Skip to content

Rosalba’s portrait of John Law

29 October 2020

No one who reads this blog needs to be told who Rosalba Carriera (1673–1757) or John Law (1671–1729) were, nor why their encounter in Paris in 1720, during the Régence, was of such significance – and why the lost portrait she made of him (J.21.0632  in the online Dictionary of pastellists, in this article) is mentioned in so many publications that I cannot possibly list them all.[1] Of course none of these discussions is of much help in locating the work beyond the idea, almost universally repeated, that Horace Walpole once owned it (the version he owned is J.21.06341 in the Dictionary). And the iconography of Law is so confused that it too provides little assistance.

Walpole’s version

As we shall see, the date of his acquisition is important – the reference in the first (1774) edition of his A description of the villa of Mr Horace Walpole (p. 66) providing a terminus ante quem:

John Law, inventor of the Missisipi-scheme [sic], and prime minister to the regent Philip duke of Orleans: one of the best of Rosalba’s works.

There are several further references in his correspondence: he mentions it in a letter of 7 November 1782 to the Scottish antiquary, the Earl of Buchan:[2]

If your Lordship should print any account of John Law the Missisippian, <and> wish to give a print of him, I have a portrait of him by Rosalba, the best I ever saw by her hand, and which must be extremely like, as it is the very image of this daughter Lady Wallingford now living. As the picture is in crayons and even let into the wainscot of my gallery, it cannot be taken down; the artist must therefore make the drawing from where it is.

He repeated this again in another letter to the earl, 12 May 1783, adding–

–an excellent head of [John Law] in crayons by Rosalba, the best of her portraits. It is certainly very like, for, were the flowing wig converted into a female head-dress, it would be the exact resemblance of Lady Wallingford, his daughter, whom I see frequently at the Duchess of Montrose’s, and who has by no means a look of the age to which she is arrived. Law was a very extraordinary man, but not at all an estimable one.

Buchan it seems was intending to write a biography of Law, no doubt as a famous Scot, but in the end produced only a short letter in The Bee in 1791, consisting just of Walpole’s anecdotes. As the correspondence indicates, it was contemplated that a print be taken from Walpole’s portrait, a project abandoned because the pastel could not be taken down – Walpole had fitted it into the wooden panelling, some 2½ to 3 metres high, was conscious of the risks of moving pastels, and had placed his famous Roman eagle in the same niche, preventing anyone placing a ladder there to get a closer view. This emerges again from a letter he wrote on 14 November 1792 to Richard Gough (1735–1809), another antiquary, who was presumably contemplating a different Law biography:

I have a portrait of Law, and should not object to letting a copy of it be taken, but I doubt that could not be done, being in crayons, by Rosalba, under a glass; and any shaking being very prejudicial to crayons, I fixed the picture in one of the niches of my gallery under a network of carving, whence it cannot possibly be removed without pulling the niche to pieces. The picture too being placed over the famous statue of the eagle, there is no getting near to it, I certainly could not venture to let a ladder be set against the statue. Indeed, as there are extant at least three prints of Law, there does not seem to be another wanting.

I am sorry, Sir, I cannot give you a more satisfactory answer about Lady Wallingford. I have met her at two or three places, but I did not visit her, nor have the least knowledge of her husband’s family, nor to whom she left anything she had; nor can I direct you at all where to inquire. I did not even know that there is an Earl of Banbury living.

Although there is a tiny glimpse of the pastel in the niche in the gallery of which an engraving was included in the second, 1784, edition, the image is completely indecipherable. We have a little more luck with the original watercolour of the Gallery at Strawberry Hill, made in 1781 by Thomas Sandby, Paul Sandby & Edward Edwards (V&A, inv. D.1837-1904).

Here is the detail visible in the niche to left of the chimney, straightened with Photoshop:

Although neither Walpole’s Description nor the 1842 sale catalogue provides dimensions for the pastel, and while it might seem impossible reliably to estimate the dimensions from this image, in fact the Roman eagle comes to our aid: its height is 77.5 cm, and being immediately underneath the pastel, we can estimate with some precision that the sight size of the portrait was 66 cm high (by perhaps 54 cm width, with less assurance). This is notably larger than most of the bust-length pastels Rosalba made in Paris at this time – they are typically 56×45 cm. This will certainly be a clue in identifying Walpole’s pastel should it resurface: it remained at Strawberry Hill until the 1842 sale, when it was sold[3] for 18 gns to “Brown, Esq., Pall Mall”, possibly General Sir John Brown, KCH; but all trace is then lost. It will presumably have been reframed in 1842 when removed from the wall.

