The variation of the dielectric constants of some organic liquids with frequency in the range 1 to 10
3
Kilocycles
Whilst it is recognised that the dielectric constant of liquids changes in the frequency range 10 4 - 10 5 kilocycles per second in accordance with the theory of Debye, no systematic examination of the variation of the dielectric constant of simple liquids with frequency appears to have been made at frequencies below 10 3 kc. per second. Exception must be made of the work of Fricke* who showed that the dielectric constant of blood did not change in the range 0.8 to 4500 kc., and of that of Bryan who recorded no change in the constant for xylene and an increase in the constant for nitrobenzene in the range 200 to1200 kc. In the case of chloroform and benzene a number of independent determinations have been made, eachat a fixed frequency. The values of the constants, however, at frequencies less than 1000kc. fluctuate considerably, for benzene the divergence between the extreme values is about 2·0 percent, of the mean, for chloroform about 12·5 percent. It is of importance, therefore, to establish whether these fluctuations are due to experimental error or the variation of the constant with frequency. The experiments now described were planned preliminary to work at higher frequencies; measurements of the dielectric constant and of the conductivity of a number of liquids have been made in the frequency range 1 to 10 3 kc. Attention has been directed to examine the variation of these quantities with frequency rather than to obtaining their absolute values. Owing to the illness of one of the authors the work had to be discontinued before the original programme had been completed, nevertheless, in view of the increasing importance of the subject the results appear to be of sufficient interest to merit publication. Since the data now reported were obtained, an extremely careful determination of the absolute value of the dielectric constant for benzene at 1000 cycles has been described by Hartshorn and Oliver ( loc. cit. ). They report no change in the constant in the audio frequency range, that is, presumably, below 5 kc., and thus confirm, in part, the data now presented.