protecteur delapatine. mineuse supérieure à 350 lux supprime l’effet point. Il a été démontré qu’une exposition lu tine sulfurée est protectrice jusqu’à un certain pa La parabolique. courbe une selon ternit se l’argent pollués, plus les environnements les Dans étudiés. été ont formaldéhyde du et formique l’acide de acétique, l’acide de concentrations des et l’HR, de température, la de ultraviolets, rayons des lumière, la de réaction la sur effets Lesd’argent. particules le chlorure d’argent se convertissant en fines visible, lumière la par affectées sont argent en médailles et pièces certaines de patines Les étudiés. été l’argentont sur mécanismes Deux accru. dégazement un à conduisant d’exposition, matériaux des dégradation la accélérer et organiques revêtements les re. La lumière visible est connue pour détruire les métaux n’étaient pas sensibles à la lumiè Jusqu’à ce jour, on pensait généralement que Résumé remove theprotective ofthetarnish. effect exposures above 350 lux have been shown to tarnish is protective to a certain degree. Light sulfide manner.The parabolic a in tarnishes environments,pollutedmore silver gated. In investi been have reaction the on trations concen methanal and acid methanoic acid, ethanoic RH, temperature, light, UV light, of effect The particles. silver fine to converting chloride silver with light, visible by affected are medals and coins silver some of mechanisms on silver have been studied. The off-gassing.Two enhanced to leading terials accelerate the deterioration of showcase ma known to break down organic coatings and to is light Visiblesensitive. light not are metals that thought been generally has it date To Abstr Keywords: [email protected] London, United Kingdom House Rangers David Thickett a

