For as long as I have known him, Peter Lynch has liked to play with pendulums. No doubt he started off with conkers on a piece of string, but by the time I met him he had progressed to the more adult pastime of gazing, Galileo-like, at any swinging object that might approximate a simple pendulum.
Then in more recent years he took to the elastic pendulum, a bob suspended by a spring and free to oscillate in both the horizontal and the vertical.
Now, Dr Lynch is assistant director of Met Eireann and a mathematician and meteorologist of quite formidable repute. So why, you may wonder, this fixation with a childish pleasure? Perhaps, as he follows the motion of the bouncing bob with darting eyes, he echoes Feste in Twelfth Night: When I was but a little tiny boy, With hey, ho, the wind and the rain, A foolish thing was but a toy, For the rain it raineth every day.
But, no. Peter's contemplation of the elastic pendulum has led him to conclude it is very difficult, if not impossible, to predict the weather.
Let's start with the pendulum. A swinging spring can oscillate up and down, or from side to side, and apparently when it settles it behaves in a fascinating way. After bouncing up and down, it gradually develops a horizontal swinging motion similar to a simple pendulum, and after further time, reverts again to bobbing up and down. This alternation between "swinging" and "springing" can be described with great precision mathematically; it turns out to be impossible to predict which direction the first horizontal excursion of the spring will take.
Regarding the atmosphere, we know thousands of feet above the ground in the midlatitudes, the air flows in a strong, steady stream from west to east around the world. But the flow is not straight; the motion has a sinuous, undulating character, and the behaviour of these long waves - Rossby waves - dictates the weather we experience underneath. If we could predict the life-cycle of Rossby waves, we could predict the weather for some time into the future.
Where is the connection? Peter Lynch noticed the mathematical equations governing the behaviour of the Rossby waves are identical to those of the motion of the bob of an elastic pendulum. We cannot predict what the pendulum will do - so how can we hope to be able to predict the weather?