Until the discovery during the preparation for the 2002 Girtin bicentenary exhibition at Tate Britain of an inscription on the back that identifies the source for this watercolour of the Temple of Vesta at Tivoli as a ‘Cork model’, it was assumed that the composition was based on a print by Giovanni Battista Piranesi (1720–78). More specifically, it seemed to be a free adaptation of plate sixty-three of Piranesi’s Vedute di Roma (Views of Rome) with the addition of an imaginary landscape and changes to the form of the temple to the right, in which case it roughly reproduced the view seen in an earlier Monro School drawing that is incorrectly identified as the ‘Temple of the Sibyl’ (TG0590). Given that Girtin did not travel to Italy and could not have studied the subject himself from life, the newly revealed inscription closes off Piranesi as the source and opens up, in turn, two alternatives. Firstly, Giovanni Altieri (active 1766–1802) produced a number of cork models of the Temple of Vesta that include part of the hill and the arched substructure of the building as depicted here by Girtin, and it is possible that his source was a more complete version of the example purchased by Sir John Soane in 1802 (see figure 1). The second, slightly more likely option is that Girtin sketched one of the exhibits shown at the museum of models that was opened by a Monsieur Dubourg in Duke Street in London in 1798. A later image indicates that the Temple of Vesta and its landscape setting (see figure 2), constructed on the scale of ‘one inch to a foot’, was one of the central features of the show, and, according to Dubourg’s Descriptive Catalogue, the ‘Sibyl’s Temple’, as the monument was then known, was part of a model of the town of Tivoli that included a recreation of the ‘grand Cascade’ (Du Bourg, 1810, p.44).

The fact that Girtin’s drawing had its origin in a cork model, even if we cannot yet pin down the precise source, nonetheless changes our understanding of the work. Clearly, it is unrelated to the commission that John Henderson (1764–1843) gave to Girtin to make copies of prints he owned by Piranesi, such as The Temple of Augustus at Pula in Istria (TG0886). Rather than being a commission, the watercolour was presumably made by Girtin with an eye on the market for his sketches, which was firmly established by 1799–1800. The rapid manner in which the washes of a limited palette of colours are swept over a still evident sketchy pencil drawing combined with the very liquid tones in the foreground, which recall the sort of accidental effects of blotting associated with working in the field, suggests a work coloured on the spot. The fact that ‘on the spot’ here means a London interior is confused by the Italian subject and a sky, which by definition must have been invented. The result is the reimagining of a secondary source as an encounter with nature and a new type of commodity: the pseudo on-the-spot sketch as a sign of the artist’s superior creativity.

On a technical note, the paper historian Peter Bower has identified the support used by Girtin as an off-white laid drawing cartridge paper by an unknown Dutch maker, that was worked on the wireside, where the surface is impressed with the lines of the mould used in its manufacture (Smith, 2002b, p.91; Bower, Report). The paper comes from the same batch that Girtin used for the distant view of Conwy Castle that he painted for Sir George Beaumont (1753–1827) from one of the amateur’s own sketches (TG1578). Both drawings probably date from around 1800, though it has not been possible to identify any further links between them or, indeed, to establish whether this work also came from Beaumont’s collection.

1794 - 1797

Tivoli: ‘The Temple of the Sibyl’

TG0590

1797 - 1798

The Temple of Augustus at Pula in Istria

TG0886

1799 - 1800

Conwy: The Town Walls

TG1578