A long standing conjecture of Bagemihl (1956) states that there can be at most eight tetrahedra in 3-space, such that every two of them meet in a two-dimensional set. We settle this conjecture affirmatively. We get some information on families of similar nature, consisting of ei…