Thanks Nick...
however thinking about it, my introduction of the concept of the net weight (as measured by a spring balance) may cause some confusion with regard to how stability works, to try to correct that...
Even when it's in the water the boat has weight, but that weight is exactly cancelled by the buoyancy force. So the spring balance registers zero. That buoyancy force is created by the boat displacing it's own weight of sea or lake water which indeed is "Archimedes Principle" as David says. The buoyancy force is always equal to the weight of the boat because if it is less, the boat sinks lower until enough water has been displaced for the buoyancy to counteract the weight.
The weight acts through the centre of mass (also known as the centre of gravity) and the buoyancy force acts through the centre of buoyancy which is located at the central point of the part of the hull's volume that is under water. People often don't realise that, apart from sailing yachts with heavily ballasted keels, the centre of mass is normally higher than the centre of buoyancy. That is particularly true for large ships where there is a lot more above the water line than below it (as we saw with that car transporter which ended up on the Brambles bank). Ships only remain upright as long as you don't tilt them too much; beyond a certain point there is no return!
What matters for stability is how the centre of buoyancy moves in relation to the centre of mass. I really need to draw some diagrams to illustrate this. When the boat is upright the two centres are directly in a vertical line. As the boat heels the centre of buoyancy moves outwards towards the downward side of the boat. Provided the ballast (including the crew) does not move across the boat, the centre of mass stays in the same place in relation to the rest of the boat. However if the centre of mass is higher than the centre of buoyancy, the heeling of the boat will also move the centre of mass further across towards the downward side of the hull. Imagine a boat where the centre of mass is some way up the mast - in that case it's obvious that it will moves outwards as the boat heels.
For stability it is important that during the heel the centre of buoyancy moves out faster than the centre of mass. This is achieved by keeping the centre of mass as low as possible (and the boat as beamy as possible). As long as the centre of buoyancy is further out, the two forces, of gravity and of buoyancy, act to rotate the boat back into an upright position.
The actual force of this righting tendency depends (as you might expect) on how great the distance is between gravity pulling down and buoyancy pushing up. The way this changes with the angle of heel is what is shown in a "gz" curve where the g is gravity and z is the distance. But for a given distance (that is a for given value of gz), it also depends on how big those two forces are, and that depends on how heavy the boat is. According to Matt's Water Craft magazine article ("What's the point of Water Ballast?", Water Craft 109, 30-33, Jan/Feb 2015) the water ballast in a Bay Raider increases the gz by about 30% (by lowering the centre of mass). But the tendency for the boat to come upright is more than doubled mainly due to the extra weight of the ballasted hull (pulling downwards) and hence the larger buoyancy force (pushing upwards).
If you add ballast to a boat in the form of a closed heavy box of a certain weight, it doesn't matter whether what's in the box is water or lead (provided neither the water or lead can move around in the box). The advantage of lead is you can use a smaller box and you can get it lower down in the hull. The advantage of water is you can add it or remove it more easily.
...wow! I only meant to write a sentence or two! If it's thought worthwhile I could take up Andy's challenge and create a library article with some diagrams. Matt covered much of this in his Water Craft article, although, I imagine to avoid confusion given limited space, he didn't introduce the concept of "centre of buoyancy".