One day in five, counted
Mercury is retrograde far more often than almost anyone assumes. We counted the days: 67 of 365 in 2027. And for fifteen centuries this was not a nuisance but the central problem of astronomy.
Anything in a violet panel we computed from DE421 ephemeris data. You can check it.
Anything on a gold rule was said by people, at a particular time. We say who, and when.
Where Mercury is now
GEOCENTRIC LONGITUDES, TROPICAL ZODIACMercury is the fastest thing in the chart after the Moon, and the least willing to stay still. Across a single year its apparent speed runs from 2.16 degrees a day forwards to 1.22 degrees a day backwards.
Mercury never leaves the Sun's neighbourhood, which is why it is so rarely seen: it orbits inside us, so from here it is always somewhere near the Sun in the sky.
Why three times a year, and not once
GEOMETRY, NOT TEMPERAMENTMercury does not reverse. It goes round the Sun the same way round as everything else, and faster than everything else, because it is closest.
That speed is the whole explanation. Mercury laps us roughly every four months, and each time it passes on the inside, between us and the Sun, we watch it slide backwards against the fixed stars for about three weeks. A planet that laps us three times a year appears to go backwards three times a year.
Venus does the same thing, and does it only once every nineteen months, for the same reason in reverse: it is further out, so it takes far longer to catch us up.
How much of the year, actually
A COUNT, NOT AN IMPRESSIONThis is the figure that is asserted constantly and measured almost never, so we measured it. Taking 2027 and asking, once a day at noon, whether Mercury was moving backwards:
The count samples one instant a day, which can be wrong by at most one day per station, so at most six days a year. That is nowhere near the size of the figure and we would rather state the method than round it away.
Nearly a fifth of every year. If Mercury retrograde explained a run of bad luck, it would have to explain about seventy days of it annually, which is a great deal of bad luck to be going on with.
| Retrograde from | At | Direct on | At | Days |
|---|---|---|---|---|
| 9 Feb 2027 | Pisces 5°59′ | 3 Mar 2027 | Aquarius 20°55′ | 21.79 |
| 10 Jun 2027 | Cancer 6°22′ | 4 Jul 2027 | Gemini 27°28′ | 24.06 |
| 7 Oct 2027 | Scorpio 4°56′ | 28 Oct 2027 | Libra 19°19′ | 20.99 |
Notice how unequal they are. The three differ by three days in length and by nearly seven degrees in the arc they retrace, which is the eccentricity of Mercury's orbit showing through.
The number that hides a fortnight
AN AVERAGE, AND WHAT IT CONCEALSReference works give Mercury a synodic period of 115.88 days, the average interval between one of these passages and the next. Measured across 11 consecutive intervals, we get 116.67, which agrees.
The agreement is the least interesting part.
Mercury has the most eccentric orbit of the planets, so how long it takes to lap us depends on where in that orbit the lapping happens. The average is real and the individual intervals scatter more than two weeks around it.
Which is worth holding on to whenever a single tidy number is quoted at you about a planet. The number is usually true and usually an average, and the average is usually the least true thing about any particular year.
For fifteen hundred years, this was not a nuisance. It was the problem astronomy existed to solve.
The circles built to explain it
THE INTERPRETATION, AND ITS HISTORYFrom here down we are not measuring. We are telling a story, with names and dates.
If the Earth stands still and everything circles it, retrograde motion is impossible: a thing going round you does not stop and reverse. Greek astronomy's answer was to give each planet two circles instead of one. The planet rides a small circle, the epicycle, whose centre rides a large one, the deferent; when the planet is on the inner part of the small circle, the two motions oppose and it appears to go backwards.
The construction is credited to Apollonius of Perga, around 200 BC, and was developed by Hipparchus. It exists for one reason. Retrograde motion is what it was built to explain.
Ptolemy formalises the whole system in the Almagest, adding a further device, the equant, to account for the planets changing speed. The result predicts the sky well enough to be used for the next fifteen hundred years, which is a very long time to be almost right.
The same man wrote the Tetrabiblos, the book that shaped Western astrology. The astronomy and the astrology are by one author and are not the same activity, a distinction his readers have been losing ever since.
Nicolaus Copernicus publishes De revolutionibus orbium coelestium, and retrograde motion stops needing an explanation at all. Put the Sun at the centre and let the Earth move, and the backwards loops are simply what overtaking looks like. The epicycles were never in the sky. They were in the frame of reference.
That is the sense in which a retrograde planet has always been about the observer. Not metaphorically, and not as a lesson about perspective. Literally, and as arithmetic.
The modern idea, that a retrograde Mercury interferes with messages, machines and travel, is much younger than the phenomenon and younger than most people assume. Astrology reached a mass audience through newspapers: R. H. Naylor's column in the Sunday Express in 1930 is the usual starting point, and sun-sign astrology as a popular form dates from there.
Nicholas Campion, a historian of astrology at the University of Wales, places the broadening of Mercury retrograde into an explanation for daily life in the 1980s, when, as he puts it, people started applying it to everything. The association with communication itself is older and comes from the god rather than the planet: Mercury carried messages, and looked after merchants, travellers and thieves.
So the phenomenon is one thing and its meanings are several, arriving in a definite order and from identifiable people. The circles of Apollonius, the equant of Ptolemy, the moving Earth of Copernicus, the newspaper column of 1930, the software of the 1980s. Underneath all of them, 116.67 days.
How these numbers were arrived at
METHOD, AND ITS CHECKNot one figure in this piece was typed by a person. All of them were computed from NASA's DE421 ephemeris data, by the same machinery that computes your birth chart inside the app.
The stations were located with a centred difference and successive bisection, down to the minute, then found a second time with Swiss Ephemeris, an entirely independent implementation.
Mercury is the hardest body to pin this way, and for a reason worth naming: the method approximates a curve, and the curvature grows with how fast a planet's speed is changing. Mercury changes speed faster than anything else in the chart. A few minutes inside three weeks change nothing, but we would rather print it than leave it out.
Sources
SO YOU CAN CHECK THEM YOURSELF- Positions and dates: JPL DE421. Cross-checked against Swiss Ephemeris 2.10.03.
- Epicycles: attributed to Apollonius of Perga (c. 262 to c. 190 BC) and developed by Hipparchus; formalised in Ptolemy, Almagest, 2nd century AD.
- Nicolaus Copernicus, De revolutionibus orbium coelestium, Nuremberg 1543.
- R. H. Naylor, Sunday Express, 1930, the newspaper column usually taken as the start of popular sun-sign astrology.
- Nicholas Campion, University of Wales, on the 1980s broadening of the idea, interviewed in Time, 2018.
Where Mercury stood when you were born, and where it is now. The chart is computed free.
Compute your chart