The main methods of defining the “quality” of an X-ray beam are discussed and criticised. It is concluded that the simple half-value method is, at the moment, least open to criticism.Spectra of high voltage radiations through Pb, Sn and Cu show the superiority of Sn as a filter. This is demonstrated in many ionisation experiments. Combinations of metals are discussed and suggestions made for practical filters.The question of the most suitable metal for h.v.l. measurements is raised, and it is suggested that tin is the most suitable material.