surrounded by Chinese and Japanese objects from Whistler's own collections, and the screen behind her was his, sold at his bankruptcy in 1879. No source consulted says the robe belonged to the Spartalis, and Marie's account says only that it was ready.

The one contemporary document that touches the sittings is Whistler's, not hers, and it records that she fell ill. Writing to Fantin-Latour from Lindsey Row early in 1865 he says the Japanese woman is taking him from morning to evening, and that she had been put aside until now because of the model's illness. The catalogue raisonne raises the possibility that a professional model stood in for part of the work and hedges it, and this article leaves that hedge alone.

What became of her, which our record gave in six words

Christine Spartali married Comte Edouard Cahen d'Anvers, of the Paris banking family, in 1868 or 1869, and the forename is worth pausing on because the literature has it wrong. The Pennells called him Edmond in 1908 and the Glasgow project still carries Edmond on the same page where its own header prints Edouard. The man who existed was Edouard Joseph Cahen d'Anvers, 1832 to 1894.

She had two sons, Rodolfo in 1869 and Ugo in 1874. She was living in Rome. She died on 27 September 1884 in the Tyrol, at thirty-eight, and she was buried on 9 October in the Hellenic enclosure at West Norwood Cemetery in London. Her husband was created hereditary Marquis of Torre Alfina by decree of the King of Italy on 8 March 1885, five months after her death, so the marquisate some sources give her was never hers. The genealogical compilation we use elsewhere puts that grant in 1881, which would place it in her lifetime; the 1885 decree is the better attested of the two and the compilation is wrong here, as it is on the husband's forename.

How she died is not settled and this article will not settle it. The entry built on Alice Lege's research describes a severe depression and says she did not survive the treatments of the era, naming morphine and chloral hydrate among them. A genealogical compilation gives the Tyrol and 27 September while offering 22 September as an alternative and asking whether it was Germany, and separately raises both suicide and an overdose of chloral. A blog post by Kirsty Stonell Walker, reporting David Elliott's biography of the Stillmans, describes an overdose of chloral, accidental or otherwise, discreetly covered up, adds that most sources call the cause unexplained, that William Michael Rossetti called it cataleptical and that a rumour of death in childbirth circulated. That last strand reaches us at two removes, through a blog reporting a book we have not read, and it is named here as such rather than laundered into the record. Chloral appears in three of these accounts and none of them is a document. Our record ended her life in six words, that she went on to marry and died in 1884, and it now carries this instead, with the disagreements named.

She was named as the sitter in her own lifetime and just after it, by three different hands. E. W. Godwin, annotating in the room itself on 16 February 1877, labelled the overmantel picture a portrait of M. Spartali in Japanese dress. Ford Madox Brown, writing in 1888, listed among Whistler's best works a portrait of the Countess de Cahen in Japanese costume, using her married name four years after her death. Whistler told his dealer in February 1892 that there was only one picture of his that he cared for in the Leyland collection, the portrait of Miss Spartali in Japanese dress. And then the naming stops, and the room takes over. The archive holds the opposite case in the Klimt at Piacenza, a portrait whose sitter has never been named at all and where there is no father, no refusal and no marriage to follow, only a face. Between the two, being named and then forgotten is not obviously the better outcome.

Four things our own record had wrong

This article was written against our record for the painting and the research found four defects in it, all corrected in the same push. The fourth, her father's consulship, is treated above. The first is the one that made the record contradict itself. It stated that the picture is signed and dated in the lower area of the canvas, attributing that to the catalogue raisonne, while its own narrative elsewhere said a buyer was put off by the size and prominence of Whistler's signature at the top of the canvas. The Smithsonian records the signature at the upper left corner. The catalogue raisonne gives no position at all, so the attribution was also wrong.

The second is the hundred pounds, treated above: the record said the original sum is not given in the sources consulted, and the Pennells give it at volume 1 page 124, one volume before the page the record was citing for the multiple.

The third is a claim about her that no source supports as phrased. Our record said that Christine was, by contemporary account, one of the beauties artists of the day competed to paint. What the sources actually say is that Whistler, like Rossetti, was struck by the beauty of both sisters and asked the younger to sit. Tsui writes that she was noted for her beauty in Pre-Raphaelite circles, which is the germ of it, but the competing is ours. The exclamation about a Spartali sister that gets quoted in this connection is Swinburne's, and every source consulted attaches it to Marie. Christine's documented sittings amount to this painting, a lost pencil drawing and two photographs, one of them by Julia Margaret Cameron in March 1868. That is not a woman artists competed over, and the record should not have said so.