Introduction

The first aim of this study is to determine the type of paper used for a pastel ascribed to Pierre Auguste Renoir and to establish the production date of this paper as accurately as possible. Second aim is to answer the question whether or not a photographic print is present under the pastel. This question is related to that of the nature of the paper coating, since the presence of a photographic image would influence the chemical composition of the coating. The final aim is to find out whether the paper of the investigated pastel is in any way related to papers of authenticated drawings by Renoir or other impressionists.

General description of the pastel

Provenance

The pastel has been in the collection of mr and mrs Charles Wolf from France up to 1960 when the collection was sold. A small plaque from this period gives the text:
Une ‘Etude’ pour: Les Baigneuses Pastel Par Pierre-Auguste Renoir 1841-1919

We also have a ticket with the name of Charles Wolf printed and in typescript added ‘collection of mr and mrs’. Probably one of Wolf’s business cards was used for the purpose. The ticked is attached to the backing of the frame in which the present owner acquired the pastel. The paper of the ticket is strongly discoloured, which cannot have been caused by the cardboard to which it is now attached. One explanation is an earlier backing, which was highly acidic. If this deduction is correct, the discoloration of the ticket is an indication that the pastel has been in the possession of Charles Wolf for a fairly long time. This is confirmed by the typographic style of the ticket which points to the 1930s but this of course does not exclude a later use. Before this period the pastel must have been in an environment where it incurred the damage it now shows: earlier research for instance established restoration of the border area which may indicate that the pastel was stored without proper protection.

The pastel

Dimensions of the pastel are: 71 x 55 cm. The technique is rubbed pastel. For a general description of the technical aspects and the pigments we refer to the report by Wagner and the data from the XRF-tests performed by McCrone.

The pastel in question is intimately connected with a study by Renoir for his painting ‘Les Grandes Baigneuses’. This study is now in the collection of the Musée d’Orsay. The study in the Musée d’Orsay is larger than the investigated pastel - 108 x 162 cm - and consists of three main figures. Yet the two bathers that are on both drawings are practically congruent, in spite of the different dimensions, which is evident from comparison by means of copies on transparent sheets; in the background there are a few small identical places. On the other hand the investigated pastel is strikingly different in the way the figures are coloured, in the media used and in the overall structure of the background. At some places changes and corrections have been made into the pastel: for instance, the left hand of the seated Bathers has been redrawn

From the fact that the background is structured around the composition of the two women, one gets the impression that the investigated pastel was intended as an independent work of art. This interpretation is confirmed by the presence of a contour line, probably drawn by means of an eraser, around the composition. Important from the point of the investigation into the paper support is that the colour of the paper surface is clearly integral part of the drawings background.

Renoirs studies for the Great Bathers of 1887

The drawing fits into a series of studies for the painting ‘Les Grandes Baigneuses’, exhibited for the first time at the Galerie Georges Petit in 1887. Renoir made a great number of studies for this painting, five of which represent the whole composition. At first sight the investigated pastel seems the only drawing with just the left half of the composition, but this is not so: the last study of the complete group of bathers has been cut in two by Renoir himself (both halves are signed) and finished in different techniques. The two halves should therefore be regarded as two drawings. Obviously Renoir felt the need to work on the left and right halves of the composition separately. The investigated pastel would however be the only composition study for the Great Bathers with all the colours filled in.

There may be a relation between the investigated pastel and the authenticated composition studies in the way the image is ‘cut off’. To elaborate upon this would lead beyond the scope of the present study. As the question involves Renoirs handling of paper, the discovered data are given in Appendix II.

The date and order of the various studies for the Great Bathers are important if we want to establish whether the paper of the investigated pastel could have been from the same period. From the literature it appears that Renoir studied the individual figures and the composition of the Great Bathers painting during a fairly long period. There is no consensus on the exact period: 1882 or 1884 till the beginning of 1887. The composition studies would have been done towards the end of that period. However, according to information from the Departement des Arts Graphiques of the Louvre, the composition study now in the Musée d’Orsay dates from 1881-1882. An early date for this study might be based upon the possibility that Renoir started work on the painting, but abandoned it in order to study the composition again. Taking into account that the paper may have been in stock for some time, this means that the paper of our pastel, if authentic, should date from the middle 1870s till the middle 1880s as the most likely period. A later period should however not be completely excluded.

Description of the paper

Physical description of the paper

On the basis of visual observation and direct measurements we can give the following description of the paper:

Surface: a coating is clearly visible on one side of the paper. In spite of this coating, the fibre structure of the paper under the coating is still noticeable. Creases are visible on the surface; these are in line which may mean that the paper has been kept on a (thin) roll.

Smoothness: the fibre layer as well as the coating are very smooth; both must have been glazed.

Backing: traces of brown coloured board are present on the verso of the paper.

