(present: TI, TH, RK, ML, MF, AA)

N.B.: The meeting time has now been changed to Friday morning, 11 o'clock.

Tayfun investigated the problem he presented last week - he had found electron runs with several percent of all events that showed a good amount energy in the EMEC (but less than the total signal peak) and very little in the HEC. Tayfun and Margret found that for all of these events the global cubic time is right on the edge of the time distribution for that run. For those events, the reconstructed energy in the second layer of the EMEC is very small. (Tamara found similar problems in pion runs she has looked at.) What is still not fully resolved are the events where there is plenty of energy in the HEC and none in the EMEC several likely theories where suggested (eg: bad timing, threshold problems), they still need to be investigated.
Tayfun looked also at the beam chamber reconstruction in a total of 23 runs. 10 of these runs had bad beam chamber reconstruction, and they mainly affect runs in which impact point D has been hit. Among the `good' runs however, there are 7 runs that have the above mentioned problems of lots of energy in EMEC and none in HEC or vice versa.

Tamara is working on minimizing the sigma/E of her data by varying the HEC weight (w) when she adds the EMEC and HEC energies: Etotal=Emec+w*Ehec. For 60GeV pions she finds w=1.045.
Last week, she mentioned that her 100GeV data point has really poor resolution. She looked into that and plotted the HEC energy vs the EMEC energy and finds that there are lots of events that have energy in one, but not in the other, and the remaining events are completely uncorrelated. This run is obviously garbage and no longer used in the analysis.
Tamara also looked at the global cubic time and plotted the energy in the EMEC, HEC and the Etotal vs time. She finds some anti-correlation between the distributions for the HEC and the EMEC: The HEC distribution has a little dip, and the EMEC distribution has a little peak at the same time. Michel pointed out that this anticorrealtion could be related to the fact that the global cubic time is obtained from actual calorimeter data.