Chapter 4

GOLD ANALYSIS: FROM FIRE ASSAY TO SPECTROSCOPY - A REVIEW

H.G. BACHMANN
Wildaustr. 5
63457 Hanau
Germany

ABSTRACT. Analytical methods and techniques for the determination of gold contents are employed in
the exploration and evaluation of gold deposits, grade control in primary and secondary production, the
assaying and hallmarking of gold and its alloys in jewellery, and in the study of coins, archaeological
finds etc. and in provenance studies.
The methods range from those known since antiquity (e.g. fire assay, touchstones, specific gravity
measurements) to modern spectroscopic (emission, absorption, induced coupled plasma etc.) and non-
destructive fluorescence (X- and gamma-rays in the widest sense) and radioactive (neutron activation)
techniques. The review includes an extensive list of references.

1. Introduction

The analysis of gold from ores to artifacts requires a variety of techniques unique to this
paramount precious metal. In terms of concentration, several orders of magnitude have to be
bridged. Obviously, no general method is capable of covering this whole field. Thus any re-
view has first to define the various analytical applications. Apart from wide concentration
ranges, the determination of gold contents - especially in objects of archaeological value - is
only too often restricted to non-destructive methods. This additional requirement adds a fur-
ther dimension (and restriction) to the list of available methods.

Any discussion and review of analytical techniques for the determination of precious me-
tals - and gold in particular - has to differentiate between the following fields of application:

- Techniques used in exploration and evaluation of gold deposits,

- Analytical methods required in control of primary and secondary gold production,
Characterisation of high-grade gold and its alloys, including provenance studies.

Following this sequence, the paper will discuss the methods specific to the various fields of
application.

2. Exploration and evaluation of gold deposits

The concentration range likely to be encountered in these applications extends from ca. 0.05
ppm to above 50 ppm (I ppm = 1 g/t).
303
G. Morteani and 1. P. Northover (eds.i. Prehistoric Gold in Europe. 303-315.
© 1995 Kluwer Academic Publishers.
304

2.1 FIELD METHODS

As kown by every prospector looking for placer gold, the panning dish or batea is a powerful
analytical tool, provided the operator has sufficient experience not to lose any material during
the concentration process. If sampling is based on a defined volume or weight of material
(e.g. using buckets of known volume), quite reliable quantitative gold determinations are pos-
sible in the field. Where the panning operation can be carried on until "nuggets" (normally
only tiny grains) are isolated from associated heavy mineral particles, a so-called grain-size
chart is a very useful aid. It helps to find the relation between particle size and weight. The
following six classes (in miner's terminology colours) of grains, recommended by Utter
(pers. communication) are (Table 1):

Table 1. Grain size table giving the relation between gold particle size and weight.
Group or Colour Size (in=) Average Weight (in mg)

o 0.062 - 0.125 0.02

1 0.125 - 0.25 0.09
2 0.25 - 0.5 0.98
3 0.5 - 0.8 2.54
4 0.8 - 1.0 8.18
5 1.0 16.45

After completion of the panning operation, sorting and counting the gold grains, the final
step in grade determination is performed as illustrated by the example in Table 2. Assuming
a total sample weight (prior to panning) of 8.8 kg, the following gold grains are supposed to
have been isolated:
Table 2. Example for a calculation of the total gold content of a sample from colour and number of
grains.
Colours Number of Grains Total Weight

o 8 8 x 0.02 = 0.16
1 1 1 x 0.09 = 0.09
2 1 1 x 0.98 = 0.98
4 4 4 x 8.18 = 32.72
Sum = 33.95

The deposit thus sampled and evaluated has a gold content of: 33.95 mg/ 8.8 kg = 3.85
mg/kg = 3.85 g/t.

Another method of determining the gold content in bulk samples is the treatment of the
concentrate with mercury. The resulting gold amalgam (more correctly gold/silver amalgam)
is decomposed by heat, preferably by distillation to recover the mercury and to avoid the
highly toxic effects of mercury vapour. Purification of the recycled mercury is carried out in
the field by dissolving the mercury in nitric acid and precipitating the metallic mercury from
its nitrate solution by inserting copper rods, as witnessed by the author in Brazilian gold ex-
ploration camps.
 305

