As children, when we start to count, we usually use our fingers. We have ten of them and so ten and multiples of ten would seem to be the most natural counting system. Yet time, and the angles in a circle, don’t use the decimal system. So, why is this?

It goes back quite a long way to the numeric system of the Babylonians in Mesopotamia (‘land between rivers’, the rivers being the Tigris and Euphrates, and essentially what is now modern-day Iraq and parts of the bordering countries), which they inherited from the Sumerians. The Babylonians used cuneiform for recording, a system of mark-making whereby a wedge-shaped piece of wood or reed was pressed into wet clay. To create numerals, they used two simple shapes: first an upright single wedge-shape for 1 – repeated for 2, 3, 4, etc – and then the same shape angled twice to make a sideways v-shape for 10, and again repeated for 20, 30 etc, as on the right.

Why the number 60 was chosen as the base isn’t known, though it is much more easily divided than 10, which can be divided only by itself, 2 and 5. The Babylonians made astronomical calculations 4000 years ago based on 60.

It was the Egyptians who made sundials, which were simply a stake of wood placed in the ground, that divided the day into smaller parts by the length and direction of the shadow. Later the instruments were calibrated to divide the time between sunrise and sunset into 12 parts. At this time, ‘hours’ were about equal, but would vary according to summer or winter.

Water clocks, or *clepsydra*, from as early as the 1600 BC in Babylon and Egypt, and probably from 4000 BC in China, consisted of a bowl or dish whereby the flow of water into or out of was measured, which indicated the time. These clocks were the most accurate and most common clocks until replaced by pendulum clocks in the 17th century. (Right, above an original ancient Greek water clock and below a reconstruction.)

The Greek astronomer Eratosthenes (right) (c.276–194BC) divided a circle into 60 parts to create an early system of latitude. One hundred years later Hipparchus devised a system of longitude lines that ran north to south from pole to pole and also encompassed 360°.

Ptolemy, in about 150 AD, subdivided each of the 360 degrees into smaller segments of 60 parts. The first division of parts was called the *partes minutae primae*, or *first minute*, which then became simply *minute*, and the second segmentation was called the *partes minutae secundae* or *second minute*, or *second*.

Richard of Wallingford gave a clock to St Albans in 1336 (right). This had a large dial showing the sun and moon’s age, phase, and node, a star map and possibly the planets. As well as this it had a wheel of fortune, and showed the state of the tide at London Bridge. Bells rang every hour with the number of strokes indicating the time. It was only about 500 years ago, though, that minutes became commonly used. Clock displays before that divided the hour into halves, thirds and quarters, or even 12 parts, but never into 60 parts.

Mechanical clocks saw the introduction of minutes, and an early one is the Dover Castle clock.

It is driven by a foliot balance rather than a pendulum. Cogs on a wheel drive a level which leads to a weighted beam, and it is this that drives the clock. The foliot balance was later replaced by a pendulum.

Nowadays, of course, time is very accurate, as a second is defined as the duration of 9,192,631,770 energy transitions of the cesium atom (!). This atomic time, though, has to keep pace with astronomical time, and to do so, not every minute contains 60 seconds. About 8 minutes every ten years consist of 61 seconds.