Reading et al. <2003> propose a transformation which takes the rectangular coordinate system ZRT to the 'skew' PSH system. The transformation isolates P-, SV- and SH-waves of same slowness on the P-, S- and H-component, respectively, and removes the free surface effect. We have the following comments to their article: A) The transformation matrix in their paragraph has misprints propagated to the text in their Figure 1c. B) The LQT rotation <Vinnik, 1977> has no leakage of P-displacement into the Q-component, although a minor SV-leakage is present on the L-component. C) The differences between PSH- and ZRT-receiver functions are large, whereas the differences between PSH- and LQT-receiver functions are very small in theory and hardly ever significant in practice.