Watteau's painting of The Fortune Teller ("La diseuse de bonne aventure") has had a curious history. The original was engraved, in reverse, by Laurent Cars for the Recueil Jullienne, and its publication was announced in December 1727 in the Mercure de France. A caption on the print specifies that it reproduces the painting's actual size (approximately 339 x 274 mm.), and that the original came from the "Cabinet de Mr. Oppenort"; Mariette further clarified that this referred to the "Architecte de M. le Duc d'Orléans," or Gilles-Marie Oppenordt (1672-1742), architect, decorator and surveyor of buildings for the Regent. Oppenordt was a leading early decorator in the rococo style and a friend of Watteau: on a visit to Watteau's studio in 1715, Count Tessin reported examining two sketchbooks by Oppenordt that Watteau owned, and two early paintings by Watteau -- including The Fortune Teller -- were inventoried in Oppenordt's estate in 1742; presumably, the two artists had exchanged works. They probably knew each other through their mutual patron, Pierre Crozat, the rich financier and art collector for whom Watteau painted a series of Four Seasons (only Summer survives in the National Gallery of Art, Washington) and for whom Oppenordt undertook various architectural projects; Watteau lived in Crozat's house at several points in his life, as did Oppenordt, whose inventaire après décès was drawn up in Crozat's townhouse on the rue de Richelieu.
The Fortune Teller proved to be one of the artist's most popular youthful works, and many references to various versions of the subject (most of these presumably copies) appear in 18th and 19th-century auction catalogues (fifteen are listed by Rosenberg and Stewart, op. cit. including one for the present painting). In addition to Cars' engraving, at least four other contemporary prints of the composition are known, three of them made for the English market. Various paintings over the years have vied for recognition as Watteau's original, most of them of such poor quality that they can be easily dismissed. Recently -- and especially since its inclusion in the great Watteau retrospective in Paris, Washington and Berlin in 1984-1985 -- a fine version of the composition on panel in The Fine Arts Museums of San Francisco (inv. 64.4) has been widely, if somewhat tepidly, endorsed as the autograph prototype. In the catalogue to the 1984-1985 exhibition, Rosenberg wrote that "despite the scant enthusiasm of those who have seen it, the San Francisco painting seems to us to be from the hand of Watteau. (In any event, we believe that of all the known versions, this is the only one that could be the original.)" Rosenberg's analysis of the painting is both measured and convincing: he notes that "the trees and the vine in the middleground and the cross-bred spaniel are done with a sure touch. But above all the little dark lines marking the face and the hands of the gypsy woman as well as the child who holds the tambourine, setting off the eyes and the mouth, seem to be characteristic of the artist. Also typical are the way of marking folds and bows and the play of light in the silks and changing colors of the satins." He acknowledges certain peculiarities in the painting however, in particular "the tiny heads and hands, the clashing colors and the slightly stiff composition" which he attributes to the young artist's inexperience.
A troubling fact for Rosenberg and others who have studied the San Francisco painting, however, is to be found in the 1742 inventory of Oppenordt's effects. This document cites a painting depicting a "soothsayer of good fortune," described as on "panel, a copy after Vatteau, partially damaged" and estimated at 24 livres, a very low sum (M. Rambaud, AN, Documents du Minitier Central concernant l'histoire de l'art 1700-1750, Paris, 1971, II, p. 907). As we know, the original of The Fortune Teller was engraved in 1727 as in the collection of Oppenordt, and yet here, fifteen years later, a painting of the same subject in that same collection is being described as a copy and minimally valued. There are, as Rosenberg and Stewart noted, several possible ways of interpreting this document. Oppenordt might have sold the original and kept a copy. In the opinion of the present author, this would have been most unlikely. It is very possible, on the other hand, that the agent who drew up the inventory was unable to distinguish between an original painting by Watteau and a copy, and that the painting inventoried as a copy after Oppenordt's death was, in fact, Watteau's original, autograph panel. A fact that cannot be easily accounted for is the description of the painting in the inventaire as "partially damaged"; as Rosenberg accurately observes, "the San Francisco panel seems to be in relatively good condition."
The reemergence of the present version of The Fortune Teller might hold the key to solving this mystery. It was in the estate of Henry G. Bohn and sold with the rest his collection (including six other paintings attributed to Watteau) at Christie's, London, 19 March 1885, lot 63, as both Christie's archives and an old label on the reverse of the panel confirm. It was acquired for 6 gns. by the agent Philpot for Henry Martyn Kennard, the great-great-grandfather of the present owners. Kennard was a Member of Parliament who had made a fortune in railroads, and thus was able to pursue his interests as a collector of principally Egyptian antiquities (many subsequently bequeathed to the British Museum). As is apparent by comparison to either the San Francisco version of the composition or the engraving, the Kennard panel has been truncated on three sides, and is missing entirely the dog on the far right and much of the foliage at the top. In fact, the loss is greater, as x-rays make clear: the original panel ends where the elegant subject of the palm reading touches hands with the gnarled soothsayer; the barefooted boy and the gypsy fortune teller were painted by another hand on a narrow strip of wood that was added on at a later date. Considerable repainting is evident all along this join, no doubt applied to help integrate the addition.
What remains of the original picture is very dirty and disfigured by extensive, old repainting; any conclusive assessment of its authorship will have to await a careful program of cleaning and restoration. However, the small areas that are clearly legible and relatively free of later interventions display the characteristics of Watteau's hand: in particular, the hesitantly curious, gently anticipatory expression of the fortune teller's beautiful client is painted with a refinement and nuance entirely consistent with Watteau's idiosyncratic genius; it is a touch that is rarely -- if ever -- reproduced by even the most sophisticated copyist. The delicate but confident execution of the faces and fabrics and subtlety of the painting's russet palette seem to the present author more convincing than the pretty but vapid faces and cold, metallic coloring in the San Francisco version. Might the present painting be Watteau's original from Oppenordt's collection? And might it be the picture that between 1727, when it was engraved intact, and 1742, when it was inventoried after Oppenordt's death, suffered the 'partial damage' that caused it to be cropped and then misidentified as a copy?