BROAD NE-VIII-LAMBDA-774 EMISSION FROM THE QUASAR PG-1148+549

We discuss the probable detection of broad Ne VIII lambda 774 emission from the z(e) = 0.978 quasar PG 1148 + 549, and we use spectral synth esis calculations to study the physical conditions in the line-forming gas. The theoretical predictions and the measured line wavelength bot h support Ne VIII lambda 774 as the most likely identification. Our ca lculations show that Ne VIII lambda 774 forms in hotter and more highl y ionized gas than previously recognized in the broad emission line re gion. If the gas is photoionized by a standard active galactic nucleus continuum, the observed Ne VIII equivalent width, the Ne VIII/O VI fl ux ratio, and the assumption of cloud stability imply ionization param eters 5 less than or similar to U less than or similar to 30 and tempe ratures 5 x 10(4) K less than or similar to T-e less than or similar t o 2 x 10(5) K. The large Ne VIII equivalent width also suggests that t he emitting clouds cover greater than or similar to 1/3 of the continu um source and have a total hydrogen column density N-H greater than or similar to 10(22) cm(-2). If the gas is instead collisionally ionized , Ne VIII could reside in stable clouds with equilibrium temperatures near 8 x 10(5) K. In either case, the Ne VIII-emitting clouds will app ear as X-ray ''warm absorbers'' if they lie along our line of sight to the X-ray continuum source.Line thermalization can greatly lower some of the line fluxes and alter the line ratios. For example, C IV lambd a 1549 can be thermalized easily, with doublet emission ratios lambda 1548/lambda 1550 near unity. However, our calculations indicate that N e VIII lambda 774 is not thermalized and that its doublet emission rat io should be lambda 770/lambda 780 x 2. Future observations of the wid ely separated Ne VIII doublet would test this prediction and constrain the space and column densities in the Ne VIII emitting region. Finall y, temperatures in the Ne VIII gas may be in the range needed for opti cally thin models of the UV continuum, i.e., the ''big blue bump.'' Ho wever, the diffuse thermal continuum flux from the Ne VIII-emitting re gion falls well short of the observed continua unless the measured lin es are severely suppressed by thermalization at densities greater than or similar to 5 x 10(12) cm(-3). This result supports Kriss's claim t hat the optically thin blue-bump models have the serious problem of ov erpredicting the UV metal line fluxes.