"... In PAGE 5: ...nd AI = 0.190. See HFR for a discussion of why these values were adopted. Table2 lists the adopted aperture corrections for the WFPC2 CCDs. We were unable to obtain reliable aperture corrections for the PC images of the G312 field due to a lack of bright stars in this field.... ..."

"... In PAGE 11: ... In fact, either system can be configured with fewer apertures, with some tradeoff in performance (Figure 12). Studies reveal many configurations are possible ( Table2 ), each with advantages and disadvantages with respect to size, weight, complexity, and performance. In an eight- aperture configuration, the radio frequency (RF) components are very simple, each beam is independent of all others, and no downlink power couples directly into the uplink receive channels.... ..."

"... In PAGE 13: ...perture data for 5 of our galaxies. On the average, our magnitudes are 0.145 mag brighter in J and 0.061 in K. Table6 gives the di erences (literature magnitudes minus ours) for the 5 galaxies. The di erences are larger than expected from the scatter in the standard stars, and consistent with the surface brightness comparisons.... ..."

"... In PAGE 12: ...nd 3C 273 are larger; for those two cases, the inner radii used were 1.500 and 2.000, respectively. The outer radii chosen in general represent how far we could see the host galaxy (see Table3 ). The aperture magnitudes calculated using an inner radius r = 1:000 are typically 0:6 mag fainter than the total magnitude for the galaxy that was estimated by tting a model (see xx 5.... In PAGE 15: ... MAGNITUDES AND POSITIONS OF HOST GALAXIES In this section, we report on the measurements of the magnitude and position of the host galaxies. In Table3 we list the magnitudes for the host galaxies measured using aperture photometry and for a few cases the surface brightness method. The inner and outer radii used in performing aperture photometry are listed as well.... In PAGE 16: ... In computing the entries in Table 5, we used the k-correction of Fukugita et al. (1995) that are listed in the next-to-last column of Table3 . In computing the absolute magnitudes, we selected the best- t model based on the morphology of the host galaxy, unless the morphology is uncertain, and for those cases we list the model that gives the smallest 2 residuals.... In PAGE 20: ...5 kpc to 25 kpc (H0 = 50 km s?1Mpc?1, 0 = 0:0) corresponds approximately to 6 kpc to 12 kpc with our chosen cosmological parameters. In Table 8 we list for each quasar the inner and outer radii in arcsec, the apparent and absolute F 606 aperture magnitude, and the absolute V magnitude in the annulus (see adopted values for F 606 ? V values in Table3 ). As stressed by V eron-Cetty and Woltjer (1990), these annular measurements can be compared to future measurements obtained by other techniques.... In PAGE 36: ... The minimum detectable host brightness depends strongly upon the assumed morphology of the host galaxy. We have shown by a series of numerical experiments, described in Table3 of Paper II, that host galaxies are, on the average, visible on our images down to about 4.2 mag fainter than the quasar luminosity.... In PAGE 36: ...2 mag fainter than the quasar luminosity. Galaxies that are smooth ellipticals are the most di cult to detect (see rows 5d and 5e of Table3... In PAGE 40: ...that only one of the initial eight quasars studied in Paper I and Paper II had an obvious host. Given the repaired, but still complex and temporally variable PSF of the HST , we presented our results (see Table3 of Paper II) for the non-detections as a morphology- dependent limit on the brightness of the host galaxy. Spiral galaxies, with their azimuthal variation in brightness and regions of high surface brightness, could be seen to considerably fainter total brightnesses (more than a magnitude) than large, extended ellipticals with their smooth, regular pro les.... In PAGE 40: ... In retrospect, it appears that the vast majority of our initial observations consisted of quasars whose hosts had smooth, regular pro les. Comparing our adopted brightnesses of the hosts (Table 12) with the brightness limits set in our earlier work by simulations ( Table3 in Paper II), we nd that the appropriate brightnesses limits (those for smooth ellipticals like 5d and 5e of Figure 5 of Paper II) were reasonably accurate. The detected brightnesses reported in Table 12 of this paper range from considerably fainter than the Paper II limit (e.... ..."