The grain refinement of aluminum dates from the early 1930s, when titanium additions were found to affect cast structure. Twenty years later Cibula found boron improved refinement. In the 1960s Kawecki produced master alloys by reacting a mixture of Ti- and B-containing salts. The resulting product was added via rod feeders, a practice still in use today. My introduction to this area was in 1980, working for the Kawecki company. For several years we studied the variability of commercial refiners, and correlated performance to the structure of particles in the master alloy. The results were never fully explained. As a co-worker at the time said: “Not all particles are created equal.” Later I worked in foundries. The performance of grain refiners is alloy specific, in unexplained and mysterious ways. These unsolved mysteries are reviewed together with our current scientific knowledge.