A technical examination of William Blake’s Virgil wood engravings

The NGV is perhaps best known to William Blake scholars for the magnificent suite of watercolours illustrating Dante’s Divine Comedy that the Gallery purchased from the John Linnell sale at Christies in London in 1918. The NGV also holds a composite group of fourteen wood engravings Blake designed and engraved for Dr Robert Thornton’s Pastorals of Virgil, which were purchased in London in 1959 and once formed part of Linnell’s collection as well.¹ For more information about the history of the NGV’s impressions, see Louise Wilson, ‘William Blake and the NGV: the history of the Virgil wood engravings’, National Gallery of Victoria, https://www.ngv.vic.gov.au/essay/william-blake-and-the-ngv-the-history-of-the-virgil-wood-engravings/, accessed 30 June 2025.

Much has been written about the history of Blake’s wood engravings, so only a brief overview will be provided here. At John Linnell’s suggestion, Thornton commissioned Blake to produce wood engravings for his third edition of The Pastorals of Virgil (1821), a schoolbook designed to teach young boys Latin. Blake’s task was to illustrate Ambrose Philips’s imitation of Virgil’s first eclogue, which follows two shepherds – the old, sagacious Thenot and the young, brooding Colinet – in various rural settings. Working in a pocket sketchbook, he created preparatory drawings in pencil, pen and ink, and wash, that he later copied with ink onto pieces of boxwood in readiness for engraving.² Blake’s uncut woodblock of The prophet Isaiah foretelling the Crucifixion and the Ascension, now in the British Museum, shows his technique of drawing lines with ink directly on the boxwood as a guide for cutting. I am indebted to Dr Bethan Stevens, for showing me this woodblock. See The William Blake Archive, https://blakearchive.org/copy/pid?descId=but773.1.penink.01, accessed 26 February 2025. (see fig. 1).

Fig. 1: William Blake, illustration for Thornton’s Virgil, And unyok’d heifers, loitering homewards, low. Image courtesy of the Princeton University Art Museum

Boxwood was the preferred timber for wood engraving because it is particularly hard, making it possible to achieve very precise and finely cut designs. Generally, imperfections in the wood were removed by a boxwood preparer who would ensure it was perfectly dried prior to cutting to size and polishing the surface to a smooth finish.³ Thomas Gilks, The Art of Wood Engraving: A Practical Handbook, Winsor and Newton, London, 1866, p. 26. However, technical examination of the Virgil wood blocks suggests they were not professionally prepared because they have an uneven, slightly rough surface.⁴Iain Bain & David Chambers, ‘The printing of the blocks: a technical note’, in The Wood Engravings of William Blake, British Museum Publications, London, 1977, p. 9.

To create his wood engraving, Blake used boxwood that had been cut across the growth rings. Known as the end grain, this feature distinguishes wood engravings from woodcuts, which are carved into timber cut longitudinally along the grain. Using familiar tools for engraving on metal, Blake cut away the compositional lines on the boxwood and printed proofs in black ink on China paper using his rolling press. The growth rings of the end grain are sometimes visible as lines of reduced printing on wood engravings, as is the case with Blake’s proof sheet now at the Fitzwilliam Museum.⁵ China paper is believed to have first come to the United Kingdom as the lining of timber chests carrying tea. It is sometimes referred to as ‘India’ paper, since the tea was imported by the British East India Company; however, this is a misnomer. P. Jenkins, ‘Plate papers and India Proof Prints’ in The Quarterly Journal of the British Association of Paper Historians, Dec. 1994, p. 1. (see fig. 2).

Fig. 2: William Blake, proof sheet showing curved lines of reduced printing corresponding with the concentric growth rings of the boxwood. Featuring plates 6, 7, 8 and 9. Photo: © Fitzwilliam Museum, Cambridge

When the woodblocks were prepared for publication, they were cut down, presumably to enable four images to fit per page using the chosen paper size with room for titles underneath each. The impressions in Thornton’s schoolbook would have been printed on a hand press, which is ideally suited to relief processes like wood engraving. The printer would have positioned four of the small woodblocks with a title in letterpress type under each one. The type and images were then wedged tightly together in a forme so they didn’t move during the printing process.⁶The individual blocks and letterpress type would have been put in a small frame-like structure called a chase, with any vacant space around the blocks packed out with small pieces of timber called ‘furniture’. Quoins (expandable metal locks) would have been placed within the chase and opened with a key until the blocks were locked tight. This locked up unit is referred to as a ‘forme’. Peter Marsh, ‘Letterpress printing’, Museums Victoria Collections, ‘Pearl Printing Press’, , accessed 12 Apr. 2020. While bound copies of Thornton’s Virgil are highly collectable and expensive today, at the time of publication it was a simple textbook, to be printed within a budget. To keep costs down, a variety of poor-quality papers (both machine and handmade) were used.⁷ I noted the variety and quality of paper used while examining a copy (1863, 1114.299) at the British Museum, on 25 Jul. 2019. Generally, impressions in the work are not fine quality, possibly because of the speed of production and the paper.

