Attachment B - Analysis Report.pdf

© 2023 Longevity Art Preservation

Date: February 9, 2023 Prepared by: Emily MacDonald-Korth

Artist: Charles Willson Peale (1741 – 1831, American) Title, Date: Washington at Princeton, c. 1779 Materials: estimated oil on canvas Dimensions: 96.25 inches x 60 inches, 244.5 cm x 152.5 cm, unframed Distinguishing marks: none

Client: Office of Cultural Heritage, U.S. Department of State, Bureau of Overseas Buildings Operations (OBO), Tobin N. Tracey, Director, traceyt@state.gov, 571-226-9585, (mobile) 202-431-1698

Analyst: Emily MacDonald-Korth, President and Chief Conservator-Scientist, Longevity Art Preservation LLC, emilymk@artlongevity.com, Miami, FL, 917-684-1065

Figure 1. Overall, unframed. Figure 2. Overall, verso, unframed.

ANALYSIS REPORT

Charles Willson Peale, George Washington at Princeton, c. 1779 Scientific analysis of materials: Easel painting

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 2

Contents

Executive Summary

Introduction ....................................................................................................... Error! Bookmark not defined.

Purpose .............................................................................................................. Error! Bookmark not defined.

Description of Analytical Methods

Analytical Procedure

Peale portraits of Washington, paintings studied and methods used

Compositional component terminology

Sampling and optical microscopy

Results

Summary of structure and materials identified in this study

Sample table

Sample location map

Condition examination report

Preventive conservation recommendations for immediate and long-term preservation

Conservation treatment recommendations

Comparison to image of the Laurens Washington at Princeton from 1946

Visible light imaging

Ultraviolet radiation imaging

Infrared radiation imaging

X-ray fluorescence spectroscopy

Cross-sectional optical microscopy

Discussion

Original materials per compositional component

USDOS Washington at Princeton compared to other paintings in the study: Infrared imaging

USDOS Washington at Princeton compared to other paintings in the study: X-ray fluorescence spectroscopy

USDOS Washington at Princeton compared to other paintings in the study: Cross-sectional optical microscopy

Literature review and correlations

Original painting structure, materials, their timeline of availability, and corresponding materials and techniques used by Charles Willson Peale

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 3

Conclusions

References

Appendix I: Analytical Techniques

Appendix II: Visible Light Images

Appendix III: Visible and UV Images

Appendix IV: Visible and IR Images

Appendix V: Visible, Ultraviolet, and Infrared Images

Appendix VI: p-XRF Data

Appendix VIII: Photomicrograph Data Sets

Appendix IX: Pennsylvania Academy of Fine Arts, Washington at Princeton, 1779

PAFA Analytical Techniques

PAFA Visible, UV, and IR Images

PAFA p-XRF Data

Appendix X: Mount Vernon, Washington after the Battle of Princeton, 1780

Mount Vernon Analytical Techniques

Mount Vernon Visible, UV, and IR Images

Mount Vernon p-XRF Data

Mount Vernon Sampling Locations

Mount Vernon Sample Table

Mount Vernon Cross-sectional optical microscopy

Appendix XI: Senate, Washington at Princeton, 1779

Senate Analytical Techniques

Senate Visible, UV, and IR Images

Senate p-XRF Data

Appendix XII: Château de Versailles, Washington after the Battle of Princeton, 1779

Versailles Analytical Techniques

Versailles Condition Examination Report

Appendix XIII: Colonial Williamsburg, Washington after the Battle of Princeton, 1780

Colonial Williamsburg Analytical Techniques

Colonial Williamsburg p-XRF Data

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 4

Executive Summary

Longevity Art Preservation carried out analysis of Washington at Princeton (attributed to Charles Willson Peale, c. 1779). The purpose of the investigation was to determine whether the USDOS painting was properly attributed; it was listed in their records as the Laurens-Albemarle painting from 1779, without provenance. Longevity was hired to carry out conservation evaluation and scientific analysis of the USDOS painting. Longevity subcontracted Carol Eaton Soltis (Project Associate Curator at the Philadelphia Museum of Art and Peale scholar) to carry out a complementary provenance and connoisseurship study;

Soltis will submit a standalone report. Longevity studied and/or analyzed paintings by or attributed to Charles Willson Peale at PMA, PAFA, Mount Vernon, the White House, the Senate, and Versailles; Colonial Williamsburg shared p-XRF data. Data from those studies created a baseline for the materials, techniques, and condition of Peale painting of Washington at Princeton c. 1779; the data was then compared to the results of analysis of the USDOS painting, with a focus on attribution. The following analytical techniques were employed: visual examination, condition examination, technical photography (digital imaging under visible, ultraviolet, and infrared radiation), local elemental analysis using portable X-ray fluorescence, sampling and sample preparation, cross-sectional optical microscopy, literature review, and reporting.

Extensive restoration was present across the painting surface including some regions of heavy overpaint obscuring the portrait of Washington. Overall, the painting was largely stable and the general appearance and condition was consistent with a heavily restored painting of this type from the 18th century. The painting was not signed and no inscriptions were visible on the front or reverse of the painting.

Distinct correlations with the last known photo of the Laurens painting, from 1946, were identified.

