Subscription log in
New Information about William Blake’s Pre-Publication Proofs of His Job Engravings

New Information about William Blake’s
Pre-Publication Proofs of His Job Engravings

Mei-Ying Sung (maysung70@hotmail.com) is the author of William Blake and the Art of Engraving (Pickering & Chatto, 2009). She is currently taking part in cataloguing and researching the printing woodblock collection at the Huntington Library, California, USA, and teaching in the History Department, FoGuang University, Taiwan.

This essay examines a unique set of impressions of William Blake’s Illustrations of the Book of Job in the Beinecke Library, Yale University. The body of evidence it presents means that our understanding of the stages of production of this series of engravings now needs to be reconsidered. The Beinecke set, hereafter known as the Rosenbloom set, reveals the meticulous detail in which an artist and craftsman worked. The essay offers previously unrecorded information about the early stage of Blake’s engraving of the Job illustrations that will augment our knowledge of his working methods. It aims to record the details of the Rosenbloom proofs and revise the current sequence of pre-publication proofs set up by Robert N. Essick. It will be necessary to summarize Essick’s important work in order to give the full context for the development of the Job plates.The research for this paper was supported by a grant from the Paul Mellon Centre for Studies in British Art, Yale University (2003) and the BSA-ASECS Fellowship for Bibliographical Studies in the Eighteenth Century from the Bibliographical Society of America (BSA) in conjunction with the American Society for Eighteenth-Century Studies (ASECS) (2008). I am grateful to Robert N. Essick for the loan of photographs of many pre-publication Job proofs. I would also like to thank both Essick and David Worrall for their suggestions, as well as Sarah Jones for her careful checking and editing work.

William Blake engraved twenty-two plates for Illustrations of the Book of Job in the last few years of his life and published them in 1826. The work arguably represents the best combination and manifestation of his lifelong training and practice as an engraver and, not least, his efforts to be recognized as an original designer. The engraving and printing processes of Blake’s Job have been discussed extensively and thoroughly, most notably by Essick in comparing and cataloguing in great detail the many pre-publication proof states.Robert N. Essick, “Blake’s Engravings to the Book of Job.” However, I examined a further set of highly significant pre-publication proofs in the Beinecke Library during my trip to see the copperplate of “Chaucers Canterbury Pilgrims” in the Yale University Art Gallery in 2003. The details of this set have not been recorded until now. The images can be viewed in the Beinecke Digital Collections database: <http://​brbl-dl.​library.​yale.​edu/​vufind/​Record/​3432519>.

This run of proofs necessitates a recataloguing of the engravings’ proof states. While print “impressions” refer to the number of pulls taken from a plate at the press, the “states” of a print refer to the changes made by the etcher or engraver. Normally, a printmaker would take impressions of each state in order to inspect how the plate had printed and to determine what further changes, if any, needed to be made. Since these changes are usually (but not always) additions to the image or its accompanying text, scholars of printmaking can use this information to work out the sequence and number of the states. Of course, the number and sequence are not apparent by examination of the final print, or even by examination of the copperplate (if it is extant), so the study of the states reveals otherwise invisible information about the artistic process.

Although Blake’s Job was frequently reproduced in the twentieth century, most studies concentrated on the development and meaning of the iconography, with Bo Lindberg’s William Blake’s Illustrations to the Book of Job (1973) perhaps the best example of this trend. These studies were supplemented by David Bindman’s 1987 Blake Trust edition, with its elegant catalogue raisonné by Essick. However, the previously unexamined Beinecke proofs, particularly as they record so many early states of the Job impressions, add significantly to our knowledge of the development process. Not least, the new states further extend our notion of the number of states that were needed and proofed by Blake in order to produce a finished engraving. In addition, the existence of these proofs corroborates my findings, outlined in William Blake and the Art of Engraving, about the hammer marks for revision on the versos of the Job copperplates, which are now in the British Museum.The hammer marks found on the versos of Blake’s engraved copperplates are traces of the repoussage technique. As described in handbooks of printmaking, repoussage is used to correct mistakes or make changes on
line-engraving copperplates by scraping and burnishing the lines and hammering up the area from the back of the plate to make it into an even surface for further engraving.

In the Blake Trust edition, Essick lists nine sets of pre-publication proofs—British Museum (two sets: Evans, Tatham), Fitzwilliam Museum (two sets: Keynes, Riches), Houghton Library, Harvard University (one set: Hofer), National Gallery of Art, Washington, DC (two sets: Linnell, White), and University of Texas (two sets: El Dieff, Hanley)—plus some odd proofs in other collections.Pl. 1, last recorded in the collection of Douglas Cleverdon, who is now deceased (the whereabouts of the plate is not known); pls. 1, 2, and 11 in Essick’s collection; pls. 17 and 19 at McGill University, Montreal; pl. 11 at the Morgan Library and Museum, New York; and pl. 13 at the Philadelphia Museum of Art (Essick, “Blake’s Engravings to the Book of Job” 54). He adds that there is “one highly significant group of untraced pre-publication proofs of the Job engravings” (54), recorded in an auction catalogue of the American Art Association/Anderson Galleries of New York, 9 December 1936, lot 62. At the time of writing he was not able to locate the set, but he realized its importance: “If indeed these are first pulls after the addition of borders, they record a crucial point in Blake’s development of the Job copperplates, perhaps the first time he could determine fully the success or failure of his combination of outline borders and highly-finished central designs. The rediscovery of these impressions could occasion some major adjustments in the known record of progress proofs” (Essick, “Some Unrecorded Proofs” 96). My study therefore bridges this gap and adjusts Essick’s otherwise authoritative catalogue raisonné of pre-publication Job proofs.

The 1936 sale catalogue has a 4¼-page description of this set of proofs, particularly concentrating on its differences from the finished state, which exactly match the features of the rediscovered set in the Beinecke Library. Especially important are plates 1 and 21 (both reproduced in the auction catalogue and also in Essick’s two essays), which have several textual variants on the border design. Plate 1 lacks the first two lines on the altar in the lower border (“The Letter Killeth / The Spirit giveth Life”), but under the altar are the extra words “Prayer to God is the Study of Imaginative Art”, which are not in any other impression. In the same place on plate 21 is another inscription, “Praise to God is the Exercise of Imaginative Art”. The description in the auction catalogue is not sufficient, however, to decide the sequential state of these proofs. Unable to locate the set, Essick could not record this early state in his 1987 catalogue, but worked only from the vague evidence of the 1936 catalogue.Essick assumes that pl. 1 is between his states 4 and 5, pl. 2 between 1 and 3, pl. 3 between 1 and 3, pl. 4 between 1 and 3, pl. 5 between 1 and 2, pl. 6 between 1 and 4, pl. 7 between 1 and 2, pl. 8 between 1 and 3, pl. 9 between 1 and 2, pl. 10 between 3 and 6, pl. 11 between 1 and 2, pl. 12 between 1 and 2, pl. 13 between 1 and 2, pl. 14 between 3 and 5, pl. 15 before state 2, pl. 16 between 2 and 3, pl. 17 between 1 and 3, pl. 18 prior to state 1, pl. 19 between 1 and 4, pl. 20 between 1 and 2, and pl. 21 between 1 and 2 (“Blake’s Engravings to the Book of Job” 56). All these assumptions match my conclusions from the examination of the proofs. G. E. Bentley, Jr., in his Blake Books Supplement (1995), mentions this set as being at “Yale” and reproduces plate 1 (194-96). In my thesis, “Technical and Material Studies of William Blake’s Engraved Illustrations of the Book of Job (1826)” (2005), I was able to announce its exact location at Yale and add considerably more detail to the accounts set out by Essick.

The former owner of the set was unknown to Essick; as he says, the 1936 catalogue does not mention the vendor’s name. The title page of the catalogue names eight sellers, but none is a known collector of Blake (“Some Unrecorded Proofs” 96). A bookplate at the back of the front cover of this new set of proofs, however, reveals the name of the former owner, Charles J. Rosenbloom, the famous Blake collector of the early twentieth century.

All illustrations of the Rosenbloom set are from the Beinecke Library, Yale University (call number 1976 +373).
The bookplate is a copy of the painting The Peaceable Kingdom by the American painter Edward Hicks (1780–1849),Hicks painted many versions of The Peaceable Kingdom; Rosenbloom owned this oil painting that Hicks made around 1837 and later bequeathed it to the Carnegie Museum of Art, Pittsburgh (accession number 74.7.13). one of Rosenbloom’s favorite paintings. There is also another plate, printed “YALE UNIVERSITY LIBRARY / BEQUEST OF / CHARLES J. ROSENBLOOM / YALE 1920.” “Yale 1920” refers to his year of graduation, rather than the year of the bequest.

According to a photocopy of the dealer Sessler’s unpublished sales records, Charles Sessler, Inc., of Philadelphia, acting as agent, bought the proofs at the 1936 auction for “Rosenbloom.”  The invoice, dated 12.19.36, indicates that the cost was $5000 plus 10% commission to Sessler.I am grateful to Essick for this information. With this information, I am able to suggest that this rediscovered set of Blake’s Job proofs should be named after the early collector, consistent with other sets, and therefore be known as the Rosenbloom proofs.

