The story of Salmacis and Hermaphroditus is taken from a source on which so many renaissance and post-renaissance artists drew, Ovid's Metamorphoses (IV, v. 271 ff.). The nymph Salmacis fell in love with Hermaphroditus, the son of Venus and Mercury. While bathing in her river he rejected her advances: but in answer to Salmacis's prayer, the gods fused their bodies.
This remarkable picture, long in store at Knole, only came to notice as a result of the comprehensive photographic survey of the collection there undertaken by the Courtauld Institute in 1958-1960. It was given to a follower of Domenichino, but, in 1990, Aidan Weston-Lewis recognised that the picture is in fact Ludovico Carracci's prototype of a composition previously known from a picture in the Rospigliosi collection in the Galleria Pallavicini, Rome (F. Zeri, La galleria Pallavicini, Rome, 1959, pp. 23-30, no. 11, as Sisto Badalocchio). His attribution has now been endorsed by other scholars in the field, including Professore Alessandro Brogi, who has inspected the picture in the original, and Dr. Daniele Benati, Dr. Babette Bohn and Professor Gail Feigenbaum, who have studied this from transparencies.
The obvious pentiments, particularly that in Salmacis's hand, confirms the priority of the picture. It ranks indeed as one of the most poetical of Ludovico's secular works, eloquent of the pictorial revolution at Bologna to which Ludovico and his cousins, Agostino and Annibale, contributed in equal measure. Not surprisingly, Ludovico turned, in particular, to Titian, whose Ferrara Bacchanals he must have studied with close attention before their removal to Rome in 1598. That a picture of corresponding description was in a collection in Ferrara in 1632 (see below) may well mean that the patron for whom that picture was intended was conscious of this debt. Professors Brogi and Feigenbaum, among others, propose a date just before 1600, while Dr. Benati places the picture circa 1600-2. Aidan Weston-Lewis considers the picture to be of 1592-5, and compares this with the Bacchus and Ariadne at Vercelli (Civico Museo Francesco Bergogna, A. Brogi, Ludovico Carracci, Bologna, 2001, I, no. A1, II, fig. 251) which he believes to be of that date.
Giovanni Gaetano Bottari (Raccolta di Lettere sulla Pittura, Scultura e Architettura, Rome, 1757-1768, VII, p. 26) records that the poet Giovanni Battista Marini (1569-1625) suggested that Ludovico paint for him a 'fantasia oscena e lasciva', showing Salmacis and Hermaphroditus 'ignudi et abbraccianti in mezzo della fontana'. Verses in Marini's La Galeria ... distinta in pittura e sculture, Venice, 1622, p. 12, quoted by Carlo Cesare Malvasia, Felsina pittrice, Bologna, 1678, p. 325, imply that Ludovico did indeed supply the poet with such a picture. No such work survives, but a picture of this subject by Albano at Turin, Galleria Sabauda, no. 498 (Puglisi, op. cit., no. 120) may reflect this, or a drawing for it.
In the Knole picture Ludovico decided to represent the earlier phase of the narrative, in which Salmacis falls in love at the sight of Hermaphroditus and prepares to plunge into the waters to seduce him. The subject clearly enjoyed a wide intellectual vogue and the poet Girolamo Preti's La Salmace was published in 1618, to be translated into English by Sir Edward Sherburne with Marini's Lydia in 1651. A canvas of the subject is recorded in the 1632 inventory of the collection of Roberto Canonici at Ferrara (cf. R. Wittkower, The Drawings of the Carracci in the Collection of Her Majesty the Queen at Windsor Castle, London, 1952, p. 109, under no. 88): this description could very well refer to the Knole picture:
La Salmaza di Ludovico Carazza, che sto mirando un gioveneto nudo, apresso lui gl' è un cane, che dorme, tra di loro corre un'aqua, e sono in una gran boscaglia (G. Campori, Raccolta di Cataloghi ed Inventari, inediti, ..., Modena, 1870, p. 114, quoted by Zeri, op. cit., p. 30).
According to Zeri, the Canonici collection was sold after 1638. Malvasia in 1678 (op. cit., I, pp. 334 and 354) recorded a Salmacis and Hermaphroditus in the Landini collection at Bologna: this was still at Casa Landini in 1769 and may, as has been suggested, correspond with that in the posthumous sale of Robert Adair, Christie's, 21 May 1790, lot 46 (see infra).
The popularity of the design is attested by two drawings, Florence, Horne Foundation, no. 5582 and Windsor, Royal Collection (Wittkower, op. cit., no. 88), the first corresponding closely with the Knole picture, the last of upright format: both are attributed by Brogi to Ludovico's pupil Francesco Brizio, to whom he also attributes the Pallavicini picture (in Studi di storia dell'arte, 4, 1993, pp. 85). A near replica of the Windsor drawing was owned by Charles Rogers, and engraved as by Agostino Carracci for his A Collection of Prints in Imitation of Drawings, II, London, 1778, p. 33.