Infrared radiation allowed original compositional changes (pentimenti) made by the artist to become visible, and the most significant locations were located along the edges of sitter’s boots and the edges of the proper right arm and hat brim, suggesting the artist slightly changed the position of the arm and boots. Compositional changes to the position of the hat, boots, proper right arm, and proper right shoulder were observed in similar positions on all full-length paintings in the study for which infrared imaging was carried out. Portable XRF was carried out to get an initial understanding of the pigments present on the painting, suggesting the following pigments: lead white (2PbCO3 · Pb(OH)2), calcium carbonate (CaCO3), iron oxide pigments (Fe2O3 · H2O + varied, iron oxide), organic red lake, bone black (C

· Ca3(PO4)2), Prussian blue (Fe[Fe(CN)6]3), vermilion (HgS), and Naples yellow (Pb3(SbO4)2). All pigments used in the original design layer were available in the 18th century, correspond to the findings from analysis of other paintings in the study, and results published in previous Peale studies (Ashworth et.al.)

The general structure of the painting, including type and material of the canvas support and ground layers, were identified on the Mount Vernon samples, and correspond to some findings published in previous Peale studies (Ashworth et.al.) Based on the findings of this study and comparable research on works by Peale, the original materials, condition, artist techniques, and design of the USDOS Washington at Princeton correspond to other Washington at Princeton paintings by Charles Willson Peale carried out

c. 1779, the provenance connects the USDOS painting to the Laurens-Albemarle painting, and the USDOS Washington at Princeton is properly attributed to Charles Willson Peale.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 5

Description of Analytical Methods

A variety of analytical techniques were utilized to research the painting. Non-destructive macroscopic and microscopic techniques of examination and analysis were used and observations were recorded in writing, with digital photodocumentation, and with digital mapping.

A conservation condition examination was carried out to document the condition and construction of the painting; this included detail photography, microscopy and photomicrography, and graphic condition mapping.

Ultraviolet radiation (UV) was used to locate and characterize varnish layers, identify fluorescent pigments, and distinguish areas of restoration.

Infrared photography (IR) was used to image layers beneath the design layer, such as underdrawings, compositional alterations, and inscriptions.

Portable X-ray fluorescence (p-XRF) was used to gather information about the elements used in the pigments; this technique results in preliminary characterization of pigments and other materials used on the object.

To obtain information about the stratigraphy of the paint layers, pigments, binders, and fibers present, minimally invasive methods of analysis were employed. Paint samples were taken on a very small scale and sampling was kept to a minimum to preserve the work in its most intact form and to limit the need for later intervention.

Cross-sectional optical microscopy (XSEC) provided information about paint layer stratigraphy and some characterization of fibers, binders, and coatings based on the morphology of layers and particles, and autofluorescence of the layers under varying wavelengths.

Analytical Procedure

From April 2022 through February 2023, Longevity carried out visual evaluation, conservation condition examination, technical imaging (under visible, ultraviolet, and infrared radiation), p-XRF, and other scientific research on paintings of George Washington by Charles Willson Peale. Longevity collected paint samples from two paintings, Mount Vernon and USDOS, to identify the original layering structure, pigments, fibers, and binding media used in the original execution.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 6

Peale portraits of Washington, paintings studied and methods used

The following paintings were analyzed using the methods listed. All work was carried out by Longevity Art Preservation except p-XRF analysis of the Colonial Williamsburg picture which was carried out by conservators Kirsten Travers (Conservator and Materials Analyst, Colonial Williamsburg Foundation) and Shelley Svoboda (Senior Conservator of Paintings, Colonial Williamsburg Foundation).

Collection Painting title Painting date

Methods of analysis carried out

Notes

Pennsylvania Academy of the Fine Arts (PAFA), Philadelphia, Pennsylvania

George Washington at Princeton

1779 Technical imaging, p-XRF

The first painting in the Princeton series

George Washington’s Mount Vernon (Mount Vernon), Mount Vernon, Virginia

George Washington after the Battle of Princeton

1780 Technical imaging, p-XRF, sampling, cross-sectional optical microscopy

Three-quarter length portrait

White House Collection, Washington, D.C.

George Washington 1776 Visual inspection An earlier portrait of Washington; not part of the Princeton series

The United States Senate, Washington, D.C.

George Washington at Princeton

1779 Technical imaging, p-XRF

Musée et Domaine national de Versailles et du Trianon, Château de Versailles, Versailles, France

George Washington after the Battle of Princeton

1779 Condition examination

Attributed to Charles Willson Peale; oil on “toile”

Colonial Williamsburg Foundation, Williamsburg, Virginia

Washington after the Battle of Princeton

1780 XRF (carried out by Colonial Williamsburg)

Oil on linen ticking

US Department of State, Paris, France

George Washington at Princeton

c. 1779 Condition examination, technical imaging, p-XRF, sampling, cross-sectional optical microscopy

Referred to as the Paris picture, Laurens painting, or Laurens- Albemarle painting

Table 1. Paintings studied and methods of analysis used in the project.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 7

Compositional component terminology

Terminology describing the various compositional components of the painting was developed to aid communication regarding the materials and techniques identified in each region of the painting.

Figure 3. Compositional component terminology diagram.

Sampling and optical microscopy

Paint samples were taken on a very small scale (generally 0.5 – 1 mm x 0.5 - 1 mm) to preserve the work in its most intact form and to limit the need for later intervention. Low power magnification was used to determine the most appropriate areas. The sampling site was first photographed, then a microscalpel was used to remove the sample from the painting, the sample was transferred to a glass slide and labeled. All samples were named and numbered according to the order in which they were collected. In the laboratory, the samples were examined with a stereomicroscope under low power magnification (10-20x) to identify the purest areas of sample and those with the most complete layering structure; accordingly, having the potential to give the most information during analysis. Remaining sample material that was not mounted for examination was retained for future studies. Samples aimed at studying the stratigraphy of the painting were cast in resin cylinders, sanded, and polished to expose the cross-section surface.