The image alone does suffice to eliminate a number of wildly speculative suggestions from among Rosalba’s œuvre as candidates for the lost portrait (among them the pastel of Poleni which I discussed here). (It is curious that no one has proposed the Dresden pastel P84, my J.21.2189, which matches the composition most closely, although as we shall see it too is wrong.)


Inevitably researchers will turn to any information on Law’s appearance to conduct this search. They might start with the much earlier description circulated in the London Gazette (7 January 1694) when he was a fugitive after killing a man in a duel:

very tall, black, lean man, well shaped, about six foot high, large pock holes in his face, big high nosed.

The description was dismissed as useless by biographers (who improbably alleged[4] that he arranged for a false description to aid his escape), but the nose at least is supported by the iconography.

To which we must now turn, although again with limited confidence. A good summary (although with an important omission) is in Ingamells (Later Stuart portraits…), p. 144, accompanying the best known portrait, attributed to Alexis-Simon Belle (NPG 191), “identified as Law at least since the end of the eighteenth century, and the elegant costume suggests it was painted in France”. It graces the cover of one of the better biographies[5] (Montgomery Hyde’s), but that offers little extra assurance of the identification.

Ingamells then lists the enamel by Charles Boit in the Royal Collection, with conventional rose-bud mouth, straight nose and brown eyes which do not seem convincing.

Of it Graham Reynolds remarks (Sixteenth- and Seventeenth-century miniatures…, no. 398)[6] that “the traditional identification as Law appears justified by comparison with the engraving by Peter Schenk, 1720.”

Schenk’s engraving, although well known, is of limited iconographic value: in black and white, and unlikely to have been taken from life. It was published in Het Groote Tafereel der Dwaasheid in Amsterdam in 1720, and is one of numerous prints circulating at the time of the fame, then notoriety of the System. We can safely ignore many of these engravings. I have not yet tracked down the miniature[7] of John Law, in red coat and blue velvet waistcoat embroidered with gold. Another miniature by Coater at Knowsley listed by Ingamells is not of Law.[8] Ingamells lists too a painting attributed to Herman van der Mijn;[9] the label identifies the sitter and artist, but as the latter is given as Rigaud the name of the sitter should be treated with equal caution.

An oil by William Verelst[10] in an American museum shows a quite different face and cannot be right.

Possibly of more interest is the widely reproduced later (1843) painting by Casimir-Victor-Alexandre de Balthasar (MV 4372) which apparently (Ingamells) was copied from a portrait still in the sitter’s family in 1843, although the source he cites (Constans 1995) merely states that it was a copy of an anonymous painting (unlocated). Without the original it is difficult to assess how much licence was taken.

A greater loss is the Rigaud painting of c.1719–20 (James-Sarasin no. 1343), perhaps unfinished, and now known only from the engraving by Georg Friedrich Schmidt (1738).

Finally, a discovery which is omitted from all modern discussions but which is crucial for this essay, a print apparently made by Quenedey in Paris, commissioned for John Philip Wood’s biography of Law in The antient and modern state of the parish of Cramond, Edinburgh, 1794, reproduced (opposite page 163) with permission of the then owner, Jean Law de Lauriston (1719–1797), after “an original portrait of his uncle, reckoned an exact likeness, in his possession”:

A second engraving copying this was made by Edward Mitchell for Wood’s 1824 life of Law (presumably because Quenedey’s engraving would not last through another print run). Both are fairly wretched as works of art, but the function of portrait engravings is often documentary rather than aesthetic.

We can confirm Wood’s assertion that the pastel belonged to Law neveu as it appears in his estate inventory[11] conducted in Paris in 1802, several years after his death.

It is referred to in the usual formula (“pour mémoire”) applied to family portraits that were not valued. The wording implies that the pastel was in a gilt frame. What is also bizarre is that all three family portraits were kept together “dans un petit cabinet”, the only other item in which was an oak chest. This suggests that the financier was not regarded as a national hero during Napoleon’s reign, and had to be hidden away.

I’m not sure that much consistency emerges from these various images, apart (at least among the serious contenders) from a peculiarly aquiline nose, a cleft chin and a protruding lower lip. The shape and density (even if colour is not revealed) of the eyebrows (often a reliable feature in portraiture) are particularly variable. There certainly isn’t enough to identify Law from a portrait of an unknown sitter without documentation. But in the case of the new discovery below the exact match with the Quenedey engraving provides the assurance we need for a portrait that would otherwise have to be rejected as dissimilar to the Rigaud and Belle.

