Direct stomatal responses to humidity have been well documented in the past decade (e.g. Schulze et al., 1975a, b). They appear to increase water use efficiency by reducing the T/P ratio (Schulze et al ., 1975a) and also reduce total water use (Schulze et al ., 19756). The stomatal response to humidity is a response (probably hydroactive) to water stress in the stomatal apparatus (paper in preparation). It was thought that a species which avoids physiological water stress by stomatal closure at relatively high bulk leaf water potentials might also show a greater stomatal humidity response than one in which stornata are less responsive to changes in the bulk leaf water potential. Two leguminous species, Macroptilium atropurpureum (DC.) Urb., which avoids physiological drought by stomatal closure (Dr M. M. Ludlow, pers. comm.), and Desmodium uncinatum DC., which has stornata relatively unresponsive to changes in bulk leaf potential (Dr M. M. Ludlow, pers. comm.), and which both grow in the same region of semi-arid tropical Australia (in Queensland), were chosen to investigate the above possibility. Single leaves of each species were excised under water from glasshouse-grown plants (five replicates for each species); the petioles were connected to a Ganong potometer and the leaves placed, individually, in a leaf chamber (Sheriff and McGruddy, 1976). Air (C02 concentration 300 p.p.m.) was passed through the chamber at a rate which would completely replace the chamber air every 30 s. Leaves were kept at 26*5 °C (±0-5 °C) and illuminated at an intensity of 940 ju E m~2 s1 by a Philips HPLR700W mercury vapour lamp. Water uptake rate from the potometer was measured approximately every 60 min, immediately before changing chamber humidity. Changes in uptake rate were assumed to reflect changes in transpiration rate under equilibrium conditions. Equilibrium was considered to have been established if two consecutive readings, a few minutes apart, were the same. Approximately half of the replicates were carried out with step increases, and half with step decreases, in humidity. From measurements of water uptake leaf diffusive conductances were calculated using the method of Cowan (1977), and expressed as cmol of water lost per unit leaf area per second. Leaf conductance was plotted against the concentration difference between the air and the evaporation sites (AH), assuming these to be saturated with water at leaf temperature (Fig. 1.). Lines of best fit for the conductances of each species were computer plotted as fourth-order polynomials. This was done because the variation in conductances with AH has been found (in this lab.) to fit a polynomial curve fairly closely and a fourthorder polynomial produced the best fit for these data.