J.M.W. Turner: Sketchbooks, Drawings and Watercolours

D08084–D08106, D08108–D08123, D08125–D08142, D08144–D08187; D08222, D08231, D08274, D25373, D25451; D40043, D40045, D40141, D41481, D40351–D40381; D41516; N02782, N02941, N02942, N03631
Turner Bequest CXV 1–29, 31–48, CXVI A–E, G–R, T–V, X–Z, CXVII A–K, M–T, W–Z, CXVIII A, C, D, F, H, J–M, O, U, d–h, CXX I, Q, CXXI R, CCLXIII 328, 251
Vaughan Bequest CXVI S, CXVII L, U, V, CXVIII B, E, G, I, N, P–T, V–Z, a–c

Note:

Two records, Tate D08107 (Turner Bequest CXVI F) and D08124 (CXVI W), have been cancelled. The first was apparently a spurious duplication, and the second, recorded as a touched etching, has long been recorded as missing. See entries for Tate D08185 (CXVIII e) and D08123 (CXVI V) respectively for details.

The Liber Studiorum, a series of Turner’s landscape and seascape compositions published as prints in etching and mezzotint, has been described as perhaps containing ‘the pith of all that is best in his life and work’,1 ‘central to Turner’s career as the most personal and carefully conceived series of prints in his entire oeuvre’,2 or at least the ‘most complete document of Turner’s attitude to his art in the first decade of the [nineteenth] century’,3 and ‘one of the most comprehensive exercises in publication ever mounted by a great artist.’4 The Latin title may be translated as ‘Book of Studies’,5 although it was not issued as a bound book, and did not contain any explanatory text beyond the titles of (most of) the compositions, and letters printed above each of them to indicate general categories of landscape discussed in detail below. The incomplete series comprised seventy engravings of a planned one hundred, published by Turner (or, at an early stage, on his behalf) between 1807 and 1819 in fourteen parts together with a frontispiece. Marcia Pointon has described it as ‘a named whole, which, on inspection, turns out to comprise only fragments.’6

For the majority of the compositions, Turner specifically prepared horizontal, ‘landscape’-format designs in brown watercolour washes. These were generally around 178 to 204 x 254 to 280 mm (7 to 8 inches in height, and 10 to 11 inches wide), on about the same scale as the engravings, which were themselves printed in shades of brown.7 Most of the original working drawings for the published plates were retained by the artist, together with some for further designs which were engraved but not published, and various drawings in a similar manner which have come to be regarded as part of the series. Together with those acquired from other sources, these are the works catalogued here, in the sequence of the standard Liber Studiorum print catalogues by W.G. Rawlinson (1878; revised 1906) and A.J. Finberg (1924) – see checklist below. The subsections in the present catalogue relate to:

Designs for twenty-two Liber compositions were bequeathed to the nation by Henry Vaughan in 19008 and numbered in sequence with the Turner Bequest works in Finberg’s 1909 Inventory to conform to Rawlinson’s established print catalogue:9 Rawlinson/Finberg nos.1 (etching with wash), 16, 36, 46, 54, 58, 60, 63, 68, 73, 74, 76, 77 (unengraved variant), 79 (etching with wash), 85, 87, 88; and Rawlinson nos.92–94, 97, 99. There are three further Liber studies at Tate, acquired from other collections, relating to Rawlinson/Finberg nos.27 (etching with wash) and 78; and Rawlinson no.98.10

Checklist:

The following checklist presents the series in Rawlinson/Finberg order. Titles of published compositions (Rawlinson/Finberg nos.1–71), with their idiosyncratic spelling and punctuation, are given precisely as lettered on impressions held by Tate; for those published without a lettered title, or unpublished (nos.72–91), the customary title is shown in inverted commas. The unengraved designs (Rawlinson nos.92–100) are less formally identified and their titles are shown in italics. The categories to which Turner assigned the designs with initial letters in the plates, such as ‘P’ for ‘Pastoral’, are also indicated. Those compositions not represented at Tate by drawings or etchings with wash are listed in brackets, and are discussed in the introductions to the subsections, with some additional information: the etcher and engraver of each plate, publication dates, Tate’s holdings of the prints, Turner’s late paintings inspired by the series, and secondary versions of the prints.

There are a few drawings directly related to Liber compositions in other collections: Rawlinson/Finberg nos.77, 80, 81; Rawlinson nos. 95, 96.11 Other variants and studies are discussed in the catalogue entries for the relevant Tate drawings (see in particular nos.20, 26, 98). Currently-untraced Liber drawings have been recorded for two subjects: Rawlinson/Finberg nos.52, 75. No Liber-type drawings are recorded in several instances, where the prints were probably derived directly from existing works and Turner may have felt it unnecessary to produce intermediate studies: Rawlinson/Finberg nos.35, 40, 44, 50, 55, 62, 69, 70, 72, 83, 84, 89.

The original drawings in the Tate Collection directly related to each plate are listed with their Tate and Turner or Vaughan Bequest numbers; those marked ‘*’ are impressions of Turner’s etched outlines worked up by him with tonal watercolour washes; no.89 is represented in the form of its engraved copper plate rather than as a drawing.

Published plates:

Part 1: probably 1807

Part 2: 1808

Part 3: 1808

Part 4: 1809

Part 5: 1811

Part 6: 1811

Part 7: 1811

Part 8: 1812

Part 9: 1812

Part 10: 1812 (see also ‘Frontispiece’, no.1 above)

Part 11: 1816

Part 12: 1816

Part 13: 1819

Part 14: 1819

Unpublished plates (categories indicated in square brackets were not specificied in Turner’s own notes, which are transcribed later in this introduction):

There are further compositions, not etched or engraved by Turner, which Rawlinson included in his 1878 catalogue: ‘The eight [sic] Drawings which follow [nos.92–99; but Rawlinson continued to no.100], never having been Engraved or Etched, cannot definitely be said to belong to Liber Studiorum. They have however always been considered to represent Turner’s preparations for the completion of the work’.12 His 1906 edition concludes with no.99, though the former no.100 is mentioned.13 In 1924, Finberg stopped short at no.91: ‘I have confined my Catalogue to the list of published and unpublished plates. I consider it a mistake to include any unengraved drawings in the catalogue of a series of engravings.’14 Forrester’s 1996 catalogue also ends with no.91, with only passing (though useful) references to the following works. As most of the drawings are at Tate, they are included here in Rawlinson’s sequence. Categories have not been suggested for these remaining designs; a few are obviously suited to the established terms, while others are less easily pigeon-holed.

Plate ‘101’

In the anonymous 1911 Miniature Edition of reproductions, further works were speculatively connected to the Liber, probably by Rawlinson, who gave ‘generous help and advice all through’,15 as alternative designs for a last plate to make up the originally projected total of one hundred plus the Frontispiece; each was captioned ‘101??’. They had already been included in the Liber sections of Finberg’s 1909 Inventory, as indicated by their Turner Bequest numbers.

Four designs on sheets which were originally quarters of a single whole are also catalogued here with the unengraved drawings, although their relationship to the Liber is uncertain; two of them were included in Finberg’s Inventory among the Liber designs:

There are a few further drawings at Tate and elsewhere which are possibly connected to the Liber, but have not been included in the Liber catalogue sections; they are discussed briefly in the introduction to the unengraved Liber drawings.

Perhaps inspired by the printing of additional sets of the Liber in 1845 to fulfil an order from John Ruskin (see below), Turner made some paintings, mostly based on its more classical subjects, possibly ‘as a final attempt to achieve some kind of completion or closure.’16 Ian Warrell has suggested that, ironically, they were provoked by Ruskin’s criticism of Turner’s precursor Claude Lorrain in the first volume of Modern Painters (1843); although these unfinished paintings remained private, Turner’s last exhibited works were also indebted to Claude.17 The Liber paintings are listed by Butlin and Joll,18 the best known being Norham Castle, Sunrise (Tate N01981).19 There are drawings relating to them all in the Bequest (except Rawlinson/Finberg no.35, Inverary Pier), but Turner probably responded instead to the new Liber impressions. Further details are given in the introductions to each subsection, and in the individual entries for Rawlinson/Finberg nos.1, 3, 8, 13, 18, 28, 43, 53 and 57. A further painting, traditionally known as The Val d’Aosta (National Gallery of Victoria, Melbourne),20 has recently been identified as being based on an unengraved design, Falls of the Rhine, Schaffhausen (Rawlinson no.93).

