Brocard's problem asks if $(n-1)(n+1)$ is ever a factorial. My question is similar: is $n(n+1)$ ever a factorial?

This can be seen as the special case $k=2$ of the question: for $2\le k\le n-2,$ when is $n!/(n-k)!$ a factorial? I know of only one case, $10!/7!=6!$ (see A109095).

I have verified the absence of solutions for $n<10^{85}$ so their absence seems certain. Can this be proved? (Has it been?) I would also be interested in information on the general problem.

Edit: Having recently regained some interest in this problem, I verified it up to $m\le10^9$ or $n<10^{4282852761}$ using modular arithmetic to 37 large primes. (Each value of $m$ required 37 modular multiplications and an average of 2 Legendre symbols.)

  • 2
    $\begingroup$ you must have meant $n>2?$ $\endgroup$ – lab bhattacharjee Jul 19 '13 at 18:28
  • $\begingroup$ @labbhattacharjee: Yes. $\endgroup$ – Charles Jul 19 '13 at 18:29
  • $\begingroup$ Is your computer verification for the $k=2$ case specifically or the general arbitrary $k$ situation? $\endgroup$ – anon Jul 19 '13 at 18:29
  • 1
    $\begingroup$ math.stackexchange.com/questions/446904/… is related to the second paragraph question. $\endgroup$ – JB King Jul 19 '13 at 18:41
  • 2
    $\begingroup$ @HagenvonEitzen: Yes, that was the method I used, except that I used prime powers rather than primes. But it turns out it's more efficient to just iterate through factorials... $\endgroup$ – Charles Jul 19 '13 at 19:10

This is an interesting problem. I don't have a solution just some observations.

$n$ and $n+1$ are relatively prime via Euclid's algorithm:

$$\gcd(n+1,n) = \gcd(n,1) = \gcd(1,0) = 1$$

The two sequential numbers, therefore, share no common factors. Only one of the numbers is even (for obvious reasons), so it must contain $2$ to some power $x$. However, it does not contain all powers of $2$. The powers it may contain follow a sequence: $1,3,4,7,8,10,11,15,16,18,19,22,23,25,26,31,32,34,35,38....$

For example, $4 \times (\text{only odd factors})$ will never produce a factorial. However, $4 \times (3 \times 5 \times 7) = (4 \times 5) \times (3 \times 7) = (20)(21).$

I don't know if it's headed in the correct direction, but if you could use this fact to cover the entire set of integers you could prove that $n\times(n+1)$ never is a factorial except for the trivial case already mentioned.

  • $\begingroup$ Will it be prudent to consider legendres equality and show that it can not hold for a number (n+1)n $\endgroup$ – ARi Jul 31 '13 at 15:58
  • $\begingroup$ Why cant the expression for inverse gamma function be used, to shew that an inverse can not exist for n(n+1) $\endgroup$ – ARi Jul 31 '13 at 16:42

Solving a quadratic for $ n $ and choosing a positive root, we get:

$ 2n=\sqrt {1+4k!}-1 $

So all we need to show is that the only cases in which $ 1+4k! $ is a perfect square are when $ k=2 $ and $ k=3 $.

P.S.: Sorry for my non-number-theory notation style.

  • $\begingroup$ How do you prove that $4k!+1$ is a perfect square only when $k=2$ and $k=3$? $\endgroup$ – Bumblebee Nov 15 '17 at 20:05

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.