The angular diameters of stars are much smaller than the limit imposed by seeing conditions in the atmosphere. It may therefore seem that there is no point in building an optical analogue of the radio interferometer in order to improve the resolution of an optical telescope. However, there are moments when the image stabilizes for long enough to give interference fringes. In 1920, an interferometer was assembled on the front of the 2.5-m telescope at Mount Wilson by Michelson and Pease. It had a 6.5-m beam and allowed them to measure the angular size of seven giant stars. They found angular diameters of one-twentieth to one-fiftieth of an arc second.

An alternate approach to the measure of stellar diameters has been developed by R.Hanbury-Brown and R.Q.Twiss in Britain and more recently in Narrabri, Australia. They use two optical telescopes up to 200m apart. The light from the stars is detected with photo-multipliers and the intensities of the signals received are cross-correlated and the correlation amplitude measured for a variety of separations of the two optical telescopes. From these data it is possible to deduce information about the angular size of the star being observed. As the correlated signals are very weak, this technique is only applicable to the brightest stars. For these, resolutions down to 5 x 10~4arc sec have been achieved.

A very different approach called SPECKLE INTERFEROMETRY has been pioneered by A.Labeyrie and colleagues. Although atmospheric fluctuations spread images over their seeing disc, very short exposures (0.1-0.001 seconds) reveal that the images consist of a large number of tiny spots or SPECKLES whose size and shape may be analysed to give structural information on bright stars with a resolution set only by the diffraction limit of the telescopes used (about 0.010arc sec with the 5-m telescope on Mount Palomar). Stars as faint as 9 mag have been measured with this technique. For faint stars, large numbers of short exposures have to be processed.

Comments are closed.