Last update 25-Apr-2004
William E. Daehn
|Originally published in The Celator, Volume 5, No. 8 (August 1991), pp. 28-33.|
|© 1991 William E. Daehn; used by permission of the author and the publisher|
“Students should be warned that some scholars have been over-optimistic in estimating their own ability to interpret the surviving evidence correctly.” So concludes the entry for “Metrology” in Jones’ Dictionary of Ancient Greek Coins.1
In reviewing the existing literature on the subject of the origins of coin weight standards, the reader encounters a confusing and often contradictory array of theories. This paper summarizes various theories that have been put forward to explain the origin and spread of Greek coin weight standards, and describes a technique used by numismatists to analyze coin weights.
The search for the origin of coin weight standards must begin long before the invention of coinage. Gardner identifies three stages through which commerce passed in the development of coinage: a pre-metallic stage in which widely recognized items of value were exchanged; a metallic stage in which precious metals were exchanged by weight through the use of scales; and a third stage in which precious metal units of fixed weight, but without any official mark, were used.2 When these units were stamped with a mark of authority to guarantee their weight, coinage was born.
In the pre-metallic stage a barter system was prevalent, whereby surplus items were traded for items needed. Although Arthur Burns suggests that simple barter probably never existed: “units of value must have arisen as soon as any kind of commerce, and depended in each district on the economic environment.”3
It is when commerce entered Gardner’s second stage, the metallic stage, that a standard system of weights became vital. Burns points out “gold and silver were so valuable, and small errors so important, that ancient people measured them by weight, the units being defined in certain well-known seeds found to be uniform in weight.”4
The true benefit in a system of coinage lies in its widely recognized standard weights. This frees the merchant from the hassle of weighing the metal prior to concluding each transaction. But Macdonald reminds us “it is probably also true that the custom of appealing to the balances in private transactions enjoyed a much longer life than one is sometimes disposed to credit it with.”5 Gardner also points out that scales may have been used “and light coins taken only at a discount.”6
Still, standard weight systems were an essential element in the development of coinage. Where did these standards originate? At this point the theories begin to diverge. One theory, which I’ll call the “Eastern” theory, is most thoroughly described by Barclay Head in his introductory chapter to Historia Numorum.7 This theory can be summarized as follows:
- The weight systems of the ancient Babylonian and Assyrian empires were based on the talent containing 60 minae and the mina containing 60 shekels.
- Two “common” standards existed: a light standard with a mina of 491 gm and a heavy standard with a mina of 982 gm (the origin of these weights is disputed, but this need not concern the numismatist).
- There also existed various “royal” minae showing incremental weight increases (ranging from 1/36 to 1/12) over the common mina, both heavy and light.
- When weighing precious metals, however, a special mina of only 50 shekels was used.
- Silver coin weights were derived from these various gold weights based on a gold-to-silver value ratio of 13 1/3:1.
The most important derivative from this system starts with the royal mina of the light standard (491 gm). This weight was multiplied by 50/60 to convert it to the precious metal mina, and then was increased by 1/36, giving a royal gold mina of 420 gm. This mina was then divided by 50 to get a gold shekel weighing 8.4 gm (130 grains). With a gold-to-silver ratio of 13 1/3:1, the silver equivalent of this gold shekel would weigh 112.1 gm. Dividing this weight into a convenient quantity of 10 units results in a silver shekel of 11.2 gm.
Running through similar calculations using the gamut of Babylonian standards (e.g., common-light, common-heavy, and the numerous variations of the royal standard) one can derive many different silver weight standards. Proponents of this “Eastern” theory suggest that the weights thus derived had been transmitted westward into Lydia and Syria, and then by means of Phoenician trading ships to Greece, Italy, and Sicily long before the introduction of coinage. Accordingly, these weights became the basis for all, or nearly all, the different standards of the Greek coins.8
In the first edition of Historia Numorum published in 1887, Head describes various routes through which these standards spread.9 Most importantly, the 8.4 gm (130 grain) Babylonian gold shekel passed over the sea to Euboea, and became the basis for the Euboic, Attic, and Corinthian standards.
