The Voynich Ninja

(24-04-2026, 10:21 AM)Aga Tentakulus Wrote: You are not allowed to view links. Register or Login to view.When I look at the writing on the other side, I don’t see any pigments in the ink or on the parchment. The ink is dissolved and isn’t pigmented.

You can't tell that at this magnification.  Not even in the higher magnification images included in the McCrone report (~180 pixels/mm).  The solid pigment particles in watercolor or lampblack ink/paint are way too small to be distinguished in those images.

But I think that the particles of the blue pigment (azurite) can be seen,just barely, in the McCrone images of the blue paint:

[attachment=15292]

Those white and dark blue dots in the blue-painted area could be individual particles of the mineral, rather coarsely ground.  Which would be consistent with the general "improvised" or even "home-made" nature of the paints (and of the painting instrument -- apparently not a brush but a broad-tip quill pen).

On the other hand, I suspect that the bluish green paint used on most of the leaves may have been a soluble dye, seeing how strongly it shows on the other side of the vellum:

[attachment=15294]

McCrone did not identify that green paint; they only determined it contained copper and was not crystalline, hence not malachite or other mineral, and guessed it would be a salt of some unidentified organic acid (a "resinate" in their words). 

My wild guess is that it is copper stearate, which can be made by mixing soap with a copper salt like blue vitriol.  It is insoluble in water but should be soluble in turpentine or any other solvent used for oil paints.  And it seems that organic solvents like turpentine, unlike water and water-based inks, can penetrate the vellum. I gather that vellum is made slightly water-repellent precisely to prevent ink "flaring" and bleed-through.

The yellowish-green paint seen on the first image above may be a different pigment altogether, or a mixture of that bluish-green paint and some yellow pigment.

By the way, the shiny nearly-black ink at the center of the second image is what I believe iron-gall ink should look like.  I could contrast it to the much fainter yellowish-brown trace below it.  But that would take me into the forbidden BEEP territory...

All the best, --stolfi

I'm slightly confused. In the interest of actually aiding in this conversation, answers to the following would be helpful:

-Is there a certain way IGI should or should not get absorbed/ bleed through the parchment? 
-Is it normal that IGI soaks through to the other side of a sheet of parchment, should it only happen in certain circumstances (such as excess moisture)? 
-Should IGI rub off onto pages it is facing once dry?
-When IGI soaks into the vellum, is it really "dissolved", or is it the same chemical structure within a porous material? (like water in a sponge is not dissolved)

If we know the answers to these we can potentially test some hypotheses using image analysis of different pages. 

(24-04-2026, 03:05 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2026, 09:57 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.The following article shows iron gall ink as nearly transparent in 900-1500nm:  ... Quoting from the article: 

Quote:The underdrawing materials that become transparent, and are not therefore visible, include red Conte crayon, charcoal and iron gall ink.

Yes, but that is what the article I quoted said too: iron-gall ink remains opaque up to 1200 nm, while ocher becomes transparent by 800 nm.  And the VMS ink is already transparent before 940 nm (the longest wavelength used in those images).

All the best, --stolfi

The crux of the issue here is that we have two different studies now with two wildly different ranges for the transparency of their specific IGI. In that second study, it says "nearly transparent in 900-1500nm". Your images clearly also do not show fully transparent IGI, even at 940nm; some of it is still visible. Therefore our observations on the VMS ink is consistent with it appearing "nearly transparent" at 900-1500nm. 

Furthermore, the existence of two ranges in two studies implies the possibility of more ranges for more types of IGI in others. The exact composition of the ink clearly matters in this regard.

(24-04-2026, 11:44 AM)eggyk Wrote: You are not allowed to view links. Register or Login to view.-Is there a certain way IGI should or should not get absorbed/ bleed through the parchment? -Is it normal that IGI soaks through to the other side of a sheet of parchment, should it only happen in certain circumstances (such as excess moisture)?

No, the surface of vellum is treated to prevent that. Not just for IGI but for any water-based ink or paint.

If IGI soaked into the vellum, the edges of pen strokes would be fuzzy.  We see that on page f112r.  Note the indentation at the top of the free vertical edge: that indentation may be the very edge of the hide sheet, or of a large hole on it.  For that or some other reason, it seems that the surface was not properly treated in an area several cm wide around that indentation.  When the Scribe wrote the first line of that page, the characters that fell in that area became fuzzy.  He avoided that area on the following lines; now and then he tried to get back into it but ran into the same problem.  Only by line 22 or so he could get back to the intended margin

[attachment=15297]

On the other hand, oily ("apolar") liquids apparently can soak into vellum even in goo areas.  The "ketchup" sauce stain on pages around You are not allowed to view links. Register or Login to view. and f103r  seems to be evidence of this:

[attachment=15299]

The creation of that "artistic intervention" seems to have been a complicated process, which tried to unravel in You are not allowed to view links. Register or Login to view.. There was a big spill of sauce between pages You are not allowed to view links. Register or Login to view. and f103r, and a smaller drop between You are not allowed to view links. Register or Login to view. and f104r.  The book apparently was closed without cleaning the spilled sauce, which was squished out into flat roundish blobs.  Later the spills were cleaned, and there was an attempt to repair their damage on f103r.

