What actually casts the shadow in each kind of eclipse, why they arrive in pairs roughly twice a year, and what astrological tradition reads into them.
A lunar eclipse happens when our planet moves between the Sun and the Moon and drops its shadow across the lunar surface. A solar eclipse reverses that: the Moon slides in front of the Sun and drops its shadow on us. Both need three bodies in a near straight line, which is why they are occasional rather than monthly.
A solar eclipse can only occur at new moon and a lunar eclipse only at full moon, the two points in the cycle where anything lines up at all. Both arrive every month and most pass without incident. The reason is a tilt. The route the Moon takes around us is not in the same plane as the route we take around the Sun, but tipped by a little over five degrees. A typical new moon therefore carries it past the solar disc rather than across it, high or low, and a typical full moon takes it clear of the shadow altogether.
The two places where the lunar orbit crosses that plane are called the nodes. An eclipse needs a new moon or full moon landing near one of them, and the Sun passes each node twice a year. Those windows are the eclipse seasons. A season runs about a month, long enough to hold a new moon and a full moon a fortnight apart, so eclipses turn up in twos, one of each kind. The gap between seasons is around 173 days, which is why the dates creep earlier year on year. A calendar year holds at least four and can reach seven.
A longer rhythm sits on top. After 18 years, 11 days and about 8 hours the geometry returns and a near identical eclipse repeats, a period called the saros. The spare eight hours mean the repeat is seen from a different part of the world.
Solar eclipses rest on a coincidence. The Sun is roughly four hundred times wider than the Moon and sits roughly four hundred times further away, so the two discs look almost the same size from the ground. That will not last, since the Moon drifts a few centimetres further from us each year.
At the near end of its orbit the Moon hides the solar disc completely and the eclipse is total. At the far end it comes up fractionally too small and leaves a bright ring around the rim, an annular eclipse, from the Latin for ring. Off centre, only a bite is missing, and that partial version is what nearly everybody sees. Totality is narrow work: the full shadow paints a track at most a few hundred kilometres wide, so from one spot the total phase lasts minutes and can never exceed about seven and a half.
The safety rule is not decorative. Looking at a partly covered Sun can burn the retina permanently and painlessly, since there are no pain receptors there to warn you. Sunglasses, smoked glass, exposed film and phone cameras are all useless. Use filters made and certified for solar viewing. The single exception is the total phase itself, and the filters go back on the instant the first sliver of sunlight returns. An annular eclipse is never safe unfiltered, because a ring of Sun is still the Sun.
The shadow we cast has two parts. The dense inner cone, where sunlight is blocked altogether, is the umbra; the weaker outer region, where the light is only partly cut off, is the penumbra. The whole Moon inside the umbra is a total eclipse. Part of it inside gives a partial one with a curved dark edge, a curve that was among the earliest evidence this planet is a sphere. A pass through the outer shadow alone is a penumbral eclipse, faint enough that people look up and notice nothing.
The famous part is the colour. A fully eclipsed Moon does not go black, it turns dull copper, and the atmosphere is responsible. Sunlight grazing the rim of the world bends into the shadow, loses its blue to scattering, and arrives red. You are watching the combined light of every sunrise and sunset in progress on Earth at that moment. Blood moon is the popular name for it. A lunar eclipse also runs for hours rather than minutes and shows from anywhere the Moon is up, which is around half the world at once, and it needs no filter or equipment.
Astrology treats an eclipse as a new moon or full moon with the volume raised. The usual readings hold, a beginning at one and a culmination at the other; what changes is the scale and the suddenness. The associations are with turning points that arrive without a build-up, and with endings that later look like starting points.
The nodes reappear here under their own names. The crossing point where the Moon heads north is the north node, read as the direction of growth; the point facing it is the south node, read as familiar ground being let go. Since the two sit directly across from each other, eclipses land in a pair of opposite signs and stay there for around eighteen months, so a run of them seems to press on one area of life.
Older readings were darker. An eclipse was an omen, usually bad and usually aimed at a ruler, and ancient records describe substitute kings placed on thrones to absorb whatever was coming. That fear makes sense when the Sun goes out at midday, and it is a poor guide once the reason is known.
A listed solar eclipse is not a promise of one where you are standing. These are local events: the track crosses a thin strip of the world and everybody else gets a partial view or none, so any listing worth trusting names the region of visibility beside the time. Lunar eclipses are more generous, since the whole night side of the planet watches the same one.
Worth pausing on how precisely this is known. Eclipse times are calculated centuries in advance, printed to the minute, and they arrive as advertised. That accuracy belongs to the mechanics of three moving bodies rather than to any meaning laid on top, and keeping the two apart is the honest way to read an eclipse page, this one included.