As Carlo Volpe recognised in 1962, Albano's Salmacis and Hermaphroditus in the Louvre derives from the Ludovico composition, which he knew from the Horne drawing (see the exhibition catalogue, L'Ideale Classico, Bologna, 1962, under no. 39). Albano stretches the composition so that the figures are shown in a wider landscape, and copies the Salmacis relatively faithfully, although his Hermaphroditus is different in design. The Louvre Salmacis, one of a series given by André le Nôtre to King Louis XIV in 1693, is described by Puglisi (loc. cit.) as 'perhaps the most lyrical of Albani's cabinet pictures'. This no doubt explains why there are so many early derivations of it (Puglisi, nos. 59.V a-g and LV a-b). Albano's Turin Salmacis and Hermaphroditus (supra) was also much copied (Puglisi, op. cit, nos. 102 V a-f and LV a-e).
When the picture reached Knole cannot be established, although it is possible that it formed part of the inheritance of Frances, daughter of Lionel Cranfield, 1st Earl of Middlesex and wife of Robert Sackville, 5th Earl of Dorset (d. 1667), on the death of her brother Lionel, 3rd Earl of Middlesex in 1674. The frame is of a characteristic late seventeenth-century type that was used for a number of pictures in the house, presumably during the forty-year tenure of Charles, 6th Earl of Dorset (1636-1706). This makes it very probable that the picture corresponds with one of two identified as of Cupid and Psyche in the Knole inventory of 1706 (in which pictures are only recorded by subject), as no alternative candidate is identifable in subsequent schedules or survives at Knole, and it is quite understandable that a relatively obscure subject was not recognised. One of the two works, 'Cupid and Sykey over the Chimney', was in 'Old Lady Dorsets Chamber', the bedroom presumably of the widow of the 5th Earl who herself died in 1687, rather than her son's third and last wife Anne, Mrs Roche (described by Wraxall as 'a woman of very obscure connections'), who had died in August 1706: while the second, listed as 'Cupid and Cykey', was one of 'three large pictures' in the 'Piaza' ('An Inventory of the Goods of the Right Honorable the Earl of Dorset taken the 24th day of October 1706, by us whose Names are underwritten' [John Holloway and Richard Cornfield], locations T. and A.B., Sackville Mss.).
Neither picture is identifiable in the 1730 inventory ('An Inventory of Goods at Knole/Taken in Nov.r. 1730.', Sackville Mss.), in which references to pictures are telegraphic, nor in the more informative one of 1765 ('An Inventory of the Goods at Knole/taken in October 1765, Kent County Record Office). The latter inventory was evidently taken following the death of Charles Sackville, 1st Duke of Dorset on 10 October. His heir was his eponymous elder son, the 2nd Duke (1711-1769), who was in turn succeeded by his nephew, John Frederick, 3rd Duke of Dorset (1745-1799), the last major picture collector of the family.
The earliest full list of the pictures at Knole was compiled on the 3rd Duke's death in 1799 ('An inventory of all the Pictures, Statues, Busts, Household Goods and Furniture -- At Knole in the County of Kent, Directed by the last Will and Testament of the Late Duke of Dorset, to be considered as Heir Looms, Taken the 12th & c. of August 1799'). With the exception of the Gainsborough's celebrated portrait of his erstwhile mistress La Baccelli (London, Tate Gallery), almost every major picture in the collection is clearly identified. The Gainsborough had presumably been put away after the Duke's marriage in 1790 to Arabella Cope; and it may be that the subject matter of the present picture was also not to her taste, for it too is not mentioned: however it should be noted that a number of otherwise unidentifiable pictures are listed merely as 'Landscape and Figures'; and it would be understandable if the compiler of the inventory had been at a loss at to the subject. The picture is not identifiable in the Knole inventory of 1828.
No old master pictures would appear to have been added to the collection at Knole after the 3rd Duke's death and so the fact that there is no mention of such a picture in the rather detailed schedule of his personal acquisitions probably excludes the possibility that the picture is that of the subject included in the sale of the Surgeon-Captain Robert Adair at James Christie's rooms, 21 May 1790, lot 46 (as Lud. Carracci, Salmacis and Hermaphroditus, 8 guineas, to Jones who also bought the companion Diana and Endymion, lot 47, for 9½ guineas), the relatively modest price of which suggests that it may have been of smaller scale.
We are indebted to Professore Alessandro Brogi, Professore Daniele Benati, Dr. Babette Bohn, Professor Gail Feigenbaum and Aidan Weston-Lewis for help in compiling this note.