After treatment, Daubigny’s painting technique could be studied in detail. The Munich Sluice Gate at Optevoz was
painted on a dark grey layer that covered the earlier composi-tion (Fig. 6). In some places this was left visible at the surface and incorporated into the composition (Fig. 12). Daubigny worked mainly with opaque paints on top of this grey layer, commencing with dark colours then moving to lighter ones.
In different phases of the work process, he also applied trans-lucent dark brown glazes, reminiscent of old master painting techniques. To speed up the drying process, small amounts of oil varnish, including Venice and Cyprian turpentine, were added to the oil paint.36 Daubigny applied the paint with flat Fig. 11 Sluice Gate at Optevoz, Munich: during removal of varnish and overpaint. (Image: Bayerische Staatsgemäldesammlungen Munich.)
Fig. 12 Sluice Gate at Optevoz, Munich: after treatment. Detail of the trees showing branches that were painted with a thin, flat bristle brush. In some areas in the centre of the picture, the dark grey underpaint is visible.
(Image: Bayerische Staatsgemäldesammlungen Munich.)
EVA ORTNER
bristle brushes of different sizes or palette knives depending on the detail that was to be depicted. In the sky, traces of both brushwork and application with a palette knife can be observed; the latter made scratches in the soft paint, which evidently did not concern the artist.
Other parts, for example the embankment and the ripples on the water in the foreground, were executed using a brush only, as were the branches and leaves of the trees and bushes, which are an excellent example of Daubigny’s efficient and lively working process. Typical of Daubigny’s painterly style are the thin, bare branches that were created by slightly turn-ing a thin, flat brush with bristles (Fig. 12). On the other hand, the rocks beside the sluice gate and the waterfall were almost exclusively created with a palette knife (Fig. 13).
From a current perspective, the results of the technolog-ical analysis and restoration leave no room for doubt that the Munich Sluice Gate at Optevoz belongs to the oeuvre of Charles-François Daubigny. The most important indi-cation of the painter is his signature found on the original landscape painting, combined with the knowledge that the depiction of the sluice painted on top lay on top of the same grey preparatory layer and was executed with a virtually identical palette. This is supported by findings that this pic-ture was also originally signed in the same place as the first landscape. In general, it is not possible to attribute a painting to any one specific artist simply based on the materials he used and the way these were applied. Technological aspects, however, can support the art historical attribution to a par-ticular painter. For this reason, the technological findings relevant to the Munich Sluice Gate being classified as a work by Daubigny should be discussed.37 Its original dimensions corresponded to one of the standard formats available on the market, as already described. There are a number of paintings in Daubigny’s oeuvre with standard formats; the version of Sluice Gate from 1859 in the Musée du Louvre, for example, is executed to match ‘format no. 20, marine basse’ for small seascapes.38 The ground, applied by hand, is found in most of Daubigny’s paintings on canvas, including the Munich, Rouen and Karlsruhe versions of Sluice Gate.39 There are also other examples in Daubigny’s work of painted canvases being reused, including the Karlsruhe Sluice Gate.40 The technical
characteristics of the Munich Sluice Gate are a good match to those in Daubigny’s work; these are, however, also to be found to the same extent in paintings by Courbet and other artists of the Barbizon School.