The modern method of BLEG-sampling (Bulk Leach Extractable

Gold) uses soil samples of 2 to 5 kg which, after leaching with cya-
nides, enable the prospector to determine the gold contents in the
I'
5C'T J.frf$ pregnant solutions by spectroscopic methods, e.g. atomic absorption
¥g 42T
J,1J7 spectroscopy (AAS). The method is very sensitive and able to detect
f{7 z,&!J gold anomalies well down into the ppb-range. As many samples are
q6 2.7V normally analysed during an area survey, the accuracy of the single
q5 2..0
qq 2.J7 determination needs not to be very high. Therefore, simple equipment
q~ t,zt which can be operated in a field-site laboratory is quite adequate
q;e 4~
'If 1,9f
(Lehne: pers. communication).
'IQ' {76'
J9 1.65 During the last, and until the beginning of the present century, a
.11 1,SZ
.J7 l,'ff knowledge of blow-pipe techniques was mandatory for mineralogists
.16 f!'9 and prospectors. With only a few pieces of equipment and some basic
.35 ~t:9
chemicals qualitative and even reliable quantitative determinations
3'1 ~
s:J 'f&I7 could be carried out under virtually any conditions everywhere. The
.1Z 4!Jf subject was a major teaching topic at many mining academies with
", dIZ
.Jt!
Freiberg/Saxony taking the lead, where C.F. Plattner (1835) pu-
blished his famous book, which went through numerous editions and
was based on several earlier publications by Berzelius and especially
Harkort. However, Plattner's "Probierkunst mit dem LOthrohre" co-
vered a much wider tield of applications and was based on extensive
personal experience. His quantitative analysis of gold ores in the tield
was a short-cut, fire assay technique (cf. below), producing tiny gold
beads which were often too small to be weighed (on a small collapsi-
ble field kit balance); they were measured with a special ruler (Fig. 1)
which enabled direct correlation of bead diameters with weight; (cf.
Krug 1914).

Plattner's book had many successors, e.g. Frick and Dausch (1932)
and Henglein (1962). Hardly any books have appeared on this subject
in recent years. During the era of spectacular gold discoveries in the
New World, this useful method was not only propagated, but even
extended in the USA (Brush 1890).

2.2 LABORATORY METHODS

Turning from analytical tield techniques to laboratory methods, the

foremost place is taken by tire assay. Known, perhaps, since the 4th
millenium B.C., this versatile procedure - essentially a laboratory
scale pyrometallurgical extraction of gold with lead as collector, sub-
sequent oxidation of the resulting lead-bullion button by cupellation,

Fig. 1: Ruler used to determine the weight of gold beads. From Plattner (1835).
306

and isolation of a bead with all the precious metals originally contained in the sample - is still
used worldwide for the analysis of high- and low-grade materials. In connection with ex-
ploration and evaluation, gold contents below even 0.5 glt in waste dumps etc. are just as
amenable to this method as all types of primary ore.

The history and development of fire assay through the ages is a subject which would merit
a monograph of its own. We not only have detailed descriptions of this technique from such
renowned authors as Biringuccio (1540) and Agricola (1556), but many less well known
writers have also given prominence to the subject and have made contributions; cf. the im-
pressive list of early books on assay methods, including many editions of the so-called
"Probierbiichlein" (Booklet on Assaying), collected by Annan (1960). Good reviews of early
assay practices applied to precious metals were given by Oddy (1983) and Nriagu (1985).

Sampling of material from surveys, exploration campaigns and producing gold mines for
fire assay requires practices set out in relevant guide-lines, e.g. by Liischow (1980, 1989). A
representative sample of gold ore is one which attempts to eliminate the "nugget effect". If,
for example, a sample is supposed to contain the metal in a concentration of ca. 4 glt, the
gold is most probably not distributed homogeneously in the rock or sediment matrix. No
matter how small, the particles in gold-ore specimens have a definite size. Thus, the larger
the sample weight, the more representative it is with regard to the statistical distribution of
grain size groups (or colours). However, for routine analysis a sample weight has to be cho-
sen which can be handled with respect to crucible size, laboratory furnace capacity, time of
processing etc.). A sample weight of 25 g is a good compromise (Bachmann 1992). With the
hypothetical gold concentration mentioned above (4 g/t), fire assay will give a bead con-
taining 0.1 mg of gold (+ silver + platinum group metals, if present). This is a bead weight
which can still be weighed on a normal analytical balance, though a micro-balance is some-
times preferable. Another method to obtain beads which can be weighed with sufficient accu-
racy - particularly when the gold content is unknown - is the intentional addition of silver to
the assay charge.

Fire assay by lead collection and followed by cupellation gives a metallic bead containing
all the precious metals contained in the sample analysed; gold and silver with high accuracy
and with some restriction most of the platinum group metals as well. If gold and silver have
to be determined, they have to be separated from each other by nitric acid, leaving the gold
unaffected and bringing the silver into solution in which it can be analysed by conventional
methods (titration etc.); the gold residue can be weighed. Complete solution of the cupella-
tion bead - preferably after being milled into a thin strip to increase its surface area - is the
mandatory preparation for most spectroscopic methods, e.g. atomic absorption spectroscopy
(AAS).
The numerous text- and handbooks on gold analysis cover many analytical procedures as
derived from practical experience. Emphasis can either be placed on the determinations of
low gold concentrations, e.g. in ores, sweeps (i.e. gold-containing non-metallic materials,
such as dust, slags, slurries), or on richer materials. Virtually every aspect of assay tech-
niques is covered in the comprehensive, critical and up-to-date books by ChemikerausschuB
(1964), Beamish (1966), Beamish and Van Loon (1972, 1977), Lenahan and de Murray-
Smith (1986) and Stoch (1986).
 307

