×Time: 14:15 on 22 Oct, 2015
Place: Speigelgasse 5, 05.002
Christian Elsholtz (TU Graz/FIM)
A solution of a problem of Ramanujan of 1915
It is well known that the number of divisors $d(n)$ is large, when $n$ is a product of many primes. Wigert determined the maximal order of magnitudeof the divisor function:\[d(n) \leq \exp ( (\log 2 +o(1)) \frac{\log n}{\log \log n},\]or in other words\[ \max_{n\le x} \log d(n) \sim (\log 2){\frac{\log x}{\log \log x}}.\]Ramanujan (1915) was the first to investigate the maximal order of magnitude of the iterated divisor function $d(d(n))$, giving an example that\[ d(d(n) \ge (\sqrt{2}\log 4 - o(1)) \frac{\sqrt{\log n}}{\log \log n}\]for infinitely many $n$.Erd\H{o}s, K\'{a}tai, Ivi\'{c} and Smati gave upper bounds,but determining an analogue of Wigert's result, i.e. a best possible upper bound, was an open problem.In this talk we give a solution to this problem:\[ \max_{n\le x} \log d(d(n))=\frac{\sqrt{\log x}}{\log_2 x} \left( c + O(\frac{\log_3 x}{\log_2 x} )\right),\]where\[c =\Bigg( 8 \sum_{j=1}^\infty \log^2 (1+1/j) \Bigg)^{1/2} = 2.7959802335\ldots.\](Joint work with Yvonne Buttkewitz, Christian Elsholtz, Kevin Ford, Jan-Christoph Schlage-Puchta.)