“…That is, the unit normal vector crossed into the dyadic magnetic Green's function, then multiplied by the surface current, yields the same result as the dyadic magnetic Green's function multiplied by the surface current, followed by a cross-product multiplication by the unit normal vector. Equation (22) is a rather remarkable result, as it allows us to write down by inspection the new Green's function for the finite-sized and finite-conductivity metal strip in our anisotropic layered structure (24) We will not pursue this approach further, other than to say that, although the form has been presented, it does not prove that, in fact, such a closed-form representation can be found. Instead, in order to have some specific rules for constructing the average, resort to a procedure found in [22].…”