Class 1 Or Class 3 Counterfeit? Please Judge!

In this case, what is proven is that the coins were made of silver refined (likely from Mexico or South America) no later than the first half of the 19th century. However, the SG also shows in one case that the coin is NOT solid silver.

Remember one fact - Class 2 counterfeits were made in the first half of the 19th century and they do contain essentially perfect silver if they were made using Mexican silver. These Class 2 counterfeits can often be identified visually - the XRF tests proves the age and means they are NOT Class 3. But the XRF test does NOT mean they are GENUINE.

I believe as I said before that you are likely dealing with two varieties of Class 2 counterfeits one made to a lower standard than the other. The real question was when they were made. So they are unlikely modern.

They were unlikely made for interior China because they are not Carolus issues, but I have very recently discovered a clue (unsupported as yet) that would indicate that Ferdinand coins were also made for other markets to take advantage of a lower premium over silver melt that was present in other areas. That is a different story one too new to go into here. It is in the working theory phase. I have been pursuing this line of research because I have several coins that appear to be Class 2 silver types but which are Ferdinand designs.

To return to my theory that both coins actually fall into a counterfeit category we need to be clearer about what XRF tests provide.

XRF does not penetrate into the core of the coin so the coin may have been made using a technique that leaves good silver at the surface.

There are two primary methods.

The Sheffield process is one such method. The core of a Sheffield coin (usually nearly pure copper) is covered with thin layers of silver. That surface layer is coin silver of the era and will test properly if Mexican silver was used. (If Mexican silver was used at a secondary location, this postulates conditions similar to the Mexican mints when melting and rolling the silver so as not to remove volatile elements).

The second and more likely method that could have been used is a debased silver alloy which is pickled AFTER the coin is struck to raise the surface % of silver. Pickling removes the surface copper from the coin and alters the color and composition of the upper layers. That is the same part of the coin that XRF measures. The heavy trace contaminants are unaffected by pickling resulting in a test showing a shift in Ag and Cu ratios but not the contaminants. Pickling can of course be overdone removing too much silver and resulting in an underweight coin which also will look sea salvaged because of loss of metal.

So that means XRF can be fooled and is worthless. Right? WRONG.

What I did not see in the XRF test results you posted are any other trace contaminants like Tin or Yttrium or Cadmium. The test results of coin A add up to 99.08% and for coin B 99.09 %. Therefore something is missing from both coins - almost 1%. XRF can be done to different levels and can detect major as well as VERY MINOR components.

XRF tests are also limited by other factors and some lightweight metals and all non-metals are not detectable.

This is where the high tech science comes in.

Were there any trace contaminants in the 20 - 80 ppm range? These are the ones that tend to identify modern silver in the mix. They are what is left in the 0.001 part of the 0.999 fine silver bars from present day refineries.

This inability to see into the core is remedied if the coin has a test cut. The readings in the test cut are often totally different than XRF tests of the surface. Worn areas at the high points are also great to look deeper into the coin. They often show completely different percentages.

This is why I have wondered if it would not be advisable to core drill every coin tested to get a sample of the interior metal BEFORE it is encapsulated as genuine. The drill hole in the edge would permanently mark a coin and prove it was authenticated. IT would prevent re-holdering and the fraud that goes with it. The hole could be very small - exactly like museums do with all ancient statues to prove authenticity. The core sample would be more accurate and could be tested in several different ways. SEM analysis of the crystals provides a lot of good information.

XRF surface tests only go so far, but they can still be used to exclude the majority of Modern fakes by the very minor trace contaminants.

The principle is that to make a fake today you need to measure elements at XRF levels of accuracy to get the alloy right. If a forger takes "pure" silver and "pure" copper the trace contaminants will be different today than they were in 1800. Pure silver in 1800 meant 95-98% silver 2-5% other metal. By the second half of the 19th century it meant 99% silver with 1% other metal. By the turn of the 20th century 99.9%.

Today if you pay extra you can get 99.999% fine silver BUT what do you mix with it? Who produces 99.999% fine copper? The answer is NO ONE DOES. Most copper today has nearly 1% trace contaminants. What those are varies and they do not match elements seen in 1800 refined Mexican copper.

If the modern silver used by our theoretical forger was produced from silver salvaged from electronics components there are a couple rare earth elements used to increase conductivity that are VERY difficult to refine out of the silver. They are usually seen even in the 0.0001 remainder of the purest silver available.

So today there is a problem even in the best metal labs which precludes cheaply manufacturing an identical alloy. The more it costs our forger to make his coins the fewer he will make. Mass production of coins will be discoverable by XRF testing.

I know many people are now thinking just melt old coins. Perfect source right? WRONG AGAIN.

Melting the coins will in fact alter the alloy by driving off the volitile elements that are always seen in the originals. Also depending on the crucible composition and even furnace linings there are contaminants that can re-enter the alloy in the refining and melting process.

