Re: Colour terms in MK IV astrometry

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

  A quick followup to Andrew's note.  I'm working on 
the new DS24 grid images, and happen to recall that
my astrometric solutions -- which do NOT include a
color term, use a cubic polynomial and fit to the
tangent plane -- yield errors in each direction
of about 0.5 arcseconds.  That's the RMS of all
astrometric errors vs. Tycho after the nominal
solution.  It seems consistent with Andrew's values.

  I see the same effect that Tom mentioned -- stars
are fainter near the center of the field, brighter
near the edges.  The magnitude of the effect is (from
memory, since all my notes at at work) only around 
+/- 0.05 mag, I _think_.  Tom mentioned that the 
effect was stronger for bright stars than faint ones;
I checked that, and found the same effect in both
V and I for stars brighter and fainter than mag 10.

  Now working on a method to derive a correction 
as a function of position on the image ...

                  Michael