Dimensions the paper has been cut by hand from a larger sheet. This can be seen from the somewhat irregular cutting edge of the paper. The paper as it is now has a common format, also in use for transfer paper: the dimensions are practically the paper format ‘jésus’ ( ca 72 x 55 cm). As the rectangular contour line is well within the limits of the paper, it is not likely that the artist has slightly trimmed the paper to adjust the dimensions of the composition. The original sheet must haven been considerably larger. This may indicate that a larger sheet or roll of paper has been cut into sheets of standard format by hand.

Chemical description of the fibre layer

The paper is unsized. It consists primarily of bleached sulphite pulp. Though the pulp has been bleached the paper is neither white, nor in our opinion, discoloured. Therefore we assume that a yellow dye-stuff has been added to the pulp. Dying in a light tone required bleached pulp. How the paper has been tinted cannot be decided; chrome yellow, rust yellow (iron oxide) or synthetic dyes were all used to give paper a yellowish colour. In sized paper the resin and alum act as mordants; for the investigated paper chrome and iron compounds may have been used as mordant, which could explain the somewhat higher amount of chrome in the paper.

Chemical description of the coating

A particular point of interest during the investigation has been the possibility of a photographic print under the pastel. On the basis of the chemical composition of paper and pastel this hypothesis must be rejected: a photographic print would show itself by the presence of certain elements in the photographic image, especially silver, chromium (for the carbon process) and iron (for the cyanotype process). In fact, there is a complete lack of silver and the presence of just a small amount of chromium compared to a carbon print, tested for reference. None of the chromium is however in the gelatine layer where it should be in case of a photographic image. Finally the cyanotype or blueprint process is based on the light sensitivity of iron salts. Again there is no iron present in the coating. Preparing the paper with a solution of iron salts would also have given the coating a dirty grey colour which is not present at the places where the pigments disappeared. Finally, investigation by infrared spectrography and on the light table showed no divergence from the pastel drawing. This would have been expected if a photographic or other type of print was present under the pastel.

The layer structure of the pastel

In summary the overall layer structure of paper and pastel is as follows:

pastel layer: a single layer of pigment; gypsum and chalk uniformly dispersed on the surface area; lithopone unevenly dispersed

coating: gelatine and starch; some calcium present (about 20 ppm) but no other metals

paper fibres: china clay present as filler; no size Relative thickness fibre layer - coating ca 20 - 1.

The determination of the type of paper

All properties of the investigated paper point to one single type: transfer paper. This kind of paper has been developed during the nineteenth century in order to make lithographs without having to draw on the stone itself. An equally important impulse came from the printing industry: drawings made on transfer paper could be transferred to a zinc plate that was then etched into a relief sufficient for letterpress printing. One of the names for this technique was gillotage after the inventor Firmin Gillot. Transfer paper was also used in the printing industry to transfer an image from one printing forme to another (stone to stone, plate to stone).

Most nineteenth century instructions for the making of transfer paper are based upon the addition of a coating to an ordinary type of paper. The essential ingredients of the coating are mostly gelatine and starch or flour. Sometimes a filler is added. The coating has to be soluble, the fibre layer absorbent and therefore not or slightly sized. The paper of the investigated pastel is clearly indicated by the description in Lemerciers handbook of lithography of ‘papier sans colle ordinaire avec encollage d’un seule côte’ [unsized paper with a coating on one side].

We can determine the paper more precisely as autographic paper for pen, brush and tusche drawings. The term authographic paper, in French ‘papier autographique’, refers more specifically to paper upon which an original drawing was made for whatever final purpose, as opposed to paper for the transport of images between printing surfaces (in French: ‘papier à report’). Two variants of autographic transfer paper existed from the 1870s on: a paper with a smooth coating, as the one used for the pastel, and a paper for crayon drawing, with a thick coating containing a filler and often an embossed texture, for instance an aquatint corn, in the coating. The fibre layer of most transfer papers was probably rather thin, but especially so when the paper was intended for transport from one printing surface to another; then China paper was often used. The fibre layer of autographic papers had to be thicker and stronger in order to allow the manipulations of the artist; it also had to be smooth for pen drawing. The properties of the paper support of the pastel are in agreement with these prerequisites.

Transfer paper could be home made by artist or printer, using any kind of suitable paper. However, we don’t believe many artists took the trouble of preparing the paper themselves. In our case the smoothness of the coating must have required a press. Therefore as far as the coating is concerned the paper could have been made in a printing establishment for its own use or be produced for the market.

