Navigation – Plan du site

AccueilNuméros44-1“All that glisters is not gold”. ...

“All that glisters is not gold”. A colorimetric assessment of the touchstone for gold ternary alloys

« Tout ce qui brille n’est pas or » : Évaluation colorimétrique de l’efficacité de la pierre de touche pour les alliages ternaires
Arnaud Manas
p. 51-62

Résumés

Même si les études techniques et historiques abondent, les principes de la pierre de touche n’ont jamais été abordés de façon scientifique. Cet article étudie si et dans quelle mesure la composition des alliages ternaires Au-Ag-Cu peut être visuellement déterminée à la pierre de touche uniquement et sans l’aide d’acides forts. Il apparaît que la pierre de touche permet de surmonter les obstacles à une comparaison de la luminance des alliages (état et géométrie de la surface, éclairage…). Une approche colorimétrique confirme le haut de degré de précision de la méthode dans l’antiquité.

Haut de page

Entrées d’index

Mots-clés :

pierre de touche, essai, CIE Lab
Haut de page

Notes de la rédaction

rec. oct. 2019; acc. june 2020

Texte intégral

The author wishes to thank Vicky Buffery for her help and two anonymous reviewers for their very helpful comments and suggestions.

1. Introduction1

  • 1 “All that glisters is not gold”, The Merchant of Venice (II, vii).
  • 2 Pierre de touche in French, pietra di paragone in Italian, Prüfstein in German...
  • 3 “Touchstone testing is based on the fact that 24K gold resists all but the strongest acids. The pu (...)
  • 4 “The true test of the value of something” (Cambridge dictionary).
  • 5 See Eluère, 1986, on the touchstone of the late Bronze Age (800-700 BC) found in France and possib (...)

1Nowadays, in many cultures the word touchstone is better known in its metaphorical sense as referring to “an established standard or principle by which something is judged” (Cambridge dictionary). Nevertheless, this age-old technology remains the easiest and most widespread method of assaying gold alloys. These days, there are scores of tutorials available on youtube explaining how to test a piece of gold with acids and a touchstone.2,3 The touchstone and acid test are now technically and semantically4 equivalent but the acid test did not exist before the 19th century whereas the touchstone has been used since prehistoric times5 and is described in antiquity.

  • 6 HNO3 and HCl were discovered at the end of Middle Ages (Eluère, 2001).
  • 7 Craddock (2000).
  • 8 See Ridgeway (1895), Lord (1937) and Halleux (1982).
  • 9 “Section 46 has usually been interpreted as the weights of impurities which could be detected in a (...)
  • 10 Oddy (1983) gives a margin of 1 to 1.5 carat which corresponds to 4-6% while Metcalf & Merrick (19 (...)
  • 11 In assaying large quantities of gold dust from Korea, Gowland (1910) put aside pieces cut from the (...)

2This raises several questions. Firstly, how could the touchstone work before strong mineral acids (nitric and hydrochloric acid) were invented?6 Could vegetable or organic acids have been used? Secondly, which alloys could be assayed with a touchstone? Craddock7 raised doubts as to its efficiency in assaying ternary gold alloys (gold-silver-copper) before the discovery of mineral acids. He posited that the touchstone probably worked well for binary combinations (gold with either copper or silver), but failed when the gold contained the three metals in unknown combinations. In his view, this problem was only resolved in the medieval period when the introduction of mineral acids allowed the copper or other base metals to be selectively removed from the streak. Thirdly, how accurate was the touchstone in reality? According to classical authors such as Theophrastus or Pliny a very high degree of accuracy (up to 0.7%) could be attained with the touchstone8. However Craddock questioned this claim9, while Oddy and Metcalf estimated that an accuracy of only 5% could only be attained in antiquity.10 For Manche (2004) touchstone assaying was at best a “bet” and a “security theater” (if not a fraud) not unlike an ordeal to deter potential fraudsters. Nevertheless, Gowland (1910) obtained an accuracy of 0.5% to 1% with an experimental approach.11 Fourthly, was it possible to detect other specially designed alloys or minerals (fool’s gold, brass, bronze, etc.)? Craddock questioned whether it is possible to be fooled by “complex artificial alloys made by the artificers and alchemists”.

3In light of these questions, there is a need to investigate the touchstone’s physical principles scientifically. The first part of this paper describes and summarizes the key features in the use of the touchstone between antiquity and the 19th century. A physical model is then suggested and confirmed by colorimetric computations.

2. Key historical features: literary sources

  • 12 “They say that a much better stone has now been found than the one used before; for this not only (...)
  • 13 Carat (K) is the traditional measurement of fineness for gold: 1 carat is 1/24th. Therefore, 24K g (...)
  • 14 See “Lydian stone” in the Zuretti Anonymous, Colinet (2000), Wilson (1936) and in the Probierbüchl (...)
  • 15 See Allen (2012), p.219, Kilburn-Toppin (2019), p. 201 and Halleux (1982), p. 39.

4Theophrastus and Pliny were among the first to report the use of the βασανος (basanos) and the lapis coticula. Both12 concurred on the fact that it was used to test ternary gold-silver-copper alloys (Au-Ag-Cu) without acid. They stressed that a high degree of precision could be attained (1/144 or 1/6th of a carat13 i.e. 0.7%). The touchstone was then mentioned in the late antiquity and in the middle-ages14 where it became associated with goldsmiths and mint masters. A “keeper of the touch [touchstone] of the mint” was commissioned in London15 in 1370. In all these accountings, the touchstone was not used with any acid or reactant. This use is confirmed by SEM/EDS analysis of a touchstone found in the La Rochelle Mint showing the lack of chemical elements associated with strong acids (see Téreygeol & Blet-Lemarquand, 2011 and Arles, 2009).

  • 16 “[Touching] will show us how great a proportion of silver or copper, or silver and copper together (...)

5In the 16th century, Agricola16 confirmed the use of touchstones for testing ternary alloys and described the needles as being made of binary and ternary alloys (Au-Ag, Au-Cu, Au-Ag-Cu, and Ag-Cu) (see table 1, fig. 1 and Agricola, 1556). These needles were bound together with silver wires in ligatures (see fig. 2 & Diderot and d’Alembert, 1779). Seven ligatures amounting to 111 needles were needed to test the gold and silver content of ternary and binary alloys.

Table 1: Agricola’s set of needles / Tableau 1 : Aiguilles d’Agricola

Needles

Type

Content

23+1

Au / Ag

1 K (4.1%) between 1K and 24K

23+1

Ag / Cu

1/24 (4.1%)

12+1

Au / Cu

1 K (4.1%) between 12K and 24K

12+1

Au / (Ag50-Cu50)

3 × 12 + 1

Au / (Ag75-Cu25)

Au / (Ag66.6-Cu33.3)

Au / (Ag58.3-Cu41.6)

  • 17 Alloy Au-Ag is called caratura alba white carature. Ternary caratura mixta.

