Talk:Julian calendar/Archive 1

Archive 1

Archive 2

Archive 3

Archive 4

WikiProject Time assessment rating comment

Want to help write or improve articles about Time? Join WikiProject Time or visit the Time Portal for a list of articles that need improving.
—Yamara ✉ 18:20, 14 February 2008 (UTC)

Date links

In January 2004, 138.88.61.249 apparently removed all links to dates from this page. While the linkage may have been excessive, I could have done with links to 46 BC and 45 BC -- understandable for a page about calendars, no? As such, I've reinstated them, although not at the level they were before. JTN 17:32, 2004 Oct 4 (UTC)

Leap day error

REGARDING the issue of leap day error every 3 years and debating who has the right sequence i would like to see this 1999 publication.

The papyrus is pOxy LXI 4175, announced in A R Jones, ZPE 119 (1997) 157, formally published in A R Jones, Astronomical Papyri from Oxyrhynchus (P.Oxy. 4133-4300a) (Phildelphia, 1999), and analysed calendrically in C J Bennett, ZPE 142 (2003) 221 and ZPE 147 (2004) 165. --Chris Bennett 20:16, 27 October 2007 (UTC)

Meanwhile then please look at this chart i have made and tell me how it would be as 44 BC instead of 42 BC.

Julian Egyptian Reformation error Egyptian 45 BC* Jan 1 Koyak 30 Jan 1 = 709 AUC*Jan 1 Koyak 30 44 BC Jan 1 Tybi 1 710 AUC Jan 1 Tybi 1 43 BC Jan 1 Tybi 1 711 AUC Jan 1 Tybi 1 42 BC Jan 1 Tybi 1 712 AUC*Jan 1 Tybi 1 41 BC* Jan 1 Tybi 1 Jan 2 = 713 AUC Jan 1 Tybi 2 40 BC Jan 1 Tybi 2 714 AUC Jan 1 Tybi 2 39 BC Jan 1 Tybi 2 715 AUC*Jan 1 Tybi 2 38 BC Jan 1 Tybi 2 Jan 2 = 716 AUC Jan 1 Tybi 3 37 BC* Jan 1 Tybi 2 Jan 2 = 717 AUC Jan 1 Tybi 3 36 BC Jan 1 Tybi 3 718 AUC*Jan 1 Tybi 3 35 BC Jan 1 Tybi 3 Jan 2 = 729 AUC Jan 1 Tybi 4 34 BC Jan 1 Tybi 3 Jan 2 = 720 AUC Jan 1 Tybi 4 33 BC* Jan 1 Tybi 3 Jan 2 = 721 AUC*Jan 1 Tybi 4 32 BC Jan 1 Tybi 4 Jan 2 = 722 AUC Jan 1 Tybi 5 31 BC Jan 1 Tybi 4 Jan 2 = 723 AUC Jan 1 Tybi 5 30 BC Jan 1 Tybi 4 Jan 2 = 724 AUC*Jan 1 Tybi 5 29 BC* Jan 1 Tybi 4 Jan 3 = 725 AUC Jan 1 Tybi 6 28 BC Jan 1 Tybi 5 Jan 2 = 726 AUC Jan 1 Tybi 6 27 BC Jan 1 Tybi 5 Jan 2 = 727 AUC* Jan 1 Tybi 6 26 BC Jan 1 Tybi 5 Jan 3 = 728 AUC Jan 1 Tybi 7 25 BC* Jan 1 Tybi 5 Jan 3 = 729 AUC Jan 1 Tybi 7 24 BC Jan 1 Tybi 6 Jan 2 = 730 AUC* Jan 1 Tybi 7 23 BC Jan 1 Tybi 6 Jan 3 = 731 AUC Jan 1 Tybi 8 22 BC Jan 1 Tybi 6 Jan 3 = 732 AUC Jan 1 Tybi 8 21 BC* Jan 1 Tybi 6 Jan 3 = 733 AUC* Jan 1 Tybi 8 20 BC Jan 1 Tybi 7 Jan 3 = 734 AUC Jan 1 Tybi 9 19 BC Jan 1 Tybi 7 Jan 3 = 735 AUC Jan 1 Tybi 9 18 BC Jan 1 Tybi 7 Jan 3 = 736 AUC* Jan 1 Tybi 9 17 BC* Jan 1 Tybi 7 Jan 4 = 737 AUC Jan 1 Tybi 10 16 BC Jan 1 Tybi 8 Jan 3 = 738 AUC Jan 1 Tybi 10 15 BC Jan 1 Tybi 8 Jan 3 = 739 AUC* Jan 1 Tybi 10 14 BC Jan 1 Tybi 8 Jan 4 = 740 AUC Jan 1 Tybi 11 13 BC* Jan 1 Tybi 8 Jan 4 = 741 AUC Jan 1 Tybi 11 12 BC Jan 1 Tybi 9 Jan 3 = 742 AUC* Jan 1 Tybi 11 11 BC Jan 1 Tybi 9 Jan 4 = 743 AUC Jan 1 Tybi 12 10 BC Jan 1 Tybi 9 Jan 4 = 744 AUC Jan 1 Tybi 12

 9 BC*	Jan 1	Tybi  9	Jan 4 =	745 AUC* Jan 1 	Tybi 12
 8 BC	Jan 1	Tybi 10	Jan 4 =	746 AUC	Jan 1 	Tybi 13
 7 BC	Jan 1	Tybi 10	Jan 4 =	747 AUC	Jan 1 	Tybi 13
 6 BC	Jan 1	Tybi 10	Jan 4 =	748 AUC	Jan 1 	Tybi 13
 5 BC*	Jan 1	Tybi 10	Jan 4 =	749 AUC	Jan 1 	Tybi 13
 4 BC	Jan 1	Tybi 11	Jan 3 =	750 AUC	Jan 1 	Tybi 13
 3 BC	Jan 1	Tybi 11	Jan 3 =	751 AUC	Jan 1 	Tybi 13
 2 BC	Jan 1	Tybi 11	Jan 3 =	752 AUC	Jan 1 	Tybi 13
 1 BC*	Jan 1	Tybi 11	Jan 3 =	753 AUC	Jan 1 	Tybi 13
 1 AD	Jan 1	Tybi 12	Jan 2 =	754 AUC	Jan 1 	Tybi 13
 2 AD	Jan 1	Tybi 12	Jan 2 =	755 AUC	Jan 1 	Tybi 13
 3 AD	Jan 1	Tybi 12	Jan 2 =	756 AUC	Jan 1 	Tybi 13
 4 AD*	Jan 1	Tybi 12	Jan 2 =	757 AUC	Jan 1 	Tybi 13
 5 AD	Jan 1	Tybi 13		758 AUC	Jan 1 	Tybi 13
 6 AD	Jan 1	Tybi 13		759 AUC	Jan 1 	Tybi 13
 7 AD	Jan 1	Tybi 13		760 AUC	Jan 1 	Tybi 13
 8 AD*	Jan 1	Tybi 13		761 AUC*Jan 1 	Tybi 13
 9 AD	Jan 1	Tybi 14		762 AUC	Jan 1 	Tybi 14

10 AD Jan 1 Tybi 14 763 AUC Jan 1 Tybi 14 11 AD Jan 1 Tybi 14 764 AUC Jan 1 Tybi 14 12 AD* Jan 1 Tybi 14 765 AUC*Jan 1 Tybi 14 —Preceding unsigned comment added by 24.167.196.43 (talk) 04:34, 21 October 2007 (UTC)

You can see an Excel spreadsheet giving Roman/Julian conversions that includes this model for the three year error at http://www.tyndale.cam.ac.uk/Egypt/ptolemies/chron/roman/roman_civil.xls

--Chris Bennett 20:16, 27 October 2007 (UTC)

Orthodox Use of the Julian Calendar

I revised the portion of the article dealing with the Orthodox Church's continuing use of the Julian Calendar. First, I wanted to distinguish sharply between the adoption of the New Calendar in Orthodox countries by the civil administrations and its (partial) adoption by some national churches. Accordingly, I put the material on Orthodox usage into a separate paragraph. Second, the article, before I changed it, left the impression that most Orthodox Christians adopted the revised Julian Calendar after 1923. In fact, world-wide, the great majority of Orthodox Christians (> 80%?) use nothing but the Old Calendar. I do not wish to debate which usage is correct (this is not the page for such sterile discussions), but I think it is important to be accurate about the numbers. jloukidelis 2004-10-17

Your description is confused and I suspect you misunderstand the 1923 proposal (you apparently regard it as virtually identical to the old Julian calendar whereas it is radically different). In your description you state that Easter and Pentecost are calculated using the "revised Julian calendar" but that the Nativity (the fixed date of 25 December) uses the "new calendar". This implies that you regard the revised Julian calendar and the new calendar as two different calendars. If they are different, then what is the "new" calendar? Is it the Gregorian calendar?

The 1923 proposal had two parts, solar and lunar. The solar part is described in revised Julian calendar. Thirteen days were removed to bring it into agreement with the present Gregorian calendar, but its leap years were different. If the year number is evenly divisible by four it is a leap year, unless it is evenly divisible by 100 then it is not, unless division by 900 leaves a remainder of 200 or 600 then it is a leap year. Because 2000 divided by 900 leaves 200 it is a leap year as it is in the Gregorian calendar. Because 2100 divided by 900 leaves 300 which is neither 200 nor 600 then it is not a leap year in agreement with the Gregorian calendar. Similarly for 2200 and 2300. Because 2400 divided by 900 leaves 600 it is also a leap year just as it is in the Gregorian calendar. But 2800 divided by 900 leaves 100 which is neither 200 nor 600 so it is not a leap year unlike the Gregorian calendar. So the revised Julian calendar will be equal to the Gregorian calendar until 2800 (which is what I said in the description you reverted). Thus for the next few centuries any country or church which adopted the solar part celebrates its fixed dates on the same dates as in the Gregorian calendar, which is what I think you were trying to say. The revised Julian calendar article states that the solar part was adopted by the Greek and Syrian Orthodox Churches (though a few Old Calendar churches in those countries continue to use the Julian calendar), but the Russian and Serbian Orthodox Churches did not, so they celebrate the Nativity on 25 December (Julian) which is 7 January (Gregorian).

The lunar part of the 1923 proposal was quite different and was not adopted by any country or church, so neither Easter nor Pentecost nor any other moveable feast uses the "revised Julian calendar" as you state — all continue to use the old Julian calendar. The lunar portion actually tried to adopt astronomical rules at Jerusalem. It would have made Easter the first Sunday after the full moon that is the first to occur on or after the vernal equinox, where day is midnight-to-midnight at Jerusalem. Here, both full moon and vernal equinox are the astronomical instants. This was so radical that no Orthodox church was willing to accept it, so all continue to celebrate Easter and Pentecost according to the old Julian calendar.

