The Swift mission, scheduled for launch in 2004, is a multiwavelength observatory for gamma-ray burst (GRB) astronomy. It is a first-of-its-kind autonomous rapid-slewing satellite for transient astronomy and pioneers the way for future rapid-reaction and multiwavelength missions. It will be far more powerful than any previous GRB mission, observing more than 100 bursts yr À1 and performing detailed X-ray and UV/optical afterglow observations spanning timescales from 1 minute to several days after the burst. The objectives are to (1) determine the origin of GRBs, (2) classify GRBs and search for new types, (3) study the interaction of the ultrarelativistic outflows of GRBs with their surrounding medium, and (4) use GRBs to study the early universe out to z > 10. The mission is being developed by a NASA-led international collaboration. It will carry three instruments: a newgeneration wide-field gamma-ray (15-150 keV ) detector that will detect bursts, calculate 1 0 -4 0 positions, and trigger autonomous spacecraft slews; a narrow-field X-ray telescope that will give 5 00 positions and perform spectroscopy in the 0.2-10 keV band; and a narrow-field UV/optical telescope that will operate in the 170-600 nm band and provide 0B3 positions and optical finding charts. Redshift determinations will be made for most bursts. In addition to the primary GRB science, the mission will perform a hard X-ray survey to a sensitivity of $1 mcrab ($2 ; 10 À11 ergs cm À2 s À1 in the 15-150 keV band ), more than an order of magnitude better than HEAO 1 A-4. A flexible data and operations system will allow rapid follow-up observations of all types of high-energy transients, with rapid data downlink and uplink available through the NASA TDRSS system. Swift transient data will be rapidly distributed to the astronomical community, and all interested observers are encouraged to participate in follow-up measurements. A Guest Investigator program for the mission will provide funding for community involvement. Innovations from the Swift program applicable to the future include (1) a large-area gamma-ray detector using the new CdZnTe detectors, (2) an autonomous rapid-slewing spacecraft, (3) a multiwavelength payload combining optical, X-ray, and gamma-ray instruments, (4) an observing program coordinated with other ground-based and space-based observatories, and (5) immediate multiwavelength data flow to the community. The mission is currently funded for 2 yr of operations, and the spacecraft will have a lifetime to orbital decay of $8 yr.

Soft ␥-ray repeaters (SGRs) emit multiple, brief (ϳ0.1-s), intense outbursts of low-energy ␥-rays. They are extremely rare 1 -three 2-4 are known in our Galaxy and one 5 in the Large Magellanic Cloud. Two SGRs are associated 5-7 with young supernova remnants (SNRs), and therefore most probably with neutron stars, but it remains a puzzle why SGRs are so different from 'normal' radio pulsars. Here we report the discovery of pulsations in the persistent X-ray flux of SGR1806 ؊ 20, with a period of 7.47 s and a spindown rate of 2:6 ؋ 10 ؊ 3 s yr ؊ 1 . We argue that the spindown is due to magnetic dipole emission and find that the pulsar age and (dipolar) magnetic field strength are ϳ1,500 years and 8 ؋ 10 14 gauss, respectively. Our observations demonstrate the existence of 'magnetars' , neutron stars with magnetic fields about 100 times stronger than those of radio pulsars, and support earlier suggestions 8,9 that SGR bursts are caused by neutron-star 'crustquakes' produced by magnetic stresses. The 'magnetar' birth rate is about one per millennium-a substantial fraction of that of radio pulsars. Thus our results may explain why some SNRs have no radio pulsars.SGR1806 Ϫ 20 became extremely active between October 1996 and November 1997, when over 40 intense bursts and numerous weaker ones were detected 10 with the Burst And Transient Source Experiment (BATSE) on board the Compton Gamma-Ray Observatory (CGRO). We observed SGR1806 Ϫ 20 with the Rossi X-Ray Timing Explorer (RXTE) five times between 5 and 18 November 1996, starting five days after the first triggered burst detection with BATSE. (Information on the archival data from RXTE/PCA and ASCA is available at http://heasarc.gsfc.nasa.gov.) During these observations 11 , the source emitted series of outbursts in a 'bunching' mode, never seen before. The intensity of the outbursts, as well as the 'bunching' mode, varied significantly: mini-outbursts were interlaced with very intense ones and the rate of bursts varied from bunch to bunch (S. Dieters et al., manuscript in preparation).We made a period search of the data after excluding all bursts from the time series. The data were then energy-selected for 2-24 keV X-rays, background subtracted and binned at 0.5-s resolution. The resulting light curve was searched for periodicities between 0.03 and 1 Hz, by calculating a fast-Fourier-transform power spectrum (Fig. 1). The peaks in the spectrum are centred on the fundamental frequency of 0.13375 Hz (period of 7.47655 s) and its first harmonic at 0.26750 Hz. We find no significant power in any other frequency in the searched range. The probability that we detect a signal at the fundamental frequency this strong by chance coincidence is 1 ϫ 10 Ϫ 13 (taking into account the number of trials, 1:9 ϫ 10 6 , and the probability per trial, 5 ϫ 10 Ϫ 20 ).To determine the fundamental period, all data sets were then corrected to the Solar System barycentre and separately folded at the longest detected period of 7.47655 s, and sub-harmonics thereof. These sub-harmonic folds showed...