The present account is based largely on the comprehensive catalogues by Rawlinson and Finberg, and Gillian Forrester’s Tate exhibition catalogue,21 which are more detailed on many points (particularly with regard to the prints) than the present text, intended as a background to Turner’s Liber drawings.

Turner had been elected a Royal Academician in 1802 and, while still showing at the RA exhibitions, began exhibiting at his own gallery in 1804 – an indication of his growing reputation while he was still in his twenties. He had worked for topographical print publishers from about the age of eighteen, starting with watercolours (mainly of English scenes) engraved for the Copper-Plate Magazine between 1794 and 1798.22 In 1805 he exhibited at his gallery a major painting of The Shipwreck (Tate N00476),23 engraved and published early in 1807 as a large mezzotint by the unrelated Charles Turner (Tate impression: P79356).24 This was the first print after one of his oils; Finberg related the history of its engraving and suggested that time Charles Turner took over it did not bode well for his involvement with the Liber (see below).25 However, about 130 subscribers are recorded in the Shipwreck (1) sketchbook, indicating that there could be both financial reward and status-enhancing publicity in such projects (Tate D05377, D05379; Turner Bequest LXXXVII 2, 4).

Turner’s close friend, William Frederick Wells, a drawing master, watercolourist and etcher,26 had recently collaborated with John Laporte on a series of soft-ground etchings, A Collection of Prints Illustrative of English Scenery, from the Drawings and Sketches of Thomas Gainsborough, R.A.27 (for impressions from the 1819 edition, see Tate T11015–T11074). They combined line and tone, were of a similar size to those of the subsequent Liber, and in these senses provided a direct British model. Other precendents include late eighteenth-century series of topographical views, such as those line-engraved after Thomas Hearne or aquatinted after Paul Sandby, and most immediately, the 1803 aquatints Twenty of the Most Picturesque Views in Paris and its Environs after Turner’s late contemporary and friend Thomas Girtin.28

In a letter of 1853, Wells’s daughter Clara (Mrs Wheeler) cited a more illustrious influence, the so-called ‘Liber Veritatis’ of Claude Lorrain (?1604/5–1682):

Although it seems unlikely that the categories and contents of the first part would have been established so rapidly, the outline of events30 is generally accepted, given the lack of documentary evidence from Turner himself. Another account, apparently related by the artist David Cox, describes Turner and Wells ‘turning over Claude’s “Liber Veritatis” together, and when they were near the end, his friend exclaimed, “Turner, you could do better things if you chose.”’31 The Liber engraver Charles Turner described it slightly differently, as ‘first thought of at the house of his friend, Mr. Wells, as a companion to Claude’s Liber Veritatis, and that the first drawing was made for it in his parlour’.32

Between about 1635 and 1675, Claude had recorded each of his completed landscape paintings in a book of 195 monochrome pen and wash drawings – a ‘unique combination of a pictorial document and a work of art.’33 From the 1720s they were in the collection of the Dukes of Devonshire, mounted with five additional drawings in an album; they are now in the British Museum, London (1957–12–14–6 to 206). In 1777 John Boydell published a two-volume edition, newly named the Liber Veritatis,34 reproducing the 200 drawings in a combination of outline etching and tonal mezzotint by Richard Earlom.35 It seems likely that Turner only knew Claude’s originals through Earlom’s prints. The newly-coined Latin title, meaning ‘book of truth’, was derived from Claude’s reported use of the equivalent Italian phrase ‘Libro di Verità’, and has become the usual title for the drawings. From 1802 prints by Earlom from drawings in other English collections were issued, and gathered as a third volume in 1819.36 It may have been this new phase of publication that alerted Turner, probably through Wells, to the 1777 edition. Michael Kitson observes: ‘Various ... external characteristics, such as medium, number of plates ... and, above all, title, serve to reinforce the connection’.37 Ironically, when F.C. Lewis later engraved Claude’s drawings in the British Museum, the prints were themselves issued as a ‘Liber Studiorum’38 (see below for Lewis’s brief involvement with Turner’s Liber).

Though earlier writers emphasised what they perceived as Turner’s presumptuous, ‘retaliating’39 rivalry with Claude, Finberg felt direct comparison was inappropriate since Claude never had any intention of publishing his ‘Liber’, and pointed out the obvious differences between the projects, focusing on Turner’s direct artistic control over the production of his series.40 Kitson has noted that ‘only in one category, the E.P., did he show any obvious debt to Claude. Turner’s classification of his Claudian compositions as elevated landscapes is a significant reassessment of the 18th-century view of Claude.’41 (See below for a discussion of the ‘Elevated Pastoral’ and other categories of the Liber.)

Indeed, as in much of Turner’s work, the influences of – and implicit rivalry with – a wide range of Old Masters and contemporary British artists, from Poussin and Rembrandt to George Morland and David Wilkie, can be detected in the various modes adopted in the explicitly-differentiated categories of the Liber42 (specific instances are discussed in relation to the individual drawings). Evelyn Joll has summarised the artist’s situation and aims: ‘It was one of Turner’s central ambitions to raise the status of landscape painting onto an equal footing with other kinds of subject matter, religious, historical or mythological pictures, for instance, hitherto ranked above landscape in the pictorial hierarchy that was recognised at the time.’ The Liber ‘enabled him both to explore the full range open to a landscapist and to impart his views on the subject to a wide public’.43 Turner’s original concept of the series appears to have been as a comprehensive, personal artistic statement,44 serving partly as a summary of his relatively short career to date, and as a demonstration not just of his versatility in established modes but also of his originality:45 ‘The only excuse for the publication was the genius of the producer.’46

Andrew Wilton has suggested that the Liber’s ‘careful planning betrays Turner’s strong academic bent’;47 he was anxious to progress professionally within his beloved Royal Academy in emulation of its first President, Sir Joshua Reynolds, who had delivered his credo in fifteen learned Discourses (1769–90). Turner was appointed the Academy’s Professor of Perspective in December 1807 (see further notes in relation to the Liber’s Architectural category, below); in the last of the lectures he initially delivered in 1811, ‘Backgrounds: Introduction of Architecture and Landscape’, he argued for the acceptance of landscape painting as a genre on a par with the traditionally favoured historical subjects. The engraver John Landseer reported that Turner would have preferred to be a Professor of Landscape Painting;48 Landseer had raised the issue of such a post in his published Royal Institution Lectures on the Art of Engraving (London 1807), a copy of which he had given to Turner at about the time of the Liber’s inception,49 and subsequently wrote about Turner’s project (see below).

The original wording for the Liber’s title lettering may be the scribbled draft in the Shipwreck (1) sketchbook, enclosed in a tentative oval, perhaps indicating a palette, and flanked by what could be a cornucopia, reading (approximately): ‘No.1 of | Liber – Studiorum | being | Studies for Pictures in History | Mountains Pastoral Marine | and Architectural Landscapes | Price 15’ (Tate D05385; Turner Bequest LXXXVII 10). Finberg read the penultimate line as ‘and architecture – and Landscapes.’ This he considered ‘ludicrous’ and ‘gives some idea of the difficulties which the proposed system of classification presented to Turner’s mind.’50

Although Finberg noted that Turner’s classification ‘gave him a programme to work upon, and introduced some kind of order’, he described interest in the categories as ‘nowadays ... somewhat remote’51 and also observed: ‘Many of the designs could quite easily be classed under a different heading.’52 Walter Armstrong had considered that the ‘whole business of H. and P., and E.P. and M. [see below] was childish, and ... served no purpose at all, except to embarrass the artist and make it difficult for him to keep the promises he had made to himself.’53 As early as 1808, John Landseer had speculated on the difficulty of neatly pigeon-holing some of Turner’s compositions into his own categories.54 Rawlinson initially failed to discover ‘any connected design in the work’ as a whole,55 but later he was more positive, considering ‘how completely Turner succeeded in carrying out the original idea which was suggested to him in 1806 ... The Liber gives full expression to its author’s powers in every branch of pictorial art.’56 However, Gillian Forrester has conjectured that Turner himself ‘was unable to arrive at a definitive summary of what his own series was about’,57 thus providing scope for various self-assured Victorian interpretations but leaving later scholars less confident; in her own thesis the Liber is ‘a series of overlapping narratives, an amalgam of different models’, incorporating various commercial, didactic and theoretical elements.58