Transfer paper was manufactured for the market by a limited number of firms. These were often also major printing establishments. In France the lithographic printing firms of Lemercier and Clot are known producers. For the German market we find only four native factories mentioned around 1900: Angerer & Göschl from Vienna, Klimsch & Co, Krebs and Schaeuffelen. At least two of these, Angerer & Göschl and Klimsch were also illustration printers. As far as the producers of transfer papers were not themselves paper makers, they probably made the coating themselves and may have acquired the fibre layer directly from the paper factory, which allowed them to make special demands in regard to the paper properties on the basis of their specific experience and needs

The fibre layer of the investigated pastel must be such a raw paper especially made for the production of transfer paper because the only other type of unsized paper that could have been acquired on the market was printing paper which the researched paper definitely is not. In the technical literature on paper and paper making we found no evidence that uncoated transfer paper could be acquired through the general paper trade, though some some indications suggest that such a paper was sold in France by suppliers of lithographic articles. The most likely interpretation of the paper of the investigated pastel seems to be that it was produced for the market by a paper factory or another kind of producer. The background of this producer may explain an anomaly in the make up of the investigated paper.

The make up of the pastel paper

In order to distinguish the two sides of the paper for the convenience of the user the coating of transfer paper was often provided with a colour. This seems to have been in particular the case with autographic paper, so much so that in France it was simply called ‘papier jaune’. The fact that all recipes for transfer paper prescribe the addition of colouring materials to the coating points to the use of white paper as a support: a coating containing starch paste or a white filler would be perfectly visible on a tinted paper. This is confirmed by the fact that the still extant transfer papers of Joseph Pennell in the Library of Congress all have a white paper as basis. The pastel paper is therefore atypical in the sense that the colour is added to the pulp, not to the coating. As we concluded that the fibre layer of the paper was specially produced to make transfer paper, a reason for the anomaly must be present.

The particular make up of the investigated paper may be explained by the requirements made upon transfer paper in the 1880’s. In this period the gillotage process was further developed by Charles Gillot, son of Firmin: besides the direct transfer of images from paper to zinc, the indirect transfer by means of photography - photo gillotage or photo-etching - now became a commercial option. Indirect, photographic transfer, from paper gave a better reproduction but required that a drawing be made on a flat, white paper. A special paper was created for this purpose again by Charles Gillot, ‘papier procédé’ or scratchboard, a kind of Bristol board with a white coating. Though we do not have positive evidence, it does seem logical however that in the early period of photo gillotage transfer paper has been used for the purpose. Basically what was needed was a paper with a white coating, that is transfer paper with a coating not ‘spoiled’ by a colouring substance. The required stiffness could simply be achieved by attaching board to the paper; fixing a temporary board to transfer paper was in any case advisable, when it was used for drawing. Even after the development of papier procédé it might have been useful to have a paper that was suitable for both kinds of gillotage as they were usually done by one and the same firm. We presume that the pastel paper belongs to this category of transfer paper. The yellow colour of the coating can then be explained by the fact that under the specific circumstances, the only way to mark the right side of the paper for drawing, was by tinting the fibre.

The production date of the paper.

Appearance and chemical composition of the paper taken together point to an early production date in the last decades of the nineteenth century.

Arguments on the basis of the Chemical composition

An important point in dating the paper of the pastel is the chemical composition of the fibre. The presence of sulphite pulp gives an earliest production date of ca 1880. The fact that the pulp has been bleached may presents another opportunity for dating the paper. The most common, in some countries almost universal, method of bleaching in the early twentieth century was by means of bleaching powder (calcium hypochlorite). In this case, however, we would expect the presence of calcium in the fibre layer of the paper. When paper fibres swell in water, as during the paper production stage, they adsorb a certain amount of soluble material. In addition, because of the electronegative character of cellulose, positively charged particles, like calcium-ions, will be bound to the paper. The lack of calcium in the pastel paper, as demonstrated in the performed tests, indicates another bleaching method, viz. chlorine gas. Chlorine gas was used in the early days of cellulose production, until the 1890’s, but abandoned because too much fibre was lost in the process.

Arguments based on the make up of the paper

The colour of the fibre layer was very fashionable for certain kinds of late nineteenth century paper; it was especially typical for drawing paper. This type of paper was called ‘papier bulle’ in French. We find a similar tone in the paper of various drawings by Renoir, described as ‘papier beige’, or ‘buff paper’. A second argument based on the make up of the paper comes from the fact that the fibre layer is coloured in stead of the coating. This is best explainable by the particular circumstances in the 1880s. This argument has already been explained in the paragraph on the determination of the paper type.

The overall impression of the pastel support is that it consists of an ordinary paper to which a coating has been added. In this it has an old fashioned appearance. Modern, factory made transfer paper we have seen has a more uniform make up and a higher glaze than the paper of the pastel. In reconstructing the design of late nineteenth and early twentieth century transfer paper, we are hampered by the lack of extant papers. Transfer paper was usually destroyed during the transfer process. On the other hand we expect that the usual uniformity of industrial products of a period will also be apparent in the case of transfer paper. The modern ‘design’ of transfer paper dates back from at least the beginning of the twentieth century. A sample of autographic paper from the firm of Klimsch & Co appeared to be a thin, very smooth paper with a greyish coating. This type of colouring for the coating was perhaps also typical at the time, because we find it back in samples of a special kind of transfer paper, Steinpapier, from the firm Angerer & Göschl.