Once mounted, the cross-section samples were examined and digitally photographed under lights of different wavelengths at 40x – 400x. See Appendix I for Analytical Techniques, Appendix VIII for Photomicrograph Data Sets including sampling site images, and Appendix VI for XRF Data.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 8

Results

The results from individual analyses using the techniques listed below will be discussed in the following section: condition examination, comparative analysis with a 1946 photo of the Laurens painting, technical photography (digital imaging under visible, ultraviolet, and infrared radiation), local elemental analysis using portable X-ray fluorescence, and cross-sectional optical microscopy.

Summary of structure and materials identified in this study

Based on the results of analysis, the primary painting support is medium to heavy weight plain weave linen canvas with a thread count of approximately 14-15 threads per cm, the canvas may have been sized; over the canvas, two ground layers were applied, one light yellow ground layer with veins of translucent brown, likely binding media, and red-orange and black pigment particles followed by a whiter ground layer with little to no pigment particles visible in the sample, and over the ground layer is a possible sealant layer, perhaps toned; based on the morphology of layers and particles and autofluorescence of the layers under varying wavelengths, the results suggest an oil-based binder in the ground layer. Elemental analysis suggests the following colorants were used to make the painting: lead white, calcium carbonate, iron oxide pigments, organic red lake, bone black, Prussian blue, vermilion, and Naples yellow.

Sample table

The following samples were collected from the painting and analyzed using the methods listed.

Sample name Sample description Analysis used

S1 Varnish; estimated restoration varnish. From filled loss. XSEC

S2 Canvas and ground layer. XSEC

S3 Canvas, ground, paint, and varnish. XSEC

S4 White paint on verso. XSEC

Table 2. Sample table listing sample name, description, and types of analysis used.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 9

Sample location map

The following map shows sampling locations.

Figure 4. Washington at Princeton, c. 1779; sampling location map.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 10

Figure 5. Washington at Princeton, c. 1779; sampling location map, verso.

Condition examination report

Description The painting is rectangular and features a centrally located figure of General George Washington in a Continental Army uniform of beige pants and vest under a dark blue jacket with yellow metal buttons and complex epaulets, a cross-body blue sash with highlights giving it the appearance of silk, and knee-high black boots with white metal spurs. Washington is leaning on a cannon with his left hand, his right hand rests on his right hip holding his hat with the interior partially visible, and his weight is on his right leg with the left leg crossed in front at the ankle. The figure stands on bare turf, with Hessians flags in a leaning bundle beneath the cannon on his left, and a British flag lays on the ground to his right. In the middle ground on the sitter’s right is green turf, foliage, and two American soldiers on horseback. In the middle ground, on the sitter’s left, is a brown horse being held by a groom in Continental Army uniform positioned to show a side view portrait and the groom’s torso with a partial horse head located behind the groom. Flying above the groom and horse is a blue Continental Army flag with a circle of 13 white, six-pointed stars. In the background is a troop of marching continental and British soldiers, Princeton University buildings, and a largely gray sky with some orange and blue areas. The painting is not signed, and no inscriptions are visible on the front of the painting.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 11

Supports The auxiliary support is a heavy expandable wooden stretcher comprised of four members with two vertical crossbars and three horizontal crossbars; sixteen keys are present, and two keys are missing (upper PR and upper center PL). Several keys are wired to the stretcher with colorless monofilament line.

On the reverse, various rub marks, abrasions, scratches, dings, marks, framing holes, stains, and a moderate layer of dirt are visible across the surface. Several inscriptions and adhesive-backed labels are present and details of each are included in Figures 6-8. The frame was not examined, however labels on the reverse of the frame are included in Figure 9. The stretcher is square, stable, and in good condition.

Figure 6. Verso, detail images of adhesive-backed labels, inscriptions, and keys.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 12

Figure 7. Detail, verso, showing adhesive-backed label on the upper stretcher bar. Figure 8. Detail, verso, showing adhesive-backed label on the upper stretcher bar.

Figure 9. Left; labels on the reverse of the frame. Figure 10. Middle; detail showing the secondary support canvas under magnification; each line on the scale on the right represents 1 mm. Figure 11. Right; detail showing thicker regions of paint applied on top of the overall paint layer on the lining canvas verso.

The secondary painting support, or lining canvas, is a heavy weight plain weave canvas with a thread count of approximately 10-12 threads per cm. The lining adhesive type is unknown. The tacking margins extend from the edges of the primary canvas, wrap around the stretcher bars and are tacked on the reverse. Light beige paint is present across the entire reverse of the lining canvas with several regions of heavier paint application (see Figures 10-11). The thick layer cleaved from the paint on the reverse during sampling suggesting it was applied after the overall paint layer was dry. Delamination between the primary and secondary supports is not evident thus the lining adhesive layer appears to be sufficient. The secondary support is stable and in good condition.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 13

Figure 12. Left; loss at upper PL corner showing framing tape, ground layer, original canvas, and tacks.

Figure 13. Right; detail under magnification showing the canvas weave visible beneath the paint layer, the ground layer visible between brush strokes, and fly specks on the surface; each line on the scale on the right represents 1 mm.

