van Hulst et al. Reply: In Ref. [1] we have presented the first visualization of the phase evolution of the optical field of propagating guided modes, using heterodyne interference photon scanning tunneling microscopy (PSTM). The Comment of Vohnsen and Bozhevolnyi [2] focuses mainly on our observation of phase singularities that are expected at positions of destructive mode beating. In particular, they stress the fact that the singularities originating from TE 00 -TM 00 mode interference cannot

more »

... occur, as the local intensity is not zero. Actually this notion is not particularly new. Already in the original Letter [1] we pointed out that ordinarily TE and TM modes do not interfere due to their perpendicular polarization. It is the polarization conversion, inherent to near field optics and subwavelength probes, that leads to a quasi-interference of the mutually perpendicular fields. We analyzed the subwavelength polarization coupling and resulting interference in a previous experimental study [3] . As a result the singularities occur at points where the amplitudes of the local TE and TM fields, as projected on the detection system, are equal (nonzero) and have opposite signs. The position of TE-TM singularities is a measure for the relative amplitude of the TE and TM modes [4] . Evidently the singularities shift in a transverse direction depending on the ratio between TE and TM excitation and the amount of polarization coupling.