VMs: John M. Manly's 1922 Harpers article...

[Jorge Stolfi <stolfi@xxxxxxxxxxxxx>]

Nick Pelling quotes John Manly:

    > The ink with which the writing is done is thick, almost of the
    > consistency of printer's ink; the surface of the vellum is rough.
    > It seems almost certain that such an ink applied to such a
    > surface not by pressure, but by the sweep of a pen or brush,
    > would break up into just such shreds and filaments as the
    > microscope shows in these symbols. I cannot speak
    > positively on this point, because it is only recently that I saw
    > these marks and I have not yet been able to find an ancient
    > vellum manuscript written with ink of this consistency; but
    > a microscopic examination of old printed books undertaken
    > several years ago for another purpose lends probability to
    > this view. Moreover the strokes have a freedom of sweep
    > which does not seem compatible with the theory that they
    > are built up carefully and painfully of minute bits.

The bit "I have not yet been able to find an ancient vellum manuscript
written with ink of this consistency" is interesting. I have long
thought that there was something atypical about the VMS ink.

>From images which I have seen, the standard iron/oak-gall recipe is a
watery *solution* which, once dry, has a dense black color when
concentrated, brown when diluted; the tone is purplish when fresh, and
may change to rusty-reddish with age. It was universally used in
Europe from ~1200 on for writing on vellum and parchement, because it
binds chemically to the hide proteins and thus cannot be rubbed or
washed off. (Corretions required scraping off the vellum surface
itself.)

Being "the" standard European formula, iron-gall ink was also
generally used on paper, although it probably had no particular
advantage for that material. In fact, its iron contents tended to leak
out with age and humidity, resulting in a rusty halo aroud the strokes
and acid-like decay of the paper there.

In comparison, "India" ink, used in the Orient and possibly in the
Arab world, is a suspension of fine carbon particles (e.g. lampblack)
in water. India ink works fine on paper, because it soaks into it and
leaves the carbon particles entangled among the fibers. A bit of gum
arabic or other glue-like material makes it further resistant to
rubbing. It is not entirely waterproof, but that is not important
since the paper itself will be damaged by water. However, it is not
affected much by mere dampness. One the other hand, "india" ink is not
suitable for vellum, since the carbon is merely deposited onto the
surface and is easily rubbed off. India ink is coal-black when
concentrated, neutral gray when diluted. 

(All this is net-lore, by the way. Myself, I have seen very, very few
real manuscripts, mostly framed documents in museums.)

Now, looking at the best color images available for the VMS, I got the
impression that the ink used for the text and illustrations was
neither iron-gall nor india ink. Rather, it seems to be a suspension
of some yellowish-brown opaque powder, like a tempera or watercolor.
If true, that would be another puzzling peculiarity of the VMS...

However, that peculiar color an appearance be merely an artifact of
the image process, or due to my monitor (Sun) having a diffrent
calibration than theirs (IBM or Mac).

  ( One can tell the difference between a soluble transparent ink, like
  iron-gall (or black tea), and an opaque pigment suspension, like India
  ink (or gouache paint), from the way the color changes as the ink
  gets thiner on the page.

  With a transparent ink, light is scattered by the background, and as
  it passes through the ink layer some wavelengths get absorbed more
  than others. Doubling the thicknes squares the absorption
  coefficient for each wavelength; as a result, transparent ink
  usually changes its hue as it gets thicker, and eventually looks
  black when thick enough. (E.g. tea looks yellow on the saucer, dark
  orange-red in the cup, and black in the pot.)

  With an opaque pigment, some light is scattered by the pigment, and
  some by the background. Doubling the thickness of the paint merely
  shifts the percentage towards the pigment's side; thus a
  tempera-like ink retains its hue at all dilutions, changing only its
  saturation, and never gets darker than the pigment's own color.

  Unfortunately, the gamma distortion present in most computer images
  introduces hue shifts on its own... )

Nick then asks:

  (5) Are there any micrographs of the VMS' surface still in existence?
  (6) If so, are there any scans of them on the web?

It is not really a "micrograph", but here are some high-resolution
images of f87v that Beinecke provided us, a couple of years ago, as a
test of what their in-house scanner was capable of delivering(*).

  http://www.ic.unicamp.br/~stolfi/voynich/BeineckeScanSample-f87v-top.jpg
  http://www.ic.unicamp.br/~stolfi/voynich/BeineckeScanSample-f87v-bot.jpg
  
Unlike most images in their gallery, which were derived from color
photo negatives, these samples were digitised directly from the
manuscript, with a special IBM camera-like book scanner. At ~450
pixels per inch, these may well be the highest-resolution color images
ever made of the VMS. (I haven't checked, but presumably this scan is
available at their VMS gallery site too.)

  (7) Are there any signs of writing indentation on the pages?

What you mean by "indentation"? 

   (8) What writing instrument was used for the lettering? Quill, metal pen, 
   brush, etc?
   
If you look closely at the text, especially between the two flower
bunches, you will see several places where a single stroke splits into
two parallel tracks. At those places, apparently, the instrument was
running out of ink and/or was being pressed harder against the page.
Note also that the thicker strokes are in the NW-SE direction and have
squarish ends. These features are typical of quill pens, which were
trimmed so has to end with a rectangular "nib", and then slit
lengthwise in order to better regulate the ink flow to the tip. Brush
strokes would look *very* different.

I don't know when metal pens were invented, but they were definitely
not in common use in the 15th century.  Moreover, I suspect that 
a metal pen would have been too stiff to produce those split strokes.

Pens could be made also from hollow woody reed stems. I believe that
Arab and Hebrew scribes used those. Presumably their strokes are
similar to those of a quill pen.

In the images, note also that the general appearance of the writing
ink (color tone, sheen, the way it fills the vellum roughs, the way it
changes color as it thins out, etc.) is quite similar to that of the
fill-in paints --- which are definitely temperas (suspensions of
opaque pigments).
   
I can't tell what instrument was used to apply the fill-in colors
to the plants.  It seems to be rather stiff, with strokes
twice as wide as the text ones, at most.  It may have been a quill
pen too, perhaps with a broader nib.  Or it may have been
a rough stiff brush --- e.g. a reed with chewed-up tip.

  (9) Was the same individual writing instrument used by both Hand A
  and Hand B?

There is a report on the VMS done by a handwriting expert, available
from Jim Reeds's site (IIRC). He flatly says that the whole book is
written in a single hand.

Of course, as a Fortean Times reporter said, "for every expert there
is an equal and opposite expert"...

All the best,

--stolfi

(*) As Jim Reeds explained a while ago, Beinecke has long since given
up on doing the scanning themselves, for serveral good reasons. (One
of them was that their oldish scanner depended on strong lamps that
would have "cooked" the VMS.) AFAIK, are in the process of contracting
the services of an outside scanning firm, and eventually they may get
to do the VMS.