Decision Trees for Watermark Identification in Rembrandt’s Etchings

Erik Hinterding’s classification and chronological arrangement of hundreds of distinct watermarks in Rembrandt’s prints has resulted in a deeper understanding of the artist’s production methods and his artistic development as a prolific etcher. This article explains the decision-tree-based approach to rapid identification of watermarks in Rembrandt’s etchings that is under development at Cornell University and uses it to identify the watermarks and watermark fragments in seven Rembrandt prints in The Frick Collection in New York.

DOI: 10.5092/jhna.2020.12.2.5
Twin molds with dashed lines indicating the cutting up of full broadsheets into separate sheets of paper, all of which are batchmates
Fig. 1 Twin molds with dashed lines indicating the cutting up of full broadsheets into separate sheets of paper, all of which are batchmates (illustration by author) [side-by-side viewer]
Transmitted-light photograph of a blank sheet of laid paper from a 1536 French book. Left: Right half of a full-mold sheet containing Foolscap with Five-pointed Collar watermark. Right: Enlargement of region around watermark with chain lines and laid lines
Fig. 2 Transmitted-light photograph of a blank sheet of laid paper from a 1536 French book. Left: Right half of a full-mold sheet containing Foolscap with Five-pointed Collar watermark. Right: Enlargement of region around watermark with chain lines and laid lines (sheet from study collection of Conservation Center in the Institute of Fine Arts provided by Margaret Hoben Ellis (New York University); photographs provided by S. Schlick and P. Messier (Yale)) [side-by-side viewer]
Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, c. 1652
Fig. 3 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Herbert F. Johnson Museum, Cornell University, Ithaca, NY, etching, drypoint and engraving on paper, 21.6 x 28.6 cm (sheet), B56vi(?), 84.010, c. 1652 [side-by-side viewer]
Transmitted-light image showing watermarkRembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers,
Fig. 3-1 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Transmitted-light image showing watermark [side-by-side viewer]
Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Excerpt from beta-radiograph
Fig. 3-2 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Excerpt from beta-radiograph [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Trees, The Frick Collection, 1915.3.28. Right: Beta-radiograph of portion around watermark
Fig. 4 Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Trees, The Frick Collection, 1915.3.28. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis {NYU} and Reba Fishman Snyder {Morgan}) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Jan (Johannes) Lutma the Elder, The Frick Collection, 1916.3.37; Right: Beta-radiograph of portion around watermark
Fig. 5 Left: Rembrandt Harmenszoon van Rijn, Jan (Johannes) Lutma the Elder, The Frick Collection, 1916.3.37; Right: Beta-radiograph of portion around watermark (Provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Haybarn and Flock of Sheep, The Frick Collection, 1916.3.29. Right: Beta-radiograph of portion around watermark
Fig. 6 Left: Rembrandt Harmenszoon van Rijn, Landscape with Haybarn and Flock of Sheep, The Frick Collection, 1916.3.29. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, The Goldweigher’s Field, The Frick Collection, 1915.3.31. Right: Beta-radiograph of portion around watermark
Fig. 7 Left: Rembrandt Harmenszoon van Rijn, The Goldweigher’s Field, The Frick Collection, 1915.3.31. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Clement de Jonghe, The Frick Collection, 1916.3.36. Right: Beta-radiograph of portion around watermark
Fig. 8 Left: Rembrandt Harmenszoon van Rijn, Clement de Jonghe, The Frick Collection, 1916.3.36. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, St. Francis Praying Beneath a Tree, The Frick Collection, 1916.3.35. Right: Beta-radiograph of portion around watermark
Fig. 9 Left: Rembrandt Harmenszoon van Rijn, St. Francis Praying Beneath a Tree, The Frick Collection, 1916.3.35. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Gabled Cottages, The Frick Collection, 1916.3.30. Right: Beta-radiograph of portion around watermark
Fig. 10 Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Gabled Cottages, The Frick Collection, 1916.3.30. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Four variants of Foolscap with Five-pointed Collar. Left to right: A.a.a., H.b.a., K.a.a., N.a.b
Fig. 11 Four variants of Foolscap with Five-pointed Collar. Left to right: A.a.a., H.b.a., K.a.a., N.a.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Four subvariants of Foolscap with Five-pointed Collar, variant K. Left to right: K.a.a., K.c.a., K.e.a., K.f.a
Fig. 12 Four subvariants of Foolscap with Five-pointed Collar, variant K. Left to right: K.a.a., K.c.a., K.e.a., K.f.a (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two twins of Foolscap with Five-pointed Collar, subvariant A.a. Left to right: A.a.a., A.a.b
Fig. 13 Two twins of Foolscap with Five-pointed Collar, subvariant A.a. Left to right: A.a.a., A.a.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Feature labels for Foolscap with Five-pointed Collar
Fig. 14 Feature labels for Foolscap with Five-pointed Collar (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree for four variants of Foolscap with Five-pointed Collar in figure 11
Fig. 15 Decision tree for four variants of Foolscap with Five-pointed Collar in figure 11 (illustration by author) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Foolscap with Five-pointed Collar in Rembrandt’s etchings
Fig. 16 Decision tree by the WIRE Project from type to subvariant for Foolscap with Five-pointed Collar in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Flowchart for watermark identification
Fig. 17 Flowchart for watermark identification (illustration by author) [side-by-side viewer]
Left to right: Watermark under investigation, A.a.a.,  A.a.b
Fig. 18 Left to right: Watermark under investigation, A.a.a.,  A.a.b (middle and right images courtesy of Sound and Vision Publishers BV; red markings by author) [side-by-side viewer]
Type to subvariant path through decision tree to Foolscap with Five-pointed Collar A.a in Rembrandt’s etchings
Fig. 19 Type to subvariant path through decision tree to Foolscap with Five-pointed Collar A.a in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the collar have 4 points? Left: YES. Right: NO
Fig. 20 Does the collar have 4 points? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Are the roundels in a pyramid configuration (one on top and two on bottom)? Left: YES. Right: NO
Fig. 21 Are the roundels in a pyramid configuration (one on top and two on bottom)? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Are the peaks of the caps divided by stripes? Left: YES. Right: NO
Fig. 22 Are the peaks of the caps divided by stripes? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the collar point below the face bend underneath the chin? Left: YES. Right: NO
Fig. 23 Does the collar point below the face bend underneath the chin? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the “4” point in the same direction as the face? Left: YES. Right: NO
Fig. 24 Does the “4” point in the same direction as the face? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment test (watermark of interest over Foolscap with Five-pointed Collar A.a.a) and match
Fig. 25 Chainspace alignment test (watermark of interest over Foolscap with Five-pointed Collar A.a.a) and match (images courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Foolscap with Five-pointed Collar N.a.b
Fig. 26 Left: Watermark under investigation. Right: Foolscap with Five-pointed Collar N.a.b (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Path from type to subvariant of Foolscap with Five-pointed Collar N.a
Fig. 27 Path from type to subvariant of Foolscap with Five-pointed Collar N.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment test (Foolscap with Five-pointed Collar N.a.b over watermark of interest) and match
Fig. 28 Chainspace alignment test (Foolscap with Five-pointed Collar N.a.b over watermark of interest) and match (images courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Path from type to subvariant of fragment of Foolscap with Five-pointed Collar K.a
Fig. 29 Path from type to subvariant of fragment of Foolscap with Five-pointed Collar K.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark fragment under investigation; Right: Foolscap with Five-pointed Collar K.a.a
Fig. 30 Left: Watermark fragment under investigation; Right: Foolscap with Five-pointed Collar K.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of Foolscap with Five-pointed collar K.a.a (top) and watermark of interest and match
Fig. 31 Chainspace alignment of Foolscap with Five-pointed collar K.a.a (top) and watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; composition by author) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Paschal Lamb in Rembrandt’s etchings
Fig. 32 Decision tree by the WIRE Project from type to subvariant for Paschal Lamb in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Feature labels for Paschal Lamb
Fig. 33 Feature labels for Paschal Lamb (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Type to subvariant path for Paschal Lamb fragment subvariant A.b
Fig. 34 Type to subvariant path for Paschal Lamb fragment subvariant A.b (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Paschal Lamb A.b
Fig. 35 Left: Watermark under investigation. Right: Paschal Lamb A.b (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of Paschal Lamb A.b over watermark of interest and match
Fig. 36 Chainspace alignment of Paschal Lamb A.b over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Top: Watermark under investigation Bottom: RC
Fig. 37 Top: Watermark under investigation Bottom: RC (Countermark) RC.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of RC.a.a excerpt over watermark of interest and match
Fig. 38 Chainspace alignment of RC.a.a excerpt over watermark of interest and match (image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
One sample from each IHS variant. Top row (left to right): A.a, B.a.a. Bottom row (left to right): C.a.a, D.a.a., E.b
Fig. 39 One sample from each IHS variant. Top row (left to right): A.a, B.a.a. Bottom row (left to right): C.a.a, D.a.a., E.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two possible decision-tree realizations for classifying variants of IHS
Fig. 40 Two possible decision-tree realizations for classifying variants of IHS (illustrations by author) [side-by-side viewer]
Subvariants of IHS variant E: Left: E-a. Right: E-b
Fig. 41 Subvariants of IHS variant E: Left: E-a. Right: E-b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Samples of each subvariant of IHS variant B: Top row (left to right): B’.a, B.a.a, B.b.a. Bottom row (left to right): B.c, B.d, B.e.a
Fig. 42 Samples of each subvariant of IHS variant B: Top row (left to right): B’.a, B.a.a, B.b.a. Bottom row (left to right): B.c, B.d, B.e.a (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two possible decision-tree branches for identifying subvariants of HIS variant B
Fig. 43 Two possible decision-tree branches for identifying subvariants of HIS variant B (illustrations by author) [side-by-side viewer]
Some subvariants of IHS variant A: Top row (left to right): A.a, A.b, A.c. Bottom row (left to right): A.d.a, A.e
Fig. 44 Some subvariants of IHS variant A: Top row (left to right): A.a, A.b, A.c. Bottom row (left to right): A.d.a, A.e (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Some more subvariants of IHS variant A: Top row (left to right): A.f.a, A.g, A.h. Bottom row (left to right): A.i, A.
Fig. 45 Some more subvariants of IHS variant A: Top row (left to right): A.f.a, A.g, A.h. Bottom row (left to right): A.i, A.j (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Feature labels for IHS
Fig. 46 Feature labels for IHS (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for IHS in Rembrandt’s etchings
Fig. 47 Decision tree by the WIRE Project from type to subvariant for IHS in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Path from type to subvariant B.a of IHS Countermark
Fig. 48 Path from type to subvariant B.a of IHS Countermark (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Countermark under investigation Right: IHS B.a.a
Fig. 49 Left: Countermark under investigation Right: IHS B.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of IHS B.a over watermark of interest and match
Fig. 50 Chainspace alignment of IHS B.a over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Feature labels for Strasbourg Lily
Fig. 51 Feature labels for Strasbourg Lily (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Strasbourg Lily in Rembrandt’s etchings
Fig. 52 Decision tree by the WIRE Project from type to subvariant for Strasbourg Lily in Rembrandt’s etchings (illustration by Isabella M. Dobson (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Strasbourg Lily E’.a.a
Fig. 53 Left: Watermark under investigation. Right: Strasbourg Lily E’.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Strasbourg Lily decision-tree branch to variant E’.a
Fig. 54 Strasbourg Lily decision-tree branch to variant E’.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment of Strasbourg Lily E’.a.a over watermark of interest and match
Fig. 55 Chainspace alignment of Strasbourg Lily E’.a.a over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
  1. 1. From chapter 5 of Neil Harris, Paper and Watermarks as Bibliographical Evidence, 2nd ed. (Lyon: Institut d’histoire du livre, 2017), http://ihl.enssib.fr/en/paper-and-watermarks-as-bibliographical-evidence