Shortly after the book was published, Linnell purchased the woodblocks for his collection and later commissioned fellow artist Edward Calvert to print sets.⁸ According to John Linnell’s diary for 8 Sep. 1828, Calvert brought impressions of the Virgil wood engravings to Linnell, which Calvert had printed from the original blocks. Raymond Lister, Edward Calvert, G. Bell, London, 1962, p. 24. Linnell had eight children, five of whom are known to have been artists.⁹ His sons – John, James, William and Thomas– are all recorded in census records as artists. He commissioned his daughter Hannah to copy pictures for him while she was in Italy with her new husband, artist Samuel Palmer. Raymond Lister, The Letters of Samuel Palmer, Volume 1, 1974, p. 94. The woodblocks remained in the Linnell family for over a century with various family members undertaking printing sessions over this time.¹⁰ Personal communication with Professor Robert N. Essick, University of California, 16 Sep. 2016, and Nick Lott, Larkhall Fine Art, Bath, 7 May 2016.

Four different paper types (indicating at least four different printing sessions) were used to produce the NGV’s impressions – two subtly different China papers and two Western papers, one made on a laid mould and the other on a wove mould. (see figs 3, 4, 5 and 6)

Fig. 3: William Blake, Menalcas’ yearly wake, printed on laid Western paper; viewed by shining light through the sheet (transmitted light) showing the slightly transparent laid lines running vertically down the paper support.¹¹ The titles assigned to the Virgil wood engravings vary from institution to institution. Those used for the NGV impressions are taken from the most recent NGV Blake catalogue: Cathy Leahy, William Blake, National Gallery of Victoria, Melbourne, 2014.

Fig. 4: William Blake, Colinet resting at Cambridge by night, printed on Western wove paper; viewed under transmitted light, showing the fine woven texture of the paper

Fig. 5: William Blake, With songs the jovial hinds return from plow, printed on China paper with closely spaced transparent lines running horizontally across the sheet and vertically down the sheet at less regular intervals; viewed under transmitted light

Fig. 6: William Blake, Thenot remonstrates with Colinet, Lightfoot in the distance, printed on China paper without transparent lines; viewed under transmitted light

Perhaps most striking is the range of printing imperfections evident, potentially showing different hands and studio practices at work. To better understand these faults and to use them to identify printing sessions, I consulted eighteenth- and nineteenth-century printing treatises and sought other Virgil wood engravings for comparison. The NGV holds the only nineteenth-century impressions of these wood engravings in Australia. With no comparable material to examine locally, I travelled to the United Kingdom to study a variety of material related to the prints, including Blake’s preparatory drawings, proofs and original woodblocks; bound copies of the Pastorals of Virgil; over 150 impressions of the wood engravings and documents pertaining to the woodblocks.¹² This travel opportunity was made possible through the generous support of Peter Clemenger AO and Joan Clemenger AO.

To gain an insight into Calvert’s printing technique and choice of materials, I examined numerous impressions of his own wood engravings. Calvert is best known for his intricate wood engravings of pastoral idylls, which are heavily influenced by Blake’s Virgil designs. He ceased producing his own meticulous prints around 1831 and is unlikely to have been printing from the Virgil woodblocks after this date.¹³ Martin Butlin & Ted Gott, William Blake in the Collection of the National Gallery of Victoria, 1989, National Gallery of Victoria, Melbourne, p. 138. An extraordinary album of prints and drawings assembled by Alexander Constantine Ionides contains a complete set of Virgil wood engravings and eleven of Calvert’s fifteen pastoral compositions.¹⁴ Mark Evans, ‘Blake, Calvert – and Palmer? The album of Alexander Constantine Ionides’, The Burlington Magazine, vol. 144, no. 1194, 2002, p. 541. (see fig. 7). Ionides became acquainted with Calvert around 1828, when the latter was producing Virgil impressions for Linnell, and it has been suggested that the Virgil wood engravings in the album are from this printing session.¹⁵ibid. The scrapbook is therefore the perfect starting point for comparing Calvert’s materials and methods – in impressions of both his own and the Virgil wood engravings – with Virgil impressions in other collections.