ct light, silver, , tarnish ------treatments between coins and medals and other silver artefacts. In many many In artefacts. silver other and medals and coins between treatments differencesin conservationresultedin has patina the importanceof The to be of objects. would in of which expected found silver patinas silver sulfide are the two common, coloured chemically-stable products andChloridedesign. theaccentuates silver polished and patina dark the polishes the silver at, the high points of the decoration. The and contrast from, between patina the removes handling Coin medals. and coins silver on desirable be very can product) corrosion of a stable layer Athin (a patina and med Light-sensitivity ofchlorid e-rich p light induced changes have been investigated. potential two These effect. protective drop the can removing resistance dramatically, its illuminated, if and, sulfide silver semiconductor n-type However, an rate. is reaction the slow and transport electron classic parabolic reaction: the insulating properties of the silver as sulfide inhibit cited often is silver of sulfidation high The chloride. of contain patinas concentrations the where levels light low even under deteriorate to shown been have medals and coins on patinas significant curatorially contrast, In pollutants. of variety a of concentrations higher emit then down breaking coatings or by accelerating the deterioration effect of materials in the vicinity,indirect which strong a have can but silver, clean of Visible light has not been reported to have a direct effect on the tarnishing Introduction experienced during curatorial usage or examination by visiting scholars. patinas are stable The when in changes. darkof storage orobservations with earlier the with short-term lightanalysed, been exposures have Museum appearance of patinas on silver coins and medals the on display of in alteration The of British instances of number a decades, two past the Over visual appearance. about concerns to due conservation coin in avoided generally been have Similarly, lacquers 1983). (Loperfido and damaging unethical considered is medals or coins from patina of removal surfaces However, tarnishing. polished from the protect to utilised often are Lacquers possible. if to produce a highly polished surface on three dimensional silver artefacts layer tarnish or corrosion the remove to practice common is it museums, als silver t Effe cts oflighton arnishing a tina s oncoins 1 METALS width ofviewis0.15mm onchloridepatina surface, Silver particles Figure 1 protector delacorrosión. efecto el eliminar demostrado han lux 350 protector. Exposiciones de luz por encima de punto cierto hasta es sulfuro corrosión de La parabólica. manera de corroe se plata la dos en la reacción. En ambientes más contamina concentracioneslas metanal y de metanoico temperatura, la HR, el ácido etanoico, el ácido ha investigado el efecto de la luz, la luz UV, la Se plata. de partículas finas en convierte se plata de cloruro el y visible, luz a exposición la por afectadas resultanplata de medallas y monedas algunas de pátinas Las plata. la en mecanismos dos estudiado han Se gases. de vitrinas, lo que provoca una mayor liberación cos y acelera el deterioro de materiales de las sible descompone los revestimientos orgáni no son sensibles a la luz. Se sabe que la luz vi metales los que pensar solía se ahora Hasta Resumen - - - with a totally matt surface. This phenomena was first investigated in 1987 of gloss and, in some instances, the black or purple patinas become white, of some silver coins and medals. The alteration manifests itself as a loss However, prolonged periods on display have caused alteration to the patina most instances, smaller than this. Analyses of in areas were, of the coin particles and the medal and μm 1 approximately is SEM on analysis EDX X secondary were detected because of of penetration the electron beam and emission of patina underlying the from elements the of some although silver, were detector EDX with Oxford Link Isis software) GEM confirmed(Oxford that analysis these Elemental particles it. at efficient extremely are produced particles silver the that likely is it and process, similar a is surface a on light when suspended in the atmosphere (Spedding 1974). Light scattering visible of scattering maximum cause to shown been have diameter μm 1 approximately of Particles gloss. of loss the hence and surface, the from Scattering of light by the particles causes less light to be specularly reflected 1. Figure in shown is particles such of group a of particles could be readily removed by gentle brushing. A SEM form of small (less than 1 μm diameter) particles on the patina surface. These of a silver chloride patina to produce elemental silver. This silver took the by was the caused degradation the of alteration In cases, the majority vast the extract collected as a potassium bromide disc. dampened ether.ofmethylspectrum butyl tertiarythe withand extracted was This swab wool cotton a with wiped was surface the large, too if or, accessory in a reflectance placed diffuse either were (FTIR). Artefacts coatings was investigated using Fourier transform infra‑red spectroscopy was to compared those which were The unaltered. of presence any surface patina altered the and examined were medal or coin each of sides Both 840 scanning electron microscope (SEM) (Bradley 1988, Thickett 1998). Over 90 affected coins and medals were examined visually and with a Joel months of the exhibition opening (Bradley 1988). ‘Greeks In Italy Gallery’ were observed torefurbished have newly been affectedthe within in three case a from coins Greek BC 5th-century when originated from burial. have The altered could patinas had or lower chlorine concentrations deposition dirt or dust with associated be could elements These surfaces. patina the on detected also were silicon and to be composed mainly of silver sulfide (Franey 1985, Rice and be would and expected to handling, of years the exposure atmosphere many from accumulate patinas Such surprising. is patinas ‘cabinet’ with on coins from a burial context, their high abundance on coins and medals be would expected patinas chloride Whilst patinas. to sulfide have mainly chlorine detected. Several unaffected coins were also analysed and found of half the than less was always its concentration however, of the patinas; many in detected was based. chloridepredominantly be to found Results of four such analyses are shown in Table 1. All of the patinas were surfaces were undertaken to characterise the composition of the patinas. ‑ rays below the particles. The minimum depth resolution for resolution depth minimum The particles. the below rays on silver t Effe cts oflight

