41 thoughts on “Was the winter solstice on 25 December in the Julian calendar

  1. The winter solstice in the Julian calendar was on the same day it is on the Gregorian calendar which after all is the corrected Julian calendar: December 23rd. Now as the Julian calendar drifted out of sync with the solar year the solstice drifted earlier, not later hence during the the Ottoman occupation the people would say that on St. Spyridon’s the day had grown by one grain (spyri in Greek) since the solstice had drifted to December 10.

    In Greece we are quite aware of the relation between the Julian and Gregorian calendar because a group of quacks known as the Old Calendarists have made a religious issue out of an astronomical issue: When Jesus was born the Julian calendar was in effect, not the Gregorian calendar. The Gregorian calendar was created by the pope Gregory and since the pope is 666 it cannot but be evil like all things that come out of the Vatican. So their reasoning goes…

  2. It has nothing to do with the Gregorian calendar, because that calendar did not exist. The bruma, i.e. the festival of the winter solstice was always celebrated on VIII Kal. Ian. (Servius A. 7.720). The eighth day before the Kalends of January was always December 25th in the Julian calendar, because December always had 31 days.

  3. Roger, I don’t know but thank you so much for asking. The most puzzling thing to me is that I’d assume JC got things reset with some degree of accurately, which means the solstice should have been pretty close to the solstice for the first few centuries, at least.

    Whatever explains it all, I’d like to know if astronomers have calculated where the Julian Calendar *should have begun* (in relation to both moons and solstices) if we count backwards from Gregory to Jan. 1, 46 BC.

    Alas, I’ve not yet made friends with any professional astronomers…

  4. The Gregorian calendar was made in the 16th century in order to fix the Julian calendar that has gone out of alignment. Winter solstice was always intended to be December 23, when the day drifted due to the small inacuracy of the Julian calendar, the Gregorian calendar was made in order to fix the Julian calendar. Old Calendarists though have made a religious out of an astronomical issue. Intelligent designers are in good company…

  5. ikokki, show us the historical source that says that the Romans intended the winter solstice to be on 23 December in their Julian calendar. Frankly, I don’t know any. However, we know that it’s a historical fact that the winter solstice was celebrated on VIII Kal. Ian., which always and only corresponded to 25 December in the Julian calendar. Now your job, ikokki, is to show us that the Romans actually intended this to be two days earlier, preferrably with an explanation, why they didn’t follow their intentions and chose 25 December instead.

  6. Hans, just to be clear, are you suggesting the solstice itself occurred on the 25th, or merely that there was a distinction (not necessarily like we have today) between the event and it’s official observance? In other words, did Julius Caesar *set* the solstice to the 25th, or was the solstice on the 23rd and its observation differed for some other reason?

    The related application question in my mind is whether this affects the placement of moons on the Julian Calendar. For instance, was there really an eclipse on “March 13th, 4 BC”, or did someone miscalculate that, also, by two days?

    A further issue like Jewish calendar observance would involve many issues beyond that. For starters, I’m just wondering if we can even date full moons on the historical Julian Calendar with absolute certainty.

  7. Thanks everyone for the comments. I think the Gregorian issue is a red herring; we’re interested in how things were when Caesar set them up.

    The bit of Servius’ commentary on Vergil is:

    [720] vel cum sole novo prima aestatis parte: nam proprie sol novus est VIII. Kal. ian.; sed tunc non sunt aristae, quas ab ariditate dictas esse constat.

    (Maurus Servius Honoratus. In Vergilii carmina comentarii. Servii Grammatici qui feruntur in Vergilii carmina commentarii; recensuerunt Georgius Thilo et Hermannus Hagen. Georgius Thilo. Leipzig. B. G. Teubner. 1881.)

    Having a stab at a translation:

    Or when the new sun in the first part of the year; for properly the new sun is the 8th day before the kalends of January; but at that time there are no harvests, which ab ariditate dictas esse constat.

    Anyone care to finish it? I *must* go to bed! But I’m not yet clear that bruma is 25 Dec. / VIII kal. Jan. – I need to see some data. Julian the Apostate says something in his Hymn to King Helios about the Heliaea NOT being the solstice, but when the farmers could see the days begin to get longer.

