"Happy he who has passed his whole life mid his own fields, he of whose birth and old age the same house is witness....For him the recurring seasons, not the consuls, mark the year; he knows autumn by his fruits and spring by her flowers."

Claudian

Attributed to Romulus, himself, the Roman calendar originally was determined by the cycles of the moon and the seasons of the agricultural year. The calendar then was nominally ten months long (304 days), beginning in March in the spring and ending in December with the autumn planting. Indeed, one still can recognize the remnants of this early calendar in the numbered names for Quinctilis (July), Sextilis (August), September, October, November, and December. Six months had thirty days and four had thirty-one. The winter months were not counted because there was no agricultural work then.

According to Livy (I.19), it was Numa Pompilius, the second king of Rome (715-673 BC), who divided the year into twelve lunar months. Fifty days were added to the calendar and a day taken from each month of thirty days to provide for two new months of twenty-eight days: Januarius (January) at the beginning of the year and Februarius (February) at the end. This was a year of 354 days but, because of the Roman superstition about even numbers, an additional day was added to January to make the calendar 355 days long. Auspiciously, each month had an odd number of days: Martius (March), Maius (May), Quinctilis, and October continued to have thirty-one; the other months, twenty-nine, except for February, which had twenty-eight and was devoted to rites of purification (februa) and expiation appropriate to the last month of the year.

Although these legendary beginnings attest to the venerability of the lunisolar calendar of the Roman Republic, its historical origin probably is the publication of a revised calendar by the Decemviri c.450 BC as part of the Twelve Tables, Rome's first code of law. It is likely, too, that the lunar year had been abandoned by then. Cicero mentions a solar eclipse occurring on the Nones of June (June 5). Such an eclipse did take place in 400 BC, but it only can have occurred at the new moon, when it is in conjuction with the sun, and that would have been shortly before the Kalends, when the first day of the month was announced. In 153 BC, the new year was moved from the Ides of March to the Kalends of January (January 1), which marked the beginning of the civil year and newly elected consuls assumed office. (Several years before, Cato, writing about farming, continued to reckon time by the stars and used the civil calendar only for business contracts.)

Twelve lunar months, however, are not as long as the solar year, and the calendar had long since ceased to agree with the lunations of the moon. Nor did the college of pontiffs (from pontifex or "bridge maker"), who were responsible for regulating the calendar and the festivals that depended upon it, always intercalate the additional days necessary to sychronize the two. Intercalation was considered unlucky and, during the Second Punic War (218-201 BC), the priests were hesitant to make any changes at all. Often, too, the calendar was manipulated more for partisan political consideration than to adjust for the seasons. Legislation could be given more or less time, the tenure of an office holder extended or reduced.

The chief priest had to be in Rome during February to declare whether there would be an intercalation that year. But, although Julius Caesar had been elected pontifex maximus in 63 BC, he was preoccupied in Gaul and then with civil war, and there had been only one intercalation since he became proconsul. (In 50 BC, Cicero complains on February 13 that he still did not know whether there would be an intercalation later that month.) When Caesar finally returned to Rome from the Egyptian campaign in 46 BC and was declared dictator, a three-month discrepancy existed between the seasons and their calendar date. The harvest was being celebrated long before the crops had been taken in.

That year, a reformed calendar, based on the solar year, was introduced by Caesar, who first had heard about it, relates Lucan, while consorting with Cleopatra. With the advice of the Alexandrian astronomer Sosigenes, the missing ninety days were intercalated, extending the ultimus annus confusionis, "the last year of confusion," as Macrobius called it, to 445 and causing March 1 in the Republican calendar to fall on January 1, 45 BC in the new Julian calendar. To correct for the ten days that are lost in a lunar calendar, one or two additional days were added at the end of those months with twenty-nine days (January, Sextilis, and December would have thirty-one days; April, June, September, and November, thirty).