  2. 2. Louisa Wood Ruby, Director of Research, Frick Art Reference Library, to the author, July 24, 2019.

  3. 3. Plus a deckel, which is a removable wooden rim used by both molds in the twin pair that fits on top of each mold, contains the pulp, and sets the size of the sheet.

  4. 4. A video documenting an attempt to recreate the pre-1800 papermaking process is a solid introduction: Tim Barrett, “Chancery Papermaking at the University of Iowa’s Center for the Book,” filmed 2013 at the University of Iowa, video, 11:24, http://now.uiowa.edu/2013/05/center-book-releases-short-film-about-papermaking. A compact verbal description of the laid papermaking process appears in Cathleen Ann Baker, From the Hand to the Machine: Nineteenth-Century American Paper and Mediums: Technologies, Materials, and Conservation (Ann Arbor, MI: Legacy Press, 2010), 34: “Once the fibers were beaten to a particular degree and any non-fibrous additives, such as an internal sizing agent, color, and/or filler, were added, the formation of the paper remained the same: pulp was deposited onto the surface of a sieve-like covering and remained there until sufficient water drained away so that the sheet or web of paper could be transferred to a damp felt and pressed to compact the paper and remove some water. The paper was then dried before other wet processes, such as the application of an external sizing agent or coating, took place. Finishing procedures usually involved combinations of repeated pressings, or plating or calendaring to make the surface(s) smoother.”

  5. 5. See Dan Kushel, “Radiographic Methods in the Recording of Structure and Watermarks in Historic Paper,” in Fresh Woods and Pastures New: Seventeenth-Century Dutch Landscape Drawings from the Peck Collection (Chapel Hill, NC: Ackland Art Museum at the University of North Carolina, Chapel Hill, 1999), 117–23.

  6. 6. This watermark is discussed briefly in Andrew C. Weislogel and C. Richard Johnson, Jr., “Decision Trees and Fruitful Collaborations: The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell,” in Lines of Inquiry: Learning from Rembrandt’s Etchings, 42–43 (Ithaca, NY: Herbert F. Johnson Museum of Art, Cornell University, 2017), http://museum.cornell.edu/sites/default/files/DecisionTrees-WeislogelJohnson2017-LinesofInquiry.pdf.

  7. 7. Nancy Ash and Shelley Fletcher, Watermarks in Rembrandt’s Prints (Washington, DC: National Gallery of Art, 1998).

  8. 8. Erik Hinterding, Rembrandt as an Etcher: The Practice of Production and Distribution, 3 vols. (Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2006); Ash and Fletcher, Watermarks in Rembrandt’s Prints.

  9. 9.  C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, and Saira Haqqi, “Hunting for Paper Moldmates Among Rembrandt’s Prints,” IEEE Signal Processing Magazine 32 (July 2015): 28–37; Xuelei Xi, David Conathan, Amanda House, William A. Sethares, and C. Richard Johnson, Jr., “Automated Chain Line Marking and Pattern Matching in Radiographs of Rembrandt’s Prints,” paper presented at the 50th Asilomar Conference on Systems, Signals, and Computers, Pacific Grove, CA, November 2016, DOI: 10.1109/ACSSC.2016.7869010; and C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, Saira Haqqi, Reba Snyder, Erik Hinterding, Idelette van Leeuwen, Arie Wallert, Dionysia Christoforou, Jan van der Lubbe, Nadine M. Orenstein, Angela Campbell, and George Dietz, “Chain Line Pattern Matching in Rembrandt’s Prints,” in Rembrandt and His Circle: Insights and Discoveries, ed. Stephanie Dickey, 319–34 (Amsterdam: Amsterdam University Press, 2017).

  10. 10. The WIRE Project at Cornell is described in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” A brief overview of the WIRE Project appeared in Nancy Kenney, “Paper Trail: How Watermarks Illuminate Rembrandt’s Creative Process,” The Art Newspaper, February 26, 2019, https://www.theartnewspaper.com/news/paper-trail-how-watermarks-reveal-history

  11. 11. In order for some verbal questions to work, a specific orientation of the viewed image, such as head:top and nose:left, will be necessary. Both “yes” and “no” example images will adhere to that orientation. This will require proper image orientation, which is typically type-dependent, by the user at the start of a watermark identification session.

  12. 12. For readers who wish to go straight to the “answer,” feel free to jump ahead to the section entitled “Detailing Discoveries among Seven Prints in The Frick Collection.”