The primary painting support is medium to heavy weight plain weave linen canvas with a thread count of approximately 15 threads per cm. Because the primary support is lined, the reverse of the original canvas is not visible. Framing tape covers the tacking margins and there is little original canvas available for observation; one small area of canvas is visible at an area of loss on the upper PL corner (see Figure 12) and also reveals tacks in the original tacking margins. The canvas weave texture has been flattened during lining making thread counting challenging (see Figure 13).

Figure 14. Graphic condition map.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 14

Figure 15. Left; raking light image. Second from left; planar distortion diagram showing the location of the most significant planar distortions in red. Figure 16. Third from left; verso, raking light, grayscale, flipped horizontally to align with the orientation of normal viewing, and image adjusted to enhance the visibility of thick paint locations. Figure 17. Right; planar distortion diagram overlaying the flipped verso raking light image shown in Figure 18.

Ground layer The ground layer is exposed at paint losses and behind thin paint and it appears light beige (see Figure 19). The ground layer, though limited evaluation was possible, appears to be stable and in good condition.

Figure 19. Left; detail under magnification showing the ground layer visible behind thin paint. Figure 20.

Right; lower PL corner, showing loss of varnish, exposed ground layer, and paint loss.

Paint layer The paint layer features a direct application of fluid paint worked wet into wet with low to moderate impasto. See Figures 25-27 for detail images. Brush marks are visible throughout the paint layer. Most of the paint is thin, with some thicker regions on the epaulets and other intricate areas requiring several paint layers. A significant amount of paint on the surface is estimated to be restoration layers. Framing

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 15 paper wraps onto the front of the painting approximately ¼ inch on all sides. Stretcher bar marks in horizontal and diagonal locations near the corners are visible.

The following conditions issues were observed and marked on the graphic condition diagram (Figure 14):

loss of paint; flaking; incipient flaking; loss of framing tape, varnish, and paint; bulge; dent; impact crack;

white haze; and loss of varnish. Cracking is present throughout the paint layer. The craquelure appears to be mainly ageing cracks but there are several areas of impact cracks visible as concentric circles (see Figure 14). Areas of varnish and paint loss are located at each corner. An area of paint loss, flaking, and white haze surrounding a previous repair is located above the horse’s head near the bayonets (see Figure 21). An area of incipient flaking and white haze is located at the upper edge on the PL half (see Figure 22).

Another region of white haze is located at the central PL edge in the groom’s jacket. A line of white haze, with the morphology of a drip mark, is located in the lower PR quadrant in the red British flag. Discolored retouching and inpainting was observed across the entire painting. Evidence of multiple retouching campaigns is present; wide swaths and broad brushstrokes of retouching, fine brushstrokes of retouching, small islands and inpainting, and speckles similar to splatter paint were observed. Frame rub is present along all edges. Metallic gold residue, likely from a gilded frame, is present in some areas along the edges. Several fly specks are located across the painting. Overall, the paint layer is largely stable, though the paint losses, flaking, and incipient flaking require stabilization to keep from losing more paint;

once the paint layer is stabilized the condition will increase.

Figure 21. Left; detail, above the horse’s head near the bayonets, showing paint loss, flaking, and white haze surrounding a previous repair. Figure 22. Right; detail, upper edge, PL side, showing incipient flaking and white haze.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 16

Figure 23. Left; detail under magnification of dark varnish islands from previous incomplete cleaning.

Figure 24. Right; detail under magnification of speckle patterned inpainting.

Surface coating The painting has evidence of at least two varnishing campaigns; under ultraviolet radiation varnish was visible overall and small regions of a different varnish generation were present in many locations. Small islands of varnish from previous incomplete varnish removal are present across the painting and most visible in light hues (see Figure 23). Losses to the varnish layer are present near each corner, see Figure 24 for locations. The surface coating has a high sheen and the gloss level exacerbates the visual impact of planar deformations. More details about the varnish layer can be found in the “Ultraviolet radiation imaging” section of this report.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 17

Figure 25. Detail images of the faces of Washington, the groom, and the horse.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 18

Figure 26. Detail images of the American flag, Washington’s hand and uniform, the soldiers on horseback, and the marching troops.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 19

Figure 27. Detail images of Washington’s boots and the Hessian flags.

Preventive conservation recommendations for immediate and long-term preservation

Environmental controls have a major impact on a painting’s condition and long-term preservation.

Ultraviolet radiation exposure should be kept to a minimum by avoiding all direct sunlight and unfiltered artificial light. UV-filtering non-glare glazing should be used within the frame to mitigate damage from UV radiation. UV-filtering non-glare glazing has the added advantage of protecting the painting from mechanical damages. Environmental controls should be as consistent as possible with an aim for moderate humidity and low temperature. Relative humidity (RH) should be kept near 50%. Relative humidity levels above 75% can cause canvas to shrink, consequently straining the ground and paint layers, and encourages mold growth. Fluctuations in RH are dangerous for paintings because of expansion and contraction which can cause flaking and other condition issues, therefore consistent RH is essential.

