The phase of supersoft source (SSS) emission of the sixth recorded outburst of the recurrent nova RS Oph was observed on days 39.7 and 66.9 after outburst with Chandra and on day 54.0 with XMM-Newton. A $35 s period on day 54.0 originates from the SSS emission and not from the shock. We discuss the bound-free absorption by neutral elements in the line of sight, resonance absorption lines plus self-absorbed emission-line components, collisionally excited emission lines from the shock, He-like intersystem lines, and spectral changes during an episode of highamplitude variability. We find a decrease of the oxygen K-shell absorption edge that can be explained by photoionization of oxygen. The absorption component has average velocities of À1286 AE 267 km s À1 on day 39.7 and of À771 AE 65 km s À1 on day 66.9. The wavelengths of the emission-line components are at rest wavelengths, as confirmed by measurements of nonYself-absorbed He-like intersystem lines. We found collisionally excited emission lines from the radiatively cooling shock at wavelengths shorter than 15 8 that are systematically blueshifted by À526 AE 114 km s À1 on day 39.7 and are fading. We found anomalous He-like f /i ratios, which indicates either high densities or significant UV radiation near the plasma where the emission lines are formed. During the phase of strong variability the spectral hardness light curve overlies the total light curve when shifted by 1000 s. This can be explained by photoionization of neutral oxygen in the line of sight if the densities are of order 10 10 Y10 11 cm À3 .

Aims. We compare the observed and theoretical parameters for the quiescent and outburst phases of the recurring nova T Pyx. Methods. IUE data were used to derive the disk luminosity and the mass accretion rate, and to exclude the presence of quasi-steady burning at the WD surface. XMM-NEWTON data were used to verify this conclusion.. These values were about twice as high in the pre-1966-outburst epoch. This allowed the first direct estimate of the total mass accreted before outburst, M accr =Ṁ pre−OB ·Δt, and its comparison with the critical ignition mass M ign . We found M accr and M ign to be in perfect agreement (with a value close to 5 × 10 −7 M ) for M 1 ∼ 1.37 M , which provides a confirmation of the thermonuclear runaway theory. The comparison of the observed parameters of the eruption phase, with the corresponding values in the grid of models by Yaron and collaborators, provides satisfactory agreement for values of M 1 close to 1.35 M and logṀ between −8.0 and −7.0, but the observed value of the decay time t 3 is higher than expected. The long duration of the optically thick phase during the recorded outbursts of T Pyx, a spectroscopic behavior typical of classical novae, and the persistence of P Cyg profiles, constrains the ejected mass M ign to within 10 −5 −10 −4 M . Therefore, T Pyx ejects far more material than it has accreted, and the mass of the white dwarf will not increase to the Chandrasekhar limit as generally believed in recurrent novae. A detailed study based on the UV data excludes the possibility that T Pyx belongs to the class of the supersoft X-ray sources, as has been postulated. XMM-NEWTON observations have revealed a weak, hard source and confirmed this interpretation.

The tenth recorded outburst of the recurrent eclipsing nova U Sco was observed simultaneously in X-ray, UV, and optical by XMM-Newton on days 22.9 and 34.9 after the outburst. Two full passages of the companion in front of the nova ejecta were observed, as was the reformation of the accretion disk. On day 22.9, we observed smooth eclipses in UV and optical but deep dips in the X-ray light curve that disappeared by day 34.9, yielding clean eclipses in all bands. X-ray dips can be caused by clumpy absorbing material that intersects the line of sight while moving along highly elliptical trajectories. Cold material from the companion could explain the absence of dips in UV and optical light. The disappearance of X-ray dips before day 34.9 implies significant progress in the formation of the disk. The X-ray spectra contain photospheric continuum emission plus strong emission lines, but no clear absorption lines. Both continuum and emission lines in the X-ray spectra indicate a temperature increase from day 22.9 to day 34.9. We find clear evidence in the spectra and light curves for Thompson scattering of the photospheric emission from the white dwarf. Photospheric absorption lines can be smeared out during scattering in a plasma of fast electrons. We also find spectral signatures of resonant line scattering that lead to the observation of the strong emission lines. Their dominance could be a general phenomenon in high-inclination systems such as Cal 87.

Abstract. In the framework of a phenomenological study of the ultraviolet properties of classical novae in outburst, we have selected 12 objects among the best monitored at low resolution with the IUE satellite, and studied the temporal evolution of the ultraviolet continuum and of the O i 1300Å line flux during the early post-outburst phase. We confirm that the UV flux maximum takes place systematically after the visual maximum and that its time delay is a linear function of t3. A linear dependence on t3 is also found for the duration of the UV outburst and for the time the O i line flux reaches a maximum. This latter time marks the start of the transition phase to nebular conditions. Within the uncertainties imposed by the sample of objects used and by the observational errors, these results suggest a quite homogeneous behaviour of classical novae in the ultraviolet range.

