The most striking point brought out in this study is that as long as a definite pressure is maintained in either the pulmonary or bronchial circulations, the admixture of bloods is extremely limited. It is easily conceivable that more mixture occurs normally than under the conditions of the experiment, but there is no reason for considering this to be a large difference. If, however, in either system the pressure sinks to zero the possibility of supply by the other system becomes evident. It takes much longer for the mass injected through the bronchial arteries to penetrate to all parts of the lung than when the mass is injected through the pulmonary artery; but when accomplished, the injection reaches to all capillaries including those of the pleura, the only vessels remaining uninjected being the larger trunks of the pulmonary artery. On the other hand, the injection of the bronchial vessels by way of the pulmonary arteries is not complete with normal pressure, but occurs rapidly when a high pulmonary pressure is employed. It is therefore probable that either circulation can suffice for the simple nutritive demands of the lung if the other system is interfered with. It has been shown that embolism of the pulmonary artery, without other circulatory disturbance, does not lead to necrosis of the affected area of the lung, but it is probable that the preservation of circulation is not due to collateral bronchial circulation so much as to the free anastomosis and early division into capillaries of the pulmonary artery. In support of this statement is the fact that the appearance is not altered when the bronchials are ligated at their origin. The same ligation shows no subsequent interference with the nutrition of the bronchi up to a period of five weeks, demonstrating that the pulmonary circulation is sufficient to provide for the nutrition of the bronchi. If, however, as Virchow has shown, the pulmonary artery supplying an entire lobe be occluded, the bronchial circulation can and does suffice for the nutrition of the lobe. In the case of the occlusion of a branch of the pulmonary artery the pressure in the area interfered with does not sink to zero because of the collateral circulation in this area; whereas, if the main trunk is occluded no collateral supply is available, the pressure sinks to zero, and the bronchial artery becomes available as a source of blood supply. It must be remembered that the lung tissue, as a whole, has ready access to oxygen and this gas is the nutritive element acquired by the blood in the lungs. From these studies it would appear that the part of the lung tissue not in intimate contact with oxygen in the air is supplied by oxygenated blood of the bronchial arteries, and that the tissues through which the pulmonary blood circulates take up whatever organized nutriment they need from the pulmonary blood and possibly provide for their oxygen and carbon dioxide interchange (which must be very slight) either directly with the alveolar air, or by finding sufficient oxygen in the venous blood of the pulmonary artery. The studies of the injected specimens confirm Küttner's findings of a very rapid breaking up of the pulmonary artery into capillaries. In all the specimens studied it was found that although the pleural vessels can be injected by way of the bronchial arteries when there is zero pressure in the pulmonary arteries, yet when the two sets of vessels are injected simultaneously in the dog, the pleural vessels invariably derive their supply of injection mass from the pulmonary artery.