Charles J. Rosenbloom (1898–1973) was a Pittsburgh financier, philanthropist, art collector, and bibliophile. He developed a great interest in collecting paintings and prints, owning works by Lucas Cranach, Pieter Bruegel, Albrecht Dürer, Rembrandt, Joshua Reynolds, and Blake, as well as Rosenbloom’s contemporary Muirhead Bone.See Paintings and Prints from the Collection of Charles J. Rosenbloom. Part of his growing print collection was exhibited at the Carnegie Institute in 1938.Catalogue of a Loan Exhibition: Prints from the Collection of Charles J. Rosenbloom, with a foreword by Edward Duff Balken, who praises Rosenbloom as a discriminating collector. Blake’s works were not included in this exhibition. As Rosenbloom was a Yale graduate, much of his book and manuscript collection was bequeathed to the university.He bequeathed many other works to the Carnegie Institute in Pittsburgh. The Beinecke call number for the Job set is 1976 +373, the first number indicating the date of accession. The April 1975 issue of the Yale University Library Gazette reports Rosenbloom’s bequest of 184 items, including the Job proofs.Suzanne Kutter and Donald Gallup’s “A Check-List of the Bequest” lists at no. 120, under the category of “Illustrated Books,” “[William Blake.] Illustrations of the Book of Job … London, Published by William Blake, 1825 [i.e., 1826]. Twenty-one proof plates before all imprints. In original cloth portfolio” (330).

On initial examination, the Rosenbloom proofs, except for the title page, appear to be very early states. The printing is somewhat misregistered on the paper, sometimes tilted obliquely with awkward edges. Some of the leaves have irregular shapes. There are random marks and stains on the leaves, some of which appear to be from inky fingers. In other words, these proofs seem to have been working proofs pulled by Blake for his own reference, rather than for showing to John Linnell as commissionerAlthough the final imprints of the Job illustrations do not bear Linnell’s name, documents show that he commissioned the work, sponsored it, and sold copies (see Bryant). or to collectors as images for sale. The Rosenbloom proofs have the border design and text, including the signature and plate number, but lack the imprint and the word “Proof.” Essick assumed that many or all of the pre-publication proofs in the set are the very first states after the border was added (“Some Unrecorded Proofs” 96). In most cases, this assumption is correct, except for plates 2, 3, 10, and 14 (see descriptions of each plate below).

As described in the 1936 catalogue, two plates, 17 and 20, are printed on laid paper, and the others are on wove paper (the title page, which is in the published state, is printed on India paper laid on wove paper). Five plates, 3, 9, 12, 16, and 21, have the watermark “J WHATMAN 1823”. The watermark is consistent with the Hofer proofs, of which plates 13, 15-18 are watermarked “J WHATMAN 1823” (Essick, “Blake’s Engravings to the Book of Job” 51). According to Linnell’s accounts,There are two relevant Linnell accounts, the cash account book 1822–36 (now in the John Linnell Archive, Fitzwilliam Museum) and the Yale account book (Beinecke). See Bryant. before the payment to James Lahee in March 1826 for the published “Proof” sets on India paper there were two proving sessions, in March 1825 at Lahee’s and at Dixon’s (the other printer Linnell used on the project). The charges for the paper and printing are combined in Linnell’s record of the payments, implying that the paper came from the printers’ stock. There was also proving by Blake himself using his own stock of paper, either before the Dixon and Lahee proving sessions or between this proving and the final printing at Lahee’s. Frederick Tatham wrote in ink on one proof, Essick’s state 4 of plate 19 (Tatham proof), “proof printed by Blake himself or by Mrs. Blake at his own press” (Essick, “Blake’s Engravings to the Book of Job” 90). If a proof as late as the Tatham proofs (after the addition of an imprint) was printed by Blake or his wife, it is more likely that the Rosenbloom proofs (before the addition of an imprint) were printed by Blake himself.

A list of the known pre-publication proofs, with provenance information, is given below (abbreviations follow Essick, “Blake’s Engravings to the Book of Job”).

Pre-publication proof sets:For full descriptions of these sets (except Rosenbloom), see Essick, “Blake’s Engravings to the Book of Job” 50-54.

El Dieff (University of Texas, Austin, USA)
The House of El Dieff, the firm of the New York dealer Lew David Feldman, sold this set to the Humanities Research Center, University of Texas, Austin, in 1967. It is a complete set of twenty-two plates, with all but the title page in pre-publication proof states. They seem to be a late state pulled after the Hofer set and before the Evans set.

Evans (British Museum, London, UK)
The British Museum acquired the set in 1847 from Evans, perhaps the London print dealer Edward Evans, who had acquired copies of Job from Linnell. This is a complete set, of which plates 2, 4, 5, 8-11, 13-15, 18, and 21 are pre-publication proofs. They are probably the last pre-publication impressions pulled.

Hanley (University of Texas, Austin, USA)
The Humanities Research Center, University of Texas, Austin, acquired this set from T. Edward Hanley in 1965. It contains pre-publication states of plates 1-3, 6, 11, 14, 16, 17, and 19-21.

Hofer (Houghton Library, Harvard University, Cambridge, Massachusetts, USA)
Philip Hofer owned this set, which was for many years on deposit at the Houghton and passed to the library after Hofer’s death in 1984. It comprises twenty-one pre-publication proofs, lacking only the title page, all in the states with Blake’s imprint and the “Proof” inscription.

Keynes (Fitzwilliam Museum, Cambridge, UK)
Geoffrey Keynes bequeathed this set to the Fitzwilliam in 1982. It comprises thirteen pre-publication proofs: plates 1 (in two states), 2 (in two states), 6-8, 11, 14, 16, 17, 19, and 21.

Linnell (National Gallery of Art, Washington, DC, USA)
This set of sixty-six impressions contains many of the earliest pre-publication states; it remained in Linnell’s collection until his death in 1882. The rare-book collector Lessing J. Rosenwald bought the set from the London dealer William Robinson, who had acquired it from Linnell’s grandson Herbert. Rosenwald gave his collection of engravings to the National Gallery of Art in 1945.

Riches (Fitzwilliam Museum, Cambridge, UK)
T. H. Riches, Linnell’s grandson-in-law, bequeathed the set to the Fitzwilliam in 1935. This is a complete set of pre-publication impressions with the border designs, in the same states as the White proofs.

Rosenbloom (Beinecke Library, Yale University, New Haven, Connecticut, USA)
Charles J. Rosenbloom bought this set at auction in New York in 1936; it was bequeathed to Yale in 1975, presumably in accordance with Rosenbloom’s will after his death in 1973.See Liebert. It is a complete set of twenty-two plates, of which all but the title page are pre-publication proofs.

Tatham (British Museum, London, UK)
Tatham sold this set—pre-publication proofs of plates 2, 8, 11 (two impressions in different states), 16, and 19-21—to the British Museum in 1867.

White (National Gallery of Art, Washington, DC, USA)
The book collector William Augustus White acquired this set, which had belonged to Linnell’s daughter, Hannah; later it went to Lessing J. Rosenwald, who gave his collection of engravings to the National Gallery of Art in 1945. This set is in the same states as the Riches proofs.

Odd proof states:

Cleverdon (last recorded in the private collection of Douglas Cleverdon, London, UK)See note 4. Plate 6 was sold by Cleverdon in 1967, but is untraced since then. There is some uncertainty about its state. See Essick, “Blake’s Engravings to the Book of Job” 69.
Plates 1, 6

ESSICK (private collection of Robert N. Essick, Altadena, California, USA)Capital letters are used to distinguish the collection from the scholar.
Plates 1, 2, 11

McGill (McGill University, Montreal, Quebec, Canada)
Plates 17, 19

Morgan (Morgan Library and Museum, New York, USA)
Plate 11

Philadelphia Museum of Art (Philadelphia, Pennsylvania, USA)
Plate 13

The following section is a plate-by-plate description of the Rosenbloom proofs, outlining their features in order to determine their states in comparison with other proofs. At the end of each description is a summary of the development of the known proof states. In the summaries I pay more attention to the states before and after the Rosenbloom proofs, and note only important changes in other states.For full details of other states, see Essick, “Blake’s Engravings to the Book of Job.” It is useful to summarize the overall development of the states:
a. Central design only. b. Adding border and texts, including Blake’s signature and the plate number.
c. Adding Linnell imprint (“Published March 8:1825 by J Linnell. 6 Cirencester Place” [plate 6] and variations thereof).The Linnell imprint is found on pls. 2 (Hanley and Keynes proofs), 3 (trace on Hanley proof), 6 (Linnell proofs), 11 (Tatham proof), 16 (Tatham proof), 19 (Tatham proof), 20 (Tatham proof), and 21 (Hanley proof). d. Changing imprint to Blake’s (“London Published as the Act directs March 8:1825 by William [or Willm] Blake No 3 Fountain Court Strand”). e. Adding the word “Proof” (published state A).
f. Removing the word “Proof” (published state B).
The Rosenbloom proofs are located at the second stage, a very early and important phase of the development, where the full scale of the plates is formed.