  13. 13. Appendix 3, “Chronological list of Rembrandt’s papers,” in Hinterding, Rembrandt as an Etcher, 2:333–416, describes the methodology of building this chronology. The utility of this chronology in analysis of Rembrandt’s prints is evident in the essays in volume 1 of Hinterding, Rembrandt as an Etcher. A more recent description of this methodology can be found in Erik Hinterding, Rembrandt Etchings from the Fritz Lugt Collection (Loughborough, UK: THOTH, 2008) 1:15–36.

  14. 14. The numbers of watermarks and watermark types and of the variants and subvariants of the Foolscap with Five-pointed Collar type are drawn from Hinterding, Rembrandt as an Etcher.

  15. 15. This and other abbreviations used in Hinterding, Rembrandt as an Etcher, are listed on page 15 of vol. 2.

  16. 16. The pioneering exposition on twinmarks and suggestions for their identification appears in Alan H. Stevenson, “Watermarks are Twin,” Studies in Bibliography 4 (1951–52): 57–91. This influential work was followed by Alan H. Stevenson, “Chain-Indentations in Paper as Evidence,” Studies in Bibliography 6 (1953–54): 181–95, which with it provides a groundbreaking description of the benefits of identifying manufactured patterns in handmade laid paper.

  17. 17. Volume 3 of Hinterding, Rembrandt as an Etcher, contains Hinterding’s catalogue of watermarks. The WIRE Project Team is currently adding the fragments classified by Hinterding to the decision-tree branches. The fragments in The Frick Collection’s Rembrandt prints identified in this article use the decision-tree branches for full watermarks, and associated decision tables, as a starting point for development of a specialized branch path.

  18. 18. A mismatch at this point may indicate a mistake in answering a feature assessment question, or it may indicate the discovery of a new watermark type, variant, subvariant, or twin not in Hinterding’s catalogue. While such discoveries should be rare, they are possible, since Hinterding’s catalogue is based on a limited number of collections (approximately three dozen). The WIRE Project has examined Rembrandt prints at a variety of collections at American university museums and has discovered new watermarks not catalogued by Hinterding, as described in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.”

  19. 19. While paper shrinkage post-production, due, for example, to treatments of the paper including wetting and drying, are widely recognized as causes for altering the dimensions of the paper, few published studies exist that examine the extent to which such size changes can make originally identical papers appear to no longer be from the same mold. One such study that suggests that these changes are likely to be more modest than might be expected appears in Johnson et al., “Chain Line Pattern Matching in Rembrandt’s Prints.”

  20. 20. Attempts to identify moldmates solely from matching chain line patterns emphasize the need for prints to contain, e.g., five or more chainspaces. With chainspaces typically approximately one inch, the need for five or more chainspaces is not satisfied for many of Rembrandt’s prints that are smaller than this once they have been trimmed over time to the edge of the platemark. Furthermore, radiographic images of prints are typically limited to those with watermarks. See C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, and Saira Haqqi, “Hunting for Paper Moldmates Among Rembrandt’s Prints,” IEEE Signal Processing Magazine 32 (July 2015), 28–37; and Xi et al., “Automated Chain Line Marking.”

  21. 21. These seven Rembrandt prints with watermarks or watermark fragments in The Frick Collection were not examined in the preparation of Hinterding’s catalogue.

  22. 22. Hinterding, Rembrandt as an Etcher, 2:116–17. The Bartsch number of Landscape with Three Trees is B212.

  23. 23. Hinterding, Rembrandt as an Etcher, 2:371–72.

  24. 24. Hinterding, Rembrandt as an Etcher, 2:134. The Bartsch number of Jan (Johannes) Lutma the Elder is B276.

  25. 25. Listed as batch 94 in Hinterding, Rembrandt as an Etcher, 2:407.

  26. 26. All state designations in this article were provided by Erik Hinterding. “There may not be a difference between [states 2 and 3]. It is just that the best impressions of state 2 are so dark that it is impossible to recognize the inscription.” Hinterding to the author, September 9, 2019.

  27. 27. Hinterding, Rembrandt as an Etcher, 1:146 notes as “particularly remarkable” that three watermarks appear in both states i and ii. One of them is Foolscap with Five-pointed Collar N.a, which is similarly remarkable for its appearance in all three states. This adds to the evidence “that Rembrandt must have had several small batches of paper on hand when he was printing the plates and used them in a single session.”

  28. 28. The redundancy of the latter three of the five questions about the bottom fragment was noted by Margaret N. Canfield (Cornell) during the creation of figure 29.

  29. 29. The Bartsch number of Landscape with Haybarn and Flock of Sheep is B224. Hinterding, Rembrandt as an Etcher, 2:303–24, lists eleven impressions of B224 with other watermarks, including four with an “Unidentified” Foolscap watermark. Hinterding notes that The Frick Collection’s print is state ii and “probably the first edition, with this watermark.” Hinterding to the author, February 2, 2019.

  30. 30. Hinterding, Rembrandt as an Etcher, 2:386–89.

  31. 31. Hinterding, Rembrandt as an Etcher, 2:162. The Bartsch number of The Goldweigher’s Field is B234.

  32. 32. Hinterding, Rembrandt as an Etcher, 2:162–64.

  33. 33. Hinterding, Rembrandt as an Etcher, 2:392–94.

  34. 34. Hinterding, Rembrandt as an Etcher, 2:92.

  35. 35. Hinterding, Rembrandt as an Etcher, 2:92.

  36. 36. Hinterding, Rembrandt as an Etcher, 2:154–58.

  37. 37. Subvariants B.f and D.b are missing from the decision tree being constructed because the only available image of B.f is obstructed and the only available image of D.b is a fragment.

  38. 38. Hinterding, Rembrandt as an Etcher, 2:154–58.

  39. 39. The features from Hinterding, Rembrandt as an Etcher, are not suited to distinguishing the variants needed for a decision tree when a feature shared by all subvariants of a particular variant is also not shared by all subvariants of another variant. Answering such a question would unintentionally split the subvariants of one variant, thereby not providing a decision tree that first divides all variants of a specific type into separate groups with each subvariant of a particular variant in a single group with its related subvariants. As an example, refer to the “platform” feature for variants D, G, and H of Arms of Amsterdam in Hinterding, Rembrandt as an Etcher, 3:29–41. Some subvariants of variant D have a platform and some do not, while all of the full-watermark images of variants G and H do have a platform beneath the lions. As another example, the presence of initials is not a feature suited to variant extraction of Strasbourg Lily, as variant A has subvariants with and without initials, as is apparent from Hinterding, Rembrandt as an Etcher, 3:414–16, where the subvariant images are labeled “Strasbourg lily-A-without-initials.”

  40. 40. Periodic reassessment by the WIRE Project team of the features used, the particular wording of the questions, and their accompanying sample images is ongoing during the decision-tree development phase.

  41. 41. In Hinterding, Rembrandt as an Etcher, the apostrophe is used to indicate a chain line down the center of a watermark or countermark.

  42. 42. Hinterding, Rembrandt as an Etcher, 2:155–56.

  43. 43. Hinterding, Rembrandt as an Etcher, 2:191–231.

  44. 44. Hinterding, Rembrandt as an Etcher, 2:211–15. The Bartsch number of Landscape with Three Gabled Cottages is B217.

  45. 45. Hinterding, Rembrandt as an Etcher, 2:398–402.

  46. 46. The Frick Collection online catalogue displays a date of 1650; see http://collections.frick. org/objects/641, accessed September 10, 2019. Links to the online catalogue pages of all of the seven prints studied here from The Frick Collection are provided in the appendix to this article, “Provenance, States, and Links to Online Catalogue Entries.”

  47. 47. Illustration and discussion of this example can be found in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” Discoveries of all three types (a)–(c) are listed in Appendix 2 of Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” A companion article is currently in preparation: Andrew C. Weislogel, C. Richard Johnson, Jr., et al., “A Collaborative Approach to Rembrandt Watermark Identification and Data Collection: Research Method, Tools, and Discoveries.”

  48. 48. With respect to the watermarks catalogued in Hinterding, Rembrandt as an Etcher, and incorporated in the decision tree from the WIRE Project.