Head also believes a heavier standard used in Phoenicia found its way to Aegina, but only after centuries of degradation had lowered its weight. It was adopted by the Aeginetans and became the basis for the bronze and iron spits used as a medium of exchange before coinage was adopted.10
Head qualifies his acceptance of these theories by suggesting that the transmission of standards may in part have come through Egypt, Cyprus, and Crete during the Minoan period rather than directly through Phoenician commerce.11
An alternate theory was expounded by William Ridgeway in his book The Origins of Metallic Currency and Weight Standards, published in 1892. Ridgeway rejects the Babylonian origins of Greek weight standards and their transmission through Phoenician trade.12 He also disagrees that the Aeginetan was a degraded Phoenician standard. Degradation of this sort did not take place among the Greeks, he argues. Aegina relied heavily on trade with other cities and could ill afford the consequences of a reduced-weight currency.13
In Ridgeway’s view, the ox was universally employed as a unit of value throughout Greek lands; gold was also present throughout the same area; gold was the first commodity to be measured by weight and was weighed by the same standard (measured in seeds) throughout this region; this standard unit of gold was, in all cases, regarded as equivalent in value to an ox or cow.14
“When gold came to be substituted for goods or cattle as the accepted measure of value, the weight of the gold unit or stater naturally came to depend in each district upon the recognized value in bullion gold or gold dust of the principal unit of barter in that district.”15 Because Ridgeway believes all peoples would have valued the ox at precisely the same amount of gold (8.4 gm), wherever the ox had been the measure of value, the gold stater would be of the same weight.
“The development of the various silver standards . . . depended according to Professor Ridgeway, upon the market value in each district of silver in relation to gold.”16 For example, the Aeginetic standard he believes was derived based on a gold-to-silver value ratio of 15:1 giving a silver stater of 12.6 gm (derived as follows: 8.4 gm gold unit × 15 divided by 10 coins = 12.6 gm).
The silver standards thus derived, are related not to the standards of other cities which had been transmitted through trade, but originate from local traditional means of exchange, usually the ox. As an examination of the silver coins shows, weight standards exhibit differences not merely between one city and another, but even between one period and another in the same city. These various and fluctuating standards Ridgeway attributes to the Greeks’ “endless quest after bimetallism.”17
Bimetallism is a system in which “two metals, usually gold and silver, are both made legal tender; and a fixed relation between their values is fixed by law.”18
As evidence, Ridgeway cites a 4th century BC passage from Xenophon: “gold, whenever it turns up in quantity, becomes on the one hand cheaper itself, and on the other makes silver dearer.”19 Therefore, as the gold-to-silver value ratio changes in accordance with changing supplies of these metals, “either the silver currency must undergo certain modifications in order that a definite round number of silver units may be equal to the gold unit, or on the other hand the gold must undergo modification.”20 But since the gold unit remained unchanged down to the time of the Ptolemies, whatever changes were necessary must have taken place in the silver units.
Ridgeway’s bold theory set off a storm of debate. An immediate rebuttal came from Barclay Head in the form of a review of Ridgeway’s book published in The Numismatic Chronicle.21 Head rejects the idea of indigenous origins for some of the Greek weight standards suggested by Ridgeway. Ridgeway’s theory that the ox was universally accepted as worth 8.4 gm of gold Head finds “not quite convincing.”22 As for the constantly changing silver standards being driven by fluctuation in the relative market values of gold and silver, Head states, “In all this I confess that I am unable to follow Professor Ridgeway . . . ”23
Head does, however, admit that Ridgeway’s book “has done much to shake my faith in the time-honoured theory that all Greek weight standards were imported from the East . . . ”24 And shaken he was! For when Head published the second edition of Historia Numorum in 1911, he found himself accepting some of Ridgeway’s theories.
