Some new generalized André systems (Q1822049)

According to H. Lüneburg, a (left) quasifield \(Q(+,\circ)\) is a generalized André system if Q admits a binary operation such that \(Q(+,\circ)\) is a skewfield and the mapping f(a): \(x\mapsto a^{- 1}(a\circ x)\) is an automorphism of \(Q(+,\circ)\) for all a \(Q^*=Q- \{0\}\). A generalized André system obtained from a field \(Q(+,\cdot)\) is an \(\omega\)-system if \(Q(+,\cdot)\) has an automorphism \(\omega\) such that \(f^{-1}(\gamma)\), \(\gamma \in im f,\) is a union of cosets of \(Q^{*\omega \gamma -1}\) in \(Q^*(\cdot)\). Accordingly, a translation plane is an \(\omega\)-plane if it can be coordinatized by an \(\omega\)- system. The author analyses the four large classes of generalized André systems (D), (L), (F) and (R-Z) which are the finite Dickson nearfields, the quasifields described by Lüneburg, the finite quasifields introduced by Foulser and by Rao-Zemmer respectively, and construct two classes of finite and one class of infinite \(\omega\)-systems which are not included in (D)\(\cup (L)\cup (F)\cup (R-Z)\). Furthermore, it is shown that there is a large number of finite \(\omega\)-planes which cannot be coordinatized by any of the quasifields belonging to (D), (L), (F) or (R- Z).