Temperature should be from 16 – 25C or 60.8 – 77F at an optimal level of about 68 – 72F. High temperatures can soften some painting (and restoration) materials creating an opportunity for an otherwise durable surface to have dirt become entrained or to be marred. Very low temperatures can make some painting materials brittle or more likely to fracture. Changes in temperature also alter RH, an in some cases can result in condensation on the painting surfaces or within a frame, potential leading to serious water damage of many varieties and mold growth. Environmental sensors should be installed

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 20 near the painting and within the room where the painting is installed; some sensors can be installed within the frame. The sensor data should be available on a web-based platform for remote management and monitored regularly. Automatic alerts should be set up within the sensor software to alert preservation staff when the environment has gone outside of the safety parameters. A security boundary, stanchions for example, should be installed around the painting to keep it beyond the reach of human visitors and property caretakers. The preceding preservation recommendations apply to the target painting and are not guidelines for the care of other types of artworks.

Conservation treatment recommendations

Conservation treatment should aim to increase stability of the painting by addressing isolated areas of loss and active or incipient flaking; reinstate aesthetic harmony by reestablishing planarity, removing discolored varnish left behind in previous incomplete cleaning, and reducing restoration layers to reveal the artist’s hand; and prolong the life of the painting by increasing preservation efforts as described in the “Preventive conservation recommendations for immediate and long-term preservation” section of this report.

Comparison to image of the Laurens Washington at Princeton from 1946

An image of the Laurens painting printed in a Sotheby’s catalog from 1946, the last known photograph of the Laurens picture, was compared to the painting in person and similarities were noted. Several distinct correlations were identified. High resolution images of the USDOS painting and the 1946 image were digitally overlayed. Correlated condition features include impact cracks, stretcher bar marks, round marks that may be holes, scratches or tears, and other paint layer damages (see Figures 28-30).

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 21

Figure 28. Upper left; photograph of the Laurens Washington at Princeton from 1946. Figure 29. Upper right;

diagram showing correlations between the 1946 image and an infrared image of the Department of State Washington at Princeton annotated marked in red. Figure 30.

Lower left; diagram showing correlations between the 1946 image and an infrared image of the Department of State Washington at Princeton annotated in red.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 22

Visible light imaging

Overall recto images were captured along with detail images of the entire painting in nine sections.

The results and observations of visible light imaging were discussed in the previous section, “Condition Examination Report”. See Appendix I for Analytical Techniques and Appendix II for Visible Light Images.

Ultraviolet radiation imaging

Overall recto images were captured along with detail images of the entire painting in nine sections.

The paint layer has an overall autofluorescence consistent with certain types of binding media and pigments including oil-based paint and possibly organic red pigment. The painting has a layer of varnish across the entire surface and has evidence of at least one more varnish layer below. The autofluorescence is semi-opaque bright light blue-green. Bright white-blue flecks are visible in many locations across the painting but observed largely near detailed regions of the composition and/or areas with considerable damage such as the key fob, the sitter’s right leg, and the soldiers and buildings in the background, among others. The flecks are varnish left behind from an incomplete cleaning (see Figure 33). Across the surface of the painting, the varnish layers are not even. The varnish unevenness appears to have been caused by two things: the previous varnish layer was not entirely removed (i.e. preferential cleaning) resulting in some areas of thicker varnish thus denser autofluorescence, and the uppermost varnish layer application was not uniform showing vertical streaks. Restoration became visible under UV examination and much of the surface of the painting has been retouched. Multiple retouching campaigns became visible. Extensive retouching is present on the sitter’s face and hand (see Figures 31-32), the sky, the horse and groom, and the flags. See Appendix I for Analytical Techniques, Appendix III for Visible and UV Images, and Appendix V for Visible, UV, and IR images.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 23

Figure 31. Overall image, under UV radiation.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 24

Figure 32. Left; Detail of sitter’s upper torso showing an overall white-blue autofluorescence indicating varnish location and darker regions of paint indicating areas of restoration. Figure 32. Right; Detail of sitter’s torso and upper legs showing an overall white-blue autofluorescence indicating varnish location and darker regions of paint indicating areas of restoration; bright white-blue flecks indicate areas of varnish that were not completely removed in previous restoration campaigns.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 25

Infrared radiation imaging

Overall recto images were captured along with detail images of the entire painting in nine sections.

Most black pigments and earth pigments absorb infrared light and appear black, as seen in Figure 34 by the strengthened outlines of the costumes, objects, and facial features. Vermilion and organic red pigments do not absorb infrared light and appear transparent, which resulted in increased transparency of the pink and red flesh tones and orange hues of the sky. Infrared imaging allowed differentiation between pigments used in paints and underlayers to become visible. Because some pigments become more transparent under infrared radiation, original compositional changes (pentimenti) made by the artist became visible. The most significant locations of pentimenti are located along the edges of sitter’s boots and the edges or the proper right arm and hat brim (see Figures 35-37). The pentimenti suggests the artist slightly changed the position of the arm and boots. Areas of restoration paint also became transparent allowing for visualization of previous damage and retouching locations (see Figure 35).

Figure 34. Overall image, infrared radiation.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 26

Figure 35. IR diagram showing visible light on the left and infrared radiation on the right. The diagram points out the major compositional changes (pentimenti) and areas of restoration.

Figure 36. Left; IR image showing pentimenti along the edges of the proper right arm and brim of the hat.

Figure 37. Right; IR image showing pentimenti along the edges of the bottom and sides of the boots.

See Appendix I for Analytical Techniques, Appendix IV for Visible and IR Images, and Appendix V for Visible, UV, and IR images.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 27

X-ray fluorescence spectroscopy

Portable XRF, a method of elemental analysis, was carried out to get an initial understanding of the pigments present on the painting; 70 locations were analyzed, and the following table shows the elements detected and their inferences (Table 3). The painting was carefully examined under visible, ultraviolet, and infrared radiation to determine the best locations for analysis; areas of detectable restoration were avoided except when targeting restoration was the goal.

