top of page

Why September, the 'Seventh' Month, Comes Ninth

Writer: Feed the Reader
Feed the Reader
2 days ago
4 min read

Take a look at a calendar. Any calendar. Flick through to the back end of the year.


September. Septem is Latin for seven. October: octo, eight. November: novem, nine. December: decem, ten.


Now count. September is the ninth month. October is the tenth. November is the eleventh and December is the twelfth. Every one of them is two places out.


This isn't a hidden error tucked away in some obscure document. It's printed on every wall calendar, every diary and every phone, and it has been sitting there in plain sight for more than two thousand years. Most of us never notice.


A year that started in spring

To see how the gap opened, go back to the earliest Roman calendar, at least as later Roman writers described it.


According to tradition, the early Roman year had just ten months. It started in March and ran through to December, about 304 days in all. And the winter? The winter simply wasn't counted. Those dead weeks between December and the next March fell outside the named months altogether.


That sounds bizarre until you think about who the calendar was for. This was a farming year. Spring is when things start: ploughing, sowing, the whole working year. Midwinter, when not much happens in the fields, didn't need careful numbering. Think of a school year, which starts when school starts and not on 1 January. The Roman year started when the work did.


In a year that begins in March, the numbers line up perfectly. March first, April second, and so on, until September comes in seventh and December tenth. The names were right. For a while.


Two months squeezed in at the front

Then two new months arrived: January and February. Roman legend credits them to Numa, the second king of Rome, who is said to have added them to cover the uncounted winter.


On its own, that might not have done any damage. January and February could have sat at the end of the year, after December, and everything would still have added up.


The trouble came with the start of the year. In 153 BCE, the year the Romans used for official business, which began when the two consuls (Rome's top elected officials) took office, moved its starting point to 1 January.


And there it is. The year's starting line jumped back two months, the numbered names stayed exactly where they were, and every one of them slipped two places. September, the seventh, became the ninth. Nobody renamed anything. Why would you? Everyone already knew what September was.


Two months that got lucky

There used to be more numbered months, as it happens. The fifth month was Quintilis, from quinque, five. The sixth was Sextilis, from sex, six.


If they'd kept their names, they'd be out of place too. But they didn't. Quintilis was renamed July, in honour of Julius Caesar, and Sextilis became August, after the emperor Augustus.


So the renaming fixed two of the wrong numbers, more or less by accident. It wasn't done to tidy up the arithmetic. It was done to flatter two powerful men. But the effect was that the only numbered months left in the calendar are the four that are now out of place.


The Romans weren't too fussed

If this feels like sloppy bookkeeping, it's worth knowing that it wasn't the worst of the Romans' calendar problems.


The Roman calendar of the Republic had to be kept in line with the seasons by intercalation, which just means slipping in extra days or an extra month now and then. That job was done badly, and sometimes not at all. The calendar drifted so far that, according to the historian Livy, a solar eclipse in 190 BCE was recorded under a date in July, months away from when it actually happened.


By the time Julius Caesar overhauled the calendar, it had drifted so badly that one year was stretched to 445 days to drag it back into line. It was later nicknamed the Year of Confusion, which seems fair.


So the Romans were perfectly capable of letting both the numbers and the seasons slip. Next to a whole calendar wandering off, a few months with the wrong names barely registered.


How sure are we?

Now for an honest caveat. The ten-month year, the 304 days and King Numa's additions all come from Roman writers who lived centuries after the events they describe. Historians treat parts of that story as legend, a tidy origin tale rather than a documented reform.


The names themselves, though, are solid evidence. Whatever the precise history, the months that mean seven, eight, nine and ten only make sense in a year that once began two months later than ours. The words are the clue, even if the founding story is a little shaky.


A calendar written over

There's a lovely word for a document that's been scraped clean and written over, where traces of the older writing still show through underneath: a palimpsest.


That's what a calendar is. Every time a reform came along, the Romans wrote over the old year. They added months, moved the start and renamed a couple to honour emperors. But they never fully erased what was there before. September is one of the places where the older text still shows through.


Words are like that. They tend to outlast the systems they were made for. The calendar moved on two thousand years ago. The names never got the memo.


Sources

  1. University of Chicago (penelope.uchicago.edu), Encyclopaedia Romana, "Roman Calendar". https://penelope.uchicago.edu/encyclopaedia_romana/calendar/romancalendar.html (accessed 28 Oct 2026)

  2. timeanddate.com, "September: Ninth Month of the Year". https://www.timeanddate.com/calendar/months/september.html (accessed 28 Oct 2026)

  3. British Museum blog, "What's in a name? Months of the year". https://www.britishmuseum.org/blog/whats-name-months-year (accessed 28 Oct 2026)

  4. Wikipedia, "September (Roman month)". https://en.wikipedia.org/wiki/September_(Roman_month) (accessed 28 Oct 2026)

Featured

Get Feed the Reader's latest articles straight to your inbox. Sign up for our weekly newsletter.

© 2026 Feed the Reader.

bottom of page