It has been shown previously (Duke, 1951; Nisell, 1950)

RECENT observations in the intact animal indicate that a decrease in 02 concentration in the inspired air may act, as does CO2, on the pulmonary vessels to produce constriction [for references see Duke, 1949 a; Dirken and Heemstra, 1949 a, b, c; Nisell, 1950]. Isolated perfused lungs provide a means of determining whether these effects are independent of neuro-humoral reflexes. When this investigation was begun, Lohr [1924] had used this technique to study pulmonary vasomotor responses of the cat in response to anoxia and hypercapnia. The same problem had been investigated in anaesthetized cats [Retzlaff, 1913;Wearn, Ernstene, Bromer, Barr, German and Zschiesche, 1934;Von Euler and Liljestrand, 1946; Logaras, 1947] and in the cat heart-lung preparation [Drinker, Churchill and Ferry, 1926]. The results considered in the present paper dealing with the effects of 02 and CO2 on the isolated lungs of this species were briefly reported [Duke, 1949[Duke, b, 1950 earlier. Since this investigation was begun, Nisell [1948Nisell [ , 1950 has reported the results of a similar study. METHODS.Isolated cat lungs were set up and perfused using essentially the same technique as that described by Daly [1938] for the lungs of dogs. Cats weighing more than 2-0 kg. were anaesthetized with intraperitoneal pentobarbitone (0.03 g./kg.) or chloralose (0.1 g./kg.) and bled from the carotid artery until death occurred. Heparin (1000 I.U. "Liquemin," Roche) was injected intravenously before bleeding, and 1000-1500 J.U. were added subsequently to each 100 c.c. blood collected. Artificial ventilation (Starling "Ideal" pump) was maintained during bleeding and subsequent dissection. Cannulw were inserted into the trachea, the left auricle and the pulmonary artery, and the ventricles were compressed by tying them tightly with tape. When positive pressure ventilation was used during perfusion the lungs and heart were left in situ, but when negative pressure ventilation was used they were removed from the chest.

It has been shown previously (Duke, 1951; Nisell, 1948 Nisell, , 1950 that ventilation of isolated perfused cat lungs with gas mixtures containing a lower 02 content than atmospheric air produces an increase of pulmonary arterial pressure. Further analysis of the pressor response (Duke & Killick, 1952 a) showed that it was apparently dependent on changes in the 02 tension of the blood in the pulmonary vessels or in the alveolus and that changes in the blood 02 content were unimportant. Later (Duke & Killick, 1952b), it was found that lowering the alveolar 02 tension or the 02 tension of the pulmonary venous blood was a more effective way of influencing the pulmonary arterial pressure than lowering the 02 tension of the blood in the pulmonary artery. The present experiments were carried out to locate more accurately the site of action of anoxia on the pulmonary blood vessels of the isolated lung. Experiments were also made using the perfused left lung in the living anaesthetized animal in order to see whether anoxic pressor responses occur under more normal conditions than those of the isolated lung. The perfused left lung preparation allows a more precise control of pulmonary circulation than was possible in the experiments of Euler & Liljestrand (1946). METHODSIsolated cat lungs were set up and perfused through the pulmonary artery with the animal's own heparinized blood as previously described (Duke, 1951). Perfusion was at constant volume inflow with a Dale-Schuster pump. The venous outflow from the lungs was collected into a venous reservoir which fed the pump. Ventilation was by positive pressure using a Starling 'Ideal' pump. In some experiments the positive pressure inflation was kept at a constant value, usually 10-12 cm H20, and the volume of air not entering the lungs at each stroke of the pump was measured by the method of Konzett & Rossler (1940). This procedure gives a measure of the resistance of the lungs to inflation. Dextran ('Intradex' Glaxo) was occasionally used to augment the blood obtained from the animal. The pulmonary arterial pressure was recorded using a small capacity tambour and could also be read on a manometer filled with 0 9 % NaCl solution. The capacity of the pulmonary arterial pressure recording system was such that 0-1-0-3 ml. of blood were taken up in it for each centimetre rise in pressure.

Spontaneous respiration of small concentrations (e.g. 1-2 %) of carbon monoxide in air or oxygen causes a progressive formation of carboxyhaemoglobin, thereby reducing the blood oxygen content. Providing that the decrease in oxygen carrying power of the blood is not too great, spontaneous respiration is maintained and consequently the arterial oxygen tension can be but little affected. It is of interest, therefore, to examine the impulse activity of the chemoceptor nerve fibres in such conditions, where despite a decrease in blood oxygen content the arterial oxygen tension does not undergo a significant change.In addition, we have investigated whether carboxyhaemoglobinaemia modifies the chemoceptor response to the lowering of arterial oxygen tension produced by the inhalation of 5% 02 in N2. METHODSSeven successful experiments were carried out using cats anaesthetized with chloralose (50 mg/kg) and urethane (250 mg/kg body weight) given intraperitoneally. The trachea was cannulated, and connected to inspiratory and expiratory water valves. Suitable gas mixtures were administered via the inspiratory valve. The carotid artery or the femoral artery which was used for the registration of arterial blood pressure was connected to a condenser manometer (devised by H. W. Ead) coupled to a cathode-ray oscillograph.The carotid sinus nerve was dissected and cut centrally. Most of the baroceptor fibres were removed by thinning the nerve, the chemoceptor fibres being preserved. All dissections were performed with the aid of a binocular microscope, care being taken to avoid damage to the arterial blood supply and venous drainage of the carotid body. The nerve twig was laid on nonpolarizable electrodes and the impulse activity recorded via a resistance-capacity coupled amplifier connected to a second cathode-ray tube. Simultaneous photographic records were taken of the systemic arterial blood pressure, the electroneurogram and a time marker.The opposite femoral artery was cannulated; samples of blood were removed at intervals and

No abstract