An explanation is offered for the 30° tipping of sublattices which has been observed in the noncollinear antiferromagnet CsMnI3. In contrast to the ordinary 90° spin flop, this phase transition is due to nonlinear and relativistic corrections to the susceptibility.