Reconsidering Anders Zorn’s Omnibus Paintings

Emma Jansson

Abstract
This article reconsiders the relationship between Anders Zorn’s (1860–1920) Omnibus paintings. Two large-scale versions exist for the Omnibus motif (Omnibus I and Omnibus II), both of which have historically been viewed as official works that were exhibited by Zorn during the years 1892–1893. Drawing upon recent technical analyses of the artist’s oil paintings, together with relevant archival sources, a discussion is put forward with the aim of reconfiguring the earlier Omnibus I (1891) version as an initial sketch (esquisse) to the later Omnibus II (1892).

[1] At the beginning of the 1890s the young Swedish painter Anders Zorn (1860–1920) was continuing to make a name for himself on the Parisian art scene. Having settled in Paris in 1889, after a period of residence in London, the artist was quickly becoming an exclusive and sought-after society portraitist, with commissions including portraits of the French opera singer Jean-Baptiste Faure (1830–1914) and the comic actor Coquelin Cadet (1848–1909). In addition to undertaking portrait commissions, Zorn also devoted his time to female nudes and genre paintings, with the latter concentrating on scenes from the everyday life of the citizens of Paris. The focus of this article is a well-known example by Zorn on this everyday subject matter: the interior of a Parisian omnibus. Recounting his life’s events in his autobiographical writing, the artist makes note of the bustling carriage interior that provided the inspiration for the scene:

The interior of an omnibus in the evening with electric light thrown into the mix. The subject amused me as well. That mixture of people pressing up against each other during the short journey. The worker, who tired after the toils of the day falls asleep against his beautiful, unfamiliar neighbour’s soft shoulder. The kind modiste in the foreground, who receives a slither of the electric light on her face from the middle of the Boulevard, I think is an oft-seen type, representing the diligent, honest and good French working class.¹

Continuing, Zorn recounts part of the process that helped inform his composition:

I had to undertake many a journey with the omnibus from Montmartre before it became clear to me in my mind and later throughout the course of the work. A small glimpse of daylight through a doorway would serve as electric light in the studio.²

[2] Further aspects relating to the artist’s painting process are also documented through the remnant material records relating to the composition. Several studies and sketches exist for the omnibus motif, including preliminary works in pencil and ink on paper (Figs. 1-2), as well as value studies in oil (Figs. 3-5).

1-2 Examples of preparatory pencil/ink sketches by Anders Zorn: (left) Omnibus study, c. 1891–1892, pencil/ink on paper. Zornmuseum, Mora (photo: Emma Jansson); (right) Omnibus study (from sketchbook), c. 1891–1892, pencil on paper, 25 x 16 cm. Nationalmuseum, Stockholm, NMH 133/1939 verso (© Nationalmuseum)

3-5 Examples of studies in oil by Anders Zorn: [top] Omnibus study (oil), c. 1891–1892, oil on canvas. Zornmuseum, Mora, (photo: Emma Jansson); [bottom left] Omnibus study (oil), c. 1891–1892, oil on canvas. Zornmuseum, Mora, Z0 120 (photo: Emma Jansson); [bottom right]Omnibus study (oil), c. 1891–1892, oil on canvas. Swedish Royal Collection, Stockholm (© Kungl. Hovstaterna)

[3] In addition, Zorn is known to have painted two large-scale versions of the composition. The first version, commonly referred to as Omnibus I (Fig. 6), is signed "Zorn 1891" and currently hangs at the Stockholm Nationalmuseum.³

6 Anders Zorn, Omnibus I, 1891, oil on canvas, 99.5 x 66 cm. Nationalmuseum, Stockholm, NM 6810 (© Cecilia Heisser/Nationalmuseum)

7 Anders Zorn, Omnibus II, 1892, oil on canvas, 126 x 88 cm. Isabella Stewart Gardner Museum, Boston, P3e1 (© Isabella Stewart Gardner Museum)

