No label defined (Q4946355)

This is a written version of a talk by the author addressing the problem whether the Selbergs Theorem on the existence of a fixed lower bound \(3/16\) of the first eigenvalue \(\lambda_1\) on a large class of arithmetic Riemann surfaces is due to special properties of these arithmetically defined surfaces or whether this is a property `typically' enjoyed by generic Riemann surfaces. The answer is `yes', in the sense that there are positive constants \(C_1\) and \(C_2\) such that some randomly chosen specific surfaces have a bigger smallest eigenvalue than \(C_2\) with probability at least \(C_1\). Technically, the point is to use the combinatorial character of the gluing procedures in the construction of the surfaces and suitable results in graph theory. For the entire collection see [Zbl 0929.00011].