Ian Cowan and Graham Farquhar have proposed that, to optimize water-use efficiency, plants should regulate their stomata so that the response rates of evaporation (E) and assimilation (A) to a small change in stomatal conductance are constant over the day. That is, the partial derivative of E with respect to A equals lambda, a constant. This dissertation reports a field test of this hypothesis applied to Rhamnus californica Esch., growing in Stanford University's Jasper Ridge Biological Reserve. Evaporation, assimilation, and their relationship to stomatal conductance were followed throughout the California summer drought season, using a combination of field measurements and mathematical modeling. The daily course of each of these parameters was compared with the optimal course predicted by Cowan and Farquhar's model. The major conclusions of this study were, first, that these plants do not regulate their stomata to keep dE by dA constant. Departures from constancy were as large as an order of magnitude, and were especially pronounced near dawn and dusk, and in heavily drought-stressed plants. Secondly, the daily courses of transpiration and assimilation were nonetheless reasonably close to the optimum courses predicted by the model. Finally, seasonal changes in the daily courses of conductance were consistent with the idea that lambda should be lower for drought- stressed plants than for plants that are well watered. These results lead to the conclusion that dE by dA is too sensitive to stomatal opening to explain the behavior of Rhamnus stomata. The water relations and stomatal responses to humidity were also followed throughout the dry season in these plants. The humidity responses account for much of the behavior of the stomata observed in the daily courses of conductance, but not all of it. Seasonal courses of water potential, leaf relative water content, and leaf specific weight are presented.