The initial letters mentioned above are those which Turner placed above each published image, indicating his categories – Architectural, Historical, Marine, Mountainous and Pastoral. The last was represented in each part by two plates: one headed ‘P’ for modern-day country scenes, fundamentally derived from precedents in Dutch art, often regarded at the time as somewhat low and vulgar;59 the other headed ‘EP’ or ‘E.P.’ for idealised, classical landscapes inspired by Claude and, in the absence of a specific explanation, generally interpreted as ‘Elevated’ Pastoral, from Turner’s use of the word elsewhere;60 alternatives such as ‘Elegant’61 and ‘Epic’62 have also been mooted. An isolated list in the Finance sketchbook (which includes notes relating to transactions as early as 1807) presumably relates to the Liber: ‘<1 Marine | 2 Pastoral> | Epic Compositions 1 | Compositions 2 | Pastoral 3 | Marine 4 | Buildings’ (Tate D08321; Turner Bequest CXXII 25). The first two lines are crossed out, but John Gage has interpreted the remaining terms as relating respectively to the Historical, EP, Pastoral, Marine and Architectural categories, with ‘epic’ thus equating with ‘historical’63 – as Forrester concurs, ‘Epic is a narrative form’.64

The Sublime, the Beautiful and the Picturesque are the classic categories in aesthetic debate of the period, as first defined by theorists including Burke, Gilpin and Price.65 Andrew Wilton has suggested a literary model in the classes of poetry – Pastoral, Lyric, Didactic, Descriptive and Epic – discussed by Hugh Blair in his Lectures on Rhetoric and Belles Lettres (London, Edinburgh and Dublin 1783 and later editions), which Turner is known to have consulted in relation to his own poetry and lecturing.66 Forrester notes another precedent, provided by a book Turner owned, S. Harding’s Sculptura-Historico-Technica: or the History and Art of Ingraving (London 1747); as well as technical information, advice is given on forming a print collection in terms of categories, iconography and styles.67

Turner’s own system has been sympathetically considered by Laurent Busine as ‘a suite conceived like an alphabet or, even more so, like a herbarium. Each work is not only the reproduction of a particular motif but, because it is matched with a sign which recalls a larger definition, is the typological example of all possible subjects, of which the plate is one illustration.’68 Marcia Pointon has suggested that it ‘challenges the overall Genre system within which all academic art functioned ... The system thus by implication subsumes all artistic practice.’69

The relationship of Turner’s publication to a range of contemporary self-help drawing manuals and pattern books, including those by artists close to W.F. Wells (himself a teacher) such as John Laporte and W.H. Pyne, has been explored by Forrester,70 as has the Liber’s influence on John Varley’s Treatise on the Principles of Landscape Design (London 1816–21), which included a pair of contrasting ‘Epic’ and ‘Pastoral’ landscapes.71 Unlike such expressly educational publications, however, Turner’s Liber was, as Gerald Wilkinsion has termed it, ‘a manifesto without words’ and ‘a silent sequence of reminders of the immense diversity of expression which he saw in landscape.’72 Well-known contemporaries such as David Cox, John Sell Cotman and John Constable owned copies.73 Constable referred to it with casual sarcasm as the ‘liber stupidorum’,74 but there is a parallel to it in his own work: the Various Subjects of Landscape, Characteristic of English Scenery, often known as English Landscape Scenery. This series of twenty-two mezzotint plates, similar in format to Turner’s, was engraved by David Lucas and published by the artist between 1830 and 1832 (for impressions and catalogue texts see Tate T03983–T04063). As Forrester notes, Constable originally classified his compositions with initial letters which Andrew Wilton has interpreted as ‘pastoral’, ‘fancy’ and so on, but later characteristically sought to refute any reliance on artistic precedent in favour of Nature.75

Finberg inferred that there would have been forty Pastoral plates in total, and therefore an average of fifteen for each of the four other categories (perhaps with up to twenty Marines, allowing for less inspiration in other areas) to make up the full hundred;76 his additional candidates for some categories are set out at the relevant points below. Turner’s various lists and notes on the series have been transcribed, dated and interpreted by both Finberg and Forrester; the principal sources are found in the following sketchbooks, quoted here and in the entries for individual works:77 Tabley, No.1, 1808 (Tate; Turner Bequest CIII); Hastings, circa 1809–14 (Tate; Turner Bequest CXI); Finance (loose associated sheets), circa 1810 (Tate; Turner Bequest CXXII); Liber Notes (1), circa 1815 (Tate; Turner Bequest CXLIII); Aesacus and Hesperie, circa 1817–18 (Tate; Turner Bequest CLXIX) and Farnley, ?1818 (Tate; Turner Bequest CLIII). Most importantly, the Liber Notes (2) sketchbook (Tate; Turner Bequest CLIV(a)) includes a list of the projected first fifty subjects, probably compiled early in 1808, transcribed and discussed below, and categorised lists to which he may have added subjects as late as 1818 (see immediately below).78

The transcriptions here follow Finberg and Forrester’s readings, with minor variations resulting from further examination of the original lists. Rawlinson/Finberg Liber numbers are indicated in square brackets. Those subjects marked [-] were not developed for the series; notes on these subjects are mainly derived from Forrester’s comments, which in turn draw on Finberg’s.

Historical:

Turner lists all of these (with some unpublished or unexecuted additions) in the Liber Notes (2) sketchbook (Tate D12170, D12171; Turner Bequest CLIV (a) 30a, 31),79 under the heading ‘Historical’:

Forrester has discussed Turner’s literary interests in relation to the Liber,92 and notes possible links to subjects treated by Poussin and Richard Wilson93 in relation to a second list of historical (mythological) subjects inside the back cover of the Tabley, No.1 sketchbook (Tate D40721; Turner Bequest CIII), in use on Turner’s visit in June 1808:94

Mountainous:

Turner lists twelve of these (except the Grand Chartreuse and Inverary Castle, published 1816; and with two additions) in the Liber Notes (2) sketchbook (Tate D12166, D12167; Turner Bequest CLIV (a) 28a, 29),101 under the heading ‘Mountainous’:

Pastoral:

Turner apparently lists all fourteen of these (with some unpublished or unexecuted additions – the crosses probably indicating subjects as yet unengraved)104 in the Liber Notes (2) sketchbook (Tate D12160; Turner Bequest CLIV (a) 25a),105 under the heading ‘Pastoral’:

In seeking further subjects to make up an inferred twenty Pastorals, Finberg – having transcribed the list – incorrectly stated that Stonehenge (Rawlinson/Finberg no.81) appears on it, and also suggested Sand Bank with Gipsies (no.91).112 The designs in this category have sometimes been regarded as the least successful, with their ‘weak, sketchy, and scarecrowy’113 figures, condemned as coarse and vulgar by early commentators such as Rawlinson and Brooke, while Ruskin felt that ‘the commonplace prevails to an extent greatly destructive of the value of the series, ... introducing rather discord than true opponent emotion among the grander designs’.114 Andrew Wilton has considered the ‘scenes of children sailing toy boats, or pigs in a farmyard ... which, divorced from a general landscape context, become unexpectedly awkward’, concluding: ‘In this respect Turner is at the opposite pole from the Picturesque tradition, and from Constable, who was an innovator in deliberately abandoning the “heroic” as an essential of serious art.’115 Forrester has identified a ‘narrative of seasonal rural occupations’ as one element of the Liber (although Laurence Binyon had earlier criticised Turner’s lack of sympathy with the ‘slowly changing ritual of the seasons’),116 together with the possible ‘patriotic significance’ of such scenes, particularly in the context of the Napoleonic Wars, concurrent with much of the series.117

EP (Elevated Pastoral?):

Turner lists all fourteen of these, together with the unpublished Apullia and one other, in the Liber Notes (2) sketchbook (Tate D12162 and D12163; Turner Bequest CLIV (a) 26a, 27),118 under the heading ‘EP’:

To complete an inferred twenty EP subjects, Finberg suggested, in addition to Rawlinson/Finberg no.72 and the apparently unexecuted ‘ditto Thames’: Dumbarton (Rawlinson/Finberg no.75), The Temple of Jupiter (no.77), The Stork and Aqueduct (no.83) and Narcissus and Echo (no.90).123 However, as a subject from Ovid, the latter would have been equally if not more suitable as a Historical plate. Using the example of The Junction of the Severn and the Wye (for drawing see Tate D08132; Turner Bequest CXVII E) Michael Kitson has observed that such Liber compositions are ‘not a pastiche of Claude’s style, but rather the English landscape expressed in terms of Claudian composition. The less stylized treatment of the foliage and the more diffused English light in the drawing reflect Turner’s own observation of nature.’124 Brooke found this combination disturbing: ‘Half nature then, half convention; half Turner, half pseudo-Claude ... like a Gothic building with a Palladian porch, and as disagreeable.’125

Marine:

Turner lists all of these (together with Inverary, deleted and included instead with the ‘Mountainous’ subjects noted above; and Dunstanborough, which became an ‘Architectural’ subject, noted below) in the Liber Notes (2) sketchbook (Tate D12164; Turner Bequest CLIV (a) 27a),126 under the heading ‘Marine’:

The list continues (Tate D12165; Turner Bequest CLIV (a) 28),127 with nine further subjects, all but one apparently neither drawn nor engraved for the Liber, which Forrester, prompted by Finberg, identifies with varying degrees of certainty,128 as set out here:

In addition to the above, Finberg suggested possible additional subjects: The Felucca (Rawlinson/Finberg no.82), Moonlight at Sea (no.85) and Moonlight on the Medway (no.86).

