The National Gallery of Victoria is known to Blake scholars world-wide for its unrivalled collection of Blake’s illustrations to Dante’s Divine Comedy. The sublimity and significance of these water-colours have tended to obscure the fact that the Gallery houses other Blakean treasures. Fortunately, the veil has been lifted by the fully-illustrated catalogue of the Blake collection published by the Gallery in 1989.1 Martin Butlin & Ted Gott, William Blake in the Collection of the National Gallery of Victoria, National Gallery of Victoria, Melbourne, 1989. Although the Gallery’s impressions of plates from Blake’s illuminated books have been listed in the standard literature for some years, I suspect that few of us knew of their beauty. All were acquired in 1918 from the estate of John Linnell, who met Blake in 1818 and became his chief patron. The group consists of colour-printed impressions of plate 8 from Europe: a prophecy (1794) and plate 21 from The Book of Urizen (1794), and a hand-coloured impression of Jerusalem, plate 51.2 Copy designations, plate numbers, dating and other basic facts about the illuminated books are taken from G. E. Bentley Jr, Blake Books, Clarendon Press, Oxford, 1977 For a study of the special properties of the Urizen plate, see G. E. Bentley Jr, ‘The Shadow of Los: Embossing in Blake’s “Book of Urizen”’, Art Bulletin of Victoria, no. 30, 1989, pp. 18–23. My perusal of the colour reproduction of this last work in the Gallery’s new catalogue sparked my interest in the print and led to the theory about its production and history presented here.
Jerusalem is Blake’s longest illuminated book and is the culmination of his career as an epic poet and innovative printmaker in the media of relief and white-line etching.3 For descriptions of these processes and their stylistic evolution in Blake’s work, see Robert N. Essick, William Blake, Printmaker, Princeton University Press, Princeton, NJ, 1980, and Joseph Viscomi, The Art of William Blake’s Illuminated Prints, Manchester Etching Workshop, Manchester, 1983. The title-page bears the date ‘1804’, but all 100 plates may not have been completed until as late as 1820. None of the seven extant complete copies of the book have watermarks earlier than 1818, and every copy has some leaves with an 1820 watermark. Among the complete copies printed by Blake himself (three were produced posthumously), one stands above the rest as his chef-d’oeuvre in illuminated printing. This copy is designated as ‘E’ in the standard catalogues. Although Blake hand-coloured a copy comprised of the first twenty-five plates only, plus several extant individual leaves, Ε is the only complete and fully coloured copy of Jerusalem. In a letter to George Cumberland of 12 April 1827, Blake is no doubt referring to this copy when he states that he has ‘Finishd’ (that is, coloured) only one and that ‘it is not likely’ he ‘shall get a Customer for it’.4 David V Erdman (ed.), The Complete Poetry and Prose of William Blake, University of California Press, Berkeley, 1982, p. 784. Unfortunately, Blake’s prophecy proved all too accurate, and the volume passed to Frederick Tatham after Blake’s death. In 1951 the Blake Trust published an accurate facsimile, produced by the Trianon Press of Paris through a combination of collotype and hand-colouring through multiple stencils. The original volume has been in the possession of the American book and art collector Paul Mellon since 1952.
Each of the four chapters of Jerusalem ends with a full-page design. Three of these chapter conclusions – plates 26, 51 (fig. 1), and 100 – call particular attention to themselves because they must be turned sideways, with their horizontal major axis parallel to the book’s spine. Although linked in style, motif and theme to surrounding images and to the text, they could be printed as independent designs. Blake realised this potential in plate 51 by producing a separate print of it, now in the Keynes collection, Fitzwilliam Museum, Cambridge. For this impression Blake scratched the names of the pictured characters into the copper in white line (from left to right, ‘Vala’, ‘Hyle’ and ‘Skofeld’), printed the plate in blue-green, and hand-tinted it in sombre hues. Although the National Gallery’s impression of this same plate (fig. 2) lacks the inscribed names, it might be taken for another richly-coloured separate impression. But one feature should give us pause: the print is numbered ‘51’ in Blake’s hand lower right, a position that would place it in the conventional top right corner if the design were bound vertically along its top edge. Why would Blake number an independent work at all, and particularly why would he place this number in a position appropriate only if bound in a book? Comparison with plate 51 in copy Ε (fig. 3) will provide an answer.
