start time scan stability

[Brendan Kiburg <kiburg@npl.uiuc.edu>]

Dear Peter and Steven,
   I spent some time working with the start time scans for the nitrogen 
data.  It is clear that the run8 data becomes completely stable with the 
full kinetics function and the data easily satisfies the Kawall band 
requirements.

  For run10, the full kinetics function does a much better job producing 
stability for the additional rate than the 3-par fit, but the data still 
wanders outside the kawall bands too much.  For what it is worth, the 
error bars on the data satisfy the Kawall band containment requirements.

So I cannot currently say that the full kinetics fit is entirely stable, 
but it is considerably better than the 3-par fit did, so hopefully I 
just have some fine tuning to do with the fit/stop definitions to bring 
this to a level of acceptability.

Regards,
Brendan