# Last edited on 2025-10-04 12:52:24 by stolfi ??? The dark grey additive to blue paint may be an oil-based apint. A forced mixture of water- and oil-based paints would tend to separate while still wet, with the water-based component sticking to the hydrophillic sized parchment surface, and the oil-based conponent pushed off to the edges. The oil-based component may have taken longer to dry which would explain why it is the one responsible for the majority of the offset cases. ??? Natural atacamite is a commercial pigment now but in the past it was confined to Asia, Egypt, and the Americas, since the mineral does not occur in Europe in significant amounts. There is an old recipe for "salt green" pigment, formed by covering a copper plate with salt and leaving it exposed for months to vinergar vapours. Is that likely to have been used for the Manuscript's Painter, who did not even have a decent fine brush? ??? Maybe the Lab technician sneezed while tryingto mount a sample for XRD, and so he cribbed the results of the "green paint" analisis from a report that the Lab did previously on another manuscript. This note is a review of a Certain Report issued by Some Lab that analyzed samples from a Certain Manuscript. Any similarities with other reports, labs, and manuscripts should be assumed to be a mere coincidence. One general problem is that the Report is generally too vague about the quantity of components in the analyzed paints. For instance, it often says "minor amounts" or "traces" of something. Is that 10%, or 0.01%? The Report also mixes positive identifications of materials by the instruments with pure guesses based on nothing but what is believed to be "usual" in manuscripts. Like, the ink "must be" iron-gall, and the paint binder "must be" gum, only because "it is almost always so". But the purpose of hiring the Lab was precisely to verify whether the materials *were* consistent with each other and with other manuscripts of the time. Reminds me o the joke about the lazy cashier who had to count the $50 bills in a $1000 packet: "One, two, three, four, five, six, [...] twenty one, twenty two... bah, if it checked so far, it must check all the the way through to the end." Some glaring omissions from the Report: There was no analysis of the translucent light yellow paint that is used in almost every page. There was no analysis of the red ink used in the text of Many pages of the Manuscript have text and drawings of up to four different degrees of lightness, width, and hue. But only one sample of ink was taken from each page, and only from a couple of pages. No attempt was made to explain or dismiss those differences. There are two distinct types of blue paint. Only one sample was analyzed. There are two distinct types of green paint. Samples of both were analyzed, but the results were mashed and discussed together as if it there was just a single paint. One of the types of green paint dominates in the "ghosting" -- color from one side of the folio showing up on the other side. No attempt was made to explain why that paint, specifically, had that power. Some specific points: "Examination: The writing is a brownish-black ink of variable darkeness" More like: inks of various shades and lightness of brown, from ochre to sienna to nearly black. "When examined with ultraviolet radiation [...] the writing is a deep velvety blue-black, suggesting an iron-gall ink." The "velvety" and "blue" is just gratuituous fantasy, and the "iron-gall ink" is an unwarranted inferrence. It should have said "Under ultraviolet radiation, the writing is black." Period. "Preparation: The samples were analyzed by a combination of [Polarized Light Microscopy] PLM, [Infrared Spectroscopy] IR, [X-Ray Diffraction] XRD, and [Dispersice X-Ray Spectrometry] EDS in the [Scanning Elecron Microscope] SEM. [...] All of the inks used for text of drawing were identified as iron-gall inks." None of these methods can *positively* identify iron-gall ink. PLM and XRD can help identify *crystalline* solid particles, which iron-gall ink is not, and EDS can identify the *metals* present in the sample but not their chemical state. IR can distinguish some binders and some compounds of iron, like ocher paint from iron-gall ink; but the multispectral images in the infrared bands (700 to 940 nm) are are consistent with the former and not with the latter. ???CHECK. So the Report's conclusion of "iron-gall ink" was reached by exclusion: If there are "modern" pigments, like