The agreeable calendar

The Indian national calendar: five months of 31, six of 30, a leap day borrowed wholesale from the Gregorian — a calendar that cannot drift, and the sky it agreed to give up. ← Calendar Contraptions

Piece after piece in this corner is about a calendar that computes: leap rules, epacts, postponements, an equinox interviewed once a year. The Indian national calendar — the Śaka calendar, named for the era it counts in — is the opposite bet. Adopted in 1957, it does not compute anything. Five months get 31 days, six months get 30, and the first month, Chaitra, gets 30 days starting 22 March — except in the years the Gregorian calls leap, when it gets 31 days starting 21 March. That is the entire machine. Every month other than Chaitra starts on the same Gregorian date in every year, forever; the leap day is borrowed from the Gregorian wholesale; and the whole calendar fits in one table you could paint on a wall.

This makes it the counterweight to the previous piece here (the 33-year spring), and the comparison is a real one — the standard references make it themselves. The Solar Hijri asks the sky every year and accepts the exceptions when they come. India's committee asked the sky once, wrote the answers into a table, and never asked again. What you buy with that is a calendar that cannot break. What you pay is the subject of the second half of this page.

The Śaka date

The weekday

The year

The borrowed day

Chaitra is the adjustable spanner. The Gregorian's 29 February falls inside Phalguna, the month before the Śaka year begins — so in a leap year Chaitra starts a day earlier (21 March) and runs a day longer (31 days), and by 1 Vaisakha the two year-types have converged again. The borrowed day never leaves the building. The other eleven months start on the same Gregorian date every single year, which is why the table below never changes — and that is the design, not laziness: the first half of the year gets the 31-day months to take into account the slower movement of the sun across the ecliptic in that half, the same logic that fattens the Gregorian's summer months and the same logic the Iranian calendar encodes the hard way (see the 33-year spring).

#DaysStartsMonthDevanagariSign of the month

The weekday you already had

The seven weekdays are the same seven the Roman week spread worldwide, carried by the same planets through the navagraha: Ravivāra, Sunday, belongs to the Sun; Somavāra to the Moon; Maṅgalavāra to Mars; Budhavāra to Mercury; Bṛhaspativāra — also written Guruvara — to Jupiter; Śukravāra to Venus; Śanivāra to Saturn. The official reckoning starts the week with Ravivāra (also called Adivara) and ends it with Śanivāra, exactly Sunday-through-Saturday.

The sky it gave up

The era itself is anchored to the sky in exactly one place: its year zero is the vernal equinox of AD 78. And the calendar "begins on or near the March equinox", as the design puts it — with near doing quiet work. In 1957, the calendar's first year, the equinox fell on 21 March in Indian time and 1 Chaitra was the day after. By 2026 the equinox had slid to 20 March and the calendar started two days after it. Nothing is wrong — the start date is pinned by design, and pinned dates do not follow a moving sky. But the founding event and the anniversary have quietly drifted apart, and they will keep drifting until a skipped Gregorian leap day pulls the equinox back.

The March equinox

The calendar's answer

Why not fix it? Because the whole point of the design is that no observation is ever needed and conversion is trivial arithmetic. Pin the start to the Gregorian grid and the calendar gains permanence and loses the sky — through 2100 the computed gap stays at one or two days (one day in 15 of the 75 years from 2026 to 2100, two days in the other 60). The committee knew exactly what it was trading away.

Thirty calendars, one committee

In 1952 the astrophysicist Meghnad Saha — the man who explained the spectra of stars — assembled a Calendar Reform Committee under the Council of Scientific and Industrial Research. Its task was "to prepare an accurate calendar based on scientific study, which could be adopted uniformly throughout India", and its first finding was the size of the problem: thirty different calendars in use across the country, each tangled with religion and local sentiment. In 1954 the committee recommended one fixed tropical solar calendar for civil use — this calendar — along with a tropical lunisolar companion for religious purposes, which was not accepted. India's first prime minister, Jawaharlal Nehru, wrote in the preface to the committee's 1955 report: "They (different calendars) represent past political divisions in the country ... . Now that we have attained Independence, it is obviously desirable that there should be a certain uniformity in the calendar for our civic, social, and other purposes, and this should be done on a scientific approach to this problem."

Usage began on 1 Chaitra 1879 Śaka — 22 March 1957. The Gazette of India dates itself in both calendars, the news bulletins of All India Radio read the Śaka date, and official communications follow. But the reform never crossed into daily life: the traditional almanac-makers kept their own calendars, the month names were borrowed from the older lunisolar calendar — "so variations in spellings exist, and there is a possible source of confusion as to what calendar a date belongs to" — and today the national calendar is largely what its designers made it: a civic instrument. The republic's own dates look tidy in it:

GregorianWhatŚaka dateWeekday

The era of the Shakas

The year number counts the Śaka Samvat: year 0 is AD 78, and the years are elapsed ones — counted the way you count your own age, so the uncounted first year is year zero. What happened in 78 is genuinely open. Scholarly use separates an "Old Śaka Era", visible in some Buddhist and Jaina inscriptions and probably a millennium or more older, whose epoch is disputed — from the "Śaka Era of 78 CE" of southern inscriptions, which is the one India adopted. The modern reading ties 78 to the accession of the Indo-Scythian king Chashtana: his inscriptions at Andhau in Kutch are dated years 11 and 52, read as Śaka years 11 (89 CE) and 52 (130 CE). An older story made the epoch the accession of the Kushan emperor Kanishka — a story that has collapsed, since Kanishka now ascends around 127 CE and was no Shaka anyway.

The name most people in India actually know is a legend: the Shalivahana era, after the king who defeated the Shakas in 78 CE — a tale already told by Brahmagupta in the seventh century and by Al-Biruni nine hundred years later, and judged by the epigrapher Richard Salomon "an obvious fabrication". The historian D. C. Sircar traced it to Gautamiputra Satakarni's real victories over Shaka rulers and suspected the south built a matching legend for the north's Vikramaditya. What is not legend: the earliest dated users are the Western Satraps, Shaka rulers of Ujjain, whose coins carry the era in Brahmi numerals from the reign of Rudrasimha I (178–197); a coin of Damasena reads year 153 — 231 CE. The era was widespread by the sixth and seventh centuries — Varāhamihira and Brahmagupta both use it — appears in Southeast Asian inscriptions by the seventh, kept Javanese courts until 1633, and still runs the Balinese Saka calendar today. The Government of India calls it the Shalivahana era; the other great Indian era, Vikram Samvat, begins in 56 BCE and runs 135 years ahead.

Convert the other way

Gregorian date

Where it cheats

Almost nowhere in its own terms — that is what agreeing buys. The honest seams: the Śaka date is civic, not festive; the festivals people actually keep run on the regional lunisolar calendars, and the borrowed month names do not line up with the lunar months they came from. The month-to-zodiac pairing is the design's intent — Chaitra carries Meṣa (Aries), Phalguna carries Mīna (Pisces) — following the tropical zodiac rather than the sidereal one the Hindu and Buddhist calendars use, which is another way of saying this calendar faces the Gregorian, not the stars. The equinox machine above is a model, not an observation: it is the same validated machinery the Solar Hijri page checks against published tables, good to the minute rather than the second. And a Śaka date is a plain date — no times, no zones, no sunset and no moon; the whole country agrees on it at midnight. After the Solar Hijri's knife-edge Tehran noons, that is the luxury here: nothing waits for anything.