On the Cayley graphs of completely simple semigroups. (Q535219)

This paper deals with the geometric objects ``Cayley graphs'' over the special algebraic structure ``completely simple semigroups'' \(S=\mathcal M[G;I,\Lambda;P]\). The definition and detailed properties of the semigroup \(S\) can be found in Howie's famous book under the name of ``Rees Theorem''. Up to isomorphism, all completely simple semigroups can be defined as \(S\) above. As depicted in the introduction part of this paper, there are so many studies of Cayley graphs of (infinite) semigroups. For a completely simple semigroup \(S=\mathcal M[G;I,\Lambda;P]\) and for a non-empty subset \(A\) of \(S\), the authors present the following main results in the paper under review: i) A sufficient condition for a Cayley graph \(\text{Cay}(S,A)\) of \(S\) to be a disjoint union of complete graphs. ii) Necessary and sufficient condition for the special subgraphs (related to \(\mathcal L\)-classes, \(\mathcal R\)-classes and \(\mathcal H\)-classes) of \(\text{Cay}(S,A)\) to be strongly connected bipartite. Honestly it is quite a well written and important paper. However, I have some doubts about the first result. It is well known that the following are equivalent: \(\bullet\) \(S\) is a completely simple semigroup. \(\bullet\) \(S\) contains a minimal right (left) ideal. \(\bullet\) \(S\) is the union of all its minimal right and left ideals. (In fact, detailed references about the proof of this can be found in [Proc. Camb. Philos. Soc. 57, 234-236 (1961; Zbl 0239.20077)], written by \textit{R. McFadden} and \textit{H. Schneider}.) In this paper, the first result basically states that each of the subgraphs \(T_\lambda\) is complete. But, since \(S\) is a completely simple semigroup and since it is the union of all its left and right ideals, I suspect this first result can be seen directly by following the definitions of left (and right) ideals in any semigroup and the definition of \(T_\lambda\) as in here. I mean the problem can be simplified as ``whether the minimal left (or right) ideal of a semigroup represented by a complete graph'', and so the answer seems clear to me.