of the two 1-fluorine atoms are more nearly the same than if only one of the two rotational configurations alone were populated. The environments averaged over the oscillation, however, arc not identical. Increased frequency of oscillation with temperature elevation would have no effect on the chemical shift since presumably the frequency is sufficiently high even a t the lowest temperature studied to prohibit observation of fluorine resonances characteristic of the two configurations at frequencies of 30 MC. or less.Of the four substituted gem-difluoroethanes studied, only I ,2-dichloro-l, 1-difluoro-2-phenylethane exhibited the temperature dependent spectra discussed above. Spectra for the other ethanes a p peared to be teniperature independent over their liquid ranges. Of the lour substituted ethanes eamined, l,'-dichloro-l, l-difluoro-"-I'heiiylethane with two small chloro substituents rather than with two large bromo substituents would be expected to exhibit the lowest barrier to the 1-30' torsional oscillation postulated in Fig. 4, and consecluently to be most likely to exhibit a temperature dcpendcnt spectrum over the nccessiblc tcinperature range.U'rr.~rvc,rov, DEL The high resolution F' 9 magnetic resonance spectra of CIFa at 10, 30 and 40 Mc. are presented and interpreted in terms of Observed temperature dependencies of the spectri of CIFB and IF:, are .Activation energies for exchange of 4.8 kcal. for C1Fs and 13 kcal. for IF6 arc I t is concluded that fluorine exchange occurs in these halogen the structure of C,, symmetry for the molecule. interpreted in terms of fluorine exchange. estimated from the n-m-r temperature-dependence results.fluorides through a dimer intermediate.