I observed this property in month of July this year but unable to design a mathematical proof or mathematical way to state my observation. I need help to state this property.

We know that for certain values of 'b' in s = 1/2 + ib , ζ(s) = 0.

When I observed most of the values of 'b' here

I found that most of the primes are related to these values of 'b' in there square forms like:

    Property:    [b] = p^2        {where 'p' is a prime number and [b] is the Box Function}

example: b = 841.0363...

    so,         [841.0363...] = 29^2

below 200 there are only 6 primes which are not following the this above property.

I am searching a formula or a program to find out all those primes which follow above property but till now I didn't got any solution. I am also not good in programming please help me in this problem.


The $n$th zeta zero has $b(n) \sim 2\pi n/W(n/e)$, where $W$ is the Lambert function. The Lambert function grows as roughly the logarithm of its argument, so we see that the zeta zeroes get denser as $n$ increases, and at $n \approx 10^4$, their imaginary parts should start hitting every integer. This turns out to be the case, as the last square of a prime it misses is $103^2 = 10609$.

We can also estimate the probability of each square of a prime not having a zeta zero near it. The density of zeta zeroes is $W(n/e)/(2\pi)$, so we expect each square of a prime to not be near a zeta zero with probability $1 - W(p^2/e)/(2\pi)$. Summing this over all primes where this value is positive, we get an expected count of $7.7$, which is reasonably close to the actual value of 6.

  • $\begingroup$ Thanks for your answer but are you sure that after n=10^4 their imaginary part start hitting every integer. If yes then please tell me where I can see its proof. $\endgroup$ – Dynamo Dec 17 '18 at 16:02
  • $\begingroup$ @AdarshKumar My statement is probably a little too strong, but it is known that there is some $n$ after which $[b(n)]$ hits every integer. It is also known how to count the number of zeros with imaginary part less than $b$, with formula given here. Keeping only the dominant terms and inverting gives my estimate in the answer. It is possible there are integers missed that are significantly larger than $10^4$, but prime squares have very low density at large $n$, making them unlikely to be counterexamples anyways. $\endgroup$ – eyeballfrog Dec 17 '18 at 17:09
  • $\begingroup$ so it means that after some n which you say that will surely exist after which b(n) hits every integer than it is also possible all primes whose square is greater than n also exist . So my property is true till infinite primes. ?? $\endgroup$ – Dynamo Dec 17 '18 at 17:15
  • $\begingroup$ Yes, this property will hold for all sufficiently large primes. Though since this is true because it holds for all sufficiently large integers, of which primes are a subset, I'm not sure how much this tells us about primes. $\endgroup$ – eyeballfrog Dec 17 '18 at 17:18
  • $\begingroup$ ok thanks a lot to giving your time to solve my problem. $\endgroup$ – Dynamo Dec 17 '18 at 17:21

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.