The colouring of the two prints in question (figs 2–3) shows some basic similarities in palette not shared by the Keynes impression. There are of course some significant differences: the print in copy Ε is darker in overall tone, and its black and red foreground contrasts with the more naturalistic brown and yellow-green in the Gallery’s version. Yet the emotional expression and general conception directing the colour schemes of these two prints make them compatible with each other and with the colouring of the other plates in copy E. Surprisingly, the green foreground in the Gallery’s example finds numerous similarly tinted companions in copy Ε (for example, plates 9, 15, 19, 25, 35, 39, 41, 52, 64, 70, 85, 94, 97 and 100), whereas no other plate repeats the black and dark red landscape of the volume’s plate 51. Further, both examples are highlighted with liquefied gold, a medium otherwise unique to copy E. Other details also link the two impressions.
Both have framing lines and number drawn by hand in orange. No other copy of Jerusalem has these features in this colour. Like twenty-four other leaves in copy E, its plate 51 shows a J Whatman watermark with the date of 1820. The Gallery’s impression would appear to be on the same paper and shows the same 1820 date near the top edge of the sheet, with the manufacturer’s name apparently trimmed off.5 Bentley, Blake Books, p. 227, does not record this watermark, first noticed by Ted Gott; see Butlin & Gott, p. 184, no. 47.
Blake consistently encouraged the inherent tendency of his relief and white-line media to proliferate differences, particularly among impressions produced years apart. The similarities between plate 51 in copy Ε and in the Gallery’s collection bespeak a similar date of colouring and even suggest that they were created for similar purposes. However, it is not the similarities, but several key differences, that point most clearly to what I believe to be the true nature of the relationship between these prints. Plate 51 in copy Ε is heavily printed in black ink, much like one of Blake’s monochrome impressions (fig. 1) not intended for subsequent colouring. The Gallery’s impression is lightly printed in orange. The differences between the density and colour of the under-printing account in large part for the darker tonality of the copy Ε impression. Plate 51 is the only one in copy Ε printed in black; all others are in the same orange ink used for the Gallery’s print. The framing line and number in the latter are in a lighter, less reddish, orange than are these same features added to the impression in copy E. Like the ink and colouring of the foreground, the light orange lines and number make the Gallery’s print more compatible with other leaves in copy Ε than the impression of plate 51 presently bound in the volume.
The compound of similarities and differences between the two prints at issue strongly suggests that the Gallery’s example was originally produced for inclusion in copy Ε and that the impression now part of that volume is a substitute. This theory in turn can lead to some speculations about how this came about. From 1818 to Blake’s death in 1827 he was in regular contact with Linnell. Certainly Blake would have shown his unique coloured copy of Jerusalem to his friend and patron. Linnell had considerable respect for Blake’s abilities as a pictorial artist, but at best a limited appreciation for his writings. Thus Linnell was an unlikely customer for the complete book, but he may very well have admired some of its larger and more impressive designs. Blake may have removed plate 51– the Gallery’s impression – from the as yet unbound, perhaps even incomplete, book and sold or given the print to Linnell.6 The Gallery’s impression shows no evidence of ever having been bound. The pin-holes in three corners (the tip of the lower left corner is missing) probably resulted from an earlier mounting of the separate sheet. Blake apparently had no other impression in orange ink to use as a replacement. Rather than going to the trouble of pulling another impression in orange, Blake selected a print on the same paper already on hand, even though it was in black, and tinted it according to the general colour scheme established by other plates in copy E. This substitute, like the impression in copy C (fig. 1), may have been printed before Blake added his monogram to the copperplate in white line, clearly visible lower left in the Gallery’s impression. But the hand-colouring of the replacement print covers this area in dense black, and thus the absence or presence of the monogram cannot be determined. During the final stage of producing the second version, Blake made one slight change in conformity with the format of the book. The original print was numbered in a location indicating that it should be bound along its top margin, whereas the substitute, like the three other sideways plates in copy E, was numbered for binding along its bottom margin. Blake probably believed that this substitution would not disrupt the aesthetic unity of his book. In this he was right, for even the simple fact that plate 51 in copy Ε was printed in black has not previously been noticed.
