I was planning  to demonstrate to my calculus class that the water surface in a spinning bowl takes the  shape of a paraboloid.   A large salad bowl on an old LP player worked very well. After the water settled into rotation, the parabolic surface became smooth and perfect.

I then placed a cork on the sloped surface, just out of curiosity. I had expected the cork to stay on an incline -- it would have been a fascinating thing to watch, and perhaps to imagine floating in a spinning pool -- what an amusement ride it would have been! The cork, however, did something else: it  slowly drifted towards the bottom of the paraboloid until it came to rest there. It must be the air resistance, I thought. Just to be sure, I covered the bowl with clear plastic wrap. With the cork floating near the wall, I turned the motor on. The same thing happened again! It wasn’t the air after all: the air settles into rotation sooner than the water; the air has a higher kinematic viscosity (i.e. the ``viscosity in relation to inertia”. Any disturbance in the air slows much faster than in the water. Air is a little less viscous than water, but also a lot less inertial.)

Here is the explanation of the puzzle. Imagine the blob of water that is to be displaced when the cork is placed in the water. Now instead of lowering cork in the water, imagine the water blob swelling and becoming a cork. The swelling will cause the center of mass of the blob to get closer to the axis of rotation (the water surface is sloped!). As the result, the centrifugal force will decrease and the swollen blob = cork will drift towards the axis.

Here is a remarkable thing: in a similar way, the Earth’s rotation creates the centrifugal pull upon the icebergs away from the poles towards the equator. This pull is, however, negligibly small.