[4] The second painting, Omnibus II (Fig. 7), is signed and dated "1892", and belongs to the Isabella Stewart Gardner Museum in Boston. In previous art historical literature, it has been assumed that Omnibus I was the version exhibited at the Paris Champ de Mars Salon and the Munich Secession in 1892, whilst the later Boston version was the one purchased by Isabella Stewart Gardner at the 1893 Chicago World Fair.⁴

[5] However, despite showing similar subject matter and compositional arrangements, these two paintings are remarkably different when it comes to their facture. The earlier painting, Omnibus I, is characterised by its loose and fluid handling; lending the composition a diffuse quality that can easily be likened to the techniques of modernist painters such as the French Impressionists. Contrastingly, the later painting, Omnibus II, is relatively academic in terms of its execution, with sharper contours and careful blending between the shadows, mid-tones and highlights. In addition, the later version was painted with finer attention to detail – demonstrated by elements such as the strap buckle on the young modiste’s hat box, as well as the blurred scenery of the cityscape outside the window.

[6] Given the stylistic discrepancy between the two works, one might pose the question as to why this is the case. Why did Zorn choose to produce and exhibit two paintings with nearly identical subject matter, yet executed using such vastly different factures? This question becomes all the more pressing if we consider that Zorn’s Omnibus paintings represent one of the few instances where the artist repeated a composition in this manner. Moreover, the facture of Omnibus I appears anomalous within the context of Zorn’s broader oil painting oeuvre, particularly when compared with the artist’s other Salon paintings from 1892 (Figs. 8-9), which are much closer to Omnibus II in terms of their handling.⁵

[7] In an attempt to answer these questions, it is important to consider the possibility of an alternate relationship between Zorn’s Omnibus paintings, whereby the first version could also have served as a preliminary sketch or esquisse to the later work. Indeed, this has been suggested through recent technical analyses and archival research relating to the two paintings.

8 [left] Anders Zorn, Porträtt av Fru Veronica Heiss, 1891, oil on canvas, 120 x 90 cm. Nationalmuseum, Stockholm, NM 2668 (© Cecilia Heisser/Nationalmuseum); 9 [right] Anders Zorn, Porträtt av Direktör Frans R. Heiss, 1891, oil on canvas, 120 x 90 cm. Nationalmuseum, Stockholm, NM 2667 (© Cecilia Heisser/Nationalmuseum)

[8] Let us start with the available textual sources with regards to the paintings’ exhibition histories. The most commonly cited historical source referring to the 1892 Salon is a review written by the French critic George Lafenestre for the Revue des Deux Mondes, where the author describes Zorn’s Omnibus as follows:

[…] son [Zorn’s] étude de plusieurs personnes En Omnibus montre, en cet artiste singulier, un observateur savant de toutes les lueurs, lueurs de physionomies humaines aussi bien que lueurs de soleil, étoffes et vitrages. A côté de M. Zorn, M. Whistler, qui fut considéré autrefois comme un révolutionnaire, est un peintre tout à fait classique.⁶

[9] Based on this excerpt alone, it would appear that the association between Lafenestre’s description and the Omnibus I version is based largely on the comparison between Zorn and the American artist James Abbott McNeill Whistler (1834–1903), who exhibited his Portrait of Lady Meux (Fig. 10) at the Salon that year.⁷ Next to Zorn, Whistler is described as "once considered a revolutionary", but now "a completely classical painter". Thus, the identification of Zorn’s painting as Omnibus I can be viewed in terms of its perceived modernity – that is to say, a challenging of the prevailing classical or academic traditions of the period. As such, it is most likely that the earlier version of Zorn’s Omnibus is the one that has historically been identified as the painting exhibited at the 1892 Salon, since its summary and therefore seemingly impressionistic handling is reminiscent of the modernist facture that is usually associated with this group of painters.⁸

10 James Abbott McNeill Whistler, Portrait of Lady Meux, 1881–1882, oil on canvas, 193.7 × 93 cm. Frick Collection, New York (CC BY: Frick Collection)

