Assaying in Antiquity

Andrew Oddy
Conservation Division, The British Museum, London, U.K.
Modern chemical analysis has derived largely from the methods developed in
antiquity as a result of the need to assay gold and other precious metals for coin
authentication and other reasons. Surprisingly the three basic assay methods in use
2 500 years ago are still employed today.
In Aristophanes play 'The Frogs', first performed in 405
B.C., the leader of the chorus says:
'We do not use these (coins) which have not been adulterated, the
most beautiful of all coinage, so it seems, and which alone are struck
clearly and proven true by ringing, but these worthless bronzes,
struck yesterday and today with the foulest stamp.' (1)
This allusion to bronze coins refers to the fact that between
406 and 404 B.C. the Athenian state, hard pressed financially
by the cost of the Peloponnesian war, had officially issued silverplated copper or bronze coins (Figure 1) in place of their pure
silver coins which had hitherto been widely disseminated
throughout the eastern Mediterranean as a result of trade.
The issuing of debased or plated coins of gold or silver is
usually a major retrograde step for any government but, apart
from such infrequent instances of the 'official' manufacture of
'forgeries', ever since the invention of coinage in Asia Minor
towards the end of the 7th century B.C., forged coins have been
made by individuals in the anticipation of a quick profit.
That coin forgery was commonplace in the ancient world is
attested by the numbers of surviving counterfeits made ofsilveror gold-plated base metal. Forgeries exist for virtually all the
precious metal coinages struck in antiquity, although they are
more frequent from some periods than others. The survival rate
of forged coins is, however, bound to be variable, as one of the
main sources of ancient coins is the discovery of hoards, and
hoarders, by definition, usually select the best available coins
for burial. Nevertheless many hoards are known which have
included contemporary forgeries, a good example being the
15th-century hoard of gold coins of the kings Edward III to
Edward IV found at Fishpool in Nottinghamshire in 1966 (2).
This contained two forgeries which had been made by plating
gold onto the surface of a silver core by the method of firegilding (3) (Figure 2).
The craftsman's attitude to fraud in the Roman period seems
to have been that cheating was a battle of wits between the
customer and the artisan. Pliny, who died in 79 A.D. , records a
method of gilding base metals by sticking gold leaf onto the
surface with mercury (4); however he also adds that some
craftsmen used a fraudulent technique ofsubstituting the much
cheaper white of an egg for the mercury as the adhesive.
52

In the books of metallurgical recipes which have survived

from the Roman and Medieval periods there are numerous
descriptions of alloying processes which are said to increase the
bulk of silver or gold without changing its appearance. Needless
to say, gilding is dealt with in some detail, and the 4th-century
A.D. Leyden Papyrus X (5) cites two such recipes, one for
gilding copper with a mixture of lead and gold (6) and the other
for gilding silver by the fire-gilding technique (7); in each case
the recipe states that the gilded objects will pass the test for gold.
The governments of antiquity were extremely concerned that
their coinage should not be counterfeited and took care to
protect it by laws which prescribed severe penalties for those
convicted of forgery (8). Such laws can be traced back to Roman
times (9), but perhaps the earliest legal document which deals
with forged coins describes a law enacted in Athens in 375 or 374
B.C. and which was recently discovered inscribed on a stone
stela (10). The law provided for the establishment of a post of
coin-tester in the market place, the duties of the post being to
test coins and to pronounce on whether or not the silver was
good. In particular it mentioned that coins which were
discovered to have either copper or lead cores would be
forfeited.
Unfortunately the Athenian law gave no indication as to how
the testing was to be carried out, although it is perhaps implied
that cutting was at least one of the techniques used, and
numerous coins are known from antiquity which have test cuts
on the surface, made to ensure that they were not plated (Figure
1). It is also probable that ringing of coins by dropping them on a
hard surface was practised, as mentioned by Aristophanes (see
quotation above), and doubtless the semi-quantitative heating
method might have been employed. The Leyden Papyrus (11)
says that gold can be tested by melting it and letting it cool. Pure
gold does not change colour, but if silver is present the alloy so
treated becomes white and if lead is present it turns black. The
same process is described for testing silver (12), which blackens if
it contains lead and becomes yellow if it contains copper. Pliny
records the same procedure, but classifies the silver according to
the colour generated by heating (13). He also mentions one
other qualitative test for silver by claiming that polished pure
silver is more easily misted by breathing on it than is silver which
has been debased.

