We report the results of HST WFPC2 broadband F555W and F702W photometric and F555W polarimetric observations of the "Cloverleaf" QSO H1413+1143. This is a four-component gravitationally-lensed broad absorption line (BAL) QSO. Observations were obtained at two epochs in March 1999 and June 1999 separated by about 100 days. The goal of our program was to detect any relative changes among the components and between the two epochs. Over this time baseline we detected an approximately 0.07 mag dimming in component D of the lensed image, which we interpret as evidence for microlensing. In March 1999 we find significant evidence for a difference in the relative linear polarization of component D in comparison to the other three components; in June 1999 the combined polarization of the Cloverleaf components was lower. In March 1999 the apparently microlensed component D has a rotated polarization position angle and a somewhat higher degree of polarization than the other three components. We suggest that this difference in polarization is due to microlensing magnification of part of a scattered-light (i.e. polarized) continuum-producing region. The results indicate that in the Cloverleaf the size-scale of the polarized scattered-light region exceeds about 10^{16} cm but lies interior to the region producing the broad emission lines (< 10^{18} cm).

We report on new data, including four vacuum ultraviolet spectropolarimetric observations by the Wisconsin Ultraviolet Photo-Polarimeter Experiment ("WUPPE") on the Astro-1 (one observation) and Astro-2 (three observations) shuttle missions, and 15 new visible-wavelength observations obtained by the HPOL CCD spectropolarimeter at the Pine Bluff Observatory of the University of Wisconsin. This includes three HPOL observations made within 12 hours of each of the three Astro-2 WUPPE observations, giving essentially simultaneous observations extending from 1500 to 10500 Angstroms. An analysis of these data yields estimates on the properties of wind "clumps" when they are detectable by the polarization of their scattered light. We find that the clumps must be detected near the base of the wind, r/R* = 1.3 - 2.5. At this time the clump density is at least 10^13 cm^-3, 20 times the mean wind, and the temperature is roughly 10,000 K, about 20% cooler than the mean wind. The clumps leading to the largest observed polarization must have electron optical depth of 0.1 - 1 and an angular extent of 0.1 - 1 ster, and account for at most 2% of the wind mass loss. The H and HeI emission lines from the wind are unpolarized, but their P Cygni absorption is enhanced by a factor of 4 in the scattered light. We speculate on the possible relationship of the clumps detected polarimetrically to those seen as Discrete Absorption Features (DACs) and in interferometry.