Architectural:

Turner lists eight of these in the Liber Notes (2) sketchbook (Tate D12168; Turner Bequest CLIV (a) 29a),140 under the heading ‘Architecture’ (London did not appear in Turner’s 1808 draft of the first fifty plates; Dunstanborough was originally listed as a ‘Marine’ subject, noted above; and the Church was a later addition to the series):

Having proposed the unpublished Temple of Jupiter (Rawlinson/Finberg no.77) as an ‘EP’ subject (probably correctly), Finberg observed that ‘curiously’, no further Architectural subjects seem to have been prepared, given that up to fifteen might have been expected as an equal proportion with other categories.141 But John Gage has suggested that this was ‘a branch which ... Turner had effectively left behind him’142 before embarking on the Liber. This may account for the rather variable selection of castles, religious buildings and townscapes which Turner felt obliged to include, perhaps in the light of his application to be the Royal Academy’s Professor of Perspective in March 1807143 while he was preparing the Liber; he was appointed in December, by which time the first part had been published.

As the first step in the production of the prints, Turner would usually inscribe an outline on a prepared ground on a copper plate, using his brown watercolour wash drawing as a reference, but always leaving the sky blank at this stage.144 Thomas Lupton, one of the professional engravers employed on the series, recalled that the plate’s engraver ‘had also to lay in the etching ground and trace the subject onto the plate for the painter to etch, which was his uniform practice.’145 Sometimes no Liber-type drawing is known, and Turner may have worked directly from existing works in other formats;146 occasionally, he would then tint an impression of the etching with watercolour to indicate the tones. Luke Herrmann has speculated that Turner worked ‘freehand or from a set of separate outline drawings’ to make the etched outline, before repeating the compositions as the familiar wash drawings,147 but Forrester has dismissed the idea of such (otherwise unrecorded) outline drawings as an unlikely complication of the production process.148

In a few cases the etching has been attributed to Henry Dawe (Rawlinson/Finberg nos.54, 58, 60, 64).149 Turner retained several proofs of each etching in his studio; most remained there until the sales from his estate in the 1870s.150 Their qualities as ‘drawings’ have been highly praised by writers on the Liber from Ruskin onwards,151 though they have more recently been described as ‘not especially attractive or remarkable’,152 and the plates were probably handed over to the engravers for the biting process.153 Of the unpublished plates, several were not begun with an outline etching (nos.81, 82, 84–89), while the etchings for nos.77 and 80 are again attributed to Dawe. The outlines of nos.72–76, 78, 79 (first version), 83, 90 and 91 are generally attributed to Turner, although Finberg questioned nos.75 and 78;154 Forrester has seen no reason to doubt Turner’s involvement.155

The engraver would then work over the outline in mezzotint to establish the tones and further detail. Turner himself learned the technique and carried out this work in addition to the etching for one plate in each of parts 6 to 14 (Rawlinson/Finberg nos.28, 35, 39, 44, 50, 55, 58, 60, 66, 70; see also central ‘picture’ in the Frontispiece, no.1);156 he also seems to have at least partially engraved several unpublished plates (nos. 81, 84, 85, 87, 88, 89, 91).

Turner was in correspondence with Thomas Girtin’s engraver brother James (John) before or around the beginning of the Liber157 in relation to work the latter was undertaking for him, which Rawlinson took as evidence that Turner intended to produce the whole Liber in aquatint,158 but Finberg refuted this idea due to a lack of specific evidence.159 Alternatively, Turner may have been considering Girtin as a co-publisher.160 One of the earliest designs, Bridge and Goats (Rawlinson/Finberg no.43) was indeed undertaken largely in aquatint, by F.C. Lewis, but was not published until later in the series, after Turner and Lewis had rapidly parted company over technical issues and fees (see entry for Tate D08146; Turner Bequest CXVII R). Rawlinson took this as further proof of Turner’s original intention of using aquatint for the whole series.161

In the event, aquatint was used on a few plates, particularly in the skies (see Rawlinson/Finberg nos.13, 14, engraved by Charles Turner), but Turner apparently preferred the greater range and depth of tone which mezzotint could provide for the landscape features,162 and ‘its ability to suggest vastness, uncertainty, night, obscurity and so on – all elements of the sublime’.163 Although Turner used aquatint on the early plates he engraved himself (nos.28, 35, 39, 44), mixing the two techniques was challenging, and the later published plates which he engraved (nos.50, 55, 58, 60, 66, 70) used only mezzotint for the tones. Three of his unpublished plates combining mezzotint and aquatint in experimental ways were abandoned (nos.85, 87, 89); but he also occasionally applied aquatint to other plates as the original mezzotint wore down.164 He found himself obliged to constantly rework the plates,165 as the burr on the relatively soft copper surface would begin to lose definition after thirty or fewer fine impressions, many of which he set aside for himself.166 John Gage has suggested this was ‘an early example of that development of a representative collection of work that was to lead to the Turner Bequest.’167

Finberg pointed out the novelty of Turner’s approach: although some of the Liber designs were fairly closely based on existing paintings, as was the established practice for reproductive engravings, many were newly derived from sketches or original designs.168 Gage has discussed at length Turner’s attitude to engraving as an act of ‘translation’, rather than a simple reproduction of an image, particularly with regard to colour; the difficulties of consistent and sensitive colour reproduction at that period meant that Turner generally avoided the issue, to the extent of producing the watercolour designs for the Liber in monochrome.169 As Forrester notes, ‘there is some uncertainty as to what the prints were translating’,170 as some of the Liber drawings are quite far removed from their ostensible sources, while others were themselves developed further as images in terms of light and atmosphere through the printmaking process. Mezzotint was well established as a means of reproduction,171 generally for portraits, but Turner in effect developed it as a creative medium in close collaboration with his engravers,172 working from dark to light as the additive texture established by rocker tools was scraped and burnished away – ‘not surprising in a painter one of whose most original contributions to watercolour technique had been the extensive wiping out of lights from the fluid mass of rich, dark, wash.’173 Forrester has demonstrated a creative and educative partnership between Turner and his engravers through the sometimes lengthy proofing and development of the Liber compositions.174

In addition to Lupton, Dawe and Lewis, Turner had employed the Shipwreck’s Charles Turner for all twenty plates of the first four parts (Rawlinson/Finberg nos.2–21). The engraver also acted as publisher for parts 2, 3 and 4,175 but he requested an increase in his fee from eight guineas per plate (also covering his responsibilities as publisher and distributor)176 to ten or twelve, and was refused, leading to a rift aggravated by other issues.177 The artist also criticised him for a lack of advertising, and falling behind with the preparation of the plates – ‘everything has conspired against the work’178 – and reverted to being his own publisher for the rest of the series, supposedly ‘screwing down his engravers to the lowest possible scale of remuneration’.179 (However, plates nos.26, 57, 65, presumably engraved before the argument, are also Charles Turner’s work, while a later, unpublished engraving of his was taken from a drawing of about 1818 – see Rawlinson/Finberg no.74 – and he subsequently worked for Turner again.)

The other engravers were W.T. Annis, George Clint, Robert Dunkarton, J.C. Easling, Thomas Hodgetts, Samuel William Reynolds the Elder and William Say.180 These Turner supervised very closely, as can be seen from the annotated and touched proofs recorded by Rawlinson and Finberg. Most of their names appeared, together with Turner’s own, on a banner in the 1812 Frontispiece to the Liber, issued with part 10 (see under Tate D08150; Vaughan Bequest CXVII V). The exceptions were Lewis, as Turner perhaps sought to play down his involvement after their disagreement, and Lupton, whose first plate appeared only in part 11.

