Remarks on the membrane and buckling eigenvalues for planar domains

I present a counter-example to the conjecture that the first eigenvalue of the clamped buckling problem in a planar domain is not smaller than the third eigenvalue of the fixed membrane in that domain. I also prove that the conjecture holds for domains that are invariant under rotation by angle $\pi/2$.