5
$\begingroup$

I've seen the graph for $\varphi(n)$ vs $n$ on Wolfram, and it seems like the $\varphi(n)$ values for primes follow a constant slope, but is there a proof that states that the totient values between two consecutive primes are always smaller than the totient values of those two primes?

EDIT: I got a satisfactory answer for the case where there is a strict inequality. But what is the condition for a relaxed inequality? Is there a proof that establishes $\varphi(n)\le\varphi(p_n)$ and $\varphi(n)\le\varphi(p_{n+1})$

$\endgroup$
6
$\begingroup$

Let's assume that Oppermann's conjecture is true and for every prime $p_n$ and next prime $p_{n+1}$,$$p_{n+1}<p_n+\sqrt{p_n}$$ EDIT: This inequality has counterexample at small prime numbers. The list of known counterexample $p_n$s are $(3,7,13,23,31,113)$ (see A124129)

Now note that every number $k$ such that $p_n<k<p_{n+1}$ is composite, so it has at least one prime factor $q$ such that $q^2\le k$. Now we get$$\varphi(k)=k\prod_{p|k}\left(1-\frac{1}{p}\right)\le k\left(1-\frac{1}{\sqrt{k}}\right)=k-\sqrt{k}<p_{n+1}-\sqrt{p_n}<p_n$$ and it is proved.

EDIT: One can verify or find counterexamples of the inequality manually for small $k$s around above counterexamples.

If Oppermann's conjecture is not true, and furthermore, there exists a prime $q$ and integer $n$ such that $p_n<q^2-q<q^2<p_{n+1}$, then$$\varphi(q^2)=q^2-q>p_n$$so your conjecture has counterexample.

$\endgroup$
  • $\begingroup$ I agree with the answer, but would point out that the question asks if the totient values "are always smaller". $\varphi(3)=\varphi(4)=2$ and $\varphi(7)=\varphi(9)=6$ are counterexamples where the totient value is equal to that for the lower prime. $\endgroup$ – nickgard Apr 12 '17 at 15:13
  • $\begingroup$ Thanks, that does help, but I have a question, let's say I relax the strict inequality. Is the hypothesis known to hold with certainty if equality is allowed between two consecutive primes? $\endgroup$ – Arannya Monzur Apr 12 '17 at 15:50

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.