ABSTRACT
A widely used method for determining refractive index postulates that the derivative of the angular profile for light reflected from the sample is maximum at the critical angle for total internal reflection (TIR). It is well-known that in turbid media this "differentiation method" yields errors in refractive index. Unexplained anomalies in previous error-calculations are eliminated if one uses a recent model of TIR which departs from traditional Fresnel theory. However we find that, in practical situations, the refractive index obtained by differentiation even after error-correction is significantly different from the best estimate for the refractive index obtained by curve-fitting the reflectance data. Thus the differentiation method lacks scientific validity in turbid media.
