Since the 4782 and 4783 were ~2% low another hall proe was put into 4783 and the magnet was remeasured.  The measurements agree to about 0.34% which seems
reasonable given the alignment and hall probe calibration errors. 

The run 1 was done with probe LPT 141 serial number 001141012, use this measurement. 
The run 2 was done with probe MPT 231 serial number 01431025