MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

It is easy to see that the function:

$$\zeta(s) \pm \dfrac{1}{\zeta(1-s)}$$

has a pole at each non-trivial zero $s=\rho_n$.

However, after some experiments with this function, I would like to conjecture that:

$$|\zeta(s) - \dfrac{1}{\zeta(1-s)}|$$

only has zeros in the critical strip on the line $\Re(s)=\frac12$, but also that:

$$|\zeta(s) + \dfrac{1}{\zeta(1-s)}|-2$$

always has at least a zero in the critical strip for each $\Re(s) \ne \frac12$ (i.e. all zeros lie off the critical line).

Since both conjectures complement each other, I guess they must be connected in some way.

Using the reflection formula with $\chi(s)=2^s \pi^{s-1} \sin(\frac{\pi s}{2}) \phantom. \Gamma(1-s)$, I rewrote the functions as:

$$\left| \dfrac{\zeta(1-s)^2 \chi(s) - 1}{\zeta(1-s)\chi(s)} \right| =0 \text { and } \left| \dfrac{\zeta(1-s)^2 \chi(s) + 1}{\zeta(1-s)\chi(s)}\right| = 2$$

however this didn't help much solving e.g. the first conjecture that implies $\zeta(1-s)^2 \chi(s) = 1$ only when $\Re(s)=\frac12$ in the critical strip.

Grateful for any steers/hints on how I could best approach this problem.


share|cite|improve this question
Agno, your second conjecture is much harder :-). Couldn't find small counterexamples. Try plotting the function for large ranges and look for negative values (if they would exist). – joro Dec 1 '12 at 8:01

I suppose these are at least 58 counterexample examples for the conjecture about the zeros of $|\zeta(s) - \dfrac{1}{\zeta(1-s)}|$ in the critical strip with $0 < \Im(s) < 306$.

The first few are:

 0.83937756715930464851 + 31.699222084299394447 i
 0.16062243284069535149 + 31.699222084299394447 i
 0.11131948892305866087 + 88.155391355655174901 i
 0.88868051107694133913 + 88.155391355655174901 i
 0.89823546410519697899 + 123.67527158830200698 i
 0.24612771379601443481 + 122.093560045649621 i
 0.10176453589480302101 + 123.67527158830200698 i
 0.75387228620398556519 + 122.093560045649621 i
 0.26796662801293497366 + 134.16024368224812469 i
 0.73203337198706502634 + 134.16024368224812469 i
 0.86157401152020623657 + 140.44980499028326033 i
 0.86157401152020623376 + 140.44980499028326032 i
 0.13842598847979376624 + 140.44980499028326032 i
 0.89589742682348946277 + 158.2591758599028948 i
 0.10410257317651053723 + 158.2591758599028948 i
share|cite|improve this answer
@Thanks Joro. The master of root finding at work! Think these are indeed exceptions for the first part of the conjecture, but what about the second part? Could there be a zero on $\Re(s)=\frac12$ for $|\zeta(s) + \dfrac{1}{\zeta(1-s)}|-2$ ? – Agno Nov 28 '12 at 9:06

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.