  49. 49. For example, C. Richard Johnson, Jr., “WImBo Watermark Imaging Box Project: A Digital Art History Data Acquisition Tool,” paper from the 2018 Conference on Information Sciences and Systems, Princeton, NJ, March 2018; and John K. Delaney and Murray Loew, “Use of Infrared Hyperspectral Imaging (960–1680 nm) and Low Energy X-Radiography to Visualize Watermarks,” paper from the 2018 Conference on Information Sciences and Systems, Princeton, NJ, March 2018.

  50. 50. The initial manifestation of the WIRE decision tree was based on the set of single images of each distinctly classified watermark in volume 3 of Hinterding, Rembrandt as an Etcher. This does not assure flawless behavior of the WIRE decision tree in every occurrence of multiple impressions with different treatment and imaging histories. Among such a group, the confidence in answering some questions about distinguishing features may weaken enough that those questions should be replaced. Fortunately, the WIRE Project has access, from the archives of the Dutch University Institute for Art History in Florence, to the multisample dataset collected by Erik Hinterding in composing volume 2 of Rembrandt as an Etcher, in which multiple impressions having the same watermark are listed. As the completion of the first edition of the WIRE decision tree approaches, such multi-impression, same-watermark investigations of its behavior are beginning. Decision-tree refinements will be forthcoming.

  51. 51. Hinterding, Rembrandt as an Etcher.

  52. 52. An introduction to the utility of the classification of panel guild brands is provided in Jørgen Wadum, “The Antwerp Brand on Paintings on Panel,” in Looking Through Paintings: The Study of Painting Techniques and Materials in Support of Art Historical Research, ed. Erma Hermens, 179–198 (London: Archetype), 1998.

  53. 53. In February 2017, the author discovered—and subsequently vetted with Erik Hinterding—that the drawing The Steadfastness of Caius Fabricius Luscinus by Ferdinand Bol, which is cat. no. 41 of Thea Vignau-Wilberg, Rembrandt auf Papier: Werk und Wirkung (Munich: Hirmer, 2001), has the countermark IHS E.b in Hinterding, Rembrandt as an Etcher, 3:313, and the drawing Solomon Offering a Sacrifice of Peace Offerings by Ferdinand Bol, which is cat. no. 44 in Vignau-Wilberg, Rembrandt auf Papier, has the watermark Strasbourg Bend D’.a.a in Hinterding, Rembrandt as an Etcher, 3:373. Images of the watermarks of both of these Bol drawings appear in Vignau-Wilberg, Rembrandt auf Papier, 343. This cross-use discovery was highlighted in the display of the WIRE Project’s interactive decision tree in the exhibition “Rembrandt’s Laboratory: Rembrandt’s Technique Unravelled” at the Rembrandthuis, Amsterdam, September 21, 2019–February 16, 2020.

  54. 54. Provided by Louisa Wood Ruby.

  55. 55. State designations provided in Erik Hinterding and Jaco Rutger, Rembrandt: The New Hollstein Dutch and Flemish Etchings, Engravings, and Woodcuts 1450–1700 (Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2013).

  56. 56. Johnson et al., “Chain Line Pattern Matching.”

  57. 57. Hinterding, Rembrandt as an Etcher.

  58. 58. C. Richard Johnson, Jr., “The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell,” lecture presented at The Frick Collection, New York, December 6, 2018, https://www.frick.org/interact/wire_project_cornell.

Ash, Nancy, and Shelley Fletcher. Watermarks in Rembrandt’s Prints. Washington, DC: National Gallery of Art, 1998.

Baker, Cathleen Ann. From the Hand to the Machine: Nineteenth-century American Paper and Mediums: Technologies, Materials, and Conservation. Ann Arbor, MI: Legacy Press, 2010. https://doi.org/10.5860/rbm.13.1.370

Delaney, John K., and Murray Loew. “Use of Infrared Hyperspectral Imaging (960–1680 nm) and Low Energy X-Radiography to Visualize Watermarks.” Proceedings of the Conference on Information Sciences and Systems, Princeton, NJ, March 2018. https://doi.org/10.1109/ciss.2018.8362321

Harris, Neil. Paper and Watermarks as Bibliographical Evidence. 2nd ed. Lyon: Institut d’histoire du livre, 2017. http://ihl.enssib.fr/en/paper-and-watermarks-as-bibliographical-evidence.

Hinterding, Erik. Rembrandt as an Etcher: The Practice of Production and Distribution. 3 vols. Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2006.

—. Rembrandt Etchings from the Fritz Lugt Collection. 2 vols. Loughborough, UK: THOTH, 2008.

Hinterding, Erik, and Jaco Rutger. Rembrandt: The New Hollstein Dutch and Flemish Etchings, Engravings, and Woodcuts 1450–1700. Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2013.

Johnson, C. Richard, Jr., William A. Sethares, Margaret Holben Ellis, and Saira Haqqi. “Hunting for Paper Moldmates Among Rembrandt’s Prints.” IEEE Signal Processing Magazine 32 (July 2015): 28–37. https://doi.org/10.1109/msp.2015.2404931

Johnson, C. Richard, Jr., William A. Sethares, Margaret Hoben Ellis, Saira Haqqi, Reba Snyder, Erik Hinterding, Idelette van Leeuwen, Arie Wallert, Dionysia Christoforou, Jan van der Lubbe, Nadine M. Orenstein, Angela Campbell, and George Dietz. “Chain Line Pattern Matching in Rembrandt’s Prints.” In Rembrandt and His Circle: Insights and Discoveries, edited by Stephanie Dickey, 319–34. Amsterdam: Amsterdam University Press, 2017.

Johnson, C. Richard, Jr. “WImBo Watermark Imaging Box Project: A Digital Art History Data Acquisition Tool.” Proceedings of the Conference on Information Sciences and Systems, Princeton, NJ, March 2018. https://doi.org/10.1109/ciss.2018.8362290

Kenney, Nancy. “Paper Trail: How Watermarks Illuminate Rembrandt’s Creative Process.” The Art Newspaper 309 (February 2019). https://www.theartnewspaper.com/news/paper-trail-how-watermarks-reveal-history.

Kushel, Dan. “Radiographic Methods in the Recording of Structure and Watermarks in Historic Paper.” In Fresh Woods and Pastures New: Seventeenth-Century Dutch Landscape Drawings from the Peck Collection, 117–23. Chapel Hill, NC: Ackland Art Museum at the University of North Carolina, Chapel Hill, 1999.

Stevenson, Alan H. “Watermarks are Twins.” Studies in Bibliography 4 (1951–52): 57–91.

—. “Chain-Indentations in Paper as Evidence.” Studies in Bibliography 6 (1953–54):  181–195.

Vignau-Wilberg, Thea. Rembrandt auf Papier: Werk und Wirkung. Munich: Hirmer, 2001.

Wadum, Jørgen. “The Antwerp Brand on Paintings on Panel.” In Looking Through Paintings: The Study of Painting Techniques and Materials in Support of Art Historical Research, edited by Erma Hermens, 179–198. London: Archetype Publications, 1998.

Weislogel, Andrew C., and C. Richard Johnson, Jr. “Decision Trees and Fruitful Collaborations: The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell.” In Lines of Inquiry: Learning from Rembrandt’s Etchings, 32–5. Ithaca, NY: Herbert F. Johnson Museum of Art, Cornell University, 2017. http://museum.cornell.edu/sites/default/files/DecisionTrees-WeislogelJohnson2017-LinesofInquiry.pdf

Xi, Xuelei, David Conathan, Amanda House, William A. Sethares, and C. Richard Johnson, Jr. “Automated Chain Line Marking and Pattern Matching in Radiographs of Rembrandt’s Prints.” Paper presented at the 50th Asilomar Conference on Systems, Signals, and Computers, Pacific Grove, CA, November 2016. https://doi.org/10.1109/acssc.2016.7869010

