All of these files of mean positions are derived from my B1950.0 position 
files.  

Those labeled simply "ngcmpos_*.txt" or "icmpos_*.txt" carry unweighted mean B1950.0 
positions for the NGC and IC objects, respectively.  These are also
concatenated into single files:  "ngc1950_pos.txt" and "ic1950_pos.txt".

The files labeled "nmppos_*.txt" and "imppos_*.txt" are unweighted means precessed to
J2000.0 positions.  These, too, are concatenated into single files 
"ngc2000_pos.txt" and "ic2000_pos.txt".

Original positions from the 19th or early 20th century are not used in 
calculating the means.  If the only position available is from the original 
observer (as is the case for unrecoverable objects), I've simply carried it 
over into the mean position tables with a lower case "o" appended to the 
number of observations, always = "1" in this case.

Eventually, I'll pull together files of weighted mean positions.  (I sometimes
refer to this project as the "Asymptotic NGC/IC Project".  This assumes there
actually is an asymptote.)

The format of the mean position files is similar to that of the raw position 
files:

  1) NGC/IC number or other unique name

  2) Mean position given with the number of digits matching the most precise
     position in the raw position lists.  If you believe that we know the
     position to that accuracy, I have some interesting structures in 
     Brooklyn that I'd like to sell to you ...

  3) "N" is the number of positions used in calculating the mean, and, as 
     noted above, "o" is appended it the position is an original one used
     when the object has not been recovered.

  4) "sigRA" and "sigDE" are the standard deviations in arcseconds in RA and 
     Dec, respectively, calculated from the usual formulae.  These are 
     one-sigma errors and tell us something about the real accuracy with which 
     we know a given position.  For galaxies and planetaries, I am making an 
     effort to reduce the standard deviations to less than 4 arcsec in both
     coordinates.  For clusters and diffuse nebulae, less than 1 arcminute is
     Really Good Stuff, and I move on happily to the next object.  Where only 
     one position exists, this number can't be calculated, of course.  See the 
     "sources" file for my estimate of the standard deviation of the given 
     source.  Someday, I'll use all this to find the true standard deviations 
     for each source with more than n = 2 objects included in at least two 
     other sources.

  5) Finally, all of the source codes are listed.  See the "sources" file for
     the references corresponding to the source codes.

Here are some example lines:

<Name              ><  RA        ><  Dec       ><No <sigRA<sigDE <Sources ...
N0001               00 04 41.16   +27 25 48.3    7    3.33  2.41 HC   Pds  UZC  WS   CCA  DC   GHD1
N0002               00 04 42.38   +27 23 58.9    5    2.14  3.91 HC   UZC  WS   CCA  GHD1
N0003               00 04 42.77   +08 01 23.3    7    0.79  1.92 HC   KEBA HCds HC2m WS   UZC  CCA 
N0004               00 04 50.39   +08 05 42.5    4    0.33  2.32 HC   HCm  HCds HC2m
N0005               00 05 13.818  +35 05 02.38   4    0.71  0.59 KHJ  UZC  SBG2 CCA 
N0006=N0020         00 06 57.37   +33 01 50.0    8    0.96  2.01 WS   BPP  GHD1 GSC  Pds  UZC  CCA  DC  
N0007               00 05 48.21   -30 11 38.5    3    1.43  3.06 WS   HCds HCrs
N0008               00 06 10.61   +23 33 37.2    5    0.43  1.70 WS   HCo  GH   Pds2 HCds
N0008 w *           00 06 10.43   +23 33 39.0    1               HCds
N0008 e *           00 06 10.91   +23 33 39.3    1               HCds
I0067               00 57 45      -07 10.8       1o              Bige
I0068               00 57 50      -07 12.8       1o              Bige

hgcjr
November 2004