We show that in interpreting the conoscopic interference fringes, one should exercise care in employing approximate expressions which fail for certain crystal cuts. In this paper, we study64°- and128°-rotatedY-cut andZ-cut LiNbO3wafers. We show that the error made in using the approximate formulae for the samples is more than 25&percent; and that one has to use exact formulae in order to attain quantitative agreement with the experimental data. ©1997 American Institute of Physics.