On the Quantitative Use of the Foucault Knife-Edge Test

The work described in the foregoing paper made the development of a quantitative method of surveying surfaces necessary, that could be used for local measurements and on nonsymmetric surfaces. The theory of the method and the experimental technique are described. The accuracy of the survey is better than 1/20 wave-length with visual subjective observation and could be carried up to 1/100 wave-length with objective photographic measurements. The errors of the Foucault knife-edge test, especially the one produced by parallax, are discussed. Ways of reducing or eliminating them are described.