

Sulfur isotope analysis of cinnabar from Roman wall paintings by elemental analysis/isotope ratio mass spectrometry  tracking the origin of archaeological red pigments and their authenticity


Sulfur isotope analysis of cinnabar from Roman wall

paintings by elemental analysis/isotope ratio mass

spectrometry – tracking the origin of archaeological red

pigments and their authenticity

Jorge E. Spangenberg
1*, Jošt V. Lavrič

1y
, Nicolas Meisser

2
and Vincent Serneels

3

1Institute of Mineralogy and Geochemistry, University of Lausanne, Bâtiment Anthropole, 1015 Lausanne, Switzerland
2
Geological Museum of the Canton Vaud, Bâtiment Anthropole, 1015 Lausanne, Switzerland

3
Department of Geosciences, University of Fribourg, Pérolles, 1700 Fribourg, Switzerland

The most valuable pigment of the Roman wall paintings was the red color obtained from powdered

cinnabar (Minium Cinnabaris pigment), the red mercury sulfide (HgS), which was brought from

mercury (Hg) deposits in the Roman Empire. To address the question of whether sulfur isotope

signatures can serve as a rapid method to establish the provenance of the red pigment in Roman

frescoes, we have measured the sulfur isotope composition (d34S value in % VCDT) in samples of
wall painting from the Roman city Aventicum (Avenches, Vaud, Switzerland) and compared them

with values from cinnabar from European mercury deposits (Almadén in Spain, Idria in Slovenia,

Monte Amiata in Italy, Moschellandsberg in Germany, and Genepy in France). Our study shows that

the d34S values of cinnabar from the studied Roman wall paintings fall within or near to the

composition of Almadén cinnabar; thus, the provenance of the raw material may be deduced. This

approach may provide information on provenance and authenticity in archaeological, restoration and

forensic studies of Roman and Greek frescoes.

The finest Greek and Roman houses were frequently

decorated with mosaic floors and wall paintings (frescoes).

Studies of archaeological wall paintings often aim to

determinate the origin of the pigments and the techniques

used to apply the different colors. This is of major relevance

in cultural heritage studies because it (a) provides historical

information on early technologies (i.e., raw material proces-

sing, painting techniques) and ancient trading patterns (i.e.,

origin of raw materials, trade and commercial routes), (b)

helps detect past restoration of frescoes and authenticity of

works, and (c) gives an approximate maximal age for the

painting through the determination of the cinnabar origin.

Such studies applying different techniques – including

optical microscopy (OM), scanning-electron microscopy

(SEM), X-ray diffraction (XRD), electron microprobe

analysis (EMA), diffuse reflectance infrared Fourier trans-

form spectroscopy and gas chromatography/mass spec-

trometry (GC/MS) – were performed on Roman wall

paintings recovered from archaeological sites in France,1–4

Switzerland,5 Spain,6–10 Italy11–17 and Croatia.18 The miner-

alogical and physicochemical analyses of ancient paintings

are often difficult to interpret as they reveal only the average

composition of a mixture of materials and pigments.

