Summary. Single clover plants were grown in the vegetative state, at 20 ± 10, 85 + 5 % relative humidity, 320 + 10 ppm CO, 12-hour day, with Hoagland nutrient in Perlite, and 100 w * m-2 of photosynthetically active radiation (0.4-0.7 ,u) from mercury-fluorescent lamps. Each plant was confined within a circle 18 cm in diameter by means of a wire framework. The COO exchange rate of the whole plant was measured every second day for 3 months. There was no optimum leaf area index for the net photosynthesis rate. The respiration rate was determined mainly by the gross photosynthesis rate and only partly by the amount of non-photosynthetic or heavily shaded tissue. At the maximum leaf area index, when leaves were dying as fast as they were being produced, both photosynthesis and respiration remained at or near their maximum rates. At the end of 3 months, the whole plant was harvested and the dry weight and carbon content determined. The measured dry weight was close to that calculated from the total CO) uptake and a constant ratio of carbon content to dry weight of 39 %. Optimum leaf area indices observed in field experiments are attributed to the failure to include the material which dies between harvests, and to decreases in the gross photosynthesis rate caused by climate changes or lack of nutrient, for example. The difference between production rate and growth rate or yield is emphasized.We have shown previously (14) that young clover plants adapt their rates of respiration (R) when subjected to a change in the incident light level. The adaptation is virtually complete within 24 hours, the final respiration rate being proportional to the photosynthesis rate (P) over a wide range of light levels.In that experiment, each leaf received the same amount of light. In a plant communitv, the lower leaves are shaded by the upper ones. If the same adaptation occurs, the respiration rate of the whole community will be proportional to its photosynthesis rate. Both P and R will increase with leaf area until there are sufficient leaves to absorb all the incident radiation, and then remain constant ( fig la). This picture is quite different from the one given by most models of the net P of plant communities (6,11,15,16,18,19,23,24). In these models, R is assumed to be proportional to leaf area and independent of P, with the result that net P passes through a maximum at a point known as the optimum leaf area index1 (f g lb).