Lack Of Supporting Data
You would think that a metallurgist would leap at
the chance to provide abundant supporting data for his metallurgical argument,
and yet such is not the case in this article. There are several obvious
places where hard data are required to complete the argument, but little or nothing is offered.
The most striking omission forms a gaping hole in the
central metallurgical argument, namely that the grain-size variations in copper and antimony
control their fragment-size variations, and so remove any sense of special
properties from the WCC/MC lead. Here the argument needs hard data to move from qualitative to quantitative.
It needs to prove quantitatively
that the first one (differences in concentrations at boundaries vs. centers)
actually controls the second one (variations of copper and antimony in actual
fragments). The beginning of the article says they do, the end of the article says they do,
but does middle prove these claims?
No, it doesn't prove it, or even come close to proving it. It
does not give actual concentrations of antimony at grain boundaries of WCC/MC
lead, or even how enriched they might be relative to grain centers. It does say,
however, that "Such segregation could easily result in large relative
differences in local antimony content." I think we can all agree that "could"
and "large relative differences" are too vague to have any real import here,
however. It then follows up this statement with "Segregation, such as that shown in
Fig. 3, could easily cause local variations in antimony content of ± 100–200
ppmw [by weight] in the 50-mg MC material." Where is the quantitative
backup for this statement? There is none in the paper. Without backup, it is a
wish rather than anything scientific. The 100–200-ppm variations that grain-size variations can "easily produce" are essentially the same as the 300-ppm
variations cited from Guinn's measurements. This is presented as the end of the story as far as WCC/MC leads are concerned.
The argument for variations of copper goes the same way. It
first says:
"We would therefore expect copper to segregate as a separate phase at the
boundaries and/or form grain-boundary atmospheres in soft leads … ," and it
shows precipitated copper at grain boundaries in Figure 4. But if you are
looking for any calculations, or even estimations, of how much this might affect
the copper in actual samples, you won't find it, or any reference to it, in this
paper.
Thus the logic for antimony and copper is the same—they vary
in crystals and they vary in fragments, so the two must be cause and effect. Why
must they be cause and effect? Simply because the article says so, not because
of any backup, because it offers none.
Other parts of the metallurgical story also lack supporting
data or appear to be influenced by misleading use of data. A good example is the
section on grain size vs. fragment size. It notes that typical grain
sizes in WCC/MC lead are 500–1000
µm, but that is hard for most people to get a feeling for. Much easier is to
express it as 0.5–1.0 mm. But that is the same as the reported size of WCC/MC
samples and crime scene fragments (0.5–1.0 mm in dimension). That would make it
easy for the fragments to fit inside one or two crystals and consequently to
vary strongly in copper and antimony, depending on the proportions of boundary
vs. center that composed them. But it doesn't happen that way. As seen in the
next section, the copper hardly varies relative
to the antimony, except for a small number of erratic outliers.
The stated size for the grains, 0.5–1.0 mm, also
appears to be misleading. Since the diameter of an MC bullet is
6.5 mm, there should be only 7–13 grains across the
end of a bullet. But the article's Figure 5 (grain structures for three thin sections of
an MC bullet) tells a different story. Lines drawn across the middle of those
cross-sections intersect 30 fragments, 23 fragments, and 14 fragments in
sections a through c, respectively (from the upper picture to the lower
picture). Thus all three cross-sections have significantly more grains than
expected from the article. Only the lowest one comes close to its
projections, and it is right at their high end. These measurements mean that it
is easier for fragments to represent more crystals of lead than the article leads us
to believe. This conclusion agrees with the lack of correlation between copper
and antimony described for the sets of samples in the
next section. It is yet another indication that
these theoretical
arguments about the effect of grain size on the JFK fragments are wrong.
The article also strongly implies that WCC/MC lead is no different
from any other soft lead. The only data it offers in support of this claim is a
statement that the ranges of copper and antimony in WCC/MC lead "are quite
similar to other commercial FMJ bullets." That might well be, but ranges are
very general properties, and do not necessarily capture differences among the
leads such as standard deviations and physical scales of variation. The standard
deviations of WCC/MC lead do indeed appear to differ considerably from other
leads, as demonstrated by comparing Guinn's data with data from the NRC's recent report Forensic Analysis: Weighing Bullet
Lead Evidence. For example, intrabullet
and interbullet differences of most elements in most leads are typically <5%,
in great contrast to the 24% and 90% for antimony measured in WCC/MC lead. In
other words, WCC/MC lead differs qualitatively from the many leads of the NRC
report.
The physical scales of variation of antimony in WCC/MC lead
are treated in the sections "Failed Prediction About
Variability Vs. Size of Fragments" and "Do
We Really Need To Know Metallurgy?" They are larger than this article's
metallurgical theory
allows for.
Thus the article is presenting theoretical, incomplete arguments
about the supposed effect of metallurgy on the fragments when actual data
on the subject prove it wrong.