The most valuable pigment used in Roman wall paintings

was the red pigment Minium Cinnabaris (called also

Vermilion), containing principally cinnabar, the red mercury

sulfide (HgS).13–16 Cinnabar was used pure for a light

red pigment or mixed with Rubrica (also called Red Ochre),

composed mainly of hematite (Fe2O3), to obtain a darker

red. Cinnabar, used all over the Roman world in small

quantities for high-quality paintings, was an expensive

raw material, since this mercury ore was not abundant

inside the borders of the Roman Empire.3 The largest

known cinnabar deposits in the Mediterranean region are

Almadén in the province of Castilla la Nueva, Spain (300 000

metric tons of Hg),19,20 Idria in Slovenia (145 000 t Hg),21,22

and Monte Amiata in Grosseto, Italy (117 707 t Hg).23,24

Epigraphic, literary and archaeological evidence corrobo-

rates ore extraction in Almadén since at least 500 BC. The

other two deposits are thought to have been mined by the

Romans, but the archaeological studies are scarce, and the

earliest records are from Christopher Columbus’ notes in

1492 and 1499. Almadén cinnabar exploitation through

the ages was two and a half times greater than at Idria

and nearly four times greater than at Monte Amiata. Other

smaller cinnabar mines that were in operation during

the Roman period include Karaburun, İzmir, and Sızma

(Konya province) in Turkey,25 Medjerda in Tunisia,

Moschellandsberg in Germany,26 and Genepy (La Mure

*Correspondence to: J. E. Spangenberg, Institute of Mineralogy and
Geochemistry, University of Lausanne, Building Anthropole,
1015 Lausanne, Switzerland.
E-mail: Jorge.Spangenberg@unil.ch
y Present address: Max Planck Institute for Biogeochemistry,
Hans-Knöll-Str. 10, 07745 Jena, Germany.

1

Published in �������	
��������
���������������
�������
������������������!��"�"��which should be cited to refer to this work.

ht
tp
://
do
c.
re
ro
.c
h


region) in France.27 The exploitation history of these

deposits is poorly known. The cinnabar ore was brought

to Rome for processing; therefore, the cinnabar used in the

wall painting of the Roman Empires may be of different

origins.

Sulfur has four stable isotopes, 32S, 33S, 34S, and 35S, with

naturally occurring abundances on earth of 95.02, 0.75, 4.21,

and 0.02%, respectively. The stable isotope composition of

sulfur is reported in the delta (d) notation as the per mil (%)
deviation of the isotope ratio relative to known standards:

d ¼ [(Rsample – Rstandard)/Rstandard] � 1000, where R is the ratio
of the heavy to light most abundant isotopes (34S/32S).

The sulfur standard is the Vienna Cañón Diablo Troilite

(VCDT).

Sulfur isotope compositions have many different appli-

cations in Earth Sciences with particular attention given to

the isotopic variations among the different sulfur phases

(sulfides and sulfates). The stable sulfur isotope ratio can

be used to identify sources, mixing processes, and the fate

of sulfur species in the environment, and particularly in

deposits of metallic sulfides.28 Knowledge of the sulfur

isotope composition of the potential sources allows their

relative contribution to the sulfur at the final site to be

assessed. The sulfur isotope composition of cinnabar in the

pigments of the Roman frescoes discovered at the House

of Diana (Crosseto, Italy) were measured and used in

combination with OM, SEM and EMA data to obtain insight

into the origin of the red pigment.13 Mazzochin et al.29

compared the isotopic composition of lead present in

cinnabar of Roman wall paintings from the Xth Regio of

the Roman Italy with that of samples from the mercury

deposits at Almadén, Monte Amiata, and Idria. In this study,

we explore the discriminating potential of the sulfur isotope

composition for tracking the provenance of cinnabar present

in Roman wall paintings from Aventicum (Avenches, Vaud,

Switzerland), the most important city of central Switzerland

during the Early Roman Empire (1st and 2nd centuries AD).

The sulfur isotope composition of the red pigments was

compared with that of cinnabar samples from major

European mercury deposits.

EXPERIMENTAL

Eight fragments of painting from the collection of the Roman

Museum of Avenches were selected for sulfur isotope study.

The surface of the fragments was cleaned of visible foreign

material with organic solvents-washed stainless steel twee-

zers. Red painting samples were collected using a micro

drill to avoid contamination with material not containing

cinnabar.

The cinnabar samples from the mercury deposits were

obtained from collections of the Department of Geology

and the Museum of Geology of the University of Lausanne.

There were 13 mercury ore samples from Almadén (Spain),

24 from Idria (Slovenia), 9 from Monte Amiata (Italy), 2 from

Moschellandsberg (Germany), and 2 from Genepy (France).

The cinnabar samples were checked for impurities under

a binocular microscope, and manually homogenized using

an agate mortar and pestle.