6In the 18th century Cramer used 128 needles (see table 2 and fig. 1) bound into 5 ligatures or caraturas17 (Cramer, 1741). He stressed that sometimes intermediary needles were made to get a higher degree of precision but not for the silver bundle (i.e. Ag- Cu) where it is almost impossible to achieve greater accuracy. Cramer described the qualities of touchstone that were required:

1 It must be of the deepest Black, lest the Tincture of the Metal should be hindered, by spurious Rays of Light shining between them.

2 It must be capable of being pretty well polished: For when too rough, the Colours of the Metals rubbed against it, cannot be neatly perceived: And if it is too smooth, the Metals are but faintly and slowly scraped by it, especially Gold.

  • 18 Which is not acid-resistant.

3 It must be neither too hard, nor too soft: For Tripoly, Coal-dust, and Tin-ashes rub off the small metalline Crusts: So that, in a short Time this Stone, though ever so hard, acquires too great a Smoothness: And when it is too soft, it is easily ground into a fine Dust, and contracts Furrows. Therefore, black rough Marble18 or black soft River-Pebbles, are excellent for that Use. They must be made into a quadrangular Prism, about one Inch thick, and two or three Inches long; which is the most convenient Figure for that Purpose.

Table 2: Cramer’s set of needles / Tableau 2 : Aiguilles de Cramer

Needles

Type

Content

15+1

Ag / Cu

1/16th (6.2%) between Ag and Ag6.25-Cu93.75

27+1

Au / Ag

0.5K (2%) between 24K and 20.5K and
1 K (4.1%) between 20K and 1K

27+1

Au / (Ag66.6-Cu33.3)

27+1

Au / (Ag33.3-Cu66.6)

27+1

Au / (Ag50-Cu50)

Figure 1: Agricola’s (left) needles and Cramer’s (right) / Figure 1 : Aiguilles d’Agricola (à gauche) de Cramer (à droite)

Figure 1: Agricola’s (left) needles and Cramer’s (right) / Figure 1 : Aiguilles d’Agricola (à gauche) de Cramer (à droite)

7In the late 18th century (Walchli, 1981), the development of chemistry substantially modified assaying as the proper acid solutions were formulated for testing. The use of acid in conjunction with the touchstone and the touch-needles was standardized in 1799 by Vauquelin in his “Manuel de l’essayeur” (1812). At that time, it was mentioned that the touchstone should be acid-resistant, and marble was not suitable anymore. Furthermore, there was a change in the relative values of metals in the 19th century as silver gradually lost its monetary role. In antiquity and medieval times, silver’s value was 1/12th of that of gold, while at the end of the 18th century it was 1/15.5. In the 19th and the 20th century, it fell steadily and currently stands at 1/85th of the value of gold (i.e. approximately 1%). Copper followed the same path. As a result, it became less and less important to know the silver and copper content of an alloy. What really mattered was its gold content and the acid test was perfectly suited to that. The touchstone evolved as a form of acid test and became equivalent to the acid test. The number of needles was drastically reduced to less than 10 and several acids were needed. Furthermore, assaying became much easier and visual experience became less important, reducing the cost.

  • 19 Its hardness (4.7-6.5 on Mohs’ scale) must be higher than the hardness of the gold alloys to be pr (...)
  • 20 “Touchstones are made into Lamina’s one Line broad, one fourth Part of a Line thick, and one Inch (...)
  • 21 During an assay, it can be estimated that the loss was at least a few tenths of a mg each time (se (...)

8To summarize, according to primary sources, the touchstone has always been used to assess the gold and silver content of ternary alloys. Before the 18th century it was not used with acid (mineral, vegetable nor organic). Weak acids or salts used in cementation were useless for probing because they do not attack gold, have a very slow kinetic and would not provide visible signs of a chemical reaction. The touchstone was a purely visual method based on color. It was made of moderately hard black or dark fine grained stone19 that needed not to be acid-resistant. A large number of needles was needed (over 100) implying a substantial investment (the equivalent of 50 grams of gold).20 The expected accuracy must have been between half-a-carat and a quarter of a carat (i.e. 1-2%) otherwise it would have been costly and useless to manufacture needles with half a carat step. Nevertheless, this still raises one question: why using a touchstone to assess the gold content and not compare the needles directly with the object as this would avoid some (minimal) damage to it and some wear of the needles.21

Figure 2: Needle ligature (from Agricola) / Figure 2 : Ligature d’aiguille (Agricola)

Figure 2: Needle ligature (from Agricola) / Figure 2 : Ligature d’aiguille (Agricola)

3. Physical model of the touchstone

  • 22 Cold working and annealing have an effect stress and grain size which modify alloy hardness (dislo (...)
  • 23 See Prescott et al. (2011).
  • 24 “Spatiochromatic properties of this channel are optimized for encoding reddish or yellowish fruit (...)

9Before the 18th century, assaying with a touchstone requires several elements: a golden object, gold needles, a human observer and a touchstone. The method is mainly based on the use of vision and on color recognition. Assayers might also have used their sense of touch to gain further information as hardness is linked to the gold content22. The streak’s aspect (granularity, aspect…) and the haptic sensations23 felt when rubbing could have given them additional information. Nevertheless the touchstone was based upon vision which is the most prominent of the five senses. It is particularly efficient at detecting red-yellow colors24 because of the overlapping spectral sensitivity of long- and medium-wave cones (L M) which represents a trade-off between red-green discrimination and luminance sensitivity.

  • 25 “When light falls onto a metal it is so intensely absorbed that it can penetrate to a depth of onl (...)
  • 26 550 nm (green) for Squair and 470 nm (blue) for Eales.

10The color of a metal depends on its electronic properties.25 For alloys, the color can be difficult to predict when the metals have different structures. Fortunately, gold, silver and copper are very close together in the periodic table and are in the same column (group 11). They are mutually soluble in all proportions and the colors of ternary alloys (Au-Ag-Cu) are intermediate between gold’s yellow, silver’s white and copper’s red. It is therefore natural to assess ternary alloys on the basis of their color. Eales (1967a, 1967b and 1968) and Squair (1965) estimated the fineness of a binary gold silver alloy with a single measurement using a refractometer. They matched the reflectance and the silver content for a given wavelength.26 For ternary alloys, there are two unknown factors: the silver content (u) and the gold content (v). Two independent measurements are therefore necessary. Human vision has three independent receptors (blue, green and red cones). It is therefore possible to associate a color with an alloy. Nevertheless, some practical problems need to be overcome. As Oddy (1986) noted: “unfortunately, however, [colour assaying] is not very accurate, as the perceived colour of the surface is dependent on a number of factors, including the degree of roughness of the surface and the extent of oxidation of the alloying element which is present.” (see also Oddy, 1993). Examining gold objects without precautions can be misleading since illumination is not controlled. Furthermore, the object’s roughness, surface state and shape have an impact on its appearance and therefore on the perceived color. Lastly, human vision has some limitations that must be taken into account.

