The paper is focused on some aspects of experimental realism of Ian Hacking, and especially on his "manipulability" criterion of existence. The problem is here related to chemical molecules, the objects of interest in chemical research. The authors consider whether and to what extent this criterion has been applied in experimental practice of chemistry. They argue that "experimentation on" is a fundamental criterion of existence of entities in chemistry rather than "experimentation with". Some examples regarding studies of structures of complex compounds, taken from organic chemistry, are presented to support the authors' considerations. Chemists' laboratory practice depends strongly on the way that representations of entities on (or with) the experiments are used. The authors show that this point has not been given sufficient attention by the new experimentalists.

No abstract