All archaeological pigment samples were analyzed by

powder X-ray diffraction (XRD) using a Philips1 PW 1830

diffractometer (PANalytical, Almelo, The Netherlands)

equipped with monochromated CuKa (l ¼ 1.54056 Å) X-
radiation to determine the presence of different sulfur

phases. The scan settings were 5–658 2u, 0.58 step size, 1.5-s
count time per step.

Sulfur isotope analyses were performed using a Carlo

Erba 1108 elemental analyzer (EA, Fisons Instruments,

Milan, Italy) connected to a Thermo Fisher (formerly

ThermoQuest/Finnigan, Bremen, Germany) Delta S

isotope ratio mass spectrometer that was operated in the

continuous helium flow mode via a Thermo Fisher Conflo III

split interface30. Aliquots of the sample and of the calibration

standards (200 to 600mg) were weighed in tin cups (Säntis

Analytical AG, Teufen, Switzerland). Vanadium pentoxide

was added as an oxidation catalyst in an amount approxi-

mately twice the weight of the sample. The tin cups of

the samples and the calibration standards were closed,

crushed to a small size and loaded into an AS 200

autosampler (Fisons Instruments). They were flash-com-

busted sequentially under a stream of helium and oxygen

at 10308C in a single oxidation-reduction quartz tube filled
with high purity oxidizing (tungsten trioxide, WO3) and

reducing (elemental Cu) agents, both from Säntis Analytical

AG. Combustion-derived gases (SO2, H2O, CO2, N2) were

first dried by passing them through a 10cm long column

filled with anhydrous Mg(ClO4)2, and then directed through

a 0.8m PTFE chromatographic column packed with

Porapack 50–80mesh (Fisons Instruments) at 708C for the
separation of SO2 which was isotopically analyzed by isotope

ratio mass spectrometry (IRMS). Pure SO2 gas was inserted

into the He carrier flow as pulses of reference gas. The

reference SO2 gas was calibrated against the IAEA-S-1 sulfur

isotope reference standard (Ag2S) with d
34S value of

�0.3%.31 The overall analytical reproducibility of the EA-
IRMS analyses, assessed by replicate analyses of three

laboratory standards (synthetic cinnabar, with d34S value of

þ15.5%; barium sulfate, þ12.5%; pyrite Ch, þ6.1%; pyrite E,
�7.0%) and the Aventicum cinnabar samples, is better

Table 1. Sulfur isotope composition of the cinnabar from

Roman wall paintings in Aventicum

Sample Provenance

d
34S (%, VCDT)

Averagea s

K4605 Aventicum 1 þ10.5 (2) 0.15
K4665-1 Aventicum 2 þ10.4(2) 0.13
K4665-2 Aventicum 2 þ10.2(3) 0.26
K4665-3 Aventicum 2 þ9.7 (2) 0.10
K9510/46-1 Aventicum 3 þ9.6 (3) 0.18
K9510/46-2 Aventicum 3 þ10.9 (2) 0.27
KA4117-1 Aventicum 4 þ9.6 (2) 0.21
KA4117-2 Aventicum 4 þ8.9 (2) 0.16
K4687 Aventicum 5 þ10.7 (3) 0.25
K4686 Aventicum 6 þ10.6 (2) 0.17
K9915-1 Aventicum 7 þ9.2 (3) 0.28
K9915-2 Aventicum 7 þ8.7 (3) 0.31

þ9.9 � 0.7 (12)b
a Number in parentheses stands for number of replicate analyses.
b Number of samples.

2

ht
tp
://
do
c.
re
ro
.c
h


than �0.3% (1 SD). The accuracy of the d34S analyses was
checked periodically by analyses of the international

reference materials IAEA-S-1 and IAEA-S-2 silver sulfides

(�0.3% and þ22.7 � 0.2%, respectively, values from IAEA-
Catalogue and Documents) and NBS-123 sphalerite

(þ17.09 � 0.31%, value from NIST-Catalogue and Docu-
ments).

