A few comments are in order. First, Eqs (1) describe a free rotation of a rigid body. It seems natural to interpret the extra terms in (7) as due to a torque. Second, although the system (7) is Liouville integrable, it is not always exactly solvable. In general (6) cannot be resolved to express two variables from the triple (x1,x2,x3) as explicit functions of the third one. Third, in some instances (7) can be linked to (1) by a coordinate transformation. If the equations