The second half of the 19th century saw the rise of new corpuscular theories in physics. Piezoelectricity, discovered in 1880, underwent a development reverse to that of most fields of physics. The first molecular theories were soon replaced by a continuum-phenomeno-logical theory, which guided most research in the field, while, at the same time, physicists continued to propose molecular models. This article studies the reasons for the peculiar development of piezoelectric theory, through a close look at the models suggested and the developments in the field. It argues that the transition originated in a failure of the molecular model to account for new experimental results. Furthermore, later explanatory theories remained speculations, since they failed to derive any relations left unaccounted for by the phenomenological theory and since their basic premises were doubtful. Protagonists also raised general claims about the necessity of a phenomenological theory. Yet they did not derive their position from a doctrinal antagonism to molecular explanation. The same physicists studied both the secure phenomenological theory and speculative explanation simultaneously. They advanced in two mostly separated kinds of researches: molecular-atomistic and phenomenological-positivistic. Nevertheless, the separation between these approaches was not complete. This suggests that physicists embodied the same persona in elaborating ““positivistic”” theory as in elaborating ““realistic”” molecular theory.