By looking at the difference of two production of run12 W,  muEbX48 distribution has shown a unreasonable curve.  In order to find out which runs contributed to it, the PDF movie of muEbX48 for all the 594 runs has been made.  29 runs included in the attached .txt file have been pick out.   Among these runs, there are two L2EW empty runs, 13078037 and 13081019.

The .gif file is the plots  from sum of the 29 runs.

diff29runE.pdf is the PDF diff-movie for these 29 runs.

The same thing happened to the Barrel.  These 27 runs have made all the contribution to the difference for the mubX48 distribution.