GoldBull., 1983, 16,(2)

However, apart from these rather

rough and ready techniques for testing
precious metals the Greeks knew of three
quantitative methods for assaying gold
and silver and all three are still in use
today, two of them in the assay offices of
the world (fire assay and the touchstone)
and the third in various museum
laboratories for the nondestructive
analysis of gold coins (specific gravity
measurement by the Archimedes
method).
Fire Assay
Among the earliest references to the
assaying of gold is a document dating
from the reign of the Egyptian Pharaoh
Amenhophis IV (1377-1358 B.C.),
written on one of the cuneiform tablets
(Figure 3) known as the Amarna letters,
in which King Burraburiash of Babylon
(1385-1361 B.C.) complains to the
Pharaoh that when a recently received

Fig. 1 (Top) The small coin shown at left is an

Athenian silver-plated drachm issued in 406-404
B.C. The larger Athenian silver tetradrachmon the
right shows cuts on the surface made to test the
quality. This coin is from the 4th century B.C.

Fig. 2 (Middle) Gold-plated forged coins from

the timeof Henry VI (1422-1461) anddiscoveredin
the Fishpool hoard
Fig. 3 (Bottom) Cuneiform tablet of King
Burraburiash of Babylon (1385-1361 B.C.) which
records the assaying of a consignment of gold from
Ägypten

GoldBull., 1983, 16,(2)

53

Fig. 4 Furnace for fire assay from Juan de Arphe y

Villaphane, QuilatadordelaPlata, OroyPiedras',
Valladdid, 1572

consignment of gold was tested in the furnace, only one quarter

of the original mass was left at the end of the process (14). What
is referred to, of course, is not true assaying in the sense that a
sample is removed for analysis, but rather the quantitative
refining of an alloy. Nevertheless, this text is typical of
numerous references to the purification of gold dating from the
second and first millennia B.C. (15).
The essentials of fire assaying are the addition of excess lead to
the gold, followed by melting in a special crucible under an
oxidizing flame (Figure 4). The alloying elements (apart from
silver) and the lead are oxidized and the resulting dross is
absorbed into the ceramic crucible, leaving behind the gold and
any silver which it contains. In antiquity these two metals were
then parted by strongly heating in a crucible filled with a
mixture of brick dust, vitriols (that is to say, copper and iron
sulphates) and salt (the cementation process) which converted
the silver to silver chloride. This latter was absorbed by the brick
dust and the gold was recovered unaffected (16). Another
parting technique, described by Theophilus in the 12th century
A.D. (17) consisted of heating the gold-silver alloy with
sulphur, which reacts with silver but not with gold. A variation
on this technique was to heat the alloy with antimony sulphide,
which converts silver to its sulphide, but which alloys some of its
antimony with gold. These metals are then separated by further
heating (18). This method was first described in the early 16th
century, but its origins are uncertain (19). Finally, with the
discovery of nitric acid in the Middle Ages, the modern method
of parting by simply dissolving the silver out of the gold was
developed.
The origins of the cupellation process are unknown, but
excavations at Sardis in Asia Minor have uncovered a workshop
54