Furthermore, the question had to be answered as to whether the pigments on the Munich picture provided any clues as to Daubigny’s possible authorship. The pigment analyses conducted in 1995 detected cobalt blue within the Karlsruhe Sluice Gate and in five other works by Daubigny.41 No evidence of Prussian blue or synthetic ultramarine was found. On the other hand, no cobalt blue was found in the Munich Sluice Gate, which at that time was considered to have been painted jointly by Courbet and Daubigny, nor was it found in Courbet’s other works examined at that time; instead, Prussian blue and synthetic ultramarine were detected. In short, these findings were ultimately considered decisive factors leading to the conclusion ‘that the Munich Sluice Gate … is the work of two hands but … was painted only using Courbet’s palette’.42 As part of the recent examina-tion, selective pigment analyses were carried out for the first time on the Daubigny Sluice Gate from Rouen.43 This yielded completely different results: in samples from the blue sec-tions in the Rouen painting, Prussian blue was also found in addition to cobalt blue, sometimes mixed together.44 An addi-tional sample taken from the dark blue waves in the flowing water below the waterfall in the Munich Sluice Gate revealed that synthetic ultramarine was also used in certain areas of this canvas. In the meantime, it has been confirmed that this pigment was also used in several other works by Daubigny such as the London River Scene with Ducks.45 Today, the use of cobalt blue, Prussian blue and synthetic ultramarine can be identified in works from the 1850s by both Courbet and Daubigny so these pigments cannot be used to distinguish one artist from another. Despite the similarities in the blues used by both Courbet and Daubigny, the palette of pigments utilised for the Munich Sluice Gate certainly matches that of Daubigny’s creative period in the 1850s.
Finally, the construction of the picture and the painterly style of the Munich Sluice Gate have to be compared with works securely attributed to Daubigny. For this purpose the sluice pictures in Karlsruhe and Rouen were examined due to the similarities between the motifs.46 On a cautionary note, however, it should be mentioned that the Karlsruhe picture is a preliminary study, whereas the painting in Rouen is a fully elaborated version of the subject commissioned by the state.
To start with the differences, none of the paintings were exe-cuted on the same type of preparatory layers. In the case of the Rouen Sluice Gate, the picture was painted directly on a white ground. The Munich painting has a white ground then a dark grey layer that covered the complete surface area to prepare it for the first landscape composition; the same dark grey layer was used to obliterate the earlier landscape before the current composition was painted (Fig. 6). The depiction in the Karlsruhe Sluice Gate was painted directly over the first composition, which had been rejected. 47
The application of paint in the sky section of the Karlsruhe version was executed exclusively using a bristle brush, swiftly and energetically, with a frequent change of direction and Fig. 13 Sluice Gate at Optevoz, Munich: after treatment. Detail of the
waterfall, which was created exclusively with a palette knife. (Image:
Bayerische Staatsgemäldesammlungen Munich.)
FROM COURBET TO DAUBIGNY: THE MYSTERY BEHIND SLUICE GATE AT OPTEVOZ
dabbing in certain areas. Many of the cloud formations were painted with wavy lines or in short W-shaped strokes, char-acteristic of Daubigny’s brushwork. In contrast, the paint in the sky above the sluices in the Munich and Rouen paintings, both very similar to each other in this respect, was applied with both bristle brushes and palette knives. Both paint-ings exhibit scratches caused when the paint was applied with a knife; these are found time and again in Daubigny’s work.48 The same procedure was used when painting both the Karlsruhe and Munich versions, with the sky being executed first and the landscape in a second step, the line of the horizon being corrected later in places when the sky received a new application of paint. In the Rouen version, on the other hand, the landscape was painted before the sky.
There are considerable similarities in the configuration of the trees and bushes in all three paintings: very fine, flat bris-tle brushes were used in the execution of the branches. To vary the width of the lines, the brushes were twisted repeat-edly, but only slightly, while painting. Areas of light colour were subsequently added to the branches in the large clump of trees to suggest openings and light reflections. The areas along the banks above the standing water, painted with strik-ing long brushstrokes, are largely identical as is the water itself with its light reflections and the rocks and stones in it. All three pictures were executed in the relatively ‘dry’ paint, sup-plemented in places with dark brown glazes.