Title page: published state

There is no specific difference between the Rosenbloom title page and the published state. It is the only one in the set with the imprint—“London Published as the Act directs March 8:1825 by William Blake No 3 Fountain Court Strand”—and the only one that is not a pre-publication state. This page is very likely a later addition, as it is printed on thin India paper laid on top of thicker wove paper, and is therefore different from all the other Rosenbloom impressions, which are printed directly on wove or laid paper. India paper was an option for subscribers who bought the published “Proof” set. The presence of a published state of the title page suggests that no impression of the title page was pulled with the Rosenbloom proofs, possibly because the title plate had not yet been executed. Another possibility is that this page was simply added to make up a whole set, nearer to the time of the 1936 auction.The auction catalogue records that “the title-page, which was probably added in 1825, is the only plate on India paper.” The entry also mentions the printed Job label. It is not clear who made up this proof set and when, but it was probably assembled fairly early on by Blake or Linnell using pre-publication proofs, plus the title page and label. I am grateful to Essick for this information.

The pre-publication title page appears in the Riches and White proofs. The lack of a pre-publication title page in the Rosenbloom set indicates that the decision to add a title page to the set by either Linnell or Blake himself would have been made after adding the border designs on all the other plates.

The development of the title page:
State 1 (Riches; White): The same as in the published state, except lacking the horizontal lines between the Hebrew inscription and “ILLUSTRATIONS of”, the imprint, and some lines on the images.
*Published state (Rosenbloom, etc.): Adding the Blake imprint and some more lines.The asterisk before a description is to emphasize the Rosenbloom state.

Plate 1: state 5

Plate 1 of the Rosenbloom proofs bears the most notable differences from the other states, lacking two lines of inscription on the altar in the lower border, “The Letter Killeth / The Spirit giveth Life”, but with an extra line under the altar, “Prayer to God is the Study of Imaginative Art”. These extra words seem to be drafts scratched in drypoint on the copperplate with a casual and irregular appearance. Essick observes that photographs (of plates 1 and 21) in the 1936 auction catalogue “strongly suggest” that “the letters were scratched into the copperplates in drypoint rather than written with pen or pencil on the individual impressions” (“Some Unrecorded Proofs” 97-98). My close examination of the original has allowed me to establish that the line is indeed cut and not added in pen. The inscription “It is Spiritually Discerned” above this extra line has the same character, before being cut more deeply with the graver in state 6 (Essick’s state 5).Essick’s copy of pl. 1 in this state is reproduced in his “Some Unrecorded Proofs,” illus. 3, and William Blake, Printmaker, fig. 224 (where it is identified as the fourth state, an error later corrected in “Blake’s Engravings to the Book of Job” 59). Other text on this plate has been engraved with the graver and so has more regular, straight, and neat characters.

In Essick’s catalogue, states 1 to 4 of plate 1 have only the central design. From his state 5 on, the inscriptions are the same as in the published state. The Rosenbloom proof comes between his states 4 and 5 and is the new state 5. In his state 5 (the new state 6), “Prayer to God is the Study of Imaginative Art” has been removed and “The Letter Killeth / The Spirit giveth Life” has been added.

This plate has the highest number of proofs in the series before the addition of the border design. It demonstrates Blake’s uncertainty about how the engraved lines on the copperplate would show on the print, and thus it might have been the very first plate he engraved, as a trial. Samuel Palmer saw Blake’s work on this plate in October 1824 (Palmer 9-10, Gilchrist 1: 297).

The development of plate 1:
State 1 (Linnell B-4136): Central design only, with outlines of figures and landscape.
State 2 (Linnell B-4137): Central design only, with addition of much hatching.
State 3 (Linnell B-4138): Central design only, with more lines.
State 4 (Keynes; Linnell B-4139): Central design only, with further darkening as well as burnishing of figures. The Linnell version has a pencil sketch of the tent shape for the border design.
*State 5 (Rosenbloom): Darkening slightly all the figures by adding more lines in the central design, most noticeably on Job’s left leg; burnishing some parts, such as the hip and right leg of the daughter on the left of Job’s wife; adding the border design, including the signature and number, with inscriptions as in the published state, except lacking “The Letter Killeth / The Spirit giveth Life” on the altar and having an extra line, “Prayer to God is the Study of Imaginative Art”, under the altar.
State 6 [Essick’s state 5] (Linnell B-4140; Riches; White; ESSICK): Further darkening all the figures in the central design; cutting with the graver and darkening the words and images in the border design; extending the cloud lines in the lower border to touch the corners of the altar; adding “The Letter Killeth / The Spirit giveth Life” on the altar; erasing “Prayer to God is the Study of Imaginative Art”.
State 7 [Essick’s state 6] (Hanley; Keynes; Cleverdon): Adding the Blake imprint and “Proof” inscription; darkening as well as highlighting with burnishing figures and background.
State 8 [Essick’s state 7] (El Dieff; Hofer): Cutting the “Proof” inscription deeper with the graver; adding more lines to the central design.
Published state: Adding a few more strokes.

Plate 2: state 3

Plate 2 is very similar to Essick’s states 2 and 3, with slight differences; it comes between them as the new state 3. The fact that as a set the Rosenbloom proofs are earlier than the Riches and White proofs (both Essick’s state 3) also supports the argument that the Rosenbloom plate 2 should be the new state 3.

In state 2 the border design and text have been added, including the signature and number, but without the words “The Angel of the Divine Presence” and the Hebrew inscription that appear in the published state. The central design of Rosenbloom’s plate 2 resembles state 2; no work was added to the central design before printing the Rosenbloom proof. There are only very few additions to the border design, such as the small horizontal lines, perhaps nests, under the birds on both sides of the border. The proof lacks some features of Essick’s state 3 (the new state 4), such as the rays of light above the frame of the central design on the left side of the left cloud band (below “up” in the inscription) and on the right side of the right cloud band (below the “L” of “Likeness”). Nothing of the imprint or of the rows of dots forming crowns on the heads of the son and daughter on Job’s left is seen in the Rosenbloom proof.

The development of plate 2:
State 1 (Linnell B-4141): Central design only, with a pencil sketch of a border design with two tree trunks and an arc.
State 2 (Linnell B-4142): Adding the border design, with inscriptions as in the published state, except lacking “The Angel of the Divine Presence” and the Hebrew inscription above the central design; darkening and defining the figures in the central design (most noticeably, the halo around God has a clear edge).
*State 3 (Rosenbloom): Adding very few details. See description above.
State 4 [Essick’s state 3] (Linnell B-4143, B-4144, B-4145; Riches; White; ESSICK): Further darkening all figures in the central design; strengthening words and images in the border with the graver; adding a second line at the end of the tendril on the upper left frame of the central design; adding rays on both sides of the two cloud bands in the upper border.
State 5 [Essick’s state 4] (Hanley; Keynes): Adding part of the Linnell imprint, with only ghostly lines of his name and address visible; adding crowns on the heads of Job’s son and daughter; further darkening of images.
State 6 [Essick’s state 5] (Tatham): Changing the imprint to Blake’s; adding the “Proof” inscription.
State 7 [Essick’s state 6] (Keynes): Adding a little more hatching.
State 8 [Essick’s state 7] (El Dieff; Hofer): Burnishing away the lines of radiance in God’s halo and defining the halo with a curving line; darkening some images and burnishing others.
State 9 [Essick’s state 8] (Evans): Adding the Hebrew inscription and “The Angel of the Divine Presence”; darkening images.
Published state: Further darkening of images.

Plate 3: state 3

In the development of the proof states, plate 3 had a major revision to its border, where flames in early states were later changed to serpents, leaving large areas of repoussage on the back of the copperplate.See Sung, William Blake and the Art of Engraving 91.For an explanation of the term repoussage, see note 3. The Rosenbloom proof is between Essick’s states 2 and 3 and becomes the new state 3. In state 2, both sides of the border have steps, as in the published state, with four large tongues of flame above them on each side. In Essick’s state 3, flames and scorpions have been added in the border to the left and right of the lower corners of the central design. The Rosenbloom proof has the four large tongues of flame on both sides and some extra flames above them, but is without the scorpions and flames in Essick’s state 3. It also burnishes and highlights the quite dark state 2, most notably the right side of the column on the right side of the central design and Satan’s left leg. The central design is further burnished and made brighter in the next state, the new state 4. In the new state 4 (Essick’s state 3) two more downward tongues of flame are added on both sides of the border. All of the tongues of flame are removed in the new state 5 (Essick’s state 4), leaving only some ghost lines (which are reengraved later) of two small tongues on the right border just below the cloud band; the tongues are replaced with the scaly tails of serpents.