The average sulfur isotope values for cinnabar in the

Roman wall paintings and the mercury deposits were

compared by means of t-tests using the SAS software

(version 9.1, SAS Institute Inc., Cary, NC, USA)and MatLab1

software package (version 7.2, MathWorks Inc., Natick, MA,

USA).

RESULTS AND DISCUSSION

The results of the XRD analyses have shown that the

only sulfur-containing phase in the red painting samples

from Aventicum was cinnabar. The sulfur isotope ratios

and the standard deviation (SD) of the replicate measure-

ments of the 12 cinnabar samples from the wall paintings

recovered in Aventicun, Switzerland, are presented in

Table 1. The d34S values range between þ8.7 and þ10.9%
(average � 1 SD, þ9.9 � 0.7%). Table 2 contains the
d
34S values of cinnabar ore from the Mediterranean region

mercury deposits obtained in this study. A set of 47 sulfur

isotope ratios for Almadém cinnabar was compiled from

published data32–35 and the data obtained in this study.

The unimodally distributed d34S values range from �1.6
to þ13.0% (þ6.6 � 3.7%, n ¼ 47) (Fig. 1). The mode peaks
at about þ7% (median ¼ þ7.0%) and two smaller maxima
appears near 0 and þ12%. For the Idria deposit, a set of

187 d34S values was compiled from published data36–39 and

data from this study. The distribution of these d34S values

is summarized as histogram in Fig. 2. The d34S values for

cinnabar samples range from �19.1 to þ22.8% (þ2.2 � 6.0%),
with an unimodal distribution peaking at about þ3%
(median ¼ þ2.4%). The data sets from Almadén and Idria
(Figs. 1 and 2) represent the entire spectrum of syngenetic

(e.g., formed contemporaneously with the sedimentary

host rock) and epigenetic (e.g., formed by post-depositional

processes) cinnabar-containing rock bodies. The important

Table 2. Sulfur isotope data (d34S in % VCDT) of cinnabar from mercury deposits obtained in this study

Almadén (Spain) Idria (Slovenia) Monte Amiata (Italy)

Sample d34S Sample d34S Sample d34S

MGL25234 þ6.6 MGL51355 þ8.2 MGL-Bickel �1.6
MGL40128-1 þ6.2 MGL51357 þ5.9 MGL-Bickel �1.7
MGL40128-2 þ5.4 MGL51357 þ5.1 MGL-SGAM1 þ2.3
MGL25229 þ6.3 MGL51361 þ8.5 MGL-SGAM2 þ2.4
MGL14308-1 þ8.8 MGL51361 þ9.1 MGL-NM-1 �5.0
MGL14308-2 þ8.0 MGL51366 þ3.1 MGL-ICMA �0.8
MGL51356 þ0.6 MGL30392 þ6.0 MGL-IMP þ0.1
MGL51349 þ4.6 MGL30392 þ7.7 MGL-NM-2 �7.6
MGL34986-1 þ4.4 MGL51478 þ1.2 MGL-SGAM3 þ0.9
MGL34986-2 þ5.1 MGL51478 þ1.5 �1.0 �3.2
MGL34986-3 þ4.6 MGL51478 þ6.1
MGL34986-4 þ5.4 MGL51478 þ7.0
MGL34986-5 þ7.0 MGL51639 þ0.0 Moschellandsberg (Germany)

þ5.6 � 2.0 MGL51639 �0.9 MGL3499-1 �19.6
MGL52647 þ8.9 MGL3499-2 �15.6
MGL34995 þ1.3 �17.6 �2.8
MGL34995 þ0.5
MGL34995 þ4.1
MGL34989 þ0.0 Genepy (France)
MGL34989 �0.5 MGL58789 �2 2
MGL34976 þ7.1 1 Aupt �3.5
MGL34996 þ7.6 �2.9 �0.9
MGL34981 þ1.0
JSID26 þ3.8

þ4.3 � 3.5

Figure 1. Frequency distribution of d34S values for cinnabar

from Almadén deposit, Spain. Data from Rytuba et al.,32

Saupé and Arnold,
33

Higueras et al.,
34

Jébrak et al.,
35

and

this study.

