I HAVE made a study of the dielectric constant of nitrobenzene as a function of temperature, using a method depending on the beats of two high frequency oscillation circuits, as described by M.