  8. It’s about the festival, which always had a set date, maybe even since the times of Romulus, if we want to believe the Suda. You don’t mess with religious festival dates. So the astronomical phenomenon of the winter solstice is not relevant for arguments concerning festivals and the cultural transmission of their dates, as it is e.g. argued in the case of Christmas. Calendar asynchronicities don’t change the dates of the Roman festivals, including the Brumalia, which was always VIII Kal. Ian., as Servius still knew. And that was always 25 December in the Julian calendar, even if there were one or two intercalary days in one year. Only the festival itself is important, and we know from the sources that the Romans were weary of their calendar, or more specifically: of the priests who had been bribed by politicians to extend the calendar etc. Spring festivals were suddenly celebrated in winter, because the calendar was off. But no, they did not move the festival. They always only changed the calendar!

    Julius Caesar didn’t “set” the solstice to 25 December. As pontifex maximus he had to ensure that the prefixed festival date (VIII Kal. Ian.) remained untouched and that it was roughly corresponding to the actual solstice. Hypothetically speaking, if we do include (i.e. back-extrapolate) the Gregorian calendar, Caesar’s original VIII Kal. Ian. (= 25 December Ur-Julian) actually corresponds to our 23 December (Gregorian). That’s the two-day shift that (I think) everyone agrees on. Pretty close, don’t you think?

    Something like an “eclipse on March 13th, 4 BC” will not be in any of the sources. It goes in the other direction. The sources always use the ancient dating, e.g. with Kalends, Ides etc. plus the year of someone’s consulship, seldomly AUC. But they are valuable information nonetheless, because we can clearly compute solar eclipses with astronomical software, and then we have a specifically fixed point, a historical node for calendar references.

    Jewish calendar observance is irrelevant, because the Jews adhered to the regional calendars used in the country of their residence, i.e. for Rome: Nisan = March etc.

  9. Damn. Insomnia.

    Hans… all this is interesting but much too vague. There are simply too many claims here. I really need to see the primary sources that say these things. This is the problem; all we have are assertions, and I want to see the data. At the moment I don’t know of a source that says “brumalia/brume=25 Dec”. Do you?

    How do you get from Julian 25 = Gregorian 23? I don’t follow your thought here – sorry – and I’d like to.

    What is the line of Vergil that Servius is commenting on? He isn’t using the word “bruma” or “brumalia”, but “novus sol”. We need to break this down precisely, I think. The Romans do use the word “solstitium”.

    Do we have evidence of a festival on 25 Dec.?

    What we need is the ancient data. Once we have this, we can work out what is what. At the moment it seems very confusing. If you have this, do let us have it!

  10. Here’s the Vergil:

    710 una ingens Amiterna cohors priscique Quirites,
    Ereti manus omnis oliviferaeque Mutuscae;
    qui Nomentum urbem, qui Rosea rura Velini,
    qui Tetricae horrentis rupes montemque Severum
    Casperiamque colunt Forulosque et flumen HimellaeL,
    qui Tiberim Fabarimque bibunt, quos frigida misit
    Nursia, et Ortinae classes populique Latini,
    quosque secans infaustum interluit Allia nomen:
    quam multi Libyco volvuntur marmore fluctus
    saevus ubi Orion hibernis conditur undis,
    [720] vel cum sole novo densae torrent ut aristae
    aut Hermi campo aut Lyciae flaventibus arvis.
    scuta sonant pulsuque pedum conterrita tellus.

    Cohort of Amiternum and old-world
    Ouirites, all Eretum’s folk, and they
    Of olive-rife Mutusca; men who throng
    Nomentum’s city and the Rosean plain
    Hard by Velinus : who the rugged rocks
    Of Tetrica, and Mount Severus hold,
    Casperia’s town, and Foruli, and where flows
    Himella’s stream; who drink of Tiber’s wave
    And Fabaris, or whom chilly Nursia sends,
    With Horta’s squadrons and the Latian tribes,
    And they whom Allia, evil-boding name,
    Parts with dividing current : numberless
    As billows that on Libya’s sea-floor roll,
    When fierce Orion sets in wintry wave,
    Or thick as clustering corn-ears, that beneath
    [720] The young sun ripen
    , or on Hermus’ plain,
    Or Lycia’s yellowing fields. Their bucklers clang,
    And the earth quakes for terror as they tread.