1980). Sodium, arnishing

micrograph micrograph

2

METALS SNG 1299 1976 6-310 BMC 36 M4732 number Registration 306–285 BC Didracham ofAmastris 617 Coin II ofKhusra Antiochos VI 145–142BC Tetradrachm of 1766 James Stewart medal ofdeath of description reverse displayed white particle reverse displayed reverse displayed reverse displayed analysed Side surfaces of these coins and medals. patina are discussed in Table 2. No organic coatings were detected on the with (Tablethe boards display than the unaltered patinas retained on the side of coins or medals in contact collections. Four Greek archaic silver coins with of patinas altered collections. similar Museum in British the observed were appearance of altered causes Other restoration treatments or withchlorideinconservation reaction fromreaction chlorideburial withchloridefromreaction handling withchlorideindustparticles reaction 1995) (Robbiola withHClgas(posfromreaction incinerators) Cause Possible causesofchloridepatinas Table 2 errors quoted are in precision BD belowlimit detection EDX analyses ofcoin andmedalpatinas Table 1 of a coin as it is unusual to take an impression of one face only. Similarly, sides both on occurs high generally This white. it colourin can concentrations, and, surface the dulls presence Their surface. the from cleaned and calcite appear to be residues of these materials which were not properly silicates magnesium The medals. or coins of taken are impressions when 1996). French chalk, or talc and chalk are often used as separating agents (Lee FTIR by calcite as identified was and rich calcium be to found was and and were probably magnesium silicates. The fourth material silicon magnesium, of composed were these instances three In surfaces. were observed. Larger, 3‑20 μm diameter particles were observed on their particles but no silver based, to be found chloride were SEM. The patinas in when examined the no showed signs of appearance deterioration patina 62.4±0.5 69.6±0.6 92.3±0.8 77.3±0.7 84.2±0.4 65.1±0.9 69.2±1.0 91.3±0.9 90.7±0.8 silver Percentage Element Present 1.1±0.1 0.2±0.1 0.9±0.1 4.8±0.2 2.4±0.1 5.2±0.2 1.3±0.1 4.6±0.2 2.8±0.2 chlorine 0.2±0.1 BD 1.3±0.1 0.9±0.2 1.0±0.1 1.6±0.1 0.2±0.1 0.4±0.1 0.6±0.1 sulfur 17.2±0.6 15.2±0.5 4.6±0.5 12.9±0.6 9.8±0.3 11.7±1.0 11.5±1.1 2.5±0.5 4.6±0.5 oxygen

1). The potential causes 1). of causes The a potential chloride silver 19th century high, many coins obtainedfrom market in antiquity were never buried high for archaeological coins, butseveral coins andmedals high, gloves onlyadopted years inlasttwenty analysed insideshowcases, see Table 3 numbers, significant concentrations ofdustandchloride high, gallerieshave highdustlevels from very highvisitor using diffusiontubes, see Table 3 low, concentrations have beenmeasured below 0.9µgm Likelyhood inthisinstance BD BD BD 0.6±0.2 0.4±0.1 BD BD 0.5±0.2 BD sodium BD BD 0.3±0.1 0.4±0.1 0.3±0.1 BD BD BD 0.2±0.1 aluminium on silver t BD BD BD 0.7±0.1 0.5±0.1 0.5±0.1 BD 0.4±0.1 BD silicon Effe cts oflight arnishing 0.3±0.1 BD BD 0.4±0.2 0.2±0.1 BD BD BD BD calcium BD BD BD 0.7±0.4 0.7±0.3 BD BD BD BD

-3

3

METALS SEM and did not have significant amounts of particulate material on its its on material particulate of amounts significant have not did and SEM the in examined when alteration patina of signs no exhibited coin One only occur on the displayed surface of the coin. to be matt a would expected coin’s and this would surface dust deposition through the coupon into the measuring head. To verify this measurement measurement Tothis head. verify measuring back the into coupon the passed through and plate white the from reflected the coupon, through chloride passed silver chroma-meter the from light the arrangement, this co 1981) Billmeyer 1978, (CIE Lab CIE its measuring and plate ceramic white a standard on coupon chloride the by silver placing was determined of degree alteration SEM.The the in distinguished readily be could alteration of signs early very the which on surface flat a and homogeneity of advantages the had would better reproduce the patinas, but the commercially purchased sheet substrate to silver chloride produced by corroding silver. Such a material obtained silver chloride sheet. This was considered a superior experimental commercially from cut coupons with out carried were experiments All the and presence of the pollutant gases, acetic acid and formaldehyde. temperature humidity, relative light, were investigated factors The 1998). (Thickett problem the mitigate to used be could approach preventiveconservationa determine ifundertakenthe to alterationwas rate on factors environmental certain of influence the of study A board. unaffected when protected from light exposure, in proximity to a display silver chloride is well documented (Butts 1967), and the patinas have been of degradation photochemical The illuminated. are patinas medal or coin when occurred only has and nature in photochemical certainly almost is silver,to chloride silver of conversion the reaction, alteration main The Influenc ofthnvironmntonalter have caused a bloom in the wax coating on the coin. andcaused heating ofthecoin toover 35°C. Theseconditions appear to levels, light high very generated which light, a spot to proximity close in exhibited was coin This 1991). (Green surface the on wax a of presence surface. Only the displayed surface was affected and analysis indicated the the mean of at least five images collected from each coupon. A graph of of Agraph coupon. each from collected images five least at of mean the as determined was coverage surface percentage the and magnification ×4000 was at collected were patina Images number. atomic the lower its to whilst due grey white, appeared particles The contrast. number atomic by discriminated being particles the with utilised were software particles. The image processing abilities of the Oxford Instruments ISIS such of coverage surface percentage the measuring by determined was surface the on particles silver of were amount The SEM. using coupons examined also The exposure. after and before measured were coupons The alterations. of surface a range produced This weeks. to several hours few a from ranging periods for bulb 500W ML with tester fastness light a in were method, exposed coupons of number a chloride silver Microscal ‑ ordinates using a Minolta CR300 chroma-meter.CR300In Minolta a using ordinates on silver t Effe cts oflight a arnishing tion r a