Head still believed in the eastern origins of Greek weights, but was now ready to admit “the endless modifications of the original Babylonian and Phoenician silver standards . . . can be sometimes more naturally accounted for on the theory propounded by Prof. Ridgeway” that the Greeks’ quest for bimetallism caused frequent weight changes.25
Regarding the Aeginetic system, he still mentions a possible Phoenician origin, but now considers it “doubtful” and refers to Ridgeway’s theory (a local gold origin and a 15:1 ratio) as “perhaps the true one.”26
Shortly thereafter, in 1918, Percy Gardner published A History of Ancient Coinage, devoting much discussion to weight standards. Gardner rejected both the eastern origins of all Greek standards and the Minoan influences suggested by Head.27 Gardner says the precious metal mina of only 50 shekels described by Head never existed, and the idea of fractional increments of the “royal” mina is also dubious.28 Regarding Ridgeway’s theory (now accepted by Head) that changes in silver coin weights were driven by changes in the gold-to-silver ratio, Gardner states, “this theory I hold to be, save in a few instances, quite baseless.”29
Gardner sets forth his view “that there were three chief original monetary systems in the Greek world, whence all, or almost all others were derived:
- The gold system, exemplified in the gold staters issued by Croesus and the Persians.
- The silver system, exemplified in the silver staters issued by the people of Aegina.
- The bronze system, in partial use in the Greek cities of Italy and Sicily, and probably derived from the original inhabitants of those countries.”30
The gold system described by Gardner originated in the Babylonian system and is based on the light standard gold stater of 8.4 gm (130 grains), as well as the heavy standard Phoenician coin of 16.8 gm (260 grains). These correspond to 1/60 of the Babylonian mina, and were probably equivalent to the value of an ox.
The Aeginetic standard originates with the bronze or iron spits which were a traditional form of value in central Greece. The weights of the spits were not imported through Phoenician trade as Head originally suggested, and had no relation to a gold unit, as Ridgeway claimed. The standard in Aegina was iron – not gold. Gardner also doubts the 15:1 gold-to-silver ratio which Ridgeway used to derive his Aeginetic silver stater. Instead, the Aeginetans decided to strike silver coins which represented the value of these spits.31 Charles Seltman says the silver-to-iron value ratio used for this conversion was 400:1.32 The iron spit was converted to the silver obol, and a handful of six spits was converted to the silver drachma.
The bronze system was based on a pound of bronze (libra) which was the unit of value before the Greeks arrived in Italy and Sicily. “The Greek population issued silver coins which circulated side by side with the native bars and ingots of bronze which continued in use. In order to encourage the use of the silver coins . . . some were made so as to be equal in value to the old bronze unit of value.”33
Later authors have not added much to the debate, and in most cases they steer clear of the subject entirely. Perhaps the situation was best summarized by Arthur Burns: “The attempt to trace each silver unit in the Greek world to a local origin in the earlier unit of value . . . has failed. The only certain influences at work seem to have been a tendency to select units of a convenient size, and then a tendency for the unit to shrink. Shrinkage of the unit is what has often been explained as the adoption of a new standard.”34
Regardless of where the standards originated or how they were transmitted, many standards resulted. The most prevalent are summarized in Figure 1.
|Name of Standard||Standard Unit||Weight (gm)|
|Achaean||stater (3 drachma)||8.0|
|Corinthian||stater (3 drachma)||8.6|
|Lycian||stater (didrachm)||8.6 to 10|
|Persian||siglos||8.35 or 8.55|
|Rhodian||tetradrachm||13.2 or 15.2|
Adapted from Jones, Dictionary; and Kraay, ACGC. All silver unless noted.
Fig. 1: Greek Coin Weight Standards
It is important to stress that these standards were not always strictly adhered to in practice, and were frequently changing. Nevertheless, the study of weight standards provides a useful tool to historians and numismatists.