The following elements were identified in the XRF data: the painting: barium (Ba), bromine (Br), calcium (Ca), cadmium (Cd), chromium (Cr), copper (Cu), iron (Fe), mercury (Hg), manganese (Mn), nickel (Ni), lead (Pb), antimony (Sb), tin (Sn), strontium (Sr), titanium (Ti), zinc (Zn), zirconium (Zr). Notes and inferences for each element can be found in Table 3.

Elements detected using XRF

Notes and inferences

Barium (Ba) Very low overall signal; possible mineralogical association with strontium (Sr)

Bromine (Br) Low overall signal, high correlation with mercury (Hg); mineralogical association with mercury (Hg)

Calcium (Ca) Calcium-based pigment such as calcite (CaCO3) or chalk (CaCO3, calcium carbonate), bone black (C · Ca3(PO4)2, organic red lake

Cadmium (Cd) Present overall; restoration materials and/or mineralogical association or possible spectral effect of intense lead (Pb) signals

Chromium (Cr) Very low overall signal; likely restoration materials using chromium-based pigment such as chrome yellow (PbCrO4), orange (PbCrO4 · Pb(OH)2), red (PbCrO4 · Pb(OH)2), green Fe4[Fe(CN6)]3 + PbCrO4, chromium oxide green (Cr2O3), viridian (Cr2O3 · 2H2O), or strontium chromate yellow (SrCrO4)

Copper (Cu) Low overall signal; possible restoration materials

Iron (Fe) Iron-based pigments (Fe2O3 · H2O + varied, iron oxide), Prussian blue (Fe[Fe(CN)6]3)

Mercury (Hg) Vermilion (HgS)

Manganese (Mn) Very low overall signal with low elevation in selected areas; mineralogical association with iron-based pigments and suggests umber, manganese oxide (MnO)

Nickel (Ni) Very low overall signal; possible spectral effect of intense lead (Pb) signals

Lead (Pb) Very high signal in all locations; lead white (2PbCO3 · Pb(OH)2)

Antimony (Sb) Lead antimonate/Naples yellow (Pb3(SbO4)2)

Tin (Sn) Lead-tin yellow type I (Pb2SnO4)

Strontium (Sr) Low overall signal; correlation with calcium (Ca) and possible mineralogical association with calcium (Ca)

Titanium (Ti) Very low overall signal; possible mineralogical association with iron

Zinc (Zn) Very high signal; zinc white (ZnO) pigment (restoration)

Zirconium (Zr) Very low overall signal; possible mineralogical association with strontium (Sr)

Table #. Elements detected using XRF and the materials inferred.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 28

Careful XRF analysis of areas with apparent restoration allowed for the identification of pigments and colorants used in restoration layers resulting in the ability to distinguish restoration materials in the data, thus isolating the likely original materials for interpretation. The following elements were determined to be components of the original preparation layers and colorants used on the painting: calcium (Ca), iron (Fe), mercury (Hg), lead (Pb), antimony (Sb).

These elements suggest the following pigments: lead white (2PbCO3 · Pb(OH)2), calcium carbonate (CaCO3), iron oxide pigments (Fe2O3 · H2O + varied, iron oxide), organic red lake, bone black (C · Ca3(PO4)2), Prussian blue (Fe[Fe(CN)6]3), vermilion (HgS), and Naples yellow (Pb3(SbO4)2). Table 4 shows the distribution of the colorants in each XRF analysis location.

Location number

Location description

Secondary description

Color Ground Major Minor Trace Inferences

1 Hessian flags red

Red Pb Ca Fe Hg Red lake, iron oxide, vermilion, possibly bone black

2 Hessian

Red Pb Hg Ca Fe Vermilion, red lake, possibly bone black 3 Hessian

Red Pb Hg Ca Fe Vermilion, red lake, possibly bone black 4 British flag

Dark red Pb Fe Hg Ca Iron oxide, vermilion, red lake and/or bone black 5 British flag

6 British flag

7 GW lapel

Yellow A Pb Sb Ca Fe, Hg Lead white, Naples yellow, calcium carbonate and/or bone black, (iron oxide)

8 GW lapel

Yellow A Pb Fe Ca Sb, Lead white, iron oxide, Naples yellow, calcium carbonate and/or bone black

9 GW pants

Yellow A Pb Fe Hg Ca Lead white, iron oxide 10 GW button highlight

Yellow B Pb Fe Sb Ca, Hg, Lead white, iron oxide, Naples yellow, calcium carbonate and/or bone black

11 GW button

Yellow B Pb Fe Ca Sb, Lead white, iron oxide, calcium carbonate and/or bone black, Naples yellow

12 GW

epaulet

Yellow B Pb Fe Hg Ca Lead white, vermilion, calcium carbonate and/or bone black

13 Hessian flags yellow B

Yellow B Pb Fe Ca Hg Lead white, iron oxide, (vermilion)

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 29

14 Hessian flags green

Green Pb Fe Ca Hg Prussian blue, iron oxide, bone black, (vermilion) 15 Hessian flags green

(high calcium - bone black?)

16 Turf bone black, (vermilion)

17 Hessian flags light blue

Light blue Pb Ca Fe Hg Lead white, Prussian blue, elev calcium - bone black? Indigo??

