Magnetic Properties and Alloy Analyses of Greek Bronzes

[PtolemAE]

A few months back I noticed that some of my Ptolemaic bronzes were attracted to a strong magnet.  I couldn't find any information about this phenomenon and those I've asked about it have never heard of this or known about anyone else discovering, reporting it, or testing it.  Seemed a little mysterious so I started tracking it and wondering why.  Last week I spent some time testing about 50 Greek bronze coins with an X-Ray flourescence spectrometer which tells what metals are present at the surface.  Not a perfect test (and not cheap!) but the results turned out quite interesting.  I tested about 40 Ptolemaic bronzes of a wide variety of types and a few Sicilian bronzes and one Seleukid bronze.  I've now posted the preliminary results on www.ptolemybronze.com.

Getting this kind of alloy analysis information has been difficult.  I have found few studies of the metal alloy mix content of ancient coins so I decided to do some myself.  This was made possible by the generous support of dealers and collectors who have contributed to The PtolemAE Project by purchasing the PtolemAE automatic attribution software or ad banners on my site.  Thanks due to them all.

I can't say that the results are 100% conclusive but there are some pretty interesting bits that stand out.  E.g. the Hieron portrait bronzes tested were almost pure copper.  And a few Ptolemaic bronzes were also quite pure but some later ones contained 30% or more lead.  High lead content of some of the later coins has been reported from neutron activation analysis work by Faucher as well and helps me believe that the XRF results are, if not 100% accurate reflection of the whole coin, at least largely consistent with other methods.  XRF only tests the surface.  None of the coins tested were damaged in any way so the readings are of the coins shown as they are shown in the photos.  IOW, XRF is not a perfect testing method but it is what I could afford and do myself without a nuclear research facility at hand.  And it seems to give basically reasonable results.  I was willing try the experiments to see if it was worth pursuing further - and I'm convinced that it is.

You may view this preliminary study (more to come) by going to www.ptolemybronze.com and clicking on the first link - the one that says X-Ray Flourescence ...

You'll find a web page that lists the coins with photos, description, and results of XRF testing for alloy content.  Most were tested on both sides.

BTW - It appears that the coins most strongly attracted to a strong magnet contain at least some iron and it is enhanced by the presence of cobalt.  Maybe the answer I set out to find is at hand.

While at the SF show on Friday I showed the magnetic demonstration to a few dealers, including Harlan Berk.  Everyone was quite surprised.  It isn't really obvious that a 'bronze' coin should be picked up by a magnet.  Just for fun we ran my magnet through Harlan's 'junk-box' and pulled out a small number of coins including some little Romans.  A few other dealers let me test a couple of their coins.

I intend to follow up this initial group of experiments with a lot more when time and $ permit.  If you'd like to support this research or have a coin checked, please email me.  I plan to test many more coins and try to draw some inferences about time periods, mint locations, and get enough data to be confident in the comparisons.

PtolemAE

[tkmallon]

Dear PtolemAE:

I've always been fascinated by this topic, too.  Certain series of Kushan, Indo-Parthian, Indo-Scythian, Indian and Chinese coins have long been known to be magnetic.  In a hoard of more than 300 Vima Takto drachms I worked on years ago, only five coins were NOT magnetic.  Within the Indo-Scythian series, even some of the silver is magnetic.  If you check through the Kushan section of my site: http://www.grifterrec.com/coins/coins.html, you'll find many types listed.  Oddly, if memory serves, the Graeco-Bactrian Nickel coins of Agathokles are not magnetic.  Interesting that you got strong reaction to as little as 0.5% Fe.  Is there a way to test for the presence of Gadolinium (if that even is a component)?

John O. Nisbet, 'Magnetic Reaction In Copper Coins of the Kushana Period,' Ex Moneta: Essays on Numismatics, History and Archaeology in honour of Dr. David W. MacDowall (Volume 1), Harman Publishing House, New Delhi (1998)

Robert C. Senior, Indo-Scythian Coins and History, Volume I, Classical Numismatic Group, Lancaster (2001), 'Magnetic' Coins, pp. 141-142

I can photocopy these for you if you'd like.  Let me know.  tom

[PtolemAE]

Dear PtolemAE:

I've always been fascinated by this topic, too.  Certain series of Kushan, Indo-Parthian, Indo-Scythian, Indian and Chinese coins have long been known to be magnetic.  In a hoard of more than 300 Vima Takto drachms I worked on years ago, only five coins were NOT magnetic.  Within the Indo-Scythian series, even some of the silver is magnetic.  If you check through the Kushan section of my site: http://www.grifterrec.com/coins/coins.html, you'll find many types listed.  Oddly, if memory serves, the Graeco-Bactrian Nickel coins of Agathokles are not magnetic.  Interesting that you got strong reaction to as little as 0.5% Fe.  Is there a way to test for the presence of Gadolinium (if that even is a component)?

