The gravity of a black hole is so strong that nothing within a certain radius can escape from it - not even light. But astronomers can sometimes detect black holes by the radiation that originates from near their surfaces: as material falls into a black hole, its potential energy is converted into radiation that streams across space.

Previous studies of the motion of stars suggested that a black hole about three million times the mass of the Sun exists in the centre of the Milky Way. But the orbits of those stars had a radius 30 000 times larger than the predicted radius of a black hole that massive. In a region this large, some other entity - such as a cluster of 'dark stars' - might make up this unseen mass.

Baganoff and colleagues first confirmed that Sagittarius A* emitted X-rays in 1999, using the space-based Chandra X-ray Observatory. They detected only a weak signal initially, but similar observations a year later revealed that the flux had increased dramatically. In particular, it jumped to about 45 times its previous level during one burst, which lasted for almost three hours. During this period, Baganoff's team found that the X-ray intensity rose by a factor of five in just ten minutes, and fell in a similar period.

The frequency of these fluctuations is crucial. The conditions that led to such a radical change in the X-ray emission cannot have spread across the black hole any faster than light. But the conditions changed significantly within ten minutes, and this limits the size of the black hole to a region less around one astronomical unit - the distance from the Earth to the Sun - across. A super-massive black hole is the only realistic candidate for the invisible mass that could exist in such a small region.

Baganoff and colleagues now hope to find a similar pattern in the variations of radio waves from Sagittarius A*. If they succeed, they will have excluded the slight possibility that the X-rays came from a different source.