An Orthodox link is On the question of the "Revised Julian Calendar".
— Joe Kress 17:10, Oct 18, 2004 (UTC)

Does anyone fancy working on WikiProject Christian liturgical year? Gareth Hughes 11:05, 21 Feb 2005 (UTC)

"Revised" Julian calendar

Your points are well taken, so by all means change what I have written.

My difficulties, as I stated, were otherwise, and I would have my changes on those points preserved. That is, first, let us draw a sharper distinction between the adoption of the Gregorian Calendar by the civil administrations in various Orthodox countries vs the adoption of the New Calendar by two (actually, I think it was three) Orthodox national churches; and, second, let's be clearer about the relative numbers of Orthodox Christians who follow the New vs the Old Calendar (without being drawn into a debate on the merits of one vs the other).

JL 2004-10-18

January the first month?

Several of the articles on months (e.g., January, March) link here when mentioning the switch of March as the first month to January as the first month, saying that the switch was implemented as part of the Julian calendar reform. But this article doesn't seem to say much about it, just mentioning that Quintilius was renamed to Iulius, etc. Can someone who knows more than I do about this subject add a bit about it? Acheron 06:08, 4 Jan 2005 (UTC)

The January and March articles could use some revision since they have some errors, but the Roman calendar article has many errors and requires some serious attention. The article on the Julian calendar should only have a few comments about its year before it began in 46/45 BC. For now see Julian calendar#Year numbering for a good discussion about the beginning of the year near the beginning of the Julian calendar. In addition, the calendar year or the order of the months in a year should be mentioned, which was January to December quite early, either since Romulus (c. 700 BC) or since the beginning of the Roman Republic a couple of hundred years later (Roman writers disagree). Further revision of the articles will have to await rereading the sources. — Joe Kress 18:48, Jan 4, 2005 (UTC)

Old High German

Wouldn't "Harbistmanoth" be "autumn month", rather than "grazing month" and "Lentzinmanoth" "spring month" rather than "lent month" or which sense would be the most common during that time? (I might be wrong, but IMHO, These translations seem to derive from English false friends).

Merge from Acoreus

There is a merge request sitting in the middle of the article Acoreus that could do with some attention. Anyone here with an interest in the Julian calendar that would be interested in attempting it? --Randolph 21:50, 2 May 2005 (UTC)

Done. — Joe Kress 06:55, May 3, 2005 (UTC)

Superb. :) My organisational sensibilities have been appeased. Thanks. --Randolph 17:59, 3 May 2005 (UTC)

Hmm

Looks like this page was slashdotted: [1] :)

cp

Use of the Julian Calendar

Mount Athos still uses it: there is a place in the United Kingdom which still uses it - anyone know where? Are there any other places where the Julian Calendar is still in use?

Is it an urban legend that a group of pre-1917 Russian athletes arrived too late for the Olympic Games because they were working on the Julian Calendar or did it actually happen?

Summary of Differences (was: Length of months + Leap years + Mens intercalaris)

Seems to me the two topics of leap years and month lengths could be merged. Also, i saw the mention of the Mens Intercalaris (i didnt know that - interesting) and thought maybe something should be mentioned in the leap years correction part as well. Anybody with knowledge thereof interested? --The Minister of War 11:23, 9 September 2005 (UTC)

Merging leap years error and length of the months would be difficult because they are only tangentially related — one day is added to February in leap years. But that is not the subject of length of the months — the length of February in common years is. Intercalaris cannot be mentioned in leap years error because Intercalaris no longer existed after 46 BC. — Joe Kress 20:53, September 11, 2005 (UTC)

You are right. In reflection, my point is somewhat different; i would very much like to see a section (probably From Roman to Julian) which really lists the differences the calendar introduced. Currently, this information is somewhat scattered throughout. --The Minister of War 11:41, 13 September 2005 (UTC)

While reverting your month change, I decided to reword the introductory paragraph of length of the months to partially clarify the differences between the Republican and Julian calendars. The only thing that Julius Caesar did was abolish the intercalary month, replacing it with a quadrennial intercalary day obtained by doubling the sixth day before the calends of March, which is where Intercalaris was originally inserted. In our modern numbering, every four years there would be two February 24s. It may need more clarification. — Joe Kress 14:51, 13 September 2005 (UTC)

Another difference is that the lengths of most months were changed. However, Chris made that correction before I could. — Joe Kress 05:08, 14 September 2005 (UTC)

Hmm, i think i'll stop editing this article now :-) Although i must say that most people know the Julian calendar to run till medieval times, so perhaps it would be appropriate to name them as they were known throughout 8BC-1582AD rather than 46BC-8BC, and make it a sidenote that some changes were introduced somewhat later.--The Minister of War 06:56, 14 September 2005 (UTC)

I understand your frustation. This article has many little known details not commonly found in other sources. One possible solution is to discuss the medieval Julian calendar at the beginning of the article, and then discuss its history, including the many changes during its early years, only later in the article, which may still overwhelm it. — Joe Kress 23:03, 14 September 2005 (UTC)

To the anonymous user who wants to remove the Julian date converter

Since you don't have a talk page of your own I have to reply to you here.

You wrote:

O what do you know about what you are doing?!

More than you do apparently.

Did you even look at the date converter link which you "restored" to the Julian calendar article? Your "restoration" was quite out of place.

Did you? It converts Julian day numbers to Julian dates before 1582 and Gregorian dates thereafter. That's a perfectly reasonable external link for this page. I have used it alot in my own research on the Julian equivalents to dates in the pre-Julian calendar.

And as for my statement about days of the week: it is quite useful when you are trying to draw up a Julian calendar for a particular year, so you can line up the days with the right days of the week!!

I agree it's useful. My initial reaction to this was to leave it alone as a stub for a more comprehensive subsection on the week, since the current article doesn't discuss it, and it reasonably could, although the week isn't just an issue for the Julian calendar. However, I found that there is a perfectly good article on the history of the week which even mentions the Dominical letter. This is where your algorithm belongs. So clearly the right thing to do was to add a link to that article and to remove the stub you had added.

I'm going to put back what I wrote.

And I'm going to remove it again. If you want to add your comment about Thursday to the article on the week, go right ahead. That's the right place for it.--Chris Bennett 02:26, 19 October 2005 (UTC)

I agree that the link to Julian Date Converter is out of place here. Although it does marginally relate to the Julian calendar, that is not the primary focus of the page. The link title is Julian-Gregorian Converter, which is not the actual page title, and is deceptive. It does not convert between the Julian and Gregorian calendars, but between calendar dates and Julian Dates. Anybody attempting to use it for any modern dates will get the wrong value on the Julian calendar. Julian Dates are not dates on the Julian calendar. It is true that the page uses Julian calender for pre-Gregorian dates, but it only marginally relates to the article, is primarily focused on a different topic, there are other pages which actually convert between Julian and Gregorian calendars, and which convert between Julian Dates and the Julian calendar for any year, not just before 1582. There is already a link to the Formilab page, which actually has useful information about the Julian calendar. See also Calendrica, for example. Just because you find the link useful does not mean that it is appropriate for this article.

This link is appropriate for the Julian Day article, and can already be found there. It does not need to be here. --Nike 09:09, 15 October 2006 (UTC)

Thhe link title is immaterial to the relevance of the link itself -- if you thought it was misleading, you could have just changed it. As to the relevance of the link, I don't see how the Fourmilab convertor is any more or less relevant to this page. Why throw out this link but keep the "see also" link to "Julian year (astronomy)"? However, I haven't bothered to look at the Fourmilab convertor before. Now that I have, I think it is rather more useful than the USNO link, because it covers additional calendars, while still providing conversions between the Julian Calendar and JDN. So, while I disagree entirely with your argument, I'm willing to live with your proposal. --Chris Bennett 19:16, 15 October 2006 (UTC)

Trivia removal by Garzo

While Garzo was quite right to remove the Trivia section here, it was still pretty funny. JHCC (talk) 17:17, 8 December 2005 (UTC)

gregorian calender

This article refers to the Gregorian calender, which doesn't have a page. To create a context for the Julian calender it would be helpful to readers if someone created a page for the Gregorian calender... I may if I have time in the next few weeks.

Nudas veritas 00:06, 17 January 2006 (UTC)

The article Gregorian calendar does exist and is linked at its first reference. Just click on the hyperlink to access it. Thanks for the alert — I have created a redirect for your common misspelling. — Joe Kress 06:45, 17 January 2006 (UTC)

question about sentence

Found this article quite interesting and most helpful. However, I had trouble understanding the math in one sentence. "This intercalary month was formed by inserting 22 days before the last five days of February, creating a 27-day month."

To me this sounds as if the original February only had 5 days. That can't be correct, so I must misunderstand the sentence and if I do so misunderstand, many others will, as well. Could you clarify? Thanks! Sporophyte 03:40, 18 March 2006 (UTC)

The original February had 28 days, and it still has 28 days. But in leap years, February was split into two parts. The first part had 23 or 24 days which was still called February, even though truncated. The last five days of February lost that name and were attached at the end of a new 22-day block of days. The 27-day combination was called Intercalaris (aka Mercedonis). When the truncated February had 24 days, its last day was a duplicate of the first of the last five days of Intercalaris. See the calendar wall painting from Roman Calendar, which displays the ordinary 355-day year of the Roman Republic, including a 28-day February, as well as a 27-day Intercalaris displayed after December which was used during leap years (although inserted between a truncated February and March). — Joe Kress 07:25, 18 March 2006 (UTC)

Forgive me for disagreeing with everyone else, but I think this is confused. As I understand it (and this is what Agnes Kirsopp Michels has in "Calendar of the Roman Republic" which is pretty much the authoritative work after Mommsen) February was shortened to 23/24 days and Intercalarius followed. Its quite clear that embolistic years had 377/378 days in them. You can't have February and Intercalarius having 28 and 27 days respectively. You have two options (1) Intercalarius was short (22 days) *or* (2) February was shortened. We know, from Fasti such as the one illustrated that Intercalarius was 27 days long. More tellingly there are Fasti with the Regifugium marked as VI ante Kal Mar, but in Intercalarius, which means the Regifugium must have happened in Intercalarius, not possible if Intercalarius was interspersed into Februarius surely. Would anyone object if I amended the article to reflect this. Francis Davey 15:58, 19 June 2006 (UTC)

You are both saying the same thing, and so does the article, perhaps not as clearly as it could. It's the calendar of the Fasti Antiates that had 28 and 27 days for February and Intercalaris, not the year, for the simple reason that that calendar was painted on a wall and was (presumably) meant to be used for years at a time. The users of the calendar understood that in intercalary years February had to be truncated as Joe describes, before jumping across to Intercalaris. Brind'Amour points out that in what he presumes was the original Decemviral calendar, starting in March rather than January, February would be the last month of the year and Intercalaris would follow it immediately in a painted calendar, which would have made the relationship clearer. --Chris Bennett 21:06, 19 June 2006 (UTC)

