I understand that the Julian Calendar was introduced to align the length of the calendar to the tropical year, i.e. make the average year 365.25 days long. However, what I'm still trying to figure out is that were they trying to align the start of the calendar with? Or, put it another way, why 1 January is not aligned to a solstice, equinox or some other astronomical event?

Wikipedia currently states (emphasis mine):

The first step of the [Julian] reform was to realign the start of the calendar year (1 January) to the tropical year by making 46 BC (708 AUC) 445 days long, compensating for the intercalations which had been missed during Caesar's pontificate. This year had already been extended from 355 to 378 days by the insertion of a regular intercalary month in February. When Caesar decreed the reform, probably shortly after his return from the African campaign in late Quintilis (July), he added 67 more days by inserting two extraordinary intercalary months between November and December.

Realign the start of the year to what of the tropical year? A footnote further explains:

It is not known why he decided that 67 was the correct number of days to add. Ideler suggested (Handbuch der mathematischen und technischen Chronologie II 123-125) that he intended to align the winter solstice to a traditional date of 25 December. The number may compensate for three omitted intercalary months (67 = 22+23+22). It also made the distance from 1 March 46 BC, the original New Years Day in the Roman calendar, to 1 January 45 BC 365 days.

But why was the date 25 December special or traditional at that point in time?


10 Answers 10


After some reading about the early Roman Calendar, it is relevant to note that originally the calendar had only ten months and began on March, with an uncounted “winter” period after December. The number of days on each month were more or less flexible, and they usually tried to align the 15th of March, the mid of the month, with Ides, a full moon.

At the time of the ruler Numa Pompilius the months of January and February were introduced at the end of the year, but apparently the calendar became a mess while they tried to keep it both in sync with the moon and the seasons. Julius Caesar then fixed the lengths of the months and the year in order to more closely match it to the length of the tropical year, thus keeping it synchronised with the seasons, but at the expense of loosing synchronisation with the phases of the moon. He also wanted to move the “start” of the year to January, but this didn't last for long and the calendar was still considered as beginning on March.

It seems that, when introducing the new Julian Calendar, the first 15th of March was aligned to fall on a full moon (such alignment would, of course, be lost on the following years) and so that March would contain the Spring Equinox (which, due to the alignment with the tropical year, would remain true thereafter). As a side effect, all of this caused the “1st of January” to fall at some arbitrary point along Earth's orbit around the Sun.

Furthermore, even after the Julian Calendar was set and fixed, the “start of the year” has been celebrated at many different dates including: the 1st of May, 15th of March, 1st of January, 25th of December (which by then was significant because on the fixed calendar it aligned with the winter solstice and the christian Nativity), 25th of March (Annunciation), Easter, 1st of September, 1st of March, and others. It's only from the 16th and 17th century that most countries settled on celebrating (and legally establishing) the start of the year on the 1st of January.

  • 1
    "My guess is that, even if they tried to keep the first few 15th's of March still on sync with the full moon" Nope, they were fully aware of that going to a solar calendar they would loose the sync to the full moon. Perhaps they were syncing the first year to a full moon, though. All of the rest is good. Feb 5 '14 at 19:10
  • Sure I agree about the phases of the moon, and they knew that sync with the moon was going to be lost. My “guess” is about the fact that maybe they tried to start the new calendar with the 15th of March on a full moon and, thus, 1st of January was as a consequence fixed to some arbitrary point. / This is probably the crux of my question, to which I haven't yet found a satisfactory answer: how was the “correct” start of the year determined? Feb 5 '14 at 20:19
  • 1
    Yes, that's what I thought to, and so it is. 46 BC was the long year, so 45 BC (or year -44) was the first year of the new calendar, and indeed, the full moon is on the Ides of March. eclipse.gsfc.nasa.gov/phase/phases-0099.html Feb 5 '14 at 20:45
  • 1
    And the Spring Equinox the 23rd of March, it seems. So it seems they aligned the first year to the first day of the moon in which the Spring Equinox would happen. I'm actually quite surprised that it's the 23rd. I would have expected it to be the 25th, so the Winter solstice would have been the 25th as well. Feb 5 '14 at 21:23
  • Nice, great! Thanks a lot for the additional info. I've edited the answer accordingly. Feb 6 '14 at 8:55

The epoch of the Julian calendar (i.e., January 1st, 45 BC) was indeed synchronized with the first new moon following the previous year's winter solstice :

It was probably the original intention of Caesar to commence the year with the shortest day. The winter solstice at Rome, in the year 46 B.C., occurred on the 24th of December of the Julian calendar. His motive for delaying the commencement for seven days longer, instead of taking the following day, was probably the desire to gratify the superstition of the Romans, by causing the first year of the reformed calendar to fall on the day of the new moon. Accordingly, it is found that the mean new moon occurred at Rome on the 1st of January, 45 B.C., at 6h. 16′ P. M. In this way alone can be explained the phrase used by Macrobius: Annum civilem Caesar, habitis ad lunam dimensionibus constitutum, edicto palam proposito publicavit. This edict is also mentioned by Plutarch where he gives the anecdote of Cicero, who, on being told by some one that the constellation Lyra would rise the next morning, observed, Yes, no doubt, in obedience to the edict.

