Jim Bedient's variable star

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

  Several days ago, Jim Bedient mentioned his work on a star which 
appears to be a red, long-period variable.  He looked it up in
the TASS 'tenxcat' database, and was confused to find that two
different entries in tenxcat might both correspond to this variable.

  I think he's right.

  There are two stars of nearly equal brightness in V-band, only
37 arcseconds apart:

          RA   (2000)    Dec           GSC          USNO 2
---------------------------------------------------------------
      19:20:36.50    -03:58:18.8    5138 00446    0825-1541 3260
      19:20:35.08    -03:57:50.5       --         0825-1541 1768

  
  The second, northern star is very red, and close to an IRAS
infrared source.  It is apparently the true variable star.  

  The trouble is that the Mark III cameras have pixels 13.4 arcsec
across, so that these two stars are only about 2.8 pixels away from
each other.  The typical FWHM of Mark III images ranged from 2.5 to 4.0
pixels, so the two stars would often have been merged together.

  My guess is that when the variable was faint, the southern star's
light dominated (especially in V), but when the variable was bright,
it dominated (especially in I).  On two occasions, the two stars 
were both detected in V-band; but usually, there was only a single
V-band detection.  Sometimes, that single detection was assigned
to the constant southern star -- but occasionally it was assigned
to the variable northern star.  In I-band, only the red variable
northern star was ever detected.

  If I were going to use the V-band data, I'd treat all measurements
(from both entries in tenxcat) as the combination of light from 
a constant star and the variable star.  The constant star has 
magnitudes from USNO2 of

                           B = 13.1   R = 11.6

so you could make an estimate of its V-band magnitude.  Then you
could subtract its contribution from the reported tenxcat V-band
magnitude (which involves more than actually subtracting the
two magnitudes, of course) to retrieve the proper V-band measurement
for the variable star alone.

                                             Michael Richmond