Beyond Neptune lies the Kuiper belt, an area of the Solar System that astronomers believe is made up of billions of rocks. These are thought to be the result of collisions, and some are huge, measuring 1,000 kilometres or more in diameter. Well-known examples include Pluto and its moon Charon. Although astronomers believe they exist, the smaller rocks have been harder to find. Now, a team of researchers has hit on a new way of looking for them (see page 660).

The use of optical methods has not been particularly successful in the hunt for the smaller members of the Kuiper belt. So, a group of astronomers led by Hsiang-Kuang Chang at Taiwan's National Tsing Hua University looked for an alternative. They came up with the idea of using X-rays from Scorpius X-1, a neutron star 9,000 light years away. From time to time, the number of X-ray photons reaching an observing satellite from Scorpius X-1 dips significantly. The researchers thought that these dips might result from small bodies passing in front of the star. Such dips are known as occultations.

Much shorter occultations can be detected with X-ray than optical methods, and this is crucial for finding tiny bodies. Whereas visible-light occultation events of 0.1 or 0.2 seconds might be detectable, researchers can detect X-ray occultations as short as one millisecond. This is because the instrument they used, NASA's Rossi X-ray Timing Explorer (RXTE) satellite, can process incoming data much faster than optical instruments.

Chang and his colleagues analysed seven years' worth of publicly available archive data from the RXTE. They used a computer program to sift through each time bin — a prescribed time interval during which the total number of photons was recorded. This was a much more thorough analysis than usual. “No one has ever looked into every time bin,” Chang says.

The researchers found records of 58 occultation events. On the basis of their findings, they estimate that the Kuiper belt contains some 1015 small bodies (of 100 metres or less in diameter). Initially, however, they couldn't be sure that the effects they were seeing were not the result of instrumental error. So they turned to another batch of RXTE data, reporting observations of another major X-ray emitter, the Crab nebula, as a control.

Because of its large size, the Crab nebula is an extended source of X-rays. Scorpius X-1 is a point source. So whereas the tiny Kuiper-belt bodies would all but completely block Scorpius X-rays, they would barely obscure any of the Crab nebula's. The team reasoned that if there was an instrumental effect, both the Crab and the Scorpius data would contain similar dips in photon count. If there were none, only the Scorpius data would reveal the X-ray photon dips. They found the latter to be true.

In future work, Chang hopes to pin down the distance to these small bodies. The key could be in their diffraction effects. Although diffraction levels are much lower with X-rays than with optical light, diffraction effects are still seen in X-rays and can be used to determine how far objects are from Earth. This should lead to a better estimate of the size of these occulting bodies. Chang will also look at data from future RXTE observations in order to find more occultation events and to improve the accuracy of estimates of the size distribution of the Kuiper belt's rocks.