  • 27 The sun’s luminance is in the order of magnitude of 109Cd/m², a candle flame’s is 105Cd/m² and the (...)

11Illumination is the first obstacle. In an uncontrolled environment, the characteristics of the lighting (spectral composition and intensity, geometric distribution) can differ significantly, and this has a direct effect on the perceived colors. The issue of illumination is particularly problematic in the case of metals as we are dealing with glittering objects. Shiny metal objects behave like mirrors: light rays follow the laws of reflection. As a consequence, an object differs in color depending on its orientation. Even two identical specular objects with the same orientation and set side by side can differ in color because of a lighting parallax (see fig. 3). Another problem arises with very high contrasts. A light source, such as the sun or a candle can be directly reflected by a gold object. The luminance of these specular reflections may have a very high order of magnitude in comparison with other elements27 and might blind a human observer who can only accommodate up to 3 or 4 orders of magnitude. Another problem with specular reflection is the surface curvature; a concave surface (with positive bending radius) focuses light like a parabolic mirror and increases luminance. Conversely, a convex surface (negative bending radius) disperses light and decreases luminance. In addition, the state of the object’s surface also plays an important role since crevices or dirt that has collected in micro-asperities, reduce reflectance (see Eales, 1967a) and therefore luminance. To overcome these problems, visual comparison should be made in a light box (tent) with controlled diffuse lighting shone on objects with the same bending radius and surface state. This would probably have been somewhat difficult in antiquity.

12The last difficulty lies with the human observer who is fallible and subject to illusions (Foster, 2011). Moreover, there are several well-known psychophysical effects that modify perception (see Pridmore & Melgosa, 2015). To maximize discrimination, Hunter & Harold (1987), suggested matching specimens simultaneously against a black or gray background in order to achieve maximum sensitivity to visual differences (see table 3). Comparing the test object and the golden needles directly following Hunter’s rule would have been very cumbersome.

  • 28 Popular mechanics, 1940 “The Old Touchstone Method and the Yellow Golds”.

13The touchstone solved all these problems. As Oddy noted it “What is needed to increase the accuracy of the comparison of color is a way of standardizing the examination; the Greeks discovered that this could be done by rubbing the alloys on to the surface of a smooth black stone.” (Oddy, 1986). There are obvious benefits. Firstly, this standard procedure in mineralogy reveals the streak color which is the color of the powder of a mineral enabling its identification. For instance fool’s gold (pyrite FeS2) has gold color but has a black streak. Hematite (Fe2O3) which is black in appearance leaves a red streak. For gold ternary alloys, objects and their streaks have the same color. Secondly, rubbing may reveal surface treatments such as plating, depletion gilding “tumbaga” (Jacobson, 2000). In particular, it is suggested when suspecting a forgery or a plated item to insist with the touchstone by removing a deeper layer (or even using a file or a penknife28). Walchli & Vuilleumier (1985) mentioned the dilemma with the touchstone: on the one hand grinding must be sufficiently deep to be sure of reaching the base metal, while on the other hand excessive damage to the object must be avoided. Thirdly, it may reveal the hardness of the material used and can give an indication of its Mohs’ index (see supra). In fact the touchstone’s primary advantage comes from optical effects.

Figure 3: Specular reflection with lighting parallax (lateral background color) / Figure 3 : Réflexion spéculaire avec parallaxe d’éclairage (couleur latérale)

Figure 3: Specular reflection with lighting parallax (lateral background color) / Figure 3 : Réflexion spéculaire avec parallaxe d’éclairage (couleur latérale)

Table 3: Hunter’s rules / Tableau 3 : Règles d’Hunter

  • Place the specimens in immediate juxtaposition. As the visual dividing line separating them becomes thinner, the visual differences become easier to see.

  • Keep the intensity of illumination high. The eye can only make comparisons with maximum precision when light levels are similar to those of an outdoor overcast sky.

  • Use a background that is similar to, and if possible, grayer than the specimens so that there is no distracting contrast with the visual task.

  • Use an illumination that is spectrally representative of that normally employed in critical studies (usually actual or artificial daylight).

14From a physical perspective, touching has four major effects: it transforms a curved surface into a flat one, changes a specular surface into a lambertian one, homogenizes the state of the object’s surface and allows direct visual comparisons to be made.

15By rubbing a 3D shaped gold object onto a 2D flat surface and transferring a small layer, the curvature is removed. The gold needle can be rubbed next to the object’s streak. The object and the needle streaks on the touchstone behave like coplanar mirrors. The illumination intensity is no longer modified by the object’s shape. The needle and the object can therefore be visually compared in optimal conditions against a dark neutral background.

16The second effect is to transform a specular surface into a lambertian one. The gold specks (grains i.e. gold crystals) which are transferred from the object to be tested (and from the needle) to the touchstone are randomly oriented. The streak behaves according to the Torrance-Sparrow (Torrance & Sparrow, 1967) model which represents a surface as a distribution of perfectly specular and planar microfacets. The facet slopes are normally distributed. Roughness is measured by the standard deviation of the facets’ orientation (see Lavin, 1971). Reflection is no longer specular but diffuse. The streaks have a (quasi) lambertian reflectance, are not subject to color parallax and suffer from the same color shifts due to multiple reflections and Fresnel’s incidence. There is no need to control the illumination strictly with a light box as both streaks receive virtually the same light rays and the incoming color spectra (provided that it is not too distorted) are identical (see fig. 4).

Figure 4: Diffuse reflection without lighting parallax / Figure 4 : Réflexion diffuse sans parallaxe d’éclairage

Figure 4: Diffuse reflection without lighting parallax / Figure 4 : Réflexion diffuse sans parallaxe d’éclairage

17The third effect is to homogenize the state of the object’s surface and remove the so-called Beilby layer. Some objects can be more or less polished and worked and more or less clean. This dirt is not (or only marginally) transferred onto the touchstone. This standardizes the streaks. Since they have an identical BRDF (Bidirectional Reflectance Distribution Function) and diffuse lighting, their perceived luminance depends only on the alloy content and not on the shape or the state of the object’s surface.

18Whereas a direct comparison can only be used to compare chromaticity which represents 2 color parameters, the touchstone allows all three parameters (2 for chromaticity + 1 for luminance) to be compared. In theory, chromaticity should be enough to determine the silver and gold content of a ternary alloy. In practice, however luminance is indispensable for three reasons: it increases accuracy, discriminates between metameric ternary alloys and detects brass or bronze objects.

