The value of -1.5"/deg for TEMP_MEAN_SLOPE has been used over the weekend, but with poor results, especially this on UT20041003 :

  FE    No.data   del -vs- temperature                   Temp. range
 -----  -------   -------------------------------------  -----------
 SCUBA     20     del = (-1.9  +- 1.8 ) * (T_f - T_b)    
                  del = (-0.92 +- 0.15) * (T_f + T_b)/2    4  to  12

T_f = mean temperature of 4 probes on the front legs, etc.

Despite having gone through once before (see the report of 20040729) I suspect a sign error and fear that I am in danger of making TEMP_MEAN_SLOPE increasingly negative as a result. The value has been reset to -0.5" and we shall collect data again.