The problem of molecular structure elucidation is that of identifying a molecule
from a given set of (mostly spectroscopical) data. Thus, in order to make that
a little bit more precise, we have to make up our mind according to what we
do mean by *molecule.* This means we first of all have to tell which
mathematical *models* of a molecule we would like to use.
So lets agree upon that we usually use approximative models for molecules, and
that the first three approximations are the following ones:

- The first approximation is that of the
*chemical formula,*for example, CH for benzene. - The second step of approximating reality is the introduction of
*constitutional formulae,*for example, there are exactly 217 constitutional isomers that correspond to the same chemical formula as benzene, and so this is a considerable refinement compared with the first step of approximation, where there was exactly one model of benzene only. - The third step of approximation takes the spatial arrangement into
account. To a constitutional formula there may correspond several
*configurational isomers,*which contain besides the constitution also geometrical descriptors. For example, in the case of benzene, there are 958 such isomers for the chemical formula; again this step provides a considerable refinement of the second approximation.

Send questions to: molgen@btm2x2.mat.uni-bayreuth.de