Although the basic chemical features of coal can be qualitatively and in some instances semiquantitatively specified, "average-structure" models that purportedly reflect statistically preferred molecular structures of coal offer little that advances an understanding of coal. In part, this is due to a continuing paucity of relevant or reliable data and to the procedures used to formulate the constructs. But meaningful representations of molecular structure are currently also precluded by indications that the assumption that underlies average -structure models, namely, that there exists a more or less unique, systematic, rank-dependent, molecular chemistry of coal, is not sustained by the current evidence. Several examples, all drawn from the open literature, are presented to support the view that the chemistry of a coal is heavily influenced by its source materials and early formative history and that coals of similar rank may therefore be chemically much more diverse than is usually supposed.
SINCE PUBLICATION OF THE NOW-CLASSIC STOPES-WHEELER MONOGRAPHon coal constitution in 1918 (1), the diffuse coal band that delineates itself in C versus H plots (2) or equivalent diagrams (3) has come to be regarded as implying a genetic rather than merely generic relationship among coals of different rank; and this has prompted the view that there exists a system atic, rank-dependent, molecular chemistry of coal which, if more completely understood, would allow the chemical properties and behavior of coal in specified reaction regimes to be predicted from simple rank indicators. Con temporary basic coal research is therefore heavily concentrated on structural elucidation of a (hypothesized) characteristic "average coal macromolecule"