Ancient Electrum Coins: The Case for Manufactured Alloys

By Tyler Rossi for CoinWeek …..

In 2016, archeologists discovered a small golden bead (3.175 mm, 0.148 g) while digging in the ancient village of Pazardzhik in central Bulgaria. Between 6,500 and 6,600 years old, this bead is the oldest processed gold artifact produced by human beings and is approximately four thousand years older than the oldest coin. What relevance does this golden bead hold to the discussion of electrum coins? It is proof that by the time humans began striking coins in the seventh and sixth centuries BCE, they had many millennia of metallurgical knowledge to guide their hands.

This begs the question: was the electrum used in these early coins naturally occurring, or produced specifically for the purpose of striking coins?

In their book King Croesus’ Gold, Andrew Ramage and Paul Craddock describe how the ancient Greeks discovered the process of refining the naturally occurring gold and silver alloys of the region into effectively pure metals. The discovery of this process long before the first coins were struck mostly disproves the assumption that “natural alloys may have played in the origins of coinage.” Why would states use an alloy with dramatically varying percentages of gold and silver to strike coins that are meant to be standardized – especially when these metals are highly valuable? Additionally, these coins were “produced with an exceptionally high level of control.” This was perhaps due to their extreme value. Scholars have estimated that a stater was equal to almost one month’s salary while the smallest gold coins equal to a 1/96th stater may have “fed a laborer for a week.” As such, it is understandable that early moneyers perfected their craft to such a degree that these electrum coins vary in weight by only a few centigrams.

Kingdom of Lydia, Alyattes Ca. 610-560 BCE). EL 1/96 Stater. (0.14 g & 4 mm.) OBV: Paw of lion REV: Incuse punch. REF: Karwiese Series I, Group 1 (Ephesos); Weidauer -; SNG Kayhan -; SNG von Aulock 1805 (Uncertain Ionia, shell); Elektron II 70.
Kingdom of Lydia. Kroisos. Ca. 564/53-550/39 BCE. AU/EL Stater (16.5mm, 8.08 g) Sardes mint – Light standard OBV: Confronted foreparts of lion and bull REV: Two incuse squares. REF: Berk 3; Kurth G51; SNG von Aulock 2875; BMC 31; Boston MFA 2073; Gulbenkian 757; Traité I 401–3.

How can numismatists and scientists figure out whether or not the electrum used in a particular coin is naturally occurring or not? Since this topic has been debated for quite some time, it is important to be as accurate and meticulous as possible. In order to ascertain the origin of the alloy, we first must know the exact percentages of its component elements.

Electrum alloys with a high iron content are likely natural, while those with high proportions of lead and copper are mostly man-made. Copper was added to harden and redden the electrum, and lead was a byproduct of silver refining. Iron in electrum is naturally occurring. And while there are other trace elements, like bismuth, mercury, and tellurium, the refining process reduces these to less than 0.1%. Therefore, one can be rather certain that any electrum with significantly greater than 0.1% of these elements is naturally occurring.

Currently, the most precise technique to measure the elemental composition of electrum is Prompt-Gamma Neutron Activation Analysis. This non-destructive technique uses a beam of neutrons to takes a “picture” of the metal. Since different metals emit gamma rays of different colors when struck by subatomic particles, it is possible to measure the exact composition of an electrum coin without cutting off a small portion for chemical analysis. Previously, coins would have some metal “cut away” to be tested, leaving the coin damaged and practically uncollectible.

Published by K. Butcher and M. Ponting in their books Metallurgy of Roman Silver Coinage, Nero to Trajan, this denarius of the Roman emperor Otho (ruled Jan. – April 69 CE) from the Jyrki Muona Collection, is an extreme example. Due to the extensive damage caused by outdated invasive techniques, this new noninvasive process is preferred as it preserves a coin’s historical and numismatic value.

Roman Empire, Otho; 69 CE, Rome, AR Denarius, 2.29g. OBV: bare head of emperor with legend. REV: Securitas standing l. holding wreath and scepter with legend. REF: BM-17, RIC-8, Paris-10, C-17 (12 Fr.). NOTE:Ex Jyrki Muona Collection, 1/4 of coin cut away, so that its metallic composition could be tested and published by K. Butcher and M. Ponting, Metallurgy of Roman Silver Coinage, Nero to Trajan, 2014, p. 253.

