The name says seventh and the calendar says ninth. September is the ninth month of the modern year, yet septem is Latin for seven, and the same mismatch runs through October, November and December. It is not a translation error or a myth; it is a straightforward record of the fact that the Roman year once began in March, and the numbering simply outlived the arrangement that made it correct.
This note explains the ten-month Roman year, where January and February came from, and why nobody ever bothered to renumber the tail end of the calendar. For the practical version — which number goes with which name today — the month number page maps them both ways.
Why Is September the Ninth Month?
Because two months were added to the beginning of the year and the later names were never changed. September was the seventh month when the year opened with March. Adding Januarius and Februarius at the front pushed it, and the three months after it, two places later.
Everything else follows from that single shift. October, whose name means eighth, became the tenth. November, meaning ninth, became the eleventh. December, meaning tenth, became the twelfth. Four names, all off by exactly two, all preserved unchanged through two major calendar reforms.
The Ten-Month Roman Year
The earliest Roman calendar, traditionally attributed to Romulus, is described as having ten months and around 304 days, running from Martius through to December. Winter was left largely uncounted: a stretch of roughly sixty days that fell outside the named months entirely, because it mattered little to farming or campaigning.
In that year Martius came first, named for Mars, and the sequence ran on through Aprilis, Maius and Junius before switching to plain numbers. Quintilis was the fifth, Sextilis the sixth, then September, October, November and December as the seventh to tenth months. The numbering was perfectly consistent — it described exactly where each month stood. The origins of the named months are covered in where month names come from.
Where January and February Came From
Tradition credits Numa Pompilius, Rome's second king, with adding Januarius and Februarius to cover the unnamed winter and bring the year to twelve months. Januarius honoured Janus, god of beginnings and doorways; Februarius took its name from februa, the purification rites held in it.
The two new months went to the end of the year at first, after December, which kept the existing numbering intact and left September as the seventh month in name and in fact. Only later did they settle at the front, and that single rearrangement is why September is the ninth month today. When they did, the numbered names became wrong, and September was suddenly the ninth month with a name that still said seventh. February's position as the last month of the older arrangement is also why it ended up as the short one that absorbs the leap day, as set out in February and leap years.
When Did the Year Start Moving to January?
The Roman civil year moved to 1 January in 153 BC, when the consuls began taking office on that date rather than in March. The change was administrative rather than astronomical, and it made January the official start of the year for dating purposes.
The Julian reform under Julius Caesar in 46 BC confirmed the arrangement and fixed a much larger problem: the old Republican calendar of 355 days needed an intercalary month inserted by the priests to stay aligned with the seasons, and they had let it drift badly. Caesar replaced it with a 365-day year and a leap day every fourth year. Quintilis was renamed Julius in his honour after his death, and Sextilis became Augustus for the first emperor. Both renamings show the calendar could be changed when someone wanted it changed.
Why the Numbering Was Never Fixed
Two reforms had the authority to correct it and neither did. The Julian reform of 46 BC rebuilt the length of the year without touching the four numbered names, and the Gregorian reform of 1582 adjusted the leap rule and dropped ten days without renaming anything at all.
The reasons are mundane. Renaming a month breaks every contract, record and religious calendar that refers to it, and the mismatch causes no practical harm — nobody calculates a date from what September's name means. Here is what the four stranded names actually say:
- September: from septem, seven. Now the ninth month, with 30 days.
- October: from octo, eight. Now the tenth month, with 31 days.
- November: from novem, nine. Now the eleventh month, with 30 days.
- December: from decem, ten. Now the twelfth month, with 31 days.
Each one is off by two, and the gap of two is the exact size of the addition made at the front of the year. The mismatch is therefore evidence, not a mistake: it is the clearest surviving trace of the ten-month year.
It is also self-consistent, which is a useful check on the story. If September were the ninth month by accident, or through a mistranslation, the errors would be random. Instead all four are wrong in the same direction by the same amount, which is exactly what inserting two months at the start of a year would produce and what no other explanation would.
Does It Still Cause Confusion?
Rarely in speech, occasionally in software. Nobody misreads a date because September means seventh, but the residue of the March-first year does surface in systems that number months from zero, and in fiscal calendars that still refuse to start in January.
Several national accounting years keep an old-fashioned start date for reasons that are themselves historical — the UK tax year beginning on 6 April is a direct descendant of a pre-Gregorian calendar adjustment. Quarters bring the same issue: the third calendar quarter contains September, but a fiscal Q3 might not, which is why quarter labels always need a stated year type. That distinction is unpicked in fiscal quarters explained, and the quarter page shows which quarter a date sits in. For the current month's name and number together, use this month.
Conclusion
September is the ninth month because the Roman year once began in March, making it genuinely the seventh, and because Januarius and Februarius were later added at the front without anyone renumbering what followed. October, November and December carry the same two-place offset. Neither Julius Caesar nor Pope Gregory XIII fixed it, since the cost of renaming outweighed the tidiness gained. The result is a calendar that quietly documents its own history. To check any month's name against its number, open the month number page, or explore the rest of months.now.