area, dating from the 6th century B.C. , which is thought to have
been used for the large-scale purification of electrum (the
naturally occurring gold-silver alloy) by cupellation and for the
parting of silver and gold by cementation (20).
Apart from these excavations, the evidence for the history of
cupellation and parting is derived from ancient authors who
allude to the process. Among the earliest are biblical references
in the books of Isaiah (8th century B.C.) Jeremiah (7th century
B.C.), Ezekiel (6th century B.C.) and Malachi (5th century
B.C.) (21), but the earliest technical description appears to be
an Indian Sanskrit document, known as the Artha'ssta of
Kautilya, which describes the purification of gold by melting it
with four times as much lead as there is impurity in the gold (22) .
The date of the Artha'sstra is uncertain; most scholars think that
the earliest parts of it are from the 3rd century B.C., but with
later additions. The same process is mentioned in the 2nd
century B.C. by Diodorus Siculus (23), and one century later
Strabo (24) states that silver can be burnt out of gold by fire, but
does not directly mention the addition of salt, although earlier
in the text he mentions vitriol, which comprises one of the
ingredients of the parting process, and so this latter is what must
have been meant.
Pliny describes cupellation with lead and the use of vitriol and
salt for parting gold and silver (25). He also mentions the use of
special crucibles, but says that they were made ofwhite clay (26),
whereas by medieval times cupels contained a large amount of
bone ash. Other references in Diodorus Siculus and Cassiodorus
to the use of special cupels have been collected together by
Forbes (27). Some 16th- or 17th-century cupels have recently
been excavated on the site of the medieval mint in the Tower of
London (28) (Figure 5).
GoldBull., 1983, 16, (2)

The 4th-century A.D. Leyden Papyrus has a recipe for

colouring gold which is, in reality, a method of parting gold and
silver by heating with vitriol and salt, or vitriol, salt and alum
(29). The Papyrus also describes the cupellation of silver as a
method of purifying the metal, but contains no mention of
quantitative measurements (30). In fact the quantitative use of
cupellation is not described by any of the classical authors,
although there seems little doubt that the process was used
quantitatively as swinging balances had been available since the
Bronze -Age. By the Roman period these were capable of
weighing to about plus or minus 0.1 g.
One of the earliest medieval accounts of cupellation is
contained in the work of the great Islamic alchemist Abu Musa
Jabir ibn Haiyn, known to Western alchemists as Geber. He
described the manufacture and use of cupels and clearly stated
that they should contain a large proportion of bone ash (31).
Geber lived in the 8th century A. D., but the works attributed to
him were added to considerably in later centuries. In the 12th
century A.D. Theophilus described cupellation as a means of
purifying silver (32), but not of measuring purity. However, in
the same century Richard Fitz Nigel (the Bishop of London)
clearly alluded to the process as a quantitative method of assay
when discussing the quality of the coin of the realm (33). The
process is again mentioned for the assay of the English silver
coinage in 1280 (34) and for the assay of the newly introduced
gold coinage of Edward III around 1350 (35). These dates bring
us within the known history of the Worshipful Company of
Goldsmiths of London which was in existence by at least 1179,
received its first Royal Charter in 1327 and appointed its first
full-time assayer in 1478 (36) (Figure 6). Soon after this the first
printed descriptions of the techniques of assaying appeared,
spreading knowledge of the techniques of cupellation to all who
could read (37, 38) (Figure 7).
.

Archimedes Method
This method of assaying depends on the fact that the specific
gravity of gold (19.3 g/cm3) is nearly twice that of silver (10.5)
and more than twice that of copper (8.9). Thus, as gold is
debased with these metals its specific gravity is progressively
reduced, and may be used as an indication of the extent of the
debasement. The discovery of this fact is traditionally ascribed
to the philosopher Archimedes (about 287-212 B.C.) who is
supposed to have leapt from his bath and run home naked
shouting `Eureka' when he realised that a solid immersed in a
liquid suffers an apparent loss in mass equal to the mass of
displaced liquid. Archimedes used this principle to check
whether a craftsman, who had made a crown for King Hiero,
had added any base metals to the gold (39).
It is, however, a long step from establishing the principle to
using it as an accurate method of analysis, and the first
indication of the use of Archimedes' method as a quantitative

GoldBull., 1983, 16,(2)

technique is in the 6th century A.D. (40). It is then described in

several manuscripts of the period from the 10th to the 13th
centuries (41) including the manuscript known as Heraclius'
which is a list of craftsmen's recipes for making pigments and
carrying out various metallurgical operations (42). Another,
more accurate, description is to be found in the book of recipes
known as the 'Mappae Clavicula' (43), which were probably
first collected together in the 8th century A.D., although the
description of Archimedes' assaying method appears to be a
12th-century addition. In this early period both pans of the
balance were immersed in a tank of water, and the gold-silver
alloy had to be weighed with pure silver weights (to avoid errors
caused by the difference in the specific gravities of silver and
brass, the latter being the usual metal for balance weights)
(Figure 8).
One mention of Archimedes' method in connection with
gold coins occurs in a 14th-century Islamic book written in
Morocco, which specifically states that forged coins can be
detected by weighing first in air and then in water (44). In the
post-medieval period there was considerable interest in
automating specific gravity determinations and, among others,
both Galileo and Robert Boyle designed specially graduated
balances to simplify the calculation of the results (45).
Nevertheless, in spite of knowledge of Archimedes' principle
for over 2 000 years, it is only in modern times that it has been
used for accurate assay work, one application being for the
analysis of ancient gold coins by numismatists and museum
scientists (46).