4. Colorimetry and accuracy

  • 29 L*=0 darkest black and L*=100 brightest white). The origin (a* = 0 and b* = 0) represents true neu (...)
  • 30 ∆ELab=√(∆L*)2+(∆a*)2+(∆b*)2, see Mokrzycki & Tatol (2011).

19In a colorimetric approach, colors have two independent attributes: lightness or luminance and chromaticity. Luminance is related to luminous intensity while chromaticity consists of two parameters the color’s hue and saturation. The CIE (Commission Internationale de l’Éclairage) defined a standard color space L*a*b* that has three parameters L* lightness, a* (red-green component) and b* (blue-yellow component).29 The simultaneous comparison of color has been formalized and can be measured as a distance30 between two sets of coordinates. Observers’ lower limit is ∆E=1 (see table 4).

20Using Manas (2018a) color model for gold ternary alloys, it is possible to assess whether ternary alloys can be assayed with a touchstone and without acids, and to what degree of accuracy. In the L*a*b* color space, the chromaticities of the ternary alloys form a “wizard’s hat” with a pointed tip and furled rim.

Table 4: Color difference / Tableau 4 : Différence de couleur

0 < ∆E < 1

observer does not notice the difference

1 < ∆E < 2

only an experienced observer can notice the difference

2 < ∆E < 3.5

an unexperienced observer also notices the difference

3.5 < ∆E < 5

a clear difference in color is noticed

5 < ∆E

observer notices two different colors

21The shape is more complex when luminance is taken into account (see fig. 5). It should be noted that metameric alloys located in the “furled rim” cannot be distinguished by their chromaticity alone. Conversely when luminance is taken into account, it is easy to distinguish them as ∆L*=2.38 (see table 5). The same remarks apply to non-gold alloys. Craddock noted that the pre-acid touchstone could be fooled by specially designed alloys made by an alchemist. The primary goal of alchemy was surface coloring and the tincturing of metals. Of the 99 chemical recipes contained in this Leiden Papyrus X, processes for coloring metal surfaces, mostly to resemble gold or silver, feature prominently (Jacobson, 2000, see Cobb et al., 2014 for the golden penny experiment). Brass, copper-zinc alloy, has the same chromaticity as ternary alloys. Even though zinc was unknown as a metal to Greek alchemists, they achieved to make brass through the medium of καδμεια (cadmia), an impure oxide found in the flues of smelting furnaces (Taylor, 1930). For instance Cu70-Zn30 and Au70-Ag9-Cu21 have the same a* and b* values (see table 5). Similarly tin bronze and low carat gold alloy (10K) cannot be distinguished by chromaticity alone (∆E < 1, see table 5). Visually, one could be fooled by their identical chromaticity. However, with the touchstone the metals appear to be different due to their luminance. As gold and silver alloys have a higher reflectance than zinc, tin and copper, the “complex artificial alloys of the artificers and alchemists” mentioned by Craddock could not fool an assayer.

Figure 5: Color L* for gold ternary alloys / Figure 5 : Couleur L*

Figure 5: Color L* for gold ternary alloys / Figure 5 : Couleur L*
  • 31 Maybe with lifelong experience or special gifts an assayer could have routinely noticed smaller co (...)

22For Cramer’s needles, the color difference in each ligature (decreasing gold content but constant proportions between silver and copper) (∆A) is on average above or close to 1. When, the difference is higher than 1, the assayer can easily surround the gold object with two needles with noticeably different colors (since their color difference is above 1) and determine which one is the closest to the object (see table 7). For instance, if an object has a color between 22.5K and 23K, it will either be between 22.5 and 22.75K or between 22.75 and 23K. Accuracy is therefore 0.25K. When the difference is smaller than one (specially for assaying the silver content where ∆B can be as low as 0.3), the accuracy is the same but the methodology is slightly different: a larger number of needle must be used, and the range is given by the maximum number of needles having the same color, i.e. where no difference in color can be noticed (see fig. 6). The accuracy of 1/4K (i.e. 1/96th, around 1%) is therefore the minimal accuracy given by the touchstone. In special cases, it is possible to have a much higher degree of accuracy using overlapping needles for alloys with high color variations. For instance, for Au 78.4-Ag21.6 alloys (18.7K) the rate of change in noticeable color (∆E = 1) is maximal (see table 6 and table 7). Achieving this greater accuracy with all alloys would probably require better vision and the ability to go under the ∆E = 1 threshold. A variation of half a carat (0.5K) can be visually noted by assayer. The accuracy is 1/184th and this could be increased using overlapping needles. It just required “a good judgement, gotten rather by yeares and experience” and “a perfit eie to vewe31” (Kilburn-Toppin, 2019, p. 213). Neither Pliny nor Theophrastus displayed Mediterranean braggadocio.

Table 5: Metameric alloys (identical or similar chromaticity) / Tableau 5 : Alliage métamériques (chromaticité semblable)

  • 32 CIELAB’s coordinate were computed for Cu70-Zn30 alloy using M. R. Querry. Optical constants, CRDC- (...)
  • 33 See Radivojević et al., 2018.

Alloy

L*

a*

b*

Au38.5 - Cu61.5

92.39

4.89

18.46

Ag1 Ag16-Cu84

94.78

4.89

18.46

Cu70-Zn30 (brass32)

90.80

-2.52

28.59

Au70-Ag9-Cu21 (17K)

93.44

-2.52

28.59

Cu89-Sn11 (tin bronze33)

85.28

8.13

17.97

Au40-Cu60 (10K)

90.28

7.70

17.08

Table 6: Maximal color variation / Table 6 : Variation maximale de couleur

Alloy

L*

a*

b*

Au78.4-Ag21.6

98.02

-5.04

25.47

Au77.9-Ag22.1

98.06

-4.71

24.52

Figure 6: Overlapping needles accuracy / Figure 6 : Précision d’aiguilles se recouvrant

Figure 6: Overlapping needles accuracy / Figure 6 : Précision d’aiguilles se recouvrant

Table 7: Needles Color coordinates and differences / Tableau 7 : Coordonnées colorimétriques des aiguilles

Au

Ag

Cu

L*

a*

b*

∆B

∆A

24

91.5

1.2

42.0

 

 

23.5

0.50

92.1

-0.6

43.2

 

2.3

23.5

0.33

0.17

91.9

-0.3

42.8

0.5

1.8

23.5

0.25

0.25

91.9

-0.2

42.6

0.3

1.5

23.5

0.17

0.33

91.8

-0.0

42.4

0.3

1.3

23

1.00

92.7

-2.4

44.3

 

