Precession of the Equinoxes
The Earth’s rotation axis is not fixed against the stars. It traces a cone, completing a circuit in roughly 25,800 years, and the equinox points where the celestial equator cuts the ecliptic travel westward along the ecliptic as it does so. The rate of general precession in longitude is about 50.3 arcseconds a year, which is one degree in about 71.6 years 1.
That single fact is the mechanism behind two things this site keeps returning to: the drift of any calendar anchored to the stars, and the separation between the seasonal and stellar zodiacs.
Two year lengths
The clearest way to state precession is as a difference between two definitions of a year.
The tropical year, the interval between successive March equinoxes, runs about 365.2422 days. The sidereal year, the interval taken for the Sun to return to the same position against the fixed stars, runs about 365.2564 days 1. The gap is about twenty minutes. The Sun returns to the same season about twenty minutes before it returns to the same star, every year, and the arrears accumulate.
Twenty minutes a year is one day in about seventy-one years and a full circuit of the ecliptic in about 25,800. Over the lifetime of a written tradition it is substantial: a star-based agricultural marker recorded in the eighth century BCE now falls roughly a month later in the seasonal year than it did when it was written down. See Heliacal Risings.
Who measured it
Hipparchus of Rhodes, working in the second century BCE, compared his own measured longitudes for a set of stars with observations made roughly a century and a half earlier and concluded that the longitudes had increased, at a rate he put at not less than one degree per century. The work is lost; the report survives through Ptolemy, who treats it in two chapters of the Almagest and adopts the same rate 2 bk. VII chs. 2-3.
One degree per century is too slow. The true figure is closer to one degree in seventy-two years, and the underestimate propagated through the tradition for centuries. Later Islamicate astronomers improved on the rate empirically. The Latin West meanwhile took up trepidation, the theory that the equinoxes oscillate rather than accumulate, which held the field long enough to be embedded in the standard tables and to be entertained by Copernicus. A worse theory displacing a better observation is a rare event.
Correction
Commonly saidThe Babylonians knew about precession, and encoded it in their star lore.
Babylonian astronomy worked in a star-fixed frame, dividing the ecliptic by reference to stars rather than to the equinox 3. In such a frame the effects of precession accumulate invisibly, because the frame moves with the stars and nothing inside it registers the motion. The effects are present in the data, in the difference between sidereal and tropical year values and in the slow obsolescence of heliacal-rising dates. No surviving cuneiform text identifies the phenomenon or states that the equinoxes move.
The defensible sentence, and the one this site uses, is: Babylonian astronomy used a star-fixed frame in which the effects of precession accumulate without being identified. That is a real and interesting fact about how a frame determines what its users can see, and it does not require the stronger claim.
What can be handed back is considerable. Babylonian astronomy produced arithmetic lunar theories of great power, systematic observational records over centuries, and the twelve-sign division of the ecliptic that the whole later tradition inherited. None of that needs precession attached to it.
Why it separates the two zodiacs
A zodiac has to be anchored to something. Two anchors are available and they move relative to one another at 50.3 arcseconds a year.
Anchoring 0° Aries to the March equinox gives the tropical frame, in which the signs are permanently locked to the seasons and permanently sliding against the stars. Anchoring it to a stipulated stellar reference gives the sidereal frame, in which the reverse holds. The angular difference between the two zero points is the ayanamsha, and it grows by about a degree every seventy-two years.
The tropical convention was adopted explicitly and with a stated reason, by authors who knew precession thoroughly. Ptolemy takes the equinoctial and solstitial points as the beginnings of the signs on the ground that those points are naturally determined whereas the constellation figures have no definite boundaries 4 bk. I ch. 11, and he had set out the precession of those very points in the Almagest 2 bk. VII chs. 2-3. The choice was a choice, and it was made with the relevant fact in hand 5.
The offset is small enough to be invisible near the epoch at which the two frames coincide and large enough to be conspicuous a millennium away from it. That is why the divergence became a public argument only in the modern period, and it is why the offset is a good thing to display: computed for any date, it shows a reader the exact size of the disagreement between two systems that are otherwise hard to compare. Under the Lahiri convention, adopted as the Indian national standard by the Calendar Reform Committee in the 1950s and the default this site computes with, it now stands at roughly 24 degrees, about four-fifths of a sign. Under Fagan/Bradley it is roughly 25. The figure does not identify itself, so the convention has to be named alongside it.
For what the two frames are for, and for the third frame that is neither of them, see Three Reference Frames.
Gap in the evidence
The epoch at which the tropical and sidereal zero points coincide is not a discovered date. It follows from which ayanamsha is stipulated, and the values in current use differ by roughly two degrees, which is about a century and a half of precession. Conventions in wide use place the coincidence in the third century of the common era; the most widely used of them is calibrated so that the coincidence falls near 285. This site names the convention it is using wherever it displays an offset, because without a named convention the number is not defined. No scholarly source is cited here for the calibration dates of individual ayanamshas; the entry states the dependence and leaves the comparative table for a source it can cite.
References
- Claudius Ptolemy, trans. G. J. Toomer. Ptolemy's Almagest. London: Duckworth, 1984 (composed c. 150 CE). bk. VII chs. 2-3. find a copy (catalogue search) recordprimary source
Reissued by Princeton University Press, 1998. Book VII chapters 2-3 for precession.
- Jean Meeus. Astronomical Algorithms. Richmond, Virginia: Willmann-Bell, 1998. ISBN 0943396611 recordscholarly
Second edition, 1998. The standard reference for the algorithms behind solstice, equinox and solar-longitude computation, including the equinox and solstice chapter this site's seasonal calculations follow.
- Claudius Ptolemy, trans. F. E. Robbins. Tetrabiblos. Loeb Classical Library 435. Cambridge, MA: Harvard University Press, 1940 (composed c. 150 CE). bk. I ch. 11. find a copy (catalogue search) recordprimary source
The standard English reference and the chapter-numbering scheme this site adopts and names. Non-Robbins reprints number the chapters differently, so every locus here is given as Robbins numbering.
- Nicholas Campion. A History of Western Astrology, Volume I: The Ancient and Classical Worlds: Continuum, 2009. ISBN 9781441127372 recordscholarly
The standard single narrative survey. Volume I first appeared as The Dawn of Astrology, Hambledon Continuum, 2008. The subtitle is "The Ancient and Classical Worlds", not "The Ancient World"; the shortened form is widely miscited.
- Hermann Hunger, John Steele. The Babylonian Astronomical Compendium MUL.APIN. Scientific Writings from the Ancient and Medieval World. London: Routledge, 2019. find a copy (catalogue search) recordpeer reviewed
The current standard edition. Some listings show 2018 for the first printing. Use this and Rochberg 2004 for Babylonian star names and figures, not the popular compilations that circulate on astrology sites.