On Fri, 26 Jun 2009 00:22:09 -0700, in rec.crafts.jewelry Carl 1 Lucky Texan
wrote:


I wonder if the piece was analyzed by X-ray Fluoresence? Perhaps that
yields a volume-type ratio? Or did they physically 'assay' a piece?

XRF analysis basically looks at the intensity of a selective spectra emitted by
the sample's surface, with any given element responding differently. With
computer processing, it then is possible to derive a quantitative analysis of
the amounts of the various elements in the surface. The basic information would
be the relative numbers of each elements atoms, so the basic yeild would be
atomic ratios, rather than either weight or volume. Volume gets tricky to
determine just from this, since the information does not give you data on
density, or on the spacing of atoms in an alloy (which can differ from that in
individual pure metals, meaning that volume, a function of density/specific
gravity versus the mass/weight, needs additional information to calculate from
just what the XRF data gives you. No doubt that could be done if needed. But
weight percentages simply need to know the numbers of each type of element or
metal's atoms, which is directly given from the XRF data. Whether the
instrument gives you results as weight percentages or atomic ratios would depend
entirely on what it's programmed to give as the results of the computer analysis
the instrument has to perform to make sense of the spectral results.

However, the subject of XRF brings up an interesting possibility, and potential
answer to Teds problem. XRF analysis only looks at the surface layers of the
sample. It doesn't penetrate much. If the sample is a uniform material, then
the result is accurate for the whole of the sample. However, if the surface
layer of the sample is NOT representative of the whole mass, then one might draw
incorrect conclusions. We all know that operations like pickling, or other
chemical leaching, can deplete some metals from an alloy surface, while
enriching others. A bronze artifact buried for a long time would be exposed to
ground water, corrosive influences, and the like. It seems quite likely to me
that the composition of the surface of the artifact might not exactly match that
of the whole mass of the metal. Whether this would result in more or less
copper or or tin at the surface, I don't know. But if XRFwas the means used to
examine the shield, it's quite possible that the published results of that
examination could have lead to an incorrect assumption as to the total
composition of the alloy. And I can imagine even that the British Museum, in
publishing those results, might simply assume that other researchers, reading
that report, would simply understand this limitation of XRF analysis, while a
craftsperson, not used to the different methods of analysis, might not realize
this limitation, and could draw the wrong conclusion...

But that's just conjecture, of course.

It's also possible they took a tiny sample, perhaps a tiny drill hole somewhere
on the thing where some repair or conservation work was already needed. Then
they'd have a better sample of the overall alloy, and the likelyhood of the
stated analysis being correct for the whole mass of metal, is higher.

Dunno either...

Peter