The correlation of spectroscopic and chemical investigations in recent years has been highly beneficial of many reasons. Around 1950, no valid explanation was available of the colours of compounds of the five tran­ sition groups. Later, it was possible to identify the excited levels with those expected for an electron configuration with adefinite number of electrons in the partly filled shell. I t is not generally recognized that this is equivalent to determining spectroscopic oxidation states related to the preponderant electron configuration and not to estimates of the fractional atomic charges. This brings in an entirely different type of description than the formal oxidation numbers used for characterizing compounds and reaction schemes. However, it must be realized that collectively oxidized ligands, formation of cluster-complexes and catenation may prevent the oxidation state from being well-defined. The writer would like to express his gratitude to many, but first of all to DR. CLAUS SCHÄFFER, University of Copenhagen, who is the most efficient group-theoretical engineer known to the writer; his comments and discussions have been highly valuable. The writer's colleague, Pro­ fessor FAUSTO CALDERAZZO (now going to the University of Pisa) has been most helpful in metallo-organic questions. Thanks are also due to Professors E. RANcKE-MADsEN and K. A. JENSEN for correspondence and conversations about formal oxidation numbers.