18 GW sash

Light blue Pb Fe Hg Ca Lead white, Prussian

19 Sky light

Light blue Pb Hg Br Fe, Ca Lead white, vermilion, Prussian blue, (bone black)

20 GW jacket

Dark blue Pb Fe Ca Hg, Prussian blue, bone

21 GW jacket

Dark blue Pb Fe Ca Hg Prussian blue, bone

22 American flag

Medium

Pb Fe Ca Hg Prussian blue, bone black

23 American flag

Pb Ca Fe Hg Bone black, Prussian blue

24 Groom jacket

Pb Ca Fe Hg Bone black, Prussian blue

25 Groom jacket

Pb Ca Fe Hg Prussian blue, bone

26 British flag

Pb Fe Ca Hg Prussian blue, bone

27 British flag

Pb Fe Ca Hg Prussian blue, bone black, vermilion

28 Buildings

Purple Pb Hg Br Ca, Fe

Vermilion, red lake and/or bone black, (iron oxide), (CaCO3)

29 Buildings

Vermilion, red lake and/or bone black, (iron oxide), (CaCO3)

30 Buildings

Vermilion, red lake and/or bone black, iron oxide and/or Prussian blue, (CaCO3)

31 Sky orange

Orange Pb Hg Ca Fe Vermilion, red lake and/or bone black, (CaCO3)

32 Sky orange

Orange Pb Hg Ca Fe Vermilion, red lake and/or bone black, (CaCO3)

33 Key fob gold key

Light brown

Pb Fe Ca Hg, Sb

Lead white, iron oxide, bone black, (CaCO3)

34 GW button

Pb Fe Ca Sb, Hg

Lead white, iron oxide, bone black, (Naples yellow), (CaCO3)

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 30

35 Cannon

Pb Fe Hg Ca Lead white, iron oxide, vermilion, bone black and/or red lake, (CaCO3)

36 Horse

Pb Fe Ca Hg Iron oxide, bone black

37 Horse

Pb Fe Ca Hg Iron oxide, bone black

38 Branches at fallen soldier

Dark

Pb Fe Ca Hg Iron oxide, bone black

39 GW boot

Black Pb Fe Ca Hg Iron oxide, bone black 40 Cannon hardware

41 Cannon hardware

42 Cannon wheel

Gray Pb Fe Hg Ca Lead white, iron oxide, vermilion, bone black 43 Key fob silver charm

Gray Pb Fe Ca Hg, Sb Lead white, iron oxide, bone black, (Naples yellow)

44 GW hair

Gray Pb Fe Ca Hg Lead white, iron oxide, bone black

45 GW scarf

White Pb Fe Ca Hg Lead white, iron oxide 46 GW scarf

White Pb Ca Fe Hg Lead white, calcium carbonate and/or bone black, iron oxide

47 GW scarf

White Pb Fe Ca Hg Lead white, iron oxide, (calcium carbonate and/or bone black)

48 GW flesh forehead

Flesh A Pb Hg Fe Ca Lead white, vermilion

49 GW flesh undereye shadow

Flesh A Pb Hg Fe Ca Lead white, vermilion

50 GW flesh chin blush

Flesh A Pb Hg Fe Ca Lead white, vermilion, iron oxide 51 Groom flesh

Flesh B Pb Fe Ca Hg Lead white, iron oxide, red lake and/or bone black

52 Groom flesh blush black

53 Groom flesh shadow black

54 Arm v pentimenti arm

Special 1 Dark blue Pb Fe Ca Hg Prussian blue, bone

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 31

55 Arm v pentimenti underlayer

Special 1 Orange Pb Hg Ca Fe Vermilion, red lake and/or bone black, (iron oxide) (CaCO3)

56 Arm v pentimenti sky

Special 1 Orange Pb Ca Fe Hg Red lake and/or bone black, iron oxide, vermilion (CaCO3)

57 Branches underlayer

Special 2 Green Pb Fe Ca Hg Prussian blue, iron oxide, bone black

58 Branches underlayer

Special 2 Green Pb Fe Ca Hg Prussian blue, iron oxide, bone black

59 Branches overlayer

Special 2 Dark brown

Pb Fe Ca Hg Iron oxide, bone black

60 Groom face PR v central eye

Special 3 Flesh B Pb Fe Ca Hg Lead white, iron oxide, red lake and/or bone black

61 Groom face PR v central nose

Special 3 Flesh B Pb Fe Ca Hg Lead white, iron oxide, red lake and/or bone black

62 Restoration

GW

forehead

Special 4 Flesh A Pb Fe Ca Hg Lead white, iron oxide

63 Restoration GW cheek

Special 4 Flesh A Pb Hg Fe Ca Lead white, vermilion

64 Restoration GW chin

Special 4 Flesh A Pb Hg Fe Ca Lead white, vermilion

65 Restoration GW hand

Special 5 Flesh A Pb Fe Ca Hg Lead white, iron oxide

66 Restoration

GW

fingernail

Special 5 Flesh A Pb Fe Ca Hg Lead white, iron oxide, red lake and/or bone black

67 Restoration GW hand shadow

Special 5 Flesh A Pb Fe Ca Hg Lead white, iron oxide, red lake and/or bone black

68 Restoration sky

Special 6 Light blue Pb Hg Ca Fe Lead white, vermilion

69 Restoration sky

Special 6 Light blue Pb Ca Hg Fe Lead white, elevated calcium

70 Restoration sky

Special 6 Light blue Pb Hg Br Fe, Ca, Lead white, vermilion, Prussian blue, (bone black)

Table 4. XRF location number, location descriptions, color, elemental distribution, and inferences.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 32

Cross-sectional optical microscopy

Cross-sectional optical microscopy gave information about the layering structure of the painting including differentiation of original and restoration layers, colorant mixtures used in ground layers, canvas fiber identification, and preliminary binding media characterization. Four samples were analyzed.

