(a) Multiple scattering and energy loss in absorber are now calculated

(a) Multiple scattering and energy loss in absorber are now calculated
    using absorber material properties (density, radiation-length, ...).
(b) All the informations about the absorber are get from the geometry rootfile
    instead of from constants in AliMUONConstant.
(c) The track extrapolation to vertex has been removed from the tracking
    procedure since the vertex is not known at this level.
(d) An option has been added to cancel the track extrapolation to vertex
    when filling ESD in AliMUONReconstructor. ESD are then filled with track
    parameters at the first chamber. Similar option have been added in
    MUONmassPlot_ESD.C and MUONefficiency.C to perform extrapolation at
    this level if not done before. (a detailled mail will follow this commit).
(Philippe P.)