WARNING:  Many of the small-aperture magnitudes for Seyfert nuclei
 listed in MGT are systematically too faint, due to saturation effects
 which were worse than we had suspected.
 The "inner magnitudes" (which refer to a 0.1 arcsecond radius aperture),
 for some of the brighter Seyfert 2 nuclei in Table 2, and for 
 many of the Seyfert 1 nuclei in Table 1, are too large (i.e. too faint)
 by anywhere between 0 and 2 magnitudes. In other words, many of them 
 are merely upper limits to the true magnitude. In the paper we mistakenly
 assumed that stars fainter than 18.0-18.5 were unsaturated in our 500-second
 PC exposures.  However, a comparison with 28 stars measured in both our
 PC images and in the US Naval Observatory sky survey catalog reveals 
 some substantial disagreements. Even when the magnitude difference between 
 the small aperture (r=0.1 arcsecond) and the larger aperture (r=0.5
 arcseconds), equals 0.5 magnitudes--the value for unsaturated point sources--
 the stellar magnitude from our PC exposures can still be too faint by more
 than 1 magnitude (in 14% of the cases).  
    Unfortunately our inner magnitude estimates for saturated Seyfert 1 nuclei,
 which were based on the unsaturated wings of their PSFs, 
 are also about 1 +/- 1 magnitude fainter than values
 measured in 10 subsequent PC images having much shorter exposure times.
 It seems likely that more of this unpredictable error is due to problems
 with saturation, than with intrinsic time variability of the Seyfert 1
 nuclei.  Inner magnitudes fainter than 19 are likely
 to be accurate, as is the case for most of the fainter Seyfert 2 nuclei in 
 Table 2, and nearly all of the non-Seyfert nuclei listed in Table 3.
 We have confirmed this by finding excellent agreement between 
 our data and photometry of non-Seyfert Markarian galaxies 
 from the literature (e.g. from Huchra).
 We apologize to any astronomers who might have relied on the accuracy
 of our Seyfert photometry, and regret that we are unable to come up 
 with a reliable algorithm for correcting the saturation errors.