For the historian, the examination of the weight standards employed at Greek cities can shed light on the flow of inter-city trade in the ancient economy. As Sutherland observes, “the rival activity of great commercial blocs emerges with extraordinary clarity from consideration of weight-standards, and the comparative distribution of standards at different times shows very significantly when the prosperity of a particular commercial bloc waxed and waned.”35 We know that Alexander the Great adopted the Attic standard for his vast coinage. Certainly, “the great prosperity and political importance of Athens in the 5th century contributed to the widespread popularity of this weight standard”36 and resulted in its adoption by Alexander.
For the numismatist, a knowledge of coin weights can help to classify a coin chronologically as well as help to spot counterfeit or plated coins. A cursory glance at the coin weights listed in the catalogs of published collections will show that the weights of individual coins do not always match the theoretical standard weight. Because these weights vary, and the evidence attesting to the origin and development of standards is open to debate, how do we know what the standards really were?
The best method for determining the weight that the ancient minters were striving to attain is to examine the coins themselves. If we calculate the average (mean) weight for a large sample of coins of the same denomination issued by the same city, we could get an estimate of the standard weight employed in that city at that time.
A sample of 142 Athenian tetradrachms minted during the archaic and classical periods was taken from catalogs of published collections. The range of observed weights was 15.56 gm to 17.99 gm.37 The mean weight was 17.02 gm. Based only on this information, it appears that the standard weight was about 17 gm. However, the excessively high or low weights found in any large sample can distort the average and we cannot be sure that the average weight equals the standard.
The normal weight for a tetradrachm struck on the Attic standard is actually about 17.2 gm. The slightly lower mean weight observed in this sample is most likely due to the amount of wear which the coins suffered during their period of circulation. The numismatist George F. Hill recommends increasing the observed mean weight by 1% to compensate for wear.38 Adding 1% to the observed mean of 17.02 gm results in an estimated standard weight of 17.19 gm, just as expected.
A better method of determining the standard weight is the frequency table (or histogram) recommended by Hill.39 The frequency table is a chart which shows, along the horizontal axis, the range of weights observed in a sample. Along the vertical axis is plotted the number of coins found to equal the weights shown on the horizontal axis. This method allows the researcher to not only see the range of weights, but to see the weight most often achieved by the minters. This maximum frequency usually represents the standard weight.40
The frequency table for a sample of 202 didrachms from Velia is shown in Figure 2. The range of observed weights was 6.34 gm to 7.98 gm. The mean weight was 7.39 gm. The maximum frequency of occurance is 7.5 gm, slightly higher than the mean and matching the Campanian standard on which these coins were struck. As can be expected, more specimens fall below the maximum frequency point than above that point. This contributes to the fact that the mean is lower than the standard weight.
Fig. 2: Velia Didrachm
Figure 3 is the frequency table for a sample of 131 staters from Aegina struck on the Aeginetan standard of 12.2 gm. The range of observed weights was 11.31 gm to 12.73 gm. The mean weight was 12.13 gm. The mean plus 1% is 12.25 gm. As the table shows, the maximum frequency is found at 12.2 gm suggesting that this is a more reliable measure of the standard, free from the distortions of abnormally high or low weights.
Fig. 3: Aegina Stater
Another important advantage of using the frequency table rather than relying on the arithmetic mean is illustrated in Figure 4. This is a table of the observed weights of 55 tetradrachms of Rhodes struck on the Rhodian standard. The range of observed weights was 12.78 gm to 15.34 gm. The mean weight was 14.36 gm. However, we would be mistaken in assuming this is close to the standard weight. As the frequency table clearly indicates, specimens are clustered in two groups: one around the maximum frequency of 13.3 gm, and one around the maximum frequency of 15.1 gm. This clearly points out the use of two different standard weights. Indeed, the standard weight of Rhodian tetradrachms was reduced circa 304 BC from 15.2 gm to 13.2 gm. Because coins struck on a lighter standard tend to be later in date than those struck on a heavier standard,41 the frequency table not only points out changes in weight standards, but can also help to establish the relative chronology of the coins under examination.