No we aren't. The article talks about "inserting 22 days before the last five days of February" -- the impliaction being that 22 new days were insterted and the days that followed formed a part of the monthy of February. That is an idea I have seen in a number of places on USENET and would imply that (a) intercalarius has only 22/23 days and (b) that the last days of the year (before March) would be part of February -- again contemporary evidence says no. It may be this is not what the article means, but in that case it should say that February was shortened and followed by Interc. of 27 days (is that what you think it says?). There's no discussion of the additional day by the way. Francis Davey 20:10, 20 June 2006 (UTC)

You did not quote the complete sentence. It said "inserting 22 days before the last five days of February, creating a 27 day month." I thought it was obvious that the "27 day month" refered to Intercalaris because 22 + 5 = 27. And the following sentence said that February was truncated to 23 or 24 days, so it obviously lost a few days. Since that did not satisfy you, I have explictly stated that the last five days of February became the last five days of Intercalaris. I also mentioned the duplicate day. Does the new wording satisfy your objections? — Joe Kress 09:24, 21 June 2006 (UTC)

Yes. I was concerned because a confused view does exist, so its important to make it clear. Out of interest, what do you mean by the 24th day of February being a duplicate day? I'm not sure I understand that. Post-Julian reform, the intercalary day was a duplicate (hence bissextile) what in practical terms did the last day of February being a duplicate mean? Francis Davey 09:49, 21 June 2006 (UTC)

I assumed that, when present, the 24th day of February would keep its characteristic of a dies nefasti, which I also assumed characterized the 23rd day of Intercalaris, Regifugium, hence a duplicate regarding its fasti characteristic. But that may be making too many assumptions, so I'll let Chris state what the archaealogical evidence supports. — Joe Kress 10:39, 21 June 2006 (UTC)

I've spun it a little further, but I removed the sentence about the duplicate day, since I don't think its necessary to mention this in order to explain how the month was constructed (especially in an article on the Julian calendar!) The issue is a very obscure detail about the operation of the Roman festival calendar. The last day of the 23-day truncated February was the Terminalia, so that the following days, in a March-based year, were considered as being after the end of the year. However, the following day was the festival of the Regifugium. In the Fasti Antiates, the Regifugium is marked twice, once in February and once in Intercalaris. Clearly, in an ordinary year the Regifugium was celebrated on the day after the Terminalia, and in a 377-day year it was celebrated on 23 Intercalaris. In both cases the date was the same: ad VI Kal Mart. What is unclear is when the Regifugium was celebrated in a 378-day year, since both marked days existed in that year, which is why the day is considered to have been duplicated. (A related issue is what was the criterion the Romans originally used to decide that the intercalary year should be 377 or 378 days long.) --Chris Bennett 15:30, 21 June 2006 (UTC)

Cycle section

Since Patrick's mailbox is overfull, I'm going to ask it here.

I'm wondering why you added this and whether it's going anywhere. Currently it just looks like a random fact, unrelated to anything else that's in the article. That makes it a candidate for deletion due to marginal relevance.

It might be a reasonable observation if the Julian calendar article had a section on dominical letters, but there is a perfectly good separate article on that.

If I don't hear a case for this by tomorrow I'm going to delete it.

--Chris Bennett 00:09, 7 June 2006 (UTC)

I started to add technical information about the Julian calendar, similar to that in the article on the Gregorian calender. I plan to mention and link dominical letter.--Patrick 06:51, 7 June 2006 (UTC)

This isn't "technical information", this is currently just a bunch of numerical factoids, which don't belong in this type of article. I see you are responsible for adding the "similar" stuff to the Gregorian calendar article, which I don't normally monitor. These are rather more developed, but they are equally unfocussed.

It looks like what you want to do is to write something about algorithms for generating valid calendars. That might be an interesting addition, or (probably better) an interesting separate article. But I think you need to change your approach, if that's your goal.

--Chris Bennett 21:35, 7 June 2006 (UTC)

It just silly to object to basic information such as the 28-year cycle.--Patrick 22:34, 7 June 2006 (UTC)

It isn't basic information, its a factoid. Give it a context, a theme and a purpose and there might be something there. Right now it has no relevance to the rest of the article, which is about the structure and history of the Julian calendar. Consider it gone.

--Chris Bennett 03:06, 8 June 2006 (UTC)

Change of Start of Year

I'm baffled by this date change - this article mentions Founder's Day on 21st April? What is that? According to http://personal.ecu.edu/mccartyr/caesar.html - in 153BC the previous start of year was March 25th. This article: http://www.tyndale.cam.ac.uk/Egypt/ptolemies/chron/roman/chron_rom_cal.htm which seems very good holds that the previous date was Ides of March (which is the 15th March), or (controversially) May 1st. Anyone got evidence for dates or can confirm the dates given in the tynedale article? Thanks, Steve

As the author of the Tyndale site, thanks for the kudos ;-))

The change to January 1 is discussed on that site in the "analysis" section under AUC 600 = 154 (or directly at http://www.tyndale.cam.ac.uk/Egypt/ptolemies/chron/roman/154bc.htm) -- two sources that are available online are Livy, Periochae 47.13 (http://www.livius.org/li-ln/livy/periochae/periochae046.html#47) and the Fasti Praenestini (http://www.forumromanum.org/literature/verrius_frag.html#3 -- in Latin). Livy repeatedly mentions the consuls taking office on 15 March (not 25 March) before that, see e.g. Livy 22.1 for 217 BC (http://wyllie.lib.virginia.edu:8086/perl/toccer-new?id=Liv3His.sgm&images=images/modeng&data=/texts/english/modeng/parsed&tag=public&part=64&division=div2). The controversy about whether consuls took office on 1 May before 222 BC is discussed on my site at http://www.tyndale.cam.ac.uk/Egypt/ptolemies/chron/roman/222bc.htm

As far as I know the statement about 21 April being the start of the AUC year is wrong. This is the date of the Parilia which was celebrated as the anniversary of the foundation of Rome, but an anniversary is not the same as being the start of the AUC year. (For example, 1 January, the start of the Christian year, is not the anniversary of the birth of Christ.) The AUC year is always equated with a consular year in any text I've ever seen. However I've never bothered to correct this since I haven't checked out the other part of the statement, that this is somehow connected with the date of Easter, which I've also never seen anywhere else, but for all I know might be true.

--Chris Bennett 05:36, 20 July 2006 (UTC)

PS: The statement about AUC years being based on the Parilia is probably based on Censorinus, who equates AD 238 with AUC 991, and states that AUC years were based on the Parilia. (De die natali 21.6: "a Roma autem condita nongentesimus nonagensimus primus, et quidem ex Parilibus, unde urbis anni numerantur"). This is certainly not how AUC years are accounted by earlier historians, but by Censorinus' time the consulate was pretty much a ceremonial office, and usually changed several times a year. It looks like scholarly pedants of the day were rebasing the system, so the statement may actually be true for the third century AD. But I think its excessive to regard it as "formally" starting on this date -- there's no evidence that Varro thought of it this way, for example. --Chris Bennett 16:42, 21 July 2006 (UTC)

AUC Year

My feeling is that this is somehow relevant to Easter, as Roman Calendar March 15th is very close to Gregorian March 20-21. Did the fixed date Ides not fall on the Equinox? May 1st is also a festival, of course. This April date I am sure have seen before but in a different context in relation to Bede and the Gregorian Calendar. Was the point of the change of calendar that March 15th was no longer the Equinox? Thanks for that article, it's made a lot of things much clearer. My educated guess would be that the previous start of year was on or close to a festival. Steve

The Ides was not a fixed date (in principle), it was just the 13th or 15th date of the month. Every month had an Ides -- see the Roman calendar article. After the Julian calendar was installed it did become fixed, but was not deliberately aligned to the equinox, which in any case was held to be March 25. --Chris Bennett 14:49, 20 July 2006 (UTC)

Egyptian Calendar

Just as a piece of information for anyone else investigating, I had read that Julius Caesar was making use of the superior Egyptian calendar, but according to the article on Wikipedia, http://en.wikipedia.org/wiki/Egyptian_calendar their calendar was not solar or lunar but worked from Sirius. It does seem likely that he was making use of their calendar, however.

The Egyptians had been playing with a Julian like calendar -- IIRC that was what the Canopus Decree was requiring -- but they never did use it. Francis Davey 13:34, 20 July 2006 (UTC)

The article already indicates that Caesar's reform had input from Sosigenes of Alexandria, which we know from Pliny. Francis is surely correct that Sosigenes got the idea from the Canopic reform of 238 BC. This was almost certainly not an Egyptian reform but a Greek reform of the Egyptian calendar (which is probably one reason it failed). The idea can be traced earlier. The Almagest contains a number of dates in the calendar of Dionysius, which has been shown to have been a 365.25-day zodiacal calendar with a new year at the summer solstice. The epoch of this calendar was the summer solstice before the accession of Ptolemy II, in 285 BC. --Chris Bennett 17:06, 21 July 2006 (UTC)

OS NS

The notation "Old Style" (OS) is sometimes used to indicate a date in the Julian calendar, as opposed to "New Style" (NS), which indicates a date in the Gregorian Calendar. This notation is used when there might otherwise be confusion about which date is found in a text.