William Smith, A Dictionary of Greek and Roman Antiquities, page 231, John Murray, London, 1875.

  • I have noticed that some of the other answers mention March 25th as supposedly signifying the initial date of the spring equinox, when, in fact, it can be easily confirmed, using either these two resources together, or this one alone, that the date in question was March 22nd/23rd.
    – Lucian
    Sep 7 '18 at 17:55

Possibly they wanted to match it to Brumalia. The Roman winter solstice festival. wikipedia: "The Brumalia was also celebrated during the space of thirty days, commencing on 24 November and ending with the "Waxing of the Light", December 25" citation

Much the same can be said about Saturnalia, they're very similar. The "Birthday of the Unconquered Sun" is another Decemeber 25th shortest day thing, but it's usage seems to be late roman empire.

The tropical year is the solar year. It's a measurement of the position of the sun. It was probably based on Hipparchus's work on equinoxes.

Also the choice of this date is to do with Pliny: http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.02.0137%3Abook%3D18%3Achapter%3D59

  • 1
    Yes, but the question is why they aligned these astronomical events to the (seemingly) arbitrary date of December 25th? Jan 3 '13 at 12:28
  • I see how that's a bit different. Jan 3 '13 at 12:36
  • @JuanA.Navarro - I think this may be an answer to your question, it is something I have heard many people say, I am not sure just how accurate it is though. Solstice literally means Sun stands still, this happens on Dec 22nd. And for 2 days after it does appear that the sun starts in nearly the same position in the sky. On the 25th it is the first day in which visually with the naked eye the sun begins to appear to rise in its starting position in the sky.
    – ed.hank
    Nov 15 '19 at 18:57

According to Feeny, "Caesar's Calendar", p.196 (preview here), the concept of aligning with the celestial objects was not even present to Roman minds in Cicero's time.

It would be interesting to know precisely where in Cicero and Ovid there are references to alignment.


Scaliger 1583 indicates that 45bc is not to be treated as a leap year. By doing so, the start of the year aligns with the dark moon conjunction.

Ideler 1825 Mommsen 1859 both indicate that 45bc should be treated as a leap year. this would mean that the start of the year falls one day before the dark moon conjunction.

However, it should be of note that the Canopic calendar year that straddles the start of 45bc is a leap year. The new canopic year that starts on Oct 22 is aligned to a dark moon conjunction, and is also the start of new 28 year solar cycle based on the use of the 7 day planetary cycle in which saturday is the first day of the 7 day cycle.

This Canopic calendar is the only 365.25 day calendar that is known to have been in operation at the time, which makes it a strong contender as being the basis for the calendar reforms of the roman calendar. Sosigenes was based in Alexandria of Egypt, in which there was at the time, one of the biggest libraries available. It was also in Alexandria, that the Canopic calendar originated.

  • 5
    This would be improved if you named the works you're referring to (i.e. "Scaliger 1583", "Ideler 1825", " Mommsen 1859")
    – Steve Bird
    Apr 28 '18 at 7:20

There are several things that have not been mentioned in the other answers. The Roman calendar in 45 BC can shown to be aligned with the start of the Roman nundinal cycle.

There is also good reason to believe that the Roman calendar alignment was closely connected with the canopic calendar. The leap year cycle of the newly aligned Roman calendar, for example was aligned with the leap year cycle of the canopic cycle.

A wrong understanding of the 4 year leap year cycle, however, resulted in the leap year cycle being 3 yearly until the reign of Augustus. He did some more refinements that put the Roman calendar back onto its originally intended alignment.

  • 5
    Sources to support your assertions would improve this answer. Sep 15 '17 at 9:17

Caesar's goal with the Julian calendar may have been to correct the errors which had occurred in recent decades and restore the important holidays and other important days of the year to their rightful times - their rightful times as he and other older Romans remembered them, and not to when they might have happened in the time of Romulus centuries earlier.

And also to change the calendar enough to minimize the actions necessary to keep it correct and so make sure there would be no more calendar drift in the future. Recent events had shown that the administration could not be trusted to make the necessary actions at the necessary times to keep the calendar in line with the seasons, so minimizing those necessary actions and making them strictly periodic and automatic seemed like the proper course of action.

This question asks about the day that the consuls began their term of office:

When did the Roman consular year begin during the Republic and Empire?1

My answer lists the various dates for the start of Consuls' terms, and thus for the star of the Roman consular year. Obviously the data for the early Roman Republic is rather suspect, but it seems clear that the 153 BC date for starting the Roman consular year on January 1 should be solid.

So by 45 BC starting the Roman consular year on January 1 was a tradition over a century old and Caesar would have needed a strong reason to change it. And Caesar obviously had to keep January 1 in his new calendar reasonably close to January 1 in the old Republican calendar, thus somewhat restricting the possible date range for January 1 in the Julian Calendar.

This factor, combined with the factors mentioned in other answers, might be enough to account for the Julian calendar placing January 1 when it did.


