# 01 — The Julian calendar (the baseline) *Instituted by Julius Caesar, effective 1 January 45 BCE (reform enacted 46 BCE). This is the baseline against which every later exception in this collection is measured.* ## What changed On the advice of the Alexandrian astronomer Sosigenes, Caesar abandoned the abusable lunisolar scheme (document 00) for a **purely solar** calendar with a **fixed, mechanical leap rule**: - A common year of **365 days** in 12 months of the lengths we still use. - One **leap day every fourth year**, giving an average year of **365.25 days**. - No more discretionary intercalation by priests — the rule is arithmetic and cannot be gamed. To reset the seasons before the new system began, **46 BCE was stretched to 445 days** (the "year of confusion"), after which the regular pattern took over. ## The leap rule ``` A Julian year is a leap year ⇔ year mod 4 == 0. Average length = 365.25 days. ``` Simple and durable — but **too long**. The true mean tropical year is about 365.2422 days, so the Julian year overshoots by roughly **0.0078 days per year**, or about **1 day every 128 years**. Over centuries the calendar date of the vernal equinox slips *earlier*. By the 16th century the equinox had drifted about **10 days** from where it sat at the Council of Nicaea (325 CE) — the error that the Gregorian reform (document 02) exists to correct. ## Why it is the baseline Every item in this collection is expressed as a **deviation from Julian**: - The **Gregorian reform** (02) keeps the Julian frame and changes only the leap rule and the starting offset — it is a *correction to Julian*, not a new calendar. - The **national adoption dates** (03–05) are the moments each polity switched *off* Julian and *onto* Gregorian; the size of the day-skip at each switch is a direct function of how long Julian error had accumulated by that date (10 days in 1582, 11 by 1752, 13 by 1918). - The **proleptic Julian calendar** (projecting the rule backward before 45 BCE) is the conventional frame historians use to place ancient dates, which is why the model in `src/` can reckon in Julian as well as Gregorian. ## Concepts that carry forward - **A fixed, non-discretionary leap rule.** The core idea — the calendar is an arithmetic function of the year number, not a political decision. - **The 365/366-day solar year and the 12-month structure.** Unchanged by every later reform. - **An accumulating, measurable error.** Because the error is exactly quantifiable (≈1 day / 128 years), the later reforms can correct it precisely rather than by guesswork. ## The Roman line as a dated, weighed record (document 08) This document isolates the Julian reform as *the baseline*. For the full Roman lineage it sits inside — from the legendary pre-foundation era (1244 BC) through Romulus's 10 months, Numa's 12, the pontifical-intercalation breakdown, the 45 BC Julian reform, the 8 BC Augustan fix, the Dionysian AD era, and on to the Gregorian correction and today's proleptic Gregorian / UTC / ISO 8601 — with each event noted, its concession recorded, and a 1–10 weighing — see [`08-roman-calendar-1244bc-to-present.md`](08-roman-calendar-1244bc-to-present.md). Its headline weighings: the Julian reform (45 BC) and the Gregorian reform (1582) both carry weight 10, the Dionysian AD era (525 CE) carries 9.