After a French survey of electrum coins excavated at the mint at Sardis, numismatists can assume that these coins have a persistent level of copper and lead. As a result, numismatists have concluded that “it is likely that all electrum coins, whatever their gold percentages, were produced from non-natural alloys.” It is unknown, however, whether silver and copper were added to naturally occurring electrum, or to previously processed gold. With our current understanding of coinage and scientific tools, it may be impossible to determine which is correct. Previously it was thought that “cementation”, the process whereby silver is separated from gold, was developed around 550 BCE concurrently with the development of coinage. However, as discussed above, this process may have been in use by the mid-sixth century BCE, well before the development of coinage.

In 1936, E.S.G. Robinson published a short piece in the British Museum Quarterly detailing two small gifts of electrum coins of Cyzicus, Phocaea, and Mytilene that the museum received from Mr. H.C. Hoskier. Robinson discusses the varying percentages of gold and silver in the base alloys used to strike the early electrum coinage. Even a cursory visual analysis of the many coins demonstrates that these percentages “fluctuated appreciably.” Some types appear to be almost pure gold, while others are pale and clearly contain significant amounts of silver.

But while this variability may have increased mistrust in the coinage by common people, these coins were of significant monetary value and must have been used for large-scale transactions. Consequently, it is probable that the percentages were adjusted based on the market value of the metals, therefore providing a stable “fixed unit of value”.

MYSIA. Cyzicus. Ca. 550-500 BCE. El hekte (2.53 gm). OBV: Young, beardless head left, tunny fish behind REV: Mill sail incuse. REF: Von Fritze 62, pl. ii, 14-15. Boston 1431
Phocaea, Ca. 600-522 BCE. El Hekte (2.58 g & 10 mm). OBV: Seal swimming left w/ head reverted REV:. Rough quadripartite incuse square. REF: Bodenstedt, Die Elektronmünzen von Phokaia und Mytilene, 1.4 (only one specimen listed).
Mytilene, Ca. 500-490 BCE. El Hekte (2.54 g g). OBV: Forepart of winged boar to right. REV: Lion’s head with open jaws to right, with rectangle behind; all in incuse. REF: BMFA 1678. Bodenstedt 10

Gold held such an important financial, social, and religious role in the ancient world that this theory of “elaborate financial” and monetary control is highly likely – especially since most gold in the monetary system was used to pay for military salaries or state-to-state transfers of wealth. Both of these transaction types would necessitate precise measurement of value due to the potential socio-political consequences if they were not successfully completed. It was in the ancient state’s interest to strictly control the value of their coinage.

Combine this with the fact that the technology existed to create electrum alloys with specific percentages, I am forced to conclude that most electrum coins were struck from manufactured alloys. Nevertheless, without an extensive and comprehensive chemical analysis of thousands of coins, it would be impossible to know what percentage was struck from naturally occurring or manufactured electrum.

Regardless of whether the electrum used in early coins was manufactured or naturally occurring, these coins are highly collectible. Early period full staters, struck in the late seventh and early sixth centuries, are very expensive. Even in Poor condition, these types can cost many thousands of dollars. Slightly later types from the mid-sixth to mid-fifth centuries in average condition range from between $1,200 and $5,000. Of course, extremely high-grade and rare types will sell for much more. The smaller fractional electrum types may be acquired for a much more reasonable price. Pieces in poor condition sell for as little as $300, with higher grade types generally priced between $2,000 and $10,000.

* * *


* * *

About the Author

Tyler Rossi is currently a graduate student at Brandeis University’s Heller School of Social Policy and Management and studies Sustainable International Development and Conflict Resolution. Before graduating from American University in Washington D.C., he worked for Save the Children creating and running international development projects. Recently, Tyler returned to the US from living abroad in the Republic of North Macedonia, where he served as a Peace Corps volunteer for three years. Tyler is an avid numismatist and for over a decade has cultivated a deep interest in pre-modern and ancient coinage from around the world. He is a member of the American Numismatic Association (ANA).


This site uses Akismet to reduce spam. Learn how your comment data is processed.