List of Illustrations

Twin molds with dashed lines indicating the cutting up of full broadsheets into separate sheets of paper, all of which are batchmates
Fig. 1 Twin molds with dashed lines indicating the cutting up of full broadsheets into separate sheets of paper, all of which are batchmates (illustration by author) [side-by-side viewer]
Transmitted-light photograph of a blank sheet of laid paper from a 1536 French book. Left: Right half of a full-mold sheet containing Foolscap with Five-pointed Collar watermark. Right: Enlargement of region around watermark with chain lines and laid lines
Fig. 2 Transmitted-light photograph of a blank sheet of laid paper from a 1536 French book. Left: Right half of a full-mold sheet containing Foolscap with Five-pointed Collar watermark. Right: Enlargement of region around watermark with chain lines and laid lines (sheet from study collection of Conservation Center in the Institute of Fine Arts provided by Margaret Hoben Ellis (New York University); photographs provided by S. Schlick and P. Messier (Yale)) [side-by-side viewer]
Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, c. 1652
Fig. 3 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Herbert F. Johnson Museum, Cornell University, Ithaca, NY, etching, drypoint and engraving on paper, 21.6 x 28.6 cm (sheet), B56vi(?), 84.010, c. 1652 [side-by-side viewer]
Transmitted-light image showing watermarkRembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers,
Fig. 3-1 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Transmitted-light image showing watermark [side-by-side viewer]
Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Excerpt from beta-radiograph
Fig. 3-2 Rembrandt Harmenszoon van Rijn, The Flight into Egypt, after Hercules Seghers, Excerpt from beta-radiograph [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Trees, The Frick Collection, 1915.3.28. Right: Beta-radiograph of portion around watermark
Fig. 4 Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Trees, The Frick Collection, 1915.3.28. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis {NYU} and Reba Fishman Snyder {Morgan}) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Jan (Johannes) Lutma the Elder, The Frick Collection, 1916.3.37; Right: Beta-radiograph of portion around watermark
Fig. 5 Left: Rembrandt Harmenszoon van Rijn, Jan (Johannes) Lutma the Elder, The Frick Collection, 1916.3.37; Right: Beta-radiograph of portion around watermark (Provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Haybarn and Flock of Sheep, The Frick Collection, 1916.3.29. Right: Beta-radiograph of portion around watermark
Fig. 6 Left: Rembrandt Harmenszoon van Rijn, Landscape with Haybarn and Flock of Sheep, The Frick Collection, 1916.3.29. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, The Goldweigher’s Field, The Frick Collection, 1915.3.31. Right: Beta-radiograph of portion around watermark
Fig. 7 Left: Rembrandt Harmenszoon van Rijn, The Goldweigher’s Field, The Frick Collection, 1915.3.31. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Clement de Jonghe, The Frick Collection, 1916.3.36. Right: Beta-radiograph of portion around watermark
Fig. 8 Left: Rembrandt Harmenszoon van Rijn, Clement de Jonghe, The Frick Collection, 1916.3.36. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, St. Francis Praying Beneath a Tree, The Frick Collection, 1916.3.35. Right: Beta-radiograph of portion around watermark
Fig. 9 Left: Rembrandt Harmenszoon van Rijn, St. Francis Praying Beneath a Tree, The Frick Collection, 1916.3.35. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Gabled Cottages, The Frick Collection, 1916.3.30. Right: Beta-radiograph of portion around watermark
Fig. 10 Left: Rembrandt Harmenszoon van Rijn, Landscape with Three Gabled Cottages, The Frick Collection, 1916.3.30. Right: Beta-radiograph of portion around watermark (provided by Margaret Hoben Ellis (NYU) and Reba Fishman Snyder (Morgan)) [side-by-side viewer]
Four variants of Foolscap with Five-pointed Collar. Left to right: A.a.a., H.b.a., K.a.a., N.a.b
Fig. 11 Four variants of Foolscap with Five-pointed Collar. Left to right: A.a.a., H.b.a., K.a.a., N.a.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Four subvariants of Foolscap with Five-pointed Collar, variant K. Left to right: K.a.a., K.c.a., K.e.a., K.f.a
Fig. 12 Four subvariants of Foolscap with Five-pointed Collar, variant K. Left to right: K.a.a., K.c.a., K.e.a., K.f.a (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two twins of Foolscap with Five-pointed Collar, subvariant A.a. Left to right: A.a.a., A.a.b
Fig. 13 Two twins of Foolscap with Five-pointed Collar, subvariant A.a. Left to right: A.a.a., A.a.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Feature labels for Foolscap with Five-pointed Collar
Fig. 14 Feature labels for Foolscap with Five-pointed Collar (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree for four variants of Foolscap with Five-pointed Collar in figure 11
Fig. 15 Decision tree for four variants of Foolscap with Five-pointed Collar in figure 11 (illustration by author) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Foolscap with Five-pointed Collar in Rembrandt’s etchings
Fig. 16 Decision tree by the WIRE Project from type to subvariant for Foolscap with Five-pointed Collar in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Flowchart for watermark identification
Fig. 17 Flowchart for watermark identification (illustration by author) [side-by-side viewer]
Left to right: Watermark under investigation, A.a.a.,  A.a.b
Fig. 18 Left to right: Watermark under investigation, A.a.a.,  A.a.b (middle and right images courtesy of Sound and Vision Publishers BV; red markings by author) [side-by-side viewer]
Type to subvariant path through decision tree to Foolscap with Five-pointed Collar A.a in Rembrandt’s etchings
Fig. 19 Type to subvariant path through decision tree to Foolscap with Five-pointed Collar A.a in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the collar have 4 points? Left: YES. Right: NO
Fig. 20 Does the collar have 4 points? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Are the roundels in a pyramid configuration (one on top and two on bottom)? Left: YES. Right: NO
Fig. 21 Are the roundels in a pyramid configuration (one on top and two on bottom)? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Are the peaks of the caps divided by stripes? Left: YES. Right: NO
Fig. 22 Are the peaks of the caps divided by stripes? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the collar point below the face bend underneath the chin? Left: YES. Right: NO
Fig. 23 Does the collar point below the face bend underneath the chin? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Does the “4” point in the same direction as the face? Left: YES. Right: NO
Fig. 24 Does the “4” point in the same direction as the face? Left: YES. Right: NO (images courtesy of Sound and Vision Publishers BV; red markings by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment test (watermark of interest over Foolscap with Five-pointed Collar A.a.a) and match
Fig. 25 Chainspace alignment test (watermark of interest over Foolscap with Five-pointed Collar A.a.a) and match (images courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Foolscap with Five-pointed Collar N.a.b
Fig. 26 Left: Watermark under investigation. Right: Foolscap with Five-pointed Collar N.a.b (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Path from type to subvariant of Foolscap with Five-pointed Collar N.a
Fig. 27 Path from type to subvariant of Foolscap with Five-pointed Collar N.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment test (Foolscap with Five-pointed Collar N.a.b over watermark of interest) and match
Fig. 28 Chainspace alignment test (Foolscap with Five-pointed Collar N.a.b over watermark of interest) and match (images courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Path from type to subvariant of fragment of Foolscap with Five-pointed Collar K.a
Fig. 29 Path from type to subvariant of fragment of Foolscap with Five-pointed Collar K.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark fragment under investigation; Right: Foolscap with Five-pointed Collar K.a.a
Fig. 30 Left: Watermark fragment under investigation; Right: Foolscap with Five-pointed Collar K.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of Foolscap with Five-pointed collar K.a.a (top) and watermark of interest and match
Fig. 31 Chainspace alignment of Foolscap with Five-pointed collar K.a.a (top) and watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; composition by author) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Paschal Lamb in Rembrandt’s etchings
Fig. 32 Decision tree by the WIRE Project from type to subvariant for Paschal Lamb in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Feature labels for Paschal Lamb
Fig. 33 Feature labels for Paschal Lamb (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Type to subvariant path for Paschal Lamb fragment subvariant A.b
Fig. 34 Type to subvariant path for Paschal Lamb fragment subvariant A.b (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Paschal Lamb A.b
Fig. 35 Left: Watermark under investigation. Right: Paschal Lamb A.b (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of Paschal Lamb A.b over watermark of interest and match
Fig. 36 Chainspace alignment of Paschal Lamb A.b over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Top: Watermark under investigation Bottom: RC
Fig. 37 Top: Watermark under investigation Bottom: RC (Countermark) RC.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of RC.a.a excerpt over watermark of interest and match
Fig. 38 Chainspace alignment of RC.a.a excerpt over watermark of interest and match (image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
One sample from each IHS variant. Top row (left to right): A.a, B.a.a. Bottom row (left to right): C.a.a, D.a.a., E.b
Fig. 39 One sample from each IHS variant. Top row (left to right): A.a, B.a.a. Bottom row (left to right): C.a.a, D.a.a., E.b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two possible decision-tree realizations for classifying variants of IHS
Fig. 40 Two possible decision-tree realizations for classifying variants of IHS (illustrations by author) [side-by-side viewer]
Subvariants of IHS variant E: Left: E-a. Right: E-b
Fig. 41 Subvariants of IHS variant E: Left: E-a. Right: E-b (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Samples of each subvariant of IHS variant B: Top row (left to right): B’.a, B.a.a, B.b.a. Bottom row (left to right): B.c, B.d, B.e.a
Fig. 42 Samples of each subvariant of IHS variant B: Top row (left to right): B’.a, B.a.a, B.b.a. Bottom row (left to right): B.c, B.d, B.e.a (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Two possible decision-tree branches for identifying subvariants of HIS variant B
Fig. 43 Two possible decision-tree branches for identifying subvariants of HIS variant B (illustrations by author) [side-by-side viewer]
Some subvariants of IHS variant A: Top row (left to right): A.a, A.b, A.c. Bottom row (left to right): A.d.a, A.e
Fig. 44 Some subvariants of IHS variant A: Top row (left to right): A.a, A.b, A.c. Bottom row (left to right): A.d.a, A.e (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Some more subvariants of IHS variant A: Top row (left to right): A.f.a, A.g, A.h. Bottom row (left to right): A.i, A.
Fig. 45 Some more subvariants of IHS variant A: Top row (left to right): A.f.a, A.g, A.h. Bottom row (left to right): A.i, A.j (images courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Feature labels for IHS
Fig. 46 Feature labels for IHS (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for IHS in Rembrandt’s etchings
Fig. 47 Decision tree by the WIRE Project from type to subvariant for IHS in Rembrandt’s etchings (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Path from type to subvariant B.a of IHS Countermark
Fig. 48 Path from type to subvariant B.a of IHS Countermark (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Left: Countermark under investigation Right: IHS B.a.a
Fig. 49 Left: Countermark under investigation Right: IHS B.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Chainspace alignment of IHS B.a over watermark of interest and match
Fig. 50 Chainspace alignment of IHS B.a over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Feature labels for Strasbourg Lily
Fig. 51 Feature labels for Strasbourg Lily (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Decision tree by the WIRE Project from type to subvariant for Strasbourg Lily in Rembrandt’s etchings
Fig. 52 Decision tree by the WIRE Project from type to subvariant for Strasbourg Lily in Rembrandt’s etchings (illustration by Isabella M. Dobson (Cornell)) [side-by-side viewer]
Left: Watermark under investigation. Right: Strasbourg Lily E’.a.a
Fig. 53 Left: Watermark under investigation. Right: Strasbourg Lily E’.a.a (image courtesy of Sound and Vision Publishers BV) [side-by-side viewer]
Strasbourg Lily decision-tree branch to variant E’.a
Fig. 54 Strasbourg Lily decision-tree branch to variant E’.a (illustration by Margaret N. Canfield (Cornell)) [side-by-side viewer]
Chainspace alignment of Strasbourg Lily E’.a.a over watermark of interest and match
Fig. 55 Chainspace alignment of Strasbourg Lily E’.a.a over watermark of interest and match (top image courtesy of Sound and Vision Publishers BV; overlay composed by Margaret N. Canfield (Cornell)) [side-by-side viewer]

