The viscosity, density, static dielectric constant and complex dielectric constant at 3.279 and 0847cm have been measured for nine simple liquids having molecules which show no intramolecular freedom of rotation, at temperatures from +50°C to the lowest to which the liquid could be supercooled. The viscosities of the liquids follow a free-volume relationship of the form q cc l/(u-b), where b is a constant for any liquid, rather than an exponential relationship to 1/T. The critical wavelengths A0 of the liquids have also been found. If these are assumed to follow a similar free volume relationship, AoTcc l/(u--b'), b' is in general different from b, the difference being greater for the more symmetrical molecules.Debye's theory 1 of the (microscopic) relaxation time z, of a polar liquid giveswhere 11 is the viscosity and p the " friction constant " of the liquid and a is its molecular radius. This equation implies that the temperature dependence of rPT is the same as that of y. Later theories have either led again to direct proportionality between z,T and q , 2 $ 3 or to an exponential relationship between z,T and 1/T (e.g., Glasstone, Laidler and Eyring 4). The last-named authors also predict an exponential relationship between yv and 1/T, where v is the specific volume of the liquid, and this or some similar exponential relationship is commonly predicted for the temperature dependence of viscosity.5Experimental measurements of dielectric properties lead to the macroscopic relaxation time z rather than zp. Various relationships have been proposed between z and zp. Collie, Hasted and Ritson6 have extended Onsager's theory7 to dielectric dispersion and find that z/z,-1. This ratio is not likely to be strongly temperature dependent, so that the above relations can be writtenwhere C, C' and B are constants.The experimental evidence for the temperature dependence of z and y is limited for most simple liquids with " rigid " molecules to the temperature range 0-60°C, and suggests 8 that the temperature dependence of viscosity is better represented by Batschinski's relationship,9 y = A/(u-b), where A and b are constants and Y is the specific voIume, than by any of the exponential formulae, and that analogously the temperature dependence of zT can be represented by p = 2kTz = A'/(ub'), where again A' and b' are constants, but b' is not necessarily equal to b, so that direct proportionality between zT and q is not implied. The physical interpretation of these relationships is that the fluidity l/y and the " rotational fluidity " l/p are dependent on a free volume, v-b or v-b', but the minimum volume 6' required for rotation is not necessarily the same as b, that required for flow.