The modern look of (German) transfer paper seems to have been quite stable: in the sample book of the Aschaffenburger Buntpapier Fabrik (1937) at the Dutch Royal Library we found very thin, highly glazed and white coated transfer paper both for the purpose of “Umdruck” (transfer from one printing surface to another) and for authography or original drawing. It was also present on the Anglo-Saxon market, where it was called Berlin paper.

The origin of the paper

It is not likely that the artist would have been in possession of transfer paper unless he worked as an lithographer or had access to a printing establishment where these papers were in stock. If the artist wanted to reach a special effect, other smooth white papers, with or without coating would have been available, though perhaps not yet in the early 1880s. We must therefore conclude that the choice for the particular transfer paper used was based both on availability and artistic purpose. The state of the paper may indicate that a printing shop has been the source. Transfer paper came usually in sheets in the paper trade. We have indications that the paper has been kept on a roll and cut to (standard) size (see the paragraph on the physical description of the paper). This seems to point to a printing shop as the source of the paper for the artist. Whether as a lithographer or through connections with other users of transfer paper, the artist would also have had access to other papers from the graphic industry. Suppliers of lithographic materials had various kinds of transfer paper in stock, among them very thin papers like China paper and thin western ‘papier pelure’. Uncoated China paper, used for printing (proofs; chine collé), was also sold. In short, if the artist had access to transfer paper, he would also have had access to a wide range of other special papers.

Renoir and transfer paper

After the determination of the type of paper, the dating and tracing of the origin, the final step is establishing whether Renoir has used the investigated paper more often. In comparing the paper to those of authenticated drawings by Renoir, we do not want to limit ourselves to the particular type of paper which may have been chosen only for the occasion because of its role in the background of the pastel. Another way to link the investigated paper to Renoir, is looking for other papers, intended for lithographic use and for other graphic printing purposes, but adopted by Renoir for the purpose of drawing paper.

The period 1879-1884

The gillotage reproductions of the drawings Renoir made in 1879 and 1883-1884 for the journal La Vie Moderne and for l’Impressioniste indicate that he had access to and worked on transfer paper during that period. A signature on one of the reproductions tells us that the firm of Gillot translated Renoirs drawings in relief and therefore is likely to be the source of the paper for the original drawings. Study of the reproductions in La Vie Moderne shows furthermore that Renoir used papers with at least two different types of surface: a smooth paper without any surface structure and a grained paper. The method of producing the relief block might have been a direct transfer from paper to zinc or an indirect transfer by way of a photographic negative. As some of the original drawings that Renoir made in 1883 are still in existence, photo gillotage was probably used in that period for the transfer process in stead of direct transfer from the paper: in the last case the image is usually destroyed together with the coating. Another clue for photo gillotage is that one of the reproductions is larger than Renoirs original drawing. As far as direct gillotage was adopted Renoir must have drawn on transfer paper, but also for the photo gillotages Renoir did not use the special board, papier procédé, that was developed for this purpose: in some reproductions we see indications of the use of frottage to give the paper an additional structure. This of course presupposes that in these instances the paper must have been fairly thin.

In all cases that the original drawing is no longer in existence, this drawing was most probably made on autographic paper and directly transferred to zinc. Also in case of photo-gillotage transfer paper may have been used; this paper, if uncoloured, would have provided the white surface needed for photographic reproduction. Regrettably we have not yet been able to study the extant drawings directly.

The second period: the 1890s and early twentieth century

In the 1890’s Renoir produced a series of lithographs. He took part in all the major albums published in those years. In at least one case, Le Chapeau Epinglé, it is explicitly stated that Renoir made the original drawing on transfer paper. This seems to have been Renoir’s usual way of working. Rather than drawing directly on stone, he was contented ‘to throw a drawing on transfer paper’ and leave the rest up to the printer.

