Keeping Time with the Skies

Long before anyone owned a watch, people already had a clock, and it was the sky. Three motions in space give us the three units of time we still use today. The spin of the Earth gives the day, the journey of the Moon around the Earth gives the month, and the journey of the Earth around the Sun gives the year. This section deals with the first of them.

The Earth turns about an imaginary line called its axis, which runs through the North Pole and the South Pole. One complete turn takes about 24 hours, and that is what we call a day. The turn is from west to east. Sunlight can only fall on the half of the ball that faces the Sun, so at any moment one half of the Earth is in daylight and the other half is in darkness. As the Earth turns, the boundary between the two sweeps steadily across the surface, and every place slides out of darkness into light at sunrise and back into darkness at sunset. Nothing has to be switched on or off; a place simply carries on turning.

Because we are being carried eastwards, the Sun seems to come up in the east, climb across the sky and go down in the west. Nothing about the Sun is actually moving in that way. It is the same feeling you get in a train that pulls smoothly out of a station: the platform seems to slide backwards. You cannot feel the spin of the Earth for exactly the same reason a passenger cannot feel a steady train. The air, the buildings, the oceans and you are all moving together, at the same speed, in the same direction. Near the equator that speed is more than 1,600 kilometres per hour, and it still goes completely unnoticed.

Two things you can watch for yourself. First, shadows. Push a straight stick upright into open ground and mark the tip of its shadow once an hour. Early in the morning, with the Sun low in the east, the shadow is long and stretches away towards the west. It shortens as the Sun climbs, becomes the shortest around midday, and then lengthens again on the other side as the Sun sinks towards the west. A sundial works on exactly that idea, with the hour marks fixed permanently around the base of the rod that casts the shadow. Second, star trails. If a camera is left pointing at the northern sky with its shutter open for an hour, the stars do not come out as dots. They come out as neat curved arcs, all centred on one point close to the Pole Star. The sky is not turning. The camera is, because the ground under it is.

There is one small surprise hidden inside the day. While the Earth spins, it is also creeping along its path around the Sun. So after one complete spin, your place has to turn a tiny bit extra before the Sun is back over the same spot in the sky. A day measured against the Sun is therefore about four minutes longer than a day measured against the distant stars. Those four minutes matter more than they sound: a given star rises about four minutes earlier each night, which builds up to close to two hours in a month, and that is why the night sky slowly changes as the seasons pass.

The Moon is our nearest natural neighbour in space, and it is the reason we have a unit of time sitting between the day and the year. The Moon travels around the Earth and needs about 27.3 days to come back to the same position against the background stars. But the month we actually count is a little longer: about 29.5 days from one new moon to the next. Those extra two days appear because the Earth has itself moved a good distance along its own path around the Sun in the meantime, so the Moon must travel a little further before it lines up with the Sun and the Earth in the same way again. This 29.5-day cycle is the lunar month, and it is the backbone of Indian calendars.

The Moon makes no light of its own. Every bit of moonlight you have ever seen is sunlight bouncing off grey rock. That one fact explains the phases completely. The Sun always lights up exactly one half of the Moon, the half turned towards it. What changes from night to night is not how much of the Moon is lit, but how much of the lit half is turned towards us.

Follow a single cycle. At new moon (Amavasya) the Moon lies roughly between the Earth and the Sun, its lit half faces away from us, and we see nothing at all. A day or two later a thin crescent appears low in the west just after sunset, and on each following evening it grows a little fatter and stands a little higher. About seven days after new moon we see exactly half a disc, called the first quarter. The lit part keeps growing through the fat gibbous shape until, about fourteen or fifteen days after new moon, the Moon is on the far side of the Earth from the Sun, the whole lit half faces us, and we get full moon (Purnima). After that the same steps run backwards, the disc shrinking through gibbous and half shapes to a thin crescent seen in the east before dawn, and then back to new moon.