Though fire assay still counts as a universal and reliable analytical method - particularly in
the analysis of low-grade materials - it has occasionally received critical comments, recently
by von Hase (1989) and Gasparrini (1993). Minor systematic errors are still under investi-
gation. Small losses through volatilisation during smelting andlor cupellation have always
been taken into account by renowned laboratories. A laboratory manual, issued 80 years ago
by the Deutsche Gold- und Silberscheide-Anstalt (1914), gives sets of data for Kupellenzug
(i.e. cupellation losses) - varying with the gold-silver ratio in the samples to be analysed -
which have to be added to the final results.
An interesting, highly sensitive combination of precipitation of gold with lead sulphide,
followed by cupellation and subsequent fusion of the bead in borax to remove any associated
silver was published by Haber (1927) in his paper on the sensational proposal to repay the
German war debts after World War I by the recovery of gold from seawater. Haber (1927)
claims a high accuracy for his method with a limit of detection near 1 x 10- 10 g of gold.
Low gold concentrations in solutions can also be determined by colorimetry of intensively
coloured gold compounds, like Purple of Cassius, gold complexed with hydrazine sulfate etc.
(v.Philipsborn 1953, Haber 1927).

3. Control in primary and secondary gold production and recovery

With gold contents from below 5 to above 5000 glt, the analytical methods employed by gold
mines (primary production) and recycling plants (secondary production) are not very different
from those already dealt with in the previous sections, especially where laboratory methods
are concerned.
The South African Chamber of Mines has developed a portable, battery-operated X-ray
fluorescence (XRF) gold analyser (Jorn 1993) equipped with a radioactive cadmium source.
This instrument is capable of detecting gold contents in ores as low as 1 glt at a distance of 2
to 6 centimetres from the rock face with a measuring time of about 1 minute. The penetration
depth of the emitted gamma-rays is 25 millimetres. The instrument is used for routine grade
control in underground mines.
A further improvement in "on site" gold analysis was recently reported also from South
Africa (Anonymous 1993). Detection limit for this advanced XRF-spectrometer, named
"Goldstream" and introduced by the Western Deep Levels Mine near Johannesburg for rou-
tine production control, is as low as 0.05 gil. This, however, can only be achieved with
measuring times of 30 minutes. No further technical details of this portable instrument are as
yet known.

4. Gold and gold alloys

With a concentration range of ca. 10 to 100 % gold, the analytical methods for metallic
materials have to be divided into destructive and non-destructive techniques.

4.1 DESTRUCTIVE METHODS

Here, too, fire assay is the most prominent technique. Metallic samples, such as ingots,
semifinished products, coin blanks and minted coins, metal artifacts, gilded surfaces etc. are
308

normally high in gold. Here the strength of fire assay lies not so much in the concentration of
low gold contents in the collection step, but more in the elegant and efficient removal of im-
purities, e.g. base metals. The importance and reliability of fire assay was convincingly
pointed out by Evans (1992), Deputy Warden of the London Assay Office, which is part of
the Goldsmiths' Company. In the United Kingdom, fire assay has been in use for hallmarking
objects made of precious metals since 1238. For accurate determinations sample weights of
only 50 to 250 mg are recommended. According to Evans (1992):
"Fire assay is the only method capable of meeting UK legal requirements for gold. "
However, fire assay is not non-destructive, though in classical assaying, the gold prill,
isolated by the assaying processes (smelting or fusion with lead and cupellation) is physically
present after weighing, i.e. no loss in precious metals is incurred. Spectroscopic methods,
which require very little material (in most cases a few milligrams of homogeneous material)
have augmented or even substituted fire assay in many laboratories. But even the removal of
samples of only a few milligrams in weight is often not acceptable when valuable objects
have to be characterised. In these cases, absolutely non-destructive approaches have to be ta-
ken. It is, therefore, a question of where to draw the line between destructive and non-de-
structive methods: a better classification criterion would be: Methods without loss of sample
material, no matter how small, and others.
Spectroscopic analyses are very sensitive and thousands of museum objects have been
analysed by emission spectroscopy and other spectroscopic methods. The removal of tiny
amounts of material was generally acceptable. The most impressive summary of analytical
data on archaeological gold objects is listed in the book on prehistoric gold finds from Eu-
rope by Hartmann (1970). Further studies along this line include investig!ltion of a pre-Hi-
spanic gold chisel by AAS, microprobe, microhardness and metallographic examination by
Scott and Seeley (1983). A paper by Otto (1939) stands out as perhaps the first critical ex-
amination of prehistoric gold objects from Germany by ESA.
Sampling of precious museum objects for analysis requires precaution; some of its aspects
are discussed by Pernicka (1989). Preparation of solutions of gold samples for those analyti-
cal methods requiring liquids, like AAS and ICP, was described by Scott (1983).
Though wet chemical analyses of gold coins are rarely carried out these days, it is of histo-
ric interest to appreciate the skill with which an elaborate procedure for coin analysis was
developed and performed at the beginning of this century by the former German Technische
Reichsanstalt (Mylius 1911).
Probably even older than fire assay is the use of touchstones for the identification of gold
and its alloys. The comparison of standards with unknowns is the basic principle of working
with touchstones: The trace of a streak of a gold alloy of unknown composition on the
smooth surface of a black siliceous schist - i.e. a touchstone - is very characteristic of the
particular alloy producing this streak. It can either be compared with streaks of standards or
attacked with acids of varying concentrations. A weak acid will remove low carat gold alloys,
while high-gold alloys are more resistant. Skilled operators can thus distinguish between a
large number of jeweller's alloys quite easily. The papers by Ahlberg et al. (1976a, 1976b),
Eluere (1985, 1986) and by Moore and Oddy (1985) give good summaries and describe very
early examples of touchstone application.
A method of analysing streaks from samples by microchemical methods was developed and
advertised not only for nearly all the base metals, but also for precious metals as well by
 309

