VMs: Re: Inks and retouching

[Jorge Stolfi <stolfi@xxxxxxxxxxxxx>]

Glen writes

  > [Glen:] While there is an indelible quality to the ink, the darker
  > portions remaining appear to be a component that is very sensitive
  > to moisture. Some time ago I examined this phenomenon on color
  > jpegs, and reached the conclusion that the two portions of the ink
  > in question did not mix well together. The indelible portion
  > appears to be an organic non-soluble, possibly an organic oil or
  > acid of some sort, capable of etching the vellum. I'm not up on my
  > ink chemistry, so the exact chemical nature of the indelible
  > compound is an assumption. The dark portion(s) in question appear
  > to be water soluble.

  
The assumption that the ink color is due to a mixture of two or more
colored components, some of them soluble and some not, is quite
reasonable. It does make color analysis more complicated (especially
without those calibration strips...).

However, I do not think that the color variations can be explained by
the *soluble* component being lost. On most pages there is no evidence
of water damage, or even moisture-induced smearing. Except for a few
isolated cases where the water stains are themselves quite visible, I
believe that the only viable explanations for ink fading are (a)
mechanical wear or (b) chemical aging of the pigment/dyes themselves.

  > The two together have separate absorption properties, and I
  > wouldn't be surprised to find that most "retouching" evidence is
  > caused by this dual-based ink and the effects of age on the two. I
  > still stand by this conclusion.

I grant that some of the dark/light variation is probably due to
mecanical ink flow effects and differential wear/fading. However there
are many, many places where that explanation does not seem to work.

It is pretty obvious that the EVA "iin" (your "m") on f1r was
retraced, and there are many similar examples all over the manuscript.

Those retracings may have been done (A) by the author himself, as he
was writing -- e.g. when the pen went dry halfway through a character.
Or (B) by the author, at a much later date, to restore text that had
faded over time; or (C) by a later owner, for the same reason.

There are indeed many examples of (A); I have been just staring at
seveal of those on f81v (top left). In these examples the under-ink is
clearly the same color as the retracing ink, only fainter, and is
clearly the end of an inking.  In fact, I have looked through most
of the biological pages, and the only retracings that I can see
on the text are of this type. (The drawings may be a diferent story,
but I am not ready to make guesses on them yet.)

However, there are many examples elsewhere that cannot be explained
as (A). That covers most of the dark traces in f56v and in the
Zodiac section. There the "old" and "new" inks are too consistently 
different, the "old" text is uniformly fainter, and the retracing
seems to be consistently confined to certain parts -- such as 
labels in a certain band, or on a certain half of a circle; 
or all those right breasts.  Moreover the retracing there clearly
seems to have been the work of someone who did not understand
the drawings --- see the moustache on f56v, the nymph legs on the zodiac,
etc. --- which seems to exclude (B) as well as (A).
  
  > If you remember in an earlier email I asked why the dark ink tends
  > to run in "veins"?
  
First, a note of caution: beware that the human eye has a tendency to
see lines in images that contain only random noise. Remember Powell's
Martian Canals? For instance, when staring at the highly magnified
images, I see all sort of drawings and characters in the blank vellum.
I mostly learned to ignore them, but sometimes they seem so compelling...

So the "veins" may be real, or may be an illusion. But I grant that
the dark ink is often concentrated in certain areas of the page. Those
may be simply raised spots on the vellum surface, which rubbed against
other pages or the table, and hence had more wear. This wear is quite
visible on sharp creases, and the vellum surface is likely to have
less conspicuous waves (say 1mm high and 1 cm wide) which would not be
visible on the images but would wear a lot more than the nearby
throughs.  

Note that this explanation fits with (B) or (C) but not with (A),
since those gentle waves should not have made much difference
while the text was beeing written.

  > Let's take f3v as demonstration that my theory
  > has basis in observable fact. There's more than one mechanical (or
  > chemical) effect observable on this folio, but right now I'm only
  > interested in one. Here we see the image of the paint from f3r,
  > and notice that the moisture from the paint on the other side of
  > the folio causes some of the soluble portion of the ink to adhere
  > better, causing the ink to appear darker.
  
Are you saying that one side was written after the other side had been
painted, and while the paint was still moist? It seems highly unlikely.

Moreover, vellum (unlike paper) seems to be mostly impermeable. The
"bleedthrough" effects that we see are largely due to it being
slightly transparent, especially where a little dent on one side
matches a little dent on the other side. (Ever wondered why "vellum
paper" got that name?)

I haven't looked for them, but I bet that there are few if any
instances where the ink actually flowed across the vellum. Certainly
no such "material" bleedthrough is visible on f3r, for example. In any
case, there doesn't seem to be any correlation between the "retraced"
characters and bleedthrough.
  > We see many dark portions on the edges of the painted image, and
  > in a few places we see a glyph split between the painted image and
  > regular vellum, where the portion of the glyph inside the painted
  > image is dark, but that outside is faded. f3r, line 5, the <ch> in
  > word 2. ...