Fig 7: Alexander Constantine Ionides Album, fol. 2r. Photo © Victoria and Albert Museum

The contents of the Ionides album are adhered to the pages, limiting the scope for paper examination. Despite this impediment, it was possible to discern that most of the paper supports used to print the Virgil and Calvert wood engravings bear the hallmarks of China paper. This very thin paper support, which many Virgil impressions that I examined are printed on, is produced from bamboo fibre. The addition of clay filler creates a beautiful opacity, and the lack of sizing makes the end-product highly absorbent of ink and yielding to woodblocks.¹⁶ Kimberley Shenk, ‘The role of China paper in nineteenth-century French printmaking’, in Looking at Paper: Evidence & Interpretation Symposium Proceedings, Toronto, 1993, pp. 33–4. The addition of clay, which is hygroscopic, makes China paper vulnerable to foxing, which is initiated by moisture. Many of the impressions examined (including those in the NGV collection) have undergone aqueous treatment to reduce foxing stains. This type of treatment can cause subtle changes to the original character of the paper, such as altering the dimensions and thickness. It is easily identified, even on very small works like the Virgil wood engravings because it commonly contains the marks of its making, such as tiny spherical pits or inclusions resulting from the addition of clay. Under raking light (a strong light source shining across the paper), distinct linear indentations can be observed on one side. These surface imperfections are attributed to the way Chinese papermakers attach newly formed sheets to a drying wall with a coarse brush, the hairs impressing lines on the surface of the wet sheet.¹⁷ibid. (see fig. 8).

Fig. 8: William Blake, Colinet departs in sorrow, a thunder-scarred tree on the right, verso, viewed under raking light, showing diagonal linear striations across the China paper support.

While most prints in the Ionides album are printed on China paper, at first sight, the impressions of With songs the jovial hinds return from plow, Colinet’s fond desire strange lands to know, Colinet mocked by two boys and Calvert’s The ploughman appear to be printed on Western laid paper. This finding is not entirely surprising, since there are Virgil impressions on Western laid paper, such as the NGV’s Menalcas’ yearly wake. Laid paper is made on a timber frame (called a mould) with a sieve-like surface consisting of closely spaced brass wires running horizontally across the frame (the indentations these leave on the paper are called laid lines) and more widely spaced wires running vertically that are twisted around the laid wires to hold them in place (the indentations these leave are known as chain lines).¹⁸ Although the laid lines are oriented horizontally on the papermaking mould, when the paper is used, the artist might turn the paper so that these lines are oriented vertically (see figs. 3 and 10) (see fig. 9).

Fig. 9: Laid papermaking mould, made by the late Serge Pirard.¹⁹ This traditional papermaking mould was purchased using the NGV’s Supporters of Conservation Projects fund and is part of the NGV’s Conservation Material Archive.

The laid lines, when viewed under raking light, form a linear texture on the wire side (like the surface of corduroy fabric). (see fig. 10). The verso of the sheet, the felt side, is relatively smooth.

Fig. 10: William Blake, Menalcas’ yearly wake, viewed under raking light, showing the ridges between laid lines running vertically down the paper support.

