Stations: the instant a body turns

Station · apparent motion passes through zero · one instant, no orb · the engine's record carries a time, a body and a direction, and no longitude

01The two ends of a retrograde

A retrograde period has two ends, and each end is a station: the instant a body's apparent motion against the background stars passes through zero and changes sign. Forward before and backward after, at the first one; backward before and forward after, at the second. Most writing about retrogrades is about the stretch in between. This piece is about the ends, because nearly everything this product does and does not print about a retrograde follows from what a station is as a measurement.

Why the motion appears to reverse at all — one moving platform overtaking another, and why the Sun and the Moon are never seen to do it — is the subject of the Mercury piece, and is not repeated here.

→ What a retrograde actually is

02An instant, not a period

Apparent motion in longitude is a signed quantity: positive while a body advances through the zodiac, negative while it appears to fall back. A station is the point where that sign changes, and a sign change happens at one point, not across a range. The date can therefore be given to the second without overstating anything, and the phrase "around the station" has no measured width attached to it.

That is worth setting against how an aspect is defined, because the two look similar on a page and are built differently. An aspect is a proximity: two bodies count as square when they are ninety degrees apart give or take a tolerance, and that tolerance — the orb — is a number this product publishes and halves for transits. A station has no orb, and not because none has been chosen. There is nothing for an orb to measure. Proximity to what value is a station near, when the value it is defined by is zero and the body passes through it exactly once.

→ What an orb is, and why a transit gets half of one

What is true is that a body near a station barely moves. For something like a day on either side of the instant, its longitude changes by a negligible amount, which is why a station looks in the sky like a standstill rather than a turn. That near-standstill is a consequence of speed being close to zero, not a second definition of the event, and this product keeps them in that order: the instant is what gets recorded, the slowness is what surrounds it.

03How the instant is found

The engine does not look up a station in a table. It scans the requested window body by body, at a step chosen for that body's speed — a day for Mercury, Venus and Mars, two days for Jupiter and Saturn, five for Uranus, Neptune and Pluto — asking the ephemeris for each body's speed at each sample. When two consecutive samples come back with opposite signs, a station lies between them, and the interval is then halved repeatedly, always keeping the half that still contains the sign change, until it has closed to far less than a second.

Two things follow from that method, and both are limits rather than features. The first is that the step has to be short enough that a reversal cannot happen and unwind between two samples; the steps above are sized against each body's own motion for exactly that reason, which is also why a fast body is sampled far more often than a slow one. The second is that the precision of the printed second is the search's precision, not the sky's. The search converges on the moment the ephemeris says the speed changes sign, and the ephemeris is a model with its own error. Printing a second is honest about the arithmetic; it is not a claim that the solar system is known to the second.

→ Which ephemeris this is, and what it can and cannot promise

The Sun and the Moon are not scanned for stations at all. Neither can appear to reverse from Earth, so a search over them would be guaranteed to return nothing, and a chart that shows either as retrograde is reporting a data fault rather than a rare event.

04The record has no degree

A station, as this engine stores it, is four things: that it is a station, when it happened, which body, and which direction the body turned. There is no longitude in it. The scan solved for a moment on the body's speed; where the body stood at that moment is a separate ephemeris call, and this scan does not make it.

So no page in this product prints the degree of a station, and none of the wording around one implies a degree is known. Where a station's sign can be established at all, it is established by inference from other records rather than read off the station itself: if two ingresses bracket a station and the body crossed no other boundary between them, the station happened inside the sign those two ingresses define. That inference is sound, and it is stated as an inference wherever it is used, because it depends on the assumption that the ingress list for the window is complete.

For the outer planets the inference is not available. A body that takes seven years to cross a sign produces no ingress inside a four-month window, so there is nothing to bracket its station with, and its sign is simply not given. That gap is not brevity. It is the difference between a number that was computed and a number that would have to be supplied, and the second one is the kind of error a reader has no way to catch, because catching it would mean running the calculation that was supposed to produce it.

05Why there is no shadow period here

A common addition to a published retrograde is a shadow: a stretch before the retrograde station, during which the body first passes the degrees it will later travel back over, and a matching stretch after the direct station, during which it covers them a third time. The idea is internally consistent and it rests on something real, described in the next section. What it does not have is a single boundary that different sources agree on, so a product that printed one would be presenting a choice as a measurement.

