This article offers a contribution to the history of scientific ideas by proposing an epistemological argument supporting the assumption made by Miller whereby Niels Bohr has been influenced by cubism (Jean Metzinger) when he developed his non-intuitive complementarity principle. More specifically, this essay will identify the Bergsonian durée as the conceptual bridge between Metzinger and Bohr. Beyond this conceptual link between the painter and the physicist, this paper aims to emphasize the key role played by art in the development of human knowledge.