(Attached plots) 3rd bullet: PID cut was done for one of the runs. Attached is the plot of momentum v.s. TOF01 before and after the muon cut.

First of all, one sees this clear momentum lobes on the P-TOF plot, which corresponds to a lower peak momentum of ~150 MeV/c. Now if one sees the avrg. momentum plot on X-Y, the green/yellow colors show up at the edge of the irises (r>=100 mm).

Secondly, after the muon cut, the lobes within r<=100 mm still exist. They have ~ 240 MeV/c momentum and correct TOF01, so the lobes are not caused by pion-muon mixing. Of course, the cut also eats in part of the questionable low P pion-like region, which could be also iris-slowed pions, but again, they do not form the tubes.

Therefore, my guess is that the beam, for some reason between TOF0 and tracker (very likely before TOF1) was collimated by something, then starts to possess two lobes and lobes rotate in X-Y during transporting. Durga discovered that the beam is wide in Y at TOF1 already, so could be before then.

Durga stopped by and suggested more robust cuts too, so attached. Previous PID cuts were done "automatically" based on a possible range of TOF01 assuming the particle is muon at each momentum. It was not robust and will keep all the ambiguous ones.

Attached are after the new cuts.
-->First, cut only based on TOF01 with a slice.
lobes still exist, particles outside r=100 mm still exist;
-->Second, cut out only the Gaussian part of the muon peak, which should be robust enough.
lobes still exist, particles outside r=100 mm disappear.

The above proves:
Lobes not formed by different PID or momentum;
Particles with r>=100 mm should be those slowed by irises, which have lower momentum.

That's a fair point. I attached plot from all the 240 MeV/c runs. Again I think the momentum cut will effectively remove the ones slowed by irises (r>=100 mm)

Victoria and Durga doubted whether the hole was created by low P_t particles that were not reconstructed (similar to straight tracks). So checking spacepoints will either confirm or reject this speculation. To be continued.

I have plotted spacepoints rather than reconstructed trackpoints at TKU station 5.

I did a TOF01 cut to select out the muon peak. Results are shown below.

Notice that the previous colormap had "lognormal" scale, so color difference was milder. The newly attached plot uses a linear colormap.

Plot with more statistics from runs
8451, 8456, 8460, 8462, 8467
(muons, spacepoints @ TKU STN 5)

Bear with me... Just consistency check -

240 MeV/c runs without absorber in early Oct. showed the same thing.

Plotted are muon-generated spacepoints at TKU STN 5.

Just to note that I also see the behaviour, only looking at a single run (8451). I also see it in transverse momentum plots.

• If you slice into momentum bins, do you still see the behaviour?

X-Y and X-Px plots attached with momentum slices. Interesting to see how the beam moves with different momenta and avoids the center.

One of the speculations from Chris was that the beam was separated by their momenta and thus different rotation frequency in a solenoid. That can be checked by back-tracking the beam upstream until it does not see the solenoidal field.