The development of plate 3:
State 1 (Linnell B-4149, B-4150): Central design only.
State 2 (Linnell B-4151): Darkening the central design; adding the border design, with two cloud bands in the upper border, four large tongues of flame and steps under them on both sides, inscriptions as in the published state, and the signature and simple flame lines in the lower border.
*State 3 (Rosenbloom): Burnishing the central design; adding extra flames in the border. See description above.
State 4 [Essick’s state 3] (Riches; White): Further burnishing the central design to brighten the tone; adding on both sides of the border two extra-long flames to the four large tongues of flame and, below, a scorpion with extended flames.
State 5 [Essick’s state 4] (Hanley): Adding part of the Linnell imprint; removing the flames under the cloud bands in the border except for two tongues on the right and replacing them with scaly tails of serpents on both sides; darkening as well as burnishing to highlight the central images.
State 6 [Essick’s state 5] (Hofer): Changing the imprint to Blake’s name and address; adding the word “Proof”; adding a third flame in the right border.
State 7 [Essick’s state 6] (El Dieff): Adding three tongues of flame in the left border; further darkening the central design.
Published state: Adding more lines and hatching.

Plate 4: state 2

Essick’s state 1 of plate 4 has the central design only, while his state 2 has the border design and text. A comparison of his state 2 with the Rosenbloom proof reveals that the latter also has the border design but lacks the cloud lines on both sides of the border, whereas there is no significant difference between the border design of Essick’s state 2 and the published state. The Rosenbloom proof is therefore a transitional state between his states 1 and 2 and becomes the new state 2. In this state the cloud outline and bolt of lightning below the lower right corner of the central design, as well as the flame above the lightning, are lightly sketched in drypoint but not yet cut with the graver. This shows how Blake used drypoint for sketching at the earliest stage of developing motifs in the border.

The relatively few proof states of this plate suggest that it was possibly engraved late, as Blake’s technique grew more skillful, with fewer revisions required and fewer proofs needed. The few repoussage marks on the verso of the copperplate may also support this assumption.See Sung, William Blake and the Art of Engraving 92.

The development of plate 4:
State 1 (Linnell B-4152, B-4153): Central design only. B-4153 has a pencil sketch of a border design with Gothic arches, completely different from the engraved design.
*State 2 (Rosenbloom): Adding more details to the central design; adding the border design, including sketches of all inscriptions, the cloud outlines as in the published state, except for the lower part on the right and left sides, the flames below, and Blake’s signature. Some lines in the border have only been sketched in drypoint. See description above.
State 3 [Essick’s state 2] (Riches; White): Adding the lower part of the cloud lines in both the right and left borders; cutting more clearly with the graver the cloud outline and bolt of lightning in the lower right border and the flame above; adding more hatching patterns to the central design.
State 4 [Essick’s state 3] (El Dieff; Hofer): Adding the Blake imprint, the “Proof” inscription, and more hatching patterns.
State 5 [Essick’s state 4] (Evans): Further darkening.
Published state: Finishing with a few more lines.

Plate 5: state 2

Plate 5 of the Rosenbloom proofs has an important difference from the published state. The auction catalogue of 1936 records that it lacks two lines of inscription in the bottom border: “And it grieved him at his heart / Who maketh his Angels Spirits & his Ministers a Flaming Fire”. This has been confirmed by my inspection. These words appear from Essick’s state 2 onwards (Essick, “Blake’s Engravings to the Book of Job” 54). The Rosenbloom proof is therefore earlier than Essick’s state 2, but later than state 1 with the central design only: it is the new state 2. In addition, the flames arching above the frame of the central design toward each other in Essick’s state 2 (below “Behold he is in thy hand: but save his Life”) are not on the Rosenbloom proof; they are added in Essick’s state 2 and removed in his state 3. In the lower border, the Rosenbloom proof lacks some of the flames above the serpents in Essick’s state 2. The central design also lacks a great deal of work added in the next state. For example, the bread in Job’s hands is depicted only in outline in the new state 2, but lines have been added for shadows in the next state (Essick’s state 2), and the mountain left of the head of Job’s wife develops from two ridges in the new state 2 to one in the next state.

The development of plate 5:
State 1 (Linnell B-4154): Central design only, with all figures in outline only except for Satan, who has more lines and dots than the others.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number, except for two lines of inscription. See description above.
State 3 [Essick’s state 2] (Riches; White): Defining and darkening the central design; adding flames below “Behold he is in thy hand: but save his Life” in the upper border, arching toward each other, and the inscription “And it grieved him at his heart / Who maketh his Angels Spirits & his Ministers a Flaming Fire” in the bottom border.
State 4 [Essick’s state 3] (El Dieff; Hofer): Adding the Blake imprint and “Proof” inscription; erasing the flames below “Behold he is in thy hand: but save his Life” in the upper border; darkening the central design.
State 5 [Essick’s state 4] (Evans): Adding a little more hatching, especially on Satan, God, and the block beneath Job’s arm.
Published state: Adding lines on the background hills.

Plate 6: state 2

Plate 6 is close to Essick’s state 3, but has fewer strokes on the grass and straw below Job’s leg in the central design, and the sun rays are less defined. Thus, it is earlier than his state 3. A comparison of Essick’s state 2 (Hanley) with the Rosenbloom proof shows that apart from the darker tone of the Hanley proof, which may have resulted from inking, there is no specific difference between the two.I worked from a photograph of the Hanley proof, so this difference could also be the result of darker lighting. The Rosenbloom proof is very likely the same state as the Hanley proof, and Essick’s states remain unchanged.

The development of plate 6:
State 1 (Linnell B-4155, B-4156, B-4157): Central design only.
*State 2 (Hanley; Rosenbloom): Darkening the central images; adding the border design and text, including the signature and number.
State 3 (Riches; White): Adding detailed engraved lines, such as the grass and the straw Job lies on; burnishing and highlighting some areas, such as the rays at the edges of the cloud and Satan’s left leg.
State 4 (Linnell B-4158, B-4159, B-4160): Adding the Linnell imprint.
State 5 (Keynes): Changing Linnell’s imprint to Blake’s; adding “Proof” inscription; burnishing some areas to highlight the contrast, such as the shafts of sunlight and the lower margins of the cloud below Satan’s arms; modifying Satan’s right heel.
State 6 (Hofer): Adding more lines of detail.
State 7 (El Dieff; Cleverdon [sold by him in 1967]): Adding a few more strokes.
Published state: Adding more hatching; defining outlines more clearly.

Plate 7: state 2

Plate 7 appears to be the state after the border and text have been added but before Essick’s state 2. The outlines of most figures in the central design are cut deeper in his state 2 than in the Rosenbloom proof. The print notably lacks the trees beside the shepherd and shepherdess in the border, which appear in Essick’s state 2. The Rosenbloom proof is therefore after state 1, which has the central design only, and becomes the new state 2. The two mourning figures in the border at the upper corners of the central design are only sketched in drypoint; they are cut more deeply and neatly with the graver in the new state 3 (Essick’s state 2). The “i”s in the two instances of “recieve” in the upper border do not have dots, which are added in the new state 3.

The development of plate 7:
State 1 (Linnell B-4161, B-4162): Central design only, with figures sketched in stipple.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number; adding nearly full details to the central design. See description above.
State 3 [Essick’s state 2] (Linnell B-4163, B-4164, B-4165; Riches; White): Adding the trees beside the shepherd and shepherdess in the border; cutting the mourning figures in the upper border with the graver; defining and highlighting the outlines of the figures, especially the faces in the central design; adding more lines to the vegetation at the bottom of the central design, especially in the bottom left corner.
State 4 [Essick’s state 3] (Keynes): Adding the Blake imprint and “Proof” inscription; adding more hatching and crosshatching to the central images; creating some sun rays by burnishing above the head of Job’s wife; burnishing the foremost friend’s feet and the grass below at the bottom left of the central design.
State 5 [Essick’s state 4] (Hofer): Further darkening the central design; cutting more deeply the border images.
State 6 [Essick’s state 5] (El Dieff): Adding some details to the central design.
Published state: Reworking the radiance above the mountains; adding more lines.

Plate 8: state 2

The Rosenbloom proof is very close to Essick’s state 2, with the border and text, but it has fewer strokes on the vegetation in the bottom border, most notably on the spiky branch below the left mushroom, and on the central image, such as the fabric at the waist of Job’s wife, Job’s face and beard, the outline of his knees and feet, and the outlines of the three friends. Thus it becomes the new state 2, before Essick’s state 2 but after state 1, which has only the central design. The “i” in the inscription “nights” on the lower border of the Rosenbloom proof lacks the dot; it is added in the next state.

