(RANKIN) of equivocation information (1-:) and interaction information (M). The method is described in the present paper for I: and in a previous paper (Orloci, 1976) for M. The results presented in this paper suggest that for Species Rank order Information Percentage of total species to be weighted according to their suitability to I· M I M r M characterize isolated groups of releves in a phytosociolo­ 5 7 54.15 2.31 17.97 0.82 gical table, the equivocation information may serve as a 9 5 49.86 23.19 16.55 8.22 3 3 9 47.79 0.56 15.86 0.20 suitable weight. The appropriate formulations are derived 6 4 8 36.18