3

ht
tp
://
do
c.
re
ro
.c
h


amount of new published sulfur isotope data for cinnabar

from Almadén and Idria motivated a reevaluation of the

d
34S values of cinnabar from frescoes in the House of Diana

in Cosa, presented by Damiani et al.13

We used box plot charts, displaying the ranges, 25th

and 75th percentiles (lower and upper quartiles; Q1, Q3),

outliers, and median (50th percentile, Q2) to show the

spread of d34S values between the cinnabar of Roman wall

paintings from Aventicum (Switzerland) and Cosa (Italy)

and the Hg ore deposits (Fig. 3). The statistical significance

of the difference between the groups of d34S values was

determined using a two-tailed Student’s t-test adjusted

after checking by Fisher test whether two samples have

equal or different variances (homo- or heteroscedasticity,

Table 3). The spread of the cinnabar d34S values of the Roman

wall paintings from Aventicum are similar to those from

Cosa, and statistically different from the d34S values of

cinnabar from the Hg deposits (Fig. 3, Table 3). The

average d34S values for Aventicum (þ9.9 � 0.7%) and Cosa
(þ11.8 � 0.3%) are close to those of Almadén (þ6.6 � 3.7%)
and Idrija (þ2.2 � 6.0%), and statistically different (p<0.05)
to the other Hg deposits (Monte Amiata, Moschellandsberg,

Genepy, Izmir) (Table 3). The beginning of underground

mining only in 1490 and the absence of superficial cinnabar

exposures in Idria indicate Almadén as the source of

cinnabar used for the Aventicum wall paintings. For

any signature to be meaningful, its value must be uniform

over the dimensions of the studied artifacts and, ideally,

show only small variations on orebody and mining district

scale. Detailed studies carried out in the Almadén district

by Saupé and Arnold33 showed the variations in d34S within

mercury ore blocks to be less than 0.5% and less than 2%

Table 3. Statistical t-test comparing the mean of the d
34
S values of the cinnabar from Roman wall paintings and cinnabar from

mercury ore deposits

a
a

n
b F-testc P-valued Accepted He

Aventicum/Cosa (House of Diana) 0.05; 0.001 13 0.2131 0.000014 (0.00086) HI; HI (HI; HI)
Aventicum/Almadén 0.05; 0.001 57 7.32E-08 0.0033 HI; HO
Aventicum/Idria 0.05 197 5.18E-08 3.64E-31 HI
Aventicum/Monte Amiata 0.05 20 4.00E-05 1.21E-06 HI
Aventicum/Moschellandsberg 0.05; 0.01 12 0.0057 0.0441(3.62E-13) HI; HO (HI; HI)
Aventicum/Genepy 0.05; 0.01 12 0.5406 0.0168 (4.43E-11) HI; HO (HI; HI)
Aventicum/Izmir 0.05 23 0.0142 4.36E-18 (2.65E-16) HI (HI)
Cosa (House of Diana)/Almadén 0.05; 0.01 48 0.0092 2.49E-12 (0.0192) HI; HI (HI; HO)
Cosa (House of Diana)/Idria 0.05; 0.001 188 0.0035 7.59E-40 (0.0062) HI; HI (HI; HO)
Cosa (House of Diana)/Monte Amiata 0.05; 0.001 11 0.0122 3.86E-07 (0.000034) HI; HI (HI; HI)
Cosa (House of Diana)/Moschellandsberg 0.05; 0.01 3 0.0155 0.0424 (0.000294) HI; HO (HI; HI)
Cosa (House of Diana)/Genepy 0.05; 0.01 3 0.1448 0.0202 (0.000088) HI; HO (HI; HI)
Cosa (House of Diana)/Izmir 0.05; 0.001 5 0.3011 3.10E-06(5.36E-10) HI; HI (HI; HI)

a
a ¼ significance level.

b
n ¼ degree of freedom.

c Comparison of variance by Fisher test; F>1: equal variance (homoscedasticity); F<1: unequal variance (heteroscedasticity).
d Probability; for F values relatively close to 1 a second t test was performed assuming F>1 (homoscedasticity) and the P-values given in
parentheses.
e Hypothesis: HO ¼ equal mean, HI ¼ different mean; For P>a HO is accepted.