Indian calendars give the two halves of this cycle their own names. The fortnight running from new moon to full moon, when the lit part grows, is the shukla paksha or bright fortnight. The fortnight running from full moon back to new moon, when it shrinks, is the krishna paksha or dark fortnight. Each fortnight is divided into fifteen tithis, and festivals are fixed to a tithi rather than to a date in the English calendar.

Two more facts are worth carrying with you. The Moon rises roughly fifty minutes later each day, which is why a full moon comes up around sunset and stays up all night, while a thin old crescent shows itself only shortly before sunrise. And the Moon always keeps the same face turned towards us, so nobody standing on the Earth has ever seen the other side without sending a spacecraft round to look. That does not mean the Moon has stopped rotating. It means the Moon turns once on its own axis in exactly the time it takes to go once around us, so the spin stays in step with the journey and the same side is left facing us the whole way round. The hidden half is sometimes called the dark side, which is a misleading name, because it receives just as much sunlight as the half we can see.

Finally, a warning about a very common mix-up. The phases have nothing to do with the shadow of the Earth. During ordinary phases the Earth is not blocking anything; we are simply looking at the lit half of the Moon from a different angle. On the rare occasions when the shadow of the Earth really does fall on the Moon, we call it an eclipse, and that is a different event altogether.

Every object lit by the Sun casts a shadow into space, and the Earth and the Moon are no exception. Because the Sun is a broad source of light and not a tiny point, each shadow has two parts. The dark inner cone, where the Sun is blocked completely, is the umbra. The lighter region wrapped around it, where only part of the Sun is hidden, is the penumbra. An eclipse is nothing more mysterious than one of these shadows landing on something we can see.

A solar eclipse happens when the Moon comes between the Sun and the Earth and the shadow of the Moon falls on us. This can only happen at new moon. The umbra of the Moon has narrowed a great deal by the time it reaches the ground, so a total solar eclipse is seen only from a narrow track across the surface of the Earth, and totality at any one place lasts only a few minutes. Everybody in the much wider penumbra sees a partial eclipse instead, with a bite taken out of the Sun. Sometimes the Moon happens to be near the far end of its path and looks slightly smaller than usual; then it cannot cover the Sun completely and leaves a thin bright ring shining all around itself. That is an annular eclipse.

Total solar eclipses are possible at all only because of a remarkable coincidence. The Sun is about 400 times wider than the Moon, and it also happens to lie about 400 times farther away, so the two look almost exactly the same size in our sky.

A lunar eclipse happens when the Earth comes between the Sun and the Moon, and the Moon moves into the shadow of the Earth. This can only happen at full moon. The shadow of the Earth is far wider than the Moon, so a lunar eclipse is a slow, stately event that can keep the Moon inside the umbra for more than an hour, and every person on the night side of the Earth sees it at the same moment. During totality the Moon usually does not disappear. It turns a dull copper red, because sunlight passing through the ring of air around the Earth is bent inwards onto the Moon, and that air scatters the blue part of the light away and lets the red part through. It is the same reason a sunset is red.

So why do we not get an eclipse every month? If the Moon went round the Earth in exactly the same flat plane in which the Earth goes round the Sun, there would be a solar eclipse at every new moon and a lunar eclipse at every full moon. But the path of the Moon is tilted by about five degrees to that plane. Five degrees sounds tiny, and yet it is more than enough: at most new moons the shadow of the Moon sails past above or below the Earth and misses it completely, and at most full moons the Moon itself passes above or below the shadow of the Earth. The two paths cross at only two points, and an eclipse becomes possible only when a new moon or a full moon happens near one of those crossing points. That works out a few times in a year, which is why eclipses arrive in short seasons rather than month after month.

