This simple, monochrome study of a building next to a stream, said to be a watermill, is one of fifteen generally slight colour sketches, all measuring roughly 8.9 × 11.4 cm (3 ½ × 4 ½ in), that appear to have come from a sketchbook worked late in Girtin’s career. Thomas Girtin (1874–1960) and David Loshak thought that these works ‘represent the fruits of local sketching trips taken during the summer of 1802’, and they argued that the fact that none of them were used as the basis for studio watercolours supported a late date (Girtin and Loshak, 1954, pp.84–85). However, only one of the scenes can be identified as a local view, Copenhagen House, Islington (TG1783), and although some of them appear to be imaginary, as in this case, others resemble the picturesque vernacular subjects sketched by Girtin in Essex three or four years earlier. Thus, whilst the sketches were evidently created at speed, it is unlikely that they were worked up on the spot, being produced instead in the studio to satisfy the market for the less formal aspects of the artist’s output. The evidence that they come from a sketchbook is also ambiguous, since, as the paper historian Peter Bower has pointed out, specialised books for the use of artists were not manufactured at this date, and they either used pocketbooks or they themselves gathered together sheets of paper (Bower, 2002, p.141). New evidence, in the form of the account of John Girtin (1773–1821) of the material that he removed from his brother’s studio at his death, suggests that the latter was the case here. John records that amongst the items that he appropriated to settle his brother’s extensive unpaid debts were ‘4 little Books partly of sketches and partly blank paper’, and it seems likely that these included the group of small drawings now in the British Museum, which would, indeed, date from late in his life (Chancery, Income and Expenses, 1804). John Girtin was thus responsible for splitting up the ‘little Books’ and selling the sketches to collectors such as Chambers Hall (1786–1855), the generous patron of the museum (Smith, 2017–18, pp.35–36).

Such is the slight manner in which the monochrome washes have been added to still-visible pencil work here that I wonder whether such modest sketches would have been attributed to Girtin were it not for the provenance and the visual links with other works in the group. Indeed, given the crude way in which the washes have been added, there is, as with a number of other sketches from the ‘little Books’, a suspicion that John Girtin himself might have been responsible for the colouring, attempting to make his stock more attractive to prospective purchasers.

A sketch of what has been interpreted as a ruined tower, on a similar small scale, appeared on the art market in 2014 (see figure 1), one of a pair of slight monochrome drawings attributed to Girtin (see TG1789 figure 1). The set of small sketches in the collection of the British Museum may be slight, but the colouring, even at its weakest, retains a sense of structure that this series of blots, without any pencil armature, singularly lacks. Other than the fact that the washes were added at speed and the image is equally small, I can see no reason to link the work with Girtin with any degree of certainty, and I am not inclined to accept the attribution of its pair either.

An unpublished mezzotint known only from an impression in the collection of the British Museum reproduces Girtin’s sketch on much the same scale (see print after, above). The print lacks a credit line, but comparing the mezzotint after The Valley of the Glaslyn, near Beddgelert (TG1324), one of only two other plates engraved by Samuel William Reynolds (1773–1835) after a sketch by Girtin, strongly suggests that he was its author too. In a document dated 1 October 1801, Reynolds refers to ‘Plates from drawings by Girtin … engraved on speculation’ and it is likely that this refers to the twelve landscapes that were eventually published as the Liber Naturae or A Collection of Prints from the Drawings of Thomas Girtin (Neill & Son, 1883) and that the plate for this work was either lost in the interim or was not recognised as being after the artist’s work. Indeed, three other prints without inscriptions were added to the published collection (Carnarvon Castle (after TG1315), Warkworth Castle (unidentified) and Tynemouth (after TG1086)) and the sale of the publisher’s widow, the owner of the ‘Entire Remaining Stock of Girtin’s “Liber Naturae”’, also included ‘Arthur's Seat, an unpublished plate’ (Sotheby's, 7 February 1893, lot 108).

(?) 1802

Copenhagen House, Islington

TG1783

(?) 1802

Farm Buildings

TG1789

1800 - 1801

The Valley of the Glaslyn, near Beddgelert

TG1324

1798 - 1799

Caernarfon Castle, from the River Seiont

TG1315

1797 - 1798

Tynemouth Priory, from the Coast

TG1086