Ballczo and Mauterer (1979). However, this approach has failed to establish itself as an alter-
native to other techniques.

4.2 NON-DESTRUCTIVE METHODS

The answer to the question, "Who performed the first truly non-destructive precious metals'
analysis?" is Archimedes, who investigated the crown of Hieron II. of Syrakus (306 - 213
B.C.). The loss of weight observed after immersion of the gold/silver crown in water could
be directly related to the alloys's composition. It is the merit of Oddy (1972a, 1972b) and his
coworkers (Oddy and Hughes 1972, Oddy and Schweizer 1972, Oddy and Blackshaw 1974,
Oddy and Mumo-Hay 1980) that the specific gravity method was critically assessed, compa-
red with other methods, and tested on series of coins. The specific gravity method is used to
an increasing extent to augment other non-destructive approaches, especially XRF, in coin
characterisation. Though problematic in its application to ternary alloys, it is a reliable and
informative method when binary alloys of gold with silver or copper have to be dealt with.
Modern electronic balances and the use of liquids with high density, e.g. perfluordecaline,
C lO F 18 , d = 1.942 gcm- 3 , increase the accuracy of the specific gravity method significantly.
The most versatile method in non-destructive precious metals analysis is X-ray fluo-
rescence spectroscopy (XRF), which has as its main disadvantage the very low depth of pe-
netration of X-rays into the sample matrix. In other words, exact analytical data obtained by
XRF are restricted to surface layers often only a few micrometers thick. If this limitation is
borne in mind, and if calibration is based on a sufficiently large number of accurate standards
in combination with physico-mathematical interelement corrections, XRF is unsurpassed in
reliability, speed and specimen protection. XRF and NAA (neutron activation analysis) as
strictly non-destructive methods have been employed in many research projects on gold
coins, jewellery etc.; either alone or in combination with other techniques to verify or to cri-
tically assess the XRF-data.
Papers published during the last twenty years of non-destructive analytical methods
include:

- Araujo et al. (1993): Comparison of energy-dispersive XRF- with PIXE (proton-induced

X-rayemission)-results,
- Bodenstedt (1976, 1981): Studies of electron coins from Phokaia and Mytilene, employing
XRF, NAA, specific gravity determinations and radiography (to check coins for voids and
inclusions) ,
- Capello et al. (1973): Use of NAA to determine gold, silver and copper in ancient and
rare coins,
- Cesareo and von Hase (1973): Radio-isotope XRF of ca. one hundred Etruscan gold ob-
jects from burials,
- Coleman and Wilson (1972): NAA of Merovingian gold coins from the Sutton Hoo ship
burial,
- Cope (1972): NAA and mass spectrometry of a Roman aureus,
- Cowell (1977): Energy-dispersive XRF of Celtic and late Roman coins,
- Darling and Healy (1971): Study of Greek gold-silver-copper alloys by XRF, radiography,
hardness tests, microprobe etc.,
310

