The Online Encyclopedia and Dictionary






Roman calendar

The Roman calendar changed its form several times in the time between the foundation of Rome and the fall of the Roman Empire. This article generally discusses the early Roman or 'pre-Julian' calendars. The calendar used after 46 BC is discussed under the Julian calendar.


History of the Calendar

To begin with it was a lunar calendar containing ten months, starting at the vernal equinox, traditionally invented by Romulus, the founder of Rome about 753 BC. However it seems to have been based on the Greek lunar calendar. The months at this time were

  • Martius (31 days)
  • Aprilis (30 days)
  • Maius (31 days)
  • Junius (30 days)
  • Quintilis (31 days)
  • Sextilis (30 days)
  • September (30 days)
  • October (31 days)
  • November (30 days) and
  • December (30 days)

Thus the calendar year lasted 304 days and there were about 61 days of winter that did not fall within the calendar.

The first reform of the calendar was attributed to Numa Pompilius, the second of the seven traditional Kings of Rome. He is said to have reduced the 30-day months to 29 days and to have added January (29 days) and February (28 days) to the end of the calendar around 713 BC, and thus brought the length of the calendar year up to 355 days:

  • Martius (31 days)
  • Aprilis (29 days)
  • Maius (31 days)
  • Iunius (29 days)
  • Quintilis (31 days)
  • Sextilis (29 days)
  • September (29 days)
  • October (31 days)
  • November (29 days)
  • December (29 days)
  • Ianuarius (29 days)
  • Februarius (28 days)

In order to keep the calendar year roughly aligned with the solar year, a leap month of 27 days, the Mensis Intercalaris, sometimes also known as Mercedonius or Mercedinus, was added from time to time at the end of February, which was shortened to 23 or 24 days. The resulting year was either 377 or 378 days long. The decision to insert the intercalary month, and its placement, was the responsibility of the pontifex maximus On average, this happened roughly in alternate years.

The system of aligning the year through intercalary months broke down at least twice. The first time was during and after the Second Punic War. It led to the reform of the Lex Acilia in 191 BC. The details of this reform are unclear, but it appears to have successfully regulated intercalation for over a century. The second breakdown was in the middle of the first century BC. This time, reform resulted in the creation of the Julian calendar instituted by Julius Caesar.


The Romans had special names for 3 specific days in each month. The system was originally based on phases of the Moon (Luna), and these days were probably declared when the lunar conditions were right. After the reforms of Numa Pompilius, they occurred on fixed days.

  • Kalends - first day of the month, from which the word "calendar" is derived. Interest on debt was due on Kalends.
  • Nones – depending on the month, could be the 5th or the 7th day; traditionally the day of the Half Moon
  • Ides – depending on the month, could be the 13th or the 15th day; traditionally the day of the Full Moon
Months with Ides and Nones occurring on the 13th/5th day: January, February, April, June, August, September, November, December
Months with Ides and Nones occurring on the 15th/7th day: March, May, July, October --
a mnemonic:
In March, July, October, May
The IDES fall on the 15th day
The NONES the 7th.
The rest besides take 2 days less
For Nones and Ides.

Days were numbered in a way that is quite different from the modern Western calendar. The Romans did not count the days of the month retrospectively, looking back to the first of the month (that is: 1st, 2nd day since the start of the month, 3rd day since the start of the month). They counted forward to their named days. Also, to the distress of moderns trying to work out dates in Roman calendar documents, they counted inclusively, so that September 2 is considered 4 days before September 5, rather than 3 days before.

The example of September

The following example spells out how days were named for the pre-Julian September, which had only 29 days. It shows the Roman form of the date, the translation, and how we would say it today. The Romans used abbreviations: "a.d." = "ante diem" = "day before", "prid." = "pridie" = "the day before", "Kal" = "Kalends" etc.

Kal. Sept. = Kalends of September = September 1
a.d. IV Non. Sept. = 4 days before the Nones of September = September 2
a.d. III Non. Sept. = 3 days before the Nones of September = September 3
prid. Non. Sept. = the day before the Nones of September = September 4
Non. Sept. = Nones of September = September 5
a.d. VIII Id. Sept. = 8 days before the Ides of September = September 6
a.d. VII Id. Sept. = 7 days before the Ides of September = September 7 and so on till
a.d. III Id. Sept. = 3 days before the Ides of September = September 11
prid. Id. Sept. = the day before the Ides of September = September 12
Id. Sept. = Ides of September = September 13
a.d. XVII Kal. Oct. = 17 days before the Kalends of October = September 14
a.d. XVI Kal. Oct. = 16 days before the Kalends of October = September 15 and so on till
a.d. III Kal. Oct. = 3 days before the Kalends of October = September 28
prid. Kal. Oct. = the day before the Kalends of October = September 29
Kal. Oct. = Kalends of October = October 1

Notice that by counting inclusively and by having a special name for the day before a named day the Roman calendar loses the possibility of saying: 2 days before a named day. Also, after the Ides, the date no longer mentions September, but is counting down towards October.

When Julius Caesar added a day to September, he didn't add it to the end of the month. Rather, the new day that got added was the day after the Ides:

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

As a result, the position of all the following dates in September got bumped up by one day. This has some unexpected effects for modern readers. For example, the emperor Augustus was born on 23 September 63 BC. In the pre-Julian calendar this is 8 days before the Kalends of October (or, in Roman style, a.d. VIII Kal. Oct.), but in the Julian calendar it is 9 days: a.d. IX Kal. Oct. Because of this change, in some parts of the Empire his birthday was celebrated on both dates, i.e. (for us) on 23 and 24 September.

