Optical Measurement of Non-uniform Electric Field Vector Distribution in a Dielectric Liquid Using Triplet Measurement System

to identify the nature of the unstated approximations used in that work. The title clearly states that the subject of the paper is “optical measurements of non-uniform electric field” yet equations (1)-(9) are only valid for uniform electric fields whose direction is constant along the light path. These equations are presented as if they x e directly applicable to arbitrary non-uniform electric field distributions. In particular no attempt has been made to show how E and a in equations (l)-(9) relate to the actual non-uniform electric field distribution under investigation. In fact the rigorously correct mathematics for descrihing Kerr electro-optic field mapping measurements in three dimensions when the direction of applied electric field changes along the light path requires new solutions to Maxwell’s equations as developed in [2]. These solutions show that two measurable parameters, analogous to E and in equations (l)-(9) 2. and which we call y and a in [2], exist only under certain conditions: if the geometry is axisymmetric or approximately if the dielectric medium is not highly birefringent. Since the geometry used by Ihori et al [ I ] is axisymmetric we do agree that actual measurements of two parameters which are analogous to E and pare meaningful and valid hut these parameters are not the electric field magnitude and direction. However we also disagree that such measurements and the numerical algorithm used by Ihori et al [ I ] are directly applicable to arhitrary 3D geometries. At best, their approach is a method for reconstructing axisymmetric geometries hut cannot he applied in their present form to the full 3D problem. In fact even extending 2D Kerr measurements to axisymmetric geometries is difficult. The distinction between Kerr measurements in 2D and axisymmetric geometries is given in [3]. The main contribution of the paper by Ihori et al [l] appears to be the “triplet measuring system” which, according to the authors’ claim, reduces the amount of measurement time for the construction of the electric field vector distribution without sacrificing much accuracy. We do agree that making three measurements at the same time instead of making three measurements in succession reduces the amount of measurement time and could he valuable for truly three-dimensional geometries. However we find conclusions based on an axisymmetric geometry to he very 1 T . this communication is to point out the mis-statements and