To make an 1800 coin precisely would require using all of the melting and processing methods of the time period. This gets very costly and would slow the process to a trickle.

In this discussion, I am getting ahead of the present capabilities of XRF because of the small data base. However, if you view this as the beginning step in a long process - it will not be to long before a good database exists - then if core sampling ever catches on and lighter elements not now picked up by XRF can be included - a whole new range of trace contaminants including carbon will be available.

I see this as just the start of bringing counterfeit identification into the 21st century. It is also the only way we will be able to exclude counterfeit coins in the future. As digital engraving improves, I can see a day when an absolutely perfect die can be re-created digitally. If that is done all the visual clues will disappear and we will have only the metalurgy left for positive proof.

I am very optomistic that Counterfeit Detection can stay one step away from the technology of forgery. Sometimes we are one step ahead, sometimes one step behind but always close.

That is the real "fun" of doing this.

The other point XRF is just a compositional tool Henry - we can't really make definitive conclusions on origin with these results ... we can only SUGGEST ... in my Chihuahua paper I have reported a good Ag/Pt/Au/Pb 1811 Chihuahua cast with no regal edge (i.e., circles & squares) with the same BASIC composition as other 1811 regal edge pieces. So in my conclusion I have SUGGESTED that for the first year of the 1811 Chihuahua casts there may have been a short lapse in time (i.e., several months) until the regal edge was applied to ALL 1811 pieces as an extra anti-counterfeiting device other than the T and Pillar of Hercules countermarks ... Henry ... all I can do is ... SUGGEST.

John Lorenzo
United States

realeswatcher You are right in pointing to the obvious problem here. We are getting sidetracked in the details. We need to focu on the primary question GENUINE or FAKE.

The 1820 coin with the low SG is most certainly NOT a Class 2 Silver restrike made for the China trade at least with the Interior parts of China. The SG also proves it is a fraud (either mint produced or counterfeit). But both coins also have INCORRECT EDGE DESIGNS. That would indicate both coins were made OUTSIDE the mint by a counterfeiter. The coin with the right SG should have the mint edge if it was made in the mint.

Both coins are also clearly transfer technology copies so they have to have been made AFTER 1830.

That much is certain.

So, to re-cap, we can rule out GENUINE first. Neither coin can be a genuine coin. One is debased and NEITHER has the correct edge. By definition, we have to also rule out Class 1 since BOTH COINS USE A TECHNOLOGY THAT DID NOT EVEN EXIST UNTIL ABOUT 1830.

At this point we have ruled out both Class 1 and Class 2. This leaves us for present with Class 3.

In my opinion, we are dealing with a VERY EARLY Class 3 coins which may in time to come be placed into a Class 2B status. But we are ahead of ourselves AGAIN.

The reason I say this is because there are other facts that I mention in the book that are pertinent but not well known. In 1865, the Portrait dollar was adopted as a new trade coin and standard in the Straights Settlements. To do this a supply of Carolus and Ferdin dollars was needed.

Where did those "new & unchopped" coins come from?

Supplying silver in a portrait form was a good method of disposing of silver when there was a world glut following the Comstock discovery of 1859 and the European shift to gold from silver after 1854 (culminated in Germany dumping silver in 1873).

China and the trading partners of China were the only places in all the world which retained a significant interest in silver as money.

In the interior of China, only Carolus dollars were accepted at a HUGE PREMIUM. But on the coast Ferdinand dollars and even the Eagle dollars of Mexico were accepted at a smaller premium.

There was, at all times, a small premium over silver paid for Ferdinand coins in costal China. The eagle coins were less popular but compared to US Trade dollars the premium always existed. The rate of the premium varied over time with a peak for Ferdinand near 10-14% but an average over time of about 4% which applied even to the Eagle dollars. That 4% may not seem like much today but it would at the time pay for the costs of fabrication and delivery from the US to China. (Delivery cost an average of 2.5 cents per coin leaving 1.5 cents per coin for manufacture.)

This meant that silver from the US mines which which had no market in the US or elsewhere as bullion or US dollars had a small premium in China and its neighboring countries provided the silver was turned into Mexican coins.

Because of these facts, I do believe that there was even a market for silver restrikes of the Ferdin and Eagle dollars - just not as easily proven a market because the margin was far smaller.

The remaining uncertainty over this part of the issue is why I decided to eliminate the silver Ferdin's from this book and hold them until some subsequent update after more data comes to light. I own about 2 dozen Ferdin coins that are silver counterfeits. I own at least as many Eagle dollars as well(Cap and Ray counterfeits made of silver).

I moved the Ferdin silver counterfeits to Class 3 to prevent possible disputes that could result over classifying them as either Class 1 or Class 2 along with the Carolus dollars.