[10] However, what is not made clear in Lafenestre’s review, is whether this perceived element of modernity relates to Zorn’s technique per se, or to the subject matter being portrayed. This is an important distinction to make, as the urban motif of the scene can arguably also be viewed within this modernist framework.⁹ Taken in this light, Lafenestre’s commentary on Zorn’s Omnibus can be seen as relating equally to both versions of the composition, as opposed to just the earlier painting. This in turn raises the question whether the 1892 Salon painting could in fact have been Omnibus II. Indeed, if we make the assumption that any claim to modernism within the work relates simply to its subject matter, then there is no reason to favour one identification over the other.

[11] The answer to this question can be found in part through a revaluation of the Salon catalogues and press comments relating to the exhibition. Writing to Stockholms Dagblad on 5 June 1892, the newspaper’s Paris correspondent provided a detailed description of the exhibited painting’s composition:

[The painting] shows one side of an omnibus interior. A young modiste holding a box on top of her knees, a worker in a white shirt, asleep and with his head resting on his shoulder etc. are the passengers. At the very front in the foreground the lower half [Sv. "extremiteterna"] of a man can be seen, whilst his upper half remains out of view.¹⁰

[12] In this description, there are several compositional elements that point to the painting being Omnibus II, such as the worker’s "white shirt", as well as the half-visible male figure in the foreground. Moreover, the identification of the 1892 Salon painting as Omnibus II was further corroborated during this research through the discovery of a contemporary Salon catalogue, which offers a photogravure reproduction of the painting (Fig. 11).¹¹

11 Extract of Gustave Larroumet, The Salon of 1892, One Hundred Plates in Photogravure and Etchings by Goupil & Co., Bacon, H. (trans.), Paris & New York, Boussod, Valadon & Co., 1892, page 59 (photograph provided by the J. P. Getty Museum, Los Angeles)

[13] Similarly, press reviews of the 1892 Munich Secession confirm that even here it was the second version of Omnibus that was exhibited. In the following section, taken from Berliner Tageblatt (23 September 1892), Zorn’s composition is described as follows:

Anders Zorn in Paris, der geniale schwedische Maler, der sich die Welt so zu sagen im Sturm erobert hat, ist diesmal durch ein Bild vertreten, das die Insassen eines Omnibus zeigt, Herren, auch eine Dame und vorn in der linken Ecke einen einzelnen Arm, zu dem die Figur fehlt.¹²

Once again, specific mention is made of the partially obscured male figure in the foreground. Based on this information, it seems likely that it was Omnibus II (i.e. not Omnibus I) that was the painting presented at the official exhibitions in Paris and Munich during that year.

[14] In addition, a more extensive review of Swedish press commentary from the 1890s offers no mention of Omnibus I in relation to any of the other official exhibition contexts during this period. In fact, the first press mention of the earlier version does not appear until after the turn of the century, in a review of Zorn’s 1906 retrospective exhibition, held in Stockholm between September and October. Writing for Dagens Nyheter, the Swedish art historian and critic Georg Nordensvan said the following about the painting:

Interesting through its boldness in handling, as well as its assuredness in execution is the large study to the omnibus interior, painted in Paris 1902 [Note: Nordensvan appears to have mistaken the original date here, which should read 1892]. The effect of the electric light from the boulevard is quickly captured and reproduced with a secure hand. The motif is known through a later etching – the painting itself has ended up in America.¹³

[15] These comments, in connection with the observation that Omnibus I does not seem to have been exhibited around the time it was first painted, suggest that the work originally functioned as a study (étude) or sketch (esquisse) to the final version. Such an assumption would further explain the summary, wet-in-wet facture that the artist adopted for Omnibus I, which, as noted, presents itself as considerably more impressionistic than his other works from this period. Moreover, the practice of using oil sketches as a preparatory stage within his compositional planning is well documented in Zorn’s case, evidenced by the remnant sketches and studies that exist for the Omnibus motif.