But the relevant point is that apparently the sauce had two components, a reddish orange "watery" one, and a not-so-reddish orange "oily" one. 

[attachment=15298]

Looking at the small stain on 103v (or f104r), it seems that the watery component stayed put where it had been squished out, while the oily part soaked into the vellum creating a fuzzy halo around the watery stain.  That oily component also bled through the vellum and stained the other side of the folios (f103r and f104v). 

The same happened on the big stain of You are not allowed to view links. Register or Login to view. and f103r, except that the oily component that bled through to You are not allowed to view links. Register or Login to view. then ofsetted onto f104r, bled though it to f104v, and even left a faint smudge on f105r. Whereas the watery component apparently stayed put  on You are not allowed to view links. Register or Login to view. and f103r (but dissolved the ink on the latter, which came off almost completely when the spill was finally wirped off).

Quote: -When IGI soaks into the vellum, is it really "dissolved", or is it the same chemical structure within a porous material? (like water in a sponge is not dissolved)

IGI is soluble when freshly made, but does not soak into the vellum.  Instead it bonds chemically to the vellum; and soon, as the iron in it oxidizes to the 2+ state (as in green vitriol) to the +3 state (as in ocher, rust, hematite, etc), the ink turns into an insoluble black polymer. 

Quote:-Should IGI rub off onto pages it is facing once dry?

IGO was used on vellum precisely because that polymer was chemically bonded to the vellum and could not be removed by rubbing or washing.  To erase IGI one had to scrape away the surface of the vellum, down to the bottom of its pores.

It could offset to another page if the latter was placed over it when the ink was still wet, of course.  But presumably the Scribe was good enough to wait until the ink was dry (if not polymerized) before that.

Quote:The crux of the issue here is that we have two different studies now with two wildly different ranges for the transparency of their specific IGI. In that second study, it says "nearly transparent in 900-1500nm". Your images clearly also do not show fully transparent IGI, even at 940nm; some of it is still visible. Therefore our observations on the VMS ink is consistent with it appearing "nearly transparent" at 900-1500nm.

There is no contradiction.  The article I quoted says that ocher becomes transparent by 800 nm, while IGI remains visible up to 1200.  The other says that IGI becomes transparent between 900 and 1500.  The VMS images show its ink becoming transparent around 850 nm or so.

All the best, --stolfi

I appreciate all of the answers regarding how the ink interacts with the parchment, thankyou!

(24-04-2026, 03:16 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.There is no contradiction.  The article I quoted says that ocher becomes transparent by 800 nm, while IGI remains visible up to 1200.  The other says that IGI becomes transparent between 900 and 1500.  The VMS images show its ink becoming transparent around 850 nm or so.

All the best, --stolfi

How does this not contradict?

You are basing the claim the the VMS ink is not IGI on the fact that it appears almost transparent at ~850nm, which supposedly shouldn't occur in IGI until ~1200nm according to your linked study. 

But the other study shows that their IGI appeared nearly transparent at around the same range as the VMS ink (~50nm difference, but this is based on the subjective description of "nearly transparent"). Therefore the VMS ink becoming nearly transparent at ~850nm does not necessarily indicate that it isn't IGI.

By the same logic, you should assert that the IGI in the second study is not IGI at all, because it is nearly transparent at ~900nm, while IGI should be visible until ~1200nm. 

I downloaded the images and checked. At 870nm, the text is still faintly visible. At 940nm, the text is fully invisible. This suggests that the IGI became transparent at around ~900nm, which is exactly consistent with the IGI in the second study. 

870nm (file Voynich_17r+MB870IR_038_F)
[attachment=15300]
940nm (file Voynich_17r+MB940IR_012_F)
[attachment=15301]

(23-04-2026, 08:18 PM)LisaFaginDavis Wrote: You are not allowed to view links. Register or Login to view.I see. Show-through can sometimes be visible under UV light, depending on how deeply it has seeped into the parchment. For example, mirrored show-through from 116r is clearly visible on the UV image of 116v:

Could that come from wet ink a rush to show it without proper time to dry it?

(24-04-2026, 03:05 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.Yes, but that is what the article I quoted said too: iron-gall ink remains opaque up to 1200 nm, while ocher becomes transparent by 800 nm.  And the VMS ink is already transparent before 940 nm (the longest wavelength used in those images).

