Salvador Dalí

A sourced technical report on his Surrealist oils of about 1929–1949. 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. The evidence is his own books, above all his painting treatise of 1948; a museum catalogue of 1941; conservators' examinations of four paintings (about 1933, 1936, 1940, 1946); and one laboratory study, of the 1946 canvas. The treatise is prescriptive and copies whole lists from a colorman's handbook; where an examination contradicts it, this note says so. It was read through Internet Archive full-text search excerpts. The 1946 analysis was read in a master's thesis and in press releases for the paper itself. The 1936 and 1930s studies were read through page summaries.

Summary

Dalí painted these oils mostly on canvas, with many small wood panels in 1932–1936 and a few on board and copper [CR]. A panel of 1936 has a white priming; the analyzed canvas of 1946 a thin white ground of calcium sulfate, titanium white and talc [TFH; ALE]. He traced his designs from drawings and photographs and outlined them in graphite or black, though his treatise forbids pencil under paint [CIV; TFH; MEAD; 50S]. The paint is thin and smooth, laid in short strokes with small brushes inside the traced contours [CIV; TFH]. The treatise prescribes a lean monochrome underpainting in lead white and ivory black, fatter overpaintings and glazes, zinc white and a bluish black at the finish, and a medium of oils with dissolved amber [50S]. The 1946 paint does contain amber, and its zinc-white layers lie over lead-white ones [ALE; DEF]. The documented pigments are listed in §4.

Terms. Ground/priming: the preparatory layers between support and paint. Size: a glue layer that seals the fabric under the ground. Imprimatura: a thin colored veil over the ground. Underpainting: the first, usually one-color, layers that set light and shade. Camaieu: painting in tones of one color. Glaze: a transparent layer. Lean/fat: paint with little or much oil. IRR (infrared reflectography): imaging that shows carbon-based drawing under paint. UV fluorescence: the glow of materials under ultraviolet light, which tells resins apart. Metal soaps: compounds formed when metal ions from a pigment react with the fatty acids of the oil.

1. Supports

• Canvas and panel. Of 332 catalogue entries dated 1929–1949 in the range surveyed, 223 are oil on canvas, 75 oil on wood panel, 18 on cardboard or board and 2 on copper [CR]. Panels are a third of the works of 1929–1936 (60 of 184) but one in ten after (15 of 148) [CR]. Survey of catalogue numbers P 215–P 659 only.
• Small formats. "Nearly half" of about 200 paintings of 1929–1936 measure 13 inches (33 cm) or less [MEAD]; in the survey, 68 of 182 have no side longer than 33 cm, against 11 of 137 later [CR]. A famous canvas of 1931 measures 24.1 × 33 cm [CR]; a panel of 1934, 7 × 5½ inches [SOBY, list of works]. Canvases of 1938 and 1940 reach 100 × 142.5 and 81.5 × 100.5 cm [CIV].
• Copper. He painted a portrait of the early 1930s "directly on copper." The burnished metal's reflections hid his drawing, so he stuck a half-inch square of white paper on the tip of his nose to light the passage [SL p.18; ETH]. Two oils on copper of 1931–32 measure about 20 × 15 and 30 × 24 cm [CR].
• Canvas fibre. The canvases of twelve paintings at Figueres were analyzed for fibre; one of 1926 is a cheap, low-quality cotton [ORI]. Read through search summaries; the other eleven were not recovered.
• The treatise. A beginner's first work should be on "a gesso-covered wood panel, nonabsorbent and as smooth as possible," washed with a hard brush and rinsed with distilled water [50S].

2. Grounds

• White priming on panel. On the panel of 1936 he "first painted the setting of the sky, water and sand over white priming" [TFH]. Composition not reported.
• The 1946 canvas. A porous glue size (animal glue with calcite, about 100 µm) lies on the fabric. Over it is a compact white ground, also about 100 µm, of calcium sulfate, titanium white (anatase) and calcite, with talc, in a drying oil and perhaps a resin [ALE pp.66–72; DEF]. An in situ study had taken it for lead white; sampling corrected that [ALE p.66]. One object. That it was a maker's preparation is an inference from its fillers.
• A thin toned layer. A canvas of 1940 is first covered all over with "a diluted and uniform oil layer" of ochre so thin that "it reveals the grain of the texture of the fabric" [CIV].
• Imprimatura (treatise). A thin colored veil spread with a silk pad, with "a suspicion of blanc d'argent" (lead white); he calls it the "veil-imprimatura" [50S]. Not observed in any examined painting.

