I have finished adding another Almagest manuscript to the collection in the Google Sheet. This time, it’s the RARESF 091 P95A manuscript1 at the State of Victoria Library in Australia and it has some interesting features.
This manuscript was reportedly produced in the early $13^{th}$ century, likely in Italy, and is in Latin. It is believed that this one was based on the late $12^{th}$ century translation done by Gerard of Cremona. If so, then this would be one of the earlier copies done after Gerard’s translation.
Turning to the star catalog, there’s several things that jump out immediately.
The first is the layout. The data is clearly laid out with two columns per page. This is largely done by providing a smaller space for the star descriptions which more frequently take multiple lines.
The second is that the values are not given in Roman numerals, but are given in Arabic numerals2. This certainly makes things easier to read, although it’s the first time I’ve encountered this.
Still, care must be taken when reading the text. While many of the numbers are relatively familiar, the character for $5$ used looks almost exactly like a modern $4$, while the character for $4$ is a loop.
Additionally, the character for $7$ is very similar to the one for $1$, differentiated only by a very thin diagonal line upwards to the right, before descending straight down like a $1$. If the diagonal line is too thin, it is easy to misread this.
| 4 |  |
| 5 |  |
However, these Arabic numerals appear in an odd place: They appear in the columns for the signs instead of the astrological symbols for the signs themselves. It appears that the signs are numbered but, instead of starting with $1$ for Aries, it starts with $0$, followed by $1$ for Taurus, $2$ for Pisces, etc…
| Number | Sign |
| 0 | Aries |
| 1 | Taurus |
| 2 | Gemini |
| 3 | Cancer |
| 4 | Leo |
| 5 | Virgo |
| 6 | Libra |
| 7 | Scorpio |
| 8 | Sagittarius |
| 9 | Capricorn |
| 10 | Aquarius |
| 11 | Pisces |
Additionally, when modifying the magnitudes (i.e., brighter than or fainter than), this is indicated in multiple ways. The first we encounter is with two dots below the given magnitude to indicate that it is fainter than. For example, here’s a magnitude $4$ with the two dots which gets listed in the catalog as $4-5$.
To indicate “brighter than” the dots are placed above the character.
However, a second way also appears, with an “em” in the margin for brighter. An “el” indicates fainter.
To denote whether constellations are north or south of the ecliptic, this manuscript uses the terms “septemptrio” and “meridies” respectively.
Getting started on the manuscript itself, it is immediately off to a rocky start. The first page alone has two transposition errors in which the scribe3 skipped a line causing the longitudes to be shifted up by one row until a correction was made.
While things seem to get a bit better from there, there are still some serious issues with this text. From stars $125-134$, for example, the values for the latitudes are all in the teens instead of the $50$’s and $60$’s as they should be. Interestingly, this same pattern of errors also occurs in a two later manuscripts, which indicates that they were copied from this manuscript, or they share the same source in which the error was present.
Those two manuscripts are the British Museum 2795 and Laurentis4 45.
Taking the two above issues together, I think that this indicates that those two other manuscripts share a common origin, but were not copied from, this one. Had they been, they would have inherited the issue of the transposition mentioned above. Since they do not indicates that their scribes were probably using a closely related source to this manuscript.
This manuscript is heavy with corrections, written in a later hand. However, the correctness of these corrections is frequently dubious. For example, on star $927$, the longitude of ♍ $1;40$ is “corrected” to ♍ $50;40$ which doesn’t make sense since the longitude within a sign can only go up to $29;59$.
This, of course, is a pretty extreme example. In many cases, this manuscript has values that are inconsistent with any other manuscripts and this later hand corrects them to something that is found in other manuscripts, but in many cases, this manuscript was already in line with others, and the later hand corrects it to something that isn’t. Thus, the later hand is quite inconsistent as well.
For more information on this manuscript in general (i.e., not just the star catalog), there’s an interesting paper here.
- I believe this manuscript is also called Sinclar $224$ by some sources.
- Pope Sylvester II is believed to be the first Christian scholar to teach mathematics using Arabic numerals sometime in the $10^{th}$ century.
- Or the scribe of an earlier text in the lineage leading for the errors to be inherited in this manuscript.
- This is the name of the scribe known to have transcribed the manuscript as opposed to the collection in which it currently resides. As I have not yet researched this manuscript, I am uncertain of its present location.