place, the propionic acid is held strongly. Propionic acid molecules not finding a strongly interacting site move on rapidly. This phenomenon contributes to the peak broadening and tailing so greatly manifest with propionic acid. As the size of the sample of propionic acid is increased, the limited number of strongly interacting sites becomes saturated; and the bulk of the propionic acid moves on more rapidly, thus resulting in the decrease in 7, observed. Effects of this type have been reported by Ottenstein (9).Column Poisoning. Previous work (/) has shown that 0.1µ samples of the test compounds injected successively into the asphalt column did not change the value of the interaction coefficient. This suggests that no irreversible reactions of the test compounds, or column poisoning, are taking place. However, because of the reproducibility problem and extreme "tailing" encountered with propionic acid, the possibility of irreversible reaction of this test compound with the asphalt was re-examined.Propionic acid was injected repeatedly on the sample asphalt column, using sample sizes 0.1 and 2.0 µ . When 0.1-µ samples were used, the 7, varied within the limits of the experimental error.When large 2.0-µ1 injections were made, the lg s dropped from 122 to 95; however, when a 0.1-µ1 sample was subsequently injected, the previous value of 122 was obtained. These data show no significant "poisoning" by propionic acid. The explanation for the lower values of lg obtained when 2.0 µ of propionic acid were used was given in the previous section. Water Displacement. Water in solute samples in other (9) D. M. Ottenstein, J. Gas Chromatogr., 6,129 (1968).systems (70) has a displacement effect on the solute, thus de-

The ozone yield from the radiolysis of gaseous oxygen has been measured to be G(O3) = 12.8 ± 0.6 at 1026–1027 eV g−1 s−1 and to be G(O3) = 6.2 ± 0.6 at 1016 eV g−1 s−1. The latter result depends on a new determination of the stoichiometry of the oxidation of iodide by ozone. It is proposed that the difference in yields between the two dose rates is due to differences in ion-recombination processes.

SUMMARY Irradiation (Cow source) of solutions of DL‐a‐tocopherol caused extensive destruction of this vitamin and gave rise to products similar to those obtained by autoxidation. Irradiation in saturated solvents (mineral oil, methyl myristate) destroyed more tocopherol than did irradiation in unsaturated solvents (methyl oleate, methyl linoleate). Destruction by 2.0 Mrad of irradiation increased with increasing tocopherol concentration, but the increase became negligible above 0.5% tocopherol in the saturated and 0.1% tocopherol in the unsaturated solvents.