As I studied more Virgil impressions and Calvert wood engravings in other collections, I noted the appearance of laid lines on numerous prints. Some of these prints are not adhered to secondary supports making it possible to examine them using transmitted light. When light was shone through them, the distinctive alternating density of laid lines was not visible, and it became clear that they are printed on China paper, not Western laid paper as their appearance suggests.²⁰ In addition to the impressions in the Ionides album at the Victoria and Albert Museum, I found many examples at the Fitzwilliam Museum, British Museum, and the National Gallery of Scotland. I thank Harry Metcalf, Paper Conservator, Fitzwilliam Museum, for carefully examining the proof, first state of Calvert’s The Ploughman (P.605-1985) and confirming that it is printed on China paper. There will of course be cases where impressions, whether adhered to secondary supports or not, appear to be on laid paper and indeed are. Under magnification, it was evident that the lines are areas of reduced printing seemingly related to the printing process rather than the paper support. (see fig. 11). The spacing of the lines was found to be the same frequency as laid lines of many Western papers,²¹ I am grateful to Yvonne (Bonnie) Hearn, former Sherman Fairchild Conservation Fellow, Thaw Conservation Centre, The Pierpont Morgan Library & Museum, currently Conservator of Paper at the NGV, for measuring lines of reduced printing on Calvert’s Edward. The Ploughman (1974.50-2) at the Morgan. which confirmed laid paper was in close contact with the primary support during the printing process. For example, it could have been used as an interleaving to help freshly printed impressions to dry, though this is improbable because China paper, highly absorbent of ink, does not require humidification prior to printing. Moreover, examination did not reveal any signs that interleaving had been removed from the recto of the prints, such as deposition of fibres on inked areas. The more likely explanation is that laid paper was placed on top of the China paper during printing, perhaps to add cushioning or assist with handling the thin primary support, resulting in slightly reduced contact between the inked block and the China paper in areas corresponding with the furrows of the laid lines.

Fig. 11: Edward Calvert, The ploughman, showing horizontal lines of reduced printing. Wood engraving, © The Fitzwilliam Museum, Cambridge.

In 1828, Calvert lent his press to Linnell, noting that ‘its great defects are in being too small and being altogether a makeshift’.²² Lister, Edward Calvert, p. 24. A ‘makeshift’ press for printing could refer to one designed for another purpose, such as a wine press. It is difficult to discern whether he is referring to a type of hand press or a rolling press; potentially either could have produced impressions with the lines of reduced printing that I observed. In both scenarios, the inked matrix would have been placed face up on the bed of the press and covered first with the paper to be printed, followed by a piece of wastepaper. In the case of Calvert’s prints with lines of reduced printing, the wastepaper was probably laid paper, with the wire side in direct contact with the verso of the printing paper, and the impressions would have been pulled using moderate pressure.²³ Intense pressure would have flattened the texture of the laid paper, resulting in more even contact between the woodblock and the China paper and thus more uniform printing. Not all Calvert prints bear this printing imperfection, because when the wastepaper required replacement, either laid paper – with the felt side in contact with the printing paper – or wove paper could have been used.

Linnell is known for his landscape and portrait paintings and engravings.²⁴24 For example, the NGV’s Wheat, painted in 1860. Like many artists, he regularly had his engravings printed professionally, negating the need for a printing press of his own.²⁵ Joseph Viscomi, Blake and the Idea of the Book, Princeton University Press, Princeton, 1993, p. 103. He purchased a few copper plates from Hughes & Kimber, who supplied presses and materials for letterpress, lithographic and intaglio printing. They were also copperplate printers, so it is possible that he used this service. National Portrait Gallery, British Artists’ Suppliers 1650-1950-H, , accessed 19 Nov. 2019. At times, however, there was at least one printing press at his studio in Cirencester Place, Fitzroy Square. The year before Linnell borrowed Calvert’s press, Catherine Blake had moved to Cirencester Place, bringing her late husband’s rolling press with her.²⁶ G. E. Bentley Jr, Blake Records, 1969, p. 351. The press remained there from August 1827 until spring 1828, when Catherine relocated to Charles Heathcote Tatham’s office and studio.²⁷ Joseph Viscomi, ‘Posthumous Blake: the roles of Catherine, C. H. Tatham and Frederick Tatham in Blake’s afterlife,’ in Blake/An Illustrated Quarterly, vol. 53, fall 2019, pp. 1–59.

The NGV’s impressions of Sabrina’s silvery flood and Thenot and Colinet sup together both have imperfections indicating they were printed on a rolling press: isolated areas of unintentional printing on their margins, referred to as ‘foul ink’. (see figs. 12 and 13). This error can occur if a rolling press is used to print thick woodblocks without the necessary adjustments being made. Blake’s Virgil woodblocks are approximately 2cm thick, so the height of the rollers should have been raised and printing blankets staggered to create a gradual incline for the rollers to travel along.²⁸ A printing blanket is made of felted wool. The slight pliability of the blanket helps to achieve good contact between the printing matrix and the paper to be printed on. The inked matrix would have been placed face up on the bed of the press and covered first with the paper to be printed, followed by a piece of wastepaper, and a printing blanket would have been laid over the top of this sandwich. While there is no way of knowing if these impressions come from the press that belonged to Blake, the presence of foul ink shows that the precautions required for printing them on a rolling press were not taken; as a result, a small jolt occurred as the woodblocks passed under the rollers, shifting the paper slightly and accidentally transferring ink to the margins.²⁹ William Faithorne, The Art of Graving and Etching, 2nd ed., 1702, p. 57.