The above is not true. For example Charles I was executed on January 30 1648 Old Style and January 30 1649 New Style. New Style does not mean that the date is under the Gregorian Calendar, rather the year is adjusted to the January 1 instead of what ever start of year was used in contemporary documents. But the Julian calender month and day date is taken as the New Style date until the country where the event took place changed to the Gregorian calender. --Philip Baird Shearer 01:14, 30 November 2006 (UTC)

Berber calendar

I added some lines concerning the Berber calendar, which is used in Nortern Africa. If someone wishes to add something about it (the relevant article is missing), you can look at the it.wiki article (which is a featured article). --Vermondo 12:33, 10 December 2006 (UTC)

I removed "Muslim countries don't follow at all." because Turkey does use the Gregorian calendar (since 1926) and virtually all Muslim countries use it for business purposes. Furthermore, I know that all Saudi Arabian laws are dated in both the Saudi version of the Islamic calendar and in the Gregorian calendar. The subject is complicated enough to warrant a separate section, but in the Gregorian calendar article, not here. — Joe Kress 02:37, 11 December 2006 (UTC)

Month lengths

The way the "Month lengths" section is written now, it leads the reader to believe that the theory by "scholar Sacrobosco" is the currently accepted theory, until it mentions that this is "certainly wrong" at the end of the section. If I hear no objections I'll attempt to restructure that section to make it clear from the start that the widely repeated theory is considered wrong. --Romanski 18:35, 15 December 2006 (UTC)

I cannot accept "from the start" because no starting point can be discerned. Sacrobosco's theory is still widely repeated, which makes it accepted theory by the readers of those sources which repeat it because they offer no discussion, opposing views, or even hint that it is wrong. — Joe Kress 00:29, 17 December 2006 (UTC)

Not entirely sure what you meant; I've now added a sentence which should hopefully make it clearer. --Romanski 20:30, 19 December 2006 (UTC)

Now I undertand that your "from the start" means from the start of the section, not from any point in time, like 1232 (Sacrobosco), 1830 (Encyclopaedia Britannica), or some more modern date when it was realized that both were wrong. Your edits and those by Chris are certainly acceptable. — Joe Kress 03:14, 20 December 2006 (UTC)

The case disproving Sacrobosco could use clarification. Most urgently, the point regarding Note #2 ("First, a wall painting of a Roman calendar predating the Julian reform has survived,[2] "). This points to an image file which has no explanation connected to it (thus, the user cannot collaborate the evidence). Morever, judging from http://www.personal.psu.edu/users/w/x/wxk116/RomanCalendar/ and http://www.personal.psu.edu/users/w/x/wxk116/RomanCalendar/romecbib.html the image is not of a pre-Julian reform calendar, but of a post-Julian reform calendar. From the latter page, for example, "The calendar represented here is from the heyday of the Empire, when the Julian calendar was well-established."Imboot 05:03, 15 March 2007 (UTC)

Also, see http://www.geocities.com/christopherjbennett/ptolemies/chron/roman/chron_rom_cal.htm

. August has 29 days, pre Julian reform. The only surviving pre-Julian calendar is "Fasti Antiates Maiores". It doesn't look like that's the one you are linking to... Imboot 05:31, 15 March 2007 (UTC)

The calendar illustrated is the Fasti Antiates Maiores. Check the heading of the 8th month (SEX) and its length (XXIX). In fact compare the lengths to the lengths listed in the Wiki article, immediately after the Fasti is mentioned.

You have misundersood what "here" means on the page you have cited. It is not referring to this illustration, but to the whole of that website. Compare this calendar to the others illustrated on the same page and you will see they are very different. The others are Julian. The source of your confusion is that the author of that page has illustrated a page about the Julian calendar with a pre-Julian calendar without explaining that that's what it is. --Chris Bennett 05:53, 15 March 2007 (UTC)

Thank you. Then, how about a) noting that note #2 points to an image of Fasti Antiates Maiores, the only surving pre-Julian calendar.

The current text reads: "a wall painting of a Roman calendar predating the Julian reform has survived[2]". All that is missing, compared to your statement, is that that calendar is known today as the Fasti Antiates Maiores. In the context, it really doesn't matter what its called -- how would it improve our understanding of Sacrobosco's theory or its errors? Or are you saying that it needs to be said that the calendar image linked at [2] is that calendar? That seems to me to be stating the obvious. --Chris Bennett 17:08, 15 March 2007 (UTC)

And b) making it clear to the who actually changed the number of days per month (in particular, the misunderstood Feb. and Sex./Aug.) pre and post the Julian and Augustus-period reforms (if not Julius, _someone_ changed the number of days in Sex. from 29 to 31).

The text under "Month Names" reads: "Immediately after the Julian reform, the twelve months of the Roman calendar were named Ianuarius, Februarius, Martius, Aprilis, Maius, Iunius, Quintilis, Sextilis, September, October, November, and December, just as they were before the reform. Their lengths were set to their modern values." What is ambiguous about this? --Chris Bennett 17:08, 15 March 2007 (UTC)

According to Augustus, Sex. had 31 days before the name change. Readers can easily wrongly come away with a wrong understanding (after reading the bit that calls Sacrobosco erroneous) if the timing of the change in Sex. from 29 to 31 is not explained clearly.

The text says: "He then said Augustus changed this to: [lengths omitted] so that the length of Augustus would not be shorter than (and therefore inferior to) the length of Iulius. ... There is abundant evidence disproving this theory. " and that evidence includes: "... the 31-day Sextilis given by the new Egyptian papyrus from 24 BC".

What part of this discussion leads you to think that "Readers can easily wrongly come away with a wrong understanding"? Seems to me that it requires not paying any attention to what is written.

Don't get me wrong, I'm sure the text can be improved, but it helps to understand what the problems actually are with the current text. As far as I can see, your difficulties have all arisen out of misunderstanding something that's on a completely different page. --Chris Bennett 17:08, 15 March 2007 (UTC)

Reviewing this, I do think that thesection on Month Lengths should begin with an explicit statement that Caesar set the month lengths to their modern values. I have added this, and made apropriate adjustments elsewhere. --Chris Bennett 18:15, 15 March 2007 (UTC)

Finally, it would also be helpful if the image at Roman_calendar is also labeled as F.A.M. Imboot 07:34, 15 March 2007 (UTC)

Julian calendar is based on sidereal year

The basis of the Julian calendar is the sidereal year - the orbital period of earth around the sun. The idea of earth orbiting around the sun has been known to the ancient Egyptian astronomers. Later it has been inherited by the Pythagoreans, and has been quite prevalent among them. Later in 3rd century BC this idea had set foundations of ancient Hellenic mathematics doctrine, especially that of Aristarchus of Samos, who is considered a predecessor of Nicolaus Copernicus.

At the time of creation of the Julian calendar (46 BC), the theory of Aristachus has been prevalent among the astronomers and scientists. Hence it is no wonder that Sosigenes of Alexandria and other creators of Julian calendar have used the time required for one complete revolution of the earth about the sun, relative to the fixed stars, i.e. a sidereal year, as a basis of the Julian year. This real astronomic unit has become a pattern for the Julian calendar.

In that respect, the Julian calendar is heliocentric, while the Gregorian is geocentric. In the calendar reformation, the Roman Catholic Church initiators were persistent followers of Ptolemy geocentric system, and have used the tropical year (the time needed for a full revolution of sun about the earth) as an absolute unit. The very idea of existence of the sidereal year has been declared a heresy by the Roman Catholic Church, and its acceptance had hardly theoretical consequences (for example, destinies of Giordano Bruno and Galileo Galilei).

The Gregorian calendar project received negative evaluations from astronomers of that time such as Michael Maestlin and Anders Celsius, and has been graded as "meaningless" by the universities of Vienna and Sorbonne. ~Mkol 14:03, 2 January 2007 (UTC)

Do you have sources, or is this just supposition? I do not doubt that Sosigenes was aware of the difference between the sidereal and tropical years, but do you have any evidence that he specifically chose to base the Julian calendar on the former, and not the latter? Especially since the match between the new calendar and the sidereal year was no better from the tropical year, and that the tropical year was of more importance to Roman agriculture than the stars.

I also doubt that the 16th century Church considered the sidereal year heresy, since it was something known for two thousand years, and easily measurable by contemporary astronomers. I do not know why the Church would consider it "heresy", since in geocentric terms it merely relates to the motion of the fixed stars, not of the earth. --Nike 01:45, 3 January 2007 (UTC)

What do these links mean? Are they vandalism? http://www.enigmasdabiblia.com/ http://cronologia.enigmasdabiblia.com/ 62.1.230.209 16:24, 9 January 2007 (UTC)

Roger Bacon's Plea for Calendar Reform 300 Years Before Gregorian Calendar

In his Opus Maius, Roger Bacon advocated reforming the Julian calendar because during his life time the vernal equinox was already nine days ahead of March 21st (the vernal equinox decreed by the Council of Nicea). Since the Wikipedia article on Roger Bacon references and links to the Julian calendar article, should not a reciprocal link be included? And perhaps some note that Roger Bacon realized the errors in the Julian calendar and the problem of celebrating Easter on the correct day. He was disciplined and largely ignored or suppressed by his superiors for his ideas and teachings.

Best wishes, Jon Moss 18:56, 15 March 2007 (UTC)

I think it would be better to put this into the Gregorian Calendar article. The History section there is missing a lot of background like this. Not just Bacon, also Sacrobosco and (IIRC) Grosseteste were early advocates of reform. --Chris Bennett 19:53, 15 March 2007 (UTC)

Hi there... I have NO idea what I'm doing so I'm sorry if this comment comes out screwy or if this isn't in the right place. Right then, having prefaced it, onto my question!! I am interested in finding out more information about the Charlemagne calendar... did it replace the Julian calendar for a time, or run along beside it? And does anyone have any references for more information on the calendar Charlemagne set up (as in, why he changed the month's names, etc.)? I have looked in the article on Charlemagne but there is no mention of him changing the calendar. 122.148.138.178 12:23, 4 April 2007 (UTC) Cadence

Eponymous years

Recent edits have challenged the statement that the last year named after consular eponyms was 541, the year the consulate was abolished.

I haven't looked at any source data for this period. I thought the statement was almost tautologous. I don't find either proposed edit satisfactory because they raise unanswered questions. If eponymous dating continued after 541, who was the eponym? And when was eponymous dating abandoned? We ought to be able to do better than "a few years later", even if it's only to say that the "last known" eponymous year is XYZ. --Chris Bennett 14:21, 31 July 2007 (UTC)

Sorry for the ambiguous edit. Byzantine emperors included a consular year in their dating formulas, which gradually disappeared but was still used by Heraclius. In the west, post-consular years were used (so many years after consul so-and-so). Bickerman states that Roman emperors after 534 (!) and Byzantine emperors after 541 were also consuls (Consuls of the Roman Empire at the end). List of late imperial Roman consuls lists the years between imperial consuls as "post consulatum". Regnal Formulas of the Emperor Heraclius (pdf, 5.45 MB) confirms this. Chronological reckoning in Byzantine Egypt indeed states that the emperor generally held the consulate only for the year of his accession—all other years in that emperor's reign were postconsular (p.285). It also notes that 542 to 566 were postconsular Basilius years. Chronological Systems in Byzantine Egypt has a table of consular years. Other sources only refer to a certain year as an emperor's numbered consular year. Some uncatalogued papyri of theological and other interest in the John Rylands Library has "In the consulship/of our most pious ruler [...?],/the sixth year". Chronological observations on later Byzantine documents states that 632 was year 22 of the consulate of Heraclius Sr (p.362), but also states that the consular year was very rare during his reign (p.361). Despite this evidence, some sources state thar Justinian abolished the consulate after 541 (Sports of the Byzantine Empire note 44). — Joe Kress 06:35, 1 August 2007 (UTC)

Phocas in consular robes [2], Mauricius in consular robes [3], Tiberius Constantine in consular robes [4], so consulate was not abolished in 541. Basilius was not even the last civilian consul because Heraclius and his father were declared consuls by the Constantinople's senate before Heraclius became emperor. See also:[5]. This source [6] says the last consul was Constans II and this [7] source says it was Leo the Wise who abolished dating by consular years (but not necessary the title itself).--Dojarca 11:53, 1 August 2007 (UTC)

OK this is all good information, why not add a summary at the appropriate place?

