How does a seismometer record an earthquake?

A seismometer is a machine that keeps one heavy weight perfectly still while the whole world shakes around it. That sounds impossible, but it is just inertia, the stubbornness every heavy object has about starting or stopping its motion.

Picture a weight hanging from a spring, and the spring attached to a frame bolted to the ground. When an earthquake shakes the ground, the frame shakes with it. The weight, thanks to its inertia, tends to stay where it is. So the frame moves and the weight lags behind. The difference between the two is the ground motion. In the old versions a pen was attached to the weight and a drum that turned with the frame, so the pen drew the shaking as a wiggly line, a seismogram.

Ground moves in three directions, so observatories usually record three axes: north-south, east-west and up-down.

The idea is old. In the year AD 132 in Han dynasty China, Zhang Heng is said to have built a bronze vessel about two metres across, with eight dragon heads around the top, each holding a bronze ball. When the ground shook, a dragon's mouth opened and dropped its ball into a bronze toad below with a sound, supposedly showing which way the quake came from. It was a seismoscope, a device that says that something shook rather than draws how.

Modern seismometers are electronic, but the principle is the same one: a mass that stays put, and a very careful measurement of how the frame moves relative to it. They have been placed on the Moon and on Mars.