    Corn ripening under the young sun?

  11. Hans, thanks for that extensive clarification. Please understand my asterisks around *set* and my scare quotes around “March 13th, 4 BC” were because I do understand the problem with what I was asking. I just didn’t know how to phrase it more succinctly.

    Roger, the two day difference is the standard gap (for first century dates I’ve looked at, at least) between a reconstructed Julian date and an artificially retrojected Gregorian date (used by computer programs and such).

    Hans again, what I’m trying to ask is this. How do we know the two day gap shouldn’t be a three day gap? Or to come back at this from my original angle, how did Julius Caesar and his associates decide the precise number of extra days to allow in that one extra long year (the one we call, if memory serves, 47 BC)?

    I would assume JC picked a day for January 1st that would allow key parts of the officialized planting and festival dates to be alligned with the equinox and solstices.

    For instance, when JC was riding out that one extra long year, and the bruma happened far in advance of the winter solstice, and then they waited until X number of days after the actual solstice, and then they proclaimed day X+1 to be “January 1st”… how did he decide what X should equal?

    All of that, as before, so to speak. 🙂

  12. At the moment I don’t know of a source that says “brumalia/brume=25 Dec”. Do you?

    Servius. VIII Kal. Ian., eight days before the first day of January, counting inclusively, of course. Converted to the Julian month this corresponds to 25 December. If you’re looking for a direct historical quote, where the author actually wrote “25th of December” in Latin (or Greek), we first need to find out, when the Romans stopped using the old system in favor of the 1-31 system, because that’s our terminus post quem.

    How do you get from Julian 25 = Gregorian 23? I don’t follow your thought here – sorry – and I’d like to.

    You don’t need to. I said it was hypothetical. The Gregorian calendar did not exist at the time. But if you use calendar converters that have the ability to back-calculate to the decades BC, then you have a two-day difference between the Julian date and the (hypothetical!) Gregorian date. This two-day shift is (as far as I remember) common opinion. I think Mommsen and many others came to that conclusion that Caesar’s calendar was off by two days when comparing it to a back-calculated Gregorian calendar.

    He isn’t using the word “bruma” or “brumalia”, but “novus sol”. We need to break this down precisely, I think. The Romans do use the word “solstitium”.

    The shortest day is also the day that the sun starts to live again: novus sol. On the word solstitium: it refers to the summer solstice (or the summer in general). See my comment on the other article.

    On the corn ripening under the young sun: Servius is not commenting on the corn, but on the “new sun” by simply stating that the new son begins at VIII Kal. Ian.

  13. In other words, it seems JC had the opportunity to pre-determine that the bruma of the first newly regulated year (“46 BC”) would fall on the equinox precisely.

    If JC took that opportunity, did he arrange for the actual solstice to fall on the day of the bruma? Or if JC arranged for the solstice to fall two days before the bruma, why did he do so?

    In the first case, I’m beginning to fear that something has gone very wrong in the standard accounting of all this. In the second case, why? Did he miscalculate? Or did he first allign the spring festivals with the equinox and then found the traditional date had already been miscalculated?

    I’m betting a dollar on that last option, now. Otherwise, what am I (or what are we) missing?

  14. I would assume JC picked a day for January 1st that would allow key parts of the officialized planting and festival dates to be alligned with the equinox and solstices.

    That’s highly probable. But it was not an all-out solar calendar. The moon also played a role, but I haven’t yet delved into the topic.

    For instance, when JC was riding out that one extra long year, and the bruma happened far in advance of the winter solstice, and then they waited until X number of days after the actual solstice, and then they proclaimed day X+1 to be “January 1st”… how did he decide what X should equal?

