JAIC 2002, Volume 41, Number 3, Article 1 (pp. 203 to 223)
JAIC online
Journal of the American Institute for Conservation
JAIC 2002, Volume 41, Number 3, Article 1 (pp. 203 to 223)




The techniques used to create silhouettes were extremely varied. The materials ranged from plain, off-white wove paper from which a hollow-cut was fashioned and mounted over a fabric scrap to the patented portable Facietrace machine used by Rembrandt Peale to trace profiles. Little is written about the historic techniques, perhaps because the technology was straightforward and available to anyone with a pair of embroidery scissors and a scrap of paper. However, when silhouettes are examined in-depth, especially those made by trained artists, it becomes clear that their manufacture can be quite involved. All the materials encountered in this project will be systematically presented. Special attention is paid to the black-coated paper used for cutout silhouettes and to tracing devices, two of the most unusual aspects of the craft. The sparse technical information that was found in the literature is also presented.


The paper used to make cutout silhouettes is relatively thin and black on one side, white on the other. Some of the coatings are irregularly or crudely applied, implying application by the artist. Occasionally papers are encountered that are colored throughout. While there is some variation in coatings from silhouette to silhouette, often papers used by different artists appear very similar. Some papers are so uniformly coated with the matte black medium that commercial production is suggested. The confirmation of commercially prepared paper would be revealing about both the economics of silhouette cutting and the business of papermaking. That is, were enough silhouettes being produced to justify a papermaker producing this paper?

The black colorant on the papers is usually a coating sitting on the surface of the paper. The coating is often relatively thick (see below), matte, and opaque, and it appears somewhat dry, as if it is leanly bound. There are very often small chunks of a black material that visually looks like bone black. Bone black is known to be difficult to grind, as opposed to the more fine lamp black, and it is not uncommon to find small chunks of charred bone. The use of bone black was confirmed by analysis, as described below. Occasional brush strokes can be seen on many of the examples.

In 1995 paper conservator Jane Smith undertook a research project at the National Portrait Gallery to investigate the black-coated paper used by Edouart, paper that fell into the category of possibly being commercially produced. Together with Smithsonian Center for Materials Research and Education scientist Walter Hopwood, Smith analyzed samples from three Edouart cutout silhouettes from the National Portrait Gallery's collection (Smith 1995a, b). The Edouart samples that Smith looked at were approximately 30 µm thick. The research revealed that the papers were coated with a combination of pigments and binder. The pigments were a mixture of bone black and Prussian blue with binders of silica and waxes, possibly paraffin and/or beeswax. The adhesive used to affix the cutout silhouette to its secondary support was probably a gum (arabic, ghattti, or tragacanth).

This author had the occasion to show some of the Yale silhouettes to Smith. The Yale Hubard Gallery and Edouart silhouettes appeared very similar to each other and to the Edouart samples that Smith tested from the National Portrait Gallery. To follow up on these observations, while the author was a Morse Fellow at the MFA, five cutout silhouettes that were visually similar to those by Edouart were chosen for analytical testing: four by the Hubard Gallery (1981.472–75)1 and one by Samuel Metford (1810–1890) (1964.1139). The reason for testing was to determine whether the papers used by different artists contained the same components as those found by Smith. If so, this finding would support the idea that the paper was commercially prepared.

MFA conservation scientists Richard Newman and Michele Derrick and the author analyzed the coatings from the five silhouettes using Fourier transform infrared spectroscopy with an attached microscope (FTIR) (table 1). The four coatings from the Hubard Gallery contained bone black, and two had

Table . FTIR Results from Black-Coated Paper Analysis
Prussian blue as well (Prussian blue was used to counter the brown hue of bone black, making the bone black appear blacker). The binders from the Hubard Gallery silhouettes were more difficult to determine using FTIR; some of the samples suggested wax and others suggested protein. Because of the small size of the objects and because preliminary identification of the pigments and binders indicated that the black coatings differed to some extent from those on the Edouart cutouts that Smith tested, it was decided not to resample to determine specific binders using a chromatography technique. The Metford coating was like the National Portait Gallery's examples in that it contained bone black and Prussian blue and was possibly bound with wax. In four of the five samples, gypsum, possibly added as a cheap filler, was also found. Smith did not find gypsum in her samples, but she did find silica, which we did not. The MFA examples differed enough from those from the National Portrait Gallery to suggest that no single commercially produced paper was used by these professional silhouettists.

