The cross section for photodetachment of C;-is investigated in the photon energy range 14000-20000 cm-' (1.75-2.5 eV). Sharp resonances due to autodetachment are observed at photon energies corresponding to transitions between high vibrational levels of the C;-X 2~ state and high vibrational levels of the C;: B 2I.+ states. The resonances are narrower than 6 GHz, and those arising from the v = 5 level of the B state are about 1/10 as strong as those arising from the v = 6 and higher levels. In addition, the signal from the allowed direct photodetachment of C;-X state, which would produce a smooth background cross section, is not observed, indicating that it is less than the l()l peak to background contrast ratio. A rotational analysis of the nine bands of the C;-B-X transition observed in this study, coupled with previous measurements provides anew, more precise set of spectroscopic constants for these states. These constants are then used to generate RKR potential curves for the B and X states of C;-. Based on all C;data, the only consistent interpretation of the observations is that the electron affinity of C 2 is bounded by 3.374 eV ~EA(C,)~3.408 eV. The autodetachment rate into (C 2 a 3il. +e) is deduced to be much faster than the rate into (C 2 X 'I,+ +e), probably as a result of strongly R dependent configuration interaction in the C;-B state. The weakness in the direct photodetachment process is attributed to poor vibrational overlap between the initial and final states.

No abstract

ery close to P", and the measured deviations from 1 may reflect the influence of the longitudinal field on stimulated scattering rather than on SF.In order to see the true critical power, we resorted to mode-locked laser operation (open circles). At 1-nsec pulse duration stimulated scattering was weak and disappeared completely at large 8. An obvious increase of P"TM/P"Ts was found this time which depends on 8 in the same way as the calculated curves. Most interesting, however, the experimental values are in between the areas ascribed to electrostriction (1.00) and its superposition with the optical Kerr effect (0. 15). They are clearly away from the region of the optical Kerr effect alone (-0. 50).Hence, one might tentatively conclude that electrostriction does in fact contribute significantly to the SF in our investigation even at pulse durations as small as 1 nsec. This result would be in good agreement with the focal-zone transient times estimated above. But it should be kept in mind that various details of the SF process have been omitted in our investigation, and more work has to be done in order to clear up completely the role of the mechanisms involved. The mere fact, however, that there is a difference in the SF of TEp, and TMp, light is clear evidence for the influence of longitudinal field components on the nonlinear propagation of light; a phenomenon which, to our knowledge, has never been observed before.The author gratefully acknowledges the contribution of E. Courtens to the mathematical. formul, ation of the problem, and the technical assistance of R.