Law family tree

Incidentally, to follow this and the later discussion, you may want to have a short pedigree of the Law family: see here for a fuller version, but the key players are:

John Law of Lauriston (1671–1729), financier {Carriera} ∞ c.1701 Katherine Knollys (c.1669–1747) {Carriera}, dau. of the 3rd “Earl of Banbury”

⇒John Law (1706–1734), soldier in the Austrian dragoons {Carriera}

⇒Mary Katherine (1711–1790) {Carriera} ∞ her cousin, William Knollys, Viscount Wallingford (1694–1740sp), MP, son of the 4th “Earl of Banbury”

William Law (1675–1752), of the Compagnie des Indes ∞ 1716 Rebecca Dewes ( –1729)

⇒John Law de Lauriston (1719–1797), chevalier de Saint-Louis 1780, brigadier d’infanterie, commandant des troupes françaises dans l’Inde, gouverneur de Pondichéry 1764–77

⇒Jacques-François, chevalier de Law, comte de Tancarville (1724–1785)

Rosalba’s account

What then of the circumstances of the commission? All sources agree that Law was a very busy man in 1720, even more so than Rosalba’s other clients. But of course he was not just any other client. Rosalba Carriera was accompanied during her visit to Paris by her mother, two sisters, and brother-in-law, the painter Giovanni Antonio Pellegrini (1675–1741). And it was the latter who lost most from the encounter: Crozat, at whose invitation Rosalba’s visit had taken place, encouraged Pellegrini to undertake the decoration of the immense ceiling of the Hôtel de Nevers which was to be the assembly room for the banque du Mississippi – some 350 m2, more than half the size of Würzburg. Pellegrini’s allegorical fresco, commissioned for 10,000 ducats, was destroyed two years after its completion – perhaps as much because of French artists jealous of the Venetian’s success as of the creditors eager to erase all trace of the system.[12]

It is however worth picking out the entries in Rosalba’s journal that mention Law and his family. These from my translation (you can find my transcription of the Italian in my edition):

APRIL 1720

The portrait of … of Law’s son;[13]

JUNE 1720

11., Tuesday. Started the portrait of Law’s daughter.[14]

  1. I asked the little Miss Law to change a bill for me.


  1. I went to Versailles, and Mme Law sent me the frames.
  2. I was visited by the wife[15] of M. Law the younger….


  1. Went to M. Law and left the portraits there.

22., Sunday.[16] Went with Bononcini to M. Law.

23., Monday. … I started [the portrait] of Law.

  1. At a concert given by M. Crozat, I saw the Regent, Law, and others.


First November. A bad day. I saw M. Law at the Bank, and talked to him.

  1. At Mme Du Revest’s I retouched the portrait of M. Law’s son. At the moment he left his house, the gun of one of his guards went off by accident, and wounded a child in the thigh. A Frenchman, who knew me in Venice, came to the Bank to ask me for some miniatures.
  2. I arranged to … finish [the portrait] of M. Law.
  3. Went to lunch at Mme Law’s, and finished her husband’s portrait. Went to the Comédie, and refused to make copies of the portraits of this family.
  4. Devaluation of the coinage.


  1. Law’s daughter came, and I gave her her own portrait.
  2. Went to … and to Mme Law. Agreed to go to the Gobelins on the 12th.
  3. Saw Mlle Law, whose father was disgraced the same day.
  4. I went in vain to Mme Law, who had gone to the Opéra; whence I went to the Comédie with Mme Boit and my sisters.

15., Sunday. I went to Mme Law, whose husband had left the same day. She gave me 20 louis.

  1. … I returned the wig and cravat of M. Law.
  2. I wrote to M. Law’s daughter. I received twenty louis of 45 livres each.


  1. I got ten louis of 45 livres for the portrait of M. Law, which remain in the hands of my brother-in-law with 62 Spanish écus he holds on my account.

Several important things emerge from a close reading of these entries. First the relationship with the family was deeper than just that of portraitist; second that the sittings were few; third that they ended so abruptly that Rosalba had to borrow the wig and cravat to complete the work without the sitter. Indeed all this was happening as the System was crashing: Law was bankrupt, and had resigned in disgrace on 9 (not 11) December, proceeding a few days later to exile at his estate near Paris, Guermantes,[17] before leaving France for good. We don’t know if the payments she received were the full amount due (Rigaud had been paid in shares in the compagnie des Indes, but they were by then worthless), but she was astute enough to resolve in November “not to make copies of the portraits of this family” (“rifiutato di far duplicati li ritratti di detta famiglia”). It would seem then that the four Law portraits she made were unique. She might already have made replicas, and may even have kept sketches from which she subsequently worked up versions (and logically we cannot totally exclude the possibility that Law himself, who would die in Venice, sat to her again) – but we simply don’t know.


