08:59 May 22, 2010 

My immediate understanding of this was that, where there is a discontinuity in the ferrous medium, there will be a disruption in the magnetic field. This disruption is detectable in the form of a modification of the spatial orientation (both within and beyond the metallic boundary) of the flux lines. This spatial modification can be visualised by the presence of tiny ferrous particles, each of which is effectively an induced dipole (remember those iron filing experiments with bar magnets?). Personally, I would avoid using the word "scatter" since this generally implies some sort of random process (eg scattering of light), which this isn't.