Documentary evidence for commercial silhouette paper was sought in the examination of trade catalogs at the American Antiquarian Society, with particular attention being paid to stationers' and artists' material trade catalogs from 1800 to 1875.2 While decorative papers, colored tissue, and lightly tinted papers in colors like violet, gray, and fawn were found, this research did not turn up any appropriate black papers. John Krill (1997) of the Winterthur Museum kindly reviewed his research on English artists' supplies from the period 1840–1900 done at the archives of Winsor & Newton and several London museums. He also found no mention of such a paper.

Based on watermark evidence, it is known that Edouart used Whatman paper at least some of the time (Coombs 1999, 2001). Inquiries posed to What-man International Limited about the feasibility that these papers were coated by the manufacturer yielded no concrete answers, but the general consensus from the company was that the Whatman paper company did not do the coating (Ward 2001a, b).

All aspects of cutting silhouettes are discussed in the literature with the exception of paper preparation or acquisition. If artists were preparing the coatings themselves using variations on the bone black and Prussian blue recipe, why is there no mention of this? In the earliest book found on the subject, Introduction to the Art of Cutting Groups of Figures, Flowers, Birds, &c. in Black Paper (1815–16), author Barbara Anne Town-shend simply writes that one should use “thin black paper, either dyed or shiny according to taste” (n.p.). Parenthetically, Townshend also writes that one should cut from the white side of the paper, so she was referring to coated rather than dyed paper. In the 1836 edition, the author states, “The paper best calculated for this use is thin black paper, either dyed black or glazed” (Townshend 1836, 1). Townshend goes on to provide detailed information about how to set up compositions and how to cut figures versus flowers, but the acquisition of paper is mentioned as if it were as easy as a walk to the local stationer.

The availability of commercially coated papers remains unsubstantiated. Regardless, based on visual similarities among papers, the absence of directions for coating paper in the literature, the even and somewhat sophisticated appearance of some of the coatings, and the wide range of paper products available, it is likely that some black-coated papers were commercially available. If so, these papers certainly would have been used by some professional silhouettists whose businesses were busy and profitable enough to afford such paper, while some artists and many amateurs simply coated or colored the paper themselves.

Twentieth-century sources reveal that in 1938 one could acquire a black-coated silhouette paper called “krome kote,” which usually had a white, pregummed verso for affixing the silhouette to a secondary support (Diehl 1938). A number of modern sources referred to “surface paper,” which was described as thin, evenly black on one side, and white on the reverse (Swannell 1929; Leslie 1939). Quimby (1953) suggested that though requesting “silhouette paper” could draw only a blank stare from the clerk at the stationery store, it did exist.

Coated silhouette paper is still available today. It is an American-made product that comes from American Craftlines (1997a, b) and is available through Dick Blick, an artist's supply company (item 2630006). The paper has a matte, rich black surface on one side and is white on the other. This paper was considered a candidate for replacing lost or very degraded hollow-cut backing papers, such as for the Yale University Art Gallery silhouette of Augustus Street (1916.2). In the end, a paper from Light Impressions, TrueCore Card Stock, was chosen instead for the Street silhouette because more was known about it and it was not alum-sized, even though it was thicker than the papers used historically. The Light Impressions paper is acid and lignin-free. Fiber examination showed both cotton and wood; Graff's C-stain indicated there was purified wood and cellulose. Testing with the surface pH electrode meter showed the pH to be near neutral, though Light Impressions lists the pH at 8–9.5. Testing was done with the Tri-test kit, an Abbey pH pen, and a Beckmann Zeromatic SS-3 surface pH electrode meter. The fiber samples were examined with a polarizing light microscope at 65x. (Testing of the American Craftlines paper revealed that it contains purified ground softwood and alum and is slightly alkaline and that the color is not water-soluble. Through personal communication with the company making the paper [American Craftlines 1997a, b], the author was told that the fibers were wood, the sizing was starch, and the coating was a water-based pigment.)