2.1

23

0.67

0.33

92.4

-1.8

43.4

1.1

1.6

23

0.50

0.50

92.2

-1.5

43.0

0.5

1.4

23

0.33

0.67

92.1

-1.2

42.6

0.5

1.2

22.5

1.50

93.2

-4.0

45.1

 

1.9

22.5

1.00

0.50

92.8

-3.1

43.9

1.6

1.4

22.5

0.75

0.75

92.5

-2.6

43.2

0.8

1.2

22.5

0.50

1.00

92.3

-2.1

42.6

0.8

1.0

22

2.00

93.7

-5.5

45.6

 

1.6

22

1.33

0.67

93.1

-4.2

44.0

2.1

1.2

22

1.00

1.00

92.8

-3.6

43.2

1.0

1.0

22

0.67

1.33

92.5

-3.0

42.4

1.1

0.9

21.5

2.50

94.2

-6.8

45.8

 

1.4

21.5

1.67

0.83

93.5

-5.2

43.8

2.6

1.0

21.5

1.25

1.25

93.1

-4.4

42.9

1.3

0.9

21.5

0.83

1.67

92.7

-3.6

41.9

1.3

0.8

21

3.00

94.6

-7.9

45.6

 

1.2

21

2.00

1.00

93.8

-6.0

43.4

3.0

1.0

21

1.50

1.50

93.3

-5.1

42.3

1.5

0.9

21

1.00

2.00

92.9

-4.1

41.2

1.5

0.9

20.5

3.50

95.0

-8.9

45.1

 

1.2

20.5

2.33

1.17

94.0

-6.7

42.7

3.4

1.0

20.5

1.75

1.75

93.5

-5.5

41.5

1.7

0.9

20.5

1.17

2.33

93.0

-4.5

40.3

1.7

1.0

20

4.00

95.3

-9.6

44.2

 

1.2

20

2.67

1.33

94.3

-7.1

41.7

3.7

1.2

20

2.00

2.00

93.7

-5.9

40.4

1.8

1.2

20

1.33

2.67

93.2

-4.6

39.2

1.9

1.1

5. Conclusion

  • 34 Qualitative methods were based upon smell, sound, color changes when heated and mechanical propert (...)
  • 35 See Caley (1949), Namer (1964), Kraut & Stern (2000), Hughes & Oddy (1970).
  • 36 See Téreygeol & Thomas (2003) and Ramage & Craddock (2000).
  • 37 The first accounts of this transition date back from 500 AD (see Craddock’s Assaying in Antiquity (...)
  • 38 Vinegar’s and lemon juice’s pH are 2.5. The average value for urine pH is 6.0 but it can range fro (...)
  • 39 Green vitriol (iron sulfate, FeSO4), and common salt.

23Touchstone was one of the three quantitative methods34 available in ancient time. The two other methods were density (or specific gravity) measurement35 and fire assay36 (see fig. 7). As Corti (2011), noted it, these three methods are still in use today but their modus operandi evolved over centuries. The touchstone that started as a purely colorimetric method became a physicochemical method at the beginning of the 19th century. The density measurement method began by measuring the excess volume overflow discovered Archimedes. It then evolved by measuring the buoyant force with a scale (Galileo, Bilancetta, 1586) in the late Antiquity37. The fire assay started with a cupellation stage followed by a cementation one to dissolve silver with weak acids38 and salts39. The slow cementation stage was replaced by a quick parting one using nitric acid at the end of the Middle-Age.

Figure 7: Classification of assaying methods / Figure 7 : Classification des méthodes d’essai

Figure 7: Classification of assaying methods / Figure 7 : Classification des méthodes d’essai
  • 40 To assay ternary alloys by density measurement a preliminary cupellation stage is needed to remove (...)
  • 41 Weighing must be done at each stage (Au-Ag-Cu before assaying, Au-Ag after cupellation and Au afte (...)

24Each method had its own benefits: density measurement and the touchstone were nondestructive but density measurement could only work for pure gold and binary alloys40 (Au-Ag or Au-Cu). Fire assay which relies on weighing made during the refining process41 was destructive but could be controlled by non-experts. The touchstone was nondestructive but relied on the judgement of an expert and its accuracy depended on the type of alloy. Therefore, the choice of the assaying method was clearly context and purpose dependent. For judiciary matters and quality control for mines or smelting furnace, fire assay was most appropriate. On the other hand, the touchstone was better suited to control quality of manufactured objects like coins or jewelry: “when gold coins are assayed in fire, of what use are they afterwards?” (Agricola).

Haut de page

Bibliographie

Agricola, G., 1556. De re metallica, J. Froben and N. Episcopius, Basel (translation: 1912 Hoover H. and Hoover L.H., The Mining Magazine, London).

Allen, M., 2012. Mints and Money in Medieval England, Cambridge University Press, Cambridge.

Arles, A., 2009. Entre monnayage officiel et faux-monnayage. La fabrication de la monnaie au marteau en France (xiiie - xviie siècles). Approche physico-chimique, expérimentale et historique. Thèse de doctorat. Université d’Orléans.

Biringuccio, V., 1540. De la pirotecnia, (translation : Vincent J., 1572, La pyrotechnie, ou L'art du feu. Claude Fremy, Paris).

Bompas, A., Kendall, G., Sumner, P., 2013. Spotting fruit versus picking fruit as the selective advantage of human colour vision, i-perception, 4(2): 84-94.

Bosten, J. M., Robinson, J. D., Jordan G., Mollon, J. D., 2005. Multidimensional scaling reveals a color dimension unique to ‘color-deficient’ observers. Current Biology, 15, R950-R952.

Burns, D. T., 1981. A Touchstone for Gold and Silver wares, Analytical Proceedings, 18: 146-151.

Caley, E. R., 1949. Validity of the Specific Gravity Method for the Determination of the Fineness of Gold Objects. The Ohio Journal of Science. 49 (2): 73-81.

Caley, E.R., Richards J.F.C., 1956. Theophrastus on Stones. The Ohio State University, Columbus.

Cauuet, B., Tama, C.G., Boussicault, M., Munoz, M., 2018. Quantités et contrôle de l’or produit à l’âge du fer en Gaule du Centre-Ouest. Mélanges de la Casa de Velázquez, 48-1: 13-42.

Cobb, C., Fetterlof, M. L., Goldwhite, H., 2014. The chemistry of alchemy, From Dragon’s Blood to Donkey Dung How Chemistry was forged, Prometheus book, New York.

Colinet, A., 2000. Les alchimistes grecs, tome X, L’anonyme de Zuretti. Les Belles Lettres, Paris.

Colinet, A., 2010. Les alchimistes grecs, tome XI, Recettes alchimiques. Les Belles Lettres, Paris.

Corti, C.W., 2011. Assaying of Gold Jewelry – Ancient and Modern. The Santa Fe Symposium on Jewelry Manufacturing Technology, 49-70.

