This paper discusses from a historical standpoint the incomplete reasoning of Book I proposition 9 of Archimedes' On the sphere and cylinder and gives a proof, which only uses means available to Archimedes himself. This paper also discusses some demonstrations which, through the centuries, have been proposed to solve this problem. One of these is the previously unpublished proof by Jakob I Bernoulli, which is here reported, recently discovered at the Universitätsbibliothek of Basel.