The green bleedthrough does indeed make the characters written over it
look a bit darker. This is an expected optical effect and does not
need to involve any moisture transfer. 

One interesting detail on page f3r (a page which I had not yet
looked at in high-res) is the cracked green paint, on the bottom
right leaf. Note that even though this area was positively
flooded with paint, it did not cause any more bleedthrough  
than other places where the green paint was applied with an 
almost dry brush.

Another interesting detail is the way this green paint darkens with
increasing thickness. I don't know, but we may be looking at a green
dye rather than a green opaque pigment. (The text ink on this page too
looks transparent.) I wish I had looked at this page yesterday: I just
spent many hours trying to separate the green ink on page f81v
(biological section), and was frustrated -- that too seems to be a
transparent dye.

If it is indeed a transparent dye, I have no idea of what substance it
could be. It may be much later than 16th century, perhaps 19th or 20th
century. A student at the Collegio Romano, perhaps?

I am well aware of this effect and I *do not* count those slightly
darkened characters over bleedthroughs as retracing evidence, not even
as "suspects". On the very first word of f3r (EVA "tsheos"), I would
say that the final EVA "s" is of the same color as the legs of the
initial "t", except for the effect of darker background.

On the other hand, there are many cases of a single character or
stroke being much darker than its neighbors, with all of them are over
the bleedthrough, or all of them are outside it. The "o" on that
same word is an example.

Incidentally the darkening of the strokes over the bleedthrough
may allow us to decide whether the brown ink is transparent dye
or opaque pigment.  In both cases we would expect a darkening,
but the numerical effects on each channel should be different.

Most of the darker letters on f3v are what I would call "suspects"
only, not evidence. The ink seems to be the same, and the "hand" too.
So indeed they may well be instances of (A) above. E.g., after the
author finished writing a paragraph, he went over it again and
reinforced those letters which had come out too faint on the first
pass -- paying special attention to those which are likely
to be confused, like "o" and "a"s, "r"s and "s"s.

There is marginal evidence of retracing is on the first "tsheos".
Note that the horizontal stroke of the "t" does not connect to
the left eye. Note that the plume is much fainter than any text
around it; so it it was added later ("crossing the sh's") it must
have been much after the end of that word. 

The first word of line 6 ("ychtaiin") also has a marginal evidence:
a bit of light ink sticking out of the right foot.  A small bit,
but note that the pen was obviously quite loaded, whreas the 
"old" stroke seems to have been done by a very dry nib. 

Other similar bits are seen on paragraph 2: at the very bottom of the
first "s" in "sols" (line 1, word 3), which may even have been an "r"
originally; and at the top of the second "i" of an "okodaiin" (line 2,
word 1).

The reason why we don't see more of such evidence could be that there
was no retracing... or that the original text has mostly faded to
invisibility. Check for instance the plume on the first word
("tsheos"), which I would guess is "old", and the EVA "y" (your "9")
on the next-to-last word of the first parag, "otchody", which I guess
had its head (but not its tail) retraced with an almost-dry pen. Under
the "retouching" hypothesis, those details were still above the
visibility threshold at the time, so the "old" characters that were
traced-over should be even fainter.

I can't reply to all of your other points here, unfortunately. (This
thread seems to be snowballing, each line of one posting calls for two
paragraphs of reply...) I will try to expand my webpage with a more
systematic and organized list of examples. For now I can only say
that, while most instances of "retracing" could be explained by
"natural" accidents with a sufficiently magic ink, there are far too
many examples to explain them that way.

On the whole, "retracing" seems to be a much simpler explanation for
all that evidence. It does not require UFOs or bizarre sects or
super-human cryptographers/forgers, nor unspecified ink chemistry --
just an ordinary book owner doing a very simple and natural thing, for
a very plausible reason...

As for the colored paints, not only the visual evidence but simple 
common sense point to them being the work of a later owner. 
That would hardly be an unusual occurrence: see

  http://special.lib.gla.ac.uk/exhibns/month/oct2002.html

for an example of a *printed* book that was completely and carefully
colorized by a later owner. (Scroll about 2/3 down the page for the
story.) On 

  http://www.med.yale.edu/library/historical/fuchs/

you will find another copy of this same book - Fuchs Herbal - where a
few pages have been fully (but crudely) colorized, a few more had only
the leaves painted green, and the rest was still unpainted.

So I think that the burden of proof is on those who claim that the
painting is original, not on those who believe it to be a later
addition. (For instance, it would be a good start to find a single
example of brown text or outline written on top of the paint, rather
than under it.)

All the best,

--stolfi
______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list