Other graphic printing papers used by Renoir

The above indicates that Renoir could have acquired the transfer paper of the investigated pastel through his contacts with the printer of La Vie Moderne. In order to test this hypothesis we have started comparing the paper of the investigated pastel with that of authenticated Renoir drawings. Aim is not only to establish whether Renoir has used the particular type of transfer paper on which the investigated pastel has been drawn, more often but also whether he has used other, related papers from the graphic printing industry

Up to this point it can be said that Renoir did use for one of his studies for the Grandes Baigneuses a paper that may originally have been intended for lithographic printing. This became clear after examination of the three studies for the Great Bathers that are now in possession of the Musée d’Orsay but kept in de Departement des Arts Graphiques of the Louvre. Of these studies, a large drawing of the left seated bather (1 m x 0.72 cm) is on a very thin paper. Because the drawing is framed, the possibility of investigating it is limited. However, the thinness of the paper is evident from the fact that the structure of the cloth, to which it has been attached, is visible from the recto. It is also apparent from the wrinkles and folds in the paper due to careless mounting to the cloth. The colour of the paper is brown

This paper can be interpreted as China paper as was used in printing establishments for proofs and for certain kind of transfer papers. It cannot be excluded that the paper of the Bather study is a very thin western paper, papier pelure, but here too we must look at lithographic printing for its original purpose (it might be a kind of be transfer paper or have been intended for the preparation of transfer paper).

Renoir also used a thin paper of the ‘papier bulle’ type for another of his studies for the Great Bathers. This paper is generally described as a page from a sketchbook. The thickness of the paper is on estimation that of the investigated paper. In this case the paper is discoloured except for an area around the borders, that has probably been covered by a frame. Originally the sheet must have had a light yellow tone. There is a general resemblance to the paper of the investigated pastel. We might even suppose that the ‘sketchbook’ paper was also connected to lithographic printing just as the paper used for the Bather study mentioned above, but this can at the moment neither be proven nor refuted.

The paper of the great composition study closest to the investigated pastel (see paragraph on the general description of the pastel) is now brown of the same tone as the aforementioned; the original colour will also have been a more light tone. It could very well be the same paper as the ‘sketchbook’paper, but too little is visible to give more than an impression.

The impressionists and transfer paper

It should be noted that when one takes a broader view and looks at the papers used by Renoirs contemporaries and fellow impressionist it becomes clear that they did use on occasion similar papers as the one we have investigated. Degas made a monotype on ivory paper. Particularly interesting is however the practice of Monet to use papier procédé or scratchboard on a number of occasions. In case crayon is used, as for Monets “Two men fishing” (in the Fogg Museum) this choice was more logical than for pastel, as the paper was developed for this kind of medium. A number of drawings by other impressionist was also done on coated paper possibly with a view to have a reproduction made of their paintings for catalogues and the like. As a type of paper, scratchboard is of course closely related to transfer paper; in fact we wonder if it could be distinguished on view from transfer paper attached to board as our investigated paper once was.

Conclusions

Our findings have demonstrated convincingly that the paper used for the investigated pastel is transfer paper. The determination of the production date is based partly on the direct evidence of the chemical composition and the colour of the paper, partly on an interpretation of the visual characteristics of the paper and a reconstruction of nineteenth century transfer paper that was produced for the market. Of course the latter arguments have a much larger margin of uncertainty than the former. However taking all the evidence together a production date for the paper before 1900, or even earlier is by far the most likely. If we accept the conclusions on the date of the paper as late nineteenth century, then it must also be regarded as likely that the pastel was made during that period. In the process of transferring the image from the transfer paper to another support, the paper or its coating were usually destroyed. The paper produced in a certain year in principle disappeared in use. At the same time transfer paper has no special features why some of it should be kept, like the fine Whatman papers. Therefore a stock of nineteenth century transfer paper in the twentieth century is unlikely.

We must, on the same premisse, also conclude that the pastel has been made in France. The picture is based on the large pastel now owned by the Musée d’Orsay, in the Louvre since 1947 and before that in possession of French private collectors. No reproduction was available before 1903.

The artist has used a sort of paper that is not really suited to hold the pigment particles of a pastel. The choice of paper seems to be determined by the wish to create a certain effect but also by availability. The artist who made the pastel was either personally active as a lithographer or had contacts with lithographic printers or other firms where transfer paper was in use. This image fits Renoir. We know that he has worked on transfer paper, both for the purpose of gillotage and for lithography. From our limited study of authentical Renoir drawings we also know that for at least one Great Bather study Renoir used an unorthodox paper of a similar origin as the transfer paper of the investigated pastel.

Renoir was not the only impressionist artist to have used papers intended for the production of prints and illustrations. From the perspective of nineteenth century drawing practice the use of transfer paper by Renoir would not be exceptional.

Acknowledgements

Various individuals and institutions have helped in the research. Henk Porck curator for paper history and conservation scientist of the Royal Librart at The Hague, was closely involved in the research. Mr Hallebeek has led the research at the Netherlands Institute for Cultural Heritage, Amsterdam; Ad Stijnman analyzed the drawing. Mrs Calkoen, Proost and Brandt Paper Laboratory did a fibre analyses and allowed us to use her laboratory; R. S. Gerritsen performed the infrared reflectography; Important information, comments and other forms of assistence were provided by the following persons and institutions: Johan de Zoete, Museum Johan Enschedé; Linda Stiber Morenus, Library of Congress; Ellen van der Grijn, paper historian; Frieder Schmidt, Deutsche Bücherei Leipzig; U. de Goede and E. Löffler, Rijksbureau Kunsthistorische Informatie; Graigen Bowen and Miriam Stewart, Fogg Art Museum, Harvard University, Cambridge Mass.; mr te Rijdt, Rijksmuseum Amsterdam; Nico Lingbeek and Bas van Velzen, paper conservators; Clara Waldthausen, conservator photography; Departement des Arts Graphiques, Louvre; the librarians of the Library of the Amsterdam University and the Van Gogh Museum.

