It seems that I’ve found a bug in version 10.5: DNA chemistry examples work incorrectly when G4Ellipsoid is used as world volume.
How to reproduce a bug: just replace G4Box with G4Ellipsoid in DetectorConstruction.cc file of chem1 or chem5 example.
It works correctly with G4Sphere and incorrectly with G4Ellipsoid even when ellispoid is exactly sphere (all semi-axes are even). Bug can be easily seen in chem1 example’s output: chemistry stops at 1 picosecond instead of 1 microsecond.
I would appreciate if you tell me how to fix this.