te 4

Effects of light on silver tarnishing METALS

change in ‘L*’ versus the percentage coverage of particles on the surface is shown in Figure 2.

The change in lightness, ‘L*’, ∆L* values correlated best with the degree of reaction. The coupons also showed a distinctive browning, which was observed as a sudden increase in the ‘a*’ value, indicating a shift towards red. This caused a corresponding shift in the ∆E* values. No such browning was observed on any coins or medals from the British Museum; however, this has been observed on coins from the National Museum of Wales (Hill 1998). Neither ∆E* or ∆a* gave a good correlation with the number of particles present, and ∆L values have been measured in all following experiments. Although L* and hence ∆L* are constrained by a* and b* and not independent variables, they only depend on one tristumulus value, ‘Y’ and in this instance, their use appears to be empirically justified.

The exposure experiments were carried out in a custom-built light bleaching chamber (Daniels 1993). Silver chloride coupons were exposed inside small perspex boxes on a raised polyethylene disc. The light exposure at the coupon surface was measured at 40165 lux with 61 µW/lumen of UV. Neutral density and UV filters were placed over the lids of the boxes to examine the effects of different light exposure levels. Atmospheres with different RHs were generated by adding water–glycerol solutions (Miner 1953) to the bottom of the boxes, avoiding any direct contact with the silver chloride coupon. The light bleaching chamber had two fans, normally giving a chamber temperature of 39°C. Using just one fan increased the chamber temperature to 52°C. The effects of temperature were investigated by repeating the 50% RH experiment with just one fan running. Ethanoic (acetic) acid, methanoic (formic) acid and methanal (formaldehyde) atmospheres were generated above 1% solutions in 50% RH producing glycerol solution. The 1% solutions would be expected to generate vapour concentrations in the order of 10,000 µgm‑3 (Donovan Figure 2 1972), much greater than would be expected in museum display cases Decrease in L* determined by colorimetry against the coverage with silver particles (Grzywacz 1993). Results of the accelerated exposure tests are shown in determined by SEM and image analysis. Figures 3 to 5. Raising the temperature had no observable effect and this Errors are + 0.3 determined from replicate graph has not been included. measurements All of the ∆L* plots rise towards a final plateau value. The greater Figure 3 Decrease in L* as a function of light the initial rate, the higher the plateau and number of silver particles exposure with and without UV present. produced. The reaction is strongly influenced by the amount of light Errors are + 0.3 determined from replicate energy illuminating the silver chloride. The presence of high UV levels measurements has a marked effect, which is not surprising as silver chloride absorbs strongly in this region of the electromagnetic spectrum (Butts 1967).

Increasing RH and the presence of carboxylic acid and methanal vapours caused small increases in the degree of reaction, in the order of 30%. These results indicated that controlling RH or eliminating carbonyl pollutants from showcases would only have a limited effect and would not significantly extend the lifetime of the patina on coins or medals. Controlling light exposure would seem to be a better option to control the rate of alteration of the patina.