Figure 3. Box plot of d34S values for the cinnabar from the

Roman paintings at Aventicum (Switzerland) and Cosa

(Italy)13 and European Hg deposits, displaying the

ranges, 25th (1st quartile, Q1) and 75th (3rd quartile, Q3)

percentiles, median, and outliers.

Figure 2. Frequency distribution of d34S values for cinnabar

from Idria deposit, Slovenia. Data from Drovenik et al.,36

Drovenik et al.,37 Lavrič and Spangenberg,38 Palinkaš

et al.,39 and this study.

4

ht
tp
://
do
c.
re
ro
.c
h


within an outcrop of most orebodies. However, the

d
34S average values between cinnabar orebodies in Almadén

district vary between þ0.2 and þ13.6%. Thus, the slight
difference between the d34S average values from Aventicum

and from Cosa could be explained by cinnabar coming

from different Hg-mineralized bodies in the Almadén mine.

An additional possible source of heterogeneity is the fact

that the cinnabar ore was brought to Rome for processing.

The average sulfur isotope composition of the two groups

of cinnabar from Roman wall paintings (Aventicum in

Switzerland, Cosa in Italy) cannot be statistically differ-

entiated. The identification of not local (exotic) sources

for cinnabar found at Aventicum adds substantially to our

understanding of regional interaction and trade during

the Roman period.

The results presented in this study indicate that the

sulfur isotope composition provides further insights on the

origin and authenticity of the red pigment produced from

cinnabar in archaeological paintings. This approach may

have important implications for archaeological, restoration

and forensic studies. Very small amounts of sulfide are

needed (down to 100 mg, 1mmol) for sulfur isotope analysis

by EA/IRMS, which is relatively fast and inexpensive, and

permits a high throughput of samples.

Acknowledgements
We thank Anne Hochuli-Gysel, Director of the Musée

Romain d’Avenches, and Michel Fuchs from the Sevice

ArchéologiqueCantonaldeFribourg, forthekindpermission

to take samples of Roman paintings from the museum collec-

tion. This study was supported by the Swiss National Science

Foundation and the University of Lausanne.

REFERENCES

1. Delamare F, Darque-Ceretti E, Dietrich JE. Revue d’Arché-
ométrie 1982; 6: 71.

2. Frizot M. Revue d’Archéométrie 1982; 6: 47.
3. Barbet A. Pigments et colorants de l’antiquite et duMoyen Age.

Colloque International du CNRS, Laboratoire de Recherche
de Musées de France: Paris, 1990; 255–271.

4. Guichard V, Guineau B. Pigments et colorants de l’antiquité et
duMoyenAge. Colloque International du CNRS, Laboratoire
de Recherche de Musées de France: Paris, 1990; 245–254.

5. Bèarat H. Archaeometry 1996; 38: 81.
6. Doménech-Carbo MI, Bosch-Reig F, Gimeno-Adelantadob

JV, Periz-Martinez V. Anal. Chim. Acta 1996; 330: 207.

