An acoustic redetermination of the gas constant

The recent value of the gas constant, determined acoustically by Quinn et al. (1976), suffered from a hitherto unsuspected systematic error which was attributable to a slight nonlinearity in the transducer used to excite their acoustic interferometer. This effect, which led to low pressure curvature in their acoustic isotherm, has now been corrected for. When allowance is also made for high pressure curvature due to the third acoustic virial coefficient as determined by new measurements at higher pressures, a linear isotherm results. It exhibits the slope recently predicted by Rowlinson & Tildesley (1977) on the basis of pressure-volume-temperature measurements and of new work on the intermolecular potential of argon. The nonlinearity corrections having been made, the original work is entirely compatible with the new high pressure results and the two sets of data have been combined to form a new isotherm. A value of the gas constant $(R=8314.48$ $\text{J K}^{-1}\text{kmol}^{-1})$ has been derived from the intercept of this isotherm, which has an overall uncertainty of only 0.21 J K$^{-1}$ kmol$^{-1}$ or, fractionally, $25/10^{6}$. This value of $R$ $\text{is}$ $8/10^{6}$ above the conventionally determined, accepted value of Batuecas, $37/10^{6}$ above the new, conventionally determined value of Pereira Cordido and $1/10^{6}$ below that recently obtained by Colclough (1979a) from the comparison of gas and acoustic thermometry in the range 4.2-20 K. All these values agree with the present one to within their respective random uncertainties.