rutile or prussian blue, it is a forgery; else If there are colored mineral pigments of the epoch, like azurite and malachite, it is "paint from the period"; else If it is writing in dark ink with iron, say "iron-gall ink"; else If it is writing in dark ink but without iron, say ... duh ... "low-iron iron-gall ink". Line 3 is normally a safe bet for the Lab, because writing on parchment only makes sense if one wants the document to last for many decades and resist wear, humidity, spills, etc. Namely for decrees, contracts, deeds, diplomas, prayer books, manuals, chronicles, etc. For things that are meant to last only a few years at most -- like letters, drafts, diaries, etc -- one would use paper instead. But writing on parchment in turn only makes sense if it is with iron-gall ink, because any other type of ink would easily come off under those abuses. Thus, any dark writing on a random parchment document is iron-gall ink with 99.99% probability. But the Manuscript in question is anything but a "random parchment document"... There are several clues that the ink used on the Manuscript, exceptionally, is *not* iron-gall ink. They include the color and the facility with which the ink was rubbed or washed away. One would have expected the Lab to notice those clues and either confirm or dismiss them *with hard data*. Instead, it insists on "iron gall" even when the data flatly contradict that conclusion. Even for quire marks and marginal notes by would-be codebreakers, which, not being intended for long term, might well have been made with lampblack ink instead.. "All of the inks examined [...] had similar microscopic characteristics. The particles were consistently brownish-black color[...] " Iron-gall ink should be an amorphous black film, not "brownish-black particles". "there was considerable variation in the opacity (darkness) of the ink; the opacity correlates with the amount of iron present in the EDS spectra". Not surprising. But was it less opaque because the film was thinner or lighter? Or because there were fewer particles per mm^2? "The particles are transparentm isotropic flakes with irregular of conchoidal fracture" Not what one expects iron-gall ink to look like. "[particles] with refractive index lower than 1.662". So? "Sample 16 also include several very darks opaque particles of more concentrated ink" Yeah? "and Sample 20 included a red-brown particle with characteristics consistent with burnt sienna, a common iron oxide" Burnt sienna is made by roasting "raw sienna" (an iron oxide-hydroxide with some manganese content) to a few hundred degrees. Why and how would that pigment appear on traces of "iron-gall ink"? It must have been part of the formula. Then why wouldn't the other samples also have mineral iron pigments? "X-ray diffraction ... in Sample 2 identfied three crystalline materials: potassium lead oxide, potassium hydrogen phosphate, and syngenite, a basic potassium calcium sulfate." Potassium hydrogen phosphate is highly soluble in water and deliqueschent (absorbs water from the air and dissolves in it). It should not be a component of iron-gall ink or vellum. How did it get there? It probably was not available in the 1400s. "Interestingly, the EDS spectrum shows no lead at all" So one of the two tests, EDS or XRD, was botched. Which one? Maybe both? "Sample 19, the quire number, is a high carbon, verly low iron ink. Microscopically it appears as a transparent light brown material without the particulate material suggestive of carbon ink ... While carbon ink cannot be completely ruled out" The carbon particles in lampblack ink (soot ink, China ink, India ink) are typically very small, around 100 nm, hence invisible with optical microscopes. It would appear as a transparent light gray film. The brown hue could be due to (degraded) binder or other impurities. "and thus consistent with an iron-gall ink of particularly low iron content [...] it is our conclusion that the ink present is simply an iron-gall ink of particularly low iron content." Where "particularly low" means zero. "Iron-free iron-gall ink" is an oxymoron, like "grape-free Chianti wine". As explained above, heir "conclusion" is simply an automatic guess based solely on their prior assumption: "dark ink on parchment can only be iron-gall". "Samples 9 and 17 also contained small amounts of mercury" How much was "small amount"? 