7. Edreira MC, Feliu MJ, Fernandez-Lorenzo C, Martin J. Anal.
Chim. Acta 2001; 434: 331.

8. Edreira MC, Feliu MJ, Fernandez-Lorenzo C, Martin J.
Talanta 2003; 59: 1117.

9. Salvado N, Buti S, Tobin MJ, Pantos E, Prag AJNW, Pradell
T. Anal. Chem. 2005; 77: 3444.

10. Franquelo ML, Duran A, Herrera LK, Jimenez de Haro MC,
Perez-Rodriguez JL. J. Mol. Struct. 2009; 924/926: 404.

11. Mirti P, Appolonia L, Casoli A, Ferrari RP, Laurenti E,
Amisano Canesi A, Chiari G. Spectrochim. Acta 1995; 51A:
437.

12. Colombini MP, Modugno F, Giacomelli A. J. Chromatogr. A
1999; 846: 101.

13. Damiani D, Gliozzo E, Memmi-Turbanti I, Spangenberg JE.
Archaeometry 2003; 45: 341.

14. Mazzocchin GA, Agnoli F, Mazzocchin S, Colpo I. Talanta
2003; 61: 565.

15. Mazzocchin GA, Agnoli F, Salvadori M.Talanta2004; 64: 732.
16. Duran A, Jimenez de Haro MC, Perez-Rodriguez JL, Fran-

quelo ML, Herrera LK, Justo A. Archaeometry 2010; 52: 286.
17. Duran A, Perez-Maqueda LA, Poyato J, Perez-Rodriguez JL.
J. Therm. Anal. Calor. 2010; 9961: 803.

18. Mazzocchin GA, Rudello D, Makaković N, Marić I.AnnaliDi
Chimica 2007; 97: 655.

19. Saupé F. Econ. Geol. 1990; 85: 482.
20. Hernandez A, Jébrak M, Higueras P, Oyarzum R, Morata D,

Minha J. Mineral. Deposita 1999; 34: 539.
21. Mlakar I, Drovenik M. Geologija (Ljubljana) 1971; 14: 67.
22. Mlakar I. Idrijski Razgledi 1974; 3–4(XIX): 1.
23. Tanelli G. Mem. Soc. Geol. Ital. 1983; 25: 91.
24. Dini A, Benvenuti M, Costagliola P, Lattanzi P.OreGeol.Rev.

2001; 18: 149.
25. Akcay M, Ozkan HM, Moon CJ, Spiro B. OreGeol. Rev. 2006;

29: 19–51.
26. Krupp RE. Mineral. Deposita 1989; 24: 69.
27. Sarrot-Reynauld J. Trav. Lab. Géol. Grenoble 1957; 135.
28. Ohmoto H, Goldhaber MB. Geochemistry ofHydrothermalOre
Deposits. John Wiley: New York, 1997; 517–611.

29. Mazzocchin GA, Baraldi P, Barbante C. Talanta 2008; 74:
690.

30. Giesemann A, Jager HJ, Norman AL, Krouse HP, Brand WA.
Anal. Chem. 1994; 66: 2816.

31. Coplen TB, Krouse HR. Nature 1998; 392: 32.
32. Rytuba JJ, Rye RO, Hernandez AM, Dean JA, Arribas A.28th
International Geologic Congress Abstracts with Program.
Washington, 1989; 2741.

33. Saupé F, Arnold M. Geochim.Cosmochim.Acta 1992; 56: 3765.
34. Higueras P, Oyarzun R, Lunar R, Sierra J, Parras J. Mineral.
Deposita 1999; 34: 211.

35. Jébrak M, Higueras PL, Marcoux E, Lorenzo S. Mineral.
Deposita 2002; 37: 421.

36. Drovenik M, Duhovnik J, Pezdić J. Verh. Geol. B-A 1978; 3:
301.

37. Drovenik M, Dolenec T, Režum B, Pezdić J. Geologija (Ljubl-
jana) (in Slovene) 1990; 33: 397.

38. Lavrić LV, Spangenberg JE. Mineral. Deposita 2003; 38: 886.
39. Palinkaš L, Strmić S, Spangenberg JE, Prochaska W, Herlec

U. Swiss Bull. Miner. Petrol. 2004; 84: 173.

5

ht
tp
://
do
c.
re
ro
.c
h