5 Effects of light on silver tarnishing METALS

Effect of light on protective effect of corrosion layers

The sulfidation of silver is often cited as a classic parabolic reaction. The reaction slows as it progresses, as the insulating properties of silver sulfide inhibit electron transport and the reaction rate is controlled by ‘diffusion’. However, silver sulfide is an n-type semiconductor and, if illuminated, its resistance can drop dramatically, reducing or even totally removing the protective effect. The potential for this to occur with patinas from a number of heritage environments has been assessed.

Silver coated piezo electric quartz crystals (6 MHz Purafill Onguard) were exposed at a range of English Heritage properties presenting very different environments:

• Apsley House – highly polluted urban

• Rangers House – polluted urban

• Brodsworth House – polluted rural

• Audley End House – unpolluted rural

• Walmer Castle – unpolluted maritime

The environments were characterised by monitoring the temperature and relative humidity with Meaco radiotelemetry system (Rotronic Hygroclip RH probes). The nitrogen dioxide, ozone and concentrations were measured with diffusion tubes provided by Gradko International. The dust deposition rate was determined by exposing glass slides. The amount of dust deposited on the slides was determined by microscopy and image analysis (Howell 2002). The dust was extracted from the slides with 18.2 MΩ water and the chloride concentration of the extract analysed with ion chromatography (Dionex 600 with AS14A column with Figure 4 0.02 M sodium bicarbonate and 0.018 M sodium hydrogen carbonate Decrease in L* of silver chloride sheet exposed to different relative humidity levels. eluent at a flow rate of 0.5 ml/minute). The environmental conditions Errors are + 0.3 determined from replicate are shown in Table 3. measurements Table 3 Figure 5 Environmental parameters at exposure sites Decrease in L* in the presence of ethanoic acid, methanoic acid and methanal. Site Amount of time in RH band (%) NO2 O3 H2S HCl Dust Errors are + 0.3 determined from replicate µgm-3 µgm-3 µgm-3 µgm-3 measurements -1 day -2 30-40% 40-50% 50-60% 60-70% 70-80% >80% in 30 days % coverage Cl deposition rate µg m Apsley 1.03 41.73 46.74 9.91 0.53 0.00 31 0.7 0.193 2.17 1.34 9 Rangers 16.04 35.93 29.05 12.55 5.48 0.61 23 1.95 0.175 0.69 0.16 3 Brodsworth 0.00 2.26 18.84 37.51 36.26 5.12 11 2.79 0.043 1.18 0.17 1 Audley End 0.03 3.59 22.70 69.93 3.67 0.00 4 0.98 0.076 <0.52 0.14 7 Walmer 1.27 18.71 31.09 20.83 25.41 2.65 8 6.23 0.130 1.23 1.14 94 British Museum 45.13 51.83 5.90 0.80 0.00 0.00 20 1.5 0.72 0.86 1.74 38 showcase