10% or 0.01%? "PLM examination of [samples 9 and 17] did not show the presence of vermillion (mercury sulfide), a common red pigment. The origin [of the mercury] remains unknown" Thanks for the very informative analysis. By the way, sample 17 was supposed to be the *black ink* from the face of a nymph on f70v, which (like most nymphs) got red dots sprinkled on cheeks at some point in the book's history. The mercury contamination may well be from a small amount that "make-up" --- which was probably vermillon after all -- picked up together with the "black ink" sample. Could it be that the black ink had been stroked after and over the "make-up"? This is a question that the report should have answered... "Blue paint: The blue paint [sample 4, from a blue flower on page f26r] was unambiguously identified as gound azurite with minor amounts of cuprite, a copper oxide" On the B 2014 scans one can distinguish two kinds of blue paint. There ia a "pure blue" paint used, for example, on f80v; and there is a "dark grayish blue" paint used, for example, on f4v. The former is indeed quite probably azurite. The latter seems to be a mixture of "pure blue" with some dark gray substance that behaves very badly: it collects into a dark irregular fringe along the edges of the painted area, which leaves a dark gray offset print on the facing page. From a 2014 image of that page, provided by a Certain Library, the blue on the flowers of f26r seems to be almost pure blue, with only a little bit of the dark gray component. Could the dark gray component be just cuprite (which *is* black)? But cuprite should be inert; by itself it does not seem to explain the dark fringe and the offsetting. That should have been investigated. To be sure, it is not a fault of the Lab, but of the selection of areas to be sampled. "Infrared spectroscopy [of the blue paint sample] was not performed, but a gum binding medium is the most likely, as consistent with other samples tested" Again, a honest report would have a period after "not performed" and omitted the rest of the sentence. Instead they provided a totally unfounded guess, based only on "most paints use gum as binder". Which is not even true, since they report that two samples had a "proteinaceous" binder. "Sample 12 [from the bonnet of a nymph on Bio f78r], the clear/white material, was identified as proteinaceous, with a large amount of calcium carbonate present. A mixture of glair (eggwhite) and calcium carbonate is likely" But the parchment itself is "proteinaceous" and its surface was almost surely smoothed by sizing it with calcium carbonate. So a possible explanation for the above result is that sample 12 was improperly collected and turned out to be a bit of parchment rather than of "white paint". In fact it is questionable whether those areas have much paint, or any paint at all. "Green paint: The green paint was identified as a copper and copper-chlorine resinate, most likely produced as a salted copper corrosion product. PLM indicated the presence of both isotropic green-stained transparent material consistent with copper resinate, [...] In all cases, the amount of copper is in significant excess to that of chlorine, and something like a copper resinate, that is, a copper-containg amorphous organic material, is likely. None of the classical resin were found in this sample by IR spectroscopy, only the gum binding medium was identified in all paints and inks." In other words, they have no idea of what the green pigment is. And therefore cannot tell whether it is something that could have been available in the 1400s, like copper stearate; or a modern pigment, like copper phthalate... But on the 2014 Library scans of the Manuscript one can see two very different kinds of green paint. One is bright bluish green, extensively used in the Herbal section, e.g. on f8r. The other is a duller yellowish olive-green extensively used for the water on the Bio section, e. g. on f75r. On some pages, like f40r, both paints are used in different areas, making the difference very obvious. Samples 3 and 7 were both taken from the bright bluish-green paint on f26r and f47r, while sample 11 was taken from the green pool water on f78r. This is what table 1 says: 3 bright green] "copper organic complex, atacamite (possible), CaSO4, CaCO3, gum binder" 7 brightgreen "copper organic complex, atacamite (probable), gum binder" 11 dull green "copper organic complex, atacamite (probable), Sn and Fe compounds, azurite and cuprite (traces)" "PLM [also] indicated the presence of smaller amounts of anisotropic green