APPENDIX I: Technical data

General Research took place at the Netherlands Institute for Cultural Heritage. Amsterdam, and at the Proost & Brandt paper laboratory, Diemen. The paper has been investigated by direct observation, partly aided by enlargement up to 50x and various types of light under different angles (docucentre, Proost & Brandt ), as well as by means of the electron microscope. To investigate the chemical composition different spectrometric techniques have been used. The presence of a number of elements in the paper, that were regarded as relevant, has been established by X-ray fluorescence spectrometry (XRF). This technique, however, does not give information where the various elements are: in the pastel layer, the coating of the paper or the fibre layer. The precise location of a number of elements was then established by SEM-EDX, X-ray spectrometry combined with an electron microscope. The coating appeared to consist of a single layer under the electron microscope. Infrared spectrometry (FTIR) showed the coating to be a mixture of gelatine and starch; the presence of starch was confirmed by means of a spottest using potassium iodide reagens. Liquid chromatography (HPCL) also confirms the presence of gelatine as well as the absence of resin. Finally gas chromatography-mass spectrometry (GC-MS) points to the absence of glycerine. Fibre analysis has been done in an earlier stage by Integrated Paper Services and during the present stage at Proost & Brandt paper laboratory.

A particular point of interest during the investigation has been the possibility of a photographic print under the pastel. On the basis of the chemical composition of paper and pastel this hypothesis must be rejected: a photographic print would show itself by the presence of certain elements in the photographic image, especially silver and chromium (for the carbon process). In fact, there is a complete lack of silver and the presence of just a small amount of chromium compared to a carbon print, tested for reference by means of XRF. No chrome was found in the coating by SEM-EDX which conclusively excludes the presence of a carbon print.

Infrared spectography was used in the search for a photographic or other type of print Investigation by infrared spectrography and on the light table showed no divergence from the pastel drawing.

Physical characteristics of the paper

The paper support of the pastel measures 55,4 x 71 cm. the paper is machine made and is glaze it is fairly thin: 0.24 mm (average of 4 measurements with hand micrometer). Smallest measurement 0.15, highest 0.3. Measurements with the electron microscope on two places give 0.1 mm on average for the fibre layer. the paper is coated on one side. the paper is yellowish, but does not seem to be discoloured; the coating is off white the paper is sensitive to humidity; it has a tendency to curl along the short side. This points to a gelatine coating on one side of the paper

The coating

The coating is very smooth, of an off white colour and somewhat shiny on the edge of the paper. The coating is very thin: 0.01 mm (SEM).

there are creases in coated side of the paper. From research by means of electron microscopy, together with FTIR (infrared spectrometry), HPLC and a spottest on starch it appeared that:
on the layer of paper fibers there is only one closed layer, the coating of the paper on top of which are the more separately distributed pigment particles the coating consists of a mixture of gelatine and starch. SEM-EDX showed the absence of a filler in the coating.
There is no glycerine present (sometimes used to keep transfer paper moist) XRF testresults as far as the possibility of a photographic print under the pigment layer is concerned: silver is completely absent from the paper, chrome is present in a smaller amount than in a tested carbon print (20-30 ppm versus 240 ppm) but in a larger amount than in ordinary drawing and printing paper (5-10 ppm). SEM-EDX proved the absence of chrome as well as iron from the gelatine coating.

Fibre composition

Bleached sulfite for the paper; the board that was once attached to the paper consists of mechanical wood, unbleached sulfite pulp and old paper.

Chemical composition of the paper

From the fibre research it appeared that the paper pulp has been bleached. This is also indicated by the large amount of sulphur in the paper, which may point to an incomplete washing of the pulp to remove cooking and bleaching products. XRF spectrometry showed the presence of the elements silicium and aluminium. These could haven been both in the paper and in the pastel. SEM-EDX proved them to be only present in the fibre layer of the paper. This is a clear indication that china clay has been used as a paper filler.
The paper has no internal sizing. We have a number of clues for this:
tests show no detectable amount of resin there is no potassium in the paper, therefore no alum in view of the quantities of aluminium and silicium in the paper, the aluminium must be completely bound to the silicium. Therefore there can be no aluminium sulfate in the paper. Gypsum (calciumsulphate), chalk (calciumcarbonate) and lithopone are part of the pastel but not of the paper or its coating. The quantitative relationship of chromium in the pastel layer and the fibre layer is 2/3 versus 1/3. We found the same relationship for iron.