6 METALS exposure. The crystals were attached to an Onguard 3000 data logger 12 with after collected were crystals The 2005). (Baboian kinetics tarnish parabolic produced environments the of Some loggers. data 3000 The increase in mass of the crystals was measured continuously with Onguard temperature of the gold leg conductors, which are thermally connected to the cleancrystalscamerathe(0.1°C).theusedwithinerrorsofthat of the The with an Infra-red camera just after light exposure. It did not increase above clean crystal surface. The surface temperature of the crystals was measured concern as the albedo of the tarnished layer is much smaller than that of the particulara real.effectwasnotThis an and increased be tarnish may rates an apparent increase in mass. If the lighting was heating the crystals, then the temperature decreases the oscillation frequencynot fullyof thecompensate quartz for crystalincreased andtemperature induces (AnkerschmitThe temperature 2004). compensationIncreased circuits on the Onguard system are reported to layers when illuminated with visible light. tarnish containing sulfide silver for nature protective the of breakdown illustrates This House). (Apsley site one for lux 350 at rate tarnish increase in slight a with lux 500 and 350 between occur to appears effect House,Rangers House,illuminated (Apsley whenWalmer Castle). This The tarnish rates increased to close to those of clean silver for some sites, Walmer Audley End Brodsworth Rangers Apsley Site Tarnish behaviour oflight andeffect exposure onpatinas formed Table 4 exposures. all during measured rate was tarnish silver The cables. optic fibre via source halogen a from lux 1000 and 500 350, 200, 50, to exposed then were crystals Both to then was run allowed settle, for crystal 60 in minutes the near darkness. Each crystal. silver unexposed new one and crystal silver exposed one then display light sensitivity. from the less polluted sites become protective with further exposure and patinas the if determine to continued being is monitoring Onguard The these locations for three years, as part of a different project, were analysed. SEM-EDX and electrochemical stripping (Costa 2002). Coupons exposed in by characterized were environments these in formed layers tarnish The explain the increase in tarnish rate measured when illuminated. sufficientto not are layerstarnished the temperatureon in increases these temperaturecalibrationbythatindicatedreported Ankerschmit the(2004) registered temperature rises below 0.03°C. Comparison of these figures with temperature probes. The temperaturedepositedsilver film, was also measured measurements,with high precision Pt100 surface on the crystal leg connectors, Parabolic Linear Linear Parabolic Parabolic Tarnish behaviour 0.23 0.47 0.54 0.16 0.17 0 lux oftarnishrateRatio at to ofexposed new crystal crystal 0.26 0.45 0.50 0.19 0.21 50 lux 0.28 0.47 0.57 0.22 024 200 lux on silver t 0.25 0.48 0.53 0.22 0.40 350 lux Effe cts oflight 0.97 0.43 0.56 0.87 0.98 500 lux arnishing

months of of months 0.95 0.48 0.62 0.86 0.98 1000 lux

7

METALS Apsley Characterisation ofpatinas formed onthree years exposure Table 5 alfaauthor.htm. Meeting (IAQ2004), quality.air In indoor assess to crystals electric piezo plated Ankerscmid Referncs periods is unknown and requires further study. time longer for exposed when characteristics passivating develop will transmitted through fibre optics. Whether patinas from less polluted sites lamp tungsten a from light with lux 350 of excess in illuminated when lost effect. This is essentially passivating have a significant environments museum polluted relatively in formed patinas containing sulfide Silver environmental other Controlling parameters has limited potential to slow this deterioration. medals. and coins many on present Relatively low light levels have been shown to bleach the chloride patinas C tarnish the through its full depth. analyse will to process likely is stripping The it µm. 1 of silver order For the sulfur, in silver. be for for deepest shallowest and be chlorine, will by and followed elements the characterise to used X-rays of by depth the of EDX energy is the characteristic determined The of information depths two these techniques. information the different to due likely most is analyses stripping and SEM the in discrepancy The Walmer Audley End Brodsworth Rangers , A., Butts, unpublished. Museum, British Research, Scientific and Conservation of Department VI Exhibition”. Bradley,S.M. PA: Wested. International. Conshohocken, Second ASTM Baboian, Science C No.2 to CIE Publication No.15 (E-2.3.1), 1971/(TC-1.3). on uniform color spaces, color difference equations, psychometric color terms. Supplement C Inc. VanCompany Jersey: New Nostrand Princeton, 200–201. o ommission onclusions sta, V. , ed. J.H. Townsend, K. Eremin, and A. Adriaens, 88–95. London: Archetype. Archetype. London: 88–95. and A. Eremin, Adriaens, K. Townsend, J.H. ed. , 2002. Electrochemistry as a conservation tool: an overview. In R. and ed t Internationale de , B., S. W 1988. Report on the condition of silver coins in the “Greeks in Italy in “Greeks the in coins silver of condition the on Report 1988.