Fig. 12: William Blake, Thenot and Colinet sup together, showing ‘foul ink’ on the lower margin

Fig. 13: William Blake, Sabrina’s silvery flood, showing ‘foul ink’ on the upper margin

The unique and beautiful NGV impression of Unyoked heifers loitering homeward, low has an unusual black square in the lower left margin, signalling that it was printed at the same time as another block (as was the case when the woodblocks were printed for Thornton’s Pastorals of Virgil). (see fig. 14). Since formes are not commonly used with a rolling press and it would have been very difficult to print multiple blocks at the same time without one, I assume that this impression comes from a hand press.

Fig. 14: William Blake, Unyoked heifers loitering homeward, low, showing partial printing of another wood block on the lower left corner

I examined numerous impressions where the ink was too heavily applied in some passages, leading to loss of definition, and under-applied in others, creating a hazy, indistinct effect (see fig. 15). There are several potential causes: inconsistent pressure when printing, because of the uneven height of the woodblock (unlikely, since the impressions that I studied did not exhibit reduced printing in the same location on each image);³⁰ This issue can be rectified by a process called ‘making right’, where the back of the woodblock is gradually built up by adhering fragments of paper or card, shaped identically, to the areas that are under-printing. uneven dampening of the paper prior to printing (improbable for impressions on China paper, as previously noted); or, the most likely cause in the case of the Virgil wood engravings, uneven application of ink.³¹ Personal communication with Michael Phillips, 2 Aug. 2019 In the nineteenth century, both daubers and composition rollers were tools for applying ink to a matrix, with daubers being the traditional implement.³² Daubers are also called ink balls or dabbers. They are the shape of a muller and can be made by rolling wool or cloth into a ball and covering it tightly with linen or leather to form a round pad. A wooden handle is bound into the open end. It is probable that, like Blake, Calvert and the Linnell family used a dauber to apply ink to the Virgil woodblocks. Achieving an even application requires considerable skill, the ink needing to be built up with consistently thin sequential layers on the surface of the matrix.

Fig. 15: William Blake, The good shepherd chases away the wolf, showing under-inking on lower left corner and over-inking in areas of the tree on the right

Menalcas’ yearly wake (see fig. 16) and Tate’s impression of the same image, For him our yearly feasts and wakes we hold (Tate, A00123) were printed with ink that does not have a well-balanced ratio between pigment and oil, leaving an oily halo. This imperfection suggests the application of an ink that was not commercially produced. Printing treatises available during Blake’s time describe the making of ink at length and discuss the use of ‘weak’ and ‘strong’ oil combined with dry pigment to create black printing ink.³³ Faithorne, pp. 62–4. Weak oil is not heated to the same extent as strong oil, resulting in differences in their flowing qualities – weak oil is more mobile and ideally suited to intaglio techniques, where it migrates down into the incised lines, while strong oil is more viscous, producing an ink with short tack that sits on the surface of the matrix and is ideally suited to relief techniques like wood engraving.³⁴ Personal communication with Adrian Kellett, 11 Oct. 2019. The appearance of the ink in these impressions reveals that weak rather than strong oil was added. Linnell produced intaglio prints and was probably more familiar with the type of ink required; Calvert was more focussed on wood engraving and would have known the nuances of preparing ink for that technique. This suggests that these two prints are more likely to have been created by a member of the Linnell family.

Fig. 16: William Blake, Menalcas’ yearly wake showing the blotchy appearance of the image relating to excess weak oil being used to make the ink

Menalcas’ yearly wake also has marked blind embossing on its verso indicating that the pressure exerted during printing was much greater than that used to pull Tate’s impression (see fig. 16).³⁵ Tate’s impressions have undergone aqueous treatment, which could have reduced blind embossing but is unlikely to have removed it completely. I am grateful to Rosie Freemantle, former paper conservator, Tate, for sharing details of the treatment history of Tate’s prints. It was probably printed on a rolling press where the pressure is easily controlled. Printers sometimes increase pressure to achieve a second pull, meaning that they do not reink the matrix and instead try to extract any residual ink.³⁶ Personal communication with Michael Phillips, 2 Aug. 2019. The NGV’s print is on a medium-weight, Western laid paper support, which is not ideally suited to achieving a good impression from a woodblock. It was possibly created as the first stage of cleaning the block after a printing session, a process requiring great care to avoid compromising the clarity of the engraved lines. Pulling prints using scrap paper lying around the studio is a quick, easy, and common method of removing the bulk of the remaining ink prior to gentle wiping with a soft cloth.³⁷ Personal communication with Ros Atkins, 27 Oct. 2019. The resulting prints are referred to as maculature impressions.

