Spinning up a millisecond pulsar.
Pulsars - rotating neutron stars - come in several types. The commonest ones spin anywhere from several dozen times a second to once every few seconds. Their rotation is left over from their birth, when the core of a massive star collapsed in a supernova explosion. Most of them are gradually slowing down as they emit particles and lose angular momentum.
Then there are the millisecond pulsars, which spin much faster - hundreds of times per second. They probably weren't born that way; a collapsing supernova core might not have that much angular momentum. Something, the theory went, probably spun them up later. The usual explanation has been that a close companion star transferred angular momentum to the pulsar by pouring mass onto it.
Now a binary system has finally been caught in the midst of this s, pin-up process. In April an X-ray source flared up for several weeks at the border of Sagittarius and Corona Australis. Rudy Wijnands and Michiel van der Klis (University of Amsterdam) examined it with the Rossi X-ray Timing Explorer satellite and found that its X-ray output pulses 401 times per second. SAX J1808.4-3658, as the source is known, is the first binary millisecond X-ray pulsar ever found.
Deepto Chakrabarty and Ed Morgan (MIT) soon found slight, periodic variations in the pulse timing, which revealed that the neutron star is closely orbited by a very low-mass star (or a massive planet) every 2 hours. This, no doubt, is the source of the hot matter that is falling onto the neutron star, causing the X-rays and adding to the pulsar's spin.
Announcements in the IAU Circulars quickly brought other astronomers into the chase. A check at visible wavelengths by Paul Roche and his colleagues (University of Sussex) found a new 16.6-magnitude speck at the X-ray source's location. Barry Giles (University of Tasmania) found that this visible source varies slightly with the same 2-hour period.
Then spectra taken at one of the Keck telescopes by Alexei Filippenko and Douglas Leonard (University of California at Berkeley) showed evidence of a hydrogen emission line with two peaks separated by a Doppler shift of 1,000 kilometers per second. Here, apparently, was the telltale sign of an accretion disk closely surrounding the neutron star, with one side rotating toward us, the other away.
The entire picture, assembled by independent teams in just 10 days, "is a holy grail of pulsar and binary evolution theory," says Victoria Kaspi (MIT). "It unambiguously confirms the most basic assumptions in the field for the first time."
When SAX J1808.4-3658's X-ray flare peaked around April 12th, the system was putting out 2,000 times as much energy in X-rays as the Sun emits at all wavelengths (assuming the pulsar lies 13,000 light-years away). By May 2nd the X-ray output had dwindled to only 15 Suns. Astronomers are watching to see whether the object becomes a radio pulsar when the stream of mass from the companion star dries up. According to theory, systems like this should flip-flop between pulsing at X-ray and radio wavelengths as mass transfer starts and stops.