What is the reason for the orbital movement of the Foucault pendulum? The Foucault pendulum, now in exhibition at the Paris Pantheon, is 67 meters high, with a period of oscillation about 16 s. Its oscillations last for more than one hour, when an employee manually resumes the initial amplitude.
The Earth rotation would lead the oscillation plane to twist for a local observer. One hour would be enough for that twist be clearly noted by the observer, what is facilitated by marks in a circle around the pendulum.
That oscillation, soon after the employee set ups, define a vertical plane. After some minutes however, a small orbital movement becomes apparent, which increases with the time. After one hour, the end points of the weight clearly change as expected, but its path, projected to the ground is no more a straight line, but seems a very elongated ellipse.
Why?


*

*Coriolis acceleration?

*Torsion in the cable between the weight and a fixed point at the roof?

*Imprecision of the set up by the employee?

*Other?

 A: What you describe is the plane of swing opening up. That effect is not specific to a Foucault pendulum, in general any pendulum that is free to swing in all directions is subject to it.
The effect goes back to the fect that it is very hard to create a pendulum setup that has exactly the same properties in all directions. 
The whole point of a Foucault pendulum is that it is left to swing for hours, so that the Foucault effect accumulates. However, that means that unwanted effect also have the time to accumulate. 
The wire has to be secured at the suspension point. Ideally the wire has the same flexibility in all directions. 
Let's consider the case where the wire is significantly more flexible in one direction, in such a way that that wire has the least flexibility in the direction perpendicular to it. Less flexibility gives a shorter period of oscillation, as there is more restoring force towards the midpoint of the swing. So there will be an unevenness of period of swing, depending on the direction of swing.
Now imagine what happens if you start a swing that is at 45 degrees to the planes of most flexibility and least flexibility. In effect that is a linear combination of two swings. And if that is left running for a prolonged time the two oscillations start to go out of phase. It's because of the accumulation that even the smallest difference gives rise to a noticable effect.
It is quite common for Foucault setups to include a device that suppresses that tendency of the plane of swing to open up. That device is called a 'Charron ring'.