32 erg s −1 . The September 11 and 17 EPIC spectra can be represented by a power-law model with photon indices of 2.25 ± 0.15 and 1.42 ± 0.09, respectively. Thus, the 0.5−10 keV spectrum hardens with increasing intensity. The low-energy absorption during both September observations is comparable to the interstellar value. The X-ray lightcurves for both September observations show energy dependent flaring which may be modeled by changes in either low-energy absorption or power-law index.

The classical nova V4743 Sgr was observed with XMM–Newton for about 10 h on 2003 April 4, 6.5 months after optical maximum. At this time, this nova had become the brightest supersoft X‐ray source ever observed. In this paper, we present the results of a time‐series analysis performed on the X‐ray light curve (LC) obtained in this observation, and in a previous shorter observation done with Chandra 16 d earlier. Intense variability, with amplitude as large as 40 per cent of the total flux, was observed both times. Similarities can be found between the two observations in the structure of the variations. Most of the variability is well represented as a combination of oscillations at a set of discrete frequencies lower than 1.7 mHz. At least five frequencies are constant over the 16 d time interval between the two observations. We suggest that a period in the power spectrum of both LCs at the frequency of 0.75 mHz and its first harmonic are related to the spin period of the white dwarf (WD) in the system, and that other observed frequencies are signatures of non‐radial WD pulsations. A possible signal with a 24 000 s period is also found in the XMM–Newton LC: a cycle and a half are clearly identified. This period is consistent with the 24 278 s periodicity discovered in the optical LC of the source and thought to be the orbital period of the nova binary stellar system.

We have carried out a detailed analysis of the IUE archival high resolution spectra of the classical nova V1974 Cyg 1992. The main UV resonance lines show P Cygni profiles in the first days, which change into symmetric pure emission lines, and then slowly become fainter and narrower. Lines of higher ionization species reach their peak luminosity later than those of low ionization. This can be explained by a fast wind which is optically thick in the early days, when the pseudo-photosphere is located inside the wind. As the mass loss decreases, the radius of the pseudo-photosphere schrinks. This has three effects that explain the observed changes: (1) the deeper accelerating layers of the wind become visible where the emission lines are formed by collisional excitation and/or recombination, (2) as the mass loss rate decreases the emission comes from deeper regions of the wind where the velocities are smaller, (3) the effective temperature and the degree of ionization increase. In addition to the P Cygni and emission lines, we could identify two shortward shifted absorption systems which originate in two separate expanding shells, outside the wind layers where the emission lines are formed. The velocity of both shells increase with time. The outer main shell, containing most of the matter ejected at the outburst, produces the so-called "principal absorption line system", and the inner faster moving second shell produces the so-called "diffuse-enhanced absorption line system". The acceleration of the two shells is the result of increasing line-radiation pressure due to the UV-brightening of the star as the effective radius decreases. Around day 60 the second shell has overtaken the slower moving principal system shell, and merged with it. This explains: the sudden disappearance of the diffuse line system near that date, the upward jump of ∆v = 240 km s −1 in velocity of the principal system and the first detection of hard X-ray emission on day 63. This velocity jump indicates that the main shell is ≈ 4 times more massive than the second shell. The deceleration suffered by the diffuseenhanced system after the shock provides a shock temperature T shock ≈ 1.6 keV, in fairly good agreement with the temperature of the observed hard X-ray emission. The UV observations are interpreted through an empirical model in which the pre-nova slow wind phase is followed by the ejection of two shells, where the principal and the diffuse-enhanced absorption systems are formed, and by a phase of fast continuous lower density wind. Our empirical expansion velocity law for the principal system, together with Hα interferometric observations of the angular radius on day 10 are used to determine the distance to the nova, which is found to be 2.9 ± 0.2 kpc, in agreement with HST imaging and with the absolute magnitude versus rate of decline relationship.

Abstract. The recent 1994-1995 active phase of AG Draconis has given us for the first time the opportunity to follow the full X-ray behaviour of a symbiotic star during two successive outbursts and to compare with its quiescence X-ray emission. With ROSAT observations we have discovered a remarkable decrease of the X-ray flux during both optical maxima, followed by a gradual recovering to the pre-outburst flux. In the UV the events were characterized by a large increase of the emission line and continuum fluxes, comparable to the behaviour of AG Dra during the 1980-81 active phase. The anticorrelation of X-ray/UV flux and optical brightness evolution is shown to very likely be due to a temperature decrease of the hot component. Such a temperature decrease could be produced by an increased mass transfer to the burning compact object, causing it to slowly expand to about twice its original size.