The Roman Campagna, with the Claudian Aqueduct

  • Share this page

Infrared examination has revealed that this painting has an underdrawing. Read on to find out where it is.

Technical notes

Oil on paper, laid down on canvas, 22.3 -22.8 x 34.0 cm (visible surface). Signed lower left: COROT. (fig.1)

Corot, 'The Roman Campagna, with the Claudian Aqueduct', probably 1826
fig.1 Jean-Baptiste-Camille Corot, 'The Roman Campagna, with the Claudian Aqueduct', probably 1826



The paper weave is prominent and affects the paint structure in the left part of the sky, where ridges are visible. The paper is mounted onto canvas, which is stretched on a rigid strainer, suggesting that the mounting was done soon after the picture was made. Robaut notes that almost all of Corot's early Italian views were painted on paper and were subsequently mounted: '…they were badly stuck onto poor quality canvases fitted with poor stretchers or executed on sorry panels, and what is more by lending them to hundreds of colleagues and friends, the master let them come and go all his life, the result being that many of them had suffered very badly.'1  It is quite possible that NG 3285 was one of those lent by Corot.

Materials and technique2

There are some traces of rudimentary drawing, possibly done with crayon, especially in the architecture and where the hills meet the skyline, visible in infra-red (fig. 2). There is no drawing in the foreground. The ground is lead white; above this the whole has been painted in one single layer of paint, swiftly and confidently brushed. Working with a small range of colours, greens, beiges and purple/violet and blues for the land, and blues, purples and whites for the sky, Corot has perfectly captured an extensive sunlit landscape under a blue sky studded with bright clouds.

Infrared photograph of Corot, 'The Roman Campagna, with the Claudian Aqueduct', probably 1826
fig. 2 Infrared photograph of Jean-Baptiste-Camille Corot, 'The Roman Campagna, with the Claudian Aqueduct', probably 1826


The foreground is a light grey-beige, almost white in appearance. Behind there is a strip of mid-green with a row of small trees or bushes. The country then extends towards the aqueduct and tower. The hills in the distance are rendered in violet and blue. Corot has left reserves for elements in the composition, such as the green bush on the right. The hills are brought around the aqueducts on left and right. In the sky the clouds were painted first, with the sky brought around them. 

The pigments used are conventional for the period, including vermilion, cobalt blue in the sky and lead white in the clouds. However, there is one notable exception, viridian, which is present, for the most part mixed with earth colours, in the green in the foreground, in the mid-green strip on the left-hand side, in the dull green foliage of the right-hand side, and in the black of the bush on the right-hand side.

Viridian is the English name for hydrated (that is, transparent) chromium oxide, vert éméraude in French, thought until comparatively recently to have been available to artists only from the 1830s. Chromium oxide, an opaque green, was discovered by the French chemist Louis Nicolas Vauquelin in 1797 and used in porcelain factories. In 1798, however, Vauquelin also described a green colour produced by heating the newly discovered chromium metal with a blowpipe and borax, which may have been chemically similar to hydrated chromium oxide pigment.3

Antoine-Claude Pannetier (1772–1859), an artist-chemist and friend and pupil of Girodet, was the first to manufacture the pigment, keeping the process secret. Church gives the date of 1838 for Pannetier’s discovery, which has been repeated by every author on the subject since.4 Charles-Ernest Guignet patented a method for producing the pigment in 1859.5

However, Louis-Charles Arsenne refers to viridian in his Manuel du Peintre et du Sculpteur of 1833, stating that it was available from the colourmen Colcomb-Bourgeois.6  He writes: 'This last colour marries perfectly well with browns and gives the greens a very intense tone. When it is known it will be of great use to the landscapist.'7 It was extremely expensive at this time but price does not seem to have been a disincentive to Corot, as by 1821 he was in receipt of a yearly income of 1,500 livres.8

Corot was already a customer at Colcomb-Bourgeois when he first went to Italy in 1825, and on 29 October 1826 wrote to his friend Abel Osmond for colours from the firm.9 The evidence from this painting suggests that Pannetier had placed his pigment with these colourmen as early as the 1820s, and that Arsenne subsequently wrote about it in his manual of 1833.

Related to this early use of viridian is the further question of whether Corot preferred to mix his greens rather than use ready-made ones, and in much of the recent technical literature on Corot it has been asserted that he preferred to use a mixture of blue and yellow. Robaut is again the source for this theory: 'When he saw the master on the 22nd of September 73 he observed once again that he didn’t use a single green or viridian; and when I asked him the reason he told me that these ready-made greens irritated him, that they are harsh, that he preferred making them in his own manner, sienna earth and burnt sienna with cobalt or yellow ochre with Prussian blue or mineral blue, yellow cadmium with ochre lake and all the tones in-between without thinking about it or formula decided upon in advance.'10

However, not only did Corot use ready-made greens he used them extremely early in his career. Viridian has also been found in a painting from Corot’s second trip to Italy, View of Riva of 1835 (Munich, Neue Pinakothek), where he has mixed it with cobalt blue.11 The crucial point is that in both the Munich painting and NG 3285 Corot has tempered the colour to mitigate its harsh tone.12

Analysis of the binding media has revealed heat-bodied linseed, poppyseed and walnut oils with the minor addition of pine resin, mastic resin (in the green impasto of the bank) and pinaceae resin, possibly fir balsam. The resin would have been incorporated into the paint in the form of varnish, which Corot presumably added on his palette as he worked.

