Question: Suppose $G$ is a group and $H$ and $K$ are subgroups of $G$ such that $|H|=39$ and $|K|=65$. Prove that $H \cap K$ is cyclic.
[It is Question-$9$ of Chapter $10$ of Abstract Algebra by Dan Saracino]

Up to this chapter, Sylow Theorem has not been introduced, and we should solve this by Lagrange Theorem or Conjugacy Classes.

I solved this, but I think my solution is too complex. I write my answer here. Would you tell me if there is a better solution for this question and whether my solution is correct or not?

My solution:

Possible order for elements of $H$ are : $1, 3, 13, 39$

Possible order for elements of $K$ are : $1, 5, 13, 65$

If $H$ or $G$ do not contain an element of order $13$ then the intersection would be $\{e \}$ which is cyclic. So, I investigate the case where there is an element of order $13$ in both subgroups.

Now I am going to prove $H$ can not contain two or three cyclic subgroups of order $13$. If it contains an element of order $13$, it must only contain one subgroup of order $13$.

In the following different cases, $o(x_i)=13$ and $o(y_i)=3$:

Case 1: $H=\{e,x_1,x_1^2,\cdots,x_1^{12},x_2,x_2^2,\cdots,x_2^{12},x_3,x_3^2,\cdots,x_3^{12},y,y^2 \}$

Case 2: $H=\{e,x_1,x_1^2,\cdots,x_1^{12},x_2,x_2^2,\cdots,x_2^{12},y_1,y_2^2,\cdots,y_7,y_7^2 \}$

Case 3: $H=\{e,x_1,x_1^2,\cdots,x_1^{12},y_1,y_1^2,\cdots,y_{13},y_{13}^2 \}$

Case 1 and 2 are not possible because:

$\langle x_1,x_2\rangle \subset H$ and $|\langle x_1,x_2\rangle|>39$ which is contradiction.

Hence, only Case 3 is possible, and, indeed, in this case, $H=\{e,x,x^2,\cdots,x^{12},y,y^2,xy,xy^2,x^2y,x^2y^2,\cdots,x^{12}y,x^{12}y^2 \}$

Now, if $K$ does not contain $x$, then $H \cap K=\{ e\}$, which is cyclic. If $K$ contains $x$ then $H \cap K=\{ e,x,x^2,\cdots,x^{12} \}$, which is cyclic.

  • 3
    $\begingroup$ The only possible orders would be $1$ and $13$. $\endgroup$
    – lulu
    May 25 at 20:37
  • 2
    $\begingroup$ That goes not just for the order of the elements in the subgroup $H \cap K$, but for the order of the subgroup itself. $\endgroup$ May 25 at 20:40

3 Answers 3


Since $H,K\le G$, we have $H\cap K\le G$; therefore, in particular, $H\cap K$ is a group. By Lagrange's Theorem, the possible orders of $H\cap K$ are $1$ and $13$. The case $1$ is trivial. If the order is $13$, then, since $13$ is prime, $H\cap K$ is cyclic.


I do not quite follow your proposed proof (how do you come to those cases? why if $K$ contains $x$ then ...?), but as pointed out in comments, this exercise has a much easier proof:

The order of the group $H\cap K$ must divide both orders of $H$ and $K$. So it can only be ... or ..., and in either case, ...

  • $\begingroup$ Let's say $C_i$ are cyclic subgroups and $\bar{C_i}=C_i - \{e \}$. I came up with those cases because I consider a group should be a union of disjoint $\bar{C_i}$ and $\{e \}$. Also if $K$ contains $x$, then it should contain $<x>$, and since $H$ contains $<x>$ the intersection is $<x>$. $\endgroup$
    – khashayar
    May 25 at 21:15
  • $\begingroup$ Even if you mean proper cyclic subgroups, your first assertion does not hold already for $\mathbb Z/4$ or $\mathbb Z/6$. For the second argument, well why is the intersection not bigger? $\endgroup$ May 26 at 16:22

$H\cap K$ being a subgroup of $H$, its order must divide $39$, and being a subgroup of $K$, its order must divide $65$ as well. The only two numbers that divide both $39$ and $65$ are $1$ and $13$. The trivial group is cyclic, and any group of prime order is cyclic, too.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.