Fig. 16: William Blake, Menalcas’ yearly wake, viewed from the verso under raking light showing blind embossing

Many of the impressions that I studied have tiny dots of ink within fine compositional lines. For example, the sky region of the NGV’s Colinet mocked by two boys and Tate’s impression of the same work, entitled Colinet with his shepherd’s pipe mocked by two boys (Tate, A00122), have numerous tiny dots within identical lines (see fig 17). They may have been produced during the same printing session, or the blocks may not have been inadequately cleaned between sessions, allowing minute nuggets of ink to dry.³⁸ Personal communication with Nick Lott, 7 May 2016.

Fig. 17: William Blake, Colinet mocked by two boys showing tiny nuggets of dried ink trapped between engraved lines around the figure of the boy on the left

The NGV’s Unyoked heifers loitering homeward, low is very lightly printed and extensively touched with bistre ink (see fig. 14).³⁹ Infrared examination confirmed the ink is bistre, a brown ink made from chimney soot. Like Menalcas’ yearly wake, it may well be a maculature impression, casually produced as a means of cleaning the block at the end of a printing session; rather than casting it away, somebody, perhaps Blake himself or one of the Linnells, applied delicate washes of ink in a manner similar to Blake’s preparatory drawings. (see fig. 1). Blake and his wife, Catherine, are known to have touched prints with ink, as are members of the Linnell family, the purpose being to either address a printing imperfection or enhance the composition.⁴⁰ The woodblocks of Thenot remonstrates with Colinet, Blasted tree and blighted crops and Colinet departs in sorrow, a thunder-scarred tree on the right were cut down roughly, leaving an upper corner angled. On an impression of Thenot remonstrates with Colinet (Fitzwilliam Museum, P.59-1950), pencil has been used to continue the composition into the margin, creating a right-angled corner. An impression of Blasted tree and blighted crops in the collection of Michael Phillips has been touched with ink on the upper-left corner to complete the composition. Michael Phillips, William Blake: Apprentice and Master, 2014, p. 193. Linnell or his sons are credited with hand-colouring seven Virgil wood engravings; Hannah Palmer (née Linnell), who coloured a set of Linnell’s lithographs, could also have had a hand in them.⁴¹ Lister, The Letters of Samuel Palmer, vol. 1, 1974, p. 94; Unknown author, Catalogue of the John Linnell Collection of Highly Important Works by William Blake Obtained Direct from the Artist and Numerous Prints, Books in Colour, Autographs, &c. The Remaining Pictures and Drawings by John Linnell, Sen. and Pictures of Old Masters, 1918, p. 10; and G. Ingli James, ‘Blake’s woodcuts illuminated’, in Apollo, vol. 94, 1974, pp. 194–5. Linnell’s set of lithographs after Michelangelo were listed in the 1918 sale catalogue as being coloured in Italy by Mrs. Samuel Palmer. I am grateful to Lauren Lott for showing me the Linnell catalogue and the hand-coloured set of lithographs. The artist responsible for the beautiful brushwork on Unyoked heifers loitering homeward, low is unlikely to have been Calvert, as he was strongly against touching wood engravings with ink.⁴² Lister, 1974, p. 18.

Technical examination of the Virgil wood engravings in the NGV and comparison with examples in other collections has provided insights into how imperfections can help us to understand the method in which an impression was produced. The NGV’s composite group includes impressions printed with different presses, papers and inks. Several that were perhaps destined for the dustbin were retained, and one was extensively touched with ink and transformed into a work of great beauty. Examination of works in collections in the United Kingdom revealed a subtle quirk of Calvert’s printing technique that can aid identification of Virgil impressions printed by him. It is now possible to appreciate the NGV’s wood engravings in their broader context and to recognize that, while many of them may not be the finest impressions, they provide a valuable window into the complex history of Blake’s Virgil woodblocks.

An earlier version of this essay was published in Blake/An Illustrated Quarterly, vol. 56, no. 2, fall 2022.

Notes