This view from Piersfield on the river Wye, looking over Chepstow to the river Severn beyond, was for much of its existence attributed to Girtin. It was exhibited as such in the centenary exhibition in 1875 and was subsequently bought by the Victoria and Albert Museum, London, as by the artist. The attribution was confirmed by the fact that it was sold by one of the artist’s descendants, with the first owner listed as the artist’s son, Thomas Calvert Girtin (1801–74). The work did not appear in Thomas Girtin (1874–1960) and David Loshak’s catalogue in 1954, however, and they recognised that although Thomas did inherit some of his father’s works, presumably through Mary Ann Girtin (1781–1843), the majority were bought on the art market (Girtin and Loshak, 1954). It seems that Girtin’s son was no more immune than any other early collector to being swayed by a subject often treated by the young Girtin, the river Wye, rather than the work’s inherent qualities as a watercolour, and a dispassionate analysis suggests that the attribution is indeed no longer tenable. Thus there is no clear differentiation between buildings, foliage and rocks, with the castle itself lost amongst the buildings of the town, whilst the schematic representation of the water neither produces attractive patterns or convinces as a river in motion. Such problems are not always absent in Girtin’s earliest works made after the compositions of his master, Edward Dayes (1763–1804), and the initial attribution was presumably entered into on the basis that the work was based on one of Dayes’ compositions. No views by Dayes of Chepstow from this viewpoint have been traced, however, and in any case the work lacks the stylistic features that the young Girtin appropriated from his master. Instead, the use of a laid paper and the presence of Girtin’s later, more generalised manner suggest that the watercolour was executed by a later artist, perhaps amateur, working in Girtin’s style on a subject that was characteristic of his earliest works.

Another unresolved question relates to the relationship between this view of Chepstow and a sketchier version of the composition (see figure 1) that was also attributed to Girtin until it was shown at the loan exhibition organised by Agnew’s in 1953 as a precursor to the publication of Girtin and Loshak’s catalogue (Exhibitions: Agnew’s, 1953a, no.61). A note in the Girtin Archive states that it is a ‘dud’ (family code for the work of another, inferior, artist), and the watercolour did not subsequently appear in the catalogue (Girtin Archive, 41). This watercolour too is large in scale and, although it features different figures and shipping on the river, it follows the composition closely. The drawing is in poor condition, complicating any judgement about its status. However, I suspect that it is a second version of the subject by the same, anonymous artist and that, despite its less finished quality, it neither was sketched on the spot nor formed the basis of Chepstow Castle, on the River Wye. The Girtin Archive includes the suggestion that the author of the watercolour was William Pearson (1772–1849), but, although that is a possibility, there is no evidence to substantiate the claim other than that he is known to have produced a number of other competent essays in Girtin’s manner – works that, as here, were not simple copies of Girtin’s compositions (Girtin Archive, 14).

Another Wye scene attributed to Girtin, showing the junction of the river with the Severn (a subject that was depicted by Dayes), was noted by Girtin and Loshak in their catalogue (Girtin and Loshak, 1954, p.210). The watercolour, titled ‘Severn and Wye: Sunset’, has not been seen since it was sold at auction in 1914 (Exhibitions: Sotheby’s, 24 February 1914, lot 72) and it too is assumed to be a work by a follower, though confirmation of this must await its reappearance.