The printing was generally by James Lahee,181 of Castle Street, off Oxford Street, London. Rawlinson claimed that ‘Turner was in the habit of constantly visiting Lahee’s printing-office to watch the results of his alterations, and the effects of new plates’,182 retouching the plate between impressions. Such incessant supervision seems impracticable, and Rawlinson’s note that the plates would be ‘sent home to him to be seriously treated’183 at nearby Harley Street (and subsequently Queen Anne Street West) seems more realistic.184

The first part appears to have been issued on 11 June 1807, published by the artist from his Harley Street address. The plates were not lettered with a date, which was first inferred by Finberg from an entry in the diary of the painter Joseph Farington, who had visited Turner’s exhibition on Saturday 13 June and seen ‘the first number of his publication of prints ... published on Thursday last.’185 This and subsequent parts comprised five plates stitched between blue paper covers with a standard, printed title on the front to which Turner added the number of the part and his initials in manuscript186 (in black or red, indicating respectively ordinary or proof copies);187 the text read:

The price of standard and proof sets was given beneath – initially fifteen shillings or a pound and five shillings, rising to one or two guineas respectively by part 11 in 1816, as indicated at that point in the only known printed advertisement for the series (for full text see introduction to catalogue section covering Rawlinson/Finberg nos.52–71). Turner subsequently continued to sell complete sets, not always comprising impressions of consistent quality, for fourteen and twenty-eight guineas respectively.189

The 1816 advertisement reaffirmed his intention to produce a series of a hundred plates, as first indicated by a prospectus displayed at the exhibition at his gallery in the spring of 1808, the wording of which – or part of it – has survived as a footnote to a review by John Landseer:

Landseer’s 1808 article is incidentally one of very few contemporary references to the launch of the Liber.191 David Cox recalled that Turner had hung a prospectus ‘on first opening an exhibition of his works, also a frame containing the drawings in sepia for the first proposed numbers’.192 Although Finberg initially related this passage to Turner’s 1807 exhibition193 and speculated that a version of the text had been displayed then,194 there is no indication of date in Cox’s account and the exhibition he saw may (as Finberg seems to have decided later)195 have been the one opening in April 1808. By this time two parts of the Liber had been published, with the third following in June. Finberg’s assertion, that ‘the sepia drawings for the five plates [Rawlinson/Finberg nos.2–6] in the first number of Liber Studiorum were included’196 in the 1807 exhibition, ignores the point that in terms of preliminary designs, Scene on the French Coast (no.4) is now known only by an untrimmed, tentative drawing (Tate D08104; Turner Bequest CXVI C), probably considered unsuitable for display, and by an impression of Turner’s outline etching with watercolour washes (Tate D08105; Turner Bequest CXVI D). Basle (no.5) is similarly known only as a washed etching (Tate D08110; Turner Bequest CXVI I). Nevertheless, since both the tinted etchings are neatly trimmed and superficially resemble drawings, it is not impossible that they might have been exhibited.

In Turner’s manuscript draft of an advertisement mainly concerning Charles Turner’s new role as the publisher of part 2 (issued in February 1808), he added: ‘[?50] drawings intended to be engraved may be seen on the 18 of April when Mr J M Turner’s Galery [sic] in Harley St will open for the Season’;197 but Landseer’s otherwise thorough review did not mention any drawings, and Cox’s recollection of ‘a frame’, if reliable and referring to the same event, suggests that only a few may have been shown. Though he acknowledged Landseer’s silence on this point,198 Finberg assumed that fifty drawings were indeed exhibited, and that they were the designs listed by Turner in his Liber Notes (2) sketchbook.199 The last of the sketches used as source material were dated by Finberg as being made in October 1807,200 and the list appears to have been written at one sitting, departing from the published order after part 2 and thus probably made between its publication in February 1808 and that of part 3 in June.201 The following transcript is based on direct examination, informed by published transcriptions:

Tate D12156; Turner Bequest CLIV (a) f.23a:202

D12157; f.24:203

D12158; f.24a:204

The publication of these designs continued until 1812, except for Tenth Plague of Egypt, not issued until 1816; London from Greenwich, relating to a painting of 1809, took its place (Rawlinson/Finberg no.26: see under Tate D08131; Turner Bequest CXVII D).

Doubt has been thrown on whether Turner would have exhibited all of these in 1808, particularly in the light of the lack of contemporary comment on such a substantial group of works.205 Finberg’s thesis is problematic: if Turner did represent all the subjects, he might have had to include the washed-in outline etchings for Rawlinson/Finberg nos.4, Scene on the French Coast, and 5, Basle (as discussed above in relation to a possible 1807 showing), while Liber drawings or tinted etchings are not recorded for nos.35, 40, 44 or 50. The surviving line drawings for the Frontispiece are presumably too provisional to have been exhibited, and its genesis is otherwise only marked by a washed etching (see entry for Tate D08150; Vaughan Bequest CXVII V) which may not yet have been in hand. Finally, ‘Windmill & Junction’, presumably the plate known as Windmill and Lock (no.27, published in 1811), for which no Liber drawing is known, only a washed-in etching, is related to a painting first exhibited in 1810; unless Turner already had it in mind as a definite subject in 1808, this suggests either that the painting existed at least two years prior to its exhibition, or that Finberg’s dating of Turner’s list is incorrect (see entry for Tate N02941).

That Turner at first intended to sell some or all of the drawings is perhaps indicated by most of them having been trimmed, probably in anticipation of exhibition,206 and also by the figure of £500 given as their value – whether for the first fifty or so or for the whole projected series is unclear – in two sets of financial calculations, of about 1809 or later and 1810. These are respectively on a page (of three concerning the Liber) in the Hastings sketchbook (Tate D07747; Turner Bequest CXI 96a – see also D07748, D07749; CXI 97, 97a),207 and a loose sheet kept with the Finance sketchbook (Tate D40900; verso of Turner Bequest CXXII (4)).208 When the Turner Bequest was first examined, it was found that some Liber drawings ‘had already been mounted before Turner’s death’209 (see discussion of early National Gallery displays, below). A few of the designs for published plates left Turner’s studio during his lifetime, some being retained by their engravers. As Forrester notes, the unengraved drawings (Rawlinson nos.92–99) were uniformly trimmed, possibly passing through the hands of Turner’s agent Thomas Griffith after the project had been abandoned, and ‘may have been purchased by collectors on the basis that they were related to the series.’210

A few Liber designs remain uncropped: Scene on the French Coast (Tate D08104; Turner Bequest CXVI C); Ploughing, Eton (Tate D08100; Turner Bequest CXV 47); and Windsor Castle from Salt Hill (Tate D08171; Vaughan Bequest CXVIII Q). The trimmed drawing for Devil’s Bridge, Mt St Gothard (Tate N03631), being squarer in proportion than other designs, retains a strip of landscape at the top which was omitted in the plate; and Mill near the Grand Chartreuse (Tate D08156; Turner Bequest CXVIII B) is proportionately larger than its engraving, as Turner worked to the full height of the Studies for Liber sketchbook page and decided to incorporate the whole design into the plate, rather than following his usual practice of working within a rough ad hoc margin and then trimming the drawing to approximately the same size as the engraved image.

Turner’s only subsequent Liber list, recording designs in hand,211 appears inside the back cover of the Aesacus and Hesperie sketchbook, watermarked 1817 (Tate D40933; Turner Bequest CLXIX). It comprises ten published and eleven unpublished subjects from later in the series, with three bracketed for part ‘13’, which was issued with part 14 as the last of the series in January 1819 (Rawlinson/Finberg nos.62–71):

The parts were initially issued at relatively regular intervals, with the first four (comprising Rawlinson/Finberg nos.2–21) appearing within two years (1807–9). After the quarrel with Charles Turner there was a gap of nearly two years; parts 5 to 10 (plate nos.22–51) then came out in just over two years (1811–12). After further delay, parts 11 to 14, the last to be published (plate nos.52–71), appeared between early 1816 and early 1819. The drawings for these three phases are considered as three corresponding groups in the present catalogue. Using Turner’s convoluted notes of costs, potential sales and profits from the prints (in the Hastings sketchbook and the sheet with the Finance sketchbook mentioned above), and taking into account known additional print runs ordered by Turner and the stocks which remained in his studio at his death – a rough average of about seventy per print – Finberg calculated that there were probably between 170 and 200 impressions of each published plate, figures with which Forrester broadly concurs.212