If my substitution theory is correct, we are presented with more evidence about Blake’s attitudes towards his bibliographic and chalcographic endeavours, attitudes that seem more than a little cavalier when judged by standards implicit in conventional book production. His works are marked by continual revision, even in the final stages of their manufacture. When, for example, he was unable to repair a crooked impression of plate 4 of The Book of Urizen for inclusion in the final copy (G) he is known to have printed, he did not even bother to provide a substitute and simply left this important plate out of the volume.7 For a discussion of this peculiarity, see Robert N. Essick, ‘Variation, Accident, and Intention in William Blake’s The Book of Urizen’, Studies in Bibliography, vol. 39, 1986, pp. 230–5. Jerusalem also bears witness to its multiple revisions. Blake scored through several lines of text on copperplates 3 and 4, but made no attempt to disguise these cancellations by masking the awkward fragments of the erased words during printing. Only copy Ε shows any attempt to fill these areas with interlinear decorations added by hand. The text pages of chapter II appear in one sequence in copies A, C and F, but in a completely different order in copies D and E. In copy F the page numbers printed from the copperplates of chapter II contradict the numbers Blake added with pen and ink. Elsewhere in all copies, the arrangement of the plates does not consistently accord with the etched catchwords, presumably executed as part of some earlier conception of the book. Jerusalem seems to have been a work in perpetual evolution without ever reaching a final form in Blake’s lifetime. Given the ever-various nature of the book, the removal of a finished, full-page design and the substitution of another version seem almost traditional. We should, however, be grateful that Blake took such liberties. The result in this instance was two dramatically coloured impressions of plate 51, whereas an unwillingness to make changes would have left the world with only one.
Robert Ν. Essick, University of California, Riverside (in 1990).
I am grateful to Ted Gott and Irena Zdanowicz of the National Gallery of Victoria for their
assistance with this essay. Morton D. Paley kindly confirmed for me the accuracy of the Blake
Trust facsimile in its representation of the ink colour in Jerusalem, copy E, plate 51, reproduced
here with the permission of The William Blake Trust.
1 Martin Butlin & Ted Gott, William Blake in the Collection of the National Gallery of Victoria, National Gallery of Victoria, Melbourne, 1989.
2 Copy designations, plate numbers, dating and other basic facts about the illuminated books are taken from G. E. Bentley Jr, Blake Books, Clarendon Press, Oxford, 1977 For a study of the special properties of the Urizen plate, see G. E. Bentley Jr, ‘The Shadow of Los: Embossing in Blake’s “Book of Urizen”’, Art Bulletin of Victoria, no. 30, 1989, pp. 18–23.
3 For descriptions of these processes and their stylistic evolution in Blake’s work, see Robert N. Essick, William Blake, Printmaker, Princeton University Press, Princeton, NJ, 1980, and Joseph Viscomi, The Art of William Blake’s Illuminated Prints, Manchester Etching Workshop, Manchester, 1983.
4 David V Erdman (ed.), The Complete Poetry and Prose of William Blake, University of California Press, Berkeley, 1982, p. 784.
5 Bentley, Blake Books, p. 227, does not record this watermark, first noticed by Ted Gott; see Butlin & Gott, p. 184, no. 47.
6 The Gallery’s impression shows no evidence of ever having been bound. The pin-holes in three corners (the tip of the lower left corner is missing) probably resulted from an earlier mounting of the separate sheet.
7 For a discussion of this peculiarity, see Robert N. Essick, ‘Variation, Accident, and Intention in William Blake’s The Book of Urizen’, Studies in Bibliography, vol. 39, 1986, pp. 230–5.