General painting structure Based on the results of analysis, the following is a description of the general original structure of the painting and materials identified in this study; the complete structure described below was observed in one sample, the other three samples contained primarily restoration materials. The painting is on linen canvas, the canvas may have been sized. Over the canvas, two ground layers were applied in the following sequence: one light yellow ground layer with veins of translucent brown, likely binding media, and red-orange and black pigment particles followed by a whiter ground layer with little to no pigment particles visible in the sample. It is possible that the second ground layer is part of the first ground layer and appears distinct because of incomplete mixing. Over the ground layer is a possible sealant layer, perhaps toned; more samples are required to confirm the presence of the sealant layer. No original paint layers were present in the samples.

Binding media type Based on the morphology of layers and particles and autofluorescence of the layers under varying wavelengths, the results suggest an oil-based binder in the ground layer; no original paint layers were present in the samples.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 33

Sample name

Sample description Layers and notes

S1 Varnish; estimated restoration varnish.

Restoration

3. Dirt layer

2. Varnish layer 2

1. Varnish layer 1

100x S2 Canvas and ground layer.

200x

Generation 1

5. Possible sealant layer

4. Ground layer 2

3. Ground layer 1

2. Canvas size

1. Canvas, linen

S3 Canvas, ground, paint, and varnish.

200x

4. Varnish layer 2

3. Varnish layer 1

2. Deteriorated paint layer Generation 1

1. Canvas, linen

S4 White paint on verso.

100x

2. Paint layer 2

1. Paint layer 1

Table 5. Layering structure and materials identified using cross-sectional optical microscopy (see Appendix VIII for more detailed information and images).

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 34

Discussion

Original materials per compositional component

The following section summarizes the findings regarding the original colorants present in each paint hue of each compositional component included in the analysis. If not listed in the section below, two pigments can be considered global in most paint hues, particularly areas of light paint hue or tone:

calcium, likely in the form of calcium carbonate (CaCO3) and lead, in the form of lead white pigment (2PbCO3 · Pb(OH)2); in any area with a light hue, lead white should be considered present as a value adjuster and calcium carbonate could be present as well.

Hessian flags, red paint Results of analysis indicate the paint hue was created using vermilion, likely organic red lake, and possibly bone black.

British flag, dark red paint Results of analysis indicate the paint hue was created using iron oxide pigments, vermilion, and likely red lake and/or bone black.

George Washington, lapel, yellow paint (A) Results of analysis indicate the paint hue was created using lead white, Naples yellow, iron oxide pigments, and calcium carbonate and/or bone black. Two locations were analyzed; one location included Naples yellow and little iron oxide pigments, the other locations included more iron oxide pigments than Naples yellow.

George Washington, pants, yellow paint (A) Results of analysis indicate the paint hue was created using lead white and iron oxide pigments.

George Washington, button highlight, yellow paint (B) Results of analysis indicate the paint hue was created using lead white, iron oxide pigments, Naples yellow, calcium carbonate and/or bone black. Two locations were analyzed; one location included Naples yellow and little iron oxide pigments, the other locations included more iron oxide pigments than Naples yellow.

George Washington, button highlight, yellow paint (B) Results of analysis indicate the paint hue was created using lead white, iron oxide pigments, calcium carbonate and/or bone black, and Naples yellow.

George Washington, epaulet highlight, yellow paint (B) Results of analysis indicate the paint hue was created using lead white, vermilion, and calcium carbonate and/or bone black.

Hessian flags, yellow paint (B) Results of analysis indicate the paint hue was created using lead white, iron oxide pigments, and possibly vermilion.

© 2023 Longevity Art Preservation/Analysis Report/Peale, Washington at Princeton, c. 1779 35

Hessian flags, green paint Results of analysis indicate the paint hue was created using Prussian blue, iron oxide pigments, likely bone black, and possibly vermilion.

Turf, green paint Results of analysis indicate the paint hue was created using Prussian blue, iron oxide pigments, likely bone black, and vermilion.

Hessian flags, light blue paint Results of analysis indicate the paint hue was created using lead white, and likely Prussian blue.

Elevated calcium levels suggest calcium carbonate and/or bone black; an unidentified organic blue colorant is possible.

George Washington, sash, light blue paint Results of analysis indicate the paint hue was created using lead white and Prussian blue.

Sky, light blue paint Results of analysis indicate the paint hue was created using lead white, vermilion, Prussian blue, and possibly bone black.

George Washington, jacket, dark blue paint Results of analysis indicate the paint hue was created using Prussian blue and likely bone black.

American flag, medium blue paint Results of analysis indicate the paint hue was created using Prussian blue and likely bone black.

Groom jacket, medium blue paint Results of analysis indicate the paint hue was created using Prussian blue and likely bone black.

British flag, medium blue paint Results of analysis indicate the paint hue was created using Prussian blue, likely bone black, and vermilion.

Buildings, purple paint Results of analysis indicate the paint hue was created using vermilion, red lake and/or bone black, iron oxide pigments and/or Prussian blue.