I have finished my review of the Vatican Graecus $1594$ manuscript and it is available in my Google Doc. If you’d like to view the original manuscript, it can be found here, starting on $155r$1.
Discussion of this catalog below the fold.
Peters & Knobel describe this as “the most beautifully written Greek manuscript of the Almagest thus far discovered” which is a sentiment I’m quite tempted to agree with.
The hand is extremely regular and easy to read. The columns are organized like many other manuscripts with ones for the description, sign, longitude, direction north or south of the ecliptic, the latitude, and magnitude. However, what’s interesting is that there are no breaks between constellations.
Instead, a small circle is placed to the side of the table to indicate the beginning of the constellation. Below is an example for Scorpio (Σκορπιο).
While the circle is denoting the constellation, somewhat frustratingly, the line the circle points to is not necessarily the start of the constellation. Rather, the small lines with the hooks seen at the left and just above left of the circle, indicate the line below which the first star of the new constellation is listed.
Breaks are taken between lines to indicate the more major divisions of the northern, zodiac, and southern constellations.
There are also some lovely illustrations enhancing the overall look of the table.
As usual, here’s a collection of exemplars for the constellation signs
| English | Character |
| Aries |  |
| Taurus |  |
| Gemini |  |
| Cancer |  |
| Leo |  |
| Virgo |  |
| Libra |  |
| Scorpio |  |
| Sagittarius |  |
| Capricorn |  |
| Aquarius |  |
| Pisces |  |
As far as relationships to other manuscripts, I strongly suspect that the Paris $2390$ manuscript is related. Since the Paris $2390$ manuscript is dated to later than this manuscript, this would suggest that it was based on this one or shared a common source.
The first big clue is that the Paris $2390$ manuscript makes use of the same circles off to the side to denote the different constellations.
There are also a number suspicious instances in which the Paris $2390$ manuscript does have some rare variants in common.
For star $122$, my reading for these two manuscripts is $20;00$ into Libra which is an extremely rare variant. However, in the Paris $2390$ manuscript, the character h ($8$) is present which should give a reading of $28;00$, but I have chosen to ignore it because it is distinctly different in character, appearing much fuzzier. Thus, I took this as a later correction and not original to the copy.
For star $160$ the latitude is $+56;30$ which is an extremely rare variant. While I have taken the reading in the Paris $2390$ manuscript as $+50;30$, I do note that the character for $6$ is faintly present2 which would make the two consistent.
We also get an entire section of similarities with stars $168-177$. This manuscript incorrectly gives the sign for each as Cancer when it should be Aquarius. For the first several of these ($168-172$) the Paris $2390$ manuscript appears to have both signs but the Aquarius is more prominent. However, for stars $173-177$, the Cancer sign is the more prominent. Regardless, that Cancer is present at all is something that, in the manuscripts I have reviewed, only these two share.
We can also consider star $218$. Both manuscripts have the longitude written in a very odd manner. The reading I have taken is $15;15$ which should normally be written as ie d‘ ($15 \frac{1}{4}$). However, instead we find ie ie‘ which should be read as $15 \frac{1}{15} = 15;04$. However, this value does not appear anywhere else in the catalog. Thus, I take the same reading as Peters & Knobel of $15;15$. However, this particular odd writing is one that I have only encountered in the Paris $2390$ manuscript. Peters & Knobel indicate that $15;15$ is the reading they take for many manuscripts, but they do not indicate whether this particular oddity is appearing in other manuscripts that led them to this reading. Thus, I can only say it does not appear in any of the other manuscripts I have reviewed.
| Vatican $1594$ | Paris $2390$ |
 |
 |
Next is star $221$. This one has a very similar error to the above. However instead of having ie, we have ig as the fractional portion. However, this time, I have encountered this value in one other manuscript: The Vatican $2056$3. Regardless, this is still a very rare variant.
| Vatican $1594$ | Paris $2390$ |
 |
 |
We can also look at star $234$. This star has the sign incorrectly given as Taurus in both manuscripts which is not something I have encountered in others that I have reviewed. That being said, the distance into Taurus it gives ($21;50$) is reported by Peters & Knobel to be in other manuscripts which makes me wonder if the sign is incorrect at them as well making this a more common error and not particularly representative of a relationship.
Next is star $247$. For this star, the magnitude is suspiciously missing in both manuscripts. Peters & Knobel do not indicate that the value is missing in any of the other manuscripts they reviewed, although the section on magnitudes is not as comprehensive as the rest of their review and only includes six period manuscripts. Thus, again, there may be others which are simply not apparent at my present stage of research.
There is also a similarity for star $543$. This manuscript incorrectly gives the longitude as $23;00º$ into Scorpio when it should be into Libra. This same error occurs in the Paris $2390$ manuscript, but is not one that occurs in others I have reviewed. Unfortunately, Peters & Knobel doesn’t indicate that this error occurs in any other manuscripts either. In their section where they discuss each star, this one is not even mentioned. The later sections only consider the distance into the sign and not the sign itself. As such, it’s possible this error is more widespread than I am currently believing, but this will have to await further review.
Star $667$ is also interesting. In both cases, I have taken the reading for the latitude as $14;45$ which is consistent with other texts. However, this is a bit specious. The reason is that the value is written as id La‘. The a is the odd character here. If I were to interpret this whole value literally, it would be $14 \frac{1}{2} \frac{1}{1}$4. However, $\frac{1}{1}$ doesn’t make a lot of sense since this would just be $1$ and thus should be moved to the ones place changing the d (4) to e (5). Thus, the reading doesn’t make sense. However, a is very similar to d5. Thus, the intended reading can easily be interpreted as $14 \frac{1}{2} \frac{1}{4} = 14;45$. But only in these two manuscripts do I find this error.
There are certainly a fair number of cases in which case the two manuscripts do not match. However, the majority of these are the sort of errors in which it may not have been clear whether a character was intended to be a whole number or the fractional component. Thus, these are, to my mind, a weak test of relationship.
- The index seems to use a slightly different numbering system than the banner along the bottom. My reference here is the bottom banner.
- So perhaps originally included by the scribe, but removed later?
- Which I have not yet created a post about although the values are all included in the Google Sheet.
- The L I have used here represents the character for $\frac{1}{2}$ which doesn’t appear in any of the fonts I’ve found and looks a bit like a c which the bottom of simply trails off. The a would simply be treated as the inverse like most of the other fractions we routinely encounter, hence $\frac{1}{1}$.
- Indeed, in many places in this manuscript, the d even has the little hook on the top of the a making them quite hard to tell apart if it weren’t for the fact that the d has a flat bottom which extends well past the triangular top in this manuscript.