Leonardo da Vinci

A sourced technical report on his oil and oil-tempera paintings on panel. Every claim cites a source key; the sources are listed at the end. "Uncertain" marks claims that are unverified, disputed or taken from a single object. Fewer than twenty paintings survive and each was built differently [GON23 p.76], so a finding from one panel describes that panel. Read in full: a cross-section study of the London Virgin of the Rocks, a technical survey of his early Florentine panels and a study of the Mona Lisa's ground. Read through summaries: a lead-white study of the Saint Anne and the Kraków examinations of the Lady with an Ermine. Read only through press releases and search excerpts: a study of the faces in seven Louvre paintings. His notes were read in English editions of 1883 and 1802, and the 1568 biography in translation.

Summary

His panels are poplar or walnut. Several carry gesso; the Mona Lisa and La Belle Ferronnière have an oil ground laid straight on the wood. From the later 1480s every ground holds lead white in oil, and over gesso he laid one or two thin oil primings, white, grey or cream [GON23 p.76; VOR p.43; EARLY p.13]. He drew with a brush, pounced heads and hands from cartoons, and worked the design up in monochrome washes before colour [EARLY pp.13, 15, 17; VOR p.43]. The paint is walnut or linseed oil, often heat-thickened, sometimes rich in medium [VOR p.48; EARLY pp.19–20]. The palette is small: lead white, lead-tin yellow, earths, vermilion, red lake, azurite, natural ultramarine, verdigris, a translucent brown earth and charcoal black [VOR pp.43–47, p.54]. Faces are laid in pale and modelled with thin, semi-translucent browns blended until no edge shows [EARLY p.20; VOR p.47; VIG]. Work took years and many panels were never finished [VOR p.32; LIVES]; drying cracks are common [EARLY pp.19–20]. Documented pigments are listed in §4.

Terms. Gesso: a ground of gypsum and animal glue. Priming (imprimitura): a thin oil layer over the ground. Underdrawing: the drawing on the prepared panel. Pouncing (spolvero): dusting charcoal through holes pricked along the lines of a cartoon, leaving dotted lines. X-radiograph: an image made with X-rays, which shows lead-rich paint. Cross-section: a microscopic sample cut through all layers. Tempera: paint bound in egg. Heat-bodied oil: oil thickened by heating. Glaze: a transparent layer. Scumble: a thin, semi-opaque layer dragged over another. Sfumato: modelling by blended tones without lines or edges. Lead soaps: compounds formed when lead reacts with the oil. Blanching: a whitish clouding of a paint layer.

1. Supports

• Poplar. The London Virgin of the Rocks is on four poplar boards, 189.5 × 120 cm; its size may have ruled out costlier walnut [VOR pp.32–33, 43]. The Annunciation (100 × 221.5 cm) [EARLY p.14], the Mona Lisa (77 × 53 cm) [GON23 pp.76, 78] and the Saint Anne (168 × 113 cm) [LOUV] are poplar. So, by an encyclopedia, is the Madonna of the Carnation (62 × 48.5 cm) [WIKI-NEL; EARLY p.21].
• Walnut. The unfinished Saint Jerome (103 × 75 cm) is on walnut [EARLY p.12], and so is the Lady with an Ermine (54.8 × 40.3 cm) [KRK; WIKI-LE]. Walnut was usual for Lombard panels of the late fifteenth century [VOR p.43].
• His own list. "The panel should be cypress or pear or service-tree or walnut" [NB §628].
• Who made the panels. Large panels were perhaps ordered from carpenters, who gessoed them; small planks bought bare left him free to experiment with grounds [GON23 p.76]. Hypothesis.
• An early shared panel. The Baptism of Christ (Florence) was painted largely in egg tempera in his teacher's workshop, then completed and retouched by him in oil [EARLY p.6].