John O. Nisbet, 'Magnetic Reaction In Copper Coins of the Kushana Period,' Ex Moneta: Essays on Numismatics, History and Archaeology in honour of Dr. David W. MacDowall (Volume 1), Harman Publishing House, New Delhi (1998)

Robert C. Senior, Indo-Scythian Coins and History, Volume I, Classical Numismatic Group, Lancaster (2001), 'Magnetic' Coins, pp. 141-142

I can photocopy these for you if you'd like.  Let me know.  tom

I'd be very grateful for copies of the articles you have cited.

The device I used does not appear to measure Gadolinium. 

I found the effect seems to be partly due to iron but strongly influenced by even smaller amounts of cobalt.  That was a surprise.  I figured iron was the sole culprit, but glad I did some testing.  A lot more work to be done on this and it's going to take a while but I believe the preliminary results mean that this is a productive avenue of investigation. 

Thank you for your encouraging kind words.

PtolemAE

[moonmoth]

I noticed some while ago that a small proportion of "bronze" coins are attracted to a magnet.  Some strongly so, some just a little.  It does not seem to be consistent for any particular type of coin, and like you, I have found both Roman and Hellenic coins. 

Iron, cobalt and nickel are the obvious suspects, but I notice that you don't mention nickel.

Bill

[Steve Minnoch]

The original post was cross-posted here.  It's much better, imho, to have all the discussion in one place, by merging this thread with the one there and either leaving it in "Greek Coins" or moving it to "Ancient Coin Forum"

PtolemAE, do you have any objections?

[DEAD LINK REMOVED BY ADMIN]

Steve

[PtolemAE]

The original post was cross-posted here.  It's much better, imho, to have all the discussion in one place, by merging this thread with the one there and either leaving it in "Greek Coins" or moving it to "Ancient Coin Forum"

PtolemAE, do you have any objections?

https://www.forumancientcoins.com/board/index.php?topic=49788.msg309662;topicseen#msg309662

Steve

Whatever is most convenient for you and the forum board.  I didn't know which was the best venue.
No objections,

PtolemAE

[Arminius]

We  know silver coins crystallize during the centuries.
See the broken edges or rough surfaces of some mechanically stressed silver coins like this one.

The various metals in bronze also crystallize to a certain extent, forming small areas of pure magnetic iron, nickel and other magnetic components ?


Regards

[moonmoth]

I have at least two coins that respond strongly enough to be picked up by a magnet; a follis of Diocletian from Treveri, and a centenionalis of Constans from Heraclea.  There are others as well - I just happen to have made a note about those two.

In fact, I have just tested the coins currently on my desk, and an assarion of Septimius Severus from Nikopolis ad Istrum can be picked up too.  It jumps up to the magnet.  None of these coins show any obvious signs of crystallisation.

Bill

[PtolemAE]

I have at least two coins that respond strongly enough to be picked up by a magnet; a follis of Diocletian from Treveri, and a centenionalis of Constans from Heraclea.  There are others as well - I just happen to have made a note about those two.

In fact, I have just tested the coins currently on my desk, and an assarion of Septimius Severus from Nikopolis ad Istrum can be picked up too.  It jumps up to the magnet.  None of these coins show any obvious signs of crystallisation.

Bill

crystallization seemingly has different meanings in numismatics and chemistry.  virtually all solid metals are 'crystalline' in their microscopic structure, speaking scientifically (some specially prepared types are glasses).  the granular appearance that we see in some coins may or may not have anything to do with the way the word 'crystallized' is used in physical sciences since non-granular coins are also 'crystallized', perhaps differently.   Or perhaps the crystal grain structure has changed over time in some more than in others.  'Crystallized' is too vague when the discussion gets beyond superficial descriptions.