One query: If Constans II (r. 641-668) was the last consul but consular dating was ot abolished until Leo the Wise (R. 886-912) does that mean that postconbsular dates for Constans II are known for the intervening 200 years?? That seems worth a mention in itself!

I've restored Phiip Sherarer's addition (although its overly fussy in its sourcing for my taste) since this seems to have become an innocent victom of this issue. However I have left out the old statements on the end of consular dating in the expectation that something more authoritative will get added shortly. --Chris Bennett 14:30, 1 August 2007 (UTC)

Recent addition.

Some remarks. First, Julian did not reserve the consulate for the emperor. Just the opposite: before him the consul had to pay large sum of his own money for games to be held. Julian abolished this tradition. But since the change was unpopular, there were no candidates to become consuls and as such be associated with the reform.

Hmm. Seems to depend who you read. But I've now looked up Procopius, who talks about apopintment to be the office being abandoned by Justinian, rather than any explicit prohibition, which is in line with Klimshin; I'll reword. (Feel free to adjust!) --Chris Bennett 17:25, 5 August 2007 (UTC)

About the book. It is I.A. Klimshin, "Notes on our calendar". The chapter named "Year-counting by consuls". The translation of the paragraph about Leo is Successors of the creator of Constantinople's Sophia restored the custom to declare themselves consuls 1st January and throw money to people as it was used before. Hence year-counting "post consulatum" continued until IXth century. It was only emperor Leo the Wise (886-912 AD) who issued a decree finally prohibiting to count years by consuls.--Dojarca 08:04, 5 August 2007 (UTC)

Thanks. I looked up Gibbon on this (ch 40, at end) who refers to Novell xciv for it, i.e. part of Leo VI's law code, so I think the text can stand. However, your source justifies my original question, and apparently does not address it. If post-consulatum dating was in use, however rarely, till the late 9th century, and if Constans II was the last emperor to take the consular title, then who was the reference point for the post-consular dates in the intervening 200 years? Constans II? Or the reiging emperor? In the latter case then a post-consular year was effectively a regnal year.

Gibbon also says that Anno Mundi dating was introduced at the Council of 681, which is a point worth noting if I can verify it. --Chris Bennett 17:25, 5 August 2007 (UTC)

I was trying to determine whether a discussion of Anno Mundi years was appropriate to the article. But there were several types of AM years used with the Julian calendar, all in the Eastern Roman Empire, starting in the fourth century. Such a discussion would be somewhat long and convoluted. The AM year mentioned by the Council in Trullo is written 6109, which is usually emended to 6199, placing the council in the next year, AD 692, not 681. See [8]. — Joe Kress 07:02, 7 August 2007 (UTC)

Thanks for the citation, though this seems to show ecclesiastical usage rather than official adoption. There is an article on Anno Mundi eras but it doesn't say much. What would be appropriate here, apart from a brief motivating background that is already there, is to discuss when such eras started being used offically or primarily, with a more in-depth discussion in the other article. At the moment there is clearly a gap between Justinian and the 10th century. --Chris Bennett 16:17, 7 August 2007 (UTC)

The gap after Justinian is especially long because the earliest surviving unambiguous (cannot be any other era) official use of the Byzantine era occurred in AM 6496 (AD 988) by Basil II as cited by V. Grumel, La chronologie (1958), p. 127 (in French). This is also the same year that Vladimir I of Kiev adopted that era for the Rus'. The Byzantine era began September 1 because the indiction began on that date (already mentioned in the article), whereas the Rus' era began March 1 because the earlier Viking pagan year began about that date. The Byzantine era remained the dominant era used by the Byzantine Empire until its collapse in 1453. Ivan III of Muscovy shifted the beginning of the Russian era back to September 1 near the end of the fifteenth century when he adopted the double-headed eagle and other regalia of the former empire when he married the niece of the last Byzantine emperor, Sophia Paleologue. Peter the Great abandoned the Byzantine era for the Anno Domini era effective January 1, 1700 (Julian)—ironically, the same year that Western Protestant Europe was adopting the Gregorian calendar. — Joe Kress 20:32, 7 August 2007 (UTC)

It seems that Byzantine chronographers were using different AM epochs from at least the 5th century (shades of AUC). But Grumel wrote 50 years ago, I would be surprised if there were no official decrees or inscriptions that have not come to light from the period between Heraclius and Basil II that would document the transition from post-consular / regnal dating to AM. How about coinage?

Well, I'll add this to the list of interesting research topics, to be got to at some future date unless someone (hopefully) beats me to it. --Chris Bennett 00:48, 8 August 2007 (UTC)

PS did the Russian year N begin 6 months before or 6 months after the Byzantine year N? --Chris Bennett 00:50, 8 August 2007 (UTC)

I can't locate my original source, which may have been an article in the Great Soviet Encyclopedia, but История календаря в России и в СССР (Calendar history in Russia and the USSR) has a table showing that the 1 March Russian year started six months after the 1 September Byzantine year with the same number. It also states that according to church tradition the shift occurred in AD 1492 (AM 7000), meaning the March style AM 7000, which began 1 March 1492, ending six months later when the September style AM 7001 (= Byzantine AM 7001) began. — Joe Kress 19:59, 9 August 2007 (UTC)

All good and interesting stuff. Can I ask you to draft some text on it? (It would be interesting to know when other orthodox countries or churches -- Armenia, Georgia, Bulgaria, Serbia etc adopted and abandoned AM eras, but that's getting very ambitious.) --Chris Bennett 04:41, 10 August 2007 (UTC)

I have not found any info regarding the AM eras of countries other than Byzantium and Russia. — Joe Kress 21:14, 11 August 2007 (UTC)

The Mysterious 46 BC

I have a few questions regarding 46 BC (the "last year of confusion" – but I am still confused). Is the following lost in antiquity, or is there a way to get this kind of detail?:

1. What were the names of the two temporary months added between November and December?

Two letters of Cicero mention Intercalaris Prior and Intercalaris Posterior. You sometimes see them called Undecember and Duodecember in modern accounts; I don't know the basis for this but it is wrong. Will add a comment on the point when I get time.

2. What were the actual lengths for these two months (i.e., how does 22-23-22 split into two months)?

No one knows.

3. What were the naming conventions for all those one-time days within those two months?

They will have had a Kalends, a Nones and an Ides like any other month. No one knows the placement of the Nones and the Ides. I would guess that the Nones was the 7th and the Ides the 15th, with a long countdown towards the next Kalends.

4. Where did they add the one-time-adjustment (the 23-days added to the 28-day February of 46 BC)? Should I assume that the action followed the practice for ntercalaris years (which you so eloquently described in the first paragraph of the "Motivation" section)? If so, the 23 days were added "between" the 24th and 25th. Alternatively, were they added a month-end (i.e., after the 28th day)?

They are said to have been added "between" November and December, and the arithmetic of a 445 day year suggests that they did not steal days from November in the way that a reglar intercalaris stole days from February.

5. What were the naming conventions for these 23 additional days added to February in 46 BC? I am sure that they could not use the "a.d.VI.bis.Kal.Mar." designation twenty-three times. Although it was a one-time deal, there must have been some kind of naming convention. If they followed the naming convention for prior (pre-47 BC ntercalaris years – I am still clueless as to what they called them. So, if anyone knows – that would be super!

The Intercalaris inserted after the Terminalia was a regular Intercalaris following the regular numbering conventions, whatever they were. For the dates before Kal. Int., there is some doubt as to what this was. As well as the expected dates a.d. Kal. Mart, we have dates a.d. Terminalia, which could be appropriate, but also dates a.d. Feralia and a.d. Quirinalia -- i.e. to the major festivals in late Februarius (though no a.d. Lupercalia, at least not yet). We also have an a.d. Terminalia date in a year which we know was NOT intercalary. For the Intercalary month itself, it had a Nones on day 5 and an Ides on day 13 (I see this is not in the Roman calendar article, will add when I get the time.)

355 days in 47 BC. 355+22+23+22+23=445 days in 46 BC. 355+2+2+2+1+1+1+1=365 days in 45 BC. Cool! (See, he can be taught! <grin>)

Great, wonderful, exceptional article! Your hard work is appreciated. — Tesseract501 22:25, 14 August 2007 (EST)

HTH --Chris Bennett 17:25, 15 August 2007 (UTC)

47 BC versus 45 BC

I am inserting the following on two discussion pages: "Julian calendar" and "Roman calendar."

Replying here only

I appreciate all the scholarly effort that went into the writing of the "Julian calendar" and the Roman calendar" pages. I particularly appreciate your commitment in keeping other contributors focused on the parameters of the articles. It is refreshing to see pages that keep the focus on the topic at hand, instead of including satellite issues that are covered in other Wiki articles.

I do have two questions. They are concerning the calendars dealing with the Julian Reform period. I.e., the actual reform period dealing with the year 45 BC, and the base-line year of 47 BC – without detailing the 47 BC data, we are left to deduce the reform-period’s additions, modifications, and deletions). Hence, after reading both articles, my lack of deductive reasoning has me derailed.

Question #1 – my concern is regarding two monhs (Quintilis and Sextilis - July and August). The "Roman calendar" article indicates a pre-Julian Quintilis with a 31-day duration and a Sextilis of 29 days). How do we resolve this against the adjustments indicated on the "Julian calendar" article? I.e., 2-day adjustment (addition) in 45 BC for Quintilis; and, no adjustment for Sextilis.

Julian calendar article is wrong, I've fixed it, thanks.

The following shows the (753 BC?)

No, probably fourth century with the Decemvirs. The Roman calendar is weak on the early history, which is in any case still controversial. Should fix one day.

355-day calendar that resides on the "Roman calendar" page. I have added notations at the end of the months, showing the 45 BC adjustments that are referenced on the "Julian calendar" article. Read my notation on the Quintilis and Sextilis lines. Based on the two articles, I assume the following:

Martius (31 days).

Aprilis (29 days in 753? BC; 30 days (29+1) in 45 BC).

Maius (31 days).

Junius (29 days in 753? BC; 30 days (29+1) in 45 BC).

Quintilis (31 days in 753? BC). If if this is true, then it must have lost two days by the time 47 BC arrived. The "Julian calendar" article indicates that 2 days were added to Quintilis (July) on the 45 BC calendar. If something like that happened in the period between 752 and 47 BC, it would be helpful if the "Roman calendar" article indcated such. If no historical records detailing this are available, it would still be helpful to have a sentence indicating that such a change appears to have taken place "some time" before (by) 47 BC. Otherwise, easily confused folks (like me) get turned around when we try to bridge the gap between the two articles and the gray zone that is 48, 47, and 46 BC.