How then do we place the versions the Rosalba portrait?

The literature assumes that Walpole’s was the only version. But it is surely too big to be the portrait of “Mr John Law, from Life”, measuring a mere 24×19 cm, which Consul Joseph Smith, (inv. 1762, no. 19) sold to George III, but of which there is no later trace (where did Smith acquire it if not in Venice? But that could have been from artist or sitter). We don’t know if Smith’s description was accurate or if so how it fits into the narrative above.

Further, as we have seen, Walpole had his version by 1774, while the portrait Quenedey copied was with the nephew Jean Law from before 1794 and remained in Paris at least until his 1802 inventory. For the same reason, Walpole’s cannot be the portrait sold by the Law family at Christie’s in 1782. So despite his letter to Lord Buchan revealing that he knew Lady Wallingford (which must be tempered by his admission to Gough that he didn’t know her well), there is nothing to suggest that he acquired his pastel from the family. We should also perhaps interpret without today’s punctiliousness Walpole’s assertion that the picture was Rosalba’s finest: it is not entirely impossible that he would have thus referred to a copy he commissioned of a picture he thought her finest.

To understand the various routes in which Law’s pictures travelled after his death, turn to JoLynn Edwards’s 2001 study. Unfortunately, despite intensive research, the answers are far from clear. The pictures Law had with him at his death[18] were sent to Holland by boat but were damaged by water and had to be returned to Venice to be restored. It appears for example that the two groups of pictures sold as from Law at Christie’s in 1765 and 1782 (the latter including Rosalba pastels of Law and his son, the former, Lot 47, “a highly finished portrait of the celebrated Monsieur Laws, one of the best of the charming artist”, sold for 8½ gns to Wilde) may have been sold either by Law’s daughter Lady Wallingford, or perhaps on her behalf by George Middleton, a London banker, who had written to Law in 1728 explaining the difficulty of selling the pictures of mediocre quality Law had sent him.

1782 sale

We cannot exclude the possibility that the pastel purchased by “Wilde” was resold to the nephew, although this seems improbable. Lot 48, of the son, was purchased by Walton, a dealer who bought extensively from this and other sales at the time; but Wilde bought only Lot 47. His identity is uncertain; but the name next appears in London sale records as purchaser in 1799, when it is given as De Wilde – almost certainly the portraitist Samuel De Wilde (1751–1832).

Added complications arise from Law’s will, later replaced by a lifetime donation just before his death to his “wife”, Katherine Knollys, to whom he was never in fact married. His brother William was also overlooked as he was a Protestant, so the direct heirs of his estate were his two nephews Jean Law de Lauriston (1719–1797), noted above, and Jacques-François, chevalier de Law, comte de Tancarville (1724–1785).

A pastel which had belonged to the horticulturist Ellen Ann Willmott appeared on the London market about ten years ago (right: now J.21.06343 in the online Dictionary). It was shown to me as of an unknown sitter with an attribution to Lundberg; I thought it much closer to Rosalba in composition, but the technique was not hers (Dr Sani agreed).[19] And although the composition resembled the Sandby detail (and promisingly it had a modern frame), it was a little too small, and the eyes appeared to look in a different direction; we concluded that there was not enough to make the connection securely. This was of course before the discovery of the Quenedey print.

All of this was overturned when a private collector recently showed me an image of what is quite clearly a Rosalba pastel (now J.21.06325):

Of dimensions (60×45 cm) far closer to Rosalba’s normal size than Walpole’s niche, its French frame bears a later label “Rosalba/Pt de John Law” suggesting that it spent part of its life in France. Moreover the Willmott pastel is plainly copied from it. Further a comparison of both with the Quenedey engraving reveals one important difference: the left edge of the jabot where the lace bulges out allows see-through in the new pastel not present in the Willmott picture, but captured in the Quenedey print. For those reasons I concluded that this may well be the pastel Jean Law owned in 1794. This is endorsed by a provenance which shows that the pastel has remained among the descendants of Jean Law de Lauriston.


So what do we conclude having finally discovered this elusive image? Perhaps the most surprising thing is how restrained it is. Compare for example the fanciful Schenk portrait, which has been described in a recent article[20] as–

he stands in courtly dress in front of a well-manicured, formal French lawn. His dignified attire matches that of the orderly garden; his clothes are festooned with gold brocade while the garden is adorned with stately fountains, acanthian-scrolled parterres, and topiary trees. Both environment and man are contemporarily modish, consistent with the imagery of current fashion plates which showed courtiers, resplendent in silk and lace, posing in Le Nôtre-styled lawns &c.