The paper used for hollow-cuts was most often cream-colored and wove, though laid paper was also used. Watermarks, usually only partial because of the small size of the objects, are occasionally encountered. For example, from watermark evidence it is known that the Peale family of artists from Philadelphia consistently used wove paper from the Thomas Amies Mill on a Schuylkill River tributary (Sellors 1948). The author's examination of 89 hollow-cuts by the silhouettist William Chamberlin (1790–1860) at the American Antiquarian Society revealed the use of many different papers; they were all cream, but some were rough while others were smooth, some were thin and others thick. Some, though not all, of the wove papers were machine-made. From the fragmentary watermark evidence, it is also clear that the papers came from various mills. This finding demonstrates again that many silhouettists used whatever was at hand.

That some hollow-cuts were made from laid paper was somewhat surprising. One might think that wove paper, readily available from the 1790s, would have been decidedly more desirable for its more uniform, smooth texture, as it would have been less distracting visually. However, the use of both laid and wove paper simply reinforces that silhouette cutting was a common art; it was practiced by people with widely varying aesthetics and artistic concerns, access to materials, seriousness of endeavor, and financial means.

The backings for hollow-cuts were usually black and often from a household material such as a textile fragment or a scrap of paper. It is common to find backings unattached, that is, the hollow-cuts were simply laid on top of the backings. Some paper backings were matte black while others were very shiny. An occasional blue backing was also found and encountered in the literature (Yale University Art Gallery, 1947.440, 1947.441; Mègroz 1949; Laughon and Laughon 1986). It is not unusual, as well, to find unbacked hollow-cuts that have either been separated from their backings or were never backed.


Overwhelmingly, the dark colorants referred to in the literature were either lamp black or India ink. Although the use is not spelled out, most references refer to colorants used to “paint” a silhouette. Other uses may have included blackening a paper for cutouts or for backing hollow-cuts. One 1835 encyclopedia names India ink as the black of choice for painting silhouettes (Lieber 1835). Silhouette historian Neville Jackson (1938) stated that India ink made with pine soot, beer, or tallow smoke was used. Two other authors suggested that a rich, velvety black was obtainable by the same ingredients of beer mixed with lamp black or soot (Bolton 1914; Quimby 1953). Vernay (1911) cited lamp black specifically as the colorant used by an 1826 silhouettist. It is no surprise that lamp black or India ink were the colorants mentioned repeatedly; they are closely related inks made from carbon black. Historically, lamp black usually referred to carbon black (from various flame sources) and gum arabic, with glycerin added by the 1830s. India or Indian ink was most often described as a mixture of carbon black, gum arabic, and a fish ingredient like isinglass or fish bones, and after a certain date shellac (or another resin). Of course, the ingredients varied from recipe to recipe (Valuable Secrets 1795; Payne 1797; Nash 1810). Prepared dry watercolors were certainly obtainable commercially from the early 18th century, the cake form was available by the late 18th century, and moist pan watercolors were developed around 1815 (Cohn 1977). A source from 1814 recommends an iron gall ink—“a sort of black ink fit for painting figures, and to write upon stuffs, and linen, as well as on paper” containing gallnuts, white wine vinegar, and iron filings, with or without gum arabic—but says the use of iron gall ink on silhouettes is not common (Norman 1814, 82). As with the black-coated paper, it can be deduced from the material evidence that while many 19th-century silhouettists availed themselves of commercial products, many others made the dark colorants themselves, using a wide range of recipes.

Many silhouettes were enhanced by the addition of gouache or ink to add details to the portrait such as eyelashes, hair ribbons, and shirt collars. From their beginning, hollow-cuts were often detailed by inked-in hair and eyelashes, for example. Alternatively, cutout silhouettes were more frequently decorated after 1840 in order to compete with photography. Many artists decorated their silhouettes with graphite, ink, whites (especially for lacy collars), or gold color, the latter called “bronzing” (Roe 1970) or “touching” (Vernay 1911). Sometimes the treatment of the detailing is enough to secure an attribution; for example, William Chamberlin's hollow-cuts have a very consistent and distinctive shirt collar drawn in ink.