The status of the more expensive so-called proofs compared with supposedly standard impressions is one which has pre-occupied the collectors and scholars of the Liber.213 Collation of the various published states made it apparent that Turner did not draw a clear distinction when issuing sets of standard or proof impressions. Rawlinson and Finberg drew attention to the care with which Turner prolonged the life of the plates by re-engraving, sometimes yielding fresh nuances;214 presumably the drawings would have been kept to hand for comparison. The states were differentiated by additional strokes to the lettering, or marks in the margin; Forrester has transcribed Charles Stokes’s notes on the subject (Department of Prints and Drawings, British Museum, London, MM.5.12).215 Finberg doubted that the marks were evidence of duplicity,216 but others have assumed that Turner was attempting some sort of deception for financial gain. His early biographer, Thornbury, condemned the publication as ‘unbusinesslike, fitful, and peculiar’217 and criticised Turner’s lack of transparency,218 alleging that he ‘either from slovenliness or fraud, shuffled together the best states, so that no-one could get a perfect copy without buying several copies.’219

The prints in the last two parts of the Liber are dated 1 January 1819. It was later in this year that Turner first travelled to Italy on a journey which has often been regarded as the most significant watershed of his career; he nevertheless continued to work on further designs for the series. Finberg attributed its gradual abandonment over the next few years not to a lack of sufficient financial return – Turner himself had hoped to make only twenty pounds at best from each plate, and undertook it rather ‘to make his reputation safe with posterity’.220 Rather, the causes may have included a general change in the artist’s work (the use of a lighter, more colourful palette), and the introduction in the early 1820s of the steel-faced engraving plates developed by the Liber engraver Thomas Lupton221 through which many of Turner’s subsequent watercolours were interpreted. While working on the Liber, Turner continued to produce watercolours for other publishers,222 for engraving (like the Liber) on copper, such as Picturesque Views on the Southern Coast of England (published 1814–26).223 The later Picturesque Views in England and Wales (published 1827–38)224 also used copper plates. However, steel enabled his engravers ‘to lighten the key in which they worked, and to model the lights with greater subtlety and delicacy’,225 and the more durable surface facilitated longer print runs without the constant need for reworking. The Rivers of England and Ports of England (published 1823–7 and 1826–8 respectively)226 were mezzotints on steel by Lupton and other Liber engravers.

Turner’s so-called ‘Little Liber’ mezzotint plates, some engraved on copper and some on steel, ²²⁷ are also mentioned in both editions of Rawlinson’s Liber catalogue,228 though he refuted any direct link. Marcel-Etienne Dupret has discussed the twelve subjects in detail,229 making comparisons with the handling of the later, unpublished Liber designs.230 The ‘Little Liber’ episode probably dates from the early to mid-1820s, and shows the transition between the established use of mezzotint over an etched outline in the Liber towards purely tonal use of light and shade. Turner evidently found this creatively stimulating and is traditionally thought to have engraved the plates, even though the technique was stylistically incompatible with the completion of the Liber itself.

John Pye claimed that Turner’s insistence in being paid for each part of the Liber ‘in ready money’ and the refusal of a realistic trade discount to print dealers had not helped his sales.231 Forrester has shown that Turner’s prices compared unfavourably with various contemporary print part-works.232 Rawlinson concluded that ‘Liber Studiorum, in fact, had been, as far as public appreciation went, a failure, and therefore necessarily a pecuniary failure’,233 while Ruskin blamed the taste of the ‘tyrannous mob’ for its ‘total public neglect’.234 Overall, however, Finberg felt the technical experience provided by the Liber had served its purpose and laid useful foundations for Turner’s continued close involvement with engraving.235

The Liber studies’ supports are smooth, white or off-white wove paper, many sheets bearing a ‘J Whatman’ watermark or identified by Peter Bower as coming from Whatman batches.236 In relation to Gillian Forrester’s 1996 Liber exhibition and catalogue, the drawings’ media and techniques have been explored by Joyce Townsend, who has published a general article on them237 as well as compiling data on individual works.238

Although earlier writers had usually referred to the drawings as being in sepia239 (derived from the ‘ink’ of the cuttlefish), analysis has shown that Turner’s pigments are mostly inorganic, with burnt umber, burnt sienna and Indian red being the most common, although sepia itself is difficult to detect using infrared techniques.240 Another traditional assumption, that Turner used pen and ink for the linear elements of the design echoed in his subsequent outline etchings, has also been shown to be incorrect in most instances, since close examination has shown that a fine brush was used to add detail over the initial washes. This process served to define the compositions, which have only sporadic slight pencil outlines for the figures and landscape or architectural elements. Highlights were reserved, washed out, blotted or scratched out, sometimes possibly using etching needles and very occasionally a mezzotint rocker – tools at hand due to Turner’s direct involvement with the printmaking processes.241 (See individual entries for details of papers, media and techniques.)

Forrester has observed how Turner’s early work as a topographical draughtsman, with his combined use of undisguised pencil outline and tonal washes, ‘translated directly into the formula of etched line and mezzotint tone’ in the Liber.242 She also notes that his established practice of using a mid-toned wash to prepare sketchbook pages, to be worked with darker line or wash and with scratching-out or chalk for highlights, had ‘close analogies with the mezzotint process’. An example is the 1802 St Gothard and Mont Blanc sketchbook (Tate; Turner Bequest LXXV), from which several Liber compositions were developed.243

As working studies for prints, the drawings were not signed or dated. In order to present a sequence correlating with Rawlinson’s Liber print catalogues, Finberg gave a blanket date of circa 1806–10 to the two sections of his Turner Bequest Inventory covering the corresponding drawings for the first ten parts, despite the well-established publication of the first twenty prints by 1809.244 He dated a third section, covering the drawings for prints published between 1816 and 1819 and the unpublished or unengraved works, to circa 1810.245 Since then, various dates have been assigned to individual works when exhibited or discussed in the literature. The dating in the present entries generally follows Gillian Forrester’s, whose reasoning in linking various batches of drawings is summarised below.246

The project’s terminus a quo is generally accepted as October 1806, when the specific idea of the Liber was reportedly first mooted at Knockholt (see above). Clara Wells (Mrs Wheeler) recalled Turner completing the five drawings for the first part during this visit. However, on stylistic grounds it seems likely that the earliest, relatively tentative drawings were those for The Castle above the Meadows (Tate D08112; Turner Bequest CXVI K), The Junction of the Severn and the Wye (Tate D08132; Turner Bequest CXVII E) and Bridge and Goats (Tate D08146; Turner Bequest CXVII R),247 published in 1808, 1811 and 1812 respectively (Rawlinson/Finberg nos.8, 28, 43). For these and the other drawings, the absolute terminus ad quem is of course the publication date of the related print.

In 1924, Finberg tabulated the datable source material he had identified for the first fifty designs – effectively the first half or volume of the project, published at intervals until May 1812 – the last sketches being from late in 1807. This corroborated ‘the conclusion that all these drawings were made before April 1808’,248 around the time they were listed in the Liber Notes (2) sketchbook, and when he assumed they were exhibited at Turner’s gallery (as discussed above). On technical, stylistic and compositional grounds, some can be grouped and dated more specifically, for example nos.6, 7 and 12, while no.24 can be related to probably concurrent studies for later prints, nos.51 and 56.249 Some can be dated to early 1808, following an autumn 1807 excursion to Portsmouth, as nos.33, 37, 47 are derived a sketchbook used en route – and nos.22, 29, 32, 34, 45 and 49 are variously related to them in terms of style and materials.250

Finberg also listed the sketchbook material used as sources for later designs, noting that it ‘nearly all belongs to the same early period as that utilized for the first fifty drawings.’251 Turner was still using studies from the 1806–8 period for prints published in 1816 (for example nos.56, 57, 59 and 61), and even in the last published part, in 1819 (no.67).252 These were interspersed with looser, less finished designs, which, Forrester has suggested, ‘perhaps indicates that Turner’s enthusiasm for the project was declining’,253 such as nos.58, 64, 79, 85, 88.254 Occasionally a work’s watermark will indicate Turner’s continued involvement with the project long after the last published issues – see for example nos.82 and 86.255 Finberg’s concluded that almost all the drawings for the published plates dated from between 1806 and 1810: ‘This may account for the distinct individuality which the Liber Studiorum possesses. For these plates differ not only from the works of other artists, they also differ considerably from the designs which Turner produced before and after this publication.’256

For the unengraved designs numbered 92–101in the tables above, and the sheets which were originally leaves of the Studies for Liber sketchbook, watermarked 1807 (Tate; Turner Bequest CXV), a date of circa 1807–19, the span of the Liber’s active publication, has been adopted unless there is specific individual evidence.