[16] The fact that Omnibus I was signed by the artist does not necessarily disqualify its status as an original sketch, as Zorn’s practice of signing even his preparatory works has also been demonstrated in other instances. A notable example of this is the painting entitled Spetssöm (Fig. 12), which was documented by Gerda Boëthius (Zorn’s first biographer) as being a study for the large-scale painting Spetsknypplerskor (Fig. 13). Commenting on the study, which was painted in Venice in 1894, Boëthius notes how "he [Zorn] made a captivating studio study of two lacemakers, which is more immediate and painterly than the larger work, which was painted under less favourable circumstances and was later reworked in the studio in Paris."¹⁴

12 Anders Zorn, Spetssöm (study for Spetsknypplerskor), c. 1894, oil on canvas, 42 x 48 cm. Zornmuseum, Mora, Z0 111 (© Zornmuseum)

13 Anders Zorn, Spetsknypplerskor, 1894, oil on canvas, 92 x 64.5 cm. Private collection (CC BY: Wikimedia Commons)

[17] Here, Boëthius touches upon a quality or trait that can be perceived as common to all of Zorn’s preparatory works in oil; namely their loose facture, as well as its associated impressions of immediacy and spontaneity in handling. The recent technical analysis of a large group of Zorn’s oil paintings (dating from c. 1880–1920) has confirmed this aspect of the artist’s practice, whereby his sketches and studies are often considerably less finished than their corresponding final versions, whilst also being smaller in format.¹⁵ This is in turn understandable, since the majority of Zorn’s preparatory works would have served as either compositional sketches or value studies. As such, they were not intended for public view or sale, which accounts for the relatively unfinished nature of their facture and also explains why several of these smaller paintings remained unsigned.

[18] This discrepancy in handling has already been noted for Omnibus I and Omnibus II and can be particularly demonstrated in detail by comparing the modelling of the female figure’s face in both paintings (Figs. 14-15). In doing so, the quick and deft application of the brushwork in Omnibus I becomes apparent. Applied using mostly single, unblended brushstrokes, the facture in this earlier version conveys a sense of rapidity in handling, which in turn places the painting closer to the other remnant sketches for this composition. Contrastingly, the modelling used for the face in Omnibus II is more carefully constructed. Examples include the gradual transition between shadows and highlights seen in the figure’s forehead, as well as the shape of her proper right ear.

14 [left] Anders Zorn, detail of Omnibus I (© Cecilia Heisser/Nationalmuseum); 15 [right] Anders Zorn, detail of Omnibus II (© Isabella Stewart Gardner Museum)

[19] In addition to these clear differences in facture, there are other technical aspects that further indicate that Omnibus I most likely served as a sketch to the final version. Firstly, the earlier painting is notably smaller in format, measuring 99.5 x 66 cm, whereas Omnibus II is 126 x 88 cm. Furthermore, an x-radiograph (Fig. 16) made of Omnibus II shows how Zorn made alterations to an earlier version of the painting, whereby he changed the shape of the hat box from being square to rounded in shape. This in turn suggests that the artist initially conceived the composition as being more similar to the arrangement seen in Omnibus I, thus creating a clear chronology for the two works. The reasons for this change in Omnibus II remains unclear, although it is possible that the size of the hat box was reduced in order to make room for the half-visible gentleman in the foreground, since it appears that the placement of the male figure’s hand was shifted slightly upwards, as was the head of the main female figure.

[20] Another change that is visible in the x-radiograph is the placement of the incident light falling upon the cheek of the female figure’s face, whereby its previous position appears to have been more similar to the backlighting seen in Omnibus I. In the final version of Omnibus II, however, the incident light falling from behind the figure is subdued, favouring the sharp angular lighting striking her face from the front. This particular change is also visible on the surface of the painting due to some slight abrasion in this area, as well as some potential increased transparency of the upper paint layers as they have aged over time. It is also curious to note that the later angular-shaped lighting is not as prominent in the x-radiograph as the earlier backlit highlight, perhaps indicating a change in Zorn’s pigment composition between these two areas. However, further elemental analysis, such as non-invasive x-ray fluorescence or cross-section sampling, would be required in order to confirm this observation.¹⁶

