The Voynich Ninja

f67r1 has a curious centre about which it is not easy to say whether it's Sun or Moon (or both at once), but it features twelve sectors filled with stars that "blink aboon" (as the old poem goes). At least I supposed these asterisks to represent stars several years ago, and the number of twelve being readily associated with hours, months or Zodiac constellations, I picked the latter option and tried to map the sector labels to constellation names in different languages - not directly, of course, but I mean - to find some pattern that could suggest a match. That to no avail.

Since then, the mnemonics discussion arose and went on, and now I'm inclined to think that the author heavily relies on mnemonics and the labels would be anything but the constellation names. So I dropped the labels for now.

Now, there's the symbol at 11 o'clock (marked with red circle below) which I always considered an arrow pointing rightwards. I took that for the direction of rotation or of some sequence that the reader is to follow. That made sense because this sector is clearly marked as the "initial" by the "separator" in the rings of text.

However, I noticed yesterday that this symbol, if rotated 90 degrees clockwise, resembles scales, only if the "pointer" of the arrow is considered grease (of which there is some in the folio) and excluded from consideration, because, on the other hand, this "arrow" contains a strange tail to the left, which, although, if the object is rotated 90 degrees clockwise, looks like the scales' handle.

The scales immediately suggest the sign or constellation of Libra. Note that this depiction of scales does not look as the traditional depiction of the sign of Libra, it just resembles scales schematically.

The "special" point that Libra would occupy in the figure (the "initial" position) is not inexplicable; at the verge of the AD era the point of the autumnal equinox was in Libra (while that of the vernal equinox was in Aries), so the point of the autumnal equinox has been traditionally designated with the sign of Libra since then, notwithstanding that currently the point shifted to Virgo. (The same story is with Aries and Pisces for the vernal equinox).

I then supposed that the asterisks in the sectors (which I considered stars for the sake of this investigation) represent respective Zodiacal constellations. Again, from the point of mnemonics, if you have set the point of Libra and you understand the direction (e.g. counterclockwise), you don't need to take trouble to depict individual constellations, since their sequence is perfectly known and thus can be reconstructed by any reader.

But what follows is based on the hypothesis that those are Zodiacal constellations. To test that hypothesis, I counted the number of stars in each segment (shown in figure). It's not clear whether the large stars should be added to the small stars or they are magnified references. From what we see roughly at 6 o'clock I'd say they are references, but in the table I provide two counts - one is just small stars, the other is small plus large.

Now, does the count match the number of stars in respective real world constellations? The short answer is no, since the number of stars in a constellation is really very very large, and no astronomer has ever discovered them all. Even back from modern astronomy to the times of Hipparchus and Ptolemy, the number of stars in constellations as catalogued in Almagest is much larger than what we find in f67r1. So I further supposed that the author used some handy contemporary reference (I don't know which, maybe somebody has better knowledge in that respect) which listed only the "major" stars of the constellations. Which stars would be "major"? Probably those with smaller apparent magnitude (that is, those which look brighter). So I took the tables of stars of Zodiacal constellations and counted how many of them have the apparent magnitude of 4 or less (the choice of this threshold was such that the count was more or less on par with what is observed in f67r).

Since there was no perfect match (and we don't know which threshold was used by that supposed contemporary reference), I check not for the absolute values but for the trend. Does the number increase when it increases in f67r, and does it increase when it increases in f67r, while following the circle? For most part it does, except for three cases: supposed Gemini to supposed Cancer, Cancer to Leo and Virgo to Libra. The supposed direction is counter-clockwise.

I did the count very late yesterday, so there may be errors, but I think not many. The first count is the real world, the second is f67r1 (small stars). What's in brackets is small stars plus large stars.

Virgo            10     7 (8)
Libra             6      9 (10)
Scorpio         20    12 (14)
Sagittarius    16    10 (12)
Capricorn      5      8 (10)
Aquarius       9      9  (11)
Pisces           3      8 (10)
Aries            4      11 (13)
Taurus         16     11 (13)
Gemini         13      8  (10)
Cancer          2      12 (14)
Leo              14     9 (10)

Another thing that does not fit is the fact that real-world constellations with small amount of bright stars, such as Pisces, Aries and especially Cancer, are matched, with this plot, to large figures in the diagram - figures that are higher than those of some other supposed "constellations" in the diagram. Note that if we suppose that some asterisks were situated in those half-sectors which are now painted over by the blue paint (a possibility I have in mind), this will make the situation even worse, not better.

