“…Except at low J , the theoretical values are then notably greater than the experimental ones, especially for K approaching or equal to J , and the calculation strongly exaggerates the K dependence of the broadening. This overestimation is not surprising since it is generally observed for molecules with a large dipole and/or quadrupole moment such as CH 3 Cl (35), OCS (52), or HCN (53). Finally note that, for the studied lines, the ratio of H 2 to He broadening is practically constant and equal to 2.7, whereas γ Ar /γ He has a mean value of 1.55; in contrast, self-broadening has a very different behavior.…”