PtolemAE

[moonmoth]

I was using the term to mean the appearance of a crystalline structure visible to the naked eye or under low magnification, as often happens with ancient silver coins and sometimes billon coins too.  Like this rather extreme example.  I used the word "obvious" to cover the possibility that there might be a non-obvious process that I could not see!

[PtolemAE]

I noticed some while ago that a small proportion of "bronze" coins are attracted to a magnet.  Some strongly so, some just a little.  It does not seem to be consistent for any particular type of coin, and like you, I have found both Roman and Hellenic coins. 

Iron, cobalt and nickel are the obvious suspects, but I notice that you don't mention nickel.

Bill

the U.S. five-cent coin is about 75% copper and 25% nickel and is not magnetic at all.  nickel alone, esp. in the amounts present in these coins (typically less than 0.1% as measured) is apparently not a factor.

the results on my page show nickel is typically very low and that iron and cobalt are the most prevalent in coins that are magnetic.

and cobalt, btw, was not an obvious candidate to me until i tested the coins and discovered it was present in many coins.  it was a surprise that it would have as much effect on magnetism as larger amounts of iron would have.  i suspected that iron was the only candidate and i was wrong.  cobalt seems to 'boost' the magnetic properties of the iron.  some coins with little iron but a modest amount of cobalt are just as magnetic as coins with a lot more iron but no cobalt.  cobalt seems to dominate the magnetism.  

PtolemAE

[PtolemAE]

I was using the term to mean the appearance of a crystalline structure visible to the naked eye or under low magnification, as often happens with ancient silver coins and sometimes billon coins too.  Like this rather extreme example.  I used the word "obvious" to cover the possibility that there might be a non-obvious process that I could not see!

precisely my point.  anyone can have their own private meaning for a term so it is potentially confusing in this context.  i *measured* metal content was to get beyond vague.

quantitative comparison of crystal grain sizes in different types of coins would be interesting but it's probably a separate subject I encourage others to study.

PtolemAE

[moonmoth]

anyone can have their own private meaning for a term so it is potentially confusing in this context.  i *measured* metal content was to get beyond vague.
PtolemAE

That can only be a good approach to take.  My mention of crystallisation, using the generally accepted meaning as regards silver coins, was no more than a response to the suggestion that bronze coins might crystallise and so expose raw metals.

I am cautious about using the term "magnetic,"  as clearly you are too.  It is useful to distinguish between objects that attract iron to them, and objects that are merely attracted by magnets, even though the term can mean both. The coins in my possession do not attract iron, but are strongly affected by a magnetic field.  I have not tried magnetising them by either of the two easy methods - have you?

Bill

[Lloyd Taylor]

Have you examined or considered the possibility that these coins have a remnant magnetic field?

This may explain the phenomenon of variable magnetic attraction amongst coins with low levels of iron and cobalt in the surface metallurgy. 

By way of explanation, remnant magnetic fields are a product of the alignment of magnetic particles in material with the earth's geomagnetic field over a period of time.  The phenomenon is common in geology and exploited by geophysics to do many things including sea-floor mapping, polar wandering analysis, magnetic age determination and paleoenvironmental studies. Remnant magnetic fields explain the striped magnetic field response of the sea-floor and were instrumental in the unraveling of sea-floor spreading and continental drift.  But back to coins.  Over thousands of years iron in particular can be mobilized in solution and re-deposited in, or on the coin surface. The iron so deposited will align to the earth's magnetic field at the location of the coin, in effect creating a mini-magnet in the coin. This remnant field can measured in a geophysics laboratory and its one aspect you may want to check.  A simple, crude test would be to determine the consistency with which a coin is oriented on attraction to a pole of a magnet.  A statistically consistent alignment to a single pole would indicate the presence of a remnant magnetic field.

Another observation I make is that with XRF you are measuring the composition of surface patination of the coin.  This is the part of the coin most influenced by environmental factors and thus most subject to the development of remnant magnetism over thousands of years. Also the iron in the surface patina may have been deposited from solution in ground water, rather than originating from the metal of the coin.

A further possibility is that the remnant magnetism, if present is a thermal remnant magnetism arising from the cooling and solidification of the metal during the casting of the flan of the coin.  However, I would discount this possibility, given that the cooling  and solidification would have occurred very rapidly at a time when fluid vortices in the molten metal would most likely have been the dominant factor, overriding the potential for alignment of magnetic components with the geomagnetic field at the location of the mint that issued the coin.

