The rainy season is with us, but what makes the difference between persistent drizzle and sudden downpour? Here is your historical guide to atmospheric precipitation.

Aristotle took a philosophical view of rain. As the sun moves across the heavens, he believed, its motion generates heat which causes water to evaporate and rise. But when the sun moves away again, "the vapour cools because its heat is gone ... and condenses again and turns from air to water. And after the water has formed it falls down again to the earth."

He pictured the process as a circular process of rising and falling moisture that followed the course of the sun. "We must think of it as a river flowing up and down in a circle and made up partly of air, partly of water."

But he did get one thing spot on: "When the water falls in small drops it is called a drizzle; when the drops are larger, it is rain."

From Aristotle's day until recent times, a simple model of raindrop formation seemed to explain everything: as warm air cooled, its water vapour condensed into droplets, small enough to be kept up by air currents; as those droplets bumped into each other, they coalesced into larger droplets and fell as rain.

As clouds came to be studied in greater detail, however, this simple view became unsustainable. For one thing, water droplets are just as likely to fragment when they bump into each other as they are to coalesce. Another problem was that the water droplets in clouds are only about a hundredth of the diameter of raindrops - which would necessitate a million collisions for one drop of rain - and nobody could explain why they hung around as tiny droplets for a long time, then suddenly started bumping into each other at such a rate. One idea was that droplets coalesced quickly by electrical attraction, but it seems they are too far apart, and their electrical charges too small, for this to be the explanation.

Another idea was that turbulence within a cloud caused warm droplets to evaporate and cold droplets to grow as they did so. But the temperature of cloud droplets seems generally wrong for this to happen.

The mechanism of raindrop growth that is now generally accepted (called the "Bergeron-Findeisen theory") is based on two important discoveries. First, that droplets coalesce far more easily around microscopic particles in the air than they would ever do in pure, clean air. And second, that condensation occurs at a lower humidity over ice than it does over water. So if ice crystals and cold water droplets exist together in a cloud, the water tends to evaporate, leading directly to a deposit of more ice on the crystal.

So raindrops begin with the formation of water droplets around hygroscopic nuclei - very fine particles of soil, dust or some other impurity in the air. As the cloud rises through colder air, the droplets freeze into minute ice crystals, which grow, either by proximity with evaporating water droplets, or by collision with other ice crystals. Crystals coalesce into snowflakes, whose weight causes them to fall through the cloud. As they fall through enough air that is above the freezing point of water, they melt into raindrops.

Since some heavy rain (especially over tropical oceans) seems to come from clouds whose temperature is above freezing point, this theory does not explain everything, but in general it fits most observed facts. The idea of droplets forming about minute nuclei is the basis of successful cloud-seeding, which introduces dry ice or silver iodide particles into very cold clouds to encourage the formation of ice crystals.

This also provides an explanation of the use of seaweed in weather prediction. The surface of seaweed is rich in sea salts which are particularly good at encouraging condensation. Damp seaweed means over-saturated air.