Sextilis (29 days). This is a bit mysterious in itself. When did August (Sextilis) go from 29 days to 31 days? Did this happen sometime after 753 BC but before 47 BC? Maybe days were taken away from July?

Happened with the Julian reform -- again, there is an error in the Julian calendar article on this pont.

September (29 days in 753? BC; 30 days (29+1) in 45 BC).

October (31 days)

November (29 days in 753? BC; 30 days (29+1) in 45 BC).

December (29 days in 753? BC; 31 days (29+2) in 45 BC).

Ianuarius (29 days in 753? BC; 31 days (29+2) in 45 BC).

Februarius (28 days).

Question #2 – the "Roman calendar" article indicates, "When Julius Caesar added a day to September, he did not add it to the end of the month. Rather, the new day that was added was the day after the Ides:

"a.d. XVIII Kal. Oct. = 18 days before the Kalends of October = 14 September

"As a result, the position of all the following dates in September got bumped up by one day"

The "Julian calendar" article indicates, "Macrobius states that the extra days were added immediately before the last day of each month to avoid disturbing the position of the established Roman fasti (days prescribed for certain events) relative to the start of the month. However, since Roman dates after the Ides of the month counted down towards the start of the next month, the extra days had the effect of raising the initial value of the count of the day after the Ides."

So, which article is correct (or is it that the historical sources disagree: Macrobius, and the unreferenced source on the "Roman calendar" page)?

Roman calendar is wrong, Julian calendar is right. Joe Kress pointed out this problem at one time, and it got fixed in one article but not both. Thanks for pointing out the inconsistency, will fix.

Or, is this apparent disagreement concerning September alone (i.e., were the adjustment added the-day-after-Idus for September, but the end-of-month for other months)? The naming conventions that I have seen for September run consistently from Idus thorugh month-end.

Or, are they both correct? Was the September addition made by Caesar, as referenced in the "Roman calendar" article, something that predates 45 BC? Was this omething Julius did to September BEFORE the Julian period. E.g., September 47, 48, or 49 BC, or even earlier. If so, it might be worthwhile to clarify this in the article.

Thanks again for sharing all the great information that exists on the two pages. You have pulled together a truckload of valuable information from dozens of sources. Great job at bringing it all together in clear and thoughtful articles.

By the way, I would love to see listed details (month-name and day break-outs) for the year 46 BC too – but that’s just because I love lists. You know, in your spare time <grin>

see above -- can't be done for the special intercalary months.

— Tesseract501 00:05, 15 August 2007 (EST)

HTH --Chris Bennett 17:34, 15 August 2007 (UTC)

1751? or 1752?

"Note that as a consequence of change of New Year, 1 January 1751 to 24 March 1751 were non-existent dates in England."

I've understood that England did both the first-of-the-year change and the adoption of the Gregorian calendar in the same year: a few months apart in 1752. If the above sentence is correct, then I think it needs some further explanation. Thanks. --Hordaland (talk) 11:28, 6 January 2008 (UTC)

I've clarified that comment and made it a footnote because it is a minor aspect. — Joe Kress (talk) 07:26, 7 January 2008 (UTC)

Justinian II

Justinian II was declared consul:

Hic depositus est Caedual, qui et Petrus, rex Saxonum, sub die XII Kalendarum Maiarum, indictione II; qui uixit annos plus minus XXX, imperante domno Iustiniano piissimo Augusto, anno eius consulatus IIII, pontificante apostolico uiro domno Sergio papa anno secundo. [9]--Dojarca (talk) 04:03, 9 January 2008 (UTC)

Bissextile Day

Much confusion has been created by two oft-repeated fallacies about this reform. The first is that Augustus changed the lengths of the months. About this, more later. The second relates to the way intercalation was handled in the reform.

In the old system, the most simple adjustment was the insertion of Intercalaris after 23 February (the Terminalia). This meant that the following days disappeared from the calendar: the 24th (a.d. VI Kal. Mart.), the 25th (a.d. V Kal. Mart.), the 26th (a.d. IV Kal. Mart.), the 27th (a.d. III Kal. Mart.), and the 28th (Prid. Kal. Mart.). They reappeared, however, with their attendant festivals, at the end of the following month.

Thus the 23rd of Intercalaris was a.d. VI Kal. Mart., the 24th a.d. V Kal. Mart., the 25th a.d. IV Kal. Mart., the 26th a.d. III Kal. Mart., and the 27th Prid. Kal. Mart. For an explanation of the Roman dating system, see the article Roman calendar. The festival of Regifugium, normally 24 February, transferred to 23 Intercalaris. All Caesar did was abolish Intercalaris and replace it with a single extra day in the same position, i.e. immediately after the Terminalia. <ref> "Lastly, in consideration of the quarter of a day, which he regarded as completing the true year, he established the rule that, at the end of every four years, a single day should be intercalated, where the month had been hitherto inserted, that is, immediately after the terminalia; which day is now called the bissextum." Censorinus, The Natal Day.</ref> The five days which had hitherto been transferred to Intercalaris then followed.

Thus in leap years the sequence became:

a.d. VII Kal. Mart. (23 Feb.); a.d. bis VI Kal. Mart. (or a.d. VI Kal. Mart. posteriorem) (24 Feb.); a.d. VI Kal. Mart. (or a.d. VI Kal. Mart. priorem) (25 Feb.); a.d. V Kal. Mart. (26 Feb.); a.d. IV Kal. Mart. (27 Feb.); a.d. III Kal. Mart. (28 Feb.); and Prid. Kal. Mart. (29 Feb.).

Our system of numbering days in order from the beginning of the month dates from the late middle ages, although the Roman Catholic Church did not adopt it until late in the twentieth century. That is why, in living memory, the feast of St Matthias (a.d. VI Kal. Mart.) was observed on 24 February in ordinary years and 25 February in leap years.

Some people have gone to great lengths to set up the case that in leap years it was 25 February which was intercalated, appealing to obviously unreliable sources. <ref> Celsus 39, cited in the Digest of Justinian 50.16.98. [1] [10]. In this sentence the redactor claims that if someone is born on 1 March in a leap year his birthday falls six days earlier! In 98. [2] he says "It is, however, established that there are twenty - eight days in the intercalary month." (There were 27). </ref> Reference is then made to an inscription of the year 168 AD, which states that a temple was dedicated on the fifth of the calends of March of that year, which followed the bissextile day. <ref>Roscoe Lamont, "The Roman Calendar and its reformation by Julius Caesar" [11], Popular Astronomy '27' (1919) (pp. 583 - 595 (at p. 589)). The reference is the second article in the hyperlink; its last page is here.</ref>

This is perfectly correct. A.d. V Kal. Mart. follows a.d. VI Kal. Mart., which is duplicated in leap years (AD 168 was a leap year). That is why it was called the bissextum (English, bissextile).

To close the case, when the question came before the Roman judges they decided that, of the duplicated days, the intercalation was the one labelled posteriore (or bis), which separated the normally consecutive festivals of Terminalia and Regifugium.<ref> Mommsen: Roman Chronology.</ref>

As the written evidence of intercalation between 23/24 February cannot be explained away it is then claimed "In later times, however, the relative position of the two days was sometimes reversed." If this had happened it would have been unique in recorded history. People do not tinker with calendars in this fashion - they are imposed by the state. No details are offered of when, where, why and by whom the alleged reversal took place. The Roman Catholic Church is specifically mentioned, but examination of old church calendars shows that intercalation was always between 23/24 February.<ref>The Roman calendar and its reformation by Julius Caesar (op.cit., p. 590).</ref>

The inescapable inference is that intercalation between 23/24 February was universal and unbroken. A similar evidential situation arises in relation to the lengths of the months, which will be considered shortly.

Material inserted into the body of the article by an anonymous user 217.169.37.146 moved to the Talk page --Chris Bennett (talk) 15:23, 21 January 2008 (UTC)]

First, dear Anonymous, it would be nice to be debating with a human being rather than an IP address. Also, this page is the right place for argumentation, not the main article. We should debate here until we agree on what changes to make.

Second, you are promising a similar discussion of the month lengths, but one already exists in the article that clearly shows the evidence that the lengths did not change after Augustus' reform.

Third, I agree that "sometimes" is wrong, and that the prior bissextile day was standard by AD 168.

Fourth, I also agree that the quote from Celsus in the Digest is corrupt, but the corruptions are minor and obvious: "Kal Mart" ought to be "ad VI Kal Mart" -- simple omission -- and 28 ought to be 27 -- reflecting loss of knowledge of the pre Julian calendar. No such corruption can explain the core statement. Reversing the two days requires a wilful falsification. Why?

Celsus wrote in the late first century or early second century. He was a knowledgeable and respected jurist. This is the earliest statement we have on the position of the bissextile day, and it is not to be dismissed out of hand. It shows that there was uncertainty at that time as to which day was the bissextile, and also gives the official position of the state.

Fifth, Celsus' remark, in conjunction with Pompeian graffiti, allows us to explain an odd statement in Dio 60.24.7 concerning a shift of the market day to avoid a religious festival. You can read the argument at http://www.tyndale.cam.ac.uk/Egypt/ptolemies/chron/roman/ad60.htm or in my article in ZPE 147.

Sixth, calendar innovations do not necessarily happen at the direction of the state. Consider the 7-day week, not made official till Constantine, but already in use in the first century AD. Or indeed the movement of the intercalary day to 29 February.

We don't know what position Caesar placed the bissextile day in since we have no evidence on it predating the Augustan reform. That reform seems to have introduced the 48-hour day to avoid a market day conflicting with the Regifugium without interrupting the nundinal cycle. This was the official rule in Celsus' time. However, a 48 hour day is unnatural and the Regifugium, along with the other archaic festivals revived by Augustus, fell into disuse. Whether by fiat of the state or by natural selection, the prior bissextile day won out over the posterior bissextile day by AD 168.

So, I propose the following text:

The old intercalary month was abolished. The new leap day was originally inserted, like the intercalary month it replaced, after the Terminalia (23 February). It was dated as ante diem bis sextum Kalendas Martias, usually abbreviated as a.d. bis VI Kal. Mart.; hence it is called in English the bissextile day. The year in which it occurred was termed annus bissextus, in English the bissextile year.

The exact position of the day changed over time. In the first century AD, a.d. bis VI Kal. Mart, and a.d. VI Kal. Mart. were considered to be two halves of a single 48-hour calendar day, with the bissextile occurring in the second half.[2 -- Celsus] By AD 168, the relative position of the two days was reversed.[3 -- CIL VIII 6979] During the late Middle Ages when days in the month came to be numbered in consecutive day order, the leap day was considered to be the last day in February in leap years, i.e. 29 February. The Roman Catholic Church did not make this change till the late twentieth century.

