A lack of symmetry in the susceptibility tensor &khgr;yxwas found experimentally by Naiman and Lawrence over the spin‐flopping region in CuCl2·2H2O when the external magnetic field was less than 10−3rad away from the easy axis. This fact cannot be explained by considering the effective field acting on the two‐sublattice system to be composed of exchange, anisotropy, and external magnetic fields. Additionally, the Dzialoshinski‐Moriya interaction for CuCl2·2H2O appears analytically only as a shift in the critical field while the wander of the easy axis influences the shape but not the symmetry of &khgr;yx. The demagnetizing field has been included exactly using a 2×2 tensor, and it was found that for an ellipsoidal spin‐flopping specimen the major effect on &khgr;yxis to shift the value of the critical field. According to Joseph and Schlo¨mann, the demagnetizing tensor components for bodies with ``corners'' need not be constant, and this could possibly explain the observed asymmetry in &khgr;yx. To test this theory, single crystals were shaped into spheres and &khgr;yxwas remeasured. The essential agreement between the results before and after shaping lead us to conclude that demagnetizing effects which depend on the crystal shape are not important. Possible crystalline inclusions and/or domain formation may play a significant role.