Alternative derivation of electromagnetic cloaks and concentrators

In section 5.1 of the paper, a mistake in our algebra gave a relative permeability dyadic for a spherical magnetostatic cloak (ω = 0) that was different from that for a spherical time-harmonic cloak (ω > 0). Correcting the algebra, we find that the same dyadic is required for both static and time-harmonic cloaking, a result that can be shown to hold not only for spherical cloaks but also for all other cloak shapes as well. Also, in the last sentence in the second and third paragraphs of section 2.1, the words 'tangential polarization' and 'tangential polarization densities', respectively, should be replaced by the words 'the normal components of the E and H fields'. It should be noted that cylindrical cloaks, unlike spherical cloaks, can also have delta functions in the tangential polarization densities at their inner surfaces.