On a geometric interpretation of multiplicity