Constructions of large translation nets with nonabelian translation groups (Q1182419)

While some bounds on the maximum number of parallel classes in a translation net with a non-abelian translation group have been known for some time [see the reviewer, Finite geometries and designs, Proc. 2nd Isle of Thorns Conf. 1980, Lond. Math. Soc. Lect. Note Ser. 49, 172-196 (1981; Zbl 0457.05011)], there was a lack of interesting series of examples showing that a large number of parallel classes can be achieved. In the paper under review, the author uses group theoretic arguments (involving the first cohomology group of the translation group and some series of simple permutation groups, like the Suzuki and Ree groups) to construct such examples. In fact, he obtains series which admit more parallel classes than in the abelian (but not elementary abelian) case. For instance, he constructs a non-abelian translation net of order \(q^ 3\) with \(q^ 2+1\) parallel classes for every odd prime power \(q\).