Footnotes

  1. 1. From chapter 5 of Neil Harris, Paper and Watermarks as Bibliographical Evidence, 2nd ed. (Lyon: Institut d’histoire du livre, 2017), http://ihl.enssib.fr/en/paper-and-watermarks-as-bibliographical-evidence

  2. 2. Louisa Wood Ruby, Director of Research, Frick Art Reference Library, to the author, July 24, 2019.

  3. 3. Plus a deckel, which is a removable wooden rim used by both molds in the twin pair that fits on top of each mold, contains the pulp, and sets the size of the sheet.

  4. 4. A video documenting an attempt to recreate the pre-1800 papermaking process is a solid introduction: Tim Barrett, “Chancery Papermaking at the University of Iowa’s Center for the Book,” filmed 2013 at the University of Iowa, video, 11:24, http://now.uiowa.edu/2013/05/center-book-releases-short-film-about-papermaking. A compact verbal description of the laid papermaking process appears in Cathleen Ann Baker, From the Hand to the Machine: Nineteenth-Century American Paper and Mediums: Technologies, Materials, and Conservation (Ann Arbor, MI: Legacy Press, 2010), 34: “Once the fibers were beaten to a particular degree and any non-fibrous additives, such as an internal sizing agent, color, and/or filler, were added, the formation of the paper remained the same: pulp was deposited onto the surface of a sieve-like covering and remained there until sufficient water drained away so that the sheet or web of paper could be transferred to a damp felt and pressed to compact the paper and remove some water. The paper was then dried before other wet processes, such as the application of an external sizing agent or coating, took place. Finishing procedures usually involved combinations of repeated pressings, or plating or calendaring to make the surface(s) smoother.”

  5. 5. See Dan Kushel, “Radiographic Methods in the Recording of Structure and Watermarks in Historic Paper,” in Fresh Woods and Pastures New: Seventeenth-Century Dutch Landscape Drawings from the Peck Collection (Chapel Hill, NC: Ackland Art Museum at the University of North Carolina, Chapel Hill, 1999), 117–23.

  6. 6. This watermark is discussed briefly in Andrew C. Weislogel and C. Richard Johnson, Jr., “Decision Trees and Fruitful Collaborations: The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell,” in Lines of Inquiry: Learning from Rembrandt’s Etchings, 42–43 (Ithaca, NY: Herbert F. Johnson Museum of Art, Cornell University, 2017), http://museum.cornell.edu/sites/default/files/DecisionTrees-WeislogelJohnson2017-LinesofInquiry.pdf.

  7. 7. Nancy Ash and Shelley Fletcher, Watermarks in Rembrandt’s Prints (Washington, DC: National Gallery of Art, 1998).

  8. 8. Erik Hinterding, Rembrandt as an Etcher: The Practice of Production and Distribution, 3 vols. (Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2006); Ash and Fletcher, Watermarks in Rembrandt’s Prints.

  9. 9.  C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, and Saira Haqqi, “Hunting for Paper Moldmates Among Rembrandt’s Prints,” IEEE Signal Processing Magazine 32 (July 2015): 28–37; Xuelei Xi, David Conathan, Amanda House, William A. Sethares, and C. Richard Johnson, Jr., “Automated Chain Line Marking and Pattern Matching in Radiographs of Rembrandt’s Prints,” paper presented at the 50th Asilomar Conference on Systems, Signals, and Computers, Pacific Grove, CA, November 2016, DOI: 10.1109/ACSSC.2016.7869010; and C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, Saira Haqqi, Reba Snyder, Erik Hinterding, Idelette van Leeuwen, Arie Wallert, Dionysia Christoforou, Jan van der Lubbe, Nadine M. Orenstein, Angela Campbell, and George Dietz, “Chain Line Pattern Matching in Rembrandt’s Prints,” in Rembrandt and His Circle: Insights and Discoveries, ed. Stephanie Dickey, 319–34 (Amsterdam: Amsterdam University Press, 2017).

  10. 10. The WIRE Project at Cornell is described in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” A brief overview of the WIRE Project appeared in Nancy Kenney, “Paper Trail: How Watermarks Illuminate Rembrandt’s Creative Process,” The Art Newspaper, February 26, 2019, https://www.theartnewspaper.com/news/paper-trail-how-watermarks-reveal-history

  11. 11. In order for some verbal questions to work, a specific orientation of the viewed image, such as head:top and nose:left, will be necessary. Both “yes” and “no” example images will adhere to that orientation. This will require proper image orientation, which is typically type-dependent, by the user at the start of a watermark identification session.

  12. 12. For readers who wish to go straight to the “answer,” feel free to jump ahead to the section entitled “Detailing Discoveries among Seven Prints in The Frick Collection.”

