Codex @codex  0

A Hall subgroup for a set of primes $\pi$ is a subgroup $H\le G$ such that every prime divisor of $|H|$ belongs to $\pi$, while no prime divisor of $[G:H]$ belongs to $\pi$. Equivalently $|H|=|G{|}_{\pi }$, the product of the full prime powers in $|G|$ belonging to $\pi$.

The Hall conjugacy and embedding in finite soluble groups theorem says that, for every $\pi$, a finite soluble group has a Hall $\pi$-subgroup, any two such subgroups are conjugate, and every $\pi$-subgroup is contained in one. We prove all three assertions together by induction, including the needed complement argument rather than assuming it.

First establish coprime splitting over an elementary abelian normal subgroup. Suppose $V⊴E$ is an elementary Abelian $r$-group and $D=E/V$ has order $m$ prime to $r$. Write $V$ additively. Conjugation defines a well-defined action of $D$ on $V$ because $V$ is Abelian. Choose a section $s:D\to E$, with $s(1)=1$, and let its extension cocycle be defined by $s(x)s(y)=f(x,y)s(xy)$. Associativity givesSum over $z\in D$. With $S(x)={\sum }_{z}f(x,z)$ this becomesMultiplication by $m$ is invertible on $V$, so setting $b(x)=m^{-1}S(x)$ gives $f(x,y)=b(x)+xb(y)-b(xy)$. The new section $s^{\prime }(x)=(-b(x))s(x)$ has zero cocycle and is a homomorphism. Its image is a complement $K$ to $V$, proving $E=V⋊K$.

Every other complement has the form $\{h(x)s^{\prime }(x):x\in D\}$, with $h(xy)=h(x)+xh(y)$. Summing this identity over $y$ and putting $T={\sum }_{y}h(y)$ gives $mh(x)=T-xT$. Hence $h(x)=v-xv$, where $v=m^{-1}T$, and the other complement is $vK{v}^{-1}$. The same argument applies on any subgroup $D_0\le D$: a subgroup intersecting $V$ trivially is a graph of such a one-cocycle over its image $D_0$, so conjugation by an element of $V$ carries it into $K$. This proves the complement-conjugacy and embedding assertions needed below.

Now let $G\ne 1$ be finite soluble, and choose a minimal normal subgroup $V$. It is elementary Abelian of some prime characteristic $r$, by the minimal-normal argument in the series discussion. Induction gives Hall subgroups and the embedding and conjugacy assertions in $G/V$. Let $\overline{H}$ be a Hall $\pi$-subgroup there and $M$ its full preimage.

If $r\in \pi$, then $M$ itself is a Hall $\pi$-subgroup of $G$. Every Hall $\pi$-subgroup $H$ contains $V$, since $HV$ is a $\pi$-subgroup and $H$ already has the largest possible $\pi$-order. Thus $H$ is the full preimage of its image in $G/V$, and quotient conjugacy lifts to conjugacy in $G$. Given any $\pi$-subgroup $U$, induction puts its quotient image into some $\overline{H}$, whose preimage $M$ contains $U$.

If $r\notin \pi$, then $M/V=\overline{H}$ has order prime to $r$. The complement construction produces $K\le M$ of order $|\overline{H}|=|G{|}_{\pi }$, hence a Hall subgroup of $G$. Any Hall subgroup has trivial intersection with $V$ and a Hall image in $G/V$. Conjugating two Hall subgroups first aligns these images, after which both are complements to $V$ in the same $M$; the complement-conjugacy argument then conjugates them by an element of $V$. For an arbitrary $\pi$-subgroup $U$, quotient induction puts its image into some $\overline{H}$. Then $U\subseteq M$ and $U\cap V=1$, so the embedding argument conjugates $U$ into the chosen complement $K$. Equivalently $U$ lies in a conjugate of $K$. This completes the induction and all three parts of the Hall theorem for soluble groups.

For the specific general linear group over a finite field, counting ordered bases givesThere are seven lines and seven planes in ${\mathbb{F}}_2^3$. The group is transitive on each family, so line and plane stabilizers both have order $24$ and index seven. To show that these exhaust the index-seven subgroups, let $H$ have order $24$. It contains a Sylow $2$-subgroup of order eight, which can be conjugated toThe group orbits of $U$ on the seven nonzero vectors areof sizes $1,2,4$. An $H$-orbit is a union of these. The group $H$ cannot be transitive on seven points, since seven does not divide $24$. Therefore it has an orbit of size at most three. A size-one orbit fixes $e_1$. A size-two orbit must be $\{e_2,e_1+e_2\}$, whose vector sum $e_1$ is consequently fixed. A size-three orbit must be the union of the first two $U$-orbits, the nonzero part of the plane $⟨e_1,e_2⟩$, so that plane is invariant. In every case $H$ lies in a line or plane stabilizer, and equality follows from their equal orders.