16 Anders Zorn, Omnibus II, X-radiograph (© Isabella Stewart Gardner Museum).

[21] These technical observations, combined with the aforementioned archival sources, present a convincing argument for reconfiguring Omnibus I as a sketch to the final Salon version of the composition. However, before concluding this discussion it is worth mentioning a handful of subsequent press reviews relating to the earlier version, as it helps account for its later reception and eventual association as a finished Salon tableau. In a review of Tor Hedberg’s 1910 monograph on Zorn, Nordensvan offers the following reference to Omnibus I:

One is, amongst other things, reminded of one or another less well-known painting, for example the first version of the Parisian omnibus painting and of Coquelin Cadet’s portrait, which Mr. Thorsten Laurin newly had the luck of purchasing at auction in Paris.¹⁷

[22] Then, four years later, on the occasion of the Baltic exhibition in Malmö, an author signed "K.B." mentions Zorn’s inclusion of an "omnibus painting" in the artist’s section of the show.¹⁸ In the former case, it is clear that the text is referring to Omnibus I, since it is recorded as being in the ownership of Torsten Laurin, to whom the painting belonged prior to its entering the Nationalmuseum collection in 1985.¹⁹ In the later review it is a little more ambiguous as to which version of the painting is being referred to, as there is no mention of a specific owner. However, in the exhibition catalogue there are a few clues that point to the painting being Omnibus I, such as its date being listed as "1891" (Omnibus II is dated 1892) and the owner as "Director Th. Laurin".²⁰ Therefore, one can safely assume that the two different reviews are referring to the same painting.

[23] Interestingly, in both cases there is no mention of the painting’s original function as a study or sketch to the later Salon version. Since Omnibus II would have been out of circulation with regards to the European exhibition context for a period of over twenty years by this date, it is easy to envisage how the earlier version might have been mistaken for the later one. Another interesting observation to make, is that in these later, retrospective exhibition contexts, it would appear that Zorn was, in effect, appropriating Omnibus I as a finished painting – this being despite the fact that the work was never officially exhibited around the time it was painted. Indeed, a retrospective appropriation of this kind is alluded to in Ernst Malmberg’s overview of the Larsson-Liljefors-Zorn exhibition, held at Liljewalchs Konsthall in Stockholm, in 1916. In discussing Omnibus I, Malmberg makes the following comments about the painting: "The exhibited canvas, which belongs to Thorsten Laurin, is a prélude – the artist did not want to call it a sketch, as they are two independent works – to the larger canvas with the same subject, which is owned by Mrs. Jack Gardner in Boston."²¹

[24] Zorn’s use of the term "prélude", which bears with it notable musical connotations, is unsual, if not unprecedented within the context of the artist’s œuvre.²² This in turn makes it difficult to interpret precisely what he meant by this term. Whilst it is stated that by this date he considered the two works as separate, the description of Omnibus I as a prélude nevertheless suggests that it was somehow an introductory piece, and by extension of a different hierarchical order to the final Salon version of the painting. That is, although Zorn may have determined in retrospect that the earlier painting should be considered as an autonomous work (i.e., not a preliminary sketch or esquisse), it is also clear that he did not accord it the same status as a finished tableau, as was the case with Omnibus II. The fact that Zorn regarded the later painting as the official version of the composition is further suggested through the aforementioned etching that exists for the motif (Fig. 17), which follows the format of Omnibus II rather than Omnibus I.