Touching
Cupellation and Archimedes' methods rely ort the accurate
weighing of the sample to be analyzed, and their accuracy is thus
directly related to that of the balances available to the assayers.
The third method of quantitative analysis which was known in
the ancient world is, however, essentially nondestructive.
Touching involves rubbing the alloy under investigation onto
the surface of a piece of smooth, fine-grained, slightly abrasive,
black stone and comparing the colour of the streak produced
with those obtained from standard alloys (47). The accuracy of
the method depends on knowing whether the alloying element
in the gold issilver or copper and on having a suitable range of
standard alloys for comparison with the unknown sample.
Nevertheless Theophrastus (371-288 B.C.) claimed that
touchstones could be used with an accuracy of 1 part in 144 (48)
and Pliny (23-79 A.D.) said that a touchstone would `detect
silver or copper to a difference of a scruple' (49). Unfortunately
Pliny did not say in how much mass the scruple difference could
be detected, but there were 24 scruples in a Roman ounce and
288 scruples in aRoman pound, and it seems most likely that the
touchstone was accurate to 1 part in 24, that is an accuracy of
about 2 per cent. This is supported by evidence from the analysis
55

could be obtained from only one site in

Asia Minor. The touchstone method of
assay is discussed in the Sanskrit
Artha'sstra (3rd century B.C. with later
additions) which also describes the
manufacture of a set of touchneedles
(standard alloys) in which the ratio of
gold to copper is varied in steps of 1/16
(that is 1 /2 carats) (53).
After Pliny, the touchstone is mentioned in passing in the 4th-century
Leyden Papyrus X, but Western sources
are then surprisingly silent until the
appearance of a law of Edward I (promulgated in 1300 A.D.), which refers to a
standard for gold alloys known as the
touch of Paris' (54), and a document of
Fig. 5 Cupels from the site of the Tudor mint in the Tower of London
about 1350, which states that an 'assay of
gold may also be made by the
touchstone; but that assay can only be
determined by experts in the art, and
hardly by them without frequent
failures.' (55)
This latter statement is, perhaps,
rather pessimistic, and an Islamic treatise
written about the mint in Cairo, in the
early 13th century (under the Ayyubid
dynasty) describes the use of the
touchstone and the provision of touchneedles in which the ratio of gold to silver
is varied in steps of 1 / 24 (that is one carat)
(56). This is the same accuracy as is
deduced above from the incomplete
description in Pliny. Strangely, the
earliest Chinese reference to the touchstone appears to date from 1387 A.D..
'(57), but the technique was (-and is)
widely used in the East so that by the 19th
Fig. 6 Weighing (left) and fire assaying (right) are depicted through a window, dated 1624in Constance,
century Gowland, who had seen the
Schweiz
method in use in Japan, could write that
accuracies
of
better
than plus or minus 1 per cent were
of early medieval coins (50).
obtainable
by
skilled
craftsmen
(58).
The earliest references to the use of the touchstone have been
As
with
the
other
two
methods
of assaying the touchstone and
discussed by numerous authors (51) , and appear to start with the
its
use
are
fully
described
in
numerous
printed books which
Greek poet Theognis in the 6th century B.C., although Bogaert
began
to
make
their
appearance
in
the
early 16th century
has postulated its origin in Egypt in the 12th century B.C. (52).
(Figure
9),
and
it
is
still
widely
used
today,
not only as a rough
Mentions are also to be found by the historian Herodotus (5th
qualitative
check
of
whether
a
piece
of
scrap
metal contains any
century B.C.) and in 'Agamemnon' by. Aeschylus (about 460
gold,
but
also
as
an
accurate
method
of
analysis
in numerous
B.C.) and in 'Oedipus Tyrannus' by Sophocles (about 420
official
assay
offices.
B.C.). However, it is not until the book on stones by
However, unlike the other two methods, for which there is
Theophrastus (371-288 B.C.) that a detailed description of
thus
far no physical evidence for their use in antiquity,
touchstones was given, including the information that they
-

