1. Jun 18, 2013

### davidchen9568

I do not understand the concept. Let the centrifugal forces on the two objects are F1 and F2. The rod will not rotate because net torque is 0. That is, F1r1=F2r2. That Does that mean we have to carefully adjust the positions of objects (before experiment) so that m1gr1=m2gr2? Then we conclude F1/m1=F2/m2 (to very high accuracy)?

If it's like what I said, how can phycisits obtain very high accuracy? Even if m1gr1 does not exactly equal to m2gr2, I believe the system can balance somewhere, thus create a error far larger than 1/20,000,000.

So how was that done?

Article on Wikipedia: http://en.wikipedia.org/wiki/Eötvös_experiment

2. Jun 18, 2013

### 256bits

The system does balance somewhere and you record it.
Then you switch places of the two masses on the rod, or rather it is easier to turn the whole apparatus 180 degrees around.
A null result, meaning that the rod does not twist in the opposite sense, confirms the equivalence principle.

3. Jun 18, 2013

### TurtleMeister

G1 and G2 depends of the passive gravitational mass of m1 and m2 respectively. F1 and F2 depends on the inertial mass of m1 and m2 respectively.

For example: if the passive gravitational mass of m1 were made to be less than the inertial mass of m1, then gr1 would become greater, and gr2 would become less. (The balance point would change.) So applying the force F1 and F2 would cause the rod to rotate.

Edit:
Greater accuracies and sensitivities have been achieved since the EotVos experiments. You may want to check out http://www.npl.washington.edu/eotwash/.

Last edited: Jun 18, 2013
4. Jun 19, 2013