2. Grounds

• Each one different. From about 1485–90 each ground differs; all are "oil-based" and contain lead white [GON23 p.76].
• Gesso with oil primings. On the London Virgin of the Rocks a gesso and glue ground was covered by a light grey oil priming of lead white and a little carbon black. A second priming, faintly greyish-cream, of lead white with some lead-tin yellow, went on later, bound in walnut oil that had not been heat-bodied. Palm and finger prints, very probably in this layer, show in the X-radiograph [VOR pp.43, 48]. The Madonna of the Carnation has two layers of gesso (coarse, then fine) under a lead-white priming [EARLY p.13], apparently in linseed oil [EARLY p.31 n.53]. The Saint Anne has gesso under a lead-white priming [GON23 p.76; GON20].
• No gesso. The Mona Lisa has, by its X-radiograph, a single thick lead-white ground with no gesso, soaked into the poplar [GON23 pp.76–77]. Its oil is strongly saponified and holds plumbonacrite, which the authors take as the sign of oil loaded with lead(II) oxide to make a thick, fast-drying paste [GON23 pp.75, 82–83]. Inference. La Belle Ferronnière has an orange oil ground of lead white and red lead laid directly on the wood [GON23 p.76].
• The Lady with an Ermine. A thin lead-white layer [KRK], or "white gesso and a layer of brownish underpaint" [WIKI-LE]. Uncertain: the summaries differ.
• His recipe. Mastic, turpentine and white; "boiled linseed oil"; "liquid varnish and white"; then "pounce and outline your drawing finely and over it lay a priming of 30 parts of verdigris with one of verdigris with two of yellow" [NB §628]. If the first "verdigris" is a slip for lead white, as one examiner reads it, this matches the yellowish priming over the first drawing on the London panel [EARLY p.30 n.48].
• White under transparent colours. "For those colours which you mean should appear beautiful, prepare a ground of pure white. This is meant only for transparent colours" [TREAT ch. CCXXXII].

3. Drawing and lay-in

• Brush drawing. In the Annunciation the drawing was "executed with a brush and a liquid medium"; it is functional, not elegant [EARLY p.15]. He shaded with washes rather than hatching [EARLY p.13].
• Pounced cartoons. Pounce dots on Ginevra de' Benci and the Lady with an Ermine; he made and reused cartoons for single heads and hands [EARLY p.17]. The Lady also has brush corrections [KRK]; the Saint Anne shows pounced transfer [LOUV].
• Two drawings on one panel. On the London Virgin of the Rocks the first design, a different composition, was drawn in a fluid brownish-black paint on the gesso and half-hidden by the grey priming; after the second priming he drew the present design [VOR pp.42–43].
• Monochrome underpaint. The second drawing "merges seamlessly into areas of elaborate wash-like monochrome underpaint" in dark yellow, brown and blackish-brown, of a translucent brown earth, other earths and black, "relatively rich in medium" [VOR p.43]. The unfinished Adoration and Saint Jerome stop at this stage, in washes "better described as grey than brown"; in the Milanese works the brown undermodelling is oil paint [EARLY p.13].
• Tone before colour. He thought "in terms of contour and tone ... before colour" [EARLY p.13], and composed the Adoration directly on the panel [UFF].

4. Pigments

The fullest pigment evidence comes from cross-sections of one painting, the London Virgin of the Rocks [VOR pp.43–48]. Its examiners call the palette "deliberately limited" [VOR p.54].

