This deeply moving painting has, on the few occasions when it has surfaced, been recognised as a variant of Titian's celebrated composition of the subject, most familiar from the two pictures in the Prado. Although due allowance has to be made for the condition of the Virgin's mantle, the face and hands, with the veil and the sleeve of the Virgin's dress, are of commanding quality. The evidence of x-rays now establishes the position of this picture within the sequence of Titian's treatment of the subject. That the distinction of the design was widely recognised is testified by the survival of at least ten early copies of this rather than either of the Prado pictures.
The theme of the Mater Dolorosa, the Virgin in Mourning, was of northern origin. Among the earliest examples of half-length format is that at Antwerp by Massys, which forms the left half of a diptych with a frontal representation of the Redeemed Christ, and the workshop version in reverse in the National Gallery, London, which is the right wing of a diptych, and of which a number of variants are known: the half-length London composition is comparable in format with Titian's. The popularity of paintings of this type in Spain, as Miguel Falomir illustrates in his discussion of Jan Massys in the Prado (M. Falomir, Tiziano, catalogue of the exhibition, Madrid, Museo del Prado, 2003, p. 226, fig. 125), leaves little doubt that the impetus for Titian's compositions came from the Emperor Charles V and subsequently from King Philip II himself.
The two pictures in the Prado have long been recognised as among the most poignant of all Titian's creations; Crowe and Cavalcaselle observe that 'in none of his single figures has Titian ever shown more genuine feeling' than in the Mater Dolorosa with clasped hands (J.A. Crowe and G.B. Cavalcaselle, Titian: His Life and Times, London, 1877, 2, p. 270); while Wethey, not often given to hyperbole, writes fairly of the Mater Dolorosa with raised hands: 'The extraordinary pictorial beauty of this panel and its depth of emotional expression place it among Titian's finest religious pictures' (p. 116).
The fullest and most satisfactory account of the two Prado pictures is that of Falomir, from which the information in this paragraph is largely drawn (op. cit., nos. 37-8). It is generally accepted that the earlier of the two is the Mater Dolorosa with clasped hands (see fig. 1; Prado, no. 443; Wethey, no. 76). This, showing the Virgin turned almost in profile to the left, is on panel and can thus be identified with the projected picture intended for the Emperor Charles V that, with the Ecce Homo to which it was intended as a pendant, is mentioned in a letter to him from Francesco Vargas, the Spanish ambassador at Venice, of 30 June 1553: 'una tabla de Nuestra Señora ygual del ecce homo que V...tiene' (Crowe and Cavalcaselle, op. cit., p. 507): the picture was delivered in 1554. The second picture, the Mater Dolorosa with raised hands (see fig. 2; Prado, no. 444; Wethey, no. 77), is also mentioned in correspondence of 1554. This was to be painted on stone, though it is in fact on marble, a very unusual support that establishes the identification beyond any reasonable doubt. The Virgin's hands are held apart: she is seen in three-quarter view, looking down to the left.
In both the Madrid pictures the Virgin is in a purple dress and wears a blue mantle, darker and richer in the earlier picture, paler in the later one. In the former, her diaphanous veil emerges from under a swathe of pale orange drapery; in the latter the white headdress is reminiscent almost of a nun's cowl. Two tears have issued from the Virgin's visible left eye in the first, four are visible in the second.
As Falomir has argued, Titian's customary practice where replicas were concerned was to entrust to an assistant the transfer of the design of the previous composition, in this case to the panel, and then he himself 'included minor changes in order to individualize the painting' (letter of 12 May 2005). X-rays establish that the point de part for the present picture was the earlier of the Prado pictures (fig. 3). The head was originally laid out in near-profile as in that picture, while the drapery and hands also adhere to the model of those in that work and the last seem, as Falomir (oral communication) points out, to have been traced from the same template as the panel at Madrid as the hands, although repeated, are realised with a characteristic fluidity. The sleeve of the dress, here a rich pink rather than purple, was obviously important to the artist, but his especial concern was the dramatic revision of the position of the head, silhouetted so effectively by the almost transparent white of her veil.