  13. 13. Appendix 3, “Chronological list of Rembrandt’s papers,” in Hinterding, Rembrandt as an Etcher, 2:333–416, describes the methodology of building this chronology. The utility of this chronology in analysis of Rembrandt’s prints is evident in the essays in volume 1 of Hinterding, Rembrandt as an Etcher. A more recent description of this methodology can be found in Erik Hinterding, Rembrandt Etchings from the Fritz Lugt Collection (Loughborough, UK: THOTH, 2008) 1:15–36.

  14. 14. The numbers of watermarks and watermark types and of the variants and subvariants of the Foolscap with Five-pointed Collar type are drawn from Hinterding, Rembrandt as an Etcher.

  15. 15. This and other abbreviations used in Hinterding, Rembrandt as an Etcher, are listed on page 15 of vol. 2.

  16. 16. The pioneering exposition on twinmarks and suggestions for their identification appears in Alan H. Stevenson, “Watermarks are Twin,” Studies in Bibliography 4 (1951–52): 57–91. This influential work was followed by Alan H. Stevenson, “Chain-Indentations in Paper as Evidence,” Studies in Bibliography 6 (1953–54): 181–95, which with it provides a groundbreaking description of the benefits of identifying manufactured patterns in handmade laid paper.

  17. 17. Volume 3 of Hinterding, Rembrandt as an Etcher, contains Hinterding’s catalogue of watermarks. The WIRE Project Team is currently adding the fragments classified by Hinterding to the decision-tree branches. The fragments in The Frick Collection’s Rembrandt prints identified in this article use the decision-tree branches for full watermarks, and associated decision tables, as a starting point for development of a specialized branch path.

  18. 18. A mismatch at this point may indicate a mistake in answering a feature assessment question, or it may indicate the discovery of a new watermark type, variant, subvariant, or twin not in Hinterding’s catalogue. While such discoveries should be rare, they are possible, since Hinterding’s catalogue is based on a limited number of collections (approximately three dozen). The WIRE Project has examined Rembrandt prints at a variety of collections at American university museums and has discovered new watermarks not catalogued by Hinterding, as described in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.”

  19. 19. While paper shrinkage post-production, due, for example, to treatments of the paper including wetting and drying, are widely recognized as causes for altering the dimensions of the paper, few published studies exist that examine the extent to which such size changes can make originally identical papers appear to no longer be from the same mold. One such study that suggests that these changes are likely to be more modest than might be expected appears in Johnson et al., “Chain Line Pattern Matching in Rembrandt’s Prints.”

  20. 20. Attempts to identify moldmates solely from matching chain line patterns emphasize the need for prints to contain, e.g., five or more chainspaces. With chainspaces typically approximately one inch, the need for five or more chainspaces is not satisfied for many of Rembrandt’s prints that are smaller than this once they have been trimmed over time to the edge of the platemark. Furthermore, radiographic images of prints are typically limited to those with watermarks. See C. Richard Johnson, Jr., William A. Sethares, Margaret Hoben Ellis, and Saira Haqqi, “Hunting for Paper Moldmates Among Rembrandt’s Prints,” IEEE Signal Processing Magazine 32 (July 2015), 28–37; and Xi et al., “Automated Chain Line Marking.”

  21. 21. These seven Rembrandt prints with watermarks or watermark fragments in The Frick Collection were not examined in the preparation of Hinterding’s catalogue.

  22. 22. Hinterding, Rembrandt as an Etcher, 2:116–17. The Bartsch number of Landscape with Three Trees is B212.

  23. 23. Hinterding, Rembrandt as an Etcher, 2:371–72.

  24. 24. Hinterding, Rembrandt as an Etcher, 2:134. The Bartsch number of Jan (Johannes) Lutma the Elder is B276.

  25. 25. Listed as batch 94 in Hinterding, Rembrandt as an Etcher, 2:407.

  26. 26. All state designations in this article were provided by Erik Hinterding. “There may not be a difference between [states 2 and 3]. It is just that the best impressions of state 2 are so dark that it is impossible to recognize the inscription.” Hinterding to the author, September 9, 2019.

  27. 27. Hinterding, Rembrandt as an Etcher, 1:146 notes as “particularly remarkable” that three watermarks appear in both states i and ii. One of them is Foolscap with Five-pointed Collar N.a, which is similarly remarkable for its appearance in all three states. This adds to the evidence “that Rembrandt must have had several small batches of paper on hand when he was printing the plates and used them in a single session.”

  28. 28. The redundancy of the latter three of the five questions about the bottom fragment was noted by Margaret N. Canfield (Cornell) during the creation of figure 29.

  29. 29. The Bartsch number of Landscape with Haybarn and Flock of Sheep is B224. Hinterding, Rembrandt as an Etcher, 2:303–24, lists eleven impressions of B224 with other watermarks, including four with an “Unidentified” Foolscap watermark. Hinterding notes that The Frick Collection’s print is state ii and “probably the first edition, with this watermark.” Hinterding to the author, February 2, 2019.

  30. 30. Hinterding, Rembrandt as an Etcher, 2:386–89.

  31. 31. Hinterding, Rembrandt as an Etcher, 2:162. The Bartsch number of The Goldweigher’s Field is B234.

  32. 32. Hinterding, Rembrandt as an Etcher, 2:162–64.

  33. 33. Hinterding, Rembrandt as an Etcher, 2:392–94.

  34. 34. Hinterding, Rembrandt as an Etcher, 2:92.

  35. 35. Hinterding, Rembrandt as an Etcher, 2:92.

  36. 36. Hinterding, Rembrandt as an Etcher, 2:154–58.

  37. 37. Subvariants B.f and D.b are missing from the decision tree being constructed because the only available image of B.f is obstructed and the only available image of D.b is a fragment.

  38. 38. Hinterding, Rembrandt as an Etcher, 2:154–58.

  39. 39. The features from Hinterding, Rembrandt as an Etcher, are not suited to distinguishing the variants needed for a decision tree when a feature shared by all subvariants of a particular variant is also not shared by all subvariants of another variant. Answering such a question would unintentionally split the subvariants of one variant, thereby not providing a decision tree that first divides all variants of a specific type into separate groups with each subvariant of a particular variant in a single group with its related subvariants. As an example, refer to the “platform” feature for variants D, G, and H of Arms of Amsterdam in Hinterding, Rembrandt as an Etcher, 3:29–41. Some subvariants of variant D have a platform and some do not, while all of the full-watermark images of variants G and H do have a platform beneath the lions. As another example, the presence of initials is not a feature suited to variant extraction of Strasbourg Lily, as variant A has subvariants with and without initials, as is apparent from Hinterding, Rembrandt as an Etcher, 3:414–16, where the subvariant images are labeled “Strasbourg lily-A-without-initials.”

  40. 40. Periodic reassessment by the WIRE Project team of the features used, the particular wording of the questions, and their accompanying sample images is ongoing during the decision-tree development phase.

  41. 41. In Hinterding, Rembrandt as an Etcher, the apostrophe is used to indicate a chain line down the center of a watermark or countermark.

  42. 42. Hinterding, Rembrandt as an Etcher, 2:155–56.

  43. 43. Hinterding, Rembrandt as an Etcher, 2:191–231.

  44. 44. Hinterding, Rembrandt as an Etcher, 2:211–15. The Bartsch number of Landscape with Three Gabled Cottages is B217.

  45. 45. Hinterding, Rembrandt as an Etcher, 2:398–402.

  46. 46. The Frick Collection online catalogue displays a date of 1650; see http://collections.frick. org/objects/641, accessed September 10, 2019. Links to the online catalogue pages of all of the seven prints studied here from The Frick Collection are provided in the appendix to this article, “Provenance, States, and Links to Online Catalogue Entries.”

  47. 47. Illustration and discussion of this example can be found in Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” Discoveries of all three types (a)–(c) are listed in Appendix 2 of Weislogel and Johnson, “Decision Trees and Fruitful Collaborations.” A companion article is currently in preparation: Andrew C. Weislogel, C. Richard Johnson, Jr., et al., “A Collaborative Approach to Rembrandt Watermark Identification and Data Collection: Research Method, Tools, and Discoveries.”

  48. 48. With respect to the watermarks catalogued in Hinterding, Rembrandt as an Etcher, and incorporated in the decision tree from the WIRE Project.

