Ivan Shishkin, Rye (1878)

Discussions

Finite groups with 2 conjugacy classes

0 messages

Solution

Solution by Sequoia · EN

Let GG act on X:=GX:=G by conjugation, that is via the action (g,x)G×Xgx=gxg1G(g,x)\in G\times X\longmapsto g\cdot x=gxg^{-1}\in G.

Thus the orbit of xx under this action is Ox:={gxg1,gG}O_x:=\{gxg^{-1}, g\in G\}, i.e., the conjugacy class of xx, and its stabilizer is Stab(x):={gG/gxg1=x}Stab(x):=\{g\in G/ gxg^{-1}=x\}, i.e., the set of elements that commute with xx.

We know that the set of orbits forms a partition of XX, that is, there exist x1,,xrXx_1,\dots,x_r\in X such that X=i=1rOxiX=\bigsqcup_{i=1}^r O_{x_i}. But there are only two conjugacy classes here, so r=2r=2. Also, one of these classes contains the identity element ee, whose orbit is reduced to itself since it commutes with everyone. We thus have the equality X={e}OxX=\{e\}\sqcup O_x for some xXx\in X.

Let us now consider cardinalities: we find #G=1+#Ox\#G=1+\#O_x. But the cardinality #Ox\#O_x can also be written as #G#Stab(x)\frac{\#G}{\#Stab(x)}, so we finally obtain #Stab(x)=#G#G1\#Stab(x)=\frac{\#G}{\#G-1}, which must be an integer. This is only possible if #G=2\#G=2.

Conversely, if GG contains only two elements, it indeed contains only two conjugacy classes since it is abelian, which concludes the proof.

No messages yet.