This potentially opens a new field of geomagnetic numismatics! 

A by product of this is that if a remnant magnetic field is developed in the patina of coins that have been subject to appropriate burial conditions (mobile iron in the groundwater), then cleaned and re-patinated (artificially patinated) coins from the same area of burial will not exhibit remnant magnetism.  This may explain some of the variability in strength of magnetic attraction you are observing.

[Lloyd Taylor]

Addition to the preceding note:

I forgot to mention that a remnant magnetic field will provide for stronger magnetic attraction at low concentrations of iron in the coin patina than would be the case if the same amount of iron was present in a randomly dispersed orientation. This could explain the strength of magnetic attraction at low iron concentration and the variability of magnetic strength from coin to coin with similar concentrations of iron, depending on the burial conditions and burial stability over a protracted period of time. To be fully effective in creating the remnant magnetic field the coin would have to remain in a stable unvarying orientation during the period in which the patina was created. Subsurface movement, the collapse of hoard contents, etc would serve to disrupt the creation of a string remnant magnetic field - again this highlights the fact that the strength of magnetic attraction to which you refer will be variable and dependent on much more than a single variable such as the iron content alone. As a geophysicist, I know  from experience, as well as theory, that the factors involved are many and varied when it comes to the strength of the remnant magnetic field in any material.

[PtolemAE]

Have you examined or considered the possibility that these coins have a remnant magnetic field?

This may explain the phenomenon of variable magnetic attraction amongst coins with low levels of iron and cobalt in the surface metallurgy. 

By way of explanation, remnant magnetic fields are a product of the alignment of magnetic particles in material with the earth's geomagnetic field over a period of time.  The phenomenon is common in geology and exploited by geophysics to do many things including sea-floor mapping, polar wandering analysis, magnetic age determination and paleoenvironmental studies. Remnant magnetic fields explain the striped magnetic field response of the sea-floor and were instrumental in the unraveling of sea-floor spreading and continental drift.  But back to coins.  Over thousands of years iron in particular can be mobilized in solution and re-deposited in, or on the coin surface. The iron so deposited will align to the earth's magnetic field at the location of the coin, in effect creating a mini-magnet in the coin. This remnant field can measured in a geophysics laboratory and its one aspect you may want to check.  A simple, crude test would be to determine the consistency with which a coin is oriented on attraction to a pole of a magnet.  A statistically consistent alignment to a single pole would indicate the presence of a remnant magnetic field.

Another observation I make is that with XRF you are measuring the composition of surface patination of the coin.  This is the part of the coin most influenced by environmental factors and thus most subject to the development of remnant magnetism over thousands of years. Also the iron in the surface patina may have been deposited from solution in ground water, rather than originating from the metal of the coin.

A further possibility is that the remnant magnetism, if present is a thermal remnant magnetism arising from the cooling and solidification of the metal during the casting of the flan of the coin.  However, I would discount this possibility, given that the cooling  and solidification would have occurred very rapidly at a time when fluid vortices in the molten metal would most likely have been the dominant factor, overriding the potential for alignment of magnetic components with the geomagnetic field at the location of the mint that issued the coin.

This potentially opens a new field of geomagnetic numismatics! 

A by product of this is that if a remnant magnetic field is developed in the patina of coins that have been subject to appropriate burial conditions (mobile iron in the groundwater), then cleaned and re-patinated (artificially patinated) coins from the same area of burial will not exhibit remnant magnetism.  This may explain some of the variability in strength of magnetic attraction you are observing.

the results i obtained don't rule out surface (patina) iron but are consistent with other results using neutron activation analysis showing 'bulk iron', etc.  all I am actually measuring is the surface but some coins don't actually have much patina.  also i have coins weighing 50 grams that are strongly attracted to a magnet, enough to pick them right up against gravity.  no amount of thinly patinated iron contamination (even if it were 100% of the patina) could explain that. 

remnant magnetic issues are interesting but unrelated to what i'm studying.  i suspect remanant magnetization is present but am not interested in studying it.  i hope you will study it and make a contribution to better understanding of it.  and it won't affect the strength of attraction to a magnet, anyway.  a magnet (with a magnetic moment, remanant or otherwise) is no more strongly attracted to another magnet than is any ordinary unmagnetized lump of iron.