Watching safely. A lunar eclipse is completely safe to look at with bare eyes, binoculars or a telescope, because you are only looking at dim reflected light. A solar eclipse is a different matter. Never look straight at the Sun, not even when most of it is covered, and never through sunglasses, smoked glass, exposed photographic film, an X-ray sheet or a compact disc. Those things cut down the brightness so the eye feels comfortable, but they let through the invisible rays that quietly damage the retina. Use a certified solar filter, or better still do not look at the Sun at all: make a pinhole in a card, let the image of the Sun fall on a sheet of paper held in the shade, and watch the bite appear on the paper instead. An eclipse is a predictable shadow, worked out years in advance by simple geometry. There is no reason to stay indoors, throw away cooked food or skip a meal because of one.

The third motion is the long one. The Earth travels once around the Sun in about 365 days and 6 hours, and that trip is the year. The path is very close to a circle, though not perfectly so, and this is exactly where most people go wrong about the seasons. It is tempting to think we get summer when the Earth is near the Sun and winter when it is far away. That cannot be right, for two solid reasons. First, when the northern half of the world is having summer, the southern half is having winter at that very same moment, and both halves are obviously the same distance from the Sun. Second, the Earth is actually at its closest to the Sun in the first week of January, right in the middle of the northern winter.

The real cause is the tilt. The axis of the Earth is not upright with respect to its path; it leans over by about 23.5 degrees. More important still, as the Earth moves around its orbit the axis keeps pointing in almost the same direction in space, roughly towards the Pole Star. So for one part of the orbit the northern half leans towards the Sun, and six months later the same axis, still pointing the same way, has the southern half leaning towards the Sun instead.

Leaning towards the Sun changes two things at once, and both of them add heat:

• The Sun climbs higher at noon. When sunlight comes down steeply, a beam lands on a small patch of ground. When the same beam comes in at a slant, it is smeared over a much larger patch, so each square metre of ground receives less of it. Slanting sunlight is weak sunlight.
• The days grow longer. The half that leans towards the Sun spends more hours of each turn in daylight and fewer in darkness, so there is more time to warm up and less time to cool down.

Four positions in the orbit carry names. Around 21 June the northern half leans most towards the Sun: it gets its longest day, and the noon Sun stands directly overhead at the Tropic of Cancer, the line that runs right across the middle of India. Around 22 December the situation is reversed, the northern half gets its shortest day, and the noon Sun stands overhead at the Tropic of Capricorn. These are the solstices. Halfway between them, around 21 March and 23 September, neither half leans towards the Sun, the noon Sun stands overhead at the equator, and day and night are close to equal all over the world. These two are the equinoxes.

The tilt also explains what happens at the extremes. At the equator the Sun rides high in the sky all through the year and day and night stay close to twelve hours each, so there is very little swing from one season to the next. Beyond the Arctic and Antarctic circles the lean is enough to hold the Sun above the horizon for a full turn of 24 hours in the local summer and below the horizon for a full 24 hours in the local winter. That is where the midnight Sun and the long polar night come from.

In India the picture is richer still, because the tilt sets the stage but the monsoon writes much of the script. The land heats strongly through March, April and May, and the heated land draws in moist air from the ocean, which brings the rains. The older Indian scheme recognises six ritus rather than four seasons: Vasanta or spring, Grishma or summer, Varsha or the rains, Sharad or autumn, Hemanta or early winter and Shishira or winter. Each covers roughly two months, and each is tied to the same steady journey of the Earth around the Sun.

A calendar is an attempt to fit the three natural clocks, day and month and year, into one tidy table. The attempt is harder than it looks, and the difficulty is pure arithmetic: the natural units simply do not divide into one another. A year is not 365 days exactly, it is about 365 days and 6 hours. A lunar month is not 30 days, it is about 29.5 days. And twelve lunar months add up to about 354 days, some eleven days short of a year. Every calendar humanity has built is a different answer to the question of what to do with these leftovers.

