Measurement of leaf length and stem diameter in three Upper Carboniferousl.epidodendronspecies shows an exponential relationship between these parameters. Larger axes bear longer leaves even in material which had apparently completed its primary' growth. It is suggested that this is a direct product of the determinate growth pattern of the ontogeny of arborescent lycopods. Measurements ofCyclostigmafrom the Irish Devonian are not explicable in these terms, and it is suggested that some specimens were immature at fossilization. It is recommended that the term leaf cushion be applied to the expanded leaf base in the Lepidodendrales at all growth stages, and not only after leaf abscission. Problems in the terminology of ‘young’ and ‘old’ growth stages in these determinate plants are di