Automorph-conjugacy is not finite-intersection-closed

From Groupprops

This article gives the statement, and possibly proof, of a subgroup property (i.e., automorph-conjugate subgroup) not satisfying a subgroup metaproperty (i.e., finite-intersection-closed subgroup property).
View all subgroup metaproperty dissatisfactions | View all subgroup metaproperty satisfactions|Get help on looking up metaproperty (dis)satisfactions for subgroup properties
Get more facts about automorph-conjugate subgroup|Get more facts about finite-intersection-closed subgroup property|

Statement

We can have a group G with two automorph-conjugate subgroups H,KG, such that HK is not automorph-conjugate in G.

Proof

Example in the symmetric group

Further information: symmetric group:S6

(This example demonstrates the stronger fact that automorph-conjugacy is not conjugate-intersection-closed).

Let G be the symmetric group on six letters: {1,2,3,4,5,6}. Let H,K be the following 2-Sylow subgroups of G:

H=(1,3,2,4),(1,2),(5,6);K=(1,2),(3,5,4,6),(5,6)

In other words, H is the internal direct product of a 2-Sylow subgroup on {1,2,3,4} with the 2-Sylow subgroup on {5,6}, while K is the internal direct product of the 2-Sylow subgroup on {1,2} with a 2-Sylow subgroup on {3,4,5,6}.

The intersection is given by:

HK=(1,2),(3,4),(5,6).

Now, note that:

  • Both H and K are automorph-conjugate, because they are both Sylow subgroups, and Sylow implies automorph-conjugate.
  • HK is not automorph-conjugate. To see this, note that G has an outer automorphism that sends transpositions to triple transpositions. Under this automorphism, HK goes to a subgroup of G that contains three commuting triple transpositions. If this is conjugate to HK, then HK should also contain three commuting triple transpositions. But it doesn't.