- Demortier (1984a, 1984b, 1984c, 1986, 1989, 1992a, 1992b) and Demortier and Hackens
(1982): Numerous studies of ancient jewellery, brazing alloys, copper-gold-cadmium sol-
ders etc., mostly by PIXE,
- Fabris and Treloar (1980): Comparison of XRF- and AAS-analyses of gold objects from
Sarawak/Malaysia,
- Ferreira and Gil (1981): PIXE of 18th to 19th century gold coins,
- Gillies and Urch (1985): Investigation of a pre-Columbian tumbago pectoral disk
(tumbago = gold-base metal alloy) by XPS (X-ray photoelectronic spectroscopy),
- Gordus and Gordus (1974 and 1980): NAA of gold impurities in silver coins and art ob-
jects from several eras and many countries,
- Klockenklimper et al. (1990): XRF of German gold coins from the late 19th to the early
20th century,
- Kowalski and Reimers (1971, 1972): NAA, XRF and specific gravity measurements of
medieval gold coins,
- Love et al. (1980): XRF, SEM (scanning electron microscopy) and scanning electron Au-
ger spectroscopy of ancient gold coins and their modern copies,
- Mommsen and Schmittinger (1981): High-energy PIXE employed for analysis of ancient
gold and silver coins,
- Oddy and La Nice (1986): XRF of Byzantine gold coins and jewellery,
- Piette et al. (1986): PIXE of gold solders, brazing alloys and various artifacts,
- Radcliffe et al. (1980): Coin analysis by differential absorption of gamma-rays,
- Reimers et al. (1977): Gamma-ray activation analysis and its advantage compared with
NAA for the analysis of precious metal objects, including coins,
- Schweizer and Friedmann (1972): Comparison of NAA, XRF and wet chemical methods
for determination of gold and silver contents in coins,
- Voute (1985 and 1991): XRF of Celtic gold coins,
- Wilde and Paschmann (1989): Microprobe analysis of ancient Indian gold coins,
- Ziegaus (1991): XRF of Celtic gold coins.

All the papers cited refer to laboratory examinations. Modern concepts and miniaturisation
of components have resulted in the design and development of mobile XRF-units. These can
be brought to museums and art collections rather than vice versa, thereby eliminating the
worries and refusals of curators etc. who have strong objections to seeing valuable objects
removed from their collections. Results of an on-site analytical programme, including inve-
stigation of Roman gold artifacts, have recently been published (Bachmann 1993).

5. Provenance studies

Analyses of gold artifacts have occasionally been aimed at localising the source of gold from
which the object was made. These comprise, for instance, the analysis of platinum-group ele-
ment inclusions. These data have rarely been conclusive, as shown by Meeks and Tite
(1980). If trace elements are characteristic of types of gold from certain deposits, NAA is
perhaps the most appropriate method to apply.
Recently, a new, highly sensitive and accurate method of gold fingerprinting by "laser
ablation inductively coupled mass spectrometry" was developed by John Watling. So far,
 311

only press releases and reports by science journalists (Gooding 1993) give some information
about this technique. A detailed publication is still lacking. Watling was successful in esta-
blishing the exact sources of gold stolen in Australia. He predicts that the precise trace ele-
ment profile he is able to establish for any type of gold will also enable him to identify gold
sources of early coinage and thereby get a more accurate picture of human movement, trade
routes etc.

6. Outlook

Electrical conductivity and ultrasonic probes are modern developments that give accurate re-
sults when is comes to distinguishing between different metals and/or alloys, provided the
measurements are based on calibration with standards of known composition. These methods
have so far only been employed occasionally to detect frauds and falsifications, but they may
eventually be useful in artifact analysis as well.
New methods of gold analysis will almost certainly be introduced in the future. Though
each new analytical approach widens the range and offers new possibilities, they will, ne-
vertheless, only be additions to established and irreplacable methods, such as fire assay, the
most ancient, and still most universal technique in many fields of gold evaluation.

Acknowledgement
The assistance of Dr. Christiane Lutz in retrieving references on gold analysis from the
Chemical Abstracts' Data Base and her help in providing hard to obtain literature is appre-
ciated with sincere thanks.

7. References

Agricola, G. (1556): De Re Metallica.- English translation with comments by H.C. and L.H.
Hoover, London 1912 (new edition, Dover, New York, 1950): 219-265 (Book VII).
Ahlberg, M., Akselsson, R., Forkman, B., Rausing, G. (1976): Gold traces on wedge-shaped
artifacts from the late Neolithic of Southern Scandinavia analysed by proton-induced X-
ray emission spectroscopy. Part I. The wedge-shaped pendants of the Scandinavian
Neolithic.- Archaeometry 18 (1): 39-42.
Ahlberg, M., Akselsson, R., Forkman, B. (1976): Gold traces on wedge-shaped artifacts
from the late Neolithic of Southern Scandinavia analysed by proton-induced X-ray emis-
sion spectroscopy. Part II. Analysis of wedge-shaped pendants by PIXE.- Archaeometry
18 (1): 42-49.
Annan, R. (1960): Historic books on mining and kindred subjects.- London, Science Mu-
seum Book Exhibitions No.IV: 18-25.
Anonymous (1993): Goldgestein genauer nachweisbar.- Frankfurter Allgemeine Zeitung,
8.9.93.
Araujo, M.F., Alves, L.C., Cabral, I.M.P. (1993): Comparison of EDXRF and PIXE in the
analysis of ancient gold coins.- Nucl. Instrum. Methods Phys. Res., Sect. B, B 75 (1-4):
450-453.
312