56

GoldBull., 1983, 16, (2)

Fig. 7 The manufacture of cupels from Sir John Pettus, 'The Laws of Art and
Natureinknowing, judging, assaying, fining, refining... Metals' London, 1686

Fig. 8 Specific gravity measurement with both pans of the balance immersed,
as was the practice in early times, from Sir John Pettus, `The Laws of Art and
Natureinknowing, judging; assaying, fining, refining... Metals', London,1686

touchstones are to be found in the archaeological record,the

earliest known ones being from the 6th/ 5th century B.C. levels
of the Bhir mound at the ancient city of Taxila in what is now
Pakistan (59). Some of these were described by the excavator as
being black natural pebbles while others were pieces of hard
siliceous slate cut into long strips and still retaining traces of gold
streaks on their surface.
One touchstone is known to have been found at the Iron Age
hill-fort of Hengistbury Head, Hampshire (60) but
touchstones appear to have been recorded from Roman sites,
although this is likely to be due to a failure to recognise them for
what they are, and there is little doubt that many touchstones
are misidentified in museums as hone- or whet-stones used for
sharpening tools.
For the early medieval period numerous examples of
of them
touchstones are known (61, 62, 63) (Figure 10),
still retaining traces of gold on the surface, and recent
petrological analysis of some of them has shown that three main
types of rock were used, tuffs, cherts and siltstone, with various
other types being used occasionally (64).
By the Renaissance period, and later, the sets of touchneedles
had grown to include ternary alloys of gold, silver and copper in
which the silver to copper ratio was varied to facilitate the assay
of a wide range of alloys and the touchstones had grown with

Fig. 9 Touchstones (B) and touchneedles (A) from Juan de Arphe y

Villaphane, 'Quilatador de la Plata, Oro y Piedras', Valladdid, 1572, with the
streaks (C to F) produced by rubbing alloys and standards onto the stone

no

many

GoldBull., 1983, 16, (2)

57

Fig. 10 Early medieval touchstones from (lett to

right): Andernach (?6th/7th century A.D.),
Ozengell, Kent (late 7th century A.D.),
Winchester (10th century A.D.), and Winchester
(9th century A.D.)

them so that some examples were very Iarge tablets of black

stones, often fitted with handsome metallic mounts (Figure 11).
All the ancient sources are in agreement that the best
touchstones should be black, but the Sanskrit Artha'sstra
relates that the artisan can cheat the customer by secreting red
chalk behind a fingernail and introducing it onto the streak as
the metal is rubbed on the stone. The same source also adds that
if the artisan is selling gold it would be to his advantage to use a
stone with a greyish-green tinge, and if he is buying gold he
should use a stone of uneven colour (65). These remarks reflect
the medieval recipes, quoted above, by which the craftsmen of a
later period sought to cheat their customers with debased or
plated metals.

Concluding Remarks
The methods developed in antiquity for the assaying of
precious metals, in particular gold, are the ancestors of modern
chemical analysis, and the need to be able to determine the
composition of a manufactured material is the basis for all
quality control. In view of the rapid advances made in the
techniques of chemical analysis in the past thirty years it is very
surprising that the three techniques known to the Greeks of
2500 years ago are still in use.
Acknowledgements

The author's grateful thanks are due to all those Museum Curators and
excavators who have loaned touchstones for study during the past few years. lam
also grateful to Judith Swaddling for references to assaying in Greek literature

Fig.11 Large touchstone which was originally from

the Royal Mint, London and is now in the Science
Museum. Probably 18th or 19th century A.D.
Crown Copyright. Science Museum, London

58

GoldBull., 1983, 16, (2)

and translating the quotation from Aristophanes, to George Boon for loaning a
slide of the stained glass window in Constance (Figure 6), to the Ancient
Monuments Laboratory of the Department of the Environment for the slide of
the cupels from the Tower of London (Figure 5) (which is reproduced with the

permission of the Controller of Her Majesty's Stationery Office) and to the

excavator, Mr. G. Parnell, and to the Science Museum for a slide of its large
touchstone (Figure 11). The other illustrations (Figures 1-4, 7, 8, 9 and 10) are
reproduced by permission of the Trustees of the British Museum.