I think the case for mass manufacture of the Carolus dollars in silver proves itself easily from the period sources, but the added market for the lower regarded Ferdin and Eagle dollars could cause added confusion and result in peripheral arguments that would detract from the main point that Class 2 coins DO REALLY EXIST.

These Ferdin dollars, like both of Henry's coins, are definitely fake. One has a very low SG and the other has a bad edge. So Class 3 is, FOR NOW, better than nothing at all and far better than calling them REAL.

That is the biggest hurdle - real silver of the right amount does NOT MEAN A REAL COIN. We are simply too new at XRF testing to tell if these 1820 dated coins were made in 1825, 1835 or 1865. We may be able to do this in the future but not now.

But they are both forgeries of some sort.

Remember, for most collectors this subject is new ground. Too much too soon especially of these side discussions lends nothing to the main point and sets up a smokescreen of side discussions that could damage the acceptance of the main point.

Time will clarify all of this. It has taken over 100 years to get the whistle blown. Lets not sweat the periperal details now.

I really have to question that s.g. result with a Ag reading at 94%. All Sheffields (i.e., silver plated) in a circulated state will normally show some degree of silver peeling particualrly along the rim areas. I also do three spot XRF shots on the surface to make sure there is not a high spot anomaly. Normally - regal legitimate pieces measure at this level of 92-96% Ag due to silver surface enrichment. In our study or from the 200-300 pieces that were XRF analyzed for the CC8R ANS book it was mainly an exercise on determining all the different alloys and to also see the Ag ranges of debased specimens with primarily Cu & Zn being substituted in the alloy for Ag. This specimen may be an oddity of sorts. This situation here explains one thing for sure why there are hundreds of 8R Portrait counterfeits sitting in PCGS/NGC/ANACS holders today. I have had other researchers indicating to me I need to recheck my analysis method or get a better scale ... its 94% Ag on the surface with good Pt/Au & Pb.

Actually its a project I am working on now ... actually for a business type presentation with another researcher here at EDAX for a Material Analysis trade show for Ametek - EDAX in August 2013 ... we are looking at the effects of the L.Beck silver surface enrichment process on struck coins using (2) examples. L. Beck did his study using recently made cast coins with various amounts of Ag/Cu. With regals I was getting results in the 92-96% range. So to explain this to E-Bay buyers - Bob you seen my coins for sale - I had to explain WHY the Ag readings were above 90% - in this EDAX study we will be looking at (2) coins. A 1964 90% Silver Kennedy half and a 2 Reales El Cazador shipwreck coin. We are analyzing both with XRF and SEM/EDS with the later seeing the microstructure of the core and surface. The first part was done and what we saw was NO! silver surface enrichment for the Kennedy half. It seems a half a century is not long enough to indicate any Ag enrichment or any detected Ag movement to the surface. In other words the Kennedy half read ~ 90%. XRF results for the El Cazador were 98% Ag on the surface but we have not cut it in half yet to examine core/surface (only $45 coin) with XRF and SEM/EDS. The point is I do not expect much Ag % increase at the surface for a die struck coin. I would be very surprised if this Henry coin were 94% Ag at the surface and 35-45% Ag at the core ... it just can't be from my analysis so far after 300-400 CC8R analyses ... from the assay the only possibility is the core is high copper and low silver ... if this was the case we would not see a coin toned in this matter or the weight would be low.

Pre-script - have you guys played around at all with actually melting/assaying coins?


That's a good point. Doesn't seem like there could any significant amount of a white metal hiding in there... something would show at the surface, no?

One thing, though, about the "or it would be low weight" part... I assume the coin is in transit right now. Has anyone noticed/tested whether this piece is appreciably thicker?

Anyway, SG error is an easy answer... but if he got 10.18 for "Coin A", which he presumably would have done at the same time b/c these coins were considered together... shouldn't whatever drastic error he was making have shown up in that figure as well? That is, unless that "Coin A" has a lot of lead in it and actually has a much higher SG itself...

John, a suggestion - you're obviously capable, perhaps you should verify SG as part of your testing. It's a necessary reading to have (accurate) if one is going to even bother going to the effort of doing XRF.

===========

On a separate note... if you're going to do a study on, in layman's terms, the degree of copper leeching... you should analyze some "in between" stages, I would think. Just common sense, what would really happen to a '64 Kennedy? It saw little circulation wear, and would for the most part be sitting in a fairly protected environment... and for not that long a period of time. Then, at the other extreme, you're taking a specimen that was acted upon by relatively severe conditions for 200 years... to the point where copper leeching is a known phenomenon on salvaged coins if only from seeing coppery spots on them.