APPENDIX II: Renoirs treatment of paper in the studies for the Great Bathers

It has been mentioned in the report that the final composition study for the Great Bathers painting has been cut in two halves by Renoir himself. When the two separate drawings are placed together it is evident that the half with two bathers ( now in the Fogg Museum) has lost a stripe of paper of about 15 cm on the bottom and the half with the lonely splashing bather a similar stripe on the top. One can only conclude that this has been done by Renoir to change the dimensions of the two halves of the drawing. It is not unusual that an artist changed the exact proportions of a drawing by adding small stripes of paper. Obviously the opposite must also have happened: changing a drawing by cutting off smaller or larger stripes and Renoirs treatment of his last study for the Great Bathers is an example of this.

The relevancy for the present study is in the relation between the contour line on the investigated pastel and the way the drawing is cut off in the various authenticated works by Renoir. It is to be noted that on a total of five composition studies by Renoir two have a contour line. Also there is a similarity between the way the contour line in the investigated drawing cuts off the picture and the way this is done by the border of the paper in two other studies: the one in the Wadsworth Athenaeum and the great composition study of the Musée d’Orsay. In the Wadsworth study it is the right hand of the foreground bather, in the d’Orsay study the left hand of the Central bather that are cut off in similar ways. All this seems to indicate identical choices on the final dimensions of the composition. But as far as the study from the Musée d’Orsay is concerned there is a complication.

We have a reproduction of this drawing, published in 1903, in which the drawing is larger: in particular the hand of the central bather is complete. This is confirmed by the dimensions provided in 1903: 1 m 12 x 1 m 67 in stead of 1 m 08 x 1 m 62. The drawing is presently framed and part of the drawing could be hidden by the frame, but this would be very strange. The record in the database at the Departement des Arts Graphiques of the Louvre, where the drawing is kept, gives no indication of what might have happened. Therefore we must at least entertain the possibility that the drawing has been trimmed.

We don't think any collector would have done this to a Renoir drawing and that therefore Renoir himself must have trimmed the drawing when he still owned it. The reproduction of 1903 was made according to the ‘procédé Georges Petit’ a primitive type of heliogravure. This Georges Petit was the famous galery owner, who doubled as publisher of prints and inventor. He was also the one who sold the d’Orsay drawing in 1903 and the painting of the Grandes Baigneuses in 1887. It is therefore very likely that the photographic negative, that was the basis of the reproduction of 1903, was already in the archives of the Galeries George Petit. It could very well date from the 1880s when the drawing was made. After that date Renoir may have reconsidered the composition and its exact proportions, a reconsideration that led to a trimming of the drawing. If this is correct, Renoirs handling of the d’Orsay drawing would be a clear indication of his doubts on the final dimensions of the composition, a doubt that is suggestive of the contour line in the investigated pastel.