Craddock, P., 2000. Assaying in Antiquity (Appendix 5), in Ramage A. and Craddock P., 2000. King Croesus’ Gold, Excavations at Sardis and the History of Gold Refining, Harvard University Press, Cambridge: 245-250.

Cramer, J.A., 1741. Elements of the art of essaying metal, Woodward, London.

Diderot, D., d’Alembert, J., 1779. Encyclopédie ou dictionnaire raisonné des sciences, arts et métiers, Panckoucke, Stoupe et Brunet, Paris.

Eales, H. V., 1967a. Reflectivity of gold and gold-silver alloys. Economic Geology, 62: 412-420.

Eales, H. V., 1967b. Checking of optical data derived from polarization Figures, Economic Geology, 62 (5): 737-738.

Eales, H. V., 1968. Determining fineness variation characteristics in gold ores by reflectometry, Economic Geology, 63 (6): 688- 691.

Eluère, C., 1986. A prehistoric touchstone from France, Gold Bulletin, 19 (2): 58-61.

Eluère, C., 2001. Les secrets de l’or antique, La bibliothèque des arts, Paris.

Ercker, L., 1574. Beschreibung allerfürnemisten Mineralischen Ertzt unnd Bergwercks arten […], Georg Schwartz, Prague.(translation: Pettus J., 1683. Fleta minor the laws of art and nature, in knowing, judging, assaying, fining, refining and inlarging the bodies of confin’d metals, Thomas Dawks, London).

Foster, D. H., 2011. Color constancy, Vision Research, 51: 674-700.

Gowland, W., 1910. Arts of working metals in Japan, Journal of the Institute of Metals, 4: 4-41.

Halleux, R., 1982.·Méthodes d'essai et d'affinage des alliages aurifères dans l’Antiquité et au Moyen Âge. In Morrisson C. ed., Numismatique et histoire, l’or monnayé de Rome à Byzance : purification et altérations. Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 126e année, no 2: 203-223.

Hughes, M.J., Oddy, W.A., 1970. A reappraisal of the specific gravity method for the analysis of gold alloys, Archaeometry, 12:1-11.

Hunter, R. S., Harold, R. W., 1987. The Measurement of Appearance, John Wiley & Sons, New York.

Jacobson, D. M., 2000. Corinthian Bronze and the Gold of the Alchemists, Gold Bulletin, 33 (2): 60-66.

Ježek, M., 2013. Touchstones of archaeology, Journal of Anthropological Archaeology, 32: 713-731.

Jordan, G., Deeb, S. S., Bosten, J. M., Mollon, J. D., 2010. The dimensionality of color vision in carriers of anomalous trichromacy. Journal of Vision, 10(8) : 12.

Judd, D.B., 1943. Colorblindness and the detection of camouflage. Science, 97: 544-546.

Kilburn-Toppin, J., 2019. ‘A place of great trust to be supplied by men of skill and integrity’: assayers and knowledge cultures in late sixteenth- and seventeenth-century London, The British Journal for the History of Science, 52(2): 197-223.

Kraut, J.C., Stern, W.B., 2000. The density of gold-silver-copper alloys and its calculation from the chemical composition. Gold Bulletin. 33 (2): 52-55.

Lavin, E.P., 1971. Specular Reflection, Monographs on Applied Optics, no 2, Adam Hilger, London.

Lord, L.E., 1937. The Touchstone, The Classical Journal, 32, no 7: 428-431.

Manas, A., 2015. The music of gold: can gold counterfeited coins be detected by ear? European Journal of Physics, 36, no 4: 045012.

Manas, A., 2018a. Fifty shades of yellow. A color model for gold-silver-copper alloys, Gold Bulletin, 51: 205-212.

Manas, A., 2018b. Why do pirates and champions bite gold coins and medals? Revue numismatique, 175: 563-580.

Manche, G.F., 2004. Le contrôle des métaux monétaires, une technique aléatoire au service du bimétallisme : l’exemple vénitien. In Naissance de la science dans l’Italie antique et moderne : actes du colloque franco-italien des 1er et 2 décembre 2000 (Université de Haute-Alsace), Peter Lang : 93-118.

Metcalf, D. M., Merrick, J. M., 1967. Studies in the Composition of Early Medieval Coins, The Numismatic Chronicle, vol. 7: 167-181.

Mokrzycki, W., Tatol, M., 2011. Colour difference Delta E- A survey, Machine Graphics and Vision, 20(4): 383-411.

Moore, D. T., Oddy, W. A., 1985. Touchstones: some Aspects of their Nomenclature, Petrography and Provenance, Journal of Archaeological Science, 12: 59-80.

Morgan, M. J., Adam, A., Mollon, J. D., 1992. Dichromats Detect Colour-Camouflaged Objects That Are Not Detected by Trichromats, Proceedings: Biological Sciences, 248, no 1323: 291-295.

Namer, E., 1964. Le traité de « La balance hydrostatique » de Galilée. Revue d'histoire des sciences et de leurs applications. 17 no 4 : 397-403.

Nassau, K., 1983. The Physics and Chemistry of Color, The fifteen Causes of Color, John Wiley & Sons, New-York.

Oddy, W.A., 1983. Assaying in Antiquity, Gold Bulletin, 16 (2): 52-59.

Oddy, W.A., 1986. The touchstone: the oldest colorimetric method of analysis, Endeavour, New Series, Volume 10: 164-166.

Oddy, W.A., 1993. The assaying of gold by touchstone in antiquity and the medieval world. In Eluère C., Outils et ateliers d’orfèvres des temps anciens. Société des Amis du Musée des Antiquités Nationales et du château de Saint-Germain-en-Laye.

Prescott, T.J., Diamond, M.E., Wing, A.M., 2011. Active touch sensing, Philosophical Transactions of the Royal Society B, 366: 2989-2995.

Pridmore, R.W., Melgosa, M, 2015. All Effects of Psychophysical Variables on Color Attributes: A Classification System. PLoS ONE, 10(4): e0119024.

Radivojević, M., Pendić, J., Srejić, A., Korać, M., Davey, C., Benzonelli, A., Martinón-Torres, M., Jovanović, N., Kamberović, Ž., 2018. Experimental design of the Cu-As-Sn ternary colour diagram, Journal of Archaeological Science, 90: 106-119

Ramage, A., Craddock, P., 2000. King Croesus’ Gold, Excavations at Sardis and the History of Gold Refining, Harvard University Press, Cambridge.

Ridgeway, W., 1895. How far could the Greeks determine the Fineness of Gold and Silver Coins, The Numismatic Chronicle and Journal of the Numismatic Society, Third Series, 15: 104-109.

