(forwarded from John Hill, author of *Through The Jade Gate to Rome*
] as FYI on this new ^M2S^sm technique.)
Dear friends and colleagues:
Some years ago I came across the use of new scientific techniques to
establish the provenance of ancient gems which caused minimal damage to
the samples. Spectrographic analysis allowed the identification of a
gemstone in a Gallo-Roman earring (c. 3rd to 6th centuries CE) as being
an emerald mined in Swat – in northern Pakistan. (See: Giuliani, et al
(2000): “Oxygen Isotopes and Emerald Trade Routes Since Antiquity.”
/Science/, January 28, 2000, pp. 631-633. I reported on this interesting
fact in my book, /Through the Jade Gate to Rome/ (2009), p. 501.
Since then I have kept a look out for new developments. Just recently I
noticed in the the April 2012 edition of /National Geographic/ Magazine
a brief notice that a company named Materyalitics was using a new
technique for provenancing gems. I sent an enquiry by email to the
company in Texas and quickly received a friendly reply which, amongst
other things, said their technique could be used to provenance a wide
range of materials, not just gems.
“. . . And more importantly: gems have been only our starting point.
In fact, our ^M2S^sm technology can be applied to virtually any
materials that seem important to people, and we've enjoyed talking
about testing archaeological and fossil artifacts. The problem is
that archaeologists and paleontologists tend not to have the money
it takes to build the necessary Reference Collections and Reference
Databases to serve their needs, while manufacturers of high-tech
devices do. It may take a while to get to the artifacts.
On the other hand, we have done a great deal of testing of
beryls...which include emeralds, and have already created a fairly
substantial Reference Database that might be useful. In fact, we
must have samples from the Panjshir Valley. (I'm not sure, but the
name rings a bell.)
It has been pointed out to us that a good emerald mined in
Afghanistan may be worth $X...but if that same emerald can be
transported somehow to Colombia, magically being transformed into a
Colombian Emerald in the process, it may then be worth $10X. This
encourages shenanigans that our technology may help to inhibit.
Reading your first email, our Research Director commented that it
would be fascinating if a stone traditionally considered to have
come from somewhere else proved to be Colombian...perhaps in ancient
and unsuspected times. No one knows, and there must be interesting
surprises to find.
It seems possible that silk, to pick a material at random, could be
traced to its source. We haven't looked at silk, but one supposes
that worms eating mulberry leaves acquire distinctive chemical
makeup characteristic of the soil in which those trees grew, that
transfers to their cocoons. This speculation is not entirely idle,
and it's intriguing, but no database of silks exists, and the cost
of someone's creating a comprehensive database would run to hundreds
of thousands of dollars, at least. The return on investment seems
unappealingly low for anyone but an uncommonly well-to-do scholar
eager to know more about silk than anyone now knows.
What other materials that are worth tracing originated at points
along the Silk Routes? ...and so on. Just notions to think about. .
. . .
Office Phone: 254-226-9639
This was the first time I had heard of being able to provenance
materials such as silk, so I wrote back immediately asking whether
provenancing silks was a real possibility, he replied that, yes it could
be done, but it would be expensive as it would involve the collection
and testing of many samples:
“. . . The whole matter of building a database of silks…from wild
worms, cultivated worms, mollusks, or whatever, is very interesting,
though the chances of doing it are very low, considering the cost .
Just so you know what the procedure is, I’ll give you an outline.
The ^^Materialytics^^sm Sequencing System (^^M2S^^sm ) does not look
for a list of features (chemical composition, color, specific
gravity, or any of that) when it examines a material.
It looks at one sample of material from a known source, a gem from a
particular mine or district, or an apple from a particular orchard,
or a metal part produced on a particular production line, for
example, and records whatever spectral data it finds.
It doesn’t know, and doesn’t have to know, what the data represent.
It collects millions of data points from the sample.
It then looks at another well-documented sample from the same
source, the same way, recording whatever it finds.
M2S looks at a “statistically significant” number of samples from
It looks at a statistically significant number of points on each
sample. It stores all that data. It’s a LOT of data.
Traditionally, among statisticians the smallest “significant” number
of samples is thirty. We typically analyze several times than many
samples from each source. Similarly, we analyze several times thirty
points on each sample to accommodate variability in the material.
The data from all laser shots on each sample are processed to create
a ^^Quantagenetic^^sm Sequence” of that sample.
The Quantagentic Sequences of all of the samples from a single
source (mine, factory, orchard, whatever it is the samples have in
common that it of interest to us) are processed to create a
Quantagenetic Signature that characterizes materials from that source.
When we have analyzed material from all of the sources that may be
of interest to whoever is paying for the work, we have created a
Reference Database with which to compare samples of material whose
source is not known.
We test an unknown sample the same way we did the Reference Samples,
but with only a few laser shots. The data are processed the same
way. The resulting Quantagenetic Sequence of the unknown sample is
compared with the Quantagenetic Sequences in the Reference Database,
to see if there’s a match with any signature.
In the case of gems, the average accuracy of matching (if there is a
match) is over 98%. We really have been able to track the provenance
of unknowns to the specific mines they came from…IF…we have enough
data in the Reference Database. The accuracy of M2S increases as the
size of the Reference Database increases. M2S gains more
"experience” with the world.
Anyone interested in pursuing this should contact John Hill (and not me).
All good wishes,
Belforte (Si), Italy
Visit Zenobia's blog at Empress of the East