to the traumatism inflicted upon the sympathetic ganglia of the pelvis.

SECONDARY HEMORRHAGE.

Whenever large vessels in the body are secured, either in continuity of tissue or en masse, this accident may follow; and the vaginal operation is no exception to this rule. The vessels may be perfectly secure under the forceps, and yet secondary bleeding occur any hour between the time they are removed and two weeks later. The bleeding usually springs from one uterine artery, and is readily controlled by bilateral pressure. (Fig. 93.) A narrow

FIG. 93.-Péan's long retractor. I use two in making the pelvic Mikulicz packing.

retractor is introduced through the center of the column of gauze, and one-half of the gauze-that upon the side from which the bleeding comes-is pulled hard against the lateral pelvic wall. A similar retractor is entered alongside the first, and the other half of the gauze pulled to one side. When it is seen that the pressure is sufficient to stop the bleeding, the vaginal packing is increased by the introduction of additional pieces of gauze between the two retractors. The retractor which holds back the gauze over the bleeding vessel is not to be moved until the dressing is complete, but the adjustment and compression of all fresh pieces of gauze are effected

by means of the opposite blade. After waiting a few minutes to see whether the bleeding is stopped, the patient is put to bed, the foot of the bed being elevated.

If the pressure does not control the hemorrhage, the patient is placed in Sims' position and given chloroform. All dressings are removed, and the bleeding vessel sought for. Descent of the intestines is prevented by gauze pads, and the bladder is sharply retracted with the trowel, while the perineum is held back by a Sims' speculum. When the spouting vessel is seen, it is grasped with bullet-forceps, which take a firm hold on the tissues, and the stump is lifted away from the vaginal wall. It is then an easy matter to grasp the stump with forceps. The vagina is to be packed with iodoform gauze. If after searching carefully the bleeding is seen to come from above the vaginal vault, and the vessel cannot be found, the hemorrhage springs from an ovarian artery. When the operator is convinced that this is the case, he does not attempt to secure the vessel through the vagina with forceps, nor to compress it with gauze, but, after packing the vagina with gauze to prevent descent of the intestines, he throws the patient into Trendelenburg's position and opens the belly. When he has found the source of the bleeding, the artery is tied with silk and the stump trimmed. The same is done with the other ovarian artery. The ligatures are cut short, and the pelvis cleared of clots. The abdomen is closed. It is well to give a high enema of three pints of salt solution before the patient leaves the table, or to inject sterile filtered normal salt solution into a median cephalic vein. I have seen this accident but once, in one of my earliest cases.

If it be found that the bleeding comes from the azygos artery or other vaginal branch, it is best secured by passing a curved needle around it and tying en masse with silk. I can not conceive it possible that so tortuous and long a vessel as the ovarian artery can bleed after its current has been completely shut off for two days. It is probable that the ovarian artery bleeds because the occlusion has been partial and incomplete, and after the

removal of the forceps the blood stream bursts through whatever clot has formed in the vessel. It is not so with the uterine artery. After this vessel is clamped, but little of its length remains between the forceps and the internal iliac artery, and, consequently, when the forceps are removed the end of the artery feels the full force of the pressure from the iliac.

I can not explain the very late hemorrhage occasionally occurring when the patient is ready to get up, except upon the hypothesis that the repatency of the artery becomes established. That this does occur I have shown. It has been observed after abdominal hysterectomy with ligature, and has heretofore been ascribed to bleeding from anastomotic vessels. It is always from the uterine artery or its branches, and is easily checked by forceps applied through the vagina.

INTRAVENOUS INJECTION OF NORMAL SALT SOLUTION.

A seven-tenths of one per cent. solution of chemically pure sodium chlorid in soft water is made.

This is fil

15

tered into either a Florence flask-to be found in all drug stores or else into a perfectly clean agate kettle. It is then boiled ten minutes and is cooled by placing on ice. The solution is employed at a temperature of 105° F. The infusion apparatus is composed of a twelve-ounce glass funnel, eight feet of pure gum rubber tubing to fit this, and a canula (Fig. 94). The apparatus is boiled twenty minutes in plain water. The hand grasps the arm above the elbow and compresses the veins. The median basilic vein will show running across the bend of

the elbow from without in (Fig. 95). The skin is drawn. upward and is incised carefully alongside the upper border of the vein. Upon rolling the skin down into position the cut is found to be over the vein. The vein is carefully dissected out of its bed. The distal or outer end of the vein is grasped across with an artery forceps, and a half inch internal to this the vein is caught with mousetooth forceps. While this is being done an assistant whose hands are absolutely clean, has filled the infusion funnel. This he holds six feet above the patient. The clothing in the patient's axilla has been loosened. The operator severs the vein entirely across and takes the canula in his right hand while holding the bleeding end of the vein with toothed forceps. The saline solution is allowed to flow against the cut end of the vein until the solution feels warm, then the canula is inserted well into the vessel; at the same time, the pressure on the arm is loosed. The assistant watches the flow of water from the funnel, and warns the operator when he is to refill it, so that the operator may compress the tube and prevent entrance of air. To avoid this, all the water is not allowed to flow from the funnel before refilling. The speed of flow is about six ounces in three minutes, or about a quart in a quarter hour. Having introduced the desired amount of fluid, the canula is withdrawn and pressure made around the arm. The two ends of the vessel are secured by fine catgut, and the wound stitched by the same material. Iodoform gauze dressing.

Subcutaneous Injection.-The material is prepared as before. Opposite the angle of the scapula and over the margin of the latissimus dorsi muscle, the skin is cleansed. A few drops of cocain solution is injected, or the skin is frozen with a stick of ice dipped in salt and applied. It is incised for a quarter inch. While the edges are held apart, the solution is allowed to flow through the canula until warm, and the canula is plunged into the cellular tissue between the skin and muscle. Ten ounces of fluid are allowed to enter, when the canula is withdrawn and a stitch of catgut used to unite the surfaces. Iodo

FIG 95.-The superficial veins at the bend of the elbow (after Quain): 6. The median basilic vein, into which intravenous salt infusion is made. 4. Cephalic vein. 3. Basilic vein 2. Venæ comites of the brachial artery X. As these latter lie beneath the deep fascia of the arm, they are not in danger in the operation of intravenous infusion of salt solution. (The reader's attention is called to the fact that the elbow vein into which the infusion is made is sometimes the median cephalic, as the veins of the elbow are not constant in their arrangement).