A geantino finds wrong intercepts of solids inside replicas. The case is a tube divided in phi - twice. See code snippet and PostScript file in: http://www.hep.man.ac.uk/~johna/replicas.lis http://www.hep.man.ac.uk/~johna/replicas.ps
This problem has been solved since some time. I leave it to its new owner to document it.
Fixed since more than two years !! The problem was due to mis-match in the visualisation representation, if I remember well....