From analysis of the inelastic $e\ensuremath{-}p$ scattering data it is found that an important feature of the parton probability function $P(N)$ is that it should be peaked around some small number of partons. The most probable number of partons is prognosticated to be around 4.5. Contrary to the usually accepted forms of $P(N)$, where it decreases uniformly from a maximum, we find that the data also allow a behavior of $P(N)$ where it rises to a maximum first and then decreases with increase of $N$. It is conjectured that there is a connection between this peaking and the observed peaking of the multiplicity distribution in inelastic collisions.