The most striking difference in the painting technique between the Munich picture and the other Sluice Gate ver-sions is the wide use of the palette knife for the ground in the middle distance of the landscape section. However, a varied and experimental application of paint can be observed con-sistently in Daubigny’s work and are perfectly in keeping with the artist’s style. It is possible that the Munich Sluice Gate at Optevoz is an early version in which Daubigny experimented both with the technique and the motif, searching for a defini-tive composition to be used for the work commissioned by the state.49
Conclusions
Through the technological investigation and the restoration of the Munich version of Sluice Gate at Optevoz, greater know-ledge has been gained of Daubigny’s painting technique. An early work by the artist was (re)discovered and can be appreci-ated by researchers and the general public. In 1909, when Hugo von Tschudi acquired Sluice Gate at Optevoz, he had no idea that a beautiful landscape by Daubigny was hidden underneath the landscape allegedly painted by Courbet and, therefore, he unwittingly provided a starting point for a case that turned out to be very mysterious. It is quite remarkable that Tschudi’s
‘Courbet’ landscape led an untroubled existence as part of the artist’s oeuvre for about 100 years. This was unmasked only through an interdisciplinary approach and, it might be added, despite the difficult decision-making connected to the conservation treatment, it is ultimately preferable for the Neue Pinakothek to possess a real Daubigny than a false Courbet.
Acknowledgements
I would like to thank René Boitelle (Van Gogh Museum, Amsterdam), Marcia Steele (The Cleveland Museum of Art, Ohio), Sarah Herring, Gabriella Macaro, Jo Kirby and Larry Keith (The National Gallery, London), Sylvain Amic and Catherine Régnault (Musée des Beaux-Arts de Rouen), Thomas Heidenreich (Staatliche Kunsthalle Karlsruhe), Barbara Schaefer and Iris Schafer (Wallraf-Richartz-Museum, Cologne), Mario-Andreas von Lüttichau and Silke Zeich (Museum Folkwang, Essen), Stephanie Dietz (Frankfurt a. Main).
The conservation treatment of the Munich Sluice Gate was funded by the Fondation BNP Paribas and BNP Paribas Germany.
Notes
1. J.G. Prinz von Hohenzollern and P.-K. Schuster (eds), Manet bis Van Gogh. Hugo von Tschudi und der Kampf um die Moderne, exh. cat., 2nd edn, Munich and New York 1997.
2. Hohenzollern and Schuster 1997 (cited in note 1), pp. 42–59, nos. 1–8.
3. C. Fabre, ‘“L’expertise, c’est vraiment dangereux!” Théodore Duret, Verfechter und Vermittler der Avantgarde’, in Bayerische Staatsgemäldesammlungen (ed.), Courbet > Daubigny. Das Rätsel der ‘Schleuse im Tal von Optevoz’, Munich, 2014, pp.
36–45, esp. p. 45, n. 26.
4. Fabre 2014 (cited in note 3), p. 40.
5. R. Hellebranth, Charles-François Daubigny 1817–1878, Morges, Matute, 1976, nos. 520 (Rouen, Musée des Beaux-Arts), 525 (Paris, Musée du Louvre), 527 (Valence, Musée de Valence), 528 (Houston, Museum of Fine Arts) and 529 (Karlsruhe, Staatliche Kunsthalle); R. Hellebranth and A. Hellebranth, Charles-François Daubigny 1817–1878 (supplement), Morges, Matute, 1996, nos. 108 (private collection), 110 (private collection), 111 (Essen, Museum Folkwang).
6. Charles-François Daubigny, Schleuse im Tal von Optevoz, c.1855, oil on canvas, 48 × 81.5 cm, Karlsruhe, Staatliche Kunsthalle, inv. no. 2581.
7. ‘Neuerwerbungen für die Neue Pinakothek’, 10 December 1909, in Akten über Erwerbungen für die Staatsgemäldesammlungen durch Kauf, 1904–1910, Munich, Bayerische Staatsgemäldesammlungen, archive shelf 23, no. 1.
8. See, for example, L. Dittmann, ‘Courbet und die Theorie des Realismus’, in H. Koopmann and J.A. Schmoll gen. Eisenwerth (eds), Beiträge zur Theorie der Künste im 19. Jahrhundert, vol.
1, Frankfurt am Main, Klostermann, 1971, pp. 215–239, esp.
pp. 234–239; K. Herding, ‘Egalität und Autorität in Courbets Landschaftsmalerei’, Städel-Jahrbuch 5, 1975, pp. 159–199, esp.
pp. 170, 194, n. 126; W. Hofmann and K. Herding, Courbet und Deutschland, exh. cat., Cologne 1978, pp. 219ff., no. 231 (P.-K.