Sisco, A.G., Smith, C.S., 1949. Bergverk- und Probierbüchlein. A Translation from the German of the Bergbüchlein and of the Probierbüchlein with Technical Annotations and Historical Notes. The American Institue of Mining and Metallurgical Engineers, New-York.

Smith, E.A., 1913. The Sampling and Assay of the Precious Metals, comprising Gold, Silver, Platinum, and the Platinum Group Metals in Ores, Bullion, and Products, Charles Griffin & co, London.

Squair, H., 1965. A reflectometric method for determining the silver content of natural gold alloys. Transactions of the Institution of Mining and Metallurgy, 74: 917-931.

Taylor, F., 1930. A Survey of Greek Alchemy. The Journal of Hellenic Studies, 50: 109-139.

Téreygeol, F., Blet-Lemarquand, M., 2011. Une pierre de touche pour l’essai des métaux précieux. In Mille P. & Téreygeol F., L’atelier monétaire royal de La Rochelle – La place de Verdun : Archéologie d’un espace urbain. Presses universitaires de Rennes, Rennes, 136-138.

Téreygeol, F., Thomas, N., 2003. La coupellation des alliages cuivre-argent: approche expérimentale de l’essai d’argent par voie sèche. Revue d’Archéométrie, 27 : 171-181.

Torrance, K., Sparrow, E., 1967. Theory for Off-Specular Reflection from Roughened Surfaces. Journal of the Optical Society of America, 57: 1105-1114.

Vauquelin, L.-N., 1812. Manuel de l’essayeur, Klostermann, Paris.

Velde, F., 2013. On the Evolution of Specie: Circulation and Weight Loss in 18th and 19th century Coinage, Revue numismatique, 170: 605-650.

Walchli, W., 1981. Touching Precious Metals, Gold Bulletin, 14: 154-158.

Walchli, W., Vuilleumier P., 1985. Aurum, 24: 36-45.

Wilson, W. J., 1936. An Alchemical Manuscript by Arnaldus Bruxella. Osiris, 2: 220-405.

Haut de page

Notes

1 “All that glisters is not gold”, The Merchant of Venice (II, vii).

2 Pierre de touche in French, pietra di paragone in Italian, Prüfstein in German...

3 “Touchstone testing is based on the fact that 24K gold resists all but the strongest acids. The purer the gold, the stronger the acid required to dissolve it”. https://www.gia.edu/bench-tip-use-the-touchstone-method-for-testing-purity-karat-gold.

4 “The true test of the value of something” (Cambridge dictionary).

5 See Eluère, 1986, on the touchstone of the late Bronze Age (800-700 BC) found in France and possible ones dating from before the First Millennium BC, (Ježek, 2013 and Cauuet et al. , 2018).

6 HNO3 and HCl were discovered at the end of Middle Ages (Eluère, 2001).

7 Craddock (2000).

8 See Ridgeway (1895), Lord (1937) and Halleux (1982).

9 “Section 46 has usually been interpreted as the weights of impurities which could be detected in a one stater gold coin. Taken at face value, it suggests that quantities of one grain (krithë, about 0.06 g) could be determined in a stater of 8.72 g: that is, 1 part in 144, which, as Caley [Caley & Richards, 1956] remarks, seems very doubtful.” (Craddock, 2000).

10 Oddy (1983) gives a margin of 1 to 1.5 carat which corresponds to 4-6% while Metcalf & Merrick (1967) indicate 4%.

11 In assaying large quantities of gold dust from Korea, Gowland (1910) put aside pieces cut from the ingots of the gold dust, which varied from about 790 to 990 per 1000 of gold, and made touch needles from them. By the use of these he was able to determine the proportion of gold present to within 5 or 10 millièmes (i.e. five to ten parts per 1000), and thus to decide on the amount of silver to be added for parting. (Smith, 1913, p. 281).

12 “They say that a much better stone has now been found than the one used before; for this not only detects purified gold, but also gold and silver that are alloyed with copper, and it shows how much is mixed in each stater. And indications are obtained from the smallest possible weight. The smallest is the krithë, and after that there is the kolly bos, and then the quarter-obol, or the half-obol; and from these weights the precise proportion is determined.” (Theophrastus) and “In testing, workers skilled in the use of the whetstone take with it a scraping from an ore, as one might with a file, and can then tell straight off, to the nearest scruple, how much gold, silver, or copper it contains, by this wonderful method which never fails them.” (Pliny, Natural History), see Burns (1981).

13 Carat (K) is the traditional measurement of fineness for gold: 1 carat is 1/24th. Therefore, 24K gold is fine gold.

14 See “Lydian stone” in the Zuretti Anonymous, Colinet (2000), Wilson (1936) and in the Probierbüchlein (Sisco & Smith, 1949).

15 See Allen (2012), p.219, Kilburn-Toppin (2019), p. 201 and Halleux (1982), p. 39.

16 “[Touching] will show us how great a proportion of silver or copper, or silver and copper together, is in the gold, or else how great a proportion of copper is in silver. These needles are of four kinds. The first kind are made of gold and silver, the second of gold and copper, the third of gold, silver and copper, and the fourth of silver and copper.” (Agricola, 1556). “But the Gold Smiths take not so so much Pains, not are so large Expenses, but cut a piece of a Ducate and of a Crown, and of a Rhenish Gilder, and soder every on Copper piece, after this they touch their Gold: And by this they can very well see whether the Gold they have have its right content either of Ducats, Crowns or Rhenish Gold, but if there be a different Content, then they cannot know how much properly the Content is left” (Ercker, 1574, p. 139 translated by Pettus,). Contrarily, it is not mentioned in Biringuccio’s Pireotechnia (1540).

17 Alloy Au-Ag is called caratura alba white carature. Ternary caratura mixta.

18 Which is not acid-resistant.

19 Its hardness (4.7-6.5 on Mohs’ scale) must be higher than the hardness of the gold alloys to be probed (2.5 for Au, Ag and Cu) between 2.5 and 3.5-4 for alloys). The difference in hardness is not relevant as the softer material is rayed and deposited on the touchstone. On the petrography of touchstones, see Moore & Oddy (1985). According to The Great Soviet Encyclopedia, (3rd edition, 1979) touchstones are poorly metamorphosed, fine-grained carbonaceous or siliceous slates. The basic requirements for a touchstone are a hardness of 4.7–6.5 on Mohs’ scale and an 8- to 23% C content and no more than 2% admixed materials, such as Al, Fe, Ca, Mn, S, Na, and Cl. It should have no fractures and should not react with inorganic acids or their mixtures. Schungites found in the Urals and siliceous slates are suited for touchstones. Artificial touchstones are now available.

