The effect of viscosity on a second‐order acoustic field has been studied when the time independent relation between the excess pressure and the excess density is replaced by a relaxational one. The method used is similar to that of Eckart except that some terms explicitly neglected by him are included. The introduction of relaxation indicates that steady flow caused by higher order terms is not only caused by the viscosities but is affected by the relaxational relation between the pressure and density. The calculations also indicate that quartz wind measurements are not equivalent to absorption measurements because the temperature variations of the viscosity coefficients enter into the flow problem. If these calculations are correct they suggest interesting problems on classical liquids and give a method to distinguish between relaxational and viscous effects. The introduction of relaxation also explains simply Liebermann's measurements on ethyl formate.