It is pointed out that King and Sandler [1994] have overlooked that a trapped surface wave can be excited by the source for the case where the coating layer is a low-loss dielectric and the substrate is a highly con- To simplify the discussion, I will employ the notation familiar to the radio science community. The geometry of the model is the same as that shown in Figure 1 of KS, but I allow the vertical electric dipole to be in the air and located on the surface of the coating. The distance or range to the observer, also on the surface, is denoted by r. Following KS, the wave numbers for the air, coating, and substrate are k0, k l, and k2, respectively, and the adopted time factor is exp(-iwt).Restricting attention to the far field, defined by• Also where E 0 is the reference ideal ground wave field and is commonly known as the "inverse distance field" that varies as r-• exp (ikor) in the present context. In which is the form used by Norton [1937] for a homogeneous half-space, which, in the present context, corresponds to setting the coating thickness L equal to zero. In such a case the phase of the normalized surface impedance A is never less than -45 ø, in which case the phase angle of p is never negative. In the coated case, it is evident that the phase angle of A can be less than -45 ø, and, in fact, for a pure dielectric coating on a perfectly conducting 251