And, because the solar year is approximately a quarter day longer than the calendar year, a single intercalary day was to be inserted every four years, when February 24 would be counted twice. (This was a leap or bissextile year, since that date was the "twice sixth" day, bissextus, before the Kalends of March). When Caesar was assassinated the next year, Quinctilis, the month in which he had been born, was renamed Julius (July) in his honor, although this change in the name of the month was ignored until made legal after the appearance of a comet four months later, which, recounts Cassius Dio, was understood to mean that Caesar had become immortal and taken his place among the stars.

But the pontiffs mistakeningly adjusted for leap year every three years (having counted inclusively) and inserted too many intercalary days. In 8 BC, Augustus was obliged to omit any further intercalations until AD 8. Only then, after the superfluous days had been corrected and intercalation was resumed, did the Julian calendar function as intended, with February gaining an extra day every four years. In honor of his reform, Sextilis was renamed Augustus (it was that month that Augustus first had been elected consul and Egypt had become part of the Roman empire).

The four-year cycle of the Julian year averaged 365.25 days, but this still does not quite correspond to the solar year. Rather, it was a bit more than eleven minutes too long, gaining a day every 128 years, an accumulation that, by the mid-sixteenth century, amounted to approximately ten days. The vernal equinox, from which the church calculated Easter, no longer was occurring on March 21, its traditional date, but ten days earlier. (The summer soltice, which occurs on June 21, also was being celebrated ten days too soon, on the feast of St. Barnabas, from whence the rhyme "Barnaby bright / The longest day and the shortest night.")

To correct for this retrogression and bring the calendar year back to the solar year, Pope Gregory XIII omitted these extra days, ordaining in 1582 that, for that year, October 4 was to be followed by October 15. And, to prevent the discrepancy in the Julian calendar from reoccurring, three leap years were to be omitted every four centuries. A leap day would not be added in those years that ended in hundreds unless they were divisible by 400; thus, for the first time since 1600, there was a February 29 in the centurial year 2000.

The Gregorian calendar has been in use ever since, although, in practice, the Julian calendar still is used for dates prior to the Gregorian reform, and the Roman calendar for dates BC referring to Roman history.

---

References: The Oxford Companion to the Year (1999) by Bonnie J. Blackburn and Leofranc Holford-Stevens; Marking Time: The Epic Quest to Invent the Perfect Calendar (2000) by Duncan Steel; The Roman Festival Calendar of Numa Pompilius (1986) by Michael York; Festivals and Ceremonies of the Roman Republic (1981) by H. H. Scullard; The Roman Festivals of the Period of the Republic (1899) by W. Warde Fowler; Calendar of the Roman Republic (1967) by Agnes Kirsopp Michels; Chronology of the Ancient World (1980) by E. J. Bickerman; On Roman Time: The Codex-Calendar of 354 and the Rhythms of Urban Life in Late Antiquity (1990) by Michele Renee Salzman; Ovid and the Fasti: An Historical Study (1994) by Geraldine Herbert-Brown; Mapping Time: The Calendar and Its History (1998) by E. G. Richards; Calendar: Humanity's Epic Struggle to Determine a True and Accurate Year (1998) by David Ewing Duncan; The History & Practice of Ancient Astronomy (1998) by James Evans; The Oxford Classical Dictionary (1970) edited by N. G. L. Hammond and H. H. Scullard; A New Topographical Dictionary of Ancient Rome (1992) by L. Richardson, Jr.

Ovid: Fasti (1936) translated by James G. Frazer (Loeb Classical Library); Suetonius: The Twelve Caesars (1957) translated by Robert Graves (Penguin Classics); Livy: The Early History of Rome (1971) translated by Aubrey de Sélincourt (Penguin Classics); Macrobius: The Saturnalia (1969) translated by Percival Vaughan Davies; Dio's Roman History (1916) translated by Earnest Cary (Loeb Classical Library); Petronius: The Satyricon and the Fragments (1965) translated by John Sullivan; Claudian (1922) translated by M. Platnauer (Loeb Classical Library); Statius: Silvae, Thebaid (1928) translated by J. H. Mozley (Loeb Classical Library); Pliny the Younger: Letters and Panegyricus (1969) translated by Betty Radice (Loeb Classical Library); Censorinus: De Die Natale (1900) translated by William Maude.