APPENDIX III: SOURCES

SOURCES ON TRANSFER PAPERFielding, The art of engraving, 1844J. M. Herman Hammann, Des arts graphiques destinés à multiplier par l’impression, 1857Motteroz, Eassai sur les gravures chimiques en relief, 1871 Idem, 1888Journal L’Imprimerie; consulted the years 1876-Jules Adeline, Lexique des termes d’art, 1884A. M. Villon, Nouveau manuel complet du dessinateur et de l’imprimeur lithographe: traitant de toutes les operations lithographiques sur pierre et sur zinc de l’autographie, des reports, de la gravure, de l’autographie, de la chromolithographie, de la photolithographie, de la phototypie et de tous les procedes connus jusqu’a ce jour pour reproduire les ecritures et les dessins au moyen de la pierre, du zinc etc., 1891E. Duchatel, Traité de lithographie artistique, 1893Jules Adeline, Les arts de la réproduction vulgarisé, 1894A. Valette, Manuel pratique de la lithographie, 1894 [1891?]Journal Bulletin de l’Ímprimerie, 1896 - .......A. Lemercier, La lithographie francaise de 1796 à 1896 et les arts qui s’y rattachent. Manuel pratique s’adressant aux artistes et aux imprimeurs par Alfred Lemercier, artiste et imprimeur lithographe, premier president de la Société des Artistes Lithigraphes Francais, neveu et associé de R. J. Lemercier officier de la Légion d’Honneur et fondateur de la célèbre imprimerie de ce nom etc .... publié sous le patronage de la Maison Lorilleux et C ie , 1896Louis Edgar Andes, Papierspezialitäten: praktische Anleitung zur Herstellung von den Verschiedensten Zwecken dienenden Papierfabrikaten, wie Pergamentpapiere, Abziehpapiere etc etc, 1896Josef Haynié, Der lithographische Umdruck nach dem heutigen Stand dieser Technik, 1900Nouveau Larousse IllustréCarl Kampmann, Die graphischen Künste, 3., vermehrte und verbess. Aufl., 1906 [1898] Alfred Seymour, Practical lithography, 1903.G. Fritz, Handbuch der Lithographie und des Steindrucks, 1902 F. Hesse, Die Chromolithographie mit besonderer berücksichtigung der modernen, auf photographischer Grundlage beruhende Verfahren und der Technik des Aluminiumdrucks, 1906Louis Edgar Andes, The treatment of paper for special purposes: a practical introduction etc etc, 1907Charles Harrap, Transferring, 1912Eduard Valenta, Das Papier: seine Herstellung, Eigenschaften, Verwendung in den graphischen Drucktechniken, Pruefung usw, 2. verm. Und verb. Auflage, 1922Henri J. Rhodes, The art of lithography, 1922 [first impression 1914?]Louis Edgar Andes, The treatment of paper for special purposes: a practical introduction etc etc, 2 nd . ed., rev. and enh., 1923 H. Trompetter, Handleiding voor de chromo-lithograpfie, 1924Lexicon des gesamten Buchwesen’s, herausgegeben von Karl Löffler und Joachim Kirchner, unter mitwirkung von Wilhelm Olbrich, 1937, Band IIIH. Trompetter, Handleiding voor de chromo-lithografie en photo-lithografie, 1939Henri H. Trivick, Autolithography: the technique, 1960Garo Z. Antheasian, The Tamarind book of lithography: art and techniques, 1971.PRINTED SOURCES ON PAPER AND PAPER MAKINGCarl Hoffmann, Traité pratique de la fabrication du papier, 1877George Olmer, Du papier mecanique et de ses apprets dans les diverses impressions, 1882Carl Hoffmann, Praktisches Handbuch der Papier-Fabrikation, 1891-1897Ernst Kirchner, Das Papier. I. Teil. Geschichte der Papierindustrie und Allgemeines über Papier. II. Teil. Rohstofflehre, 1896-1897. III. Teil. Halbstofflehre, 1907.Dr Bruno Possanner von Ehrenthal, Die Papierfabrikation, Leipzig 1913Edwin Sutermeister, Chemistry of pulp and papermaking, 1920Max Schubert, Die Praxis der Papierfabrikation. Dritte vermehrte und verbesserte Auflage von Dr Ing E. H. Ernst Müller, 1922P. Puget, La fabrication du papier, Paris 1923Prof. Dr. B. Possanner von Ehrenthal, Lehrbuch der chemischen Technologie des Papieres, Leipzig 1923Friedrich Müller, Die Papierfabrikation und deren Maschinen. I. Band: Die Roh- und Halbstoffe sowie das Ganzzeug, 1926Carl Hoffmann, Traité pratique de la fabrication du papier. 3ieme partie: collage, charge coloration des pates a papier, 1926R. H. Clapperton and William Henderson, Modern paper-making, London 1929Dr. Ehrich Opfermann & Ernst Hochberger, Technik und Praxis der Papierfabrikation. Band III. Die Bleiche des Zellstoffs, Berlin 1935G. F. Cross and E. J. Bevan, A text-book of paper-making. Fifth edition, 1936Edwin Sutermeister, Chemistry of pulp and paper making, 1941Julius Grant, A laboratory handbook of pulp and paper manufacture, London 1942Julius Grant, Wood pulp and allied products, London 1947Karl Keim, Das Papier, 1956J.A. van Os, Warenkennis en technologie, deel I, 1956

APPENDIX IV

Research reports by Wagner, Corrigan and McCrone Before the present research was undertaken the pastel has been investigated by a number of researchers whose reports are added here.

The investigation of McCrone was done by means of x-ray spectrometry. Small samples of material were taken from the surface of the pastel, mainly in the fringe areas. Two samples were taken from one of the the main figures: one at right foot of the sitting bather (sample 3) and one from her left arm (sample 14).

On the basis of the test results the following pigments were found to be present: chrome yellow, zinc yellow, umber=iron earth, vermillion, ultramarine, Prussian blue, bone black, charcoal black (probably), lithopone, whiting, calcium sulfate and clay. The research at the Netherlands Institute for Cultural Heritage showed lithopone to be unevenly, gypsum (calcium sulfate) and calcium carbonate to be evenly dispersed in the pigment layer of the pastel.

In the fringe area of the pastel some titanium dioxide was found. The tests at the Netherlands Institute for cultural Heritage, performed over the whole body of the pastel, did not confirm the presence of titanium. This confirms the earlier conclusion already drawn from the McCrone research, that the titanium dioxide was added during later restauration and not part of the pigments used by the artist.