As I understand it, the article I quoted states the opposite - iron gall ink becomes already transparent at 900nm. In any case, I think I've found above three different sources that present findings contradictory to the claim that becoming transparent before 940nm is not possible for an iron gall ink.

To tell the truth, that’s no longer relevant.
Now that we know how it works, the question of modern forgery arises.
How many times did Wilfried have to take the book into the steam room to achieve that effect?
Answers:
Tavi: The wrong trade for modern forgery.
Koen: OK, it’s Friday.
Rich: Now I have to take parchment into the sauna too.
All for science.

(24-04-2026, 08:50 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.As I understand it, the article I quoted states the opposite - iron gall ink becomes already transparent at 900nm.

I have to go to the Univ to access that article.  But from what you quoted it seems to use broadband IR, with wavelengths from 900 to 1500 all together.  

All the best, --stolfi

(24-04-2026, 09:46 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.I have to go to the Univ to access that article.  But from what you quoted it seems to use broadband IR, with wavelengths from 900 to 1500 all together.  

Yes, and if it was absorbing half of this band it would be visible, I assume.

Another paper I quoted stated that low absorption of iron gall ink in IR makes the detection of the ink difficult:

Quote:However, the detection of iron-gall ink can be difficult, as it is not readily visualized by IR imaging due to its low absorption in the IR.

The third (publicly available) specifically talks about 840-1000nm light and demonstrates how one type of iron gall ink remains fully visible, while the other becomes faint (though still visible). The manuscript is from the XVI century, and in visible light both types of ink appear high contrast, well preserved and easily readable (as expected for the aged iron gall ink). You are not allowed to view links. Register or Login to view. (paragraph 14)

As it appears, the medieval iron gall ink is not an ISO standardized substance, it is a wide range of different preparations with different properties. If some of them are fully visible up to 1200nm, this doesn't necessarily apply to others.

(24-04-2026, 10:03 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.The third (publicly available) specifically talks about 840-1000nm light and demonstrates how one type of iron gall ink remains fully visible, while the other becomes faint (though still visible).

I still haven't got the two other papers you cited.  Maybe tomorrow.

But on that Fig.5 of that last paper we can see two inks, one still quite dark while the other is faint.

I propose that the authors made the same mistake that McCrone did: they determined that both inks contained iron, and, since they were both black and on vellum, they assumed that both were IGI.

In an attempt to explain why some of that black ink faded in the infrared, they hypothesized that it was IGI of inferior quality.  They also speculated that it could be "cheaper" IGI using copper or zinc instead of iron.  But they did not really determined that; and anyway those metals cannot replace iron in IGI, because the polymerization of the latter happens as iron changes from the +2 to +3 state -- whereas copper and zinc are +2 only, and cannot change to +3.
 
Well, I propose that the ink that faded on Fig.5 simply was not IGI.  

The reason people used IGI was, again, because it was waterproof and rubbing-proof.  But it came in only one color -- black.  

Therefore, any inks or paints of other colors that one sees on velum are not IGI, but watercolors (tempera, gouache): suspensions of solid mineral or lampblack particles in water, with a binder that would hold glue the particles to the vellum when dry.  That incluedes the red ink used for highlighting and titles in many manuscripts of the time.  But the adhesion of binder to vellum is by weak (Van der Waals) bonds, not the strong (covalent) bonds that tannin makes with proteins; and all binders that could be used at the time could be softened by water and degraded by humidty.  

Therefore, any colored inks or paints would not be waterproof or rubbing-proof.

But then, if the book or document was to have writing in colors other than black, there would have been no point in using IGI for the black text.  

But then, on such books the Scribe may as well have used a mineral or lampblack tempera color also for the black text.  Which would be much easier to prepare than IGI, and would last indefinitely.

Therefore, when reading any paper (like that one) about properties of IGI on medieval manuscripts, one must read carefully to see whether they have actually determined that the ink is IGI, or just assumed that it was.

Apart from that, note that the paper I cited was specifically concerned with distinguishing IGI from ocher-based watercolor ink in drawings under paint.  Not just in visualizing ink under paint.  So, if indeed there is a contradiction with those other papers you cited (and I still don't think there is), it is my experts disagreeing with your experts -- not the experts disagreeing with me.  Another instance of Forte's Law...

All the best, --stolfi

PS. Another puzzling detail in that article is that, on Fig.7, they indicate the sampled areas as being the black heads of three notes; but the element composition shows as much mercury (Hg) as iron (Fe).  I can't imagine why a black ink would contain mercury.  I suspect that in fact the measurement was made on the note before the area "1" and/or the one between "2" and "3", which are crossed by red staff lines.  And cinnabar (mercury sulfide) was a common red pigment.