What you Need
What to do
Tip the jug over to about 45 degrees
Pour the soap in a narrow stream down the slope, and move the stream around the side of the jug.
What may happen
You should find that occasionally the stream of soap bounces off the soap.
Sam Johnson also sent us this fantastic video of the same science in action: http://www.youtube.com/watch?v=_jx8ny8p2cY - Thanks Sam!
Why does it happen?
Normally when the stream of soap hits more soap, the two areas end up touching each other and surface tension pulls them together and they merge, so soap builds up.
But sometimes the stream of soap drags a layer of air along with it when it hits the stationary soap, this stops surface tension merging the two acts as a very good lubricant and separates the stream from the stationary soap, so that surface tension can't pull the two together.
This means that the jet flies across the surface for a certain distance, before the air layer breaks down and the two merge. If the jet hits a small ramp in underlying soap it will push the stream upwards again and it will bounce.
This effect is actually possible with normal liquids as long as the ramp doesn't collapse before the bounce has started, this means that the liquid needs to be quite thick and viscous to work, but this means that it doesn't flow as easily. It works a lot better with liquids like soaps, non-drip paint and washing up liquid, as they are shear thinning.
This means that they start off quite viscous so they don't run off your hand, or your paintbrush, but get thinner or less viscous when they are rubbed into your hands, or when you apply them to a wall. This makes the bouncing more likely, as any ramps that form in the soap last much longer, but the rapidly moving stream of liquid has a low viscosity so will bend easily.
A paper on the effect, though I am not convinced by their explanation
A paper done by students with a lot more detail
A paper describing the bouncing in normal liquids