3. Drawing and lay-in

• Tracing and transfer. He worked from drawings and photographs and traced them onto the canvas; the Figueres collection keeps his celluloid sheets, transfer papers, incised drawings and photographs [CIV]. In the 1940 canvas, IRR shows the bread "transferred using a tracing paper or a similar material," its contours "perfectly traced in pencil" [CIV]. The 1936 panel was traced from a postcard; infrared shows outlines of "the figures, the internal contours and facial features" in pencil, ink or dilute black paint [TFH].
• Graphite, against his own advice. The treatise warns against "the foreign bodies of charcoal or pencil" under paint and against carbon paper [50S]. The Kimbell study of nine small paintings of 1929–1936 found that he "consistently did exactly that," with "extensive underdrawing" [MEAD]. A graphite line shows under a magnifier in an unfinished portrait of about 1933 [FGT].
• The treatise's transfer. Daub the tracing with oil paint of zinc white and burnt umber, "a color which dries the fastest," lay it face down on the panel and press it with an engraver's roller (Secret 24) [50S].
• Some passages without drawing. In the 1940 canvas the far landscape and two small figures were "painted with no preliminary drawing, straight on to the ochre paint layer" [CIV].
• Monochrome lay-in (treatise). The underpainting (Secret 31) is in lead white and ivory black, with a hard square brush and a lean medium; the work runs "from the monochrome to the polychrome, from the lean to the fat" [50S].

4. Pigments

The treatise prints two lists from the Belgian colorman Blockx's handbook: colors "Unsuitable for Artistic Painting" (p.141), which he says corroborate "twenty years of my experience," and colors "Which Can be Used with Confidence" (p.143), plus an annotated copy in the colorman's own voice [50S]. The rows below rest on analysis or on passages in Dalí's voice; colors found only in the copied lists are left out.

Pigment	Evidence	Notes
Lead white ("blanc d'argent," silver white)	Gray paint of 1946, with carbon black [ALE pp.68–72; DEF]; treatise: begin the underpainting with it, "warm," mixed with linseed oil [50S]	The list: "carbonate of pure lead" [50S]
Zinc white	Zinc-rich layers over lead-white layers, 1946 [DEF]; treatise: "finish your overpainting" with it, "icy," mixed with poppy oil, "very sparingly" in glazes over lead white [50S]	Also three parts lead white to two of zinc [50S]
Ivory black	Gray paint with phosphate, 1946 (ivory black suggested) [ALE p.69]; treatise: for the underpainting [50S]	Tone "imperceptibly ivory" [50S]
Blue black (vine black)	A black layer of carbon alone, 1946 [ALE p.72]; treatise: "for the song," the last layers; "cold, with an imperceptible blue tinge" [50S]	"Of weaker body than ivory or lamp black" [SAL §290]. The 1946 black's kind is uncertain [ALE p.66]
Cobalt blue	1946 [DEF]; treatise: the blue "with which I sense that you will end by living" [50S]
Cerulean blue	1946 [DEF]; "coeruleum" in the list [50S p.143]
Ultramarine	1946 [DEF]; treatise: he does not feel for it "the enthusiasm" of earlier times [50S]	Natural or synthetic not reported
Chromium oxide green, hydrated	"Chromium greens," 1946 [DEF]; Secret 34, listed as hydrated chromium oxide, French vert émeraude [50S]	He also bows to two lighter chromium-alumina greens of the list [50S]. The English text's "emerald green" is probably the same vert émeraude (inference)
Cobalt green	1946, in situ study [ALE p.66]	The list ranks it permanent [50S]
Aureolin (cobalt yellow)	1946, in situ study [ALE p.66]; treatise: "the most beautiful yellows are obtained by spreading aureolin over Naples yellow" [50S]	Transparent in oil, a glazing yellow [AP1 p.38; SAL §20]
Naples yellow	Treatise: a secret of its own; "the atmospheric color par excellence," some in "every kind of shadow"; keep metal from it [50S]	The colorman made it of true lead antimonate until 1974 [AP1 p.226]
Cadmium yellow, lemon and red; deep cadmium orange	Cadmium yellow, 1946 [ALE p.66]; Secret 35: "cadmium yellow, cadmium lemon and cadmium red"; "a deep cadmium orange" in an exercise [50S]
Strontium yellow	1946 [DEF]	Banned in the copied list as "strontiane yellow" [50S p.141]. Tubes sold as cadmium yellow were sometimes mostly strontium chromate [AP1 p.101] (inference that his was)
Earth pigments	1946, kinds not reported [DEF]; the ochre plain of 1940 (visual) [CIV]
Venetian red	Treatise: "of all the reds the most excellent is Venetian red" [50S]	A red iron oxide
Mars yellow, orange and violet	Treatise: "give the rank of captain to your Mars yellow, to your Mars orange and your Mars violet" [50S]
Yellow ochres (yellow, golden, de ru)	Treatise: the palette knife "will be the friend" of them [50S]
Burnt sienna; Italian earth	Treatise: burnt sienna and ultramarine are the two most fitting transparent colors; Italian earth may replace burnt sienna [50S]	The list calls Italian earth a hydrated iron oxide earth [50S]
Vermilion	Secret 36: "of an exceptional quality," only "in extreme and strategic moments," not in mixtures [50S]
Madder lakes	Treatise: "the madder rose and carmine lacquers," despite "their fragility"; the list's unroasted madder lakes [50S]
Cobalt violet	Secret 39: "the severe shade of cobalt violet" [50S]
Massicot (lead monoxide)	Raman, with lead white, in a gray cloud of 1946 [ALE p.68]	Uncertain; perhaps a drier or by-product

