Abstract
M. Green has connected the shear and bulk viscosities with the fluctuations of the virial. It is shown here that the virial theorem is correct at every moment, including the fluctuations, if the integral of the pressure over the volume is taken. For monoatomic fluids the bulk viscosity is calculated after introducing an unknown correlation length; for fluids with internal degrees of freedom or with two-state structure, the usual equations follow very easily. For the shear viscosity one finds a formula similar to that of Eyring by making some assumptions about correlations.