I'm curious as to how you would write a Lagrangian for a torsion pendulum. Say you've got some disc that has mass and it's attached by a spring with spring constant k to a stationary "ceiling" of some kind and has gravity acting on it downwards. So, not only has it got rotational energy, but...