PtolemAE

[PtolemAE]

Addition to the preceding note:

I forgot to mention that a remnant magnetic field will provide for stronger magnetic attraction at low concentrations of iron in the coin patina than would be the case if the same amount of iron was present in a randomly dispersed orientation. This could explain the strength of magnetic attraction at low iron concentration and the variability of magnetic strength from coin to coin with similar concentrations of iron, depending on the burial conditions and burial stability over a protracted period of time. To be fully effective in creating the remnant magnetic field the coin would have to remain in a stable unvarying orientation during the period in which the patina was created. Subsurface movement, the collapse of hoard contents, etc would serve to disrupt the creation of a string remnant magnetic field - again this highlights the fact that the strength of magnetic attraction to which you refer will be variable and dependent on much more than a single variable such as the iron content alone. As a geophysicist, I know  from experience, as well as theory, that the factors involved are many and varied when it comes to the strength of the remnant magnetic field in any material.

i think you could do your own experiments to support these assertions.  i doubt them.  and I doubt remanant fields (ordinarily extremely weak) have anything to do with the reason that some coins are strongly attracted to magnets.  i'll go with the more parsimonious explanation of iron and cobalt being in the coins, which has also been shown by other methods but which were not correlated to ferromagnetism by the other researchers who simply didn't look, ttbomk, at FM.

PtolemAE

[PtolemAE]

anyone can have their own private meaning for a term so it is potentially confusing in this context.  i *measured* metal content was to get beyond vague.
PtolemAE

That can only be a good approach to take.  My mention of crystallisation, using the generally accepted meaning as regards silver coins, was no more than a response to the suggestion that bronze coins might crystallise and so expose raw metals.

the 'accepted meaning for coins' is meaningless in this context, thus best avoided completely.  the change over time in which alloys might separate into granules of their elemental components is probably best described more carefully in this context, at least in the absence of any mind-reading abilities possessed by your humble reporter


I am cautious about using the term "magnetic,"  as clearly you are too.  It is useful to distinguish between objects that attract iron to them, and objects that are merely attracted by magnets, even though the term can mean both. The coins in my possession do not attract iron, but are strongly affected by a magnetic field.  I have not tried magnetising them by either of the two easy methods - have you?

i expressed very clearly what i observed - in the first sentence of my first post.  if you try magnetizing some coins let us know how it works out.  have at it.  i've got plenty on my plate now

PtolemAE


Bill

[Joe Sermarini]

I was using the term to mean the appearance of a crystalline structure visible to the naked eye or under low magnification, as often happens with ancient silver coins and sometimes billon coins too.  Like this rather extreme example.  I used the word "obvious" to cover the possibility that there might be a non-obvious process that I could not see!

I have never seen that type of structure on a billon coin.  I am under the impression that the metal must be nearly pure (on ancient terms) silver for that "crystalline" structure to arise.  Billon coins, I think, just get rough and grainy. 

[PeterD]

The term crystalisation is used rather indiscriminately, when usually what is being described is a silver-copper coin that has had copper leeched from it.

Alloy with a silver content of less than 90% will form a two phase system when cooling due to the different melting points of the two metals. In other words parts of the coin are pure silver while other parts are pure copper. When a coin is in the ground the copper is destroyed faster than the silver, leaving a lighter honeycombed coin.

This can be seen in this picture taken from Ponting and Gitler's book on the coinage of Sept Severus. It shows a section through a coin of Geta (45% silver). The yellow parts are copper and the white (or light yellow) parts are silver. The black parts 'holes' are where the copper has been 'leeched' (not a term that really desribes what is happening).

Whether 'crystallised' is an actual condition I do not know.

[Joe Sermarini]

The term crystalisation is used rather indiscriminately, when usually what is being described is a silver-copper coin that has had copper leeched from it.

That is not crystallisation. 

[CzarMike]

Id assume they cheeped out during the coin making and added some iron
.

[El Reye]

Id assume they cheeped out during the coin making and added some iron
.

Cobalt and iron are closely associated with copper ore deposits and can be quite difficult to completely separate out during the refining process. Consider that ore processing/refining here in modern times can be the make or break on the profitability of a mine. Additionally these trace elements might help determine the location on the copper deposits used in this coinage. Do the coins with magnetic properties come from the same region?

Cameron

[PtolemAE]

Id assume they cheeped out during the coin making and added some iron
.

