Vergnaud (1985)suggests a structure of type (S, I, S) which underlies that the construction of a scientific concept is accompanied among others by the emergence of a system of invariant (s) related to that concept. We propose here to observe if, in physical sciences for the case of the concept of chemical element - at the level of fifth form in high scohool the development of the invariant within such a structure (S, I, S) is able, but the condition of the concept itself, to make it more understandable. Our results show that the construction of the concept of chemical element seems to be conditioned by its invariant. However reciprocity does not appear so obvious. We have observed similar trends on other concepts in the physical sciences, that are beeing developped.. This leads us to believe that the invariants could be raised, in this area, to a major educational status . We propose a preliminary classification of invariants in physical sciences, in order to give clarity to complex objects as are scientific concepts to which they are associated. We imagine that all disciplines could "think" their invariants and that they could be used within training organizations for what they are likely to help students during their learning period.