• Banished colors. Burnt umber goes "as categorically as bitumen," except in the transfer paint (§3); green earth only "rarely" [50S]. The copied list bans raw umber, mummy, chrome yellow, Prussian blue, Indian yellow and the copper greens [50S p.141].
• Extenders. Gypsum, barium sulfate, potassium sulfate, talc and clay minerals in the 1946 ground and paint [ALE pp.66–69].
• Maker. The 1946 amber matches the Blockx firm's amber solutions [ALE p.72; DEF]. Whose tube colors he used is not documented [ALE p.20].

5. Media and varnish

• Amber in the paint. Two paint samples of the 1946 canvas contain succinic acid, the marker of Baltic amber, in a drying oil [ALE pp.71–72]; the press account puts it mostly in the final layers [DEF]. Two samples; not shown for the whole picture [ALE p.72].
• Amber in the treatise. The only material he calls "sublime" is "the yellow liquid amber according to the formula arrived at by Jacques Blockx, which you dissolve in your oils" [50S]: one drop to five of oil for glazes, pure amber as varnish and for corrections [ALE p.21, paraphrasing 50S]. The maker's solution holds amber, spike oil, poppy oil and turpentine [ALE p.22].
• The "wasp medium" (Secret 41). Equal parts poppy oil, walnut oil and rectified turpentine; to five parts of it, yellow amber dissolved in spike oil "in which some wasps have been soaked" [50S; ALE p.21]. The wasps are rhetoric.
• Oils by layer (treatise). A lean medium for the underpainting; after it, no more turpentine; then bleached linseed oil "as thick as honey," lightened with poppy oil or given "snap" with walnut oil; final touches with Venetian turpentine [50S].
• Resin in the 1930s. The figures of the 1936 panel fluoresce green under UV, indicating "natural resin, such as dammar or mastic," used alone or "mixed with linseed oil paint to create a more liquid media" laid "with a very small brush"; the background is oil [TFH].
• The search for a medium. In 1940 he was "seeking that unfindable thing, the medium to paint in, the exact mixture of amber oil, of gum, of varnish" [SL p.383].
• Varnish. Clumps of varnish lie on the paint of a portrait of about 1933 [FGT].