The importance attached to the Liber drawings following the legal settlement of the Turner Bequest is indicated by the display of twenty-four of them in February 1857 at Marlborough House in London, among an initial total of 102 watercolours selected by a National Gallery committee; some had been found already mounted, and the others were presented in a uniform style, in gilt frames.257 However, Ruskin was quick to dismiss them:

Once Ruskin had made his more extensive selection of Turner Bequest works on paper for display, he was criticised in the press for not including notes on the Liber drawings in his 1857 catalogue.259 He reiterated his view of them as not ‘in a true sense, drawings at all’ but ‘merely washes of colour, laid roughly to guide the mezzotint engraver’ with ‘the drawing, properly so-called’ embodied in Turner’s subsequent etched outlines: ‘These brown “guides” – for they are nothing more – are entirely unlike the painter’s usual work, and in every way inferior to it’. He had even ‘put good impressions of two of the plates in the same room, in order to show their superiority’.260 Walter Armstrong, for one, later felt Ruskin’s case was ‘greatly overstated ... It would give an entirely wrong impression to anyone who had never seen the drawings. A few of them ... are delightful little works on their own account, showing both great spontaneity and an immediate solicitude. They are, in short, better than the corresponding plates. But there can be no doubt that, as a rule, Turner was thinking chiefly of the final purpose as he made the drawings.’261

Many of the works on paper were transferred to the South Kensington (Victoria and Albert) Museum in 1859.262 Fifty-one Liber drawings – those for published designs, together with the so-called Pastoral (Tate D08184; Turner Bequest CXVIII d)263 – were listed by Thornbury as on display there soon afterwards.264 Despite Ruskin’s opinion, two volumes of photographs were issued on official authority in 1861 and 1862, reproducing all fifty-one on display, while compositions represented in the Bequest by etchings worked over with watercolour were excluded.265 The mounted ‘sepia’ plates were approximately the same size as the brown watercolour originals, and in some cases show subtleties of tone and detail which the originals have lost; they apparently constitute the only publication comprising actual photographic prints of works in the Bequest, and are among the earliest books to reproduce art photographically. The Arundel Society, usually concerned with the reproduction of works by the Old Masters, issued a second edition in 1877.266

The works were irreversibly damaged by being displayed ‘every day, and all day long’267 at this time. Since most of them had been trimmed to the edge of the composition (probably by Turner), the fading, though bad, was even and perhaps not obvious at first sight. However, some of the sketchier designs were, in common with sheets from other parts of the Bequest, partly masked when mounted and the fading of pigments and darkening of the exposed paper is clear – see for example Tate D08101 and D08231 (Turner Bequest CXV 48, CXX Q).

By 1878, they were back at the National Gallery, still on show and ‘becoming utterly faded and in other ways deteriorated by constant exposure’;268 they were apparently soon withdrawn from permanent display269 but were described in 1885 as ‘the ghosts of what they were’.270 When the new Turner wing opened at the National Gallery of British Art (now Tate Britain) in 1910, a room on the main floor was devoted to the Liber drawings, with other watercolours in a showcase.271 Subsequent exhibitions are briefly discussed below.

John Ruskin’s extensive writings on the Liber affected many subsequent authors, although it should be remembered that his remarks on Turner’s imagery and draughtsmanship were based on the published plates – as explained above, he regarded the preliminary drawings catalogued here as almost worthless. As a watercolourist, he first closely studied the series in practical terms before his 1845 visit to Italy, ‘mastered its principles, practised its method, and ... was able to study from nature in full chiaroscuro, with a good frank power over the sepia tinting’.272 Drawings such as his View of Bologna (Tate N03507) show the Liber’s influence, though he apparently did not own a set at this stage,273 and while he was away his father ordered one.274 Instead of supplying existing impressions Turner arranged for fifteen additional sets to be printed.275

As Alan Davis and Ian Warrell have noted,276 it was only in the revised 1846 edition of the first volume of Modern Painters (his extended defence of Turner, first published in 1843) that Ruskin introduced passages on the Liber, commenting on the artist’s concentration on homely British subjects at the expense of grander Continental ones, while condemning the more classical subjects as too reliant on Claude and lacking in originality.277 In the last volume (1860), by which time he had been responsible for sorting through all the drawings in the Turner Bequest and his view of Turner had darkened,278 he developed a much-quoted and influential overview of the series as an expression of the artist’s pessimistic outlook on existence279 (see entries on individual works for specific comments from this passage). He concluded:

Some Victorian writers, their attitudes coloured by Ruskin’s, found the supposedly doom-laden message of the Liber rather heavy-handed and obvious: ‘We all know that strong castle and fair fame will alike fall ultimately into ruin, that death and decay are everywhere, and that bones will ever be found bleaching on the mountains.’281 But in his catalogue, Rawlinson took issue with some of Ruskin’s interpretations of the Liber’s scenes of everyday urban and rural life, and pointed out that ruins can be beautiful as well as indications of failure.282 From its alternating comic and tragic elements, with ‘Nature in her sunniest and Nature in her sternest moods’, he inferred that Turner intended the Liber as ‘a complete, not a one-intentioned work’ to be kept together in all its variety rather than as an expression of consistent pessimism, quoting a print dealer: ‘Mr. Halsted tells me that Turner, once coming to his shop in Bond Street, found fault with him for breaking up sets of the Liber; and when he heard that some plates sold habitually much better than others, he grunted out “A pack of geese! a pack of geese! Don’t they know what Liber Studiorum means?”’283 Though acknowledging the power of Ruskin’s writing, Finberg declared: ‘Turner’s sense of beauty is more perfect than Ruskin’s, his feelings deeper. There is much more in Turner than even Ruskin could discover.’284

Ruskin speculated on Turner’s motivation for producing the Liber in relation to Claude:

He went on to make a specific, if inequitable, comparison of a group of Claude’s sketchy trees as engraved by Earlom with part of a Liber mezzotint, using two enlarged lecture drawings.286 Elsewhere he criticised what he saw as Claude’s adverse influence on ‘the worst and feeblest’ Liber compositions, preferring those where he perceived the effect of the Venetian School, and Titian in particular.287

Ruskin also used the Liber as a source of practical examples in his teaching on landscape composition, drawing and chiaroscuro.288 In The Elements of Drawing (1857), he recommended a process of tracing and copying the prints, first following the etched outlines and then building up to exact emulation in watercolour of the prints’ mezzotint tones.289 (He suggested that if Liber prints were not available, photographs of landscapes could be followed in the same way, as though the prints had anticipated photography in the precision of their shading.)290 Typically, he provided lists of the ‘most desirable’ and merely ‘serviceable’ compositions; others, predictably from the Claude-like ‘EP’ category, were ‘quite useless’,.291 In a later edition, he noted that he had provided William Ward of the Working Men’s College, London, where Ruskin taught art, with photographs of the etched Liber outlines from his own collection as ‘the best lessons in pen-drawing’.292 He praised the thematic and emotional range of the Liber on several occasions,293 and when he lectured at Oxford University in 1871, he declared:

He concluded: ‘These are not finished pictures, but studies; endeavours, that is to say, to get the utmost result possible with the simplest means; they are essentially thoughtful, and have each their fixed purpose, to which everything else is sacrificed; and that purpose is always imaginative – to get at the heart of the thing, not at its outside.’295 Ruskin represented the Liber by prints in the Educational Series he arranged for Oxford in the 1870s, his catalogues providing a combination of technical notes and commentaries on the compositions.296 Despite his reservations concerning the preparatory drawings, he also included photographs of two privately-owned examples (see under Tate D08168 and D08176; Vaughan Bequest CXVIII N, V). In 1878 two actual Turner Bequest Liber studies were lent to Oxford as part of the National Gallery’s Turner Bequest loan collection, with which Ruskin was involved (Tate D08166, D08185; Turner Bequest CXVIII L, e).

