Emergence and violation of geometrical scaling in pp collisions

We argue that geometrical scaling (GS) proposed originally in the context of Deep Inelastic Scattering (DIS) at HERA works also in pp collisions at the LHC energies and in NA61/SHINE experiment. We show that in DIS GS is working up to relatively large Bjorken $x \sim 0.1$. As a consequence negative pion multiplicity $p_{\rm T}$ distributions at NA61/SHINE energies exhibit GS in mid rapidity region. For $y \ne 0$ clear sign of scaling violations can be seen when one of the colliding partons has Bjorken $x \ge 0.1$. Finally, we argue that in the case of identified particles GS scaling is still present but the scaling variable is a function of transverse mass rather than $p_{\rm T}$.