Transversal spreads (Q700068)

The author generalizes results of Knarr by an investigation of spreads in three-dimensional projective spaces that are realized as transversals to derivable nets. He shows how the extension techniques of T.~G. Ostrom can be related to the question of transversal spreads and how to interconnect these ideas with the geometric embedding process to determine that all spreads are planar transversal spreads. Using these ideas the author discusses the structure of derivable affine planes which admit various collineation groups. In particular, he shows that a finite derivable affine plane of order \(q^2\) which admits a collineation group leaving the derivable net \(D\) invariant and acting flag-transitively on flags of lines not in \(D\) must always be a semi-translation plane which admits either an elation group of order \(q\) or a Baer group of order \(q\). When the semi-translation plane is a translation plane, the structure of the spread is more-or-less determined in this situation. Relaxing the assumption on partial flag-transitivity the author also proves Theorem 2. (1) If a derivable affine plane \(\pi\) of order \(q^2\) admits a linear \(p\)-group of order \(\text{gcd}(2,q) q^5\) fixing the derivable net \(D\) then there is a subgroup acting transitively on the affine points. (2) Furthermore, the group contains either an elation group of order \(q\) or a Baer group of order \(q\) with axis a subplane of \(D\). If the order of the stabilizer \(H\) of a point is at least \(2q\) then the order is \(2q\) and \(H\) is generated by an elation group of order \(q\) and a Baer involution with axis in \(D\) or by a Baer group of order \(q\) and an elation. (3) \(\pi\) is a nonstrict semi-translation plane of order \(q^2\) admitting a translation group of order \(q^3 p^{\gamma}\). Furthermore, either \(\pi\) is a translation plane or there is a unique \(((\infty), l_{\infty})\)-transitivity and the remaining infinite points are centers for translation groups of orders \(q p^{\gamma}\). (4) If \(\pi\) is non-Desarguesian in the elation case above then \(\pi\) admits a set of \(q\) derivable nets sharing the axis of the elation group of order \(q\). The paper contains quite a number of interesting ideas and results which are too numerous to be reflected here. For these the interested reader is referred to the original paper.