New video: "Is the Voynich Manuscript fake?"

[Jorge_Stolfi]

But the McCrone report states "Titanium compound" for Sample 20 and "Titanium compound (particle)" for Sample 17.
Generalizing that finding as a common impurity minimizes the analytical weight given by the report's language and it ignores the expected protocols of a report from a professional source like McCrone.  They were specific in their terminology.  As one expects in a report from a professional laboratory like McCrone, they use explicit qualifiers when constituents are present at low levels, such as "Iron gall ink (very low iron)" , "Cuprite (minor)", or "Azurite and cuprite (traces)", etc. The fact that the McCrone report listed "Titanium compound" for Samples 17 and 20 without any such qualifier (like "trace" or "minor") strongly suggests that the titanium was present at a concentration deemed significant and measurable by professional analysts.

We have to acknowledge that the McCrone report explicitly identified a "Titanium compound," which, despite not being identified as (and likely not being) a specific crystalline form of "titanium white" raises a distinct inconsistency with the expectations of a medieval source for the pigment. (Attempts to get an explanation for this inconsistency from McCrone and other sources at Beinecke have failed).

I have read the McCrone report, and must say that I don't share your respect for their competence in interpreting and describing the results of their tests.

The report has four EDS spectra of sample 17.  EDS ("Energy Dispersive Spectroscopy") works by measuring the X-ray fluorescence emitted by the sample when hit by the electron beam of a scanning electron microscope.  Thus it can reveal the elements (but not the compounds) present at each spot of the sample.

According to Table I, sample 17 is "black ink from drawing (woman's face)" from page f70v.
Not clear whether that was from what I consider original drawing, restoration round 1, or retouching rounds 2 and 3. (Probably they picked the darkest ink, which would be my round 3). 

They collected two other samples of the "black" ink, 13 from the text and 14 from the drawing, both on f80v; and those had no sign of titanium.  They also collected samples of the "black" ink from You are not allowed to view links. Register or Login to view. (16), from the folio number of You are not allowed to view links. Register or Login to view. (15), from the quire mark on You are not allowed to view links. Register or Login to view. (19), and from the "a" in the right margin of You are not allowed to view links. Register or Login to view. (20).  Sample 20 is the only one of these that showed any titanium.  But all these writings are known to be much later than the original writing.

It does not help that they mix up "folio" and "page", and call "black" ink that is actually various shades of brown. It does not help either that they totally fail to mention the large variations in ink density and color across the VMS, which would at least deserve a short explanation for why they did not consider it worth sampling and comparing them.

The report includes four spot spectra of sample 17, and only two of them shows definite signs of titanium.  Their interpretations of spectra 2/3 say 1.3/1.8% of titanium, compared to 5.7/5.8% of iron, 2.3/2.1% of copper, 2.9/2.9% of mercury, 0.9/0.0% of arsenic, and 0.5/0.5% of phosphorus; the rest being common elements that could have come from the vellum, binder, or dirt.  Thus only some particles in that sample seem to contain any measurable amount of titanium, and even then in only small amounts. 

Needless to say, titanium is present in those samples as compounds, not as the pure metal. So their use of the word "compound" means only that they cannot identify the chemical form.  And their data does not let one tell whether the only metal in the "compound" is titanium, or whether the compound could be, for instance, ilmenite FeTiO3, a fairly common black mineral that was occasionally used as black solid pigment.

  You are not allowed to view links. Register or Login to view.
  "Raman spectroscopic analysis of an important Visigothic historiated manuscript"
  "Of particular interest was sample AMS7, dated 1649–1662, where iron-gall ink was found mixed with another black pigment ilmenite (FeTiO3) [17]."
 
For spectrm 4 of sample 17, on the other hand, their interpretation gives tin Sn as 7.7%, more than iron 3.1%, copper 1.3%, and mercury 1.4% combined.  Tin is not an element that one expects to find in old pigments or inks (except maybe greens made by chemical processing of bronze scrap).  However, in the spectrum itself the peak labeled Sn coincides with a large peak also labeled Ca (calcium), which probably comes from the chalk paste usually applied to vellum in order to make it smoother and more receptive to ink.  So the 7.7% of tin may well be just an error of of the software that interprets the spectrum as a mixture of elements spectra.

Another one of the many things in the report that spoils my confidence in them is their claim of "palmierite" being detected in sample 8 (red-brown paint from the root of f26r). This statement is based onthe X-ray diffraction (XRD) spectrum, that depends on the spacing and angles of atoms in a crystal.  To identify a mineral, the spectrum is automatically compared to a large library of spectra, and the best match is taken to be the mineral present.

But palmierite, a sulfate of lead and potassium K2Pb(SO4)2, is an extremely rare colorless mineral found as tiny grains in the deposits around the fumaroles of Mount Vesuvius and a few other volcanoes worldwide, mixed with several other minerals.  It may be theoretically possible that pamierite formed over time from reactions between the red lead pigments used in the paint; or that the pigment used in the root was produced by a chemical process that happened to create some palmierite.  But what most probably happened is that the software McCrone used to identify the patterns from X-ray diffraction data found that the palmierite one happened to match the observed spectrum better than that of the mineral or mineral mixture actually present in the sample.  The technician just copy-pasted "palmierite" from the program's output, without bothering to check what "palmierite" was.  I bet he would have done the same if the output was "technetium perxenate"...

All the best, --stolfi