Cobalt and iron are closely associated with copper ore deposits and can be quite difficult to completely separate out during the refining process. Consider that ore processing/refining here in modern times can be the make or break on the profitability of a mine. Additionally these trace elements might help determine the location on the copper deposits used in this coinage. Do the coins with magnetic properties come from the same region?

Cameron

Both notions have occured to me and I can't be certain which may be the case.  I suspect in some later Ptolemaic bronzes that 'going cheap' may be part of the picture.  But some of the coins are early and some are Syracusan (both Hieron portrait types and contemporaneous Sicilian imitative Ptolemaic bronzes) with unusually high copper content that I suspect have the magnetic components of iron and cobalt due to the ore source.

I've run XRF on a bunch more coins since the first XRF experiments and the additional data support the conclusions gained from the first run, i.e. that iron esp. if a small amount of cobalt is present, is the reason why these coins are attracted to a strong magnet.  Haven't had time to post the data from the 2nd run but I hope to do so soon.

PtolemAE

[PtolemAE]

This is an interesting discussion. As Lloyd T notes it is possible that the coins at the time the flan was made and subsequently heated if the temperature exceeded the Curie temperature of certain iron compounds (around 700 or so C) the metal(s)' the domains will align to the field and remain that way once the temperature drops below the Curie point. It is exactly analogous to what Lloyd T noted for the ocean rift zones - as the molten rock cools below the Curie point the extent field is frozen into the rock. When the coin is struck momentarily the metal becomes very hot and literally flows into the die to produce the features at that time one can cause the metal to become magnetic. In the days of the first metal hauled ships they had to compensate for the residual magnetic field in the ship's haul because caused when the metals where heated to be bent and formed. Still a problem today but they have learned to do a better job of compensation and GPS helps!

c.rhodes

all that about curie temp etc. is true, but irrelevant to the phenomenon under discussion.  what i was trying to explain is the strong ferromagnetic properties of actual coins, not their potential (miniscule) residual magnetization.  iow, if a coin can be picked up with a magnet it has (little or) nothing to do with the coin's own tiny (if any) residual magnetization.  i have not measured for residual magnetization of coins, only whether they are attracted to a strong magnet (with a field thousands of times stronger than any potential residual magnetization of the coin).  i haven't found a single coin that was itself strongly enough magnetized (due to the residual magnetization resulting from cooling in the earth's field) to exhibit 'repulsion' from either pole of the magnets i'm using for testing this phenomenon.  the rare earth magnets i used to test a coin's ferromagnetism ('attraction') have field strengths of thousands of gauss.  residual magnetism might be interesting, but it's an entirely separate subject that would have to be measured by a completely different method.

the gross ferromagnetism measurements as well as subsequent XRF tell us that bronze coins that have a small iron content (above about 1/2% on XRF tests) are attracted to strong magnets - a phenomenon enhanced by the presence of even smaller amounts of cobalt along with the iron.  i have made no tests whatever whether such coins have 'compass' properties (residual magnetism) and plan to leave that to others who have speculated here about its possible existence.  i have no doubt that it may be possible to 'magnetize' some coins that have a suitably high iron content.  the whole point of this was to discover why some bronze coins were attracted to magnets *at all*, because that seemed unexpected in bronze.  the conclusion is that modest iron content, esp. with a bit of cobalt, causes it.  once we realize that iron is present at all, then the other phenomena (residual magnetization, magnetic susceptibility, etc.) are not very mysterious and so I will leave that to others.  now that we know it is plausible (because enough iron is present in some Ptolemaic and Sicilian bronzes I discovered were attracted to magnets) i would expect someone will easily detect residual magnetization and find that some coins can deflect a compass needle and even be deliberately magnetized.  the main new knowledge is that there is enough iron (and cobalt) present in some Ptolemaic and Sicilian bronzes to make a difference in the easily observed magnetic properties of these coins.  the residual magnetism stuff is sort of down in the weeds compared to the main phenomenon I've been looking at.

i encourage everyone with bronzes to let us know if other types (in addition to those already reported) are attracted to magnets.  anyone who checks for residual magnetization, please report as well.  i'll post more on the additional recent XRF tests as soon as i can.  there's plenty more to learn (e.g. if coins from certain locations, mints, or time periods, etc. display ferromagnetism more than others).

PtolemAE