Fig. 4: Rhodes Tetradrachm
In attempting to determine the standard weight for a series of coins, it is important to measure the weights of the main unit (e.g., stater or tetradrachm) rather than that of the fractional denominations. An examination of the minor coinage often shows these coins to be of slightly lighter weight than that called for by the standard.
An examination of the coins of Corinth confirms this. A sample of 159 Corinthian staters showed a maximum frequency of 8.5 gm (just 0.1 gm short of the 8.6 gm Corinthian standard). However, Figure 5 illustrates a sample of 124 Corinthian drachms with a maximum frequency of just 2.5 gm, well short of the 2.9 gm (one-third of the 3 drachm Corinthian stater of 8.6 gm) expected of coins struck on the Corinthian standard.
Fig. 5: Corinth Drachm
Probably the low weight was intentional and was meant to cover the cost of smelting, refining and coining. This minting charge “may have represented a larger proportion of the value of a small coin than of a larger one.”42 After all, a slightly underweight minor coin probably did not concern the users of coins, for as Kraay points out “small change may have been accepted on trust in local retail trade, whereas the larger pieces had to serve the needs of inter-city commerce”43 and therefore had to be nearer to full weight to avoid rejection in the marketplace.
Even for the principal units, which on average tend to be of full weight, a range of weights is evident. Again, to a certain extent, low weights may be intentional. “By keeping the intrinsic worth of a coin a little below its nominal value, the authorities made it more profitable to retain it as a coin than to put it into the crucible.”44 This practice may have led to gradually changing standards. As Hill points out, “The standard as thus reduced might be regarded as a new standard. Another state copying the standard in this reduced form, and making its own reduction, would bring about a further fall in the standard.”45
Various factors contribute to the range of weights which have been observed in the samples. Sometimes “the blanks for gold and silver coins were made individually, and thus the checking of their weight at that time al pezzo, i.e. piece by piece, is altogether likely.”46 However, a careful weighing of the silver for each blank in granular form would have been too time consuming for wide-scale use.47
More commonly, the only check on the weights was “al marco, i.e. by a guarantee that a specific number of coins were made from a specific quantity of metal, but with acceptance of weight fluctuation of any individual coin in one or the other direction from the norm, always provided that the overall total of the coins of a given issue should be equal to the weight of the ingot intended for their manufacture.”48
Although this “al marco” method did result in inconsistent weights, a recent experiment by David Sellwood concluded that “the blanks for silver of the tetradrachm size could easily and efficiently be obtained by pouring from small crucibles into open moulds” and that “only a little practice would be needed” to attain satisfactory results.49
The resulting variation is usually within a narrow range as the data discussed above indicates. Because counterfeit coins are often excessively low in weight, an accurate scale is a valuable tool for the coin collector. A coin whose weight falls outside the range of weights observed in a large sample of known genuine coins may be regarded as suspect.
The collector with an aptitude for statistics (or access to a computer) may want to calculate the standard deviation from the mean weight for an observed sample. Assuming a normal distribution of specimens around the mean, 99% of the specimens should be within plus or minus three standard deviations from the mean of a large sample.
In the sample of staters from Aegina plotted in Figure 3, the standard deviation was 0.24 gm. Adding or subtracting three standard deviations to or from the mean weight of 12.13 gm gives a range of 11.41 gm to 12.85 gm. Any coin falling outside this range should be examined carefully. This sample of 131 coins showed two coins (1.5% of the sample) which fell below this range, although this may be attributable to an excessive amount of wear on these coins.
A basic understanding of coin weights and weight standards not only lends insight into the workings of the ancient economy, but also gives the numismatist a greater appreciation for the technical problems faced by the ancient minters, and opens up new avenues for exploring the coins he collects.
1 John Melville Jones, entry for “Metrology,” A Dictionary of Ancient Greek Coins (London: B.A.Seaby Ltd., 1986), p. 143.
