A 321.5 s modulation was discovered in 1999 in the X-ray light curve of HM Cnc. In 2001 and 2002, optical photometric and spectroscopic observations revealed that HM Cnc is a very blue object with no intrinsic absorptions but broad (FWHM 1500 km s^-1
) low equivalent width emission lines (EW 1-6A), which were first identified with the HeII Pickering series. The combination of X-ray and optical observations pictures HM Cnc as a double degenerate binary hosting two white dwarfs, and possibly being the shortest orbital period binary discovered so far. The present work is aimed at studying the orbital motion of the two components by following the variations of the shape, centroid and intensity of the emission lines through the orbit. In February 2007, we carried out the first phase resolved optical spectroscopic study with the VLT/FORS2 in the High Time Resolution (HIT) mode, yielding five phase bins in the 321 s modulation. Despite the low SNR, the data show that the intensity of the three most prominent emission lines, already detected in 2001, varies with the phase. These lines are detected at phases 0.2-0.6 where the optical emission peaks, and marginally detected or not detected at all elsewhere. Moreover, the FWHM of the emission lines in the phase resolved spectra is smaller, by almost a factor 2, than that in the the phase-averaged 2001 spectrum. Our results are consistent with both the pulsed optical component and emission lines originating in the same region which we identify with the irradiated surface of the secondary. Moreover, regardless of the origin of the un-pulsed optical continuum, we note that the EWs of the emission lines might be up to -15 / -25A, larger than thought before; these values are more similar to those detected in cataclysmic variables. All the findings further confirm that the 321s modulation observed in HM Cnc is the orbital period of the system, the shortest known to date.
We report on the intense burst ``forest recorded on 2006 March 29 which lasted for ~30s. More than 40 bursts were detected both by BAT and by XRT, seven of which are rare intermediate flares (IFs): several times 10^{42} ergs were released. The BAT da
ta were used to carry out time-resolved spectroscopy in the 14-100keV range down to 8ms timescales. This unique dataset allowed us to test the magnetar model predictions such as the magnetically trapped fireball and the twisted magnetosphere over an unprecedented range of fluxes and with large statistics (in terms of both photons and IFs). We confirmed that a two blackbody component fits adequately the time-resolved and integrated spectra of IFs. However, Comptonization models give comparable good reduced chi^2. Moreover, we found: i) a change of behavior, around ~10^{41} erg/s, above which the softer blackbody shows a sort of saturation while the harder one still grows to a few times 10^{41} erg/s; ii) a rather sharp correlation between temperature and radii of the blackbodies (R^2 prop kT^{-3}), which holds for the most luminous parts of the flares (approximately for L_{tot} > 10^{41} erg/s). Within the magnetar model, the majority of these findings can be accounted for in terms of thermalised emission from the E-mode and O-mode photospheres. Interestingly, the maximum observed luminosity coming from a region of ~15km matches the magnetic Eddington luminosity at the same radius, for a surface dipole field of ~8 x 10^{14} G (virtually equal to the one deduced from the spindown of SGR 1900+14).
Previous studies of the X-ray flux and spectral properties of 1RXS J170849-400910 showed hints of a possible correlation with the spin glitches that occurred in 1999 and 2001. However, due to the sparseness of spectral measurements and the paucity of
detected glitches no firm conclusion could be drawn. We retrieved and analysed archival XTE pointings of 1RXS J170849-400910 covering the time interval between January 2003 and June 2006 and carried out a detailed timing analysis by means of phase fitting techniques. We detected two relatively large glitches Delta nu / nu of 1.2 and 2.1 10^-6 occurred in January and June 2005. Interestingly, the occurrence times of these glitches are in agreement with the predictions made in our previous studies. This finding strongly suggests a connection between the flux, spectral and timing properties of 1RXS J170849-400910.
