Proposition A

Every non-trivial finitely generated group has a maximal proper subgroup.

Proof

Let GG be a non-trivial finitely generated group. Let C\mathcal{C} be a chain of proper subgroups of G,G, with snCs_n \in \mathcal{C} and s0s1s2.s_0 \subseteq s_1 \subseteq s_2 \subseteq \ldots. Clearly S=CS = \bigcup \mathcal{C} is a subgroup of G.G. Suppose GS.G \subseteq S. In this case, there must exist nNn \in \mathbb{N} such that {g0,g1,,gm}sn,\{g_0, g_1, \ldots, g_m\} \subseteq s_n, where g0,,gmg_0,\ldots,g_m constitute a finite generating set of G.G. In this case sns_n fails to be a proper subgroup of G,G, contradicting our premises. Thus, SS is a proper subgroup of G.G.

Since every chain of proper subgroups of GG has an upper bound which is a proper subgroup of G,G, Zorn’s lemma applies. Thus there exists a maximal proper subgroup of G.G.

\blacksquare

Proposition B

The group (Q,+)(\mathbb{Q},+) has no maximal proper subgroup.

Proof

TODO

\blacksquare