Ruskin also played a part in the appearance of two series of prints reproducing or interpreting selected plates from the Liber after Turner’s death, by Thomas Lupton and Frank Short, of which Short’s are the better known. References will be found in the checklists in the introductions for each section, and in the entries for the relevant drawings. A mezzotint engraver, Lupton had worked for Turner on several of the original Liber plates and on later projects. Between 1858 and 1865 he re-engraved the designs listed below on steel plates for the London dealers Colnaghi; the projected series of up to thirty-six (to have been dedicated to Ruskin, who lent Lupton impressions of Turner’s prints), remained unpublished.297 Rawlinson provides lists of Lupton’s engravings in both editions of his catalogue;298 they are also noted individually in Finberg’s. Tate does not hold any impressions of these prints, which Ruskin described as ‘good and serviceable’:299

Frank Short (1857–1945) was a distinguished etcher in his own right, having taught himself through emulation of Turner’s Liber.300 He was encouraged by Ruskin and other Liber collectors and scholars, and Martin Hardie devoted the whole of the first volume (of three) of his catalogue raisonné of Short’s prints to his forty-seven Liber plates.301 These included interpretations of the drawings identified by Rawlinson as unengraved Liber designs (Rawlinson nos.92–100). As noted above, Finberg excluded these drawings from his catalogue and regarded it as ‘unfortunate’ that Short’s engravings from them, ‘so fine, and so individual’ as they were, had come to be regarded as ‘a part, a continuation, and even a completion’ of the series.302 Tate holds impressions of many of them.303

Short was also involved306 in the compilation of the substantial 1890 folio volume, A Selection from the Liber Studiorum,307 in the official ‘South Kensington Drawing-Book’ educational series, which ranged from elementary studies for young children to examples from Michelangelo and Raphael.308 Four photogravures from published plates were retouched by Short, who contributed a chapter on copying the compositions and technical notes on the printmaking processes,309 and thirty of the etched outlines were also reproduced for copying. The editor, John Ward, fully acknowledged the influence of ‘[o]ur greatest Art-writer, Mr. Ruskin’ in viewing the Liber as ‘the best school of Landscape Art.’310 Only one of Turner’s drawings was reproduced, on a small scale (Tate D08171; Vaughan Bequest CXVIII Q), despite Short’s view (somewhat counter-productive in this ‘utilitarian rather than aesthetic’311 context) that ‘there are qualities in a sketch which a finished work can never have.’312

In terms of the published prints, the Liber Studiorum has been one of the most extensively documented aspects of Turner’s career. The combination of its wide but finite scope, its various subjects and approaches, and the high quality of the best impressions, counterbalanced by an idiosyncratic production history and the puzzles surrounding the significance of the various states of each plate and the upublished designs, made the series an object of fascination for nineteenth-century collectors and connoisseurs, and the subject of a number of exhaustive studies.

As noted, photographs of many of the Liber drawings in the Bequest were published in 1861–2 and again in 1877. Henry Vaughan commissioned similar photographs of two (and probably three, if not more) of the drawings in his collection, at least one of which was published in 1862 (see under Tate D08168, D08173 and D08176; Vaughan Bequest CXVIII N, S, V). An Australian amateur, James Macdonald Larnach, published the first study of the series in 1865.313 There was a steady trade not only in the prints, but also in secondary reproductions of them. In 1854 the lithographers Day and Son published a set of fifteen plates,314 which were followed in successive decades by various large-format editions of the prints by photographic engraving processes, including a ‘very indifferent’315 three-volume folio set by the Autotype Fine Art Company in 1871.316

In 1872 J.E. Taylor and Henry Vaughan317 arranged an exhibition of the entire series, represented by prints and a few drawings from private collections; they numbered the compositions in the catalogue order which has been followed ever since (crucially by Rawlinson and Finberg in their respective Liber catalogues raisonnés), with the Frontispiece as no.1, followed by plates in the successive published parts (nos.2–71).318 Their numbering of the plates within each part was essentially arbitrary, as Turner did not specify a sequence. A further twenty compositions which had been engraved to various states of completion but not published were included, following the titles and sequence originally defined by Turner’s friend Charles Stokes in his unpublished notes.319

The cornerstone of modern Liber studies is W.G. Rawlinson’s catalogue (1878; revised 1906). Meanwhile, extensive ‘Notes and Memoranda’ gathered by Turner’s friend and engraver John Pye320 were published in 1879 in a rather rambling form, described by Marcia Pointon as ‘contradictory and ruptured ... because it addresses the conflicting concepts of the aesthetic and the commercial’321 but containing much first-hand information from Turner’s Liber engravers and other contemporaries.322 Rawlinson included several likely drawings as candidates for the completion of the series, reaching a high point of one hundred including the frontispiece; the anonymous 1911 Miniature Edition of Liber reproductions tentatively offered six further Turner Bequest drawings, each captioned ‘101??’, seemingly at Rawlinson’s suggestion.323 These alternative designs made up the full projected total including the Frontispiece.

The three-volume second edition of Autotype Company reproductions appeared in the early 1880s with notes by the Rev. Stopford Brooke.324 Brooke’s notes were revised and re-issued in a smaller single-volume format in 1885, and comprise the only sustained plate-by-plate commentary on the published designs to concentrate on the compositions as such. Brooke was a London-based clergyman, writer and teacher whose main specialism was English poetry;325 he freely admitted his debt to Ruskin, to the extent of usefully gathering and reprinting the latter’s writings on the Liber at the end of his own.326 He also followed Ruskin in regarding the original drawings as ‘almost in every case, and naturally so, inferior to the prints’.327 Digressions and unfounded biographical speculation notwithstanding, he subjected the published compositions to minute scrutiny; since his observations are often useful in relation to the drawings, the 1885 edition has been cited in the present catalogue. (A digestible selection of his comments was published in 1994.328) The equally Ruskinian South Kensington Drawing Book devoted to the Liber followed in 1890, as described above.

In 1924, A.J. Finberg’s large, illustrated history and catalogue of the published plates followed the most comprehensive exhibition of the drawings, together with their associated prints and copper plates, at the Tate Gallery and in Manchester between 1921 and 1923,329 by which time some thirty books and exhibition catalogues had focused on the Liber, though usually in terms of the prints alone.330 Finberg had earlier published the Inventory of the Turner Bequest and drew on his detailed knowledge of Turner’s sketches and notes to make the fundamental links from these sources to the Liber designs, seeking to redress the Victorian view of the series as ‘something almost superhuman’.331 As had Rawlinson, Finberg concentrated on establishing the history of the series and the complex sequence of proofs and published states, only mentioning the original drawings in passing and largely ignoring the meaning of particular designs.

More recently, following his extensively illustrated compendiums of drawings in the Turner Bequest,332 Gerald Wilkinson’s book on the Liber included attempts to decode the significance of Turner’s selection for each five-print part in terms of unifying themes, both formal and theoretical, though ultimately Turner’s system – or lack of it – proved impenetrable.333 A 1996 exhibition at the Tate Gallery was accompanied by Gillian Forrester’s detailed catalogue, providing the first systematic review of the series and its history since Finberg’s. It incorporated a technical survey of the drawings (although few were exhibited), and addressed the significance of the series and each design in relation to modern art-historical issues. Drawing on research for Tate’s on-line catalogue, the present author contributed commentaries on all the published plates to the catalogue for the exhibition at Waiblingen in 2008, and this source is cited in each entry; some of the watercolour designs were also exhibited.

With an upturn in interest in the Liber during the last decade of Turner’s life, prints began to appear at auction;334 more appeared on the market soon after his death, from the collections of Charles Turner and the artist’s agent Thomas Griffith,335 and prices rose steadily.336 Many sets and single engravings from the Liber were auctioned in 1873, in the first of a series of six sales of prints from Turner’s estate by order of the Court of Chancery (1873–4), comprising the many impressions found in his studio.337 The copper plates for the published plates were supposed to be ‘destroyed’ before the sale,338 but the forty-nine for Rawlinson/Finberg nos.2–50 were resoldered339 and donated by C.W.M. Turner, a collateral descendant, to the British Museum, London (1945–12–8–321 to 367).340 Eleven plates for unpublished prints were also in the sale,341 including the one for Flounder Fishing catalogued here (Tate N02782), and the locations of some of the others are known.342 Impressions were taken from them before the sale or by subsequent owners.

Several of Turner’s contemporaries – including the Liber’s printer, James Lahee343 – and subsequent scholars formed important collections of Liber prints, including many touched and annotated proofs.344 These were the foundations of museum collections such as those of the British Museum, the Royal Academy, London (A.A. Allen collection),345 and the Museum of Fine Arts, Boston (Bullard Bequest, including prints from W.G. Rawlinson’s collection).346 Through the National Art-Collections Fund, A.A. Allen presented a representative set of etchings and published mezzotint states to the Tate Gallery in 1925 (Tate A00911–A01015; A01112–A01149; A01188), together with later printings or facsimile photographs of the unpublished plates (A01150–A01159).