. 72.14 89.79 89.16 88.35 84.79 Ag SEM-EDX wt% C.D Padova, Italy, 10–12 November 2004. http://www.isac.cnr.it/iaq2004/ 2005. . C Corrosion tests and standards: application and interpretation. and Corrosion interpretation. tests and standards: application oxe, atts, 0.12 0.01 0.02 0.05 0.16 S eds and . 1967. V.C 0.31 0.06 0.01 0.03 0.23 Cl l o ’Eclaira Silver: economics, metallurgy, and use metallurgy,and economics, Silver: sta . 2004. The use of copper and silver 61 14 18 56 148 AgS Cathodic strippingnm ge. on silver t 6th Indoor Air Quality 2004 2004 Quality Air Indoor 6th 1978. Recommendations Effe 126 20 6 12 46 AgCl cts oflight arnishing Conservation Conservation 113 24 64 60 160 AgO

8 ,

METALS Dono 122–124. London: James and James. bleaching. Daniels, V., British Museum, unpublished. display on patinas medal Thickett unpublished. display Spedding, D Robiola L. Indoor corrosion of metals. Rice, Corporation. Lee, L.R., ed. J.H. Townsend, K. Eremin, and A. Adriaens, 8–11. London: Archetype Publications. In houses. historic in dust Monitoring 2002. Ho Hill, P. Museum, unpublished. British Research. Scientific and Conservation of Department 1991/8 display. on bloom L.R. Green, environment. silver by atmospheric sulfurous gases. Franey,T.P., Engineers. 1969 Amsterdam vapours. acid Thickett Miner, 96: 706–709. J. Loperfido, unpublished. display. Grzy well, wa . D 1996/11 Department of Conservation and Scientific Research. British Museum, Museum, British Research. Scientific and Conservation of Department 1996/11 1998/8 Department of Conservation and Scientific Research. British Museum, Museum, British Research. Scientific and Conservation of Department 1998/8 vaP.n, D , National Museum of Wales, personal communication, 27 June 1998. C.S., .W., Preprints of Conservation Science in the United Kingdom United the in Science Conservation of Preprints cz, and , , D D Scottish Society for Conservation and Restoration Journal D Proceedings of the Fourth International Congress on Metallic Corrosion, Corrosion, Metallic on Congress International Fourth the of Proceedings ., P.., Ll BrimblecoH. mbe, R.J. , Université de Toulouse, personal communication to Ian Macleod, 2005. 2005. Macleod, Ian to communication personal Toulouse,de Université , and . . 1991. Investigation of silver coins and medals which developed a white white a developed which medals and coins silver of Investigation 1991. and .J. 1998. 1998. Investigation of the factors affecting the alteration of silver coin and C.M.,

, ed. N.E. Hamner, 537–544. Texas: National Association of Corrosion Corrosion of Association National Texas: 537–544. Hamner, N.E. ed. , D 1998. Investigation of alteration of patina of silver coins occurring on occurring coins silver of patina of alteration of Investigation 1998. G.W. C. . Thickett ., 1974.

, W.Cappell, N.

1983. Airborne particulates: the silent nemesis. silent the particulates: Airborne 1983. I. Mcintyre. and and Kamml . 1998/15 Department of Conservation and Scientific Research. Scientific and Conservation of Department 1998/15 Dalt Air Pollution J. S J. Journal of the Electrochemical Society

C.D tringer. on. . o . S . 1996. tt tulik. 1953. Kinsolving, 1993. An apparatus for studying conservation light light conservation 1993. for studying An apparatus Corrosion Science , , 8. Oxford: Clarendon Press. and Investigation of white deposits on silver coins on on coins silver of deposits white Investigation Glycerol 1972. The corrosion of metals by organic by metals of corrosion The 1972. Proceedings of Conservation Science 2002 Science Conservation Proceedingsof T.E. 1993. Carbonyl pollutants in the museum museum the in pollutants Carbonyl 1993. oyd, Graedel. , 269 and . Frame, K. . 25: 133–134. New York: Reinhold Publishing on silver t J.J. LaskoJ.J. Effe 1985. The corrosion of of corrosion The 1985. 128: 894–895. , ed. N.H. Tennent, N.H. ed. , and cts oflight The Numismatist The arnishing 4: 16–19. B. wski. Knight 1981. 1981.

9 . ,