Search

We describe ROSAT observations of a new X-ray transient and its probable optical companion in the Small Magellanic Cloud. The transient, designated RX J0117.6-7330, appeared at a position approximately 5 min southeast of the X-ray pulsar, SMC X-1, in 1992 October 1-2 PSPC observations centered on the pulsar. It was detected again in a similar observation 246 days later at a counting rate diminished by a factor of 270, which corresponds to an average epsilon-folding decay time of 44 days. No periodic pulsations have been detected. The average 1992 flux level would be produced by a source radiating isotropically with a luminosity of 1.6 X 10(exp 37)(D/50 kpc)(exp 2) ergs/s in the energy range 0.2-2.5 keV. The 5 sec radius positional error circle includes the probable optical counterpart in the form of a magnitude V = 14.2 star with the spectral characteristics of type Be.

The X-ray phenomena of the binary system SMC X-1/Sk 160, observed with the Ginga and ROSAT X-ray observatories, are compared with computed phenomena derived from a three-dimensional hydrodynamical model of the stellar wind perturbed by X-ray heating and ionization which is described in the accompanying paper (Blondin & Woo 1995). In the model the B0 I primary star has a line-driven stellar wind in the region of the X-ray shadow and a thermal wind in the region heated by X-rays. We find general agreement between the observed and predicted X-ray spectrum throughout the binary orbit cycle, including the extended, variable, and asymmetric eclipse transitions and the period of deep eclipse.

We present a three-dimensional hydrodynamic simulation of the disrupted stellar wind in the high-mass X-ray binary system SMC X-1. The three dominant processes that determine the geometry of the wind in high X-ray luminosity systems such as SMC X-1 are the X-ray suppression of the stellar wind from the X-ray irradiated face of the primary star, the focusing of the radiatively driven wind in the X-ray shadow by the effects of stellar rotation, and the rapid X-ray heating of gas in the vicinity of the X-ray source, including the X-ray illuminated surface of the primary star. The resulting distribution of circumstellar gas provides a successful explanation for the asymmetric, extended eclipse transitions and the intensity of the deep eclipse X-ray emission in SMC X-1, as well as a possible explanation for the X-ray dips seen near superior conjunction of the X-ray source in Cyg X-1.

We have measured the decaying dust-scattered X-ray halo of Cen X-3 during its binary eclipse with the ASCA solid-state imaging spectrometer (SIS). The surface brightness profile (SBP) of the image in the low-energy band (0.5-3 keV) lies substantially above the point-spread function (PSF) of the X-ray telescope, while the SBP in the high-energy band (5-10 keV) exhibits no significant deviation. By contrast, the SBPs of Vela X-1 during its eclipse are consistent with the PSF in both the low- and high-energy bands -- strong evidence that a dust halo is indeed present in Cen X-3. Accordingly, we modeled the SBP of Cen X-3 taken from six consecutive time segments under the principal assumptions that the dust is distributed uniformly along a segment of the line of sight, the grains have a power-law size distribution, and the low-energy source flux was the same function of orbital phase before as during our observation. The best-fit set of parameters included a grain density value of 1.3 g/cu cm, substanially less than the density of 'astronomical silicate.' This result supports the idea that interstellar grains are 'fluffy' aggregates of smaller solid particles. We attribute the failure to detect a halo of Vela X-1 during its eclipse phase to extended strong circumsource absorption that probably occurred before the eclipse and allowed the halo to decay away before the observation began.

From measurements of the X-ray eclipse phenomena of the binary X-ray pulsar 4U 1538-52, we derive properties of the stellar wind of its B0 I companion, QV Nor, and a constraint on models of interstellar grains. Estimates of the wind density as a function of the distance from the center of QV Nor are obtained from an analysis of the variation of X-ray attenuation during an eclipse egress. A Monte Carlo computation of the absorption and scattering of X-rays in the X-ray-ionized wind accounts for approximately two-thirds of the spectrum of X-rays with energies above 4.5 keV observed during the eclipse. Our upper limit on R(sub XV) is 0.06/mag, which implies that the X-ray scattering efficiency of interstellar dust is less than expected for solid grains with a size distribution of the form n(sub g)(a) approximately a(exp -3.5) in the range from 0.005 to 0.25 microns and composed of silicate (R(sub XV) = 0.22/mag) or a silicate-graphite mixture (R(sub XV) = 0.11/mag) as derived from the calculations of Martin & Rouleau (1991). This lends support to the idea (Mathis & Whiffen 1989) that interstellar grains are 'fluffy' aggregates with an average bulk density less than that of their constitutent particles. Such aggregates would have a smaller ratio of X-ray scattering efficiency to optical extinction efficiency compared with solid grains of the same material.

A cyclotron absorption line near 20 keV has been found in the spectrum of the massive eclipsing binary X-ray pulsar 4U 1538 - 52 in observations with the Ginga observatory. The line is detected throughout the 529 s pulse cycle with a variable equivalent width that has its maximum value during the smaller peak of the two-peak pulse profile. It is found that the profile of the pulse and the phase-dependence of the cyclotron line can be explained qualitatively by a pulsar model based on recent theoretical results on the properties of pencil beams emitted by accretion-heated slabs of magnetized plasma at the magnetic poles of a neutron star. The indicated field at the surface of the neutron star is 1.7 (1 + z) x 10 to the 12th G, where z is the gravitational redshift.