6. Layers and application

• Thin, smooth, small strokes. In the 1940 canvas the bread was painted "with Dalí's characteristic short brushstrokes applied precisely within the contours" [CIV]. On the 1936 panel there is "not a single mislaid stroke or error" [TFH]. A 1946 gray paint layer is about 40 µm thick [ALE p.68].
• Brushes (treatise). A hard square brush, not sable, for the underpainting, "for your paint must be as dry and compact"; "a fine sable brush" for the overpainting [50S].
• Handling varies in one picture. The 1940 canvas sets a flat, nearly transparent ochre plain and a meticulous still life against "dark, delicate, liberated and spontaneous brushstrokes" in the distance [CIV].
• "Handmade photography." The 1941 catalogue describes a surface "similar to that of glossy photographic prints" and early flat tones like hand-tinted prints [SOBY pp.14–15]. By the late 1930s he set that bland finish against "passages of radiant and thick texture," "glittering impasto" and "long, slashing high lights" [SOBY pp.21, 27]. He aimed to paint "with the most imperialist fury of precision" [SOBY p.15].
• Whites and blacks in order. Lead white and ivory black below, zinc white and blue black above; "by glazing with ivory black on blue black, your grays will be dirty and pasty" [50S]. The 1946 canvas has zinc-rich layers over lead-white layers [DEF].
• Glazes. Successive overpaintings and glazes follow the underpainting; "you may even glaze with blue black on a camaieu painting painted à la prima" [50S].
• Drying. One secret is "that a painting should dry slowly" [50S].

7. Working sequence and reworking

• Order (examined works). White priming; the setting of sky, water and sand; then figures in a more fluid resinous paint (1936) [TFH]. A uniform ochre layer; the traced still life in short strokes; a free distant landscape (1940) [CIV].
• A setting first, an image later. He had painted a twilit landscape of rocks and an olive tree "to serve as a setting for some idea"; one evening he "saw" two soft watches, "prepared my palette and set to work," and two hours later the picture was finished [SL p.317]. His own account.
• Changes. A small painting of 1930 shows "extensive underdrawing and changes to the composition" [MEAD]. In the 1940 canvas IRR reveals a freely sketched monochrome lion's head hidden in the foliage [CIV].
• Corrections. The treatise keeps pure amber for correcting errors [ALE p.21, paraphrasing 50S].

8. Condition and aging

• Zinc over lead, 1946. Within twenty years the 1946 canvas showed patches of increased transparency, roughness and micro-cracking. They follow zinc-rich layers laid over lead-white layers; zinc white directly on the ground is unaltered [DEF]. The causes given are mobile resin acids from the amber, a reactive zinc oxide, uncured lead-white underlayers and chlorine contamination; the damage is now stable [DEF]. Zinc and lead soaps were found in the ground and in a paint drip [ALE p.71]. Read through press releases.
• Weak adhesion. Under magnification the resinous figures of the 1936 panel adhere weakly to the paint below; the panel is otherwise in excellent condition, kept behind glass [TFH].
• Acidic cotton. The 1926 cotton canvas is acidic and prone to splitting [ORI].

9. The tubes here

The historical names in this note are mapped to these tubes: lead white (blanc d'argent, silver white) → lead white; zinc white → zinc white; ivory black → bone black; blue black → charcoal black (vine black: the cool, weaker carbon black); Naples yellow → Naples yellow; aureolin → aureolin (a new tube); strontium yellow → lemon chrome (a proxy: both are pale chromate yellows); cadmium lemon → pale cadmium; cadmium yellow → cadmium yellow; deep cadmium orange → deep cadmium; yellow ochres → yellow ochre; Mars yellow → Mars yellow; Italian earth → raw sienna (a proxy); Mars orange → Mars orange; Venetian red → red earth; vermilion → vermilion; cadmium red → cadmium red; madder lakes → rose madder; Mars violet → Indian red (a proxy: the nearest purplish iron oxide); burnt sienna → burnt sienna; cobalt blue → cobalt blue; cerulean → cerulean blue; ultramarine → ultramarine blue; hydrated chromium oxide green and the 1946 "chromium greens" → viridian; cobalt green → Rinmann's green (the same cobalt-zinc oxide); cobalt violet → cobalt violet. The 1946 earths fall under yellow ochre and red earth.

