We will start with the known identity that


and assume for contradiction that $e=\frac pq$. Then we have that $p=qe$, where $p$ is an integer. So then $q!e$ is also an integer number (specifically $p(q-1)!$). This means that


Now, we can split this sum into two parts:

$$q!e=\sum_{n=0}^q \frac{q!}{n!}+\sum_{n=q+1}^\infty\frac{q!}{n!},$$

and note that each term in the first sum is an integer (because $n!|q!$ when $n\leq q$). On the other hand, in the second sum,


where there are $n$ terms in $q!(q+1)\cdots(q+n)$. This gives us our contradiction. Since the first sum in $q!e$ are all integers and the second sum is positive and less than 1, $q!e\notin\mathbb{Z}$ and thus $e$ is irrational.

I can't specifically say anything that's wrong with it, but I still don't feel good about it. In particular, I'm skeptical about rewriting $n!=q!(q+1)\cdots(q+n)$ when taking a sum to infinity, as well as the inequality, but I can't specifically say what might be wrong about it.


1 Answer 1


You should have written $$\sum_{n=q+1}^\infty \frac{q!}{n!} = \sum_{n=1}^\infty \frac{q!}{q!(q+1)\cdots(q+n)} \le \sum_{n=1}^\infty \frac{1}{(q+1)^n} = \frac{1/(q+1)}{1 - 1/(q+1)} = \frac{1}{q} < 1$$ since $q > 1$. Note the index of summation shifts. The argument you made is sound.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .