The Compton profiles (CP) of H:('Z,') are calculated using the exact Born-Oppenheimer (BO) wavefunction for various internuclear distances R. By using the rotational-vibrational ( RV) state distribution factors, calculated according to the Franck-Condon ( FC) principle or measured from experiment, the vibrationally averaged values of C P for H;('L,') and D;(2Zl) in the adiabatic approximation are also calculated. Various theoretical values are given for comparison with the CP measured from H;-Ar collisions. The experimental data are most compatible with the exact BO values at R = 2.0 au but the data, with an error of IS%, fail to resolve various theoretical values at either small or large momenta q. It is important to have accurate experimental data (of absolute measurements) with an error of less than 10% in order to determine the accuracy of the theoretical values reported here.

The revised Bethe theory due to Inokuti et al for the inelastic scattering of charged particles by atoms and molecules is applied to Hl(lsog). In the 'reflection' approximation, the cross sections for the sum of all the discrete excitations oex, ionisation ui and total excitations o,,, given in an asymptotic expression with three constants (the so-called Born asymptote) instead of two constants (the so-called Bethe asymptote) are accurately calculated. The Bethe asymptotes are given for comparison with the less accurate ones reported by Peek. It is shown that the ionisation cross sections given by the Born asymptote including the free-electron exchange correction are most compatible with the experimental data, but the ones given by the Bethe asymptote are only valid at high incident energies. The calculated o,, and uto, are also shown to be in excellent agreement with the experimental data at incident energies above 50 eV. This work also points out that the obtained cross sections (oto,+o,), vex and o,,,, are very sensitively dependent on the accuracy of the rotational-vibrational ( R V ) state distribution of Hl(lsog) used to take the vibrational average. The calculated cross sections (olor+u,), including an error of 5% due to the vibrational average over the measured distribution, are also shown to be in agreement with two sets of experimental data. In general, the agreement between theory and experiment is consistent with that reported for one-electron atomic targets. Thus it is concluded that the assumptions made to account for the nuclear vibrational and rotation motions, such as the Born-Oppenheimer (BO) approximation, the reflection approximation and the closure argument, are valid for H: by electron impact.