Schuster).
9. E. Ortner, ‘Die “Schleuse im Tal von Optevoz” – Ein Bild entsteht und wird verändert’, in Bayerische Staatsgemäldesammlungen 2014 (cited in note 3), pp. 62–97.
10. A. Burmester and C. Denk, ‘Comment ils inventaient ces verts chatoyants? – Blau, Gelb, Grün und die Landschaftsmalerei von Barbizon’, in A. Burmester, C. Heilmann and M.F.
Zimmermann (eds), Barbizon. Malerei der Natur – Natur der Malerei, Munich, 1999, pp. 295–329; K. Laudenbacher,
‘Beobachtungen zur Maltechnik an ausgewählten Bildern der Schule von Barbizon’, op. cit. pp. 393–402, esp. pp. 398–
399; J. Koller and U. Baumer, ‘Die Bindemittel der Schule von Barbizon’, op. cit. pp. 343–369.
EVA ORTNER
11. C. Heilmann, M. Clarke and J. Sillevis, Corot, Courbet und die Maler von Barbizon, exh. cat., Munich 1996, pp. 144–145, no. B 32; see also A. Pophanken in Hohenzollern and Schuster 1997 (cited in note 1), p. 46, no. 2.
12. H.W. Rott, ‘Courbet, Daubigny und die “Schleuse im Tal von Optevoz”’, in Bayerische Staatsgemäldesammlungen 2014 (cited in note 3), pp. 12–35, esp. pp. 22–23.
13. I would like to thank René Boitelle (Van Gogh Museum, Amsterdam) and Marcia Steele (The Cleveland Museum of Art, Ohio), for kindly making comparable material available.
14. Doerner Institut, Munich, pigment analyses report (ident no.
5131), 2007–2014 (unpublished).
15. Ibid.
16. Linseed oil and larch turpentine were analysed as being original components in both layers of varnish. Doerner Institut, binding media analyses report (ident no. 5131), 2007 (unpublished).
17. Ibid.
18. The reconstruction of the painting’s original dimensions is explained later in the text.
19. B. Haaf, ‘Industriell vorgrundierte Malleinen. Beiträge zur Entwicklungs-, Handels- und Materialgeschichte’, Zeitschrift für Kunsttechnologie und Konservierung 1(2), 1989, pp. 7–71;
P. Labreuche, ‘The industrialisation of artists’ prepared canvas in nineteenth century Paris: canvas and stretchers. Technical developments up to the period of Impressionists’, Zeitschrift für Kunsttechnologie und Konservierung 22(2), 2008, pp. 316–328.
20. Comparisons of the motif with other depictions of the sluice gate at Optevoz by Daubigny support this reconstruction of the original format.
21. Haaf 1989 (cited in note 19); Labreuche 2008 (cited in note 19).
22. During restoration it emerged that the first two layers of varnish had been removed from about 20% of the sky area.
23. In some areas, the overpaint extended right up to the brown paper strips used to mask off edges of the painting that had been added during lining.
24. Other areas of overpaint were found, especially in the filled areas and along the left-hand and bottom edges of the picture.
25. The newspaper fragments were found under the brown paper strips masking the edges and a double-layered paper gauze that was glued on during lining (see note 23).
26. See, for example, É. Rostain, Rentoilage et transposition des tableaux, Puteaux, 1981, pp. 48–52.
27. Rott 2014 (cited in note 12), pp. 28–29.
28. When the painting was first examined it was assumed that the signature revealed in the IRR was directly underneath these layers of overpaint; see Laudenbacher 1999 (cited in note 10), pp. 398–399. It is in fact on the first landscape, the existence of which was not known at that time.
29. A. Riegl, ‘Der moderne Denkmalkultus. Sein Wesen und seine Entstehung’, in A. Riegl, Gesammelte Aufsätze, Augsburg and Vienna, Braumüller, 1929, pp. 144–193.