The various media used for detailing are diverse. Whites were described in the literature as being done in watercolor and sometimes more specifically in Chinese white (Boehn 1928; Rifken 1987), a form of zinc white that was available commercially around 1834 (Gettens and Stout 1966). Jane Smith found the whites on the National Portrait Gallery Edouart cutout silhouettes she analyzed to be an inorganic silicate similar to kaolin or talc with traces of organic material, so clearly dry white media were also being used (Smith 1995a). Accents in gold were often added in shell gold, which is usually gold powder mixed with gum arabic or egg white to form a watercolor. Shell gold has been available since medieval times (Gettens and Stout 1966; Wehlte 1967). Bronze powder may have been used as well, though less frequently. Bronze powder is a metal flake pigment made from numerous copper alloys, the combinations of which give varying tonalities of gold. Bronze powder has been available for a long time, but only inexpensively from 1860s, a date rather late for silhouettes (Gettens and Stout 1966).


Most of the sources that mention silhouette scissors in any detail date from the 20th century. The one exception, an early-19th-century book on black paper cutting, states that the scissors must have long shanks with short, sharp points (Townshend 1815–16). The 20th-century references stress the importance of choosing good-quality scissors that are slightly loose on the hinge to give the cutter flexibility and dexterity to change direction with ease (Carrick 1928; Simms 1937; Leslie 1939; Dorcy 1944). Specially constructed silhouette scissors were reportedly hard to acquire in this country (Dorcy 1944). Fiskars, a Finnish company that has been making scissors since 1830, reports making special-order silhouette scissors from carbon steel on rare occasions and general paper-cutting scissors more regularly (Linden 1998). Nel Laughon, a silhouette historian and modern-day practitioner, reports that scissors made specifically for silhouette cutting were made in Germany. She also suggests that there was little consistency in silhouette scissors because people used whatever was available to them (Laughon 1997). This observation is no doubt true for amateur silhouettists; people picked up whatever they had at hand. The professional artists, however, had supplies devoted to their trade. The scissors most likely used were those for needlework, as these tended to be small and sharp with long handles relative to the short tips, again for ease of manipulation and for short, well-controlled cuts. Again and again in the literature, embroidery scissors are mentioned, and they were the known choice of Edouart (Bolton 1914; Diehl 1938; Leslie 1939; Piper 1970; Oliver 1977).

That knives were used to do some cutting is well established (Carrick 1928; Dorcy 1944; Rubi 1972). Coke, a silhouette historian, found examples where the cutter had written “Cut with a Knife” on the artwork (Coke 1913, 187), and he proposes that much of early paper cutting, in monasteries in particular, was done with a knife though he does not specify a date for this practice. There are also references to the use of a stiletto, which is an awl or stylus, in combination with scissors (Coke 1913; Roe 1970). Needles were apparently also used for very fine work (Quimby 1953). While it is often possible to tell which side of the paper a silhouette was cut from, based on the edge's curl and compression, it is more difficult to determine what tool was used. In an experiment by the author, a knife and scissors made clean, satisfactory cuts, while a needle tended to drag the paper causing a frayed edge; but perhaps the needles used historically were more suited to the task. Examination of some of the finest cut details in silhouettes of the period does suggest a tool other than scissors must have been used, simply based on the small scale of some details within which it would have been very difficult to manipulate even a pair of very small scissors. It is highly likely that a small knife or sharp needle was often employed at least for details.


One of the most interesting aspects of silhouette history is the mechanical means of capturing a profile.3 The original apparatus, developed in France by Gilles-Louis Chrétien in 1786, was brought to this country in the 1790s by a group of French émigrés. Many of the early tracers were used for capturing a profile for small engravings or to get down the general countenance prior to painting a miniature. Charles Balthazar Julien Fevret de Saint-Mémin (1770–1852), for example, used a device called in French a physionotrace for tracing and reducing a profile for portraits and very small engravings (Miles 1994).

The best-known example of a tracing device used in this country was the physiognotrace developed by John Isaac Hawkins. Hoping to promote his invention, Hawkins gave the device to Charles Willson Peale for use in the Peale Museum (Sellors 1948). Peale, a man of many talents and extraordinary energy, opened a museum in Philadelphia in 1785, from which he hoped to fashion a series of national museums for democratic education in arts and sciences. Two more museums in New York and Baltimore eventually opened. The Philadelphia version of the museum was extremely popular and well attended. By 1802, in addition to exhibiting portraits of distinguished Americans, minerals, fossils, a mastodon skeleton, wax figures of Indians, war equipment, and anatomical deformities, Hawkins's physiognotrace was located at one end of the Long Gallery.