20 “Touchstones are made into Lamina’s one Line broad, one fourth Part of a Line thick, and one Inch and a half long, upon each of which you engrave a Mark, indicating their Purity, or the several Mixtures of the Metals which they are made of.” Cramer: 2 mm x .5 mm x 50 mm = .05 cc for gold 1 g average content 0.5 g hence 50 g of gold for 100 needles and 25 g of silver.

21 During an assay, it can be estimated that the loss was at least a few tenths of a mg each time (see Velde, 2013, on coin wear).

22 Cold working and annealing have an effect stress and grain size which modify alloy hardness (dislocations and Hall-Petch effect). This topic deserves further research. For the use of other senses in gold assaying see Manas (2015 and 2018b).

23 See Prescott et al. (2011).

24 “Spatiochromatic properties of this channel are optimized for encoding reddish or yellowish fruit or leaves on a background of foliage at relatively small viewing distances” (Bompas et al., 2013).

25 “When light falls onto a metal it is so intensely absorbed that it can penetrate to a depth of only a few hundred atoms, typically less than a single wavelength. Since the metal is a conductor of electricity, this absorbed light, which is, after all, an electromagnetic wave, will induce alternating electrical currents on the metal surface. These currents immediately reemit the light out of the metal, thus providing the strong reflection. [...] The reason for the variation in color among different metals and alloys lies in the variation of the absorption coefficient k with the wavelength. Silver has a very slightly reduced reflectivity at the extreme violet end of the spectrum [...]; this gives it a yellowish ”warm” sheen which permits one to distinguish it from the ”cooler” stainless steel.” (Nassau, 1983, p. 165).

26 550 nm (green) for Squair and 470 nm (blue) for Eales.

27 The sun’s luminance is in the order of magnitude of 109Cd/m², a candle flame’s is 105Cd/m² and the sky’s is 103Cd/m². A typical object would be between 101 and 103Cd/m².

28 Popular mechanics, 1940 “The Old Touchstone Method and the Yellow Golds”.

29 L*=0 darkest black and L*=100 brightest white). The origin (a* = 0 and b* = 0) represents true neutral gray.

30 ∆ELab=√(∆L*)2+(∆a*)2+(∆b*)2, see Mokrzycki & Tatol (2011).

31 Maybe with lifelong experience or special gifts an assayer could have routinely noticed smaller color changes. Regarding these special gifts, genetic anomalies that improve visual perception are possible. In particular, tetrachromacy which is the presence in the retina of a fourth type of cone helps to improve color detection as it enables the eye to see a wider spectrum of colors (Jordan et al., 2010). Nevertheless this type of anomaly is extremely rare and therefore unlikely. Another possibility is deuteranomaly which is a weak form of color blindness (see Bosten et al., 2005). In order to explain the surprisingly high frequency of this congenital red–green color blindness, it has been suggested that dichromats might have a form of evolutionary advantage in that they are better able to break certain kinds of color camouflage (Morgan et al., 1992 and Judd, 1943). In particular, deuteranomaly provides an advantage in luminance detection which is crucial for touchstone assaying. Therefore, it is possible that assayers with a higher sensitivity might suffer from deuteranomaly. Since assaying was mainly a male activity that was passed down through generations, this could be an explanation…

32 CIELAB’s coordinate were computed for Cu70-Zn30 alloy using M. R. Querry. Optical constants, CRDC-CR-85034 (1985). https://refractiveindex.info.

33 See Radivojević et al., 2018.

34 Qualitative methods were based upon smell, sound, color changes when heated and mechanical properties like ductility (extenso per malleum).

35 See Caley (1949), Namer (1964), Kraut & Stern (2000), Hughes & Oddy (1970).

36 See Téreygeol & Thomas (2003) and Ramage & Craddock (2000).

37 The first accounts of this transition date back from 500 AD (see Craddock’s Assaying in Antiquity in Ramage & Craddock, 2000, 247-250).

38 Vinegar’s and lemon juice’s pH are 2.5. The average value for urine pH is 6.0 but it can range from 4.5 to 8.0. The pH of the urina pueri virginis used to probe the “seven metals” that include gold (R12, p. 4 in Colinet, 2010) is not specifically known.

39 Green vitriol (iron sulfate, FeSO4), and common salt.

40 To assay ternary alloys by density measurement a preliminary cupellation stage is needed to remove copper.

41 Weighing must be done at each stage (Au-Ag-Cu before assaying, Au-Ag after cupellation and Au after cementation or acid parting). Adding silver for quartering does not change the method.

Haut de page

Table des illustrations

Titre Figure 1: Agricola’s (left) needles and Cramer’s (right) / Figure 1 : Aiguilles d’Agricola (à gauche) de Cramer (à droite)
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-1.jpg
Fichier image/jpeg, 854k
Titre Figure 2: Needle ligature (from Agricola) / Figure 2 : Ligature d’aiguille (Agricola)
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-2.jpg
Fichier image/jpeg, 406k
Titre Figure 3: Specular reflection with lighting parallax (lateral background color) / Figure 3 : Réflexion spéculaire avec parallaxe d’éclairage (couleur latérale)
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-3.jpg
Fichier image/jpeg, 193k
Titre Figure 4: Diffuse reflection without lighting parallax / Figure 4 : Réflexion diffuse sans parallaxe d’éclairage
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-4.jpg
Fichier image/jpeg, 191k
Titre Figure 5: Color L* for gold ternary alloys / Figure 5 : Couleur L*
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-5.jpg
Fichier image/jpeg, 318k
Titre Figure 6: Overlapping needles accuracy / Figure 6 : Précision d’aiguilles se recouvrant
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-6.jpg
Fichier image/jpeg, 277k
Titre Figure 7: Classification of assaying methods / Figure 7 : Classification des méthodes d’essai
URL http://journals.openedition.org/archeosciences/docannexe/image/7247/img-7.jpg
Fichier image/jpeg, 524k
Haut de page

Pour citer cet article

Référence papier

Arnaud Manas, « “All that glisters is not gold”. A colorimetric assessment of the touchstone for gold ternary alloys »ArcheoSciences, 44-1 | 2020, 51-62.

Référence électronique

Arnaud Manas, « “All that glisters is not gold”. A colorimetric assessment of the touchstone for gold ternary alloys »ArcheoSciences [En ligne], 44-1 | 2020, mis en ligne le 02 janvier 2023, consulté le 24 mars 2023. URL : http://journals.openedition.org/archeosciences/7247 ; DOI : https://doi.org/10.4000/archeosciences.7247

Haut de page

Auteur

Arnaud Manas

Banque de France - Head of Heritage Division and Numismatic Collections. Paris I - Sorbonne University - Research Fellow (arnaud.manas@banque-france.fr)

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Creative Commons - Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International - CC BY-NC-ND 4.0

https://creativecommons.org/licenses/by-nc-nd/4.0/

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search