If I have a bipartite graph $G(V_1 \cup V_2)$ and I wanted to find the maximum matching. Why is it that the ford-fulkerson algorithm will take at most $min{(|V_1|,|V_2|)}$ iterations? I know that in each iteration it will find an augmenting path from source to sink, but I do not understand the limits on the number of iterations.
Any help is greatly appreciated.