30. Rott 2014 (cited in note 12), pp. 27–30.
31. Underneath the area around the signature, the previous over-paint on the lower edge of the picture has also been retained in places as have all varnish layers.
32. See also L. Keith and R. White, ‘Mixed media in the work of Charles-François Daubigny: analysis and implications for con-servation’, National Gallery Technical Bulletin 23, 2002, pp.
42–49, esp. p. 42. Isopropanol, ethanol and Shellsol A were used. Glazes, as described in the essay by Keith and White on Daubigny’s work, were not found in the sky section of the Munich Sluice Gate at Optevoz.
33. To remove the upper, readily soluble, varnish layer, a mixture of isopropanol and iso-octane was used in different ratios. To minimise penetrating the original layers the solvent was thick-ened with ethylcellulose to form a gel. The lowermost layer of
varnish that contained linseed oil (cited in note 16) could be made to swell by using a mixture of 4 parts isopropanol + 1 part ethanol + 2 parts iso-octane + 1 part Shellsol A.
34. The best cleaning results were achieved using a ceramic blade.
35. While exposing these sections the magnification setting on the stereomicroscope was continously changed from ×40 for mechanical work and ×60 for an optical inspection of the result.
36. Nut oil, or (a mixture of) linseed and poppyseed oil were analy-sed as binding agents. Doerner Institut, binding media analyses report (cited in note 16); Koller and Baumer 1999 (cited in note 10), pp. 357–369.
37. To date, the painting techniques used by Daubigny in his early works have hardly been researched.
38. See note 5 and Haaf 1989 (cited in note 19); Labreuche 2008 (cited in note 19). The Rouen and Karlsruhe versions of the sluice gate motif, on the other hand, do not have a standard format. Of 13 paintings by Daubigny on canvas in the collection of the Van Gogh Museum, Amsterdam and the Mesdag Collection in the Hague, five works are of a standard size. I would like to thank René Boitelle (Van Gogh Museum, Amsterdam) for this infor-mation.
39. In individual cases, however, Daubigny also used industrially manufactured, pre-grounded supports; see R. Boitelle and C.
van der Elst, ‘A new look at old definitions: examining notions of finish in Barbizon School landscapes’, in Burmester et al.
1999 (cited in note 10), pp. 153–159, esp. p. 156.
40. See, for example, Laudenbacher 1999 (cited in note 9), p. 398;
Boitelle and Van der Elst 1999 (cited in note 39), p. 156.
41. Burmester and Denk 1999 (cited in note 10), pp. 312–315.
42. Ibid., p. 305.
43. The pigments were analysed at the Doerner Institut in 2014.
44. Doerner Institut, pigment analyses report (cited in note 14). In addition, viridian green and emerald green (copper arsenate) were found in the Rouen Sluice Gate at Optevoz. These were new pigments at that time and were not used in the Karlsruhe or Munich versions.
45. Bords de l’Oise près d’Auvers, 1859, The National Gallery, London, inv. no. 2622. Before 1995, synthetic ultramarine had been found in Courbet’s paintings, but Daubigny’s works had not yet been thoroughly researched.
46. For an evaluation of differences in motifs in the various versions of Sluice Gate at Optevoz see Rott 2014 (cited in note 12), pp.
30–33.
47. As the application of paint in the sky area does not completely cover the painting underneath, predominantly black and dark madder shades of paint can be seen in a number of minute ope-nings left by the brushstrokes.
48. Clearly evident scratches caused by a palette knife can be seen, for example, in Daubigny’s painting Le Ru de Valmondois, c.1870/75, The Mesdag Collection, The Hague, inv. no. hwm 98 and Soleil couchant, 1871, The Mesdag Collection, The Hague, inv. no. hwm 95.
49. Rott 2014 (cited in note 12), p. 33.
Author’s address
Eva Ortner, Bayerische Staatsgemäldesammlungen, Doerner Institut, Barer Straße 29 D 80799, Munich (eva.ortner@doernerinstitut.de)