?Flint Castle

Andrew Wilton noted that the ‘figure on horseback and train of animals suggests a location such as Lancaster Sands’1 – see the watercolour Lancaster Sands of about 1818 (Birmingham Museums and Art Gallery),2 and another of about 1826 (British Museum, London),3 engraved in 1828 for the Picturesque Views in England and Wales (Tate impressions: T04534, T04535).

However, the principal focus in both Lancaster Sands works is a stagecoach, not hinted at here. Eric Shanes has instead, tentatively at first, compared this colour study to the watercolour Flint Castle, North Wales of about 1834–5 (National Museum Wales, Cardiff),4 engraved in 1836 for England and Wales (Tate impressions: T05873, T06121).5 The castle, which Turner had first visited in 1794,6 stands on the Welsh west bank of the broad, tidal Dee Estuary. Later Shanes was more conclusive, noting ‘the beached boats in the centre distance, the adjacent horses and carts (but facing in the opposite direction), a rather stormy sky and the castle represented in misty, light silhouette.’7 Assuming a connection, the overall effect here is relatively simple and naturalistic compared with the complex virtuoso rendering of the sun, clouds and reflections in the England and Wales watercolour.

For a possible Lancaster Sands ‘colour beginning’, see Tate D25132 (Turner Bequest CCLXIII 10). A ‘colour beginning’ comprising only sea and sky which has been linked to a variant watercolour of about 1830 showing Flint Castle (private collection)8 is Tate D35986 (Turner Bequest CCCLXIV 143). See also the introductions to the present subsection of identified subjects and the overall England and Wales ‘colour beginnings’ grouping to which this work has been assigned.

Technical notes:
The missing, torn bottom right corner noted by Finberg, extending to the bottom centre and about a quarter of the way up the right-hand side, was made good by 19751 and filled in with watercolour to match the surrounding area sand in that part of the composition. The ‘J Whatman | 1828’ watermark noted by C.F. Bell2 is not readily legible.