Bachmann, H.-G. (1992): Es ist nicht alles Gold .... - Extra Lapis Nr.2, Miinchen: 86-89.
Bachmann, H.-G. (1993): Analyse ausgewiihlter Metallartefakte.- In: Kellner, H.-J., Zahl-
haas, G. (eds.): Der Romische Tempelschatz von Weissenburg i. Bay., Zabern, Mainz:
147-159.
Ballczo, H., Mauterer, R. (1979): Nondestructive ultramicroanalysis of archaeological ob-
jects. 1. Complete analysis of streak samples of antique metal artifacts (ca. 100 micro-
gram). C. Gold, silver and "electron".- Fresenius' Z. Anal. Chern., 299: 46-48.
Beamish, F.E. (1966): The analytical chemistry of noble metals.- Academic Press Oxford,
London, Edinburgh, New York, Toronto, Paris, Frankfurt a.M.
Beamish, F.E., Van Loon, J.C. (1972): Recent advances in the analytical chemistry of the
noble metals.- Academic Press Oxford, New York, Toronto, Sydney, Braunschweig.
Beamish, F.E., Van Loon, J.C. (1977): Analysis of noble metals; overview and selected me-
thods.- Academic Press New York, London, Toronto, Sydney, San Francisco.
Biringuccio, V. (1540): De la pirotechnia.- English tranlation by C.S. Smith, M.T. Gundi,
M.LT. Press New York 1942: 188-205.
Bodenstedt, F. (1976): Phokaisches Elektron-Geld von 600 - 326 v. Chr.- Zabern, Mainz:
32-41,51-53, 116, 133-162.
Bodenstedt, F. (1981): Die Elektronmiinzen von Phokaia und Mytilene.- Wasmuth Tiibingen:
313.
Brush, G.J. (1890): Manual of determinative mineralogy with an introduction on blow-pipe
analysis.- 12th edition, rev. and corr., 1. Wiley New York.
Capello, C., Katz, S.A., Padilla, L., Williams, C. (1973): Determination of gold, copper,
and silver in ancient and rare coins by neutron activation analysis with Californiurn-
252.- Bull. N.Y.Acad.Sci., 18 (2): 30-32.
Cesareo, R., von Hase, F.W. (1973): Nondestructive radioisotope XRF analysis of early
Etruscan gold objects.- Kerntechnik 15 (12): 565-569.
ChemikerausschuB der Gesellschaft Deutsche Metallhiitten- und Bergleute e.V (1964):
Edelmetall-Analyse; Probierkunde und maJ3analytische Verfahren.- Springer Berlin,
Gottingen, Heidelberg.
Coleman, R.F., Wilson, A. (1972): Activation analysis of Merovingian gold coins.-
Spec.Publ.- R.Numis.Soc. 8: 88-92.
Cope, L.H. (1972): Complete analysis of a gold aureus by chemical and mass spectrometric
techniques.- Spec.Publ.- R.Numis. Soc. 8: 307-313.
Cowell, M. (1977): Energy dispersive X-ray fluorescence analysis of ancient gold alloys.-
PACT (Rixensart, Belg.), 1: 76-85.
Darling, A.S., Healy, J.F. (1971): Microprobe analysis and the study of Greek gold-silver-
copper alloys.- Nature 231: 443-444.
Demortier, G., Hackens, T. (1982): Milliprobe and microprobe analysis of gold items of an-
cientjewelry.- Nucl. Instrum. Methods Phys.Res., 197 (1): 223-236.
Demortier, G. (1984a): Milli- and microprobe PIXE for the analysis of ancient brazing of
gold.- Microbeam Anal., 19th: 249-252.
Demortier, G. (1984b): Contribution of physical methods in the analysis of ancient gold and
silver jewelry artifacts.- Rev.Quest-Sci., 155 (1): 19-40.
Demortier, G. (1984c): Analysis of gold jewelry artifacts. Characterisation of ancient gold
solders by PIXE.- Gold Bull., 17 (1): 27-38.
 313

Demortier, G. (1986): LARN experience in nondestructive analysis of gold artifacts.-

