X-ray Imaging to Identify Ancient Fakes

As long as there have been coins, there have been forgers. A corollary to this fact is that any collection of sufficient size will contain some fakes. Many ancient fakes used a false core of copper with a silver or gold casing to pass as an authentic coin¹. These fourrées (French for ‘stuffed’) are often hard to visually distinguish from authentic coins given the skill of the forgers or their illicit use of official dies. In ancient times this led to the use of test cuts to physically open or cut coins to reveal their inner metal quality.

Collection of modern and ancient coins.

X-raying coins under increasing energy levels.

To verify the authenticity of our collection, the Madison Art Collection has often partnered with the Madison Accelerator Lab to X-ray objects to verify their inner structure. When applied to coins, fourrées and authentic coins will “respond” differently under X-ray as they have differing material composition. Objects made of lighter atomic mass, such as copper, will permit X-rays to pass more easily as compared to materials with heavier atomic mass, such as silver. Applying this principle to coin analysis, if two objects are roughly the same size and weight, an object that is mostly copper, such as a fourrée, will “disappear” more quickly under X-ray imaging than objects of solid silver.

We can see this behavior dramatically on the right in Figure 1. The modern US coins on the left (mostly made of copper, tin, or light amounts of silver), and the known fourrées on the right, rapidly “disappear” as the energy level of X-rays increases. However, the “good” denarii in the middle last longer and remain until the X-rays reach their maximum energy levels.

The results from this experiment are clear. The known fakes, Didrachm of Cales (item D) and Denarius of Augustus (item F), were re-confirmed to be fake. However, a suspected Denarius of Augustus (item E), was verified as authentic!

The Madison Art Collection is grateful for the assistance of Dr. Harold Butner, Professor of Physics, in conducting these experiments and Dr. Adriana Banu, Professor of Physics and Scientific Director of the Madison Accelerator Lab, for continued access to this crucial equipment.

For those with additional interest in the numismatics or science behind this research, please continue reading below!

Table of Contents

• What Are Fourrées?
• Use of X-rays to Identify Fourrées
• Experimental Setup and Results
• Summary

What Are Fourrées?

The imitations we wish to discover are fourrées (French for ‘stuffed’), which may appear as authentic coins but are actually a copper core wrapped in silver. These can often be challenging as they are intended to be of the correct weight, and at times may have been made with stolen official dies. A cross-sectional diagram for these coins is shown in Figure 2 below along with examples of a known fourrée and solid denarius:

Figure 2: A cross-section of fourrée and good coins, along with examples

Use of X-rays to Identify Fourrées

The fundamental tool in our analysis is X-rays generated by a medical imaging device. The same machine used to identify broken bones can be used for broken coins too. The central idea is that when an X-ray strikes an atom, there is a chance the X-ray is absorbed based upon the atomic number of the object and the energy of the X-ray. This absorbed X-ray will not appear on the detector and will show up as “white” in the image. This is why a person’s bones are shown in an X-ray as they are more “dense” than the surrounding tissue. In general, the probability of absorption \(\approx{} \frac{Z^3}{E^3}\), where \(Z\) is the atomic number of an element and \(E\) is the energy of the X-ray.

Applying this principle to our ancient coins, if two coins are of similar mass and are exposed to equal X-ray energies, then the one composed of “lighter” atomic material will pass more X-rays and thus appear more transparent in the X-ray image when compared to the “heavier” one. Thus, a good denarius of solid silver (Ag with atomic number 47) should be less transparent than a fourrée filled with copper (Cu with atomic number 29). As energy increases, however, both materials will eventually be unable to stop the X-rays. The figure below shows a cross section of the X-ray behavior for each object.

This behavior can be seen most clearly with coins of vastly differing compositions. Figure 4 below shows how various modern US coins respond under increasing X-ray intensity. The three coins, from the top, going clockwise, are a 1907 Barber Dime (2.5g at 90% Silver), Buffalo Nickel, (5g at 75% Copper and 25% Nickel), and a 1943 Jefferson Nickel (5g at 56% Copper, 35% Silver, 9% Manganese). These are shown as “A” in Figure 1.

Given their different compositions, we should expect the “lighter” coins (both in mass and atomic composition) to permit X-rays to pass more easily. As seen in the images, even though the Barber Dime is made of silver (a heavier material), its small mass of only 2.5g easily allows X-rays to pass through at medium energies, whereas the Buffalo Nickel at 5g still remains. Finally, at the highest energy, only the Jefferson Nickel remains.

Collection of modern US coins

Response under low X-ray

Response under medium X-ray

Response under high X-ray

Figure 4: Modern US coins under various X-ray levels

Experimental Setup and Results

Applying these methods to our ancient coins, we selected several “good” denarii² as control objects from the Curator of Coins personal collection (Items B and C) and several suspected fourrées from the collection (Items D, E, and F). The coins are shown in Figure 5 and enumerated in the table below:

• Item B: Denarius of Sicinia, 3.50g
• Item C: Denarius of Neria, 3.49g
• Item D: Didrachm of Cales (2024.2.5), 6.80g
• Item E: Denarius of Augustus (2024.1.54), 3.78g
• Item F: Denarius of Augustus (2024.1.46), 3.33g

Following from our examples with moderns, our suspected fourrée should be “lighter” than good coins, and thus more optically transparent to the X-rays. Assuming the coins are genuine, we should expect that all the denarii (Items B, C, E, and F) will respond in a similar manner. The didrachm does not have a control coin, but as it weighs nearly twice that of the denarii, it should permit fewer X-rays than the denarii. Figure 5 below shows our experimental setup and the response of each coin under X-ray. As a reminder, black on the image is where an X-ray has passed through the object, and white where it has not.

Ancient coins to be tested

Low energy X-ray

Medium energy X-ray

High energy X-ray

Examining the response of the various coins, it can be seen that two suspected fourrées (Items D and F) permit more X-rays to pass as compared to genuine denarii (Items B and C). While they all should be solid silver, this disappearing act confirms they are in fact fourrées with copper cores. Item E was suspected to be a fourrée given its condition, but as it responds similarly to our control denarii, it is in fact good! Finally, our didrachm is convincingly a fourrée as it is nearly twice as massive as the good denarii (Items B, C, and E) but rapidly becomes transparent.

Summary

Within any collection there are likely to be fakes. The ones that are far too small, light, or simply do not match the known types are likely to be identified quickly. For others, a more detailed numismatic or scientific analysis might be needed. While not everyone will have access to an X-ray machine, the methods employed here provide powerful complementary evidence to assess the validity of coins.

Footnotes