Secondly, even by Rosalba’s standards, the tonality is subdued, in line with the costume. Law had no chivalric order – no Saint-Esprit, no Garter: and one wonders if he made a virtue of this by dressing in the plainest manner possible, as in later years Franklin did. Did he feel that such austerity would inspire confidence in his investors, perhaps by demonstrating that he had no need of ostentation to flaunt his wealth? Or was he simply too busy to involve himself in the selection of elaborate costume, perhaps fearing it would merely extend the length of the sittings?

Walpole’s pastel is still missing – almost certainly larger than the Willmott copy, but still presumably in saleable condition after being ripped from the walls of his gallery: but perhaps Sir John Brown found it had suffered more than he realised at the sale; it may be lost forever. That of course seems to have been the fate of the other three Rosalba portraits of Law’s family.

Finally the iconography can be greatly slimmed down. He was quite clearly blond and blue-eyed, not black as in the Gazette description. The Belle portrait remains credible, particularly if made six or seven years before the Rosalba; the engraving of the Rigaud has never been in doubt; and of the other prints, only the Quenedey is close to the real face of this fascinating character.


[1] You can even find it in an article by Trollope’s sister-in-law. We can rely on Arsène Houssaye to produce extravagant versions, such as his “Figures de la Régence”, Revue du xixe siècle, vii, 1.xii.1867, pp. 327ff, which begins promisingly: “A Venise, j’ai découvert un portrait de Jean Law, un pastel de cette Rosalba…” but fails to deliver. Remember that to search French sources, you need to know that his name was pronounced Lass (as in l’As), and Voltaire spells it thus.

[2] The W. S. Lewis edition of Walpole’s correspondence is now available online; the relevant letters are in volume 15, pp. 167, 180f, 192 and volume 42, pp. 386f.

[3] The price of 15 gns appears in the secondary literature, but 18 gns is clearly marked in the annotated sale catalogue; 15 gns is the price of the next lot (Rosalba’s portrait of Lord Hertford); the Liotard of Lord Holland reached on 4 ½ gns. But the Reynolds of Lord Waldegrave sold for 70 gns. Such were the tastes of the time.

[4] See, for example, Notes & queries, 2.iv.1864, p. 284f here.

[5] For the general reader; economists will want to consult the various writings of Antoin Murphy (e.g. this).

[6] The following entry (no. 399), for Boit’s enamel of the young Louis XV, suggests that Lundberg may have been the source when in fact the enamel copies Rosalba’s pastel (J.21.0697).

[7] On ivory, 7.5×5 cm (Phillips; 10 June 1865. Henry George Bohn, cat. 1884, no. 386; London, Christie’s, 19 March 1885).

[8] I am most grateful to Stephen Lloyd for providing me with an image; Ingamells evidently hadn’t seen one.

[9] Three-quarter length, standing, in a brown coat with a richly decorated blue silk waistcoat, right hand on hip; Woolton House sale, 6–7 October 1993, Lot 584.

[10] Albrecht Kemper Museum; sd 1727; sold Christie’s 16.xii.1966, Lot 291, ex Sir H. Steward but justly disregarded by Ingamells.

[11] MC/ET/LXVIII/699, conducted 5 July 1802.

[12] Pierre Rosenberg, De Raphaël à la Révolution, 2005, pp. 14, 121ff.

[13] John Law Jr (1706–1734), son of the celebrated financier John Law of Lauriston (1671–1729). He became a soldier in the Austrian dragoons.

[14] Mary Katherine Law (1711–1790), later Lady Wallingford. The suggestion that this is the girl with a monkey in the Louvre J.21.0575 is widely found in the literature, but without foundation (she was in fact the future marquise d’Havrincourt, née Antoinette-Barbonne-Thérèse Languet de Gergy (1717–1780)).

[15] Rebecca Dewes ( –1729), wife of John Law’s younger brother, William Law (1675–1752).

[16] The numbering follows Vianelli and Sensier. Sani follows the manuscript and has 23 September a Sunday, which it was not. All sources synchronise again on 1 October, a Tuesday.

[17] I don’t think this is what Proust meant by du côté de chez Guermantes.

[18] There were 488 pictures listed in the inventory (a transcription is in the Getty Provenance Index); only two pastels, both heads of saints by Guido Reni, were included.

[19] I should note that Silvia Davoli spotted in a later sale as resembling the Sandby image.

[20] Camille Mathieu, “An effortless empire: John Law and the Imagery of French Louisiana, 1683–1735,” Journal18, Issue 10, 1720, 2020,

From → Art history

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: