Glenn and Arne had _already_ produced 0.03 mag photometry

[Stupendous Man <richmond@a188-l009.rit.edu>]

  Glenn asked me last night, "How did the night-to-night scatter 
in the datafile created by the 'difcal' program compare to the
night-to-night scatter in my new results?"  I was shocked to realize
that I hadn't bothered to look at the standard deviation of 
magnitudes listed in his "photom3.out" compilation.  

  So, I did.  And it turns out --- it's about 0.02-0.03 magnitudes
for bright (7 < V < 10) stars.  Exactly the same as I produced after
spending two days fighting with the data.

  Doh!

  Technical Note 41 now shows a plot of the 'difcal' scatter, as
well as the scatter produced by my ensemble photometry.  The ridge
line is the same in both plots.  The only difference is that my
plot has fewer stars with very large scatter night-to-night ...
which is due (almost?) entirely to the fact that I discarded
portions of "bad" nights in my solution.  This will facilitate
a search for variable objects ... and that's about all.  

  Well, that will teach me to ignore previous work.

                               Michael Richmond