John Skoyles emphasizes the importance of practical experimental work for the developing scientist (Nature 508, 319; 2014). But theory is crucial too — for interpreting the results and for advancing research.

In 1928, British physicist Paul Dirac came up with his equation, which predicted the existence of the positron. The particle was discovered independently four years later by US physicist Carl Anderson at the California Institute of Technology in Pasadena. This is a good example of “theory guides; experiment decides” (the motto of analytical chemist Izaak Maurits Kolthoff).

Or as the astrophysicist Arthur Stanley Eddington declared: “It is also a good rule not to put overmuch confidence in the observational results ... until they are confirmed by theory.”

Perhaps the last word should come from Soviet physicist Peter Leonidovich Kapitza. He said: “Theory is a good thing, but a good experiment is forever.”