17 Anders Zorn, Omnibus (after Omnibus II), 1892, etching on paper, 27.9 x 19.7 cm. Nationalmuseum, Stockholm, NMG 35/1894 (© Cecilia Heisser/Nationalmuseum)

[25] Without further comments from the artist himself it is difficult to determine the exact nature of the relationship between the two works. Moreover, as has been made apparent through this discussion, it would seem that the status of Omnibus I shifted over time in-line with its changing exhibition context – moving from its original function as a preliminary sketch towards its retroactive appropriation as a prélude or introductory work to the official Salon version. In relation to this, one might ask why this later recognition of a previous sketch takes place only towards the latter half of Zorn’s near forty-year long career. A possible explanation is that by the beginning of the twentieth century there was a growing acceptance amongst Swedish critics towards the type of loose, impressionstic facture that characterises Omnibus I – an observation that can be deduced through reviewing the shifting attitudes of Swedish press commentators between the years c. 1880–1920.²³ In addition, the works of modernist painters such as the Impressionists were also becoming increasingly accepted in an official institutional sense by this date, demonstrated by the purchase of several of these works by the Stockholm Nationalmuseum in 1913.²⁴

[26] Regardless of Zorn’s motivations, the reconsideration of the painting’s original function, presented here, allows us to more fully understand the context of a work that has hitherto appeared as deviant within the context of his œuvre. For, although a work such as Omnibus I may not seem conspicuous when placed next to an Impressionist painting by Claude Monet or Auguste Renoir, when considered within the context of the Paris Salon – even the post-1890 Secessionist Champ de Mars Salon – it does not quite fit the style or aesthetic standards that continued to be upheld by these official exhibition spaces during the fin-de-siècle.

[27] As has been noted by Robert Jensen, these official exhibition contexts did not promote the same kind of modern aesthetic that was being represented by the Indépendants around this period. Instead, annual shows such as the Paris Salon, the Venice Biennale, as well as the Berlin and Munich Secessions, were more often dominated by artists such as John Singer Sargent, James Abbott McNeill Whistler, Jules Bastien-Lepage, Max Liebermann, Joaquín Sorolla and Giovanni Boldini; a group that might loosely be termed the juste milieu ("middle way").²⁵ Zorn is also counted within this loose affiliation of artists, a group which, although not connected in the sense of a school or movement, can nevertheless be characterised as a collective through their shared pursuit of official honors – the kind offered in the form of Salon medals, for example.

[28] The styles and techniques adopted by these juste milieu painters in turn differ slightly from the loose and sketch-like facture employed by artists such as the Impressionist, albeit subtly. The Swedish art historian Margareta Gynning has reflected on this distinction within a Scandinavian context, describing the Nordic juste milieu painters as "those artists who adhered to some of their academic foundations, whilst also adopting several features from the avant garde" – a characterisation that can also be extended to the abovementioned broader list of artists.²⁶ This eclecticism in handling, which Jensen refers to as "aesthetic pluralism", is something that has been relatively understudied from a technical perspective within art history, although recent studies on Sargent’s, Boldini’s and Sorolla’s painting techniques are beginning to change this.²⁷

[29] It is in turn this hybridized form of juste milieu modernism that provides us with the most useful framework for contextualising Zorn’s painting technique and practice during this period. In the case of Omnibus I and Omnibus II, it is easy to see how the relatively academic execution of the latter would have been more acceptable within the official Salon context, whilst still retaining certain elements of modernist facture, albeit in their appropriated juste milieu format. The present article has aimed to highlight some of these differences in facture, whilst also consulting relevant archival sources in order to further shed light on this discrepancy. The result has been the discovery of a new relationship between Zorn’s Omnibus paintings, which has also provided a useful case study for re-examining the artist’s placement within the broader modernist canon. Speaking more generally, the shifting functions of a painting such as Omnibus I encourage us to reconsider the ways in which works travel and move through history, with their classification being subject to both contemporary and later interpretations.

Reviewers
Frouke van Dijke, Kunstmuseum Den Haag
Jørgen Wadum, Statens Museum for Kunst, Copenhagen / Faculty of Humanities, University of Amsterdam

Local Editor
Rieke van Leeuwen, RKD – Nederlands Instituut voor Kunstgeschiedenis, Den Haag

License
The text of this article is provided under the terms of the Creative Commons License CC-BY-NC-ND 4.0