    The intercalary month was in autumn, I think. So the brumalia before the first Julian Kalends of January were already correctly aligned with the astronomical phenomenon—more or less. On Caesar’s decisions: He was a smart man, a genius, and he had lots of intelligent people from Egypt helping him with this thing. 😉

  15. If JC took that opportunity, did he arrange for the actual solstice to fall on the day of the bruma? Or if JC arranged for the solstice to fall two days before the bruma, why did he do so?

    The old calendar was a good approximation, so when Caesar introduced his calendar, he knew that the festivals would be more or less at the right astronomical and/or meteorological times, and by fine-tuning he ensured that the calendar would not be out of sync to quickly.

    But to answer which festival or which natural phenomenon was important for him, i.e. the pivot of his calendar, would be a theological question. It could have been the sun, spring equinox, a moon (full or new?), planetary alignment (e.g. Venus, his divine mother) etc. I can’t answer this at the moment.

  16. I suppose reckoning the first Julian calendar really does boil down to counting backwards from Gregory. This calendar for the year 1582 on timeanddate.com (one of the coolest calendars online) shows that Thursday October 4th was followed by Friday October 15th. Unfortunately, timeanddate.com does not offer years in BC.

    Another site, called abdicate.net agrees with you that Gregorian Dec.23 is Julian Dec.25. I’ve also compared it with http://www.hebcal.com in years gone by, and it seems to be accurate, if a bit awkward to use.

    Since the retroactive Gregorian would be more accurate in pointing out the date of the astronomical solstice, it seems Julius Caesar DID make sure that the first realigned brumalia DID allign properly with the actual solstice. My own question should then be, why did Gregory “set” the solstice to fall on the 23rd? At that point, however, I’m willing to let be for now. 🙂

    That may all be precisely what you were getting at earlier, Hans. Have I got my brain straight, finally? 😉

  17. Yes, that’s what I was getting at. However, by setting the brumalia this way (VIII Kal. Ian. [Roman] = 25 Dec [Ur-Julian] = 23 Dec [retrojected Gregorian]), Caesar was still off by roughly one day, because astronomically the winter solstice is between 20 and 22 December (Gregorian). So we have several possibilities: (a) Caesar’s calendar was an approximation, because he wanted to fit all festivals under one roof, so to speak, which would be a bit illogical in a theological context, because you need a definite beginning; (b) Caesar’s calendar was right on target because with the brumalia he wasn’t going for the solstice itself (i.e. the shortest day) but for the first day of the sol novus, which would be 25 Dec (Ur-Julian) = 23 Dec (retroj. Gregorian); (c) Caesar’s calendar was on target for another reason, and we simply don’t have the source to support the additional one-day difference.

    [NOTE: If I remember correctly, there is a minority theory on the web, which tries to align the Roman sources with the Egyptian calendar etc., and there the original Caesarian Julian calendar was off by one day, when compared to the Julian calendar after the Augustan reform. According to this theory the first day of Caesar’s first reformed year (i.e. the Kalends of January) corresponded to what was later 31 December. This is highly speculative of course, and it goes against the common opinion, which holds that Caesar’s calendar (compared to our Gregorian calendar) was shifted by two days, and not three. We do know that before the Augustan reform there were years, when the calendar was off by another one or two days until true Julian synchronism was established around 6 AD. But to assert that Caesar started his first reformed year one day early is an extraordinary theory that needs extraordinary proof—and a sound reason, which in the context of the calendar was surely religious. And I’m not sure if the brumalia were reason enough for Caesar to begin the year one day early.]

  18. Hans, thank you so much. This is wonderfully helpful.

    Can you offer a source for the 6 AD comment? After that, I’m sure I’ll be perfectly satisfied! 😉

  19. If I remember correctly, AD 6 was the first year with two corresponding calendar sources, i.e. alignment, meaning e.g. 1 September = 1 September, and not 1 September = 31 August or so. We know that the Augustan reform of Caesar’s calendar took place in February 8 BC and lasted until AD 4. So it’s also possible that the synchronism was established in February AD 4. But afaik the first corroborating evidence for this is from AD 6. This is also roughly the time when the 7-day was beginning to catch on.