Conservation and condition

NG 3285 was cleaned by the Gallery in 1965.The condition is good. The top right corner of the paper is folded and a small section is lost and there is a small diagonal tear in the upper right edge. What may be pin marks are visible in the two corners on the left side, the corners on the right side are too damaged for such to be found.


1. Bibliothèque Nationale de France, Cabinet des Estampes,  Alfred Robaut, Documents sur Corot,  4oY b 3 949 I p. 36. Quoted in A. Roquebert, ‘Quelques observations sur la technique de Corot’, in Corot, un artiste et son temps. Actes des colloques organisés au Musée du Louvre par le Service Culturel les 1er et 2 mars 1996 à Paris et par l’Académie de France à Rome, Villa Médicis, le 9 mars 1996 à Rome, Paris and Rome 1998, pp. 73–97 (79).


2. For a full discussion of materials and technique see article by S. Herring, ‘Six Paintings by Corot in the National Gallery: Methods, Materials and Sources’, National Gallery Technical Bulletin, 30, 2009.

3. L.N. Vauquelin, ‘Second Memoir on the Metal Contained in the Red Lead of Siberia’, Philosophical Magazine, I, 1798, pp. 361–7, and ‘Mémoire sur la meilleure méthode pour decomposer le chromate de fer, obtenir l’oxydé de chrome, préparer l’acide chromique, et sur quelques combinaisons de ce dernier’, Annales de Chimie, 1st ser. 70, 1809, pp. 70–94.  See R. Newman in E.W. Fitzhugh, ed., Artists’ Pigments. A Handbook of Their History and Characteristics, Washington DC and Oxford 1997, pp. 273–93. Newman does not go into the question of the date of its introduction.


4.A.H. Church, The Chemistry of Paints and Painting, London, 3rd edn, 1901, pp. 194–5.  See also R.J. Gettens and G.L. Stout, Painting Materials. A Short Encyclopedia, New York 1942, pp. 173–4.


5. Gettens and Stout 1942, pp. 173–4. In England the colour maker George Field was also making the colour, and his journal entry for 1815 gives a recipe for chromium oxide. Most assume that the green supplied by Field from 1815 was of the opaque variety, but this cannot be certain, as he did not make any chemical distinction between the opaque and transparent green.  In his Chromatography, an account of pigments available at the time, he talks of a true chrome green, ‘of various degrees of transparency or opacity’. G. Field, Chromatography. A Treatise on Colours and Pigments, London 1835, p. 129.


6. The reference in Arsenne is noted  in D. Bomford, J. Kirby, J. Leighton and A. Roy, Art in the Making. Impressionism, exh. cat., London 1990–1, p. 219, and A. Burmester and C. Denk, ‘Blau, Gelb, Grün und die Landschaftsmalerei von Barbizon’, Barbizon. Malerei der Natur – Natur der Malerei, eds A. Burmester, C. Heilmann and M.F. Zimmermann, Munich 1999, pp. 295–329 (298).


7. L.-C. Arsenne,  Manuel du Peintre et du Sculpteur, 2 vols, Paris 1833, II, p. 248.  However,  J.-F.-L. Mérimée, De la peinture à huile, Paris 1830, p. 190, only mentions the opaque variety of the green, not true viridian, which suggests that he was unaware of it.


8. Moreau-Nélaton in Robaut 1905, I,  p. 24. See Burmester and Denk 1999, p. 307.


9. Moreau-Nélaton in Robaut 1905, I, pp. 38ff. See Burmester and Denk 1999, p. 298.


10. Alfred Robaut, Documents sur Corot, Bibliothèque Nationale de France, Cabinet des Estampes,  Yb3 949, II, p. 15. Quoted in Roquebert 1998, p. 90. While acknowledging that on occasion Corot did make use of certain greens, for example vert Veronèse and vert éméraude, Roquebert repeats this in her essay ‘La technique de Corot’ in Madrid/Ferrara 2005–6, pp. 57–71, French translation, pp. 343–7, Italian edn, pp. 59–71, esp. pp. 66–7. In both essays she cites the findings of  J.- P. Rioux that Corot hardly ever used ready-made greens, backed up by R. Woudhuysen-Keller, J. Cuttle and Ch. Hurst, ‘Zwei Gemälde von Camille Corot. Gedanken zur Maltechnik und  zur Restaurierung’, Restauro, 1993, no. 3, pp. 306–15.


11. See Burmester and Denk 1999, p. 302.

12. Two later paintings in the collection where viridian has been found are NG 6439, Peasants under the Trees at Dawn,  and NG 6466, The Oak in the Valley, both of which are discussed below. At the same time Corot also created greens using a mixture of Prussian blue and chrome yellow, as in NG 2625, The Leaning Tree Trunk.