There is a second and blunter reason, and it is this article's subject. A shadow is defined by the degrees a station happens at. Those degrees are not in the record. Even granting the idea every benefit, this engine's event data cannot produce a shadow period, because the quantity the definition starts from was never stored. Declining to publish it is not caution about the astrology; it is a statement about what was computed.

06The shape the data does show: three crossings

The real geometry underneath the shadow idea survives without any degree at all. Between its two stations a body travels back over a stretch of zodiac it has already covered and will cover again, so every degree inside a retrograde arc is crossed three times — forward, backward, forward — where every other degree is crossed once.

When that arc straddles a sign boundary, the three crossings appear in the event list as three ingress records for the same boundary. Venus does this across the second half of 2026: it enters Scorpio on 10 September moving forward, re-enters Libra on 25 October moving backward, and enters Scorpio again on 4 December moving forward. One boundary, three records, in two directions, produced by a scan that was never told to look for the pattern.

The same window shows the other case just as plainly. Mercury's stations fall on 24 October and 13 November, and the engine records no Mercury ingress between them, which places the whole of that loop inside a single sign. Whether a retrograde crosses a sign boundary is therefore answerable from the structure of the event list — how many ingress records sit between two stations — without knowing a single degree.

07Half a loop, and what not to do about it

A station is one end of something. When a window contains only one station for a body, the period it belongs to has its other end outside the window, and its length is unknown. That happens constantly at a window's edges: a body showing only a direct station was already retrograde before the window opened, and a body showing only a retrograde station is still retrograde when it closes.

The right thing to print in that case is the station and nothing more. An average retrograde duration for that body exists and is easy to find, but the average is not this period, and substituting one for the other converts an unknown into a figure that looks measured. The product leaves the length unstated instead.

08Why a station outranks an ingress

The sky calendar has to rank its own events, or a month's list becomes the Moon changing signs a dozen times with everything else buried in it. The engine scores a station higher than an ingress for the same body — a base of 0.55 against 0.3 — and then adds half of the body's rarity, where rarity is one minus the weight that same engine assigns the body for aspect strength. Slow bodies carry the lightest weights, so they carry the most rarity, and a Pluto station scores higher than a Mercury one.

The base difference is a claim about the two kinds of event rather than about the bodies. An ingress is a threshold: the body crosses it and keeps going, and a moment later the crossing is over. A station is a change of direction that holds for weeks or months afterward. Both are instants; only one of them changes what the body does next.

09Where you see it in your own chart

A natal retrograde flag and a transiting station are different facts and are shown in different places. The flag on a natal placement is the sign of that body's speed at the moment of birth, fixed for good; a station on the calendar is a moment in the sky that belongs to nobody in particular.

One case joins them. Someone born within a day or two of a station has a body that was barely moving, and a body that barely moves holds its degree — so a transit contacting that placement stays inside orb unusually long, and the engine's applying and separating labels, which are read off the direction the separation is changing, lose their meaning near zero speed. That is a real feature of such a chart, and it is a statement about duration, not about character.

→ Draw the chart and read the speeds

→ Sun, Moon and Rising, each with its degree

10What a station date is not

It is not a warning. A station is the same instant for everyone alive during it, which is what makes it publishable years ahead and also what limits it: a shared moment is not a reading of anyone. Whether a retrograde touches a particular chart is an orb question about the degrees it passes over, and this product answers that against a chart rather than against a calendar.

It is also not a physical event. Nothing stops in space at a station. What passes through zero is the rate at which a sightline from a moving Earth sweeps along the zodiac, and the body it points at keeps its own speed throughout.

And the missing degree is not a rounding or an editorial cut. It is absent from the data, said out loud here so that a page which does not print it reads as a limit that was measured rather than a detail that was skipped.

11Related reading

Mercury retrograde carries the mechanism at full length and the natal condition that goes with it. What a transit is covers the step this one stops short of — how a shared sky is checked against one fixed chart, and at what tolerance. Swiss Ephemeris explains what the underlying calculation is, which is the ceiling on every second printed above.

Updated 2026-09-10.