--Chris Bennett (talk) 16:51, 21 January 2008 (UTC)

Scholars disagree regarding the location of Celsus' intercalary day. His applicable words were quoted within Justinian's Digest at 50.16.98. However, at least the first portion of his words were mistranslated by S. P. Scott as "When anyone is born on the kalends of a bissextile year". His words in Latin were "Cum bisextum kalendis est", so he did refer to a. d. bis VI. Kal. Mart. Regardless, he distinguished the two days of the doubled a. d. VI. Kal. Mart. as posteriore and priore. Although he explicitly stated that the posteriore day was the bissextus or intercalated day, he did not state its position relative to the February days before and after, that is, relative to a. d. VII. Kal. Mart. and a. d. V. Kal. Mart., so which of the two sixths was his posteriore day is subject to interpretation. Thomas Hewitt Key (pp. 231–32, 1875) stated that Ideler (1806/26/29?) thought that the posteriore day was immediately after a. d. VII. Kal. Mart. by ordering the two sixths in accordance with the Roman backward count of the days of the month, opposite to the direction of time (also claimed by Anonymous). Although Roscoe Lamont (pp.589-590, 1919) also stated that that was Ideler's opinion, Lamont agreed with Mommsen's interpretation (1858) that it was the day immediately before a. d. V. Kal. Mart. by ordering the two sixths in accordance with the modern forward count of the days of the month (Anonymous thinks that Mommsen placed posteriore one day earlier). Assuming that the translation "the bissextile day" mentioned for inscription of 168 really was bissextus, which, according to Celsus, was only one of the two days of the doubled day, then it could not mean the day to be doubled as Anonymous claims.

From 238 onward (with one possible exception) the intercalary day was between a. d. VII. Kal. Mart. and a. d. VI. Kal. Mart.. In 238, Censorinus stated (De die natali, tr. William Maude, 1900, chp. IX/XX, p. 30): "Caesar ordered that after each revolution of four years, there should be added, after Terminalia, instead of the ancient month, an intercalary day, which is now called leap-year day (quod nunc bissextum vocatur)". About 430, Macrobius stated (Saturnalia, tr. Percival Vaughan Davies, (1969), chp. 14, p. 97): "[Julius Caesar] ordained that the priest in charge of the months and days should insert one day every fourth year in that month, and in that part of it, in which of old an intercalary month used to be inserted, that is to say, immediately before the last five days of February. This intercalary day he ordered to be called bissextus [as doubling the sixth day before the Kalends of March]." I presume that the "last five days of February" means a. d. VI, V, IV, III Kal. Mart. and pridie Kal. Mart.

In 725, Bede copied verbatim the words of Macrobius regarding the intercalary day according to Faith Wallis, Bede: The reckoning of time (1999), chp. 12, p.50. However, her English translation differs somewhat from that of Davies: "[Julius Caesar] ordained as well that in the fourth year, the priests, who were responsible for months and days, should intercalate one day in the same month and place where men of old had intercalated the month, that is, before the fifth day from the end of the month of February, and he ordered that it be called the bissextus." I do not have access to either Latin text. In 1016, Byrhtferth stated (Byrhtferth's Enchiridion, ed. Peter S. Baker & Michael Lapidge, (1995), II.1, p. 61–63): "[the reverend abbot Dionysius] said that Julius, the ingenious emperor, discovered or found out this bissextile day and ordained that it sould be in the location where we now hold it, on 24 Feb. [.vi. kalendas Martii.] The five days between 24 Feb. and 1 March [.vi. kalendas and kalendas Martii] are called interkalares in Greek and interpolares uel additi in Latin—that is, the added or in-between days; and in Greek they are also called epagomene, with the same meaning as we have just said. ... the bissextile day sits on its throne on the first sixth kalends of March [forman .vi. kalendas Martii, 24 Feb.], and on the second sixth kalends [aeftran .vi. kalendas, 25 Feb.] one must hold the feast of the mass-day of St Matthias." In the portion before the ellipsis, he apparently thinks the calendar had two successive days both labeled .vi. kalendas whereas in the latter portion he distinguishes them with forman (former) and aeftran (after) (both Old English).

The possible exception is the statute De anno bissextili, 21 Henry III, 1236: "the Day increasing in the Leap-Year ... and the day next going before". All writers assume that "the day next going before" was associated with the intercalary day, that is that the intercalary day 'grew' from the day before, making it an exception. But Henry III never identifies the day before, so it could have been a. d. VII. Kal. Mart., rather than the day to be doubled. If the day before was a. d. VI. Kal. Mart., Henry would be flouting established Church dictum that the day before was a. d. VII. Kal. Mart. Alternatively, "before" could be used not in the sense of time but in the sense of the continuing Latin use of counting days backward. In any case, most writers treat the day before in the modern sense as February 28.[12][13][14] One source stated that despite the decree the dominical letter still changed on 24 February, the first half of the doubled day.[15] — Joe Kress (talk) 05:42, 28 January 2008 (UTC)

Joe, thanks for taking the trouble to do the proper work of the Anonymous IP Address (hereafter AIA) for him. Why AIA insists on conducting polemic in the body of the article and refuses to debate the topic on the talk page is beyond me. There is a legitimate, if very minor, question here, which could have been debated in an ordinary civilised fashion. AIA's latest edits are now attacking me by name for a position on the bissextile that I have never said (and have an open mind about), and implying that I hold a position on Caesar's month lengths that I have never held and in fact have gone to some lengths to refute, both here and elsewhere. Whoever he is (I have a guess), this is just ignorant and boorish. I'm very grateful to you for placing the topic in a proper context.

Now, although I've been aware of this issue, and I'm aware that respectable scholars (Ideler vs Mommsen) have taken opposing positions on interpreting Celsus, I've never had any reason to research the literature in depth. As I explained to Mark Passehl (Appietas), I ran into it tangentially when I was looking into the relationship of the nundinal cycle in the Julian calendar to that of the pre-Julian calendar. The Pompeian graffiti CIL IV 4182+8863, together with Celsus' text and Dio 60.24.7, provide a neat solution to this problem. Collaterally, that solution implies the bissextile to be the second of the two days, though I suppose it would not be impossible to reconcile it with the view that it was the first, on the same logic that Ideler used to argue the "posterior" day was in fact the earlier of the two.

I've run down the Latin text of Justinian's extract from Celsus (now at http://webu2.upmf-grenoble.fr/Haiti/Cours/Ak/Corpus/d-50.htm#16). It reads as follows:

Celsus libro 39 digestorum

pr. Cum bisextum kalendis est, nihil refert, utrum priore an posteriore die quis natus sit, et deinceps sextum kalendas eius natalis dies est: nam id biduum pro uno die habetur. Sed posterior dies intercalatur, non prior: ideo quo anno intercalatum non est sexto kalendas natus, cum bisextum kalendis est, priorem diem natalem habet.

1. Cato putat mensem intercalarem additicium esse: omnesque eius dies pro momento temporis observat extremoque diei mensis februarii adtribuit Quintus Mucius.

2. Mensis autem intercalaris constat ex diebus viginti octo.

As you note, there is nothing wrong with the bissextum. THe more interesting comment is that the intercalary month was 28 days. This looks like it was Celsus' opinion, not Cato's (or Quintus Mucius'). In principle, it could have been so in 378 day pre-Julian years. But this conflicts with Livy 43.11.13, which records an intercalation inserted on the second day after the Terminalia. It is hard to imagine how there could have been any ambiguity about this: the day after the Terminalia was either Kal Int or it was not. Even if, as it seems to me, almost all intercalations in the last pre-Julian century were 378 day years, the Fasti Ant Mai is conclusive proof that the Intercalary month was still only 27 days at that time. Looks like Celsus got that wrong. But the bissextile day was a contemporary phenomenon.

Now, to my mind it is crystal clear that Mommsen is right and Ideler is wrong. Celsus is not using dates, he is simply describing the sequence of days and using the terms "prior" and "posterior" in their ordinary adjectival sense. Clearly, Ideler was looking to reconcile this text with the later evidence, and his solution is ingenious, but sophistical. I'm not sure if he was aware of CIL VIII 6979, but that inscription does clearly state that ad V Kal Mart was the day after the bissextile (V K Mart qui dies post bis VI K fuit), which makes the bissextile the second half (I managed to confuse myself earlier). One could be equally sophistical and argue that this is a reference to the 48-hour bissextile day and therefore does not tell us the position of the intercalated day within the biduum (and I guess that has been done). But there is no doubt that by the time of Censorinus the bissextile was the first of the two days, and you have excellently quoted the chain of evidence.

Ideler was a great reductionist. He made a similar argument wrt to the Alexandrian calendar, arguing that its leap year was defined to track (and precede) the Roman leap year and therefore was originally a triennial calendar. It is only recently that enough evidence has emerged to conclusively refute this view. We aren't there on the bissextile, and we may never get there since it was a rare date, but I think the balance of evidence is moving against him.

Be that as it may, the relevant question here is how to handle this issue for an encyclopedia article.

I propose the following revision of my earlier suggested text:

The old intercalary month was abolished. The new leap day was originally inserted, like the intercalary month it replaced, after the Terminalia (23 February). It was dated as ante diem bis sextum Kalendas Martias, usually abbreviated as a.d. bis VI Kal. Mart.; hence it is called in English the bissextile day. The year in which it occurred was termed annus bissextus, in English the bissextile year.

The exact position of the leap day may have changed in the first three centuries of the Julian calendar. In the first century AD, a.d. bis VI Kal. Mart, and a.d. VI Kal. Mart. were considered to be two halves of a single 48-hour calendar day. It is debated whether the bissextile day was the first or the second half.[2 -- Celsus+Ideler+Mommsen] An inscription of AD 168, appears to show it as the later day.[3--CIL VIII 6979] By the third century, the bissextile day was certainly the earlier of the two days, and it remained that way henceforward.[3 -- Censorinus etc.] During the late Middle Ages when days in the month came to be numbered in consecutive day order, the leap day was considered to be the last day in February in leap years, i.e. 29 February. The Roman Catholic Church did not make this change till the late twentieth century.

Let me know what you think.