Guests could make their own hollow-cuts, or they could get assistance from Moses Williams, a former slave of Charles Willson Peale, who ran the physiognotrace. One broadside for the museum read, “The Profile Cutter attends every day and evening— Frames furnished at the door” (Peale's Museum 1818). Eventually almost all visitors wanted Williams to make the silhouette for them, a service for which he charged eight cents. It should be said that Charles Willson Peale made only a few silhouettes himself, but the family name (or “Peale Museum”) is generically used as an attribution for these objects, as often the actual cutter is not known. Williams cut from 1802 until approximately 1813, and at least two others worked the physiognotrace after Williams— Elizabeth Hampton and Elizabeth Meigs (National Portrait Gallery Curatorial File; Reese n.d.).

According to Peale, having one's silhouette taken by the device was the rage from 1802 to 1805, and it remained fashionable through the first decade of the 19th century (Miller 1988). After that time, enthusiasm waned somewhat, though silhouettes made with tracing devices remained moderately popular until supplanted by the camera in the 1840s.

John Hawkins's machine differed from Chrétien's and others in that it traced around the actual face with a small bar (an “index” made of brass) connected to a pantograph that simultaneously reduced the silhouette to less than 2 in. Examination of a reproduction of the physiognotrace in the exhibition The Peale Family: Creation of an American Legacy, 1770–1870, at the Corcoran Gallery of Art in 1997, revealed that the opening where the paper fit in was about 4 × 4 in. It was not uncommon for four hollow-cuts to be done at one time by folding the paper twice (Sellors 1948; Miller 1997a), and hollow-cuts exist that have never been cut apart. Peale explained his device (fig. 6) in a letter to Thomas Jefferson in January 1803:

The person to be traced, setting in a Chair, rests their head on the concave part, and the hollow of the board below imbraces the shoulder. The Physiognotrace is fixed to the board, A at a, and in the center of the joint b, is a conic steel point with a spring to press it against the paper. … This index moving round to trace any subject that the edge is kept too, as it moves, the steel point of the upper joint, gives a diminished size a perfectly correct representation (quoted in Miller 1988, vol. 2, pt. 1: 481–82).

Though Hawkins and the Peales tried to protect the physiognotrace from being copied through patents, there were many versions being used all over the East Coast, some of which predated Hawkins's. Benes writes, “Scores of portraitists, artists-entrepreneurs, and mechanicians rushed to take advantage of the popularity of cheap, machine-made profiles” (1994, 139). Many contraptions were very similar, though their users made fraudulent claims of improvements over Hawkins's patented version, while others were truly different. Some used optical projection, like the camera obscura, to capture a profile on paper that, while being traced, was reduced with an attached pantograph at the same time. This operation had the advantage that the sitter was not “scraped with the machine” (Benes 1994, 141), as happened when the brass index of Hawkins's machine passed over one's features. One such invention is described below:

The operator placed the sitter in a darkened room and projected the profile against a paper-covered pane of glass by means of a single light source positioned at the far end of a five-foot “trunk” or box. The pantograph traced the sitter's shadow … the other end traced a smaller image on a sheet of paper (Benes 1994, 140).

Fig. 6. The drawing and text of a letter sent by Charles Willson Peale to Thomas Jefferson, January 28, 1803. Thomas Jefferson Papers, Library of Congress, Washington, D.C.

Though “physiognotrace” was the commonly used term for tracing mechanisms, there were other names for the various inventions, including the Ediograph, Limomachia, Pasigraph, Prosopographus, Profilograph, Charles Schmalcalder's Delinator, Copier, Proportionometer, and William King's “patent delineating pencil.”

When examining hollow-cuts, it is often possible to find evidence of a tracing apparatus. Either a graphite or metal tip was used, and often traces of graphite or indented lines from a stylus are visible along the outline of the face. However, since the profiles were cut after tracing, evidence of the tracing tip has often been trimmed away. Also, since multiple hollow-cuts were made at once, frequently it is only the top piece of paper that bears the evidence. Hollow-cuts have another characteristic that can reveal their traced roots: they tend to be more generically and formulaically handled than other silhouette types, a characteristic that becomes apparent after looking at silhouettes even briefly.

Copyright © 2002 American Institution for Conservation of Historic & Artistic Works