  20. At the moment I don’t know of a source that says “brumalia/brume=25 Dec”. Do you?

    Servius. VIII Kal. Ian., eight days before the first day of January, counting inclusively, of course.

    Hans, Servius does not say this. You’re skipping way too fast over the text to the conclusion. He quotes Vergil saying “new/young sun” (which is ripening crops, so hardly the solstice) and says that this is the first part of the year. He does not talk about either brumalia or bruma. Neither word appears, as far as I can see. (I’m not querying that VIII kal. Jan = 25 Dec.; I think we can take that as read).

  21. The shortest day is also the day that the sun starts to live again: novus sol.

    Hans, I am getting rather frustrated by your comments. They mostly consist of bare statements by yourself, without reference. You may be well informed; or not – I can’t tell! But I want to see the data. If you want to say “novus sol = ” we need to see the sources that say so.

    This seems to apply to all the material posted here.

    This perhaps belongs to the other thread, I agree. The question here is whether the solstice on the Julian calendar fell on Dec. 25. I’m not clear that we’re further forward? Anyone?

  22. Since my latin is non-existant and all I know about Ceasar’s calendar reform is what Plutarch says in his biography (very little) I have been attacking this issue form another front: When the 1st Ecumenical Council Easter as the first full moon after the spring equinox it assumed that the Spring equinox is March 21st, not March 23th as it would be if the solstice was on December 25th. Now I have been trying to find primary sources on the decision of the calculation of Easter that would say that spring equinox is March 21, it seems that they are Eusebius of Caesaria, Ecclesiastical History PG 20,1077 and Socrates the Church Historian, but I do not seem to be able to find them online.

  23. Roger, there’s no need to become frustrated. We have a source that clearly says that the bruma was the novus sol: Censorinus, De Die Natali 21.13. Cf. Ov. Fast. 1.163. Therefore Servius’ comment on the “new sun” is about the brumalia, therefore it is on VIII Kal. Ian., which is the 8th day before the first day of January, which equals 25 December in the Julian calendar, which was still in effect in the time of Servius. Therefore: bruma = 25 December. Q.E.D.

    Now, what about the crops? The French translation of Censorinus is clear that the bruma is about the rise of the new sun, which matches what I wrote earlier. The “new sun” begins (rises) beginning with the winter solstice, and remains the new sun until the summer solstice, after which it becomes a dying sun. So that would be an explanation of the crops ripening under the new sun. In any case, it’s not necessary for our argument, because with Servius, Ovid and Censorinus alone we can establish that the bruma was the novus solus. We don’t need Virgil for that. For all we know, Virgil could have meant something different, and Servius completely misunderstood his intention there. But that wouldn’t change anything: Servius’ comment remains valid and important.

    This thing is clear as a whistle. We should debate other things, look for sources on the rituals etc., maybe also find some stuff on the whole “bruma-advent” from the end of November, and we should try to distinguish betwen the original bruma and what became of it in the Byzantine Empire.

  24. And by the way: We should also keep in mind that there are lots of winter cereals, sown in fall, germinating before the first frost, which then ripe immediately under the new young sun.

  25. Now, I’ve also just found a source that directly links the bruma with VIII Kal. Ian.: Plin. NH 18.221. From Bostock’s English translation: “The winter solstice [bruma] begins at the eighth degree of Capricorn, the eighth day before the calends of January, in general”. (They even have a footnote: “Twenty-fifth of December”.)

  26. This is a fascinating discussion (sorry I am coming to it a few weeks late!) Thanks for the link of how Brumalia occurred on Dec. 25 (8 days, inclusively, before Jan. 1).

    Regarding early Julian calendation, check out some research we’ve done here: http://www.4angelspublications.com/Books/HoL/History_of_a_Lie.pdf When the Julian calendar transitioned from an eight-day week to a seven-day week (in honor of the sun cult, Mithraism, from Persia) Saturday was the first day of the week! Saturday Sabbatarians who try and link Saturday to the Bible Sabbath thus make a huge, and erroneous, assumption. Saturday did not exist in the early Julian calendar, whose days of the week were designated: A, B, C, D, E, F, G and H.