Solar calendars follow the Sun and allow the months to drift away from the Moon. The Gregorian calendar, the one used for official dates almost everywhere today, is a solar calendar. It gives an ordinary year 365 days, and then, to use up the six leftover hours, it adds one whole extra day, 29 February, once in four years. Four lots of six hours make exactly one day, so the trick works, but only very nearly. The true leftover is a few minutes less than six hours, and those few minutes would build up to a whole extra day in about 130 years. The repair is the century rule: a year divisible by 4 is a leap year, except a century year, which counts as a leap year only if it is divisible by 400. So the year 2000 was a leap year, while 1900 was not and 2100 will not be.

Lunar calendars do the opposite. They follow the Moon faithfully and let the months drift through the seasons. The Islamic or Hijri calendar counts twelve lunar months, about 354 days in all, so each of its months begins about eleven days earlier in the solar year than it did the time before. That is why Ramzan and Id shift steadily backwards through the seasons and work their way through the whole cycle in roughly 33 years.

Lunisolar calendars refuse to give up either clock. The traditional Indian panchang, like the Hebrew calendar, keeps months that begin and end with the phases of the Moon, but it also drags them back into step with the Sun by inserting a whole extra month, the adhik maas, roughly once in about three years. Twelve lunar months fall behind by about eleven days a year; after three years the shortfall is close to a full month, and slipping in one extra month puts everything back where it belongs. This is why festivals fixed to a tithi, such as Diwali on a new moon night, Holi on the full moon of Phalguna and Raksha Bandhan on the full moon of Shravana, always turn up in the same season, and yet land on a different English date each time.

India also has an official solar calendar of its own. The Indian national calendar, counted in the Saka era, was adopted in 1957 after a committee studied and reformed the many calendars then in use across the country. It has twelve months beginning with Chaitra, which normally starts on 22 March, and it is printed alongside the Gregorian date in government publications. Beside it, many regions keep their own reckoning: the Vikram Samvat in the north and west, and a family of solar new year days in the middle of April celebrated as Baisakhi, Puthandu, Pohela Boishakh, Vishu and Bihu.

One unit in the calendar has no parent in the sky at all. The seven-day week does not divide the month or the year evenly. It is a human invention, and the names of its days quietly record what people long ago thought moved among the stars. Ravivar belongs to the Sun, Somvar to the Moon, Mangalvar to Mars, Budhvar to Mercury, Guruvar to Jupiter, Shukravar to Venus and Shanivar to Saturn. The same seven bodies, in the same order, lie behind the English names of the days.

Every clock ever built works on one principle: find something that repeats at a steady rate, then count the repeats. The sky supplied the first repeating event of all, and the earliest instruments that read it were made of shadow and water.

The oldest is the sundial. A rod called a gnomon casts a shadow that swings round as the Earth turns, and marks on the base turn the direction of that shadow into an hour. In a carefully made sundial the gnomon is not upright but slanted, so that it lies along the direction of the axis of the Earth and points at the Pole Star; that is what makes the shadow travel round at an even rate through the day. The masonry instruments of the Jantar Mantar observatories in India carry the idea to an extreme: the great sundial there is a stone triangle with a slanting edge of exactly that kind, tall enough to carry a staircase up its slope, and its shadow falls on long curved scales so widely spaced that the local time can be read far more finely than on a small dial. A sundial has one obvious weakness, though. It is useless at night and on a cloudy day. So people also built clocks that needed no Sun: water clocks, in which a bowl with a small hole steadily sinks or fills in a fixed time, and sand clocks, in which sand trickles from one glass bulb into another. The traditional Indian unit measured this way was the ghati. Sixty ghatis made one day, so one ghati works out to 24 minutes.

Later clocks kept the counting idea and only improved the repeating event. A pendulum of a fixed length takes the same time for every swing, and that made mechanical clocks accurate to a few seconds. A quartz crystal in a wristwatch or a wall clock vibrates tens of thousands of times a second when a tiny current is passed through it, and a circuit counts those vibrations. The steadiest clocks we have are atomic clocks, which count vibrations inside atoms themselves and are so regular that they would not gain or lose a second in millions of years. In India the national physical laboratory keeps the standard time using such atomic clocks, and every official clock in the country is meant to agree with them.