Pigment	Evidence	Notes
Lead white	Primings, sky, flesh of the Virgin of the Rocks [VOR pp.36, 43, 47]; grounds of the Mona Lisa, Saint Anne, La Belle Ferronnière [GON23 pp.76, 80]; the Lady with an Ermine [KRK]	In the Saint Anne a finer grade in the distant haze than in the priming; he once ordered two lead whites, one 30% dearer [GON20]
Lead-tin yellow	Second priming, the Virgin's yellow lining, leaf edges [VOR pp.43, 45–46]; the Lady with an Ermine [KRK]	"Giallolino"; the 1802 translation [TREAT ch. CCXXX] calls it "Naples yellow" [GON23 pp.84–85]
Yellow ochre (yellow earth)	Lining ("golden ochre"), foliage, the angel's lining [VOR pp.45–46]; "green with yellow ochre" [NB §618]
Brown ochre	Foliage; "very fine brown ochre" in a flesh shadow [VOR pp.45, 47]
Red and orange earths	Warm rocks; an "orange-toned earth" in a flesh shadow [VOR pp.45, 47]; "red chalk or burnt ochre for the dark shadows" [NB §618]
Vermilion	Flesh [VOR p.47]; the Lady with an Ermine [KRK]; "for the lights pure vermilion" [NB §618]
Red lake	The Virgin's underdress and the angel, extended with starch; a different lake in the flesh [VOR pp.46–47, p.55 n.30]; "then glaze with fine lake" [NB §618]	Dyestuff not identified in the sources read
Red lead	The orange ground of La Belle Ferronnière [GON23 p.76]; particles in the yellow lining, perhaps formed in the paint [VOR p.46]
Azurite	Sky underlayer, mantle underpaint, greenest foliage, underdress, angel [VOR pp.36, 39, 45–46]	Greenish beside ultramarine; coarse grinding is darker [AP2 pp.25–26]. None in the Lady with an Ermine [KRK]
Natural ultramarine	Mantle, sky, angel [VOR pp.39, 45–46]; the Madonna of the Carnation's mantle [EARLY p.13]; the Lady with an Ermine [KRK]; a blue layer at the Mona Lisa's edge [GON23 pp.77, 82]	Extracted from lapis lazuli; in oil "a translucent glazing color," very dark when thick [AP2 p.44]
Verdigris	Rocks and foreground foliage [VOR p.45, p.55 n.24]; in his notes for greens [NB §§626–627; TREAT chs. CCXXVIII–CCXXIX]	Copper in some faces' shadow glazes [VIG]
Umber	Shadow modelling in the Lady with an Ermine [KRK]; manganese and iron in the Mona Lisa's shadow glazes [VIG]	Umber for the Mona Lisa is an inference from manganese
Translucent brown earth	Underpaints and rocks, "of the Cassel earth type": iron plus probable organic matter [VOR pp.43, 45, p.55 n.21]	The examiners liken it to what a 1586 treatise calls "terra d'ombra" (umber) [VOR p.55 n.21]
Carbon black, charcoal	Carbon black in the first priming; charcoal black in drapery underpaints, foliage, underdress and a flesh shadow [VOR pp.43, 45–47]	Wood charcoal is "of weaker body than ivory or lamp black" and "of a cool neutral tint" [SAL §290]; ivory black tends "to brown in its pale washes" [SAL §293]
Bitumen	"Take green and mix it with bitumen and you will make the shadows darker" [NB §618]	Not found by analysis; "may sometimes" have been used in underpaint [EARLY p.31 n.52]. Uncertain

• Other materials. Chalk in a grey underpaint and starch in a red lake [VOR p.46]; tin and gold leaf on a raised cross [VOR p.47]; no powdered glass [EARLY p.31 n.54].
• Lead oxide. Grains of lead(II) oxide survive in priming and paint of his Milan wall painting, apparently as a drier [GON23 pp.85–86]; in the Mona Lisa it is inferred, not found (§2).

5. Media and varnish

• Walnut and linseed. The Madonna of the Carnation is bound mainly in walnut oil, "not necessarily heat-bodied" [EARLY p.19; p.31 nn.53–54]. The London panel has both oils, mostly heat-bodied: walnut in the sky, flesh and some rocks, linseed in some underpaints and a late blue, with no pattern by stage [VOR p.48].
• Making the oil. Peel the walnuts, or the skin "tinges the oil" and "rises to the surface of the painting" [NB §631].
• Rich in medium. He "added an unusually high proportion of medium to his paint in order to obtain smoothly blended tonal transitions" [EARLY p.20].