I would be curious to know if anything happened to: a heavily worn Bust Half.... an untoned AU-UNC 200 year old piece (maybe a nice half R or 2R would be a good candidate, since the idea of how old a given 8R is is being thrown into question)... How about a crusted land hoard find? A heavily toned cabinet coin? How about a heavily whizzed old crown, where severe hairlines have affected the surface? Surely with a little effort, you guys could find some cheap enough specimens to sacrifice... Things like that would give various data points and variables to draw some conclusions from.

On that note - have you observed whether excessive wear does anything to alter the metals ratio you find with XRF? (anything to do with relative hardness of the component metals?)

Just a few words about the S.G. test and XRF.

I have done S.G. test for over 150 coins. Every time I do it, I use a 9999 silver as the 100%Ag standard and a Cartwheel 2d as the 0%Ag standard. Recently, I add a Canadian maple silver coin (pure silver) and a Cartwheel 1 d in-between my meaurements as a high and low controls. Others include supermarket bottled distilled water and an 2 decimal electron scale. I find an error of about 3-7% down from the target values in my system. I am very proud to (thanks to swamperbob's guidance) perform this test from Aug last year because I am able to confirm the Ag content of my coins but I also found some unexpected coins fake to me accidentally. This debased 1820 NG 8R(Coin B) is one of them. The only limitation of my system is doing a coin under 10 grams. The error incurred will be very big. Then after about two months, I bought Coin A (mainly for comparison purpose).

On Sep 15, I measured the S.G. of coin A and B in the same batch and A got S.G.10.185 while B got 9.659. I am absolutely sure Coin B must be a debased specimen. The ring tone test between the two coins show difference, they show no observable difference in volume to me.

S.G. presents a pure physical property of an alloy, it would involve errors in measurement but they are mostly predictable and sometimes avoidable. But for XRF, it involve a delicate instrument, would it incur some sort of irregularities during measurement? I don't know.

There are my worthless suggestions to XRF.

Is there a calibration procedure for the instrument? Any controls in your system?

If someone immerse a debased coin in a silver compound solution before XRF measurement, can you recognize, John? So do you think there is a process for washing the coin before measurement?

XRF measures the surface of a coin. Do you think we can make a deeper edge scratching (a few μm) and make this the area of measurement.

Through this thread and have my coins sent to John for XRF measurement, I understand more about XRF now. Thank you John and swamperbob. I am a bit disappointed XRF can't pinpoint the era of minting of a coin. It really doesn't? Is it a problem of insufficient data?

Anyway, XRF can give a more complete metal account in a coin and as least it shows the precious metal elements that won't be added by a modern counterfeiters. XRF still has its value so far.

Forgive my poor English, hope you can understand my meaning.

P.S. I am pondering and digesting the lengthy explanations of swamperbob so I will present my opinion of his points here.

Henry -
Coin B Analyses as you indicated: Coin B: Cu 3.11, Pb 0.45, Ag 94.4, Zn 0.02, Ni 0.01, Fe 0.08, Bi 0.05, Tin 0.27, Pt 0.89, Au 0.24. If you were to show me this analyses I would consider it a real 8R. Good Ag/Pb/Au/Pt. But as I indicated we now need to examine the coin. Bob is more experienced than me in this fashion. With Ag = 94.4 other than a Sheffield (silver plated) 8R I can offer no other explanation. It showed no signs of being a Sheffield (silver-plating). Based on the Pt/Au its not modern and the lead level shows no indication of it being re-melted. Copper the most common metal used in Sheffields as the INTERNAL debased alloy read only 3.1% so I would be shocked of a 20-40% Cu core! Copper also leaches to the surface. No harm was done to the coin on the edge. XRF analysis is totally non-destructive to the entire coin.

See this Sheffield:

http://www.ebay.com/itm/MEXICO-SPAN..._1334wt_1328

Note the extreme copper leaching to the surface. A 30-40% Cu core whether a Sheffield or not is going to give you some signs. I see NO signs - no surface copper toning. XRF = Cu = 3.4%. High weight. No signs of a mercuric/amalgam/silver washing.

Coin B: Its a 19thC v.high Ag forgery. Case closed - Please - I have a headache <BG>.

Realeswatcher - I do own three 1733/1734 fake Klippes. They are also considered fakes. Current thinking is most of these good fakes were made from remelted Ag of genuine 8Rs. After my Chihuahua article I plan to submit to the MNA this next 1733/1734 study article. So I have looked at the remelting issue ... but I do not have enough data to confirm or track Pb levels based on remelting - its just common knowledge in Material Analysis projects.
The Kennedy half and 2 Reales El Cazador coin study in August 2013 is just a simple article showing two contrasting coins and also suggesting that time and environmental conditions are necessary for die struck coins to experience silver surface enrichment. This article will view the microstructure of the coins core and surface showing the different Cu/Ag regions and there migrations internally/externally using SEM/EDS and associated XRF compositional data.