We report evidence that the asymptotic low-energy power law slope alpha (below the spectral break) of BATSE gamma-ray burst photon spectra evolves with time rather than remaining constant. We find a high degree of positive correlation exists between the time-resolved spectral break energy E_pk and alpha. In samples of 18 "hard-to-soft" and 12 "tracking" pulses, evolution of alpha was found to correlate with that of the spectral break energy E_pk at the 99.7% and 98% confidence levels respectively. We also find that in the flux rise phase of "hard-to-soft" pulses, the mean value of alpha is often positive and in some bursts the maximum value of alpha is consistent with a value > +1. BATSE burst 3B 910927, for example, has a alpha_max equal to 1.6 +/- 0.3. These findings challenge GRB spectral models in which alpha must be negative of remain constant.Comment: 12 pages (including 6 figures), accepted to Ap

Long-duration gamma-ray bursts (GRBs) accompany the deaths of some massive stars and hence, because massive stars are short-lived, are a tracer of star formation activity. Given that GRBs are bright enough to be seen to very high redshifts and detected even in dusty environments, they should therefore provide a powerful probe of the global star formation history of the Universe. The potential of this approach can be investigated via submillimetre (submm) photometry of GRB host galaxies. Submm luminosity also correlates with star formation rate, so the distribution of host-galaxy submm fluxes should allow us to test the two methods for consistency. Here, we report new JCMT/SCUBA 850-µm measurements for 15 GRB hosts. Combining these data with results from previous studies, we construct a sample of 21 hosts with <1.4 mJy errors. We show that the distribution of apparent 850-µm flux densities of this sample is reasonably consistent with model predictions, but there is tentative evidence of a dearth of submm-bright (>4 mJy) galaxies. Furthermore, the optical/infrared properties of the submm-brightest GRB hosts are not typical of the galaxy population selected in submm surveys, although the sample size is still small. Possible selection effects and physical mechanisms which may explain these discrepancies are discussed.

NGC 1313 X-2 is one of the brightest ultraluminous X-ray sources in the sky, at both X-ray and optical wavelengths; therefore, quite a few studies of available ESO VLT and HST data have appeared in the literature. Here, we present our analysis of VLT/FORS1 and HST/ACS photometric data, confirming the identification of the B ∼ 23 mag blue optical counterpart. We show that the system is part of a poor cluster with an age of 20 Myr, leading to an upper mass limit of some 12 M for the mass donor. We attribute the different results with respect to earlier studies to the use of isochrones in the F435W and F555W HST/ACS photometric system that appear to be incompatible with the corresponding Johnson B and V isochrones. The counterpart exhibits significant photometric variability of about 0.2 mag amplitude, both between the two HST observations and during the one month of monitoring with the VLT. This includes variability within one night and suggests that the light is dominated by the accretion disk in the system and not by the mass donor.