...distribution
The decay times are measured through the decay paths, and the time difference t is given by 160#160 , where 161#161 (162#162) is the decay path into charged (neutral) pions and 163#163 cm is the mean decay path of the 164#164 (see the appendix).
...mm.
We acknowledge V. Patera for providing us with the simulation of Fig. 2.
...(#gamab#460>)
If CPT is not conserved the only change in time evolution equations concerns the normalization factor, which becomes 225#225.
...intervals
The general constraints on the time intervals for 226#226 tagging are: 227#227 and 228#228, with 229#229. A good choice is given by 230#230.
...gets
In the following equations we include also the CPT-conserving higher-order terms, namely the ones proportional to 294#294.
...convention,
The 346#346 decay amplitudes are parametrized in each isospin channel like the semileptonic ones, 347#347 is the CPT-conserving part and 348#348 the CPT-violating one. The Wu-Yang convention is 349#349.
...negligible
In 408#408 we will neglect possible CPT-violating effects, thus in the following we will assume 409#409.
...by
In the following we neglect possible violations of CPT and of the 525#525 rule.
...text
In all the previous formulae 698#698 denote the proper times. In this appendix, 699#699 indicate the proper times and 700#700 the laboratory times.
...[#maianib##1#].
We are very grateful to L. Maiani for clarifying discussions, and for providing us with the appendix of his work [43].
...amplitude
We choose the final state 724#724, however the results can be easily generalized to any 725#725 final state.

Carlos E.Piedrafita