As to AIA's latest text, I'm going to have to revert it again, not least because he not only makes statements about me that are not appropriate for an encyclopedia article, these statement are completely untrue. If he starts to behave, we should be able to bring this to a reasonable conclusion. --Chris Bennett (talk) 18:23, 28 January 2008 (UTC)

The Bissextile and the Regifugium

I'm splitting this off as a separate topic in case 217.169.37.146 starts debating on the Talk page instead of the body of the article. --Chris Bennett (talk) 22:34, 22 January 2008 (UTC)

Why should Augustus have any problem with a market day coinciding with the Regifugium? Such conjunctions will have been plentiful in the past; and both Dio and Macrobius expressly mention market day conjunctions which WERE actively avoided, but the Regifugium is not among them. So perhaps the effect of the 48-hr day was also the aim; to keep the sequence of nundinal days stable. Lepidus appears to have sought the same thing (or at any rate done it) with his triennial intercalation, but Caesar didn't. Appietas (talk) 12:45, 22 January 2008 (UTC)

First, the key point is that, if Celsus' rule is followed then the Pompeian graffiti allow us to infer that the market day fell on ad VI Kal Mart in AD 44, when Dio says there was a change due to a religious festival, a change that happened a few times. The trickery in Celsus' rule allows the physical day to be changed without interrupting the nundinal cycle. Agreed that this is a mathematical theorem, not historical proof, but it fits together remarkably well.

As to why this might have been done, one has to be a little speculative. As you know, there is a well-documented superstition in the first century BC that a market day on Kal Ian was a bad omen, and we know from Dio that in 41 BC the Julian calendar was tampered with to avoid this happening. The phase and alignment of the triennial cycle that meets other contemporary data also has the effect of ensuring that the market day never falls on Kal Ian. After the Augustan reform, it again became possible for this coincidence to happen. But the day afer the Terminalia, the Regifugium, was the first day of the sacred year. So, one can explain the reform as transferring the omen from the civil to the sacred year. Shifting the market day to the second half of the biduum every eight years avoids the physical coincidence of days without interrupting the nundinal cycle. In some support of this, Ammianus Marcellinus notes that the bissextile day was still considered unlucky in AD 364.--Chris Bennett (talk) 16:57, 22 January 2008 (UTC)

I don't understand the mechanics (calendric nor psychological) of shifting a dire omen attested as Kal.Jan. to Regifugium (especially when new year's day of the sacral calendar was Kal.Mart., rather than Regifugium)and find this the weakest part of your otherwise excellent analysis.

I think the mechanics and the psychology of the shift itself are clear, and for the purposes of what was being achieved it doesn't really matter whether the Regifugium was considered to be between years or the first day of the sacred year, so long as the Terminalia marked its end. What is unclear is the mechanics and the psychology by which the community invested in the superstition was persuaded to accept it. But in any case I fully concede there may be other explanations, this is the best I could come up with (and I still think its a good one). After all, you never see any reference to the old superstition after 41 BC. --Chris Bennett (talk) 22:34, 22 January 2008 (UTC)

It occurs to me that Dio's data re 44CE is met if your model was initially applied (in 4CE) but subsequently changed (for Feb 12, 20, 28, 36, 44, etc.) in the nundinal G leap year so that in those years (unlike the nundinal C leap years; 8, 16, 24, 32, 40, etc.) the bissextil was included in the market week sequence (thus changing the letter of the market day), but the extra day subsequently reversed out of the sequence later in the year (perhaps by counting 30-31 December as a single market week day) or in Feb of the following year (by counting Terminalia and Regifugium in the regular year as a single market week day?). Appietas (talk) 20:22, 22 January 2008 (UTC)

Well you could do that, but it seems to me totally unnecessary to resort to such a reconstruction, given that Celsus' rule, applied to every year, has the effect of not changing the nundinal letter of the market day in bissextile years, for both C and G years. Occam. --Chris Bennett (talk) 22:34, 22 January 2008 (UTC)

You don't see any ref. to the "old" Kal.Jan. superstition after 41 BC owing to the meagerness of the sources, but especially because it wasn't really an old superstition at all; it was a novel (and bogus) invention of Lepidus to save face in the Caesarian context, as shown by Dio's account which attributes several qualities to the Kal.Jan. nundinal conjunction which it did not possess even according to Dio's own record (i.e. the nundinal Kal.Jan. of 52). All the more reason to doubt that the community cared about Lepidus' bogus omen, and above all that Imp.Caesar would have treated such a Lepidan invention with any respect, when he treated the man himself (after 35 BC) with loathing and bitter contempt. The importance of Lepidus' action is that in addition to the (bogus) stated purpose of his modification of Caesarian intercalation, it also prevented the occurrence of a much more likely and really ancient conjunction omen which the community truly feared; as did Lepidus himself, while not wishing to publically admit the thing owing to its political nature and Caesar's contempt for it.

Your appeal to signore Occam is circular. IF Augustus transferring calendric Kal.Jan. to the Regifugium, and thenceforth avoiding any nundinal conjunction with Regifugium, was a sufficient act of omen avoidance and assuaged community superstitions and fears, THEN you are quite correct that there is no need to posit any other explanation of the Dio 60.24.7 and Celsus ap.Iustinian 50.16.98. However in my view this is a religious nonsense which wouldn't and couldn't work. I also doubt whether the one-off movement of the market day in Feb within the sextile biduum but no change for the rest of the year would be sufficient to catch Dio's attention, and prefer a version (any version) according to which the market day designation was shifted for most of the year. The beauty of the nundinal G/nonal conjunction thesis is that it seems to accommodate all the evidence very well, and the only major matter not in sources which needs to be assumed is that Macrobius' account of the avoidance of the conjunction of all market days with every nones became (sometime during 5th-3rd centuries BC) focussed alone or especially upon the nundinal G day; probably owing to its convergent position with the nones of the first month of the traditional Kal.Mart. year and calendar. However this is probably not the proper forum for such a debate, and we may as well resume it on the RF list (where it began some months ago), if you have the time and inclination. I'm especially interested in investigating the extent to which avoiding nundinal-G/nonal conjunctions may have contributed to the pattern of intercalation in the period before the introduction of the Kal.Jan.year. It may be a key to determining correct intercalary years until 154, after which it was excluded permanently from "play" so long as the year remained 355 days. I also think that this exclusion was deliberate and that one of the main reasons why the Kal.Jan. year was never changed (a burning issue in itself; rarely or never addressed) was to permanently avoid any convergence of nones and nundinal G, forever and ever. Perhaps because it was a religious imperative that began to interfere with the more "modern" and pressing reasons for the 153 reform (i.e. logistical needs of imperial rule, and desires to keep the calendar year solar aligned and beginning close to the winter solstice). Appietas (talk) 09:19, 23 January 2008 (UTC)

Mark, I agree that we should debate this in one of the more usual places if you really want to pursue this matter. (I don't. I'm not seeing anything new or more convincing in your statements, I've yet to see a comprehensible statement of your "nundinal G" conjecture, and I'm trying to do other things.) This topic is really off track for Wikipedia. This is an article about the Julian calendar. The only relevance this topic has at all is as supporting evidence that the cite from Celsus is an accurate representation of the position of the bissextile in the first century AD, and for that purpose the motivation is irrelevant. --Chris Bennett (talk) 15:50, 23 January 2008 (UTC)

Bissextile Desiderata

1) There is no Roman text in existence with a date of "a.d. VI Kal.Mart. posteriorem". The only text to use the term "posterior" in connection with the bissextile is Celsus, and what he says is quoted above. This supposed "date" does not exist.

2) Ideler's discussion of this point can be read at [16] and [17]. Its in German of course, but it's pretty straightforward. Two points are clear:

(a) He does not use "posterior" as a date either, he just says that the term has to be interpeted as being the day which is furthest from the Kalends of March, rather than in the ordinary sense of "after" or "later".

(b) He gives no justification for this unusual interpretation, he simply says that this is what it means. How does he know this? He doesn't say. I think the only way it could be proved would be to find a text mentioning two other days between the Ides of a month and the following Kalends which describes the earlier date as "posterior" or the later date as "prior". Failing that, the only reason I can see not to accept the ordinary interpretation of the word is that it conflicts with the later position of the bissextile day. That is circular reasoning.

Coincidentally, I just ran into a casual mention of the bissextum as being 25 February in the first century AD, with no indication that the point is debated, in an article written in 1997. Nevertheless, I still believe it is true that there is no scholarly consensus on the question. However, I think this reflects lack of interest in a small and unimportant issue, and a desire to avoid wasting time on small unexplained problems about which there is very little evidence. "Dispute" is probably too strong a word.

3) The feast day of St Mathias is not evidence for the position of the bissextile in the first three centuries AD. There is no reason to believe that this feast existed at that time, and very good reason to believe it didn't. Here is what the Catholic Encyclopedia has to say in its article on Ecclesiastical Feasts at [18]:

Prototypes and starting-points for the oldest ecclesiastical feasts are the Jewish solemnities of Easter and Pentecost. Together with the weekly Lord's Day, they remained the only universal Christian feasts down to the third century (Tertullian, "De Bapt." 19: Origen, "Contra Celsum", VIII, 22). Two feasts of Our Lord (Epiphany, Christmas) were added in the fourth century; then came the feasts of the Apostles and martyrs, in particular provinces.

Whether or not the church calendar is a "continuous record", the fact of the matter is that it does not go back far enough to be relevant.

Further to this: I have found an extensive discussion of this whole issue in a paper by Sternkopf (JCP 41 (1895) 718-733) available at [19]. Sternkopf's conclusion is that the term "bis sextum" originally referred to the biduum of Celsus, i.e. that it didn't actually identify a specific intercalary day, and the point was deliberately left vague. In this way, February always had 28 days.

He thinks that there were only a couple of issues on which it mattered which day was intercalary. The one on which we have direct evidence was legal. Another text in the Digest (IV.4.3 -- [20]) shows that Celsus' need to distinguish an intercalary day arose from a very specific issue -- if a man turned 25 on a.d. VI Kal Mart, which day represented the moment at which he attained his majority? He regards Ideler's argument as specious (as do I). He also makes a circumstantial argument, based on pre-Julian intercalary practice, that the Regifugium was celebrated on the second of the two days in a leap year (I am not convinced by this, I think it conflicts with the notion that February always had 28 days). He does not consider the effect of the biduum on the nundinal cycle.

His view is that the reduction of the 48-hour day to two distinct days in ordinary use was fairly recent in Censorinus' time, noting that he says that the intercalary day is "now" called the bis sextum (quod nunc bis sextum vocatur).

I don't agree with all of this but the basic concept makes a lot of sense to me. --Chris Bennett (talk) 02:57, 3 February 2008 (UTC)

28-day intercalary month

On the other matter: The "28-day intercalary month" should be debated in the Roman calendar article, not here. But I can't help noting that, when Mr Anonymous first commented on Celsus, he specifically pointed out that Celsus' 28 days was an error for 27. I agreed -- and ever since he has insisted Celsus was correct after all. He gave no reason for changing his mind, but it is completely clear that he has no knowledge of the literature beyond what is accessible through Wikipedia. Apparently he took this position because I didn't. --Chris Bennett (talk) 01:41, 2 February 2008 (UTC)

Note to the Nameless One: All further communications to you on the 28-day intercalary month takes place on the Talk:Roman calendar page, where it belongs. --Chris Bennett (talk) 16:07, 6 February 2008 (UTC)