The development of plate 8:
State 1 (Linnell B-4166, B-4167): Central design only, with hatching for most of the background.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number; filling details in the central design.
State 3 [Essick’s state 2] (Riches; White): Adding a few more strokes to the vegetation in the bottom border and to some details in the central design, such as Job’s face and beard. See description above.
State 4 [Essick’s state 3] (Tatham): Adding the Blake imprint and “Proof” inscription; darkening the central design; making the crack on the column above the friends more prominent; adding more contrast to the whole image.
State 5 [Essick’s state 4] (Keynes): Adding further lines to the central design, especially the boils on Job’s body.
State 6 [Essick’s state 5] (El Dieff; Hofer): Adding further lines to the central design, especially Job’s hair and beard and the clothes of Job’s wife.
State 7 [Essick’s state 6] (Evans): Adding further lines to the central design, especially the clothes of Job’s wife.
Published state: Some burnishing of the central design, e.g., on Job’s wife.

Plate 9: state 2

Plate 9 lacks details of the tree branches in the right and left borders and the cloud line and tree branches surrounding the inscription in the lower border, all of which are in Essick’s state 2 and the published state. It also lacks the vertical lines that appear between tree trunks below the cloud outline in both the right and left borders in his state 2. This evidence indicates that the Rosenbloom proof is before Essick’s state 2, becoming the new state 2. The “i”s in “his” (three times) and “Saints” in the upper and right borders and “Spirit” and “hair” in the lower border all lack the dots, which are added in the new state 3 (Essick’s state 2). The central designs in the Rosenbloom state and the following state have very few differences. In the Rosenbloom proof, some rays of light penetrate the cloud outline above the raised hand of the figure lower right. In the next state (the new state 3), these lines have been burnished away and the rays terminate at the outline of the cloud.

The development of plate 9:
State 1 (Linnell B-4168): Central design only, with many details already in place. Blake sketched a border design, completely different from the engraved version, in pencil on this proof.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number; adding more hatching to details in the central design, most notably the faces of all figures, and to the black rays that are below God’s garment and on the cloud band he stands on; burnishing the area right of God’s black mandorla to form a cloud band or another lighter mandorla.
State 3 [Essick’s state 2] (Riches; White): Adding more branches to the trees in the border design and the vertical lines between tree trunks; engraving the inscriptions more deeply with the graver; adding dots to the “i”s. See description above.
State 4 [Essick’s state 3] (Hofer): Adding the Blake imprint and “Proof” inscription; adding more hatching and crosshatching to the central design.
State 5 [Essick’s state 4] (El Dieff): Burnishing part of God’s garment.
State 6 [Essick’s state 5] (Evans): Adding more lines around the image of God.
Published state: Adding more lines to the image of God.

Plate 10: state 5

States 1 to 3 of plate 10 have the central design only. The Rosenbloom proof is very close to Essick’s state 4, where the border design has been added; both lack one of the two curving lines on the right side of the upper border in his state 5. In comparison with Essick’s state 4, the Rosenbloom proof has some areas of burnishing that are consistent with later states, such as the foremost friend’s legs and the straw under Job’s left knee. There are significant differences in the outlines of the images in the central design: in the Rosenbloom state, Job’s waist is narrowed, his beard is lengthened, and the shape of the garment folds of the foremost friend is changed. Job’s narrowed waist and extended beard are present in all the remaining proofs from Essick’s state 5 to the published state. Accordingly, the Rosenbloom proof has to be after Essick’s state 4; it becomes the new state 5.

The many changes to the central design of plate 10 correspond to the numerous proof states both before and after the addition of the border design; these changes are consistent with the large areas of repoussage on the verso of the copperplate.See Sung, William Blake and the Art of Engraving 99-100. This plate is probably among the earliest to be engraved, perhaps after plate 1.

The development of plate 10:
State 1 (Linnell B-4169): Central design only, with outlines of the figures and some hatching for the background.
State 2 (Linnell B-4170): Central design only, with more hatching and crosshatching for the figures and background, Job’s right arm widened, his garment altered, and the foreground enlarged (the lower edge is increased about 2 mm.). Blake sketched a border design in pencil on this proof.
State 3 (Linnell B-4171, B-4172, B-4173): Central design only, with more darkening with hatching.
State 4 (Linnell B-4174): Adding the border design and text, including the signature and number; adding more hatching to the central design, as well as the tears and forehead wrinkles to Job’s wife; burnishing the outline of the mountain in the upper right corner.
*State 5 (Rosenbloom): Burnishing some areas; changing image outlines in the central design. See description above.
State 6 [Essick’s state 5] (Riches; White): Adding a second line to the right curve in the upper border; adding more lines to the border design; partly darkening and partly burnishing the central images; changing once again the shape of the garment folds of the foremost friend in the central design.
State 7 [Essick’s state 6] (Hofer): Adding the Blake imprint and “Proof” inscription; further darkening and highlighting the central design; adding more lines to the border design.
State 8 [Essick’s state 7] (El Dieff): Partly darkening and partly burnishing the central design.
State 9 [Essick’s state 8] (Evans): Further darkening and adding lines to the central design.
Published state: Adding more lines to the central design.

Plate 11: state 2

Under my inspection, plate 11 of the Rosenbloom proofs has a mark that looks like an additional plate indentation to the right of the current one, indicating that the copperplate might have been passed through the press more than once, perhaps the first time to test the position before inking. Alternatively, this supposed second platemark may be just a crease in the paper. The slanting position of the print also reveals that the printing was not by a professional printer or for sale. The number of proof states of this plate is the highest in the series, and the hammer marks on the verso of the copperplate are also among the most numerous.See Sung, William Blake and the Art of Engraving 100.

The Rosenbloom proof lacks many lines that define the flames on each side of the border design, as well as the details of the flames in the upper and lower borders, that appear in Essick’s state 2. It is therefore the new state 2, before Essick’s state 2, but after state 1, which has the central design only.

The development of plate 11:
State 1 (Linnell B-4175, B-4176): Central design only, with many details already in place.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number, and more hatching to the central design.
State 3 [Essick’s state 2] (Riches; White; Morgan): Adding many more lines for the flames on all sides of the border; adding more hatching to the central design; burnishing the left flame at the lower right corner of the central design to reveal the devil’s fingers.
State 4 [Essick’s state 3] (Tatham): Adding the Linnell imprint; adding more details to and darkening the central design; adding more lines to the flames in the upper border.
State 5 [Essick’s state 4] (Tatham): Changing Linnell’s imprint to Blake’s; adding the “Proof” inscription; darkening the central images.
State 6 [Essick’s state 5] (Hanley): Adding more hatching to the central design.
State 7 [Essick’s state 6] (Hofer): Adding some flames above Job’s thigh and stomach; adding more crosshatching and hatching.
State 8 [Essick’s state 7] (ESSICK): Further darkening the central design, especially the thunderbolts in the upper left corner.
State 9 [Essick’s state 8] (Keynes): Adding a few more hatching and crosshatching lines to details.
State 10 [Essick’s state 9] (El Dieff; Evans): Adding a few more hatching and crosshatching lines to details.
Published state: Adding vertical strokes to the left leg of Job’s bed.

Plate 12: state 2

Plate 12 of the Rosenbloom proofs is the new state 2: it adds the border design to state 1, which has the central design only, but lacks the stars in the top border under the flying figures and the grass in the lower border (above “If thou sinnest what” and under “thou unto him” in the bottom right corner) in Essick’s state 2. The flying figures in the border are only sketched in drypoint on the Rosenbloom proof; they are made more defined with the graver in the next state (Essick’s state 2). The hand of the figure in the left border touching the inner frame is traversed by the framing line, whereas the line within the hand is burnished in the states that follow.

The development of plate 12:
State 1 (Linnell B-4177, B-4178): Central design only, with many hatching lines on the background as well as on figures.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number; adding more lines to details of figures in the central design. Many flying figures in the border design are only sketched in drypoint.
State 3 [Essick’s state 2] (Riches; White): Adding stars above the frame of the central design and grass in the lower border; cutting the flying figures in the border more deeply with the graver; adding more hatching to the central design, e.g., the ground around Elihu’s feet; defining the outlines of all figures more clearly; darkening all figures.
State 4 [Essick’s state 3] (El Dieff; Hofer): Adding the Blake imprint and “Proof” inscription; adding hatching and crosshatching to the central design almost to the level of the finished state.
Published state: Adding more hatching to the foremost friend’s knee below his hand.

Plate 13: state 2

The Rosenbloom proof has less detail on the tree in the lower border and fewer cloud lines in the left border than Essick’s state 2: it is the new state 2.

Plate 13 is an extraordinary plate with unexpected developments among its proof states, and this proof reveals a dramatic change as Blake developed his idea of the design. The upper frame of the central design is closed in the Rosenbloom proof, suddenly opened, with the image extending to the upper border, in the new state 3 (Essick’s state 2), and closed again in the new state 4 (Essick’s state 3). It appears that Blake at first created the more restrained whirlwind, then dramatically extended it, then restrained it again in all later states. The evidence of the Rosenbloom state shows Blake’s shift of mind in the most remarkable way: without it, the logical—but wrong—assumption would be that Blake “began” with the exuberant whirlwind.

