Improper intersections and a converse to Bezout's theorem (Q1324189)

Let \(X,Y \subset \mathbb{P}^ n\) be algebraic varieties, the Bezout theorem states that \(\deg X \cdot \deg Y = \deg (X \cap Y)\) where \(X \cap Y\) is the intersection cycle defined by Fulton and MacPherson [see \textit{W. Fulton}, ``Intersection theory'' (1984; Zbl 0541.14005)]. There is an ``easy'' case when \[ e:=n + \dim (X \cap Y) - \dim X - \dim Y \] is zero and the difficult one when \(e>0\). The main problem is to construct the intersection components and define multiplicities. There are essentially two approaches: One is given in the book cited above and the second one [see \textit{W. Vogel}, ``Lectures on results on Bezout's theorem'' (1984; Zbl 0553.14022)] states that \(\deg X \cdot \deg Y = \sum_{C \in {\mathcal C}} j(X,Y;C) \deg C\) where \({\mathcal C}\), \(j(X,Y;C)\) are suitable defined. \textit{L. van Gastel} had found the relation between these theorems showing that \(\deg (X \cap Y) = \sum_{C \in {\mathcal C}} (X,Y;C) [C \cap L_ C]\) where \(L_ C\) is a generic linear subspace of dimension \(\dim X + \dim Y - \dim C\). An interesting contribution of this paper is the definition of \(k(X,Y;C)\) a new intersection multiplicity and a criterion which implies \(e=0\), i.e. proper intersection.