SEVERAL years ago it was found by one of us, together with P. Mariens1 and L. Thys2, that the experimental values found for the ultrasonic absorption coefficient is much larger than the theoretical value computed by means of the equation of Kirchhoff–Stokes (viscosity and heat conductivity). This deviation was explained as due to the lag in the establishment of equilibrium between the rotational and the translational energy. Since these measurements were made, new and more accurate measurements have been carried out by one of us together with R. Vermaelen3, and also on heavy hydrogen. All those determinations were made by means of an acoustical interferometer, in which different quartz crystals and different resonator tubes were used. Later, the existence of this relaxation for the rotational energy was proved by dispersion measurements, respectively by E. S. Stewart, J. L. Stewart and J. C. Hubbard4. On the other hand, the absorption coefficients found by the use of the acoustical interferometer have been criticized by P. E. Krasnooshkin5 and more recently by J. F. W. Bell6. Their argument is that an excess of absorption should be caused by the non-uniform vibration of the crystal surface, which produces excitation of the transverse modes of resonance in the interferometer tube. This excess of absorption can be determined by plotting the observed absorption coefficient αobs as a function of the pressure p, whereby (α0 is absorption coefficient corresponding to atmospheric pressure p0, and α1 is the parasite absorption coefficient). In order to check this theory, we have carried out a series of very accurate measurements (frequency = 500.0 kc./s.) as a function of pressure and temperature (liquid oxygen temperatures) in normal- and para-hydrogen and helium gas.