On the Rosenbloom proof, the upper frame of the central design is defined with two straight lines, but there are two or three curved lines extending from God’s whirlwind between the upper line of the central design’s frame and the clouds in the upper border. More of these whirlwind lines are added in the new state 3 (Essick’s state 2), but the part between the two lines framing the central design is burnished out in the new state 4 (Essick’s state 3), leaving only one curve on the left corner; this curve and the whirlwind lines between the upper line of the central design’s frame and the clouds in the upper border are all erased in the new state 5 (Essick’s state 4).

The development of plate 13:
State 1 (Linnell B-4179, B-4180): Central design only, with a single-line frame on all sides.
*State 2 (Rosenbloom): Adding the second line to the frame of the central design; adding the border design and text, including the signature and number; extending two or three lines from God’s whirlwind between the upper line of the central design’s frame and the cloud above.
State 3 [Essick’s state 2] (Linnell B-4181): Burnishing and opening the two lines of the top frame of the central design to extend the whirlwind lines into the upper border; extending the horizontal lines representing sky into the area between the two lines that framed the top of the central design.
State 4 [Essick’s state 3] (Linnell B-4182, B-4183; Riches; White; Philadelphia Museum of Art): Closing the top frame of the central design; clearing the whirlwind lines from the frame except for one curve extending from the top left corner of the central design; adding hatching lines to the central and border designs. Some of the whirlwind lines extended from the central design to the upper border in state 3 remain in the border above the two lines of the frame of the central design.
State 5 [Essick’s state 4] (El Dieff; Evans; Hofer): Adding the Blake imprint and “Proof” inscription; clearing the last curved line from the left corner of the area between the two lines framing the top of the central design; removing the whirlwind lines between the upper line of the central design’s frame and the cloud above; further darkening the central design.
Published state: Adding some more lines to the whirlwind in the upper border, e.g., the top left corner, left of the inscription “Who is this ….”

Plate 14: state 5

States 1 to 3 of plate 14 have the central design only; the Rosenbloom proof is similar to Essick’s state 4, which has the border design added. A comparison of the two reveals that the outlines of the figures in the central design, especially the halo around God’s head, are more defined in the Rosenbloom proof, and the border design is cut more deeply with the graver, as is particularly noticeable on the cloud lines above the central design. This evidence determines that the Rosenbloom proof comes after Essick’s state 4 and is the new state 5.

The proof lacks the small flames that appear in the right border of the published state in the interstices above and below the clouds framing the central vignette. These flames are added in the new state 6 (Essick’s state 5).

The development of plate 14:
State 1 (Linnell B-4184): Central design only.
State 2 (Keynes; Linnell B-4185): Central design only, adding many details; burnishing the areas with stars above and around God’s halo.
State 3 (Linnell B-4186): Central design only, adding more details. There is also a pencil sketch of the border design, which looks very different from the later version.
State 4 (Hanley): Adding the border design and text, including the signature and number, as well as more details to the central design.
*State 5 (Rosenbloom): Defining the outlines of figures in the central design; redarkening the burnished areas around God’s halo in states 2, 3, and 4; cutting the border design more deeply with the graver. See description above.
State 6 [Essick’s state 5] (Riches; White): Adding the small flames in the right border in the interstices above and below the clouds framing the central vignette; defining more clearly the outlines of all figures in the central design.
State 7 [Essick’s state 6] (Hofer): Adding the Blake imprint and “Proof” inscription; adding many more hatching lines to the central and border designs.
State 8 [Essick’s state 7] (El Dieff): Further darkening and highlighting the central design.
State 9 [Essick’s state 8] (Evans): Further darkening and highlighting the central design.
Published state: Highlighting the central design.

Plate 15: state 1

Since there is no earlier proof with the central design only, plate 15 of the Rosenbloom proofs stands out as the very first state in existence. It is close to Essick’s state 1, which has the border design and text, but has fewer lines in the lower border defining waves just above the bottom outer frame line and on the body of Behemoth in the central design (the heavy lines in the middle of his body that help define his haunch).

The development of plate 15:
*State 1 (Rosenbloom): Much developed central design with the border design and text, including the signature and number. Hatching lines on the central design are quite near the last state.
State 2 [Essick’s state 1] (Riches; White): Adding more lines under the waves in the lower border above the bottom outer frame line; burnishing as well as adding lines to Behemoth to make the muscles on his face and body more prominent; defining more clearly the outlines of all figures in the central design.
State 3 [Essick’s state 2] (El Dieff; Hofer): Adding the Blake imprint and “Proof” inscription; adding more hatching lines in order to darken the central images.
State 4 [Essick’s state 3] (Evans): Adding more lines to the central design.
Published state: Adding more lines to the central design.

Plate 16: state 3

As mentioned in the 1936 auction catalogue, plate 16 of the Rosenbloom proofs does not have the two lines of text on the left side of the border that are present in the published state: “The Accuser of our Brethren is Cast down / which accused them before our God day & night”. The first line appears in Essick’s state 3; the Rosenbloom proof has to come before that state, and since states 1 and 2 have the central design only, it is situated between Essick’s states 2 and 3 and becomes the new state 3. What the auction catalogue does not mention, however, is that on the Rosenbloom proof there are traces of the erasure of an inscription at the place of “The Accuser of our Brethren is Cast down”. Blake might have put a different line of text, then changed his mind. The catalogue also notes a difference in the second line of inscription in the lower border: “And he said unto them” in the Rosenbloom proof becomes “Jesus said unto them” in the next state (Essick’s state 3). This also supports the conclusion that the Rosenbloom proof is the new state 3.

As with plate 13, the central design of this plate had important changes to its upper frame: Blake dramatically closed and opened the frame and then closed it again. The upper frame of the central design on the Rosenbloom proof is defined with two straight lines, which are burnished and opened up to extend God’s mandorla to the upper border in the new state 4 (Essick’s state 3), then this extension is burnished and closed with the double frame in the new state 5 (Essick’s state 4). In the Rosenbloom state, the upper border has a line of text under the cloud line and the two mourning angels’ wings on the two corners of the inner frame. It lacks the two cloud lines (and extended rays between the clouds) from the central design below and the uppermost child angels by God extending from below in the new state 4. These child angels are burnished and replaced with the two framing lines at the top of the central design in the new state 5, but the two cloud lines are retained in the upper border above the frame.

The child angels in God’s mandorla on the Rosenbloom proof appear to be the same as in state 2; their wings, only sketched lightly in drypoint, are engraved in the new state 4 (Essick’s state 3). There are stipple dots at the sides of the lower child angels, removed in the new state 4. The proof lacks a few hatching lines that are present in the next state in the large flame around the three falling figures.

On the border design, the proof lacks the small flames in the upper parts of the left and right borders and some lines defining water lower left and right that appear in the new state 4 (Essick’s state 3).

The development of plate 16:
State 1 (Linnell B-4187, B-4188): Central design only, with many figures in stipple outline.
State 2 (Linnell B-4189): Central design only, with its size increased slightly on the right and left sides; adding more hatching.
*State 3 (Rosenbloom): Adding the border design and text, including the signature and number (there are four tongues of flame on the lower left side and five on the right side of the border); adding more details on the central design. The frame of the central design is defined with two straight lines. See description above.
State 4 [Essick’s state 3] (Linnell B-4190): Burnishing and opening the upper frame of the central design; extending the rays of God’s mandorla and the cloud lines on both sides of God in the central design to the upper border; adding more child angels in God’s mandorla and wings for them; adding in the border another large tongue of flame on the right, more small flames on the right and left sides, lines defining water lower right and left, and the inscription “The Accuser of our Brethren is Cast down”; changing the inscription “And he said unto them” in the lower border to “Jesus said unto them”.
State 5 [Essick’s state 4] (Linnell B-4191, B-4192; Riches; White): Darkening the central design; clearing the upper frame of the central design and defining it with two straight lines as in state 3; removing the rays of God’s mandorla and the child angels in this area, but leaving the cloud lines above the upper frame of the central design.
State 6 [Essick’s state 5] (Tatham): Adding the Linnell imprint and, in the left border, a final inscription, “which accused them before our God day & night”; removing the cloud outline and the rays under it above the central design; darkening as well as burnishing details on the central design, including burnishing the lines between Satan’s head and the flames below.
State 7 [Essick’s state 6] (Keynes): Changing the imprint to Blake’s; adding three letters (“Pro”) of the “Proof” inscription; further darkening the central design; further burnishing the area between Satan’s head and the flames at the bottom.
State 8 [Essick’s state 7] (Hanley; Hofer): Completing the “Proof” inscription; adding more hatching and crosshatching lines to the central design; completing the border design to the level of the published state.
State 9 [Essick’s state 8] (El Dieff): Further darkening and highlighting details of the central design, especially around the figure of Satan.
Published state: Further darkening and highlighting details of the central design, especially around the figure of Satan.

