X-RAY POLARIMETRY: THE MEASUREMENT OF THE POLARIZATION OF SOLAR FLARE X-RAYS AND THE DESIGN OF A COMPTON POLARIMETER

A Compton polarimeter has been designed for use in X-ray astrophysics. This device utilizes the anisotropic Compton scattering of polarized X-rays to determine their polarization. This instrument incorporates an intrinsic coincidence rejection scheme to reduce the non X-ray background. The Compton polarimeter is sensitive to X-rays from 30-100 kiloelectronvolts.; Another class of X-ray polarimeter, the photoelectric polarimeter, is discussed. This class of devices detects the anisotropic distribution of photoelectrons ejected by polarized X-rays. The energy range for this instrument is 5-100 kiloelectronvolts. Previous work on this class of device, which indicated some promise of a useful instrument, is reviewed. An analysis of electron scattering, a complicating effect for these devices, is presented. This analysis indicates a severe limitation on the utility of photoelectric polarimeters. This theoretical work is reconciled with the previous experiments. One implementation, using an imaging proportional counter, was tested in the laboratory. No polarization effects were seen.; The polarization of solar flare X-rays is reviewed from both a theoretical and observational viewpoint. The production mechanisms in competing theories predict different amounts of polarization. Previous experiments indicated significant polarization, however, these results have been called into question. The Columbia Astrophysics Laboratory Solar Flare X-Ray Polarimeter is described. This instrument, a Thomson scattering device, was flown in March 1982 aboard the Space Shuttle Columbia on its third flight.; Various background rejection schemes were tested during the flight. A multi-wire rejection scheme and an active plastic anti-coincedence shield were found to be comparable.; The data from this flight were complicated by pre-flight contamination. This effect was removed using an in-flight calibration. An independent quality test was available because the polarimeter overdetermined the polarization. This allowed an unambiguous determination of the relative polarization of the flares observed. No significant polarization was found. The theoretical implications of this null result are discussed.