The photomultiplier is a device in which each incident phton ejects an electron from a photosensitive surface. The electron is then accelerated via a potential difference and ejects a shower of electrons where it strikes a screen, The impact of a electron is thys converted into a cascade of electrons, which is then converted into a current.
 
Photomultiplier tubes consist of a transparent evacuated tube containing a photoemissive material, a photocathode, and a series of secondary electron emitters called dynodes that provide current gain.

A simple tube without any dynodes is called a phototube or photocell. Oft times the tube is filled with gas to provide secondary electrons from collisions with the photoelectron as it is accelerated torward the anode by the tube potential.

The first dynode is usually biased at a positive potential of about 75-150 V with respect to the photocathode, which has a potential of -1000 ti -2000 V. Thus an electron emitted from the photocathode is accelerated, and, on impact with the dynode, generates additional electrons that are then accelerated torward the second dynode for further amplification.

Common dynode materials include Be-Cu and Cs-Sb alloys, which generate 2 to 6 electrons per incoming electron.
Typically, 5 to 15 dynode stages are used in photomultipliers to obtain gains that can be as high as 108