  49. 49. For example, C. Richard Johnson, Jr., “WImBo Watermark Imaging Box Project: A Digital Art History Data Acquisition Tool,” paper from the 2018 Conference on Information Sciences and Systems, Princeton, NJ, March 2018; and John K. Delaney and Murray Loew, “Use of Infrared Hyperspectral Imaging (960–1680 nm) and Low Energy X-Radiography to Visualize Watermarks,” paper from the 2018 Conference on Information Sciences and Systems, Princeton, NJ, March 2018.

  50. 50. The initial manifestation of the WIRE decision tree was based on the set of single images of each distinctly classified watermark in volume 3 of Hinterding, Rembrandt as an Etcher. This does not assure flawless behavior of the WIRE decision tree in every occurrence of multiple impressions with different treatment and imaging histories. Among such a group, the confidence in answering some questions about distinguishing features may weaken enough that those questions should be replaced. Fortunately, the WIRE Project has access, from the archives of the Dutch University Institute for Art History in Florence, to the multisample dataset collected by Erik Hinterding in composing volume 2 of Rembrandt as an Etcher, in which multiple impressions having the same watermark are listed. As the completion of the first edition of the WIRE decision tree approaches, such multi-impression, same-watermark investigations of its behavior are beginning. Decision-tree refinements will be forthcoming.

  51. 51. Hinterding, Rembrandt as an Etcher.

  52. 52. An introduction to the utility of the classification of panel guild brands is provided in Jørgen Wadum, “The Antwerp Brand on Paintings on Panel,” in Looking Through Paintings: The Study of Painting Techniques and Materials in Support of Art Historical Research, ed. Erma Hermens, 179–198 (London: Archetype), 1998.

  53. 53. In February 2017, the author discovered—and subsequently vetted with Erik Hinterding—that the drawing The Steadfastness of Caius Fabricius Luscinus by Ferdinand Bol, which is cat. no. 41 of Thea Vignau-Wilberg, Rembrandt auf Papier: Werk und Wirkung (Munich: Hirmer, 2001), has the countermark IHS E.b in Hinterding, Rembrandt as an Etcher, 3:313, and the drawing Solomon Offering a Sacrifice of Peace Offerings by Ferdinand Bol, which is cat. no. 44 in Vignau-Wilberg, Rembrandt auf Papier, has the watermark Strasbourg Bend D’.a.a in Hinterding, Rembrandt as an Etcher, 3:373. Images of the watermarks of both of these Bol drawings appear in Vignau-Wilberg, Rembrandt auf Papier, 343. This cross-use discovery was highlighted in the display of the WIRE Project’s interactive decision tree in the exhibition “Rembrandt’s Laboratory: Rembrandt’s Technique Unravelled” at the Rembrandthuis, Amsterdam, September 21, 2019–February 16, 2020.

  54. 54. Provided by Louisa Wood Ruby.

  55. 55. State designations provided in Erik Hinterding and Jaco Rutger, Rembrandt: The New Hollstein Dutch and Flemish Etchings, Engravings, and Woodcuts 1450–1700 (Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2013).

  56. 56. Johnson et al., “Chain Line Pattern Matching.”

  57. 57. Hinterding, Rembrandt as an Etcher.

  58. 58. C. Richard Johnson, Jr., “The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell,” lecture presented at The Frick Collection, New York, December 6, 2018, https://www.frick.org/interact/wire_project_cornell.

Bibliography

Ash, Nancy, and Shelley Fletcher. Watermarks in Rembrandt’s Prints. Washington, DC: National Gallery of Art, 1998.

Baker, Cathleen Ann. From the Hand to the Machine: Nineteenth-century American Paper and Mediums: Technologies, Materials, and Conservation. Ann Arbor, MI: Legacy Press, 2010. https://doi.org/10.5860/rbm.13.1.370

Delaney, John K., and Murray Loew. “Use of Infrared Hyperspectral Imaging (960–1680 nm) and Low Energy X-Radiography to Visualize Watermarks.” Proceedings of the Conference on Information Sciences and Systems, Princeton, NJ, March 2018. https://doi.org/10.1109/ciss.2018.8362321

Harris, Neil. Paper and Watermarks as Bibliographical Evidence. 2nd ed. Lyon: Institut d’histoire du livre, 2017. http://ihl.enssib.fr/en/paper-and-watermarks-as-bibliographical-evidence.

Hinterding, Erik. Rembrandt as an Etcher: The Practice of Production and Distribution. 3 vols. Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2006.

—. Rembrandt Etchings from the Fritz Lugt Collection. 2 vols. Loughborough, UK: THOTH, 2008.

Hinterding, Erik, and Jaco Rutger. Rembrandt: The New Hollstein Dutch and Flemish Etchings, Engravings, and Woodcuts 1450–1700. Ouderkerk aan den Ijssel, Netherlands: Sound & Vision, 2013.

Johnson, C. Richard, Jr., William A. Sethares, Margaret Holben Ellis, and Saira Haqqi. “Hunting for Paper Moldmates Among Rembrandt’s Prints.” IEEE Signal Processing Magazine 32 (July 2015): 28–37. https://doi.org/10.1109/msp.2015.2404931

Johnson, C. Richard, Jr., William A. Sethares, Margaret Hoben Ellis, Saira Haqqi, Reba Snyder, Erik Hinterding, Idelette van Leeuwen, Arie Wallert, Dionysia Christoforou, Jan van der Lubbe, Nadine M. Orenstein, Angela Campbell, and George Dietz. “Chain Line Pattern Matching in Rembrandt’s Prints.” In Rembrandt and His Circle: Insights and Discoveries, edited by Stephanie Dickey, 319–34. Amsterdam: Amsterdam University Press, 2017.

Johnson, C. Richard, Jr. “WImBo Watermark Imaging Box Project: A Digital Art History Data Acquisition Tool.” Proceedings of the Conference on Information Sciences and Systems, Princeton, NJ, March 2018. https://doi.org/10.1109/ciss.2018.8362290

Kenney, Nancy. “Paper Trail: How Watermarks Illuminate Rembrandt’s Creative Process.” The Art Newspaper 309 (February 2019). https://www.theartnewspaper.com/news/paper-trail-how-watermarks-reveal-history.

Kushel, Dan. “Radiographic Methods in the Recording of Structure and Watermarks in Historic Paper.” In Fresh Woods and Pastures New: Seventeenth-Century Dutch Landscape Drawings from the Peck Collection, 117–23. Chapel Hill, NC: Ackland Art Museum at the University of North Carolina, Chapel Hill, 1999.

Stevenson, Alan H. “Watermarks are Twins.” Studies in Bibliography 4 (1951–52): 57–91.

—. “Chain-Indentations in Paper as Evidence.” Studies in Bibliography 6 (1953–54):  181–195.

Vignau-Wilberg, Thea. Rembrandt auf Papier: Werk und Wirkung. Munich: Hirmer, 2001.

Wadum, Jørgen. “The Antwerp Brand on Paintings on Panel.” In Looking Through Paintings: The Study of Painting Techniques and Materials in Support of Art Historical Research, edited by Erma Hermens, 179–198. London: Archetype Publications, 1998.

Weislogel, Andrew C., and C. Richard Johnson, Jr. “Decision Trees and Fruitful Collaborations: The Watermark Identification in Rembrandt’s Etchings (WIRE) Project at Cornell.” In Lines of Inquiry: Learning from Rembrandt’s Etchings, 32–5. Ithaca, NY: Herbert F. Johnson Museum of Art, Cornell University, 2017. http://museum.cornell.edu/sites/default/files/DecisionTrees-WeislogelJohnson2017-LinesofInquiry.pdf

Xi, Xuelei, David Conathan, Amanda House, William A. Sethares, and C. Richard Johnson, Jr. “Automated Chain Line Marking and Pattern Matching in Radiographs of Rembrandt’s Prints.” Paper presented at the 50th Asilomar Conference on Systems, Signals, and Computers, Pacific Grove, CA, November 2016. https://doi.org/10.1109/acssc.2016.7869010

Imprint

Review: Peer Review (Double Blind)
DOI: 10.5092/jhna.2020.12.2.5
License:
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Recommended Citation:
C. Richard Johnson Jr., "Decision Trees for Watermark Identification in Rembrandt’s Etchings," Journal of Historians of Netherlandish Art 12:2 (Summer 2020) DOI: 10.5092/jhna.2020.12.2.5