• Stiff paint, no thinner. There is "no real evidence" of volatile thinners used to dilute paint in the fifteenth century. His oil paint in the Baptism is "visibly stiff and sticky" [EARLY p.20].

• Varnish. His notes give varnishes from cypress or juniper resin, one "mix[ed] ... with nut-oil" [NB §§635–636], and say verdigris must be varnished at once [TREAT ch. CCXXVIII]. "He tried the strangest methods in seeking out oils for painting, and varnish for preserving works when painted" [LIVES]. No original varnish was identified in the sources read [VOR p.32].

6. Layers and application

• Faces. He laid in a very pale base containing much lead white, reserving the lips, then modelled the shadows "with delicate semi-translucent browns and pinks." X-radiographs show the heads as masks with openings for eyes and mouth [EARLY pp.20, 22].
• Shadows in flesh. In the Virgin of the Rocks the flesh is two opaque layers of white with very little vermilion, red lake and black. The shadow is "a single thin layer of partially translucent paint" of charcoal black, very fine brown ochre and a little orange earth, thinned and blended so the light layer below shows through by degrees [VOR p.47]. In the Mona Lisa the shadows are glazes of 1–2 µm each, at most 30–40 µm in all, low in pigment, with manganese and iron, thicker where darker; in La Belle Ferronnière they are an opaque dark paint [VIG]. Search excerpts.
• His flesh recipe. "Use black in the shadow, and in the lights white, yellow, green, vermilion and lake. Medium shadows; take the shadow as above and mix it with the flesh tints" [NB §619].
• Blue drapery. Grey underpaint with charcoal, modelled to the final tone; a thick layer of azurite; sometimes a dark modelling layer; then ultramarine, with or without white, thick in the lights and "an extremely thin glaze" in the shadows [VOR p.45].
• Plum and mauve. Charcoal black with red lake over grey; for the angel, ultramarine and azurite scumbled over black and red lake [VOR p.46].
• Yellow drapery. One layer of lead-tin yellow, white and golden ochre over a grey underpaint [VOR p.46].
• Greens. Verdigris, yellow earth and charcoal black, with a scumble of black to darken [VOR p.45]. His recipe darkens green with bitumen, lightens it with yellow ochre and yellow, "for the high lights pure yellow," then glazes all with green and turmeric [NB §618].
• Reds. "Cinnabar or red chalk or burnt ochre for the dark shadows and for the lighter ones red chalk and vermilion and for the lights pure vermilion and then glaze with fine lake" [NB §618].
• White over dark. "White scumbled upon a dark ground appears blue" [TREAT ch. CCXXXIII].
• Fingers. In the Baptism the paint is "worked and blotted with his fingers," the lead white "stringy," dragged across the panel [EARLY p.20]. The Lady with an Ermine has fingerprints in the face [WIKI-LE].
• Dark grounds and blacks. His biographer: "he sought for blacks which might make deeper shadows and be darker than other blacks, that by their means he might make his lights the brighter," until his pictures looked like "an effect of night" [LIVES].

7. Working sequence and reworking

• Order. On the London panel: gesso; drawing; grey priming; cream priming; second drawing; monochrome wash underpaint; then colour area by area (underpaint, body colour, glaze or scumble) [VOR pp.42–47]. The Madonna of the Carnation's blue sits on gesso, lead-white priming, a translucent brown and a thin blackish layer [EARLY p.13].
• Years on one panel. The London panel took from about 1491 to 1508, and he recast its whole design on the way [VOR pp.32, 42]. After four years on the Mona Lisa "he left it unfinished" [LIVES]; the Saint Anne was incomplete at his death [LOUV]; the Adoration and Saint Jerome stop at the monochrome stage [EARLY p.13].
• Uneven finish. The London panel is "notably uneven in its level of completion" [VOR p.41]. Ultramarine at upper right was blocked in casually in the last campaign, after framing; brown paint laid over finished blue folds records contour changes he never resolved [VOR pp.39–40]. Changes were made by painting over, not by scraping (inference; scraping is not documented in the sources read).
• Too little drying time. Drying cracks on the rocks come from "building up several layers with poorly drying pigments, and perhaps inadequate drying times between applications" [VOR p.48].
• Varnish first. "He begins by thinking of the end of the work, before the beginning," said the pope who saw him distilling varnish [LIVES].