References
1 Aristophanes, Frogs, 721-726
2 M.M. Archibald, 'Fishpool, Blidworth (Nottinghamshire), 1966 Hoard:
Inerim Report', Numismatic Chron., 7th Series, 1967, 7, 133-146
3 W.A. Oddy, unpublished data
4 H.Rackham, 'Pliny. Natural History', Vol. IX, Loeb Classical Library,
London, 1952 (reprinted 1968), p. 77
5 R. Halleux, 'Les Alchimistes Grecs' , Tome 1, Papyrus de Leyde, Papyrus de
Stockhom, Fragments de Recettes, Societe d'Edition Les Belles Lettres, Paris,
1981. See pp. 22-24 for a discussion of the date
6 See (5), No. 37, p. 93
7 See (5), No. 55, p. 97
8 D.M. Metcalf and F. Schweizer, 'The Metal Contents of the Silver Pennies of
William I and Henry 1(1087-1135)', Archaeometry, 1971, 13, 177-190
9 P. Grierson, 'The Roman Law on Counterfeiting' in 'Essays on Roman
Coinage Presented to Harold Mattingly', edited by R.A.G. Carson and
C.H.V. Sutherland, Oxford, 1956. pp. 240-261
10 R.S.Stroud, 'An Athenian Law on Silver Coinage', Hesperia, 1974,43,(2),
157-188
11 See (5), No. 42. p. 95
12 See (5), No. 43, p. 95
13 See (4), pp. 95-97
14 J.A. Knudtson, 'Die El-Amarna- Tafeln', Leipzig, 1915, pp. 90-95
15 M. Levey, ' The Refining of Goldin Ancient Mesopotamia' , Chymia, 1959,
5, 31-36
16 J. Needham, 'Science and Civilisation in China' Vol. 5, Part II, Cambridge
University Press, Cambridge, 1974, p. 39
17 J.H. Hawthorne and C.S. Smith, 'Theophilus - On Divers Arts', The
Chicago University Press, Chicago, 1963 (reprinted Dover, New York,
1979), p. 147
18 D. Stockdale, 'Historical Notes on the Assay of Gold' , Sci. Prog., 1924, 18,
476-479
19 See(16)
20 G.M.A. HanfmannandJ.C. Waldbaum, 'The Eleventh andTwelfth Campaigns at Sardis (1968, 1969)', in 'Bulletin of the American Schools of
Oriental Research', No. 199, 1970, 7 ff.
21 See (16), p. 40
22 M.K. Pal,'Crafts and Craftsmen in Traditional India', Kanak Publications,
New Delhi, 1976, p. 255
23 J.F. Healy, 'Greek Refining Techniques and the Composition of GoldSilver Alloys', Revue Beige de Numismatique, 1974, 120, 19-33, especially
25 ff.
24 See (23), 27 ff.
25 See (4), pp. 49 and 65-67
26 See (4), p. 55
27 R.J. Forbes, 'Studies in Ancient Technology, VIII', Leiden, 1971, p. 177
28 These were shown tome by the excavator, Mr. G. Parnell, and by Mr. L. Biek
and Miss J. Bayley of the Ancient Monuments Laboratory, London
29 See (5), No. 24, p. 91
30 See (5), No. 25, p. 91
31 E.J. Holmyard, 'The Works of Geber Englished by Richard Russell, 1678',
J.M. Dent & Sons Ltd., London, 1928, pp. 183-184
32 See(17), p. 96
33 C.Johnson, 'The Course of the Exchequer by Richard, Son of Nigel', T.
Nelson & Sons Ltd., London, 1950, p. 37
34 C.Johnson, 'The De Moneta of Nicholas Oresme and English Mint
Documents', London, 1956, p. 67
35 See (34), p. 84

