The water droplets are so small that they can be kept afloat by small wafts of upward wind, more like a feather in the air that will float and float and float but will never fall to the ground. The droplets are kept from running into each other by the same air, which acts more like a foam cushion that keeps a high-jumper from bumping into the floor. You can imagine these droplets danced around by the wafts and yet not bumping into each other.
There is something more to clouds. Clouds are also very cold. So cold that it can be below -40 deg C (now you are having second thoughts about being in the cloud). You must have read my earlier blog and wonder: how can water remain unfrozen? The answer also lies in the same blog. For water to freeze, it needs points of crystallization. Up in the clouds these are provided by suspended particles (dirt) in the air. But as you can imagine, a cloud living so high is also far from the dust and dirt of the Earth! So it is very hard to find such dirt that can make crystallization to happen. So water droplets remain water droplets but at temperatures much below zero.
So all these droplets go as clouds from one place to another, to another. When do they ever come down as rain? Will they fall on me? The first question is easier than the second. So, we will answer the first.
As we understand it, supercool water (water below zero deg C) becomes unstable as the temperature decreases. Only the slightest disturbance is needed for it to freeze. So lower the temperature the less impurities are actually needed to cause the disturbance. After all, even in the clouds there are impurities, though not as many as on the surface of the Earth.
We know that warm currents and cold currents all have a thing going of their own, independent of the clouds. Where they come from, how they come, we don't care. But if enough cold winds come to make the supercool water droplet's temperature drop below -40 deg C, then crystallization can start. So droplets turn to ice and slowly the surface of ice (flake rather) becomes the source of further disturbance and the crystallization spreads through the cloud. There is an additional effect that happens here. There are wafts of wind blowing between the droplets. This air is rich in moisture (it contains water vapors) and on meeting ice, the moisture in the air, condenses and then freezes to form more ice.
We remember well that the droplets are floating away from each other. So, as the ice starts to form, the cloud does not become a block of ice but a cloud of snow flakes, all ready to fall. As they fall, they melt and drop as rain droplets, only much bigger ones than the ones in cloud. How so? The air added more ice to them from the moisture it was carrying, remember?
Most of the drops evaporate as the fall down, due to the resistance from the air, just the same resistance we feel when riding a bike on the road. The clouds contained in a cube of one kilometer for each side, weight about a few hundred tonnes, so heavier the clouds the more drops that actually fall down on it, heavier the rain.
So that is the story of rain. You may prefer to stand in the rain than to be among the clouds.
We can tweak science to create rain in places that don't traditionally receive rainfall. If you remember the relation between supercool water and disturbance needed for crystallization, we needed a lot of cold drafts because there was not much "dirt" up above. Now this is the reason why clouds pass by without raining. So what scientists propose is to introduce "dirt".
What I cannot explain well is that one kind of "dirt", when introduced, can actually trigger crystallization at a higher temperature than some other "dirt". Silver Iodide can do this at -20 deg C where as usual crystallization needs well below -40 deg C. The process of introducing Silver Iodide in clouds to create the rain formation artificially is called seeding the clouds. Silver Iodide needs to be sprinkled fairly well inside the cloud, so crystallization will be uniform. Or rather, the upward drafts must help mix silver iodide fairly well in the clouds. There are hundreds of factors that need to go right so rain can be formed from the clouds. Here is two important ones
1. If the cloud isn't precisely downwind from the target area, the rain will fall in the wrong place anyway.
2. If the seeding is done at the wrong time, or on the wrong cloud, it may cause "cratering," or large holes in the cloud that cause it to fall apart. So an errant effort can destroy the clouds that might otherwise produce rain.
 Its not all water. There is some ice too. But not so many that can induce crystallization. But mostly water.
 The reason why dirt and dust are less at higher levels is rather simple. Anything to rise to that level against gravity must have really small weight, or can be vaporised in gas form. Most dirt don't become gas, and as the threshold of weight becomes smaller, fewer kinds of dirt can qualify.
Stratus, Cirrus, Cumulus are three common varieties of cloud.
I spare you from reading this rather longish article, by quoting some interesting excerpts from it:
The researchers have found: Rainfall from seeded clouds lasted longer than rain from unseeded clouds, the rainfall covered a larger area, and total precipitation was higher, sometimes even doubled. And in many cases results began just 20 minutes after the seeding.
Mr. Bruinties is now in the United Arab Emirates conducting a three-month feasibility study to determine whether conditions there are right for a cloud seeding program. Remember, not all clouds can be seeded.
Only certain clouds, early in their formation, are useful, and timing is everything.
Other seeding agents have been researched for the last 40 years, along with the importance of other conditions that help seeding to arrive at a better understanding of the process.
This resource provided the most material for the above write up. The following is another excellent very understandable discussion on everything there is to know about clouds. Recent News on Cloud seeding in Wyoming (Jan '06), and other ones in Australia, South Africa, Texas, and Israel.
Trivia: The Arctic ground squirrel hibernates at a body temperature below freezing--and yet doesn't freeze. The ground squirrel is able to do this because before hibernation it goes through a sort of internal purification that gets rid of any particle that might seed the freezing process. Liquids require such a particle (sometimes called a nucleus) in order to freeze. By purging itself the squirrel avoids becoming an ice cube, even if the temperature drops below freezing. (Link to Source Courtesy: Pradzie)