8. Condition and aging

• Drying cracks and wrinkles. Wide cracks in the Annunciation's hair; cracks in the blues and wrinkling even in the flesh of the Madonna of the Carnation, from slow-drying underlayers and much medium [EARLY pp.19–20]; cracked, wrinkled passages on the London panel from its medium-rich underpaints [VOR pp.34, 43].
• Blanched ultramarine. The London ultramarine has blanched patchily; where it was a thin glaze the greenish azurite below now shows [VOR p.45, p.55 n.25].
• Lead soaps. Lead soaps have made the second priming of the London panel more translucent than it was [VOR p.43].
• Overpainted background. The background of the Lady with an Ermine, probably bluish-grey, was painted over in black, perhaps in the eighteenth century [WIKI-LE; KRK].
• A spoilt panel. A small Madonna was "much spoilt," perhaps "because of his innumerable and capricious mixtures of grounds and colours" [LIVES].
• Well preserved. Ginevra de' Benci, with "perfectly smooth and evenly blended flesh tones," seems free of drying problems [EARLY p.20].

9. The tubes here

The historical names in this note are mapped to these tubes: lead white (biacca, both grades) → lead white; lead-tin yellow ("giallolino") → lead-tin yellow; yellow ochre, yellow earth, golden ochre → yellow ochre; brown ochre → brown ochre; red earth, orange-toned earth, red chalk, burnt ochre → red earth; vermilion and cinnabar → vermilion; red lead → red lead; red lake ("lake," "fine lake") → rose madder (a proxy: the dyestuff is not identified); azurite → azurite; natural ultramarine → natural ultramarine; verdigris → copper green; umber → raw umber; the translucent brown earth of the Cassel type → bone brown (a proxy for its transparency and slow drying: the tube's own pigment, roasted bone, is not documented for him); carbon black and charcoal black → charcoal black; bitumen → bitumen. Bitumen rests on his own words alone (§4).

10. At this easel

• Support. Plain-weave linen with ground layers laid from the tubes.
• Paint. Pigment from the tubes with an oil-medium share.
• Taking paint off. A brush picks up wet paint as it passes; that is how paint comes off the canvas.
• Everything else the easel does is in the easel guide.

