The third manuscript I’ve reviewed is the Paris $2390$. A copy of it can be found here. Details below the fold.
History
Peters & Knobel give far less history on this manuscript. They note it was likely written some time in the $12^{th}$ century and describe it as “neatly written.” This is somewhat at odds with the quite they give from Halma which states1 it is “very difficult to read because of the large number of ligatures and abbreviations in the script.”
For those not familiar, ligatures are the joining of two characters. This is extremely common in the manuscript, Here’s an example:
The reading for this is $11;30$. Here, we have the ι for $10$ and then the α and symbol for $\frac{1}{2}$ have melded together. Although it took a few pages, I didn’t find these overly cumbersome.
Peters & Knobel note that Halma incorrectly stated that he viewed this manuscript in Florence as indicated by the manuscript he reviewed having the same catalog number ($2390$) as well as an identical set of errors.
The Hand
This text is written in Greek. However, the hand here is much more stylized – almost cursive than we saw with Paris $2389$.
One of the largest challenges in reading this was that the scribe apparently changed which symbols they preferred to use. As you’ll see below, several numbers had at two variants. This introduced additional characters that could be mistaken for others.
Signs
In this manuscript, the scribe has opted to forego abbreviations for the signs, instead using zodialogical symbols. These too are highly stylized. Below is a table where I’ve captured an exemplar of each:
| English | Character |
| Aries |  |
| Taurus |  |
| Gemini |  |
| Cancer |  |
| Leo |  |
| Virgo |  |
| Libra |  |
| Scorpio |  |
| Sagittarius |  |
| Capricorn |  |
| Aquarius |  |
| Pisces |  |
Longitude/Latitude
This manuscript is particularly frustrating because the scribe frequently changes between how some characters are drawn.
| Number | Symbol |
| $0$ | o (With a line over it) |
| $1$ |  |
| $2$ |  or  |
| $3$ |  |
| $4$ |  or  |
| $5$ |  or  |
| $6$ |  |
| $7$ |  |
| $8$ |  or  |
| $9$ |  or  |
| $10$ |  |
| $20$ |  |
| $30$ |  |
| $40$ |  |
| $50$ |  |
| $60$ |  |
| $70$ |  |
| $80$ |  |
While the images above are some of the cleanest examples, I came across, many characters are not well formed and hard to decipher. Without having other manuscripts and critical editions to compare to, I daresay it may prove impossible in some cases.
In particular, the difference between a $7$ and an $8$ proved challenging. While the $7$ was easy enough to interpret immediately, the $8$ varies between forms, often with the left horizontal stroke on the second form much shorter. Similarly, the small stroke at the top left of the $7$ can become extended leading these characters to look very similar.
Similarly, the characters for $40$ and $50$ are frequently indistinguishable. The tell tale difference is found on the thin tail on $40$. However, there are instances in which it is not clear if it’s present or not2.
For the fractions, we again have the problem were they are inverses of the whole numbers above. Thus, the presence or absence of the accent mark again significantly changes the reading.
There are two special characters that stand for fractions beyond a simple inverse.
| Fraction | Character |
| $\frac{1}{2}$ |  |
| $\frac{2}{3}$ |  |
Additional Notes
One of the things that struck me about this manuscript is that the table is much more condensed. With the two previous manuscripts I’ve reviewed, spaces were left between the constellations to give the total of the stars of each magnitude in each as well as the header for the next. This text does not seem to do so. Instead, I suspect this information is placed in the margins in small circles offset from the text:
However, the placement of these is not always quite aligned with where the constellations begin and end. Unfortunately, while I’ve been able to interpret the numbers, I can’t read what these say.
Aside from these additions, there is a great deal of additional notes in the margins. Frequently these appear to be corrections or clarifications to the figures in the table. Unlike previous manuscripts, we frequently find values scratched out or overwritten in the tables. Thus, these notes are frequently quite helpful.
Another important note is that Peters & Knobel did a far less accurate job of recording the values from this manuscript. They give values when they are different from the ones reported by Baily. However, I found that they frequently failed to report variants and incorrectly reported them several times as well. Thus, if you’re relying on Peters & Knobel for the reading of the text, take it with a grain of salt. I have added notes on where I disagree to my copy of the table.
Ultimately, I think my favorite little touch in this manuscript’s star catalog is the flourish placed on the final value in the table:
Compare this delta to the one right above it. It speaks so clearly from across $800$ years to declare, “Finally! I’m finished!”
Lineage
This manuscript shows a number of relationships to the Vatican $1594$ manuscript. The most obvious is the note above in which the beginnings of constellations are given in circles, offset from the table body. This is identical to the manner in which the Vatican $1594$ manuscript does this. However, this manuscript does so without as much care as the two circles are often not quite concentric.
We can also look at stars $122$ and $125$. Both of these have characters in their longitudes that are different than the rest of the text. They are fuzzy as opposed to the crisp characters of the rest of the text. This suggests these were added at a later time. If we ignore these characters, they are identical to the ones in the Vatican $1594$ manuscript.
There is also a distinct relationship in stars $199-206$. In the Vatican $1594$ manuscript, the sign for each of the longitudes is incorrectly given as Cancer. In this manuscript, this appears to be the case as well, although several of them have been overwritten and are thus, indistinct.
We can also see that for star $160$, the original value or the latitude was $56;30$ which matches the value in the Vatican $1594$ manuscript. However, the ϛ ($6$) was rubbed out and without knowing to look for it in response to the potential relationship, it would have been extremely easy to overlook.
There are several other instances in which the potential reading for this manuscript is unclear. For example, on star $3$, there is a tick mark above the ϛ which would change it from a $6$ to a $\frac{1}{6}$. However, the tick mark is well removed from the character making it unclear if it was intended or not. Suspecting the relationship makes it clear that the mark is intentional.