Nucl.Instrum.Methods Phys.Res., Sect. B, B 14 (1): 152-155.
Demortier, G. (1989): Accelerator-based spectroscopic techniques for analysis of archaeolo-
gical gold jewelry.- Spectroscopy (Eugene, Oreg.), 4 (6): 35-40.
Demortier, G. (1992a): Elemental analysis of gold jewelry.- Phys. unserer Zeit, 23 (1): 13-
21.
Demortier, G. (I 992b): Ion beam analysis of gold jewelry.- Nucl. Instrum. Methods Phys.
Res., Sect. B, B 64 (1-4): 481-487.
Deutsche Gold- und Silberscheide-Anstalt vorm. Roessler (1914): Probier-Ordnung fur
Scheidgutplanschen, Riickstande, SchHimme und Gekratze.- Private Printing, Frankfurt
am Main: 6 pp.
Eluere, C. (1985): Attention aux pierres de touche! - Bull. Soc. Prehist. Franc., 82: 203-206.
Eluere, C. (1986): A prehistoric touchstone from France.- Gold Bull. 19 (2): 58-61.
Evans, D.W. (1992): Assaying and hallmarking.- Interdisc. Science Rev., 17 (4): 339-342.
Fabris, G.J., Treloar, F.E. (1980): X-ray fluorescence and atomic absorption analysis of Sa-
rawak gold artifacts.- Archaeometry, 22 (1): 93-98.
Ferreira, G.P., Gil, F.B. (1981): Elemental analysis of gold coins by particle induced X-ray
emission (PIXE).- Archaeometry, 23 (2): 189-197.
Frick, C., Dausch, H. (1932): Taschenbuch fur metallurgische Probierkunde; Bewertung und
Verkaufe von Erzen fur Geologen, Berg-, Hiitten-Ingenieure und Prospektoren.- F. Enke
Stuttgart.
Gasparrini, C. (1993): Fire assay: its potential and its limitations.- Mining Mag., July 1993:
31-35.
Gillies, K., Urch, D. (1985): XPS depth profile analysis of the surface decoration on a pre-
Columbian gilded gold-copper alloy pectoral.- Hist.Metall., 19 (2), 176-185.
Gooding, K. (1993): End of the lode for gold.- Financial Times, 20-21 March 1993.
Gordus, A.A., Gordus, J.P. (1974): Neutron activation analysis of gold impurity levels in
silver coins and art objects.- Adv.Chem.Ser., 138: 124-127.
Gordus, A.A., Metcalf, D.M. (1980): Neutron activation analysis of the gold coinage of the
Crusader states.- Spec.Publ. - R.Numis.Soc., 13 (MetaII.Numis., V.l): 119-150.
Hase, H.-W. von (1989): Einige Probleme der Edelmetallanalytik.- Erzmetall, 42 (4): 159-
162.
Haber, F. (1927): Das Gold aus dem Meerwasser.- Z. angew. Chemie, 40: 303-314.
Hartmann, A. (1970): Prahistorische Goldfunde aus Europa; Spektralanalytische Untersu-
chung und deren Auswertung.- Studien zu den Anfangen der Metallurgie, Band 3 (SAM
3), Gebriider Mann Berlin: 16-21.
Henglein, M. (1962): LOtrohrprobierkunde.- 4th ed., de Gruyter Berlin.
J6rn, F. (1993): Goldsuche mit Hilfe von Fluoreszenz-Analysator und Computer.- Frankfur-
ter Allgemeine Zeitung 9.2.93.
Klockenkamper, R., Becker, M., Otto, H. (1990): X-ray spectrometric analysis of genuine
and counterfeit gold coins of the German Empire.- Spectrochim. Acta, Part B, 45 B (9):
1043-1051.
Kowalski, H., Reimers, P. (1971): ZerstOrungsfreie Analyse mittelalterlicher Goldmiinzen.-
European Community, Eurisotop, Report No.62: 25 pp., Brussels.
Kowalski, H. (1972): Kernstrahlen entlarven Kunstfalscher und Falschmiinzer.- Eurospectra,
11 (2): 47-55.
314

Krug, C. (1914): LOtrohrprobierkunde; Anleitung zur qualitativen und quantitativen Untersu-