Sources

• [VOR] Keith, L., Roy, A., Morrison, R. and Schade, P., "Leonardo da Vinci's Virgin of the Rocks: Treatment, Technique and Display," National Gallery Technical Bulletin 32 (2011), pp.32–56. Full text read. https://www.nationalgallery.org.uk/media/15698/keith_roy_morrison_schade2011.pdf ; article page: https://www.nationalgallery.org.uk/research/research-resources/technical-bulletin/leonardo-da-vinci-s-virgin-of-the-rocks-treatment-technique-and-display
• [EARLY] Dunkerton, J., "Leonardo in [his teacher's] Workshop: Re-examining the Technical Evidence," National Gallery Technical Bulletin 32 (2011), pp.4–31. Covers the Baptism, the Annunciation, Ginevra de' Benci, the Madonna of the Carnation and the underdrawings of the Adoration and Saint Jerome, with the Munich medium analyses quoted in the notes. Full text read. https://www.nationalgallery.org.uk/media/15697/dunkerton2011.pdf
• [GON23] Gonzalez, V., Wallez, G., Ravaud, E., Eveno, M., Fazlić, I., Fabris, T., Nevin, A., Calligaro, T., Menu, M., Delieuvin, V. and Cotte, M., "X-ray and Infrared Microanalyses of Mona Lisa's Ground Layer and Significance Regarding Leonardo da Vinci's Palette," Journal of the American Chemical Society 145 (2023), pp.23205–23213. https://doi.org/10.1021/jacs.3c07000 . Read in full as chapter 3 of Fazlić, I., PhD thesis, University of Amsterdam (2025); page numbers are the thesis's. https://pure.uva.nl/ws/files/259566314/Chapter_3.pdf
• [GON20] Gonzalez, V., Hageraats, S., Wallez, G. et al., "Microchemical analysis of Leonardo da Vinci's lead white paints reveals knowledge and control over pigment scattering properties," Scientific Reports 10 (2020), 21715. Read through a summary of the open-access text. https://doi.org/10.1038/s41598-020-78623-5 ; https://pmc.ncbi.nlm.nih.gov/articles/PMC7730476/
• [VIG] de Viguerie, L., Walter, P., Laval, E., Mottin, B. and Solé, V. A., "Revealing the sfumato Technique of Leonardo da Vinci by X-Ray Fluorescence Spectroscopy," Angewandte Chemie International Edition 49 (2010), pp.6125–6128. https://doi.org/10.1002/anie.201001116 . Not read; known through the ESRF press release of July 2010 and search excerpts. https://www.esrf.fr/news/general-old/general-2010/new-light-on-leonardo-da-vinci2019s-faces
• [NB] The Literary Works of Leonardo da Vinci, compiled and translated from the original manuscripts (London 1883); Project Gutenberg edition The Notebooks of Leonardo da Vinci (Complete), cited by the editor's section numbers (§§618–637 "The artist's materials"; §728). https://www.gutenberg.org/ebooks/5000
• [TREAT] A Treatise on Painting, English translation of the treatise compiled from his notes (London 1802), cited by the translation's chapter numbers. https://www.gutenberg.org/ebooks/46915
• [LIVES] Lives of the Most Eminent Painters, Sculptors and Architects (2nd ed. Florence 1568), life of Leonardo, trans. G. du C. de Vere, vol. 4 (London 1912–15). https://www.gutenberg.org/ebooks/28420
• [KRK] National Museum in Kraków, "What Does the Paint Hide? History of the Examination of 'Lady with an Ermine'," Google Arts & Culture (summarizing examinations of 1945, 1952–55, 1991–92 and 2012). Read through a summary. https://artsandculture.google.com/story/what-does-the-paint-hide-history-of-the-examination-of-%E2%80%9Clady-with-an-ermine%E2%80%9D-the-national-museum-in-krakow/QQXx_0jmLd0bJg
• [WIKI-LE] "Lady with an Ermine," Wikipedia (technical sections citing the 1992 Washington examination). Read through a summary. https://en.wikipedia.org/wiki/Lady_with_an_Ermine
• [WIKI-NEL] "Madonna mit der Nelke," Wikipedia (German). Read through a summary. https://de.wikipedia.org/wiki/Madonna_mit_der_Nelke
• [UFF] Gallerie degli Uffizi, "Leonardo's Magic Cosmos: The Adoration of the Magi Restored" (2017), on the restoration by the Opificio delle Pietre Dure. Read through a summary. https://www.uffizi.it/en/events/leonardo-s-magic-cosmos-the-adoration-of-the-magi-restored
• [LOUV] Musée du Louvre, collections record for the Virgin and Child with Saint Anne: oil on poplar, 168 × 113 cm, with notes on the 2010–12 restoration. Read through a summary. https://collections.louvre.fr/en/ark:/53355/cl010066107
• [AP2] Roy, A. (ed.), Artists' Pigments vol. 2 (1993): azurite pp.25–26; natural ultramarine p.44. https://www.nga.gov/sites/default/files/migrate_images/content/dam/ngaweb/research/publications/pdfs/artists-pigments-vol2.pdf
• [SAL] Field's Chromatography, revised by T. W. Salter (London 1869), §290 "Blue Black," §293 "Ivory Black." https://www.gutenberg.org/ebooks/20915