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

• [50S] Dalí, S., 50 Secrets of Magic Craftsmanship, trans. H. M. Chevalier (New York: Dial Press, 1948; Dover reprint 1992). Read through Internet Archive full-text search excerpts of the reprint (a lending-only item); cited by Secret number where the excerpt gave it, and by page for the two color lists, which he copies from the colorman's Compendium. https://archive.org/details/50secretsofmagic0000dali
• [SL] Dalí, S., The Secret Life of Salvador Dalí, trans. H. M. Chevalier (New York: Dial Press, 1942). Full OCR text read. https://archive.org/details/in.ernet.dli.2015.76092
• [ETH] Etherington-Smith, M., The Persistence of Memory: A Biography of Dalí (1992), on the portrait "painted on burnished copper, which glared in the light." Read through an Internet Archive search excerpt. https://archive.org/details/persistenceofmem00ethe
• [SOBY] Soby, J. T., Salvador Dalí: Paintings, Drawings, Prints (New York: The Museum of Modern Art, 1941). Full text read. https://assets.moma.org/documents/moma_catalogue_2085_300061969.pdf
• [CIV] Civil, I., "The technical examination of Dalí's painting Two Pieces of Bread Expressing the Sentiment of Love reveals unknown aspects of Dalí's creative process," Avant-garde Studies 3 (2018), The Dalí Museum, St. Petersburg, Fla. (oil on canvas, 1940, Fundació Gala-Salvador Dalí). Full text read. https://thedali.org/wp-content/uploads/2018/05/Civil_Final.pdf
• [TFH] Tate, "Salvador Dalí's Forgotten Horizon," paintings conservation project page (oil on wood, 1936, 22.2 × 26.7 cm). Read through a page summary and search excerpts. https://www.tate.org.uk/about-us/projects/salvador-dalis-forgotten-horizon
• [MEAD] Meadows Museum, SMU, Dallas, exhibition page and press release for Dalí: Poetics of the Small, 1929–1936 (2018), on the Kimbell Art Museum's technical study of nine paintings (C. Barry), published in the catalogue (M. A. Roglán, ed.). Read through page summaries and a reprint of the release; the catalogue was not seen. https://meadowsmuseumdallas.org/exhibitions/dali-poetics-of-the-small/ ; http://arthistorynewsreport.blogspot.com/2018/09/dali-poetics-of-small-19291936.html
• [ALE] Aleccia, D., Characterization and monitoring over time of oil-based pictorial products manufactured by the Belgian company Jacques Blockx and their use in the 20th century painting The Temptation of Saint Anthony by Salvador Dalí, MSc thesis, Ca' Foscari University of Venice, 2022–23: pp.20–22 on the treatise, pp.64–72 on the painting (oil on canvas, 1946, Royal Museums of Fine Arts of Belgium), including the earlier in situ study by S. Hardy (Liège, 2021). Full text read. https://unitesi.unive.it/retrieve/fcf356f7-85aa-43b6-868a-51171c1ce748/867814-1274606.pdf
• [DEF] Defeyt, C. et al., "Early Degradation Behavior of Amber-Based Paint Layers in The Temptation of St Anthony by Salvador Dalí," Heritage 9, no. 2 (2026), 85, doi:10.3390/heritage9020085. Read only through press releases and news summaries; the publisher refused access. https://www.eurekalert.org/news-releases/1122628 ; https://phys.org/news/2026-04-dal-sublime-amber-medium-unusual.html
• [ORI] Oriola, M. et al., "Looking beneath Dalí's paint: non-destructive canvas analysis," Analytical Methods 6 (2014), 86–96, doi:10.1039/c3ay41094c. Read through search summaries.
• [FGT] Fundació Gala-Salvador Dalí, "Documenting the painting technique" (a portrait of about 1933). Read through a page summary. https://www.salvador-dali.org/en/artwork/conservation-and-restoration/documenting-the-painting-technique/
• [CR] Fundació Gala-Salvador Dalí, online Catalogue Raisonné of Paintings: "Technique" and "Dimensions" fields of entries P 215–P 659 (59 numbers returned no page), retrieved October 2026 and counted for works dated 1929–1949. https://catalogues.salvador-dali.org/catalogues/en/
• [SAL] Field's Chromatography, revised by T. W. Salter (London 1869), §20 "Aureolin," §290 "Blue Black." https://www.gutenberg.org/ebooks/20915
• [AP1] Feller, R. L. (ed.), Artists' Pigments vol. 1 (1986): cobalt yellow pp.37–40; a "cadmium yellow" that is strontium chromate, table note p.101; strontium chromate p.205; the colorman's lead antimonate until 1974, p.226. https://www.nga.gov/sites/default/files/migrate_images/content/dam/ngaweb/research/publications/pdfs/artists-pigments-vol1.pdf