chung mit Hilfe des LOtrohres.- Springer Berlin.
Lenahan, W.C., de Murray-Smith, R. (1986): Assay and analytical practice in the South
African Mining Industry.- The South African Institute of Mining and Metallurgy, Mo-
nograph Series M 6, Johannesburg.
Love, L.J.C., Soto, L., Reagar, B.T. (1980): Surface studies of ancient gold coins and mo-
dern copies by X-ray fluorescence, scanning electron microscopy and scanning Auger
spectroscopy.- Applied Spectroscoppy, 34 (2): 131-139.
Liischow, H.-M. (1980): Probenahme von Edelmetallen.- In: Probenahme, Theorie und Pra-
xis, Heft 36, Schriftenreihe GDMB, Weinheim.
Liischow, H.-M. (1989): Zur Probenahme von Edelmetallen, Teil I: Probenahme von metali-
schen Materialien.- Erzmetall, 42 (4): 153-159.
Meeks, N.D., Tite, M.S. (1980): The analysis of platinum-group element inclusions in gold
antiquities.- J. Archaeol. Sci., 7 (3): 267-275.
Mommsen, H., Schmittinger, T. (1981): Test analysis of ancient Au and Ag coins using high
energy PIXE.- Archaeometry, 23 (1): 71-76.
Moore, D.T., Oddy, W.A. (1985): Touchstones: Some aspects of their nomenclature,
petrography and provenance.- J. Archaeol. Sci., 12: 59-80.
Mylius, F. (1911): Quantitative Goldanalyse mit Ather.- Z. anorg. Chemie, 70: 203-231.
Nriagu, J.O. (1985): Cupellation: The oldest quantitative chemical process.- J. Chern. Educa-
tion, 62 (8): 668-674.
Oddy, W.A. (1972a): Analysis of four hoards of Merovingian gold coins.- Spec. Publ. - R.
Numis. Soc., 8: 111-125.
Oddy, W.A. (1972b): Analysis of gold coins. Comparison of results obtained by non-de-
structive methods.- Archaeometry, 14 (1): 109-117.
Oddy, W.A. (1983): Assaying in antiquity.- Gold Bull., 16 (2): 52-59.
Oddy, W.A., Hughes, M.J. (1972): Specific gravity method for the analysis of gold coins.-
Spec. Publ. - R.Numis.Soc., 8: 75-87.
Oddy, A., La Niece, S. (1986): Byzantine gold coins and jewellery.- Gold Bull., 19 (1): 19-
27.
Oddy, W.A., Schweizer, F. (1972): Comparative analysis of some gold coins.- Spec. Publ. -
R. Numis. Soc., 8: 171-182.
Oddy, W.A., Blackshaw, S.M. (1974): The accuracy of the specific gravity method for the
analysis of gold alloys.- Archaeometry, 16 (1): 81-90.
Oddy, W.A., Munro-Hay, S.C. (1980): The specific gravity analysis of the gold coins of
Aksum.- Spec. Publ. - R. Numis. Soc., 13 (Metall. Numis., V.l): 73-82.
Otto, H. (1939): Die chemische Untersuchung des Goldringes von Gahlstorf und seine Bezie-
hungen zu anderen Funden.- Jahresschrift d. Focke-Museums Bremen: 48-62.
Pernicka, E. (1989): Zur Probenahme von archaologischen Metallobjekten.- Arbeitsblatter f.
Restauratoren, RGZM, Heft 1: 138-148.
Philipsborn, H. von (1953): Tafeln zum Bestimmen der Minerale nach auBeren Kennzei-
chen. - Schweizerbart Stuttgart: 186-187.
Piette, M., Demortier, G., Bodart, F. (1986): PIXE analysis of solders on ancient gold arti-
facts with a nuclear microprobe.- Microbeam Anal., 21st.: 333-336.
 315

Plattner, C. F. (1835): Die Probierkunst mit dem L6throhre, oder Anleitung, Mineralien,
Erze, Hiittenproducte und verschiedene Metallverbindungen vor dem Uithrohre, mit
theilweiser Anwendung des nassen Weges, qualitativ fast auf aile Bestandtheile, und
quantitativ auf Silber, Gold, Kupfer, Blei und Zinn in kurzer Zeit zu untersuchen.- J. A.
Barth Leipzig.
Radcliffe, C.D., Angle, B., Marcias, E.S., Gaspar, P.P. (1980): Gold analysis by differential
absorption of gamma-rays.- Archaeometry, 22 (1): 47-55.
Reimers, P., Lutz, G.J., Segebade, C. (1977): The non-destructive determination of gold,
silver and copper by photo-activation analysis of coins and art objects. - Archaeometry,
19 (1): 167-172.
Schweizer, F., Friedman, A.M. (1972): Comparison of methods of analysis of silver and
gold in silver coins.- Archaeometry, 14 (1): 103-107.
Scott, D.A., Seeley, N.J. (1983a): The examination of a pre-Hispanic gold chisel from
Colombia.- 1. Archaeo!' Sci., 10: 153-163.
Scott, D.A. (1983b): A note on the dissolution of ancient gold alloys.- Archaeometry, 25 (2):
214-216.
Stoch, H. (1986): A manual of analytical methods used at MINTEK.- Special Publication
No.7, Council for Mineral Technology (MINTEK), Randburg: 81-88.
Vofite, A. (1985): Die Feingehaltsbestimmung der Goldmiinzen.- In: Castelin, K. (ed.):
Keltische Miinzen, Bd.2, Eidg. Drucksachenzentrale, Bern: 55-68.
Vofite, A. (1991): Die Analysenverfahren fur Goldgegenstande.- In: Gold der Helvetier
(Exhibition Catalogue), Schweizer Verlagshaus, Ziirich, 49-51 and 164-167.
Wilde, P.M., Paschmann, W.F. (1989): Quantitative microprobe analysis of ancient Indian
gold coins.- Z.Chemie, 29 (9), vo!. 348: 1618-1622.
Ziegaus, B. (1991): Der Miizfund von GroBbissendorf.- Diss. University of Passau; publica-
tion in preparation.