The provocative experiments of Staub (1) on the color of blood in arterioles of rapidly frozen lungs prompted us to investigate the quantitative aspects of gas exchange in vessels larger than capillaries in a living animal. His study raised the question of whether the amount of precapillary gas exchange is sufficient to invalidate the physiologic measurement of a) intrapulmonary right to left shunt while breathing oxygen (2), b) pulsatility of pulmonary capillary blood flow (3), and c) pulmonary arterial blood volume (4). We shall show both by measurements utilizing the ether plethysmographic method (4, 5) in a living dog and by calculations of the diffusing capacity of the pulmonary arterial system in a mathematical model that precapillary gas exchange is insufficient to invalidate the aforementioned techniques. MethodsTheory. The injection of ether dissolved in alcohol into the pulmonary artery of an apneic dog enclosed within a body plethysmograph produces a rise in plethysmographic pressure. This is due to the dissolved ether in the blood becoming gaseous in the alveoli and signals the arrival of ether at the site where the major portion of gas exchange takes place. If ether is dissolved in Lipomul,' a fine emulsion of cottonseed oil in water, ether evolution in the lungs is not delayed because the particles of the emulsion are so small that they pass through the pulmonary capillaries (4). However, if ether is dissolved in kerosene, the rate of ether gas evolution from the kerosene is retarded, presumably because kerosene lodges in a pulmonary vessel that limits the diffusion of ether (5). The rapid injection of one liquid into another liquid with which it is immiscible produces droplets whose size is related to the velocity of injection and the kinematic viscosities of the liquids (6). The size of the kerosene droplets containing the ether can be estimated micro-

No abstract