The updated version was archived on 12/9/98.
Note:  File 10 and 22 are from the original dataset.

Comments from PI:
-----------------

The files currently in the archive are the ONLY copies of original data.  
I used an HP-85 for this experiment and the media cannot be read here 
any more.  Please keep those files in the archive (they are not very big).

The updated data is a cleaner version, which also removes erroneous data in 
the lat/lon files, so most users would want it, but it does not have all 
the data either for lat/lon or for CO2.  I don't think it would be hard to 
explain that in a documentation file.

To clean the CO2 files I axed the obvious flyers (just a few points for 
any flight) and then removed points more that 3-sigma from the trend line 
over a minute or two.  That's not very many points either, but enough to 
make a difference in the acuity of the plots relative to atmospheric 
structure.

This was a very early generation instrument, there is some value in these 
data but they are no better than the day they were taken.

scw