In your quote mentioning Ideler you come close to answering your own question. The only mistake is that Ideler's guess that the year was aligned to the Winter solstice. In fact, Caesar aligned the calendar to the Spring equinox and we know this from offhand contemporaneous mentions by Cicero and Ovid. In other words, the date VIII Kalends Aprilis (25th of March) was made to fall on the Spring equinox.

The beginning of the civil year in Rome was always the Kalends of January. You may hear of March as the beginning of the year, but that was the sacred year. Caesar only published new dates for use in the civil calendar.

  • 1
    Thanks for your answer! It makes a lot of sense. Just a few questions: on this source 25th of March is stated as a.d.VIII.Kal.Apr. but you have it as VI, which one is correct? And, again the big question, why this date in particular? What was special about it at the time? Also would be glad if you could provide some sources or references! May 1 '14 at 7:30
  • It is 8, I miscounted. I have updated the answer. May 1 '14 at 14:16
  • @JuanA.Navarro The Romans celebrated the vernal equinox always on the same day, the festival of Hilaria on March 25th (our word "hilarious" comes from this joyful holiday). Why this date happened to be associated with the equinox is probably accidental. In other words, when the calendar was first used the equinox happened to fall on this day and thereafter it became traditional. It was a very important civil holiday and was the day of marriages so it had a high social and economic importance in Rome. Everybody, including slaves, got the day off. May 1 '14 at 14:34
  • 1
    Thanks again. This seems to point in the right direction, and is becoming my preferred theory thus far. However, as Roger Pearse notes in his answer, could you point to a source where Cicero or Ovid make references to this alignment (either to Hilaria or possibly to Brumalia). May 2 '14 at 8:23
  • 1
    That's the point. How do we know your answer is the correct one?
    – timur
    Dec 31 '17 at 18:49

Sol Invictus was said to have died and gone to hell, and three days later rose from the grave on that day, celebrated in the Roman festival of Dies Natalis Solis Invicti, or "Birthday of the Unconquered Sun". Both the nox maximus, or Winter Solstice, and Dies Natalis Solis Invicti were said to have occured on December 25th in the Julian calender.

For more information than you asked for about later changes to the celebration, this page presents a comprehensive study of how Christmas relates to Dies Natalis Solis Invicti.

  • 5
    Deleted a series of near snark from multiple parties. I will posit that we like to avoid being a primary reference here. For that reason, a lot of our users will look askance at answers without a lot of links backing up their assertions.
    – T.E.D.
    Feb 10 '14 at 22:02
  • 1
    @LennartRegebro I changed my answer to reflect your contention and my inability to provide good sources. Apr 29 '14 at 15:07
  • Yeah, I suspect you have things backwards here. It's not that the calendar is fixed to this festival, it's this festival that is fixed to the winter solstice. It probably happens the 25th because the calendar is fixed to the spring equinox. But why the spring equinox was fixed to the 25th of March is unclear. Apr 30 '14 at 7:38
  • @LennartRegebro Maybe, but, since the asker is asking WHY the solstice was affixed to Dec 25th in the original question, "he intended to align the winter solstice to a traditional date of 25 December.", and the solstice actually falls on the 21st, your contending that the day is important because the solstice falls on that day doesn't hold up. The celebration pre-dates the time when the solstice was affixed to that day, and it's one of the earliest I can find which does. Apr 30 '14 at 17:41
  • 1
    So to recap: I'm sure that the Julian Calendar was intended to be aligned with the solstices and equinoxes in some way. But I don't know why they are aligned so they end up the 25th of the months. It would make more sense to put the 1st of April on the spring equinox. Apparently the 25th of March is the traditional date, but the question then only becomes "Why the 25th of March?" May 2 '14 at 13:34

i am unable to give references to support my mention above that the possible source to which the roman calendar was being aligned .... was the Canopic calendar .... except to say that I have sat down and attempted to reconstruct how the Canopic calendar would have looked ... starting from back in the time of the Ptolemies.... and projecting it down to roman times.

I have noted the historical names of important people who have looked into the starting date of the reformed calendar who have indicated this first year of the reforms should be a leap year.

I have when looking at the Canopic calendar ... looked at the use of the 7 day planetary week used by the Greeks in association with this new 365.25 day calendar ... which results in a 28 year calendar cycle ...

It is to be noted that in 45 BC .... the start of the Canopic calendar ... in the latter part of 45 BC .... is the first year of such a 28 year cycle ... meaning that the prior year ... which would be the last year of the 28 year cycle ... would be a leap year ... and as such would overlap with the first part of the reformed roman calendar year of 45 BC ....

This would put the leap day of the Canopic calendar year 28 ... to be in the same year as the projected leap year of the first year of the roman reformed year ....

This would then indeed give historical grounds for Ideler 1825 Mommsen 1859 to conclude that the first year of the roman calendar reforms of Julius Caesar to be a leap year.

Another interesting confirmation of this... is that if you take the corrections of Augustus ... which put the Roman reformed calendar back on to its intended cycle ... and project the corrected calendar back wards to 45 BC ... you again find that 45 BC ... should indeed be a leap year ....

  • 2
    You should probably edit this into your previous answer, rather than posting it as a new answer. Aug 10 '19 at 12:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.