particles. The presence of chlorine in the EDS spectrum suggests that the crystalline material may be atacamite or other copper-chlorine compound. X-ray dffraction of the sample produced no pattern whatsoever, suggesting that the bulk of the material is non-crystalline. These statements do not specify which samples had the "chlorine" in the EDS spectrum and which did not. Atacamite Cu2Cl(OH)3 is indeed a yellowish green mineral, which *could* be the yellowish green pigment of sample 11 (but not of the other two samples). It is however a very rare mineral that was described only around 1800. While it does occur in patinas and copper-based paints exposed to salt, it seems ulikely that it would be used as the primary pigment of a green paint in the 1400s. (The Wikipedia page says so but cites an unreliable source that may not say what it is claimed to say.) But the Lab did not really identify those crytalline particles as as atacamite, since (again) the XDS test seems to have been botched. Again, their "identification" of those particles as "atacamite" is only a guess, with no justification other than it too being a chlorine- and copper-containing mineral. In fact, it seems that the Lab failed to notice that sample 11 was of a different paint, and thus forcefully mashed together the results into a single "analysis" of the "green paint". "Red-Brown Paint: The red-brown paint [sampele 5 from f26r and sample 8 from f47r] was identified as a red ochre by PLM and EDS in the SEM. XRD characterized the crystal phases present as consisting of hematite, iron sulfide, possibly minor amounts of lead sulfide and palmierite (a lead-potassium-sulfur compound), in most likely, a gum medium" Again it failed to specify whether the iron sulfide and the "minor amounts" were 10% or 0.01%. Other than that, this paragraph was doing well until the "and palmierite". Palmeirite K2Pb(SO4)2 is an exceedingly rare mineal detected in small amounts around volcanic fumaroles. It is so rare that the only thing that a Certain Online Encyclopedia says about it is "its presence was detected on a Certain Manuscript" -- citing this Report as the source. ARRRRHG!!! What must have happened is again a blotched test. X-ray diffraction (XRD) shoots a bean of X-rays at the sample. If its heavier atoms are in a regular crystal lattice, the X-rays are scattered out in specific angles that depend on the spacing of the atom layers in different directions. The mineral can then be identified by looking up those numbers in a library of mineral XRD data. If the sample is big and clean, the best match is probably correct. But if the sample is tiny and contaminted by other stuff, the data may be noisy and the best match from the library may be quite wrong and nonsensical. It is like looking at a blurry high school album photo from the 1940s and "identifying" one of the kids as the star of /Home Alone/. TABLE 1 1 f26r $35 $102 "black ink from text" "iron-gall, gum" 2 f26r $73 $55 "black ink from drawing" "iron-gall KPbO, KHPO4, CaOHSO4, gum" 3 f26r $142 $67 "[bright] green leaf" "copper organic comp, atacamite?, CaSO4, CaCO3, gum" 4 f26r $23 $40 "blue flower" "azurite, cuprite (minor)" 5 f26r $197 $91 "red-brown root" "red ochre (hematite), lead oxide, K compounds" 6 f47r $84 $126 "black ink from text" "iron-gall (low iron), CaSO4, CaCO3, gum" 7 f47r $43 $54 "[bright]green leaf" "copper organic complex, atacamite?, gum" 8 f47r -$8 $88 "red-brown root" "red ochre (hematite), iron sulfide, palmierite K2Pb(SO4)2" 9 f78r $36 $127 "black ink from text" "iron-gall (low iron), CaSO4, Hg (traces) gum" 10 f78r $74 $28 "blue water from pipe" "azurite, cuprite (minor)" 11 f78r -$36 $48 "green pool" "copper organic comp, atacamite (prob) Sn and Fe comp, azurite&cuprite (traces)" 12 f78r -$36 $49 "clear bather hat" "CaCO3, proteinaceous" 13 f86v? $25 $67 "black ink from text" "Iron gall, K comp, CaSO4m CaCO3, gum" 14 f86v? $40 $10 "black ink from drawing" "iron gall (low iron)" 15 f26r $8 $8 "black ink from page num" "iron gall (high P, Fe), gum" 16 f116v $69 -$29 "black ink from text" "iron-gall (high iron) gum (no photo)" 17 f70v $125 -$143 "black ink from nymph face" "iron-gall ink, Hg comp, Ti comp, Sn cmp particle, gum" 18 f70v $125 -$143 "white paint fr nymph face" "proteinaceous, starch traces" 19 f8v -$2 -$128 "black ink from quire mark" "iron-gall ink (very low iron), gum" 20 f1r $25 $6 "black ink fr 'a' in stain" "iron-gall ink (very low iron), bone black, titanium cpd, binder"