Plate 17: state 2

Plate 17 is quite close to Essick’s state 2, most notably without the lower cloud line in the upper border and the radiance around the angel in the lower border in his state 3. The central images are less defined than those in Essick’s state 2, especially God’s long beard, where some lines were burnished from state 1, then more lines were added again in Essick’s state 2. The Rosenbloom proof is therefore before Essick’s state 2 and after state 1, which has only the central design. There is no difference in the border between the Rosenbloom proof and Essick’s state 2, whereas differences in the borders are crucial in establishing proof sequences for most other plates.

On the Rosenbloom proof, there is an indentation mark that looks like another tilted platemark on the left margin of the current platemark. As with plate 11, the paper was perhaps passed through the press a first time with the copperplate before inking to test the position, then a second time after the copperplate was inked. Alternatively, it may be just a crease in the paper.

The development of plate 17:
State 1 (Linnell B-4193, B-4194): Central design only, with many hatching lines in places.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number, with seven lines indicating rays in the upper border; adding more hatching and crosshatching in the central design; burnishing God’s long beard and the frontal lines of his legs and garment.
State 3 [Essick’s state 2] (Keynes): Re-adding lines to God’s beard and the frontal lines of his legs and garment.
State 4 [Essick’s state 3] (Linnell B-4195; Riches; White): Defining the inner halo around God; adding more hatching and crosshatching to the central images; adding the rays around the angel and more lines to her in the lower border; removing all but one of the rays in the upper border and replacing them with a lower cloud line and some short rays beneath.
State 5 [Essick’s state 4] (Hanley): Adding the Blake imprint, partly sketched in drypoint; further darkening the central design and border design.
State 6 [Essick’s state 5] (McGill): Completing the imprint; adding the “Proof” inscription, with the first letter cut with the graver and the rest in drypoint; adding more lines to the central design.
State 7 [Essick’s state 6] (El Dieff; Hofer): Cutting the “Proof” inscription with the graver; further darkening the central design.
Published state: Adding more hatching to the cloud left of God’s garment in the bottom left corner of the central design.

Plate 18: state 1

Since no earlier state with only the central design exists, plate 18 of the Rosenbloom proofs is the first state, before all the others listed by Essick. In comparison with Essick’s state 1, it lacks the radiance at both sides of the upper border under the cloud band, the horizontal but slightly wavy line underneath the grain stalks and above the block in the right border, and the dot on the “i” of “Friends”. The grain stalks on both sides of the border immediately next to the outer frames, the scroll lower right, and the horizontal lines (maybe of a table) beneath it are less defined in the Rosenbloom proof. A little work was also added in the new state 2 (Essick’s state 1) to the angels upper right.

The development of plate 18:
*State 1 (Rosenbloom): This proof is already very close to the published state, with the central design and the border design and text, including the signature and number.
State 2 [Essick’s state 1] (Riches; White): Darkening and defining more clearly images in the central design and border; adding some more lines in the border, e.g., the plants, the scroll, the angels, and two sets of rays in the upper border. See description above.
State 3 [Essick’s state 2] (Hofer): Adding the Blake imprint and “Proof” inscription; adding more lines to the central design, most notably on the cloud band above the altar; burnishing areas such as the head of Job’s wife; adding more lines to the border design.
State 4 [Essick’s state 3] (El Dieff): Adding a few more hatching lines to the central design.
State 5 [Essick’s state 4] (Evans): Defining the sun rays at the top of the central design more clearly; further darkening and highlighting the central design; adding five arrows in the white area at the top of the central design, above the rays.
Published state: Final darkening.

Plate 19: state 2

There is no obvious difference from Essick’s state 2, where the border design and text have been added to the central design of state 1. Since there is no new state for this plate, all of Essick’s proof states remain unchanged.

The development of plate 19:
State 1 (Linnell B-4196, B-4197): Central design only, with many hatching lines already in places.
*State 2 (Keynes; Rosenbloom): Adding the border design and text, including the signature and number; darkening the central design.
State 3 (Riches; White): Adding more leaves to the palm trees in the border design and more hatching lines to the central design.
State 4 (Tatham): Adding the Linnell imprint, more leaves to the palm trees in the right and left borders, and further hatching lines to the central design.
State 5 (Hanley): Changing the imprint to Blake’s (partly sketched in drypoint); adding further hatching lines to the central design.
State 6 (McGill): Completing the imprint with the graver; adding a few more lines.
State 7 (Hofer): Adding the “Proof” inscription; further darkening the central and border designs.
State 8 (El Dieff): Darkening as well as burnishing the central design.
Published state: Partly darkening and partly highlighting the central design to create more contrast.

Plate 20: state 2

Plate 20 of the Rosenbloom proofs lacks many details of Essick’s state 2, so comes before his state 2 and after state 1, which has the central design only. It has fewer leaves and grapes on the vines on all sides of the border than in Essick’s state 2. More importantly, there is a significant change in the first line in the lower border: the first word is “Were”, rather than “There were” in Essick’s state 2 and all the following states. Blake might have noticed the awkward syntax of “Were not found Women fair as the Daughters of Job” after the Rosenbloom proof and decided to add “There”, which is itself awkwardly squeezed in the left margin among the vines. This important change is not mentioned in the 1936 auction catalogue. Although there are traces of the erasure of the “W” on later states and even on the published state under close examination, they are not noted in Essick’s otherwise meticulous observations.

In the central design, there are also several differences between the Rosenbloom proof and the new state 3 (Essick’s state 2); the latter has much more definition of the lower reaches of the gown worn by the daughter on the left and the line defining the neckline of the gown of the daughter on the right. In the new state 3, hatching strokes were also added to the ground pictured in the roundel upper right.

The development of plate 20:
State 1 (Linnell B-4198, B-4199): Central design only, with many hatching lines already in places.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number. The first line of inscription in the lower border reads “Were not found Women fair as the Daughters of Job”. There are only six leaves in the lower border and ten in the upper border. The floor in the central design has been partially burnished from state 1.
State 3 [Essick’s state 2] (Riches; White): Adding more leaves and grapes to the border, as in the published state; changing the first line of inscription in the lower border to “There were …”; darkening the figures in the central design.
State 4 [Essick’s state 3] (Tatham): Adding the Linnell imprint, the middle girl’s belt, the right girl’s right hand, and other details in the central and border designs.
State 5 [Essick’s state 4] (Hofer): Replacing Linnell’s imprint with Blake’s; adding the “Proof” inscription and more lines.
State 6 [Essick’s state 5] (Hanley): Darkening and burnishing the central design; adding more details to the border design.
State 7 [Essick’s state 6] (El Dieff): Further darkening and burnishing the central design.
Published state: Further darkening the central design.

Plate 21: state 2

Plate 21, like plate 1, is a crucial proof in the Rosenbloom set. As noted by the 1936 auction catalogue and by Essick (“Some Unrecorded Proofs” 97-99, “Blake’s Engravings to the Book of Job” 55), this proof has an extra inscription under the altar, “Praise to God is the Exercise of Imaginative Art”, which is erased in Essick’s state 2. As with plate 1, I can confirm that the inscription was cut rather than added in pen. This line, as well as the two lines on the altar, “In burnt Offerings for Sin / Thou hast had no Pleasure”, was written in drypoint. In Essick’s state 2, the two lines of text on the altar have been cut with the graver and have a neater look, and “Thou” has been changed to “thou”. Other differences between the Rosenbloom proof and the next state include the lightly sketched tongues of flame on top of the altar, sharply angled left and right of the main cluster, which are cut more clearly in Essick’s state 2, and the blank scroll held by the central girl, which is filled with lines in his state 2. The Rosenbloom proof is, therefore, the new state 2: obviously before Essick’s state 2, but immediately after state 1, which has the central design only.

The development of plate 21:
State 1 (Linnell B-4200, B-4201): Central design only, with outlines of all figures and some hatching.
*State 2 (Rosenbloom): Adding the border design and text, including the signature and number, and more hatching on the central design.
State 3 [Essick’s state 2] (Riches; White): Burnishing away the inscription “Praise to God is the Exercise of Imaginative Art”; changing “Thou” to “thou” in the inscription on the altar; cutting with the graver the lines “In burnt Offerings for Sin” and “thou hast had no Pleasure” and the two small flames on each side of the altar; adding more lines on the central design. See description above.
State 4 [Essick’s state 3] (Hanley): Adding the Linnell imprint; adding more hatching and crosshatching, as well as some burnishing, to the central design.
State 5 [Essick’s state 4] (Keynes): Removing Linnell’s imprint and replacing it with Blake’s; adding some stipple and burnishing some areas to highlight them.
State 6 [Essick’s state 5] (Tatham): Adding the “Proof” inscription and fine hatching.
State 7 [Essick’s state 6] (Hofer): Further darkening and highlighting the central design.
State 8 [Essick’s state 7] (El Dieff): Further darkening the central design.
State 9 [Essick’s state 8] (Evans): Adding some more dots and lines to the central design.
Published state: Finishing with a little burnishing on the central design.

