The necessity for abdominal section in this condition is unquestioned. Mann sums up the operative treatment of ectopic gestation as follows:

1. Before rupture, coeliotomy.

2. Soon after rupture, coeliotomy.

3. After rupture with hemorrhage, coeliotomy.

4. Encysted hematocele, early, coeliotomy; late, colpotomy. 5. Encysted hematocele, late or septic, colpotomy.

I desire to enlarge the recommendation for colpotomy to all cases of encysted hematocele. It is not my purpose to open up that apparently irreconcilable discussion about the existence of extra-peritoneal hematocele. But from the accumulated testimony of many scores of observers I must declare my belief in this Occurrence. And also that nearly all pelvic hematoceles have their origin in tubal pregnancy. It is to the pathological anatomy that I invite your attention.

Garrigues, in describing this phase, says: "The blood is at first pure and thin, but becomes coagulated, inspissated, tarry, and still later, sometimes mixed with pus and sanies. Through adhesive peritonitis the intestinal knuckles are glued together and plastic lymph is poured out and converted into tissue, forming a roof over the extravasated blood, which, in places, is finger thick, and shuts it off from the peritoneal cavity."

In discussing primary intra-peritoneal rupture, Bland Sutton says: "When the bleeding is not excessive the blood collects

in the recto-vaginal fossa and floats up the coils of intestines. These, with the omentum, gradually form a covering to the fossa by adhering together so that the blood in the pelvis is isolated from the general peritoneal cavity."

Thus it will be seen that the geography of many hematoceles make them extra-peritoneal, from an adventitious sac of inflam matory exudate, or from primary rupture, into the folds of the broad ligament. The relative frequency of this occurrence, compared to intra-peritoneal hematocele, has been estimated as one to three.

In the majority of cases the fetus dies, and when encysted with the other products it is applicable to the vaginal operation. Here, as in all surgical work, the selection of cases is the subtlest indication of skill and the surest element of success.

In

individualizing cases as to choice of route, the general requirements favorable to vaginal operations should enter. Parous women with broad, roomy pelves being the most suitable, and in this particular condition the situation of the tumor low down is essential. As Henrotin plainly puts it, "If the tumor is low down go at it from below, if high up, from above." He also considers the route in unruptured cases according to their location. The vaginal operation is also advised by Hanks in unruptured cases, but not after rupture; he having had two uncontrollable hemorrhages by that route requiring consecutive abdominal section.

I think with the increasing number of reported cases of operation in unruptured tubal pregnancy we should be more on the alert for it, and when recognized, the simplicity of its removal by abdominal section is only exceeded by its blessed benefaction.

Encysted intra-peritoneal hematocele is differentiated from extra-peritoneal hematocele, or hematona, by the latter being usually smaller, unilateral to the uterus, pushing it over to the opposite side of the pelvis, and unaccompanied by signs of intraperitoneal inflammation. The tumor reaches much lower down, and is more closely attached to the uterus, simulating intraligamentary cysts. This attempted differentiation is only a pedantic refinement that is impractical and immaterial. The treatment of both is identical.

The fact that very few hematoceles undergoing suppuration ever bleed when evacuated, lead us to inquire if they may not be opened prior to suppuration with equal immunity from hemorrhage. The vessels are usually filled with firm thrombi, and the evacuation of such a sac is very simple.

When the hematocele has undergone suppuration its evacuation per vaginam becomes imperative. To all intents it practically becomes a pelvic abscess, and I think the modern employment of vaginal section for this condition has placed its rationale upon a sound and enduring basis. It has been computed that the mortality attending suppurating hematoceles treated suprapubically, is between 20 and 30 per cent., and the death rate of the lower operation, with the advantages of rapid execution, absence of shock, abeyance of threatening sepsis, is practically nil. Adequate preparation for every contingency is a real but

silent factor in the uniform success of latter-day surgery. The resourceful man is the one who has carefully thought out and provided for any emergency. It is largely a matter of prearrangement, and not of intrepid genius that enables the surgeon to meet the unexpected with equanimity and ease. I have long been in the habit of preparing the abdomen and the requisites for its section when undertaking vaginal operations, with the same routine that I have a transfusion canula sterilized with the instruments in other operations of magnitude.