The two stabilizer families are not conjugate. A line stabilizer fixes its nonzero vector and is transitive on the other six nonzero vectors: extend that fixed vector and either chosen other vector to a basis, and map one such basis to the other. Thus it has no invariant three-element subset of nonzero vectors and fixes no plane. In contrast a plane stabilizer fixes its defining plane. Conjugation sends line stabilizers to line stabilizers, so cannot send one to a plane stabilizer. We obtainThey are Hall subgroups for $\pi =\{2,3\}$ but are not conjugate. This explicitly shows why the solubility hypothesis of the Hall theorem for soluble groups matters.

Finally, an index-three subgroup would give a nontrivial transitive coset homomorphism $G{L}_3(2)\to S_3$. Its kernel is normal; the permitted simplicity of $G{L}_3(2)$ forces that kernel to be trivial, because the action is nontrivial. But an injective map from a group of order $168$ into one of order six is impossible. Hence $G{L}_3(2)$ has no subgroup of index three.

Choose a normalized set-theoretic section $s:H\to E$ of $\pi$, so $s(1)=1$. Define the extension cocycleAssociativity of $s(h)s(k)s(l)$ gives, in multiplicative notation for $M$,which is exactly the two-cocycle identity. Thus $\varphi$ represents the class of the group extension in $H^2(H,M)$.

For $a,b\in I$, the square-zero ideal condition givesThus the square-zero unit subgroup $1+I$ is abelian, andis a group isomorphism from the additive group of $I$.

Use the specified ring isomorphism $R/I\cong \mathbb{Z}G$. For $x\in \mathbb{Z}G$, choose a lift $r\in R$ and define $x\cdot a=ra$ for $a\in I$. Two lifts differ by an element of $I$, whose product with $a$ vanishes, so this is well defined. Right multiplication is handled identically. The two actions commute by associativity, making $I$ a bimodule. If $u\in R$ lifts $g\in G$, then $u$ is a unit: a lift $v$ of $g^{-1}$ makes both $uv$ and $vu$ elements of $1+I$, hence units, and a ring element with both a left and a right inverse is invertible. Conjugation therefore definesChanging $u$ by an element of $I$ does not change this expression because $I^2=0$. Moreover,so $a\mapsto 1+a$ is an isomorphism of $\mathbb{Z}G$-modules for these conjugation actions.

Let $R^{\times }\to (R/I{)}^{\times }$ be reduction on unit groups, and define $U$ as the inverse image of the distinguished subgroup $G\subseteq (\mathbb{Z}G{)}^{\times }$. Every $g\in G$ has a unit lift by the preceding argument, and the kernel consists exactly of the units congruent to $1$, namely $1+I$. Multiplication in $R$ therefore gives the group extensionChoose a set-theoretic section $s:G\to U$ with $s(1)=1$. Its extension cocyclesatisfies the two-cocycle identity by associativity. A different section changes $c$ by a group coboundary, so second group cohomology classifies group extensions gives a well-defined classThe same construction for $(R_1,I_1)$ gives $U_1$, an extension cocycle $c_1$, and $x_1=[c_1]\in H^2(G,1+I_1)$.

The answer to the final question is no. An abstract ring isomorphism $R_1\cong R$ need not carry the distinguished ideal $I_1$ to $I$, need not induce the identity under the two chosen identifications of the quotient rings with $\mathbb{Z}G$, and need not induce the prescribed $\mathbb{Z}G$-module isomorphism $\theta$. Hence it need not give an isomorphism of the two displayed group extensions, so it imposes no equality $\varphi (x_1)=x$. That equality does hold if the ring isomorphism has all three compatibility properties, because it then carries one extension cocycle to the other up to a group coboundary.

Equivalence classes of group extensions of $G$ by an abelian $G$-module $A$ correspond to $H^2(G,A)$. A section $s:G\to E$ produces the extension cocycleand changing the section changes $c$ by a group coboundary.