Yes, I'm aware that apparent magnitude may change with time. And yes, I'm aware that magnitudes e.g. in Almagest are different from what we are used to today. And that borders of constellations changed with time. With Almagest, or other catalogue of yore, one would have different counts (and probably would need a different threshold). I haven't time to do the Almagest count, but a screening look tells me that it will not improve the situation for Cancer.

What I can imagine is that the author used some chart with graphical representations of constellations where the stars reproduced were chosen only in part in accordance with their apparent magnitude, while the other consideration was to "complete the picture" (of the constellation).

All this, of course, if my hypothesis is valid, to begin with.

[attachment=4330]

You've been busy!  


I can't read this until Saturday, still working, but will take a good look at it then.

I always thought as these stars as simply filling up the available space, so it surprised me that there was so much difference between their numbers, from 7 to 12.

Then I wondered if maybe there are size differences between the white sections, allowing for more or fewer stars to be placed. This is notoriously hard to assess with the naked eye, so I overlaid  the 7 and 12 from your image in Photoshop. I tried to trace the pen lines and avoid painted areas.

[attachment=4333]

I'm almost certain that numbers matter, the amount of attention to the tiniest detail is seen through all the MS.

In some sectors he took effort to fit a star very close to the center (e.g. 2 o'clock), while e.g. at 3 o'clock there evidently is room for one more star which is not used though.

At six o'clock, as I wrote above, there are suspicious lines "connecting" small stars to large stars, look like pointers.

Also, through differences in ink intensity, it is seen that the sectors were not filled in simple clockwise (or couterclockwise) fashion. That would be unrealistic if it were a simple filler.

One more thing is that there seems to be some writing here and there under the blue paint, maybe some reminders which are not needed in the finished drawing.

A useful link: You are not allowed to view links. Register or Login to view.

To find Bootes constellation look to the north. 
the 12 o'clock position translates to "Achtoru" or Archturus. 
How many stars are visible without a telescope? 9

11:50 position constellation Orion
translates to "Razulu" or "Rigel"
How many stars are visible without a telescope? 12

Hi Monica,

The number of stars visible without a telescope much depends on the individual person's eyesight, but it's an interesting idea that these groups might stand for non-Zodiacal constellations.

If we consult the Almagest, Bootes will indeed give nine stars of magnitude less than four; Orion will produce fourteen, however.

But I thought the non-Zodiacal constellations were not of interest in Astrology, were they?

They might have been useful for navigation

I have been pondering the list of 15 stars given by Diane in the You are not allowed to view links. Register or Login to view. thread. You are not allowed to view links. Register or Login to view.
They were from an astrolabe dated 1596.

You are not allowed to view links. Register or Login to view.

I took the trouble to translate these into more recognisable modern names, they were, in order:

NAVIGATION STARS

1 Sirius
2 Procyon
3 Alphard
4 Regulus
5 Denebola
6 Spica
7 Arcturus
8 Antares
9 Altair
10 Algenib
11 Baten Kaitos
12 Pleiades
13 Aldeberan
14 Rigel
15 Betelgeuse

If you check this site, the list is not unique, but only one of many:

You are not allowed to view links. Register or Login to view.

Although these are stars and not constellations, in reading around the subject there often  seems (to me anyway) to have been some conflation.

But why should these appear in a roundel of 12? One possibility is the use of the stars in telling time!

In that same thread there was a post by Marco of the 15 Behenian Stars, each linked to a specific stone and herb:
You are not allowed to view links. Register or Login to view.

(New entries marked with *)

ASTROLOGICAL STARS

1 Aldeberan
2 Pleiades
3 Algol*
4 Capella*
5 Sirius
6 Pegasus
7 Regulus
8 Algorab*
9 Spica
10 Arcturus
11 Alkaid*
12 Alphecca*
13 Antares
14 Vega
15 Deneb Algedi*

I thought to write these lists out in full for easy comparison, as with all the historical names identification can be difficult.

I did check the declinations from the astrolabe against the modern ones quoted in Wikipedia, of course they are different today due to precession, but still within 1-2 degrees.

Hoping this exercise may be useful for somebody!