In this paper I argue for the idea that, throughout the history of science, there are some cases of theory change that would show how science develops with no referential continuity. For this purpose, I analyze Psillos� proposal of a theory of reference used to account for referential continuity in conceptual transitions. This kind of continuity is requested by Psillos �as by other philosophers� in his defense of scientific realism. By means of a historical case, the theory of germplasm of August Weismann, I intend to demonstrate that Psillos� demand for reference continuity can be found to be inadequate.Neste artigo, argumento em favor da ideia de que, através da história da ciência, há alguns caos de mudança de teoria que mostrariam como ...