That raises a fair question: whose time? Noon is a purely local event. It arrives when the Sun is highest at your own place, and that moment sweeps westwards across the country as the Earth turns. Since the Earth turns through 360 degrees in 24 hours, it turns through 15 degrees in one hour, which is the same as saying that one degree of longitude is worth four minutes of time. India stretches from about 68 degrees east to about 97 degrees east, a spread of roughly 29 degrees, so the Sun rises in the far north-east of the country nearly two hours before it rises in the far west.

If every town kept its own local time, trains, offices, television news and flight schedules would collapse into confusion. So a country picks one line of longitude and sets all its clocks to the local time of that single line. India uses 82.5 degrees east. Multiply 82.5 by four minutes and you get 330 minutes, which is 5 hours and 30 minutes, and that is exactly how far Indian Standard Time runs ahead of the time at the zero line of longitude passing through Greenwich. The world is divided into about 24 such time zones. The odd half hour in IST is not odd at all once you see where it comes from; it is simply what the chosen line of longitude works out to.

On a clear night away from street lights you can pick out patterns among the stars: a giant hunter, a plough, a letter W, a scorpion with a curved tail. These patterns are constellations, and nearly every civilisation has drawn its own figures on the same sky and told stories about them.

The important scientific point is that a constellation is a pattern of direction, not a group of neighbours. The stars in one figure are usually nowhere near one another in space. They can lie at very different distances, some of them many times farther away than others, and they appear side by side only because they happen to lie in nearly the same direction as seen from here. Viewed from a planet circling some other star, those same stars would be scattered about and the pattern would simply not exist. Astronomers today divide the whole sky into 88 such regions with agreed boundaries, so that every star has a fixed address in the sky.

Three patterns are worth learning, because once you know them you can find your way about the sky.

• Saptarshi, known elsewhere as the Great Bear or the Plough, is a set of seven bright stars in the northern sky, four of them forming a bowl and three a curved handle. It rides high on spring and summer evenings.
• The Pole Star, called Dhruva, is found using Saptarshi. Take the two stars at the outer end of the bowl, join them with an imaginary line and stretch that line onwards about five times its own length. You arrive at a moderately bright star that sits almost exactly above the North Pole. Because it lies almost exactly in the direction in which the northern end of the axis of the Earth points, it barely shifts all night while the whole sky wheels around it, and that is why sailors and travellers used it to find north.
• Orion, called Kalpurush or the hunter, dominates winter evenings. Three bright stars in a short straight row make his belt, with a bright star at one shoulder and another at a foot. Extend the line of the belt onwards and you come to Sirius, the brightest star in the whole night sky. On the far side of the Pole Star from Saptarshi, a slanting W or M of five stars marks Cassiopeia, also best seen on winter evenings.

Why is the winter sky different from the summer sky? Because of the yearly journey of the Earth. Night falls on whichever side of the Earth is turned away from the Sun, and as we travel around the orbit that side comes to face a different part of space. Orion is a winter object simply because in June it stands up there behind the Sun in the daytime sky, where its light is drowned out. The same effect shows up night by night in the four minutes we met earlier: each star rises about four minutes earlier than it did the previous night, which is close to two hours in a month and about 24 hours across a full circuit of the Sun, bringing the sky back to where it began.

Indian astronomy divided the sky in a different way, into 27 nakshatras spread along the path the Moon follows against the stars. Since the Moon takes about 27.3 days to complete that circuit, it crosses roughly one nakshatra a day, which made the nakshatras a natural way of recording the date. The names of the lunar months come from them, Chaitra from Chitra and Vaishakha from Vishakha for instance, because the full moon of that month falls near that nakshatra. One last piece of sky-watching. Stars twinkle because their light arrives as a fine point and is bent about by the restless air above us, while planets show a tiny disc rather than a point, so they shine more steadily, and they slowly wander from one constellation to the next. The word planet comes from a Greek word meaning wanderer.