Composition of interstellar clouds in the disk and halo. 2: Gamma(sup 2) Velorum
High-resolution observations of gamma(sup 2) Vel with the Goddard High-Resolution Spectrograph (GHRS) echelle on the Hubble Space Telescope reveal the presence of seven narrow absorption components, with LSR velocities between -23 and +9 km s(exp -1). Three of these show column density ratios N(S(++))/N(S(+)) and N(P(++))/N(P+)) of about 1 or more, and can be identified as H II regions, while the other four are H I regions, consistent with the O I profile and with the overall H(sup 0) column density of 5.9 x 10(exp 19) cm(exp -2), given the usual assumptions that S is undepleted while O has a depletion D(O) = -0.3 dex. The depletions of Fe, Si, and Mn, which could be measure accurately for two of the four H I regions (components 6 and 7), differ somewhat from the values of D(sub ws) found for slowly moving warm clouds in HD 93521; in particular, for the component at 4.0 km s(exp -1) (No. 6), abosolute of D exceeds absolute of D(sub ws) by 0.1-0.4 dex, while for that at 9.3 km s(exp -1) (No. 7), absolute of D equals absolute of D(sub ws) on the average. The observed ratio of Fe + Mg atoms to Si atoms in the grains of component 6 is 2.04 +/-0.10, consistent with an olivine grain composition; the Fe/Mg ratio is 1.5 +/- 0.2. The electron density in component 6, determined from the C II(sup *) feature, is 0.075 +/- 0.013 cm (exp -3), about two-thirds of that found for clouds of this velocity in HD 93521. In the two conspicuous H II regions, components 3 and 4, n(sub e), determined from the Si II(sup *) feature, is about 1 cm(exp -3). From the column density of S(+) + S(++) in these two components, the total H II path length is about 40 pc. With the radius of a wind-blown bubble around gamma(sup 2) Vel set equal to 60 pc, the effective Stromgren radius is about 100 pc, requiring that T approx. equal to 50,000 K for the Wolf-Rayet component of the gamma(sup 2) Vel binary. Since zeta Pup is a comparable source of ionizing radiation, this temperature is an upper limit. The profiles of the strongest H2 absorption features, from Copernicus archives, indicate that the absorbing molecules have a mean velocity identical with that of the strongest H II component (No. 4). We have no explanation for the possible presence of these H2 molecules in a region of ionized H. Alternatively, the H2 profiles can be explained by molecules in the two adjacent (in velocity) H I regions, components 2 and 5, provided their H I gas has densities and temperatures typical of normal cold clouds. The GHRS data show absorption by highly ionized atoms Si(3+) and C(3+), N(4+) in broad features, in addition to the narrow-line absorption by Si(3+) and C(3+) observed in the dominant H II components, Nos. 3 and 4. The broad C(3+) and N(4+) features have widths corresponding to T in the range (4-8) x 10(exp 5) K, consistent with the broad O(5+) line shown in Copernicus data. Despite some observational uncertainties, the ratios of column densities in the broad C(3+), N(4+), and O(5+) features agree to +/- 0.1 dex with theoretical values for warm gas, heating and evaporating by thermal conduction from an adjacent hot region. Outward evaporation from an isolated cloud in a hot ambient gas cannot be distinguished, on the basis of these data, from inward evaporation of a warm shell, compressed by an expanding, hot stellar-wind bubble. For several halo stars, the C IV/O VI ratio has a quite different average value, perhaps consistent with cooling of infalling hot gas instead of conductive heating and evaporation.