| MadSci Network: Science History |
Hi,
Thank you for your question! People use a timescale to divide a long period, such as a year, in more convenient blocks - a day, an hour, a minute or a second. In this case, we are not extremely interested in accuracy: if your watch is slow for a second a day, you probably won't notice it. However, a definition of time has a second, different, use: to be able to accurately measure very small time differences. This is in particular important for physics. We'll discuss both ways of measuring time.
On longer timescales, time chiefly depends on the concept of the year. A year is the time it takes the Earth to revolve around the Sun1,2. This means that the Sun will be at the same place on the horizon as exactly one year ago, provided we look at the same latitude. This can be measured at the equinox (the day in which both day and night last 12 hours) or the solstice (the shortest or longest day). Another, slightly different method is looking at the stars. The stars move only extremely slowly, and if we can establish the stars are at the same point in the sky as one year ago, one full year has passed.2, 3. A practical day of doing this is taking the first day at which a particular star can be seen at dawn, for instance the bright star Sirius 4.
This way of keeping time was discovered quite early in human history, presumably because it is of vital importance for farmers to know when to plant their crops. However, this way of keeping time is not that accurate, for various reasons. For instance, the axis of the earth changes its tilt, causing the day of the solstice to change. This change is not present in the siderial year - the stars don't care the tilt of the Earth's axis changes. Furthermore, the speed of the Earth around the Sun is not constant. The fact the orbit of the Earth around the Sun is tilted, too the fact the orbit is elliptical don't help, either 1, 2, 3.
In short, we need a different way to keep time accurately. As you point out in your question, we do this by looking at the oscillations of an atom, in particular, cesium. The reason is that we can measure this extremely accurately. We do this by means of an atomic clock 5, 6.
An atomic clock works on the following principle. A cloud of cesium atoms is produced and put into a vacuum chamber. Using lasers, the cloud is moved and cooled 7. So, we have a cloud of ultracool cesium. We are now going to use a special trick. You see, atoms have a set of discrete 8 frequencies at which they vibrate. These frequencies are different for every atom and act like a "fingerprint" of the atom. If the atoms move around a lot, are hot, or interact a lot with each other, this fingerprint gets a bit blurred, but in our case, the atoms are free, extremely cool and nearly stationary. The fingerprint is sharp. The way the clock operates now is by moving the cesium through a microwave source - a bit like the kitchen microwave - with a gentle nudge of the laser. This microwave source is supposed to match exactly with one of the natural frequencies of the cesium - 9,192,631,770 hertz to be precise 9. It is worth mentioning that this number is exact- the second, by definition, is this amount of cycles of the cesium atom, no more, no less 10. The energy transfer from the microwave is optimal when its frequency exactly matches the natural frequency of cesium. After the cesium falls again and moves through the microwave a second time, a laser is used to induce fluorescence in the cesium atoms that have absorbed energy from the microwave. This fluorescent signal is measured. The time can now be measured by tuning the microwave so the signal is optimal. What is left then to determine the length of a second is to count 9,192,631,770 cycles of the tuned microwave.In summary, man has used two rather different methods of measuring time: astronomy for using large amounts of time, and, recently, atomic clocks. While the use of astronomical data has been invaluable for man for millennia, atomic clocks have made an extremely accurate measurement of time possible.
I hope this answers your question
Regards,
Bart Broks
Try the links in the MadSci Library for more information on Science History.