Problems of the rotating-torsion-balance limit on the photon mass

We discuss the problems (and the promise) of the ingenious method introduced by Lakes, and recently improved on by Luo, to detect a possible small photon mass $\mu$ by measuring the ambient magnetic vector potential from large scale magnetic fields. We also point out how an improved ``indirect'' limit can be obtained using modern measurements of astrophysical magnetic fields and plasmas and that a good ``direct'' limit exists using properties of the solar wind.