[adrian_b]

Actually the solstice occurred on the 25th during the years immediately after the calendar reform of Julius Caesar. This is when the 25th has become the traditional date for what has later been reinterpreted as Christmas.

Before the Julian calendar, the solstices occurred on random dates of the Roman calendar, because the duration of the Roman year differed very much from the duration of the solar year.

Today it occurs around the 22th as a consequence of the Gregorian calendar reform, which has not restored the solstices and equinoxes from the beginning of the Christian era, but those from around 325 AD, when the Christian algorithm for computing the date of Easter has been established (First Council of Nicaea).

Between Julius Caesar and the 4th century AD, 3 days of offset in the solstices and equinoxes had accumulated, due to the difference in duration between the Julian year and the tropical year.

[thaumasiotes]

> Actually the solstice occurred on the 25th during the years immediately after the calendar reform of Julius Caesar.

So?

I took my information from https://kiwihellenist.blogspot.com/2015/12/christmas-and-its... :

> Around the time of Jesus' birth, the solstice was on 23 December; by the time of the Chronography of 354, it was 20 December.

Are you saying the right date to determine the solstice for purposes of assigning Christmas was 40 years before Christ was notionally born?

Also, why are you saying that the solstice fell on the 22nd in 325 when it actually fell on the 20th? In 325, the solstice hadn't occurred on the 22nd for at least 90 years.

Also, your original point wasn't even true:

> Digression #2: why do Roman writers report the date of the solstice as 25 December? The textual parallels between Columella and Pliny, and the fact that the date was already wrong by their time, suggest they are both based on an older source. A few pages earlier Pliny discusses three treatises by Sosigenes [] (Nat. hist. 18.212). Sosigenes can’t be the ultimate origin of the date either: in his time, in 46 BCE, the solstice had already drifted to 23 December.

> The last time the solstice actually fell on 25 December, using the retrojected Julian calendar, was in [214] BCE

[adrian_b]

Your own link says: "This discussion has now been superseded in many respects. See these follow-up pieces".

At the "follow-up piece" linked from there, it says the same thing that I have said, i.e. that the current solstices have been fixed by the Gregorian calendar to be the same as those of 325 AD.

In my posting of above I have simplified, by writing as "22" the present date of the solstice.

In fact, during the 4 years of the Julian year cycle, from one leap year to another, the date of the solstice fluctuates between the 20th and the 22th, the same as it did around the year 325 AD. As I have said, the Gregorian calendar has been chosen so that the date of the Spring equinox, which is used in the algorithm with which the Church computes the date of the Easter, is the same as in 325 AD.

I do not know from where you got "In 325, the solstice hadn't occurred on the 22nd for at least 90 years", but this is obviously incorrect.

If you are right and in 325 AD the Winter solstice was on the 20th, within 3 years before 325 and after 325 there must have been solstices on the 22th, because these occurred every 4 years.

By the time of Julius Caesar the solstices occurred 3 days later, so they must have occurred between the 23th and the 25th, with the same 4-year period. It is likely that in some year when some astronomer has actually determined the date of the solstice by measurements, it happened to be a year when it occurred on the 25th, so most later Roman authors who have reported the date from hearsay, like Pliny the Elder, have said that the Winter solstice falls on the 25th.

As also explained at the link provided by you, the dates quoted in various sources for the solstice dates supposedly reported by ancient Roman authors are frequently off by a day or two due to translation errors between the Roman calendar and the modern calendar, so they must be interpreted with care.

[thaumasiotes]

> I do not know from where you got "In 325, the solstice hadn't occurred on the 22nd for at least 90 years", but this is obviously incorrect.

It's linked. https://www.timeanddate.com/calendar/seasons.html?year=300&n...

> Dates are based on the Julian calendar.

Within four years in either direction of 325, the latest occurrence of the solstice is around 9 AM on Dec 21. It seems safe to say that this could not have been interpreted as Dec 22.

In the year 325, the most recent winter solstice of Dec 22 had occurred in 235, 90 years prior, except that with a precise time of 12:15 am, I have no idea what calendar day would have been assigned to it.

As far as I can see, it is completely impossible for the solstice to wander across three calendar days in any four year period. Only two dates can ever be candidates in any given year or most stretches of twenty years.

> As also explained at the link provided by you, the dates quoted in various sources for the solstice dates supposedly reported by ancient Roman authors are frequently off by a day or two due to translation errors between the Roman calendar and the modern calendar, so they must be interpreted with care.

Well, what it actually says is that ancient Roman authors report the date as Dec 25 and that this is so obviously wrong that (a) those authors feel compelled to qualify Dec 25 as an approximate date; and (b) more modern translators intentionally give different dates, despite what the Latin says.

It also says that the Romans could only measure the solstice to within about one day, which is more in line with your comment, but not related to "translation errors". Those are measurement errors.

But we don't need to rely on reports to know when the solstice occurred. We can calculate it directly. With measurement errors of up to one day, the solstice around 325 AD could have been reported anywhere between Dec 19 and Dec 22. With actual occurrences ranging from 2:20 pm on the 20th (328) to 9:08 am on the 21st (323), the 19th and the 22nd are unlikely reports. Frankly, it seems safe to assume that the date was measured independently in several locations every year, which should give excellent estimates.