The finite Bruck loops. (Q555561)

Let \((X,\circ)\) be a loop, for all \(x\in X\) let \(\rho_x\colon X\to X\); \(y \mapsto y \circ x\) and let \(K:=\{\rho_x\mid x\in X\}\subseteq\text{Sym}(X)\). If we set \(G:=\langle K\rangle\leq\text{Sym}(X)\) and \(H:=\text{Stab}_G(1)\) then the set \(K\) satisfies the property: \[ 1\in K\text{ and }K\text{ is a transversal to all conjugates of }H\text{ in }G,\tag{*} \] and the triple \((G,H,K)\) is called the Baer envelope of the loop \((X,\circ)\). Vice versa, if \((G,H,K)\) is a triple with \(G\) a group, \(H\leq G\) and \(K\subseteq G\) fulfilling condition (*) then \(K\) can be equipped with a structure of loop called the loop related to \((G,H,K)\). This connection between loops and transversals in groups (introduced by \textit{R. Baer} [Nets and groups, Trans. Am. Math. Soc. 46, 110-141 (1939; Zbl 0022.01105)]) turns out to be fundamental in this paper in order to describe the structure of a finite Bruck loop. For such a class of loops the authors introduce the notion of a Bruck folder and of a twisted subgroup (see Sec. 2) and they prove that every finite Bruck loop \(X\) is the direct product of a Bruck loop of odd order with either a soluble Bruck loop of 2-power order or a product of loops related to \(\text{PSL}_2(q)\) with \(q=9\) or \(q\geq 5\) a Fermat prime. The previous structure theorem allows the authors to generalize the classical group theoretic theorems (Sylow, Lagrange and Hall) to finite Bruck loops. These theorems are proven firstly for the case of finite Bruck loops of 2-power exponent and then in full generality. -- The paper is interesting, well written and the results are presented in a very clear way.