A @[email protected] that involves both the creation of a new σ-bond between atoms previously not directly linked and the breaking of an existing σ-bond. There is normally a concurrent relocation of π-bonds in the molecule concerned, but the total number of π- and σ- bonds does not change. The term was originally restricted to @[email protected] pericyclic reactions, and many authors use it with this connotation. It is, however, also applied in a more general, purely structural, sense. If such reactions are @[email protected], their @[email protected] may be visualized as an @[email protected] of two fragments connected at their termini by two partial σ-bonds, one being broken and the other being formed as, for example, the two allyl fragments in (a'). Considering only atoms within the (real or hypothetical) cyclic array undergoing reorganization, if the numbers of these in the two fragments are designated \(i\) and \(j\), then the @[email protected] is said to be a sigmatropic change of order [\(i\), \(j\)] (conventionally \(\text{[}i\text{]}\leq \text{[}j\text{]}\)). Thus the @[email protected] (a) is of order [3,3], whilst reaction (b) is a [1,5]sigmatropic shift of hydrogen. (N.B. By convention square brackets [...] here refer to numbers of atoms, in contrast with current usage in the context of @[email protected]) The descriptors a and s (@[email protected] and @[email protected]) may also be annexed to the numbers \(i\) and \(j\); (b) is then described as a [1s,5s] sigmatropic @[email protected], since it is @[email protected] with respect both to the hydrogen atom and to the pentadienyl system: The prefix '@[email protected]' (meaning one extra atom, interrupting @[email protected] — cf. '@[email protected]') has frequently been applied to sigmatropic rearrangements, but is misleading.