Joins of $\sigma$-subnormal subgroups

Let $\sigma=\{\sigma_j\,:\, j\in J\}$ be a partition of the set $\mathbb{P}$ of all prime numbers. A subgroup $X$ of a finite group $G$ is~\textit{$\sigma$-subnormal} in $G$ if there exists a chain of subgroups $$X=X_0\leq X_1\leq\ldots\leq X_n=G$$ such that, for each $1\leq i\leq n-1$, $X_{i-1}\trianglelefteq X_i$ or $X_i/(X_{i-1})_{X_i}$ is a $\sigma_{j_i}$-group for some $j_i\in J$. Skiba~[12] studied the main properties of $\sigma$-subnormal subgroups in finite groups and showed that the set of all $\sigma$-subnormal subgroups plays a relevant role in the structure of a finite soluble group. In [5], we laid the foundation of a general theory of $\sigma$-subnormal subgroups (and $\sigma$-series) in locally finite groups. It turns out that the main difference between the finite and the locally finite case concerns the behaviour of the join of $\sigma$-subnormal subgroups: in finite groups, $\sigma$-subnormal subgroups form a sublattice of the lattice of all subgroups [3], but this is no longer true for arbitrary locally finite groups. This is similar to what happens with subnormal subgroups, so it makes sense to study the class $\mathfrak{S}_\sigma^\infty$ (resp. $\mathfrak{S}_\sigma$) of locally finite groups in which the join of (resp. of finitely many) $\sigma$-subnormal subgroups is $\sigma$-subnormal. Our aim is to study how much one can extend a group in one of these classes before going outside the same class (see for example Theorems~3.6, 3.8, 5.5 and 5.7). Also, $\sigma$-subnormality criteria for the join of $\sigma$-subnormal subgroups are obtained: similarly to a celebrated theorem of Williams (see [15]), we give a necessary and sufficient conditions for a join of two $\sigma$-subnormal subgroups to always be $\sigma$-subnormal; consequently, we show that the join of two orthogonal $\sigma$-subnormal subgroups is $\sigma$-subnormal (extending a result of Roseblade [11]).