Lu, I think it's correct to say that the centripetal forces hold the subject in his seat, and as the centrifuge starts to move, his bodily fluids are accelerated along with the rest of him by forces from the walls of arteries, etc.
As the centrifuge whirls, the things that are fixed to it, such as the subject's body, are continually forced to move in a circle by centripetal forces.
The fluids are not directly 'connected', though, so at one point in time, they will be moving in a particular direction (in a straight line tangential to the circle). The subject's body, however, is moved off that tangential path, and so the fluids will tend to continue along their straight path as far as they can, i.e. flowing along paths of least resistance towards the outside of the circular path.
Then, when all the available spaces have been filled, so to speak, you will end up with the person going round in a circle and his bodily fluids bunched up towards the local gravity (the outside of the circle).
So really, it's not a force that's making the fluids go to the outside, it's just their natural tendency (as with any particular substance) to keep going in a straight line until acted on by another force (in this case, centripetal force provided by the push towards the middle of the circle provided by the subject's outer 'walls'.
My two bob's worth!
Gaseous,
I think what you're talking about is just that less dense things will float on top of dense things (with 'downwards' being the direction of local gravity).
The bus goes around the corner. The air, not being directly connected to the bus, tries to keep going in a straight line, until it hits the walls and has to bunch up.
Now because there is more air towards the outside of the turn (not a lot, mind you, but enough to have an effect), the helium balloon (being less dense) gets crowded out of that side and deflects to the inside of the turn.
You can make a simple accellerometer using this principle - put a bit of styrofoam in a jar full of water and put the lid on. Whichever way the jar accellerates, the foam moves that way too.
[ 28 December 2001: Message edited by: Arm out the window ]</p>