—Infection of the blood-clot is the complication most to be feared.
—Small effusions of blood become encysted and are gradually ab sorbed, leaving behind them an area of pleura (or pericardium) obliterated by adhesive inflammation. Infection is in frequent. Not so, however, with the more considerable effusions. For these the danger of infection is paramount. It threatens from all sides; from the thoracic wound, from the bronchi (though the smaller bronchi are germ free), from any inflamed areas in the lung (and inflammation of a lacerated lung is common), possibly from the blood—and the bacteria, once intro duced, demand no better culture medium; hence an empyema of sud den onset (see below). But infection may be averted. Then the blood is slowly absorbed as such. Clotting seems not to occur in the pleura of man.
The two therapeutic indications are to stop the bleeding and to prevent infecton. The treatment of cardiac haemorrhage has already been described (see Section II), and the ne cessity of identifying and checking parietal haemorrhage has been insisted upon. There remains hiemorrhage from the great vessels and from the lung. For the former nothing can be done be yond keeping the patient absolutely quiet, using morphine liberally to that end, and possibly by the use of gelatin in 1-per-cent. solution hypodermically or in the rectum. For haemorrhage from the lung the same expectant course is probably the best, though venturesome surgeons would have us cut down upon and pack the visceral wound. To per form this operation through an exten sive wound in the side is all very well; but to make the wound is asking too much of the patient's strength. If left to itself the lung will continue to bleed until it has completely retracted, unless there are adhesions, and the pleura is full of blood. The tension of the blood in the pleura will then tend to compress the lung, and thus to favor clotting, aided by the syncope into which the pa tient has fallen. To hasten this syncope by venesection, thereby drawing the blood of a patient who is already bleed , ing dangerously, is a quaint custom proved by certain English writers. It
is safer to wait, though perhaps "splint ing" the lung by the injection of air into the pleura might occasionally be beneficial (see below).
When the wound of the lung is giving only slight luemorrhage, the external wound should be closed with gauze and the physical signs of bleeding should be watched for. When the hahnorrhage is more marked, the chest is to be opened. a small drainage-tube inserted, and the admission of air regulated according to the difficulty of respiration in the patient. When the hiemorrhage is large and its symptoms alarming, the chest should he opened and a large drainage-tube in serted so as to form a rapid and com plete pneumothorax, and at the same time, when necessary, salt solution should be injected into a vein. When this does not control the }hemorrhage, one or more ribs are to be resected and the bleeding radically dealt with. In severe 'hemor rhage from a lung, the first object of treatment should be to get pressure on that lung. By opening the chest, air will do this as well as blood in the pleura; it will do it instantly instead of waiting until a sufficient amount of blood has been poured out; it will save to the patient the amount of blood necessary to exert this mechanical pressure; it will permit the vessels to close by clots; and the remote benefits are that it eliminates the dangers of a pyothorax or of univer sal adhesions of the pleura. The danger from engorgement of the right side of the heart must be slight where the pa tient has lost much blood. Because it is a great danger in a full-blooded dog is no reason why it should be so in an ex sanguined man, for the conditions are different and the cases not parallel. These are personal views, and are not indorsed by the majority of surgeans, • but they are the result of experience, and seem logical and correct. II. G. le Conte (Phila. Med. Jour., Apr. 14, 1900).