The tables below summarize the sequence, from earliest to latest, of the sets of pre-publication proofs: Linnell, Rosenbloom, Riches and White, Hofer, El Dieff, and Evans. The states of three sets—Tatham, Keynes, and Hanley—are not consistent in their order, but mostly they are Tatham-Keynes-Hanley. The inconsistency is probably because each set was not printed at the same time, but was assembled by collectors or sellers from different proof states.

The states of Blake’s pre-publication Job proofs, title page to plate 10

ProofsTitle12345678910Plates
Linnell 1, 2, 3, 4111111111, 2, 3Central design only
  Keynes 4         
   Linnell 2Linnell 2      Linnell 4Border and text added
Rosenbloom 5332222225
  Linnell 6Linnell 4    Linnell 3   
  ESSICK 6ESSICK 4   Hanley 2    
Riches;
White
16443333336
Tatham        4  Imprint and “Proof” inscription added
       Linnell 4    
Keynes 75   545  
Hanley 755       
  Cleverdon 7Tatham 6        
   Keynes 7        
Hofer 8864465647
       Cleverdon 7    
El Dieff 8874476658
Evans  9 55  769
The states of Blake’s pre-publication Job proofs, plates 11 to 21

Proofs1112131415161718192021Plates
Linnell1111, 2 1, 21 111Central design only
    Keynes 2       
    Linnell 3       
    Hanley 4       Border and text added
Rosenbloom22251321222
       Keynes 3 Keynes 2  
   Linnell 3, 4  Linnell 4, 5Linnell 4    
 Morgan 3 Philadelphia 4        
Riches;
White
33462542333
Tatham4, 5    6  44 Imprint and “Proof” inscription added
Keynes     7     
Hanley6    85 5 4
       McGill 6 McGill 6 Keynes 5
           Tatham 6
Hofer74573873757
 ESSICK 8        Hanley 6 
 Keynes 9          
El Dieff104583974878
Evans10 594  5  9

The significance of the Rosenbloom proofs is that they contain many early states of the engravings and thus represent some of Blake’s earliest ideas for Job, almost at the point at which he began the printmaking process. Although, like other sets, the proofs are a mix of states, ranging from 1 to 5, they may be seen to stand for Blake’s first printing of the designs as a whole. Seventeen of the twenty-one impressions in pre-publication states (all except plates 2, 3, 10, and 14) are the first states after the border design has been added and the design can be seen as an entirety. These proofs are, however, not Blake’s first thoughts about the border design. Six of the Linnell proofs have the engraved central design with a pencil sketch for the border design: plate 1 (state 4), plate 2 (state 1), plate 4 (state 1), plate 9 (state 1), plate 10 (state 2), and plate 14 (state 3). Most of these sketches are very different from the border designs on the later engravings. The lack of any other pencil sketches of the border designs and the presence of many traces of drypoint sketching on the copperplates indicate that Blake designed and executed the border directly on the copperplates in drypoint, then followed with the graver. In this sense of the use of drypoint sketches and revising or confirming with the graver, the execution on the copperplates, as shown in the Rosenbloom proofs, is not a first thought but a second thought, an evolution of conception, or reconception. This is also true with the many revisions of the central design.

As Essick suggests, the textual revisions on plates 1 and 21 from the Rosenbloom proofs to the later proof states (the erasure of “Prayer to God is the Study of Imaginative Art” on plate 1 and “Praise to God is the Exercise of Imaginative Art” on plate 21, and the addition of “The Letter Killeth / The Spirit giveth Life” on plate 1) demonstrate the practical aspect of Blake’s ideal of the unity of conception and execution. In Essick’s interpretation, the extra texts on the Rosenbloom proofs suggest “underlying similarities between passive prayer and active praise. The latter may be superior to the former, but the difference is in means, not ends” (“Some Unrecorded Proofs” 102). The revised version of plate 1 stresses instead the opposition of letter and spirit. In light of this change, Essick thinks that Blake’s conception of the Job engravings came into being at a late stage, instead of fully preceding the execution on the copperplates. At the very least, the Rosenbloom set should alert us to the complex processes, both visionary and revisionary, that Blake used in one of his last published works. Even the sketches in drypoint on the copperplates for the border designs have been altered and burnished before or after the cutting with the graver.

The Rosenbloom proofs suggest that the drypoint sketch represents the conception and the graver represents the execution, between which there exists a gap. They show an intermediate process, with some images sketched only in drypoint and others already defined with the graver. Essick says that “only as he slowly developed his images in the difficult medium of line engraving did Blake come to realize that the ‘perfect and upright’ man of ‘the land of Uz’ (Job 1:1) was not an archetype of the prayerful student of the arts but one who had already fallen in thrall to the dead letter of the law” (“Some Unrecorded Proofs” 102). He also mentions “the errors which adumbrate Satan’s entry into Job’s life” as a theme of the Job series (102). To Blake, these errors are not only the wrong means that Job used to worship God, but also the wrong notion of self-righteousness. As a practicing Christian as well as an artist, Blake illustrates the border of plate 18 with a painting brush and palette. The notion of perfection implied in the unity of execution and conception turns into the reality of a man who makes mistakes and constantly struggles with his satanic and godly natures. The Job engravings seem to be a realization of the imperfection of human nature and a history of growth, not only in Christian belief but also in artistic practice.

Whether this realization is conscious in Blake’s work, however, is not clear. The text on the published plate 1 stressing “spirit” and opposing “letter” seems to go back to Blake’s early identity as a visionary poet and artist who receives heavenly dictation and creates works without premeditation.Blake once claimed that he wrote a “Poem from immediate Dictation twelve or sometimes twenty or thirty lines at a time without Premeditation & even against [his] Will” (E 729). In the Job illustrations, the spiritual and visionary artist, such as the Job in plate 21, is not born but is the result of long-term growth. Perhaps in his later life Blake realized that the visionary artist must achieve a state of maturity.

Despite the difference between the text on plates 1 and 21 of the Rosenbloom proofs and the later revision, the versions are similar in meaning more than a shift of ideas. Since the extra phrases on these two plates are the only texts not from the Bible in the Job engravings, they could have been changed simply because Blake and Linnell wanted to keep all the texts biblical in order to suit a wider audience, and thus decided to remove Blake’s words. The Rosenbloom version, therefore, provides the hidden implication in the published state. “The Letter Killeth” on the published plate 1 implies “the Study of Imaginative Art” on plate 1 of the Rosenbloom proofs, and “The Spirit giveth Life” implies “the Exercise of Imaginative Art”. As a practicing artist, Blake regards engraving as a high imaginative art that cannot be achieved by the study of “letter” or skills, but only by the spiritual exercise of a visionary hand.

Bibliography

Bentley, G. E., Jr. Blake Books: Annotated Catalogues of William Blake’s Writings. Oxford: Clarendon Press, 1977.

---. Blake Books Supplement: A Bibliography of Publications and Discoveries about William Blake 1971–1992. Oxford: Clarendon Press, 1995.

---. Blake Records. 2nd ed. New Haven: Yale University Press, 2004.

Bindman, David, ed. William Blake’s Illustrations of the Book of Job. London: William Blake Trust, 1987.

Bryant, Barbara, comp. “The Job Designs: A Documentary and Bibliographical Record.” Bindman 103-47.

Catalogue of a Loan Exhibition: Prints from the Collection of Charles J. Rosenbloom, February 8th to March 31st, MCMXXXVIII. Pittsburgh: Carnegie Institute, [1938].

Erdman, David V., ed. The Complete Poetry and Prose of William Blake. Newly rev. ed. New York: Anchor–Random House, 1988. [Abbreviated as E.]

Essick, Robert N. “Blake’s Engravings to the Book of Job: An Essay on Their Graphic Form with a Catalogue of Their States and Printings.” Bindman 35-101.

---. “Blake’s Job: Some Unrecorded Proofs and Their Inscriptions.” Blake 19.3 (winter 1985–86): 96-102.

---. William Blake, Printmaker. Princeton: Princeton University Press, 1980.

Gilchrist, Alexander. Life of William Blake, “Pictor Ignotus.” 2 vols. London: Macmillan and Co., 1863.

Kutter, Suzanne, and Donald Gallup. “A Check-List of the [Rosenbloom] Bequest.” Yale University Library Gazette 49.4 (April 1975): 311-43.

Liebert, Herman W. “The Charles J. Rosenbloom Bequest.” Yale University Library Gazette 49.4 (April 1975): 309-10.

Lindberg, Bo. William Blake’s Illustrations to the Book of Job. Åbo: Åbo Akademi, 1973.

Paintings and Prints from the Collection of Charles J. Rosenbloom … February 7–March 24 1946. [Pittsburgh]: Carnegie Institute, [1946].

Palmer, A. H. The Life and Letters of Samuel Palmer. London: Seeley & Co., 1892.

Sung, Mei-Ying. “Technical and Material Studies of William Blake’s Engraved Illustrations of the Book of Job (1826).” Diss. Nottingham Trent University, 2005.

---. William Blake and the Art of Engraving. London: Pickering & Chatto, 2009.



Individual Subscriber?
Log In

Department of English
© Blake/An Illustrated Quarterly