Go<dBull., 1983, 16, (2)

36 T.F. Reddaway and L.E.M. Walker, 'The Early History of the Goldsmiths'
Company' , E. Arnold, London, 1975
37 J.P. Phillips,' 16th CenturyTexts onAssaying' , Chem. Educ., 1965, 12, (7),
393-394
38 C.S. Smith, 'Metallurgy and Assaying', in 'A History of Technology: Vol.
III: from the Renaissance to the Industrial Revolution', edited by C. Singer
et al., Clarendon Press, Oxford, 1957, pp. 27-71
39 F. Granger, 'Vitruvius on Architecture', Vol. 2, London, 1970, pp. 203-207
40 M. Clagett, 'The Science of Mechanics in the Middle Ages', Madison, 1959
41 C.S. Smith and J.G. Hawthorne, 'Mappae Clavicula: A Little Key to the
World of Mediaeval Techniques', Trans. Am. Philos. Soc., New Series,
1974, 64, (4), 56 see footnote 131
42 M.P. Merrifield, 'Original Treatises..... on the Arts of Painting in Oil',
London, 1849, pp. 226-227
43 See (41), p. 56
44 C. Toll' Minting Technique According to Arabic Literary Sources', Oriental
Suecana, 1970-1971, 19-20, 125-139, especially 135
45 See (41)
46 W.A. Oddy, 'Goldin Antiquity: Aspects of Gilding and of Assaying' , J.R.
Soc. Arts, 1982, 130, (5315), 730-743, especially 741
47 W. Wlchli,'TouchingPreciousMetals', GoldBull., 1981,14,(4),154-158
48 E.R. Caley andJ.C. Richards, 'Theophrastus on Stones', The Ohio State
University, Columbus, OH., 1956, p. 155
49 See (4), p. 95
50 J.P.C. Kent, 'The Date of the Sutton Hoo Hoard', in R. Bruce-Mitford,
'The Sutton Hoo Ship Burial', Vol.I, The British Museum, London, 1975,
pp. 588-607, especially pp. 590-591
51 See(23),pp.30ff.
52 R. Bogaert, 'L'essai des monnaies dans l'antiquite', Revue Beige de
Numismatique, 1976, 122, 5-34, especially pp. 8 ff.
53 R.P. Kangle, 'The Kautiliya Artha'sstra (Part II)', Bombay, 1963, pp.
128-129
54 S. Hare, 'Touching Gold and Silver: 500 Years of Hallmarks',(Catalogueof
ttn exhibition held at Goldsmiths' Hall, Foster Lane, London, in November,
1978), London, 1978, p. 14
55 See (34), p. 85
56 M. Levy, 'Mediaeval Arabic Minting of Gold and Silver Gains, Chymia,
1976, 12, 3-14, especially 7
57 J. Needham, 'Science and Civilisation in China' Vol. 3, CambridgeUniversity Press, London, 1959, p. 672
58 W. Gowland, 'Japanese Metallurgy I: Gold and Silver and their Alloys', J.
Soc. Chem. Ind., 1896, 15, 404 ff.
59 J. Marshall, 'Taxila', Vol. II, Cambridge University Press, Cambridge,
1951, p. 502
60 W. Gowland, 'Appendix II: Report on the Metals and Metallurgical Remains', in J.P. Bushe-Fox 'Excavations at Hengistbury Head, Hampshire,
in 1911-12', Reports of the Research Committee of the Society of Antiquaries of London, No. 3, London, 1915, pp. 72-83
61 See (46), 740-741
62 V. Zedelius, 'Coticulae, Merowingerzeitliche Probiersteine im nrdlichen
Rheinland', in 'Das Rheiniche Landesmuseum, Bonn: Berichte aus der
Arbeit des Museums', No. 4, 1979, pp. 58-59
63 V. Zedelius, 'Merowingerzeitliche Probiersteine im nrdlichen
Rheinland', DerAnschnitt, 1981, 31, (1), 2-6
64 D.T. Moore and W.A. Oddy, 'Touchstones: Some Aspects of their
Nomenclature, Petrography and Provenance', in preparation for J. Archaeol. Sci.
65 See(53),p.128

59