5
$\begingroup$

Theorem. there are infinitely many primes of the form $6k+1$.

I've just proved that there are infinitely many primes of the form $6k+1$.

Could you please check my proof?

At first, I proved that

$$ for \ \ p:prime, \ p \ge 5 \\\ \\ \left(\frac{-3}{p}\right)= \begin{cases} 1,& \ p \equiv 1 \pmod 6 \\ -1, & \ p \equiv 5 \pmod 6 \end{cases} $$ (I will use this lemma for proving Theorem.)

$$$$ NOW assume that there are finite primes of the form $6k+1$.

then we can say $ \ p_1, \ p_2, \cdots, p_k \ $ : all the primes of the form $6k+1$.

Let

$$ n=(p_1\cdot p_2\cdots\ p_k)^2 +3, $$

then (by Fundamental Theorem of Arithmetic) there is prime factor $p$ of $n$.

$$ $$ Id est, $$(p_1\cdot p_2\cdots\ p_k)^2 \equiv -3 \pmod p $$

So, $$p \equiv 1 \pmod 6$$

Thus $$p=p_i \ for \ some \ i=1, \cdots , k$$

This time

$$p=p_i \ \ \ divides \ \ (p_1\cdot p_2\cdots\ p_k)^2 \\ p=p_i \ \ \ can't \ \ divide \ \ 3 \\ $$

So, $$p=p_i \ \ can't \ \ divide \ \ n. \\$$

It is contradiction with "$p$ is prime factor of $n$"

$\therefore \ $ There are infinitely many primes of the form $6k+1$.

$Q.E.D.$

$$ $$ What about my proof?

After proving, I saw someone's proof,

BUT he set $$ n=(2p_1\cdot p_2\cdots\ p_k)^2 +3 $$

I don't know why he set $ n=(2p_1\cdot p_2\cdots\ p_k)^2 +3 $, instead of $ n=(p_1\cdot p_2\cdots\ p_k)^2 +3 $.

Is my proof wrong?

$$ $$ Please give me some hand. Thanks in advance.

$\endgroup$
  • 1
    $\begingroup$ Your $n$ is congruent to $4$ mod $6$. $\endgroup$ – Ted Shifrin Jun 8 '14 at 15:12
  • $\begingroup$ @TedShifrin, Ah.. You mean if $n \equiv 4 \pmod 6$, then 2 is prime factor of n... Thus this is wrong, right?? $\endgroup$ – user143993 Jun 8 '14 at 15:19
  • $\begingroup$ I wouldn't say wrong, but incomplete. Note that you can also solve the issue by noting that $x^2\equiv1\pmod8$ for odd $x$, which means your $n$ cannot be a power of $2$, hence has an odd prime divisor. $\endgroup$ – rabota Jun 8 '14 at 15:24
  • $\begingroup$ @barto, Ummm $x^2\equiv1\pmod8$ iff $x \equiv1, 3, 5, 7\pmod8$ iff $x \equiv1 \pmod2$.. How does it go "It means your n cannot be a power of 2" Could you explain more , please? $\endgroup$ – user143993 Jun 8 '14 at 15:36
3
$\begingroup$

There is a minor issue. Not every prime divisor of your $n$ is necessarily of the form $6k+1$. (Because, as you noted, the theorem with the Légendre-symbol is only valid for $p\geqslant5$.) You would have to argue that $2$ and $3$ do not divide $n$.
This can be fixed by letting $n=(2p_1,\cdots,p_k)^2+3$ as you noted, or alternatively by observing that $x^2\equiv1\pmod8$ for odd $x$, which means your $n$ cannot be a power of $2$, hence has an odd prime divisor.

$\endgroup$
  • 1
    $\begingroup$ THANKS a lot for answering. I agree "argue that 2 do not divide n", BUT why I argue that '3 do not divide n'? $n=(p_1\cdot p_2\cdots\ p_k)^2 +3$ is the form 6k+1+3=6k+4 .. isn't it? Do I get wrong thing? Give me some advice, please:-) $\endgroup$ – user143993 Jun 8 '14 at 15:26
  • $\begingroup$ It is indeed obvious that $3$ does not divide $n$, but it's worth noting because it's still a crucial step in your proof. $\endgroup$ – rabota Jun 8 '14 at 15:28
  • $\begingroup$ THANK you sincerely. :-) AND Just to be sure...I'm sorry if it offended you. $\endgroup$ – user143993 Jun 8 '14 at 15:33
  • $\begingroup$ Not at all ;-), you're welcome. $\endgroup$ – rabota Jun 8 '14 at 15:34
  • 1
    $\begingroup$ Well done! Apologies for not answering your questions immediately, I wasn't at home last evening. $\endgroup$ – rabota Jun 9 '14 at 7:03
1
$\begingroup$

If we cross out from sequence of positive integers all numbers divisible by $2$ and all numbers divisible by $3$, then all remaining numbers will be in one of two forms:

$S1(n)=6n−1=5,11,17,...$ or $S2(n)=6n+1=7,13,19,....n=1,2,3,...$ So all prime numbers also will be in one of these two forms and ratio 0f number of primes in the sequence $S1(n)$ to number of primes in the sequence $S2(n)$ tends to be $1$. see [link]

$\endgroup$

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.