  27. I have a working theory that the solstices and equinoxes occurred a month earlier back then. If you notice, there are Roman holidays on Nov, Feb, May and Aug 24th, but none on Dec, Mar, Jun and Sep. Brumalia was a thirty day holiday that ran from Nov 24 to Dec 25. Couldn’t the last day of the holiday represent the coldest day of the year, not the shortest? What sense would it make to begin the holiday thirty days before the solstice? I’d appreciate any thoughts – also I don’t understand Latin.

  28. I’m just curious, especially to Hans, were all of these statements about the Bruma being the 8th calends of January, in reference to the Julian Calendar? You sure none of them were in reference to the old Roman Calendar? Because if that was ever the case, then wouldn’t it be December 24th as well, since the Roman Calendar had only 30 days in December?

  29. Another question, wouldn’t eight days BEFORE the calends of January, be the 24th of December?
    25th/1
    26th/2
    27th/3
    28th/4
    29th/5
    30th/6
    31st/7

    Only seven days BEFORE. So I can only guess inclusive reckoning was used?

  30. The references are mainly imperial, so yes, would involve the Julian calendar. I have read that Roman counting was inclusive — never looked into the evidence.

  31. Sir Ragamuffin, the Roman counting was inclusive, and (as Roger said), the references are mainly imperial. The eigth day before the Kalends of January is 25 December (Julian). This was not real date of the solstice—that was 23 December at the time when the Julian calendar was introduced (with December having 31 days). The 25th was the fixed festive date, which stayed fixed, even when the solstice slowly moved to 21 December due to the slight errors of the Julian calendar. I think there’s a later source that says that the date of the real solstice was known, but the festival was nevertheless on 25 December. I think they explained it that it had something to do with observing for the first time that the days grew longer again.

    A hypothesis: This festive date was originally on the day of the real solstice, i.e. the 23rd day of December, when December still had 29 days. But the festival calendar of course didn’t say “23”, but VIII Kal Ian, so when the month was expanded to 31 days, the festive calendar was not touched, which in turn meant that astronomically the festive day moved forward by two days, away from the real solstice.

  32. Ah, okay, thank you guys for your replies.

    And @Hans, I saw on another of Roger’s threads that Antiochus’s calendar said something similar what you’re saying, that the solstice was actually on the 22nd while the festivalof the sun was still on December 25th.

  33. Yes, on the 23rd or 22nd. But I fear that my hypothesis is incorrect. Caesar added more days to some months (including December), but we know that he added them at the end of the months, so that the feriae would not move, counting from the Kalends of every respective month. Here – http://penelope.uchicago.edu/~grout/encyclopaedia_romana/calendar/antiates.html – we have the Fasti Antiates Maiores, and we see that VIII Kal Ian reports the Larentalia, which were later noted as X Kal Ian. Idem for the Divalia: X Kal Ian pre-Julian, but XII Kal Ian Julian. The important computation was apparently the one starting with the Kalends toward the end of the month, not the other way around. We have some more evidence by Macrobius, Sat. 1.10.2 – http://penelope.uchicago.edu/Thayer/L/Roman/Texts/Macrobius/Saturnalia/1*.html#10.2 : Apud maiores nostros Saturnalia die uno finiebantur, qui erat a. d. quartum decimum Kalendas Ianuarias: sed postquam C. Caesar huic mensi duos addixit dies, sexto decimo coepta celebrari. But there were those who celebrated according to the old system, some according to the new: Ea re factum est ut, cum vulgus ignoraret certum Saturnaliorum diem, nonnullique a C. Caesare inserto die et alii vetere more celebrarent, plures dies Saturnalia numerarentur. But this is beside the point because (a) it was not official, and (b) it didn’t mean that the Saturnalia were moved, but only that they were extended for some people.

  34. Very late to this discussion, but I found it due to a question someone asked on an astronomy group on FB. “Why did December 25th” become a holiday? I find Hans Damp’s input fascinating. I did not know about the Roman celebration of the Winter Solstice. I do know that many astronomical calculations (positions of stars etc.) do a calculation to find the the Julian date and time of an event.

Leave a Reply