2 Percy Gardner, A History of Ancient Coinage 700-300 BC (Oxford: Clarendon Press, 1918), pp. 21-24.
3 Arthur R. Burns, Money and Monetary Policy in Early Times, 1927 (rpt. New York: Augustus M. Kelley, 1965), p. 442.
4 Ibid., pp. 442-443.
5 George MacDonald, The Silver Coinage of Crete: A Metrological Note, 1919 (rpt. Chicago: Obol International, 1974), p. 23.
6 Gardner, p. 56.
7 Barclay V. Head, Historia Numorum, second edition, 1911 (rpt. Chicago: Argonaut, 1967), pp. xxxiv-xxxix.
8 Ibid., p. xxxix.
9 Barclay V. Head, Historia Numorum, first edition (Oxford: Clarendon Press, 1887), pp. xlvii-xlviii.
10 Ibid., p. xxxviii.
11 Head, Historia Numorum, second edition, p. xxxix.
12 William Ridgeway, The Origin of Metallic Currency and Weight Standards (Cambridge: The University Press, 1892), pp. 226-228.
13 Ibid. p. 224.
14 Ibid., p. 124.
15 Barclay V. Head, review of William Ridgeway’s Origin of Metallic Currency and Weight Standards in The Numismatic Chronicle, Third Series Vol. XII (London: Royal Numismatic Society, 1892), pp. 247-250.
16 Ibid., p. 248.
17 Ridgeway, p. 338.
18 Gardner, p. 53.
19 Ridgeway, p. 338.
20 Ibid. p. 339.
21 Head, review of Ridgeway, pp. 247-250.
22 Ibid., p. 248.
23 Ibid., p. 249.
24 Ibid., p. 249-250.
25 Head, Historia Numorum, second edition, p. xxxix.
26 Ibid., p. 395.
27 Gardner, pp. 26-27.
28 Ibid., p. 25.
30 Ibid., pp. 28-31.
31 Ibid., p. 115-117.
32 Charles Seltman, Greek Coins, 1955 (rpt. London: Spink & Son Ltd., 1977), p. 37.
33 Burns, p. 231.
34 Ibid., p. 222.
35 C. H. V. Sutherland, Ancient Numismatics: A Brief Introduction (New York: American Numismatic Society, 1958), p. 21.
36 David R. Sear, Greek Coins and Their Values, Vol. l: Europe (London: Seaby Publications, 1978), p. xxix.
37 In compiling the data on coin weights, I excluded broken coins and plated coins (fourrees) found in the published collections to prevent distorting the statistics with items known to be of less than standard weight.
38 George F. Hill, “The Frequency Table,” The Numismatic Chronicle, Fifth Series, Vol. IV, (London: The Royal Numismatic Society, 1924), pp. 76-85.
40 For a detailed discussion of the use of frequency tables, see Stephen N. Cope, “The Statistical Analysis of Coin Weights by Computer and a Rationalized Method for Producing Histograms,” The Numismatic Chronicle, Seventh Series Vol. XX (London: The Royal Numismatic Society, 1980), pp. 178-184. Alternate methods of graphically analyzing coin weights are discussed by Warren W. Esty, “Percentile Plots and Other Methods of Graphing Coin Weights,” The Numismatic Chronicle, Volume 149 (London: The Royal Numismatic Society, 1989), pp. 135-147.
41 Jones, entry for “Weight Standards,” Dictionary, p. 241.
42 Ibid., p. 240.
43 Colin M. Kraay, Archaic and Classical Greek Coins (Berkeley and Los Angeles: University of California Press, 1976), p. 9.
44 George F. Hill, Ancient Greek and Roman Coins: A Handbook, 1899 (rpt. Chicago: Argonaut, 1964), p. 31.
46 A. N. Zograph, Ancient Coinage Part I: The General Problems of Ancient Numismatics, transl. H. Bartlett Wells (Oxford: British Archaeological Reports, 1977), p. 55.
47 David Sellwood, “Some Experiments In Greek Minting Technique,” The Numismatic Chronicle (London: Royal Numismatic Society, 1963), p. 225.
48 Zograph, p. 55.
49 Sellwood, pp. 225-229.