Days of the week

The Roman Republic, like the Etruscans, used a "market week" of eight days, marked as A to H in the calendar. A market was held on the eighth day. For the Romans, who counted inclusively, this was every ninth day, hence the market became called "nundinae". Since the length of the year was not a multiple of 8 days, the letter for the market day (known as a "nundinal letter") changed every year. For example, if the current letter for market days was A and the year was 355 days long, then the letter for the next year would be F.

The market cycle was a fundamental rhythm of daily life, and the market day was the day that country people would come to the city. For this reason, a law was passed in 287 BC (the Lex Hortensia) that forbade the holding of meetings of the comitia (for example to hold elections) on market days, but permitted the holding of legal actions. In the late republic, a superstition arose that it was unlucky to start the year with a market day (i.e. for the market day to fall on 1 January, with a letter A), and the pontiffs, who regulated the calendar, took steps to avoid it.

Because the market cycle was absolutely fixed at 8 days under the Republic, information about the dates of market days is one of the most important tools we have for working out the Julian equivalent of a Roman date in the pre-Julian calendar. In the early Empire, the Roman market day was occasionally changed. The details of this are not clear, but one likely explanation is that it would be moved by one day if it fell on the same day as the festival of Regifugium, an event that could occur every other Julian leap year. When this happened the market day would be moved to the next day, which was the bissextile (leap) day.

The modern seven-day week came into use during the early imperial period, after the Julian calendar came into effect, apparently stimulated by immigration from the Roman East. For a while it coexisted alongside the old 8-day nundinal cycle, and fasti are known which show both cycles. It was finally given official status by Constantine in 321.

The days of the week were dedicated to gods, the sun or the moon. They were (note the similarities of some of the days with French and Spanish and other Romance languages):

  • Sunday – Dies Solis (day of the sun)
  • Monday – Dies Lunae (day of the moon)
  • Tuesday – Dies Martis (day of Mars)
  • Wednesday – Dies Mercuri (day of Mercury)
  • Thursday – Dies Iovis (day of Jupiter)
  • Friday – Dies Veneris (day of Venus)
  • Saturday – Dies Saturni (day of Saturn)

Character of the Day

An aspect of the Roman calendar that is quite unfamiliar to us is that each day had a "character", which was marked in the fasti. The most important of these were dies fasti, marked by an F, on which legal matters could normally be heard, dies nefasti, marked by an N, on which they could not, and dies comitiales, marked by a C, on which meetings of the public assemblies known as comitia were permitted, subject to other constraints such as the Lex Hortensia. A few days had a different character, e.g. EN (endotercissus or perhaps endoitio exitio nefas), a day in which legal actions were permitted on half of the day only, and NP, which were public holidays.


In the Roman Republic, the years were not counted. Instead they were named after the consuls who were in power at the beginning of the year (see List of Republican Roman Consuls). For example, 205 BC was The year of the consulship of Publius Cornelius Scipio Africanus and Publius Licinius Crassus.

However, in the later Republic, historians and scholars began to count years from the founding of the city of Rome. Different scholars used different dates for this event. The date most widely used today is that calculated by Varro, 753 BC, but other systems varied by up to several decades. Dates given by this method are numbered ab urbe condita (meaning after the founding of the city, and abbreviated AUC). When reading ancient works using AUC dates, care must be taken to determine the epoch used by the author before translating the date into a Julian year.

The first day of the consular term, which was effectively the first day of the year, changed several times during Roman history. It became 1 January in 153 BC. Before then it was 15 March. Earlier changes are a little less certain. There is good reason to believe it was 1 May for most of the third century BC, till 222 BC. Livy mentions consulates starting on 1 July before then, and arguments exist for other dates at earlier times.

Converting Pre-Julian Dates

Finding the Julian equivalent of a Roman date can be quite tricky. Even early Julian dates, before the leap year cycle was stabilised, are not quite what they appear to be. For example, it is well known that Julius Caesar was assassinated on the Ides of March in 44 BC, and this is usually converted to 15 March 44 BC. While he was indeed assassinated on the 15th day of the Roman month Martius, the equivalent date on the modern Julian calendar is probably 14 March 44 BC.

Finding the exact Julian equivalent of a pre-Julian date is much harder. Since we have an essentially complete list of the consuls, it is not difficult to find the Julian year that generally corresponds to a pre-Julian year. However, our sources very rarely tell us which years were regular, which were intercalary, and how long an intercalary year was. For this reason, pre-Julian dates can be very misleading.

We do have a number of clues to help us. First, we know when the Julian calendar began, although there is some argument about it. We have detailed sources for the previous decade or so, mostly in the letters and speeches of Cicero. Combining these with what we know about how the calendar worked, especially the nundinal cycle, we can accurately convert Roman dates after 58 BC. Also, the histories of Livy give us exact Roman dates for two eclipses in 190 BC and 168 BC, and we have a few loose synchronisms to dates in other calendars which help to give rough (and sometimes exact) solutions for the intervening period.

Before 190 BC the alignment between the Roman and Julian years is determined by clues such as the dates of harvests mentioned in the sources. This allows us to estimate approximate Julian equivalents of Roman dates back to the start of the First Punic War in 264 BC. However, the number of years before 45 BC for which we can accurately and precisely convert Roman dates to Julian dates is very small.

See also


  • Plutarch - Numa Pompilius
  • Ovid - Fasti
  • Bickerman, E.J. Chronology of the Ancient World. London: Thames & Hudson, 1969, rev. ed. 1980.
  • Michels, A. K. The Calendar of the Roman Republic Princeton, 1967

External links

Last updated: 10-15-2005 23:28:58
The contents of this article are licensed from under the GNU Free Documentation License. How to see transparent copy