Titian was clearly at pains to supply a work that was in no sense a duplicate of the earlier pictures, and intriguingly the position of the face he adopted echoes that of another compositional type evolved by, or in the circle of, Massys, the Virgin in the Virgin with the Body of Christ at Munich (M. J. Friedländer, Early Netherlandish Painting, VII, Quentin Massys, ed. H. Pauwels, Leyden and Brussels, 1971, no. 15, pl. 20, as Massys and Willem Key), a design which seems to have been particularly popular in Spain. Massys's Virgin is dry-eyed, but the cheek of Titian's bears five tears. Not the least innovative aspect of the picture is its color. The artist's intention clearly was that the mantle should be purple, an appropriate although original colour for a picture of the grieving Virgin. Unfortunately the final layer, lake, darkened so that it was mistaken with a discoloured varnish and was substantially lost in a restoration that must have taken place between 1880, when the mantle was described as 'brownish' in the Royal Academy exhibition catalogue, and 1922, when 'red robes' were referred to in the sale catalogue. A photograph of the picture taken when it was with Nicholson in 1922-4 is in the Witt Library (negative number T9774). At some stage after this was taken and used for the publication of 1924, a subsequent restoration further distorted Titian's intentions by sacrificing what survived of the area of deep shadow on the right and other areas of shadow which defined the folds of the mantle. The former, in particular, can be shown by analogy with the earlier of the Prado pictures and the print by Bertelli (see infra) to have been of significance in establishing the volume of the form. Recently a glazing layer of lake has been replaced and as a result Titian's intentions can once more be understood.
The dramatic change in the position of the head of the Virgin in this panel establishes beyond reasonable doubt that this was the prime original of the type. The question that therefore arises is whether it can be the painting often thought to have been supplied in 1564 to King Philip II of Spain, but in fact painted for his secretary, Gonzalo Pérez, as Dr. Charles Hope has now established. This and its pendant, Ecce Homo, are mentioned in a document of late August 1564, published by La Duquesa de Berwick y de Alba, Documentos escogidos del Archivo de la Casa de Alba, Madrid, 1891, p. 465 (the compiler is indebted to Dr. Hope for this reference). Pictures by Titian of both subjects were engraved by Luca Bertelli, possibly in the 1560s, with the dedication 'PHILIPPO Regi Catholico hispanarum TITIANVS pictor clarissimus DD.' This dedication led Wethey to presume that the originals on which Bertelli depended were supplied to the King: he considered that the 1564 pictures corresponded with a pair recorded in the Sacristy of the Escorial in 1599 by Padre Sigenza, who specifically refers to engravings and copies after these, which might be taken to imply that the Virgin of the pair in question was indeed of the same type as this panel although he could, as Hope points out, equally well have associated the prints with the pictures of the subject supplied to Charles V: although not recorded in seventeenth-century accounts of the Escorial, the two pictures are, according to Wethey, stated to have been in place there by Palomino in 1724 and are recorded there as late as 1800, which would exclude the present Mater Dolorosa which was in the Colonna collection by 1783. The fact that the present picture is on panel also establishes beyond reasonable doubt that it cannot have corresponded with one of the pair first recorded in July 1600 in the King's Apartment at the Alcazar, Madrid, which was subsequently in the King's Oratory there (inventories of 1623 and 1636, in the latter of which canvas is specified as the support, see Wethey, p. 89, under no. 35): as Hope points out the mantle of the Mater Dolorosa in question is described as being blue. The fact that another version of the Mater Dolorosa on an unspecified support, also of Spanish royal provenance, was offered in the same sale at Christie's in 1811 as this panel is a further reminder of the difficulty of unravelling the provenance of such variants from incomplete records. And it should be stressed that there is no evidence that Perez's picture was ever either in the Spanish royal collection or at the Escorial, or, indeed, that it corresponded in design with the Bertelli engraving.
That Bertelli's print (see fig. 4) was based on the present panel (or a very closely related version of this) seems, however, to be incontrovertible: the pose and costume correspond very closely, although the character of the face is somewhat hardened, and the nose flattened, a similar change is evident when the copy of the Ecce Homo in the Escorial (Wethey, no. 35, copy 3, pl. 220) is compared with the corresponding print. Wethey's statement (p. 116) that the headdress of the Virgin distinguishes the composition from his 'Teniers' type (infra) is incorrect. Paul Joannides, however, considers that the omission of a halo in the present panel may exclude its identification with the engraved picture, and argues that the major alteration in the face of the Virgin does not necessarily establish that the panel was the first version of the revised type to be completed. The panel nonetheless undeniably represents the third, and presumably definitive, phase in the development of one of Titian's most successful religious images.