On the existence of primary methods of measurement

 There is much discussion in chemical metrology about the definition of primary methods of measurement, just as a couple of years ago there was debate about its predecessors, absolute methods and definitive methods. It is argued in this paper that the designation of certain methods as being primary only makes sense if there is an outstanding property identified that is common to all primary methods, and not present for all non-primary methods. The aim to identify primary methods should not blur our notion that it is the good practice of analytical chemistry that produces good results, not a particular method of analysis.