13. Corot returned to France via Venice. He was again in Italy in May – October 1834, and May – August 1843. 

14. Baehr made two further trips to Rome, in 1827–8 and 1834–5. There exists from the last trip a painting, Campagna Landscape, which also shows the aqueducts (Dresden, Gemäldegalerie Neue Meister, no. 2698).

15. See Jullien 1987, pp. 112–13, for a discussion of the view.  

16. See T. Ashby, The Aqueducts of Ancient Rome, ed. I. A. Richmond, Oxford 1935, p. 232.


17. In this view he is situated four kilometres from the Porta San Giovanni.


18. In this latter they are viewed from next to the tomb of Cecilia Metella. See P. Galassi, Corot in Italy, 1825–8, PhD thesis, Columbia University, 1986, p. 419, where he links R96 with R101 and R180, stating that two of them are copies of the third, but as he has not been to able to study them at first hand, is not able to say which is the original. 


19. Christie’s, London, 2 July 2008, lot 175. It is to be included in the sixth supplement to Robaut, which is in preparation.

20. Galassi 1991, p. 157.

21. Galassi 1991, pp. 170–2. 

22.  It does not have, as Jullien suggested, a golden evening light. 


23. The compiler is grateful to Claire Bechu, Centre Historique des Archives Nationales, Paris, for this suggestion. Martin Davies identified Saint-Barbe as the Collège Sainte-Barbe in Paris, but was not able to find mention of Panis in either J. Quicherat, Histoire de Sainte-Barbe, Collège etc, Paris 1860/4, or in the Almanach Royal for that period. He is also not mentioned in E. Nouvel, Le Collège Sainte-Barbe: la vie d’un collège parisien de Charles VII à nos jours, Paris 1948. There is no mention of Panis among the college’s archives in the Sorbonne, which in fact date mostly from before the Revolution. The compiler is grateful to Jacqueline Artier, conservateur du service du livre ancien, Bibliothèque de la Sorbonne, for checking this. 


24. R3103, carnet 66, vers 1825.


25. Collection de M. Le Comte Armand Doria. Tome Premier, tableaux modernes. Préface et Catalogue analytique par M.L. Roger-Milès. Précédés d’un essai sur la Vie du Comte Armand Doria par M. Arsène Alexandre, Paris, Imprimerie Georges Petit, 1899. If indeed Corot simply lent the painting to Panis, then it would be plausible to suggest that Count Armand Doria, who was an important collector of Barbizon and Impressionist pictures and who invited artists to stay with him, bought this painting directly from Corot. Degas paid 1,150 francs for this painting. Information from L’Album Souvenir of the collection, which appeared a few months after the sale, conveyed by Comte Arnauld Doria in a letter of 20 April 1956 in the dossier. The National Gallery Library also holds an Album Souvenir, which records the price as 1,820 francs.


26. NG 3285 was one of seven paintings by Corot in Degas’s collection, six of which were landscapes, the seventh being a figure study of a woman. Although Degas was not alone in collecting Corots at this period, it was unusual to concentrate, as he did, on the artist’s early work. It has been noted that Degas might have been attracted to Corot’s Italian scenes because of their reminder of Degas’s own time in Italy in the 1850s, from where he wrote: ‘A feeling of antiquity survives in the countryside, that is wild, empty, cursed like the desert, with its great mountains carrying aqueducts and its herds of cattle spread far and wide. This is really beautiful, with the kind of beauty that is like a dream of antiquity.’ Quoted in A. Dumas, Degas as a Collector, exh. cat., National Gallery, London 1996, p. 36. It is probable that it was one of the `two small Views of Italy by Corot’ which hung in Degas’s bedroom. See P. Lafond, Degas, Paris 1918, pp. 177–21. Holmes writes that it was ‘the vivid little early landscape by Corot, which Degas kept hanging over his bed’.  C.J. Holmes, Self and Partners (Mostly Self), London 1936, p. 338.

For the Degas collection and sales see  London 1996, p. 36; New York 1997–8, Holmes 1936, pp. 336–42; D. Sutton, ‘The Degas Sales and England’, Burlington Magazine, April 1989, pp. 266–72; Anne Emberton, ‘Keynes and the Degas Sale’, History Today, January 1996, pp. 22–8. NG 3285 does not seem to have been considered the most important work by Corot in the collection, as The Limay Bridge was singled out by more commentators on the sale. However, Bouyer gives it this praise: ‘As for the Ruins of the Claudian Aqueduct, its pink patches in the harsh Roman countryside rendered in unusual greens, it is one of those loyal nature studies that justify Poussin’s bluish backgrounds and vast cloudy skies.’ See R. Bouyer, ‘Mouvement des arts: la Collection Edgar Degas’, La Chronique des Arts, March 1918, pp. 85–7, reprinted in Dumas, Tinterow et al., 1997–8, pp. 310–12.