This forethought is not in the nature of a confessed weakness of the vaginal operation, but the recognition of an inherent contingency that should be provided against.

Cases III and IV in my series are illustrations of the occasional necessity for opening the abdomen after the vaginal incision. If the abdomen has to be opened subsequent to the vagina, nothing will have been lost in the attempt to do the operation by the safest method, without sacrifice of any structure, and without subjecting the woman to a serious abdominal operation. It has been contended that the damaged tube is sometimes left, but if a woman becomes well, remains well for several years and has another baby, her cure would seem to be complete enough. This is the history of a case in my knowledge. Kelly's twelve cases

all remained well.

The diagnosis will be confirmed. The clots can be removed with greater facility and will favor the completion of the work of securing any bleeding points abdominally. The safest drainage avenue will be established, and if as Bland Sutton says: "Where blood has remained in the peritoneal cavity for several weeks after rupture it is invariably necessary to drain," it is much easier to make the vaginal opening from below than to open Douglas' space through the abdomen by cutting down on the finger, in the vagina or thrusting a pair of scissors or a puncture-machine through the vaginal vault. Bovee attributes a death after an operation for an intra-ligamentary pregnancy to

inadequate glass drainage that he thinks would have been saved by vaginal drainage.

Many cases have been reported by foreign operators. Herman's classical collection of 33 cases, Martin 58, Masseti in Italy reported a large series in 1891. Many operations have

been done by Dührssen, Schroeder, Péan, Elisher, Schauta and Kossman. Kelly and Watkins in this country have reported 13 and 8 respectively, Noble 2 (suppurating), Frankenthal, Reynolds, Hanks (4), Mann, Beckett, Newman, Bovee, Coe, and many others.

I beg to add the history of two cases of my own, and a third to illustrate the necessity for sometimes having to open the abdomen secondarily.

CASE I.-Mrs. Annie G., white, et 30, multipara, the youngest two years old, uneventful labors; ten-day puerperia. Menstruation began at 13, recurred regularly with twenty-eight day intervals. Reappeared after lactational amenorrhea of over a year in August 1896. In November 1896 she flowed for three weeks at a monthly period attended with bearing-down pains which lasted half an hour at intervals during several days, sometimes confining her to bed. It was presumably a miscarriage. The lower abdomen was tender and enlarged, and the patient supposed she had a tumor. Examination by my father, Dr. W. D. Haggard, disclosed no abnormality. Menstruation continued regular until April 23, 1897. Then it was absent for six weeks or until June 3rd, when she was seized with a sudden sharp pain while stooping over cutting out a garment. It caused her to lie down for a while, after which she resumed her work. Two days afterward the flow began and continued intermittently until the operation, June 30th. June 19th she took her bed. I saw her first June 26th. She had a dozen or more cramping spells" since the first one three weeks before. Her temperature was 99.3°, and pulse 98. The retro-uterine pouch was filled with a tense, round, bulging mass. The uterus could be made out forward and was movable. The mass appeared as large as a cocoa-nut. She came to my infirmary June 29, and was operated on the next day. Diagnosis: Ruptured ectopic gestation. It almost seemed begging for exit. The uterus was curretted and the cul-de-sac opened. Over a pint of blood-clot was scooped out. With hand in the vagina and the fingers in the sac I could map out its relations perfectly. The sac was felt as a distinct roof over the blood-clot. A finger inadvertently made a little aperture in the sac wall. The ruptured tube could not be isolated without breaking through the adhesions, and it

was deemed best to leave them undisturbed. There was no free bleeding. The cavity was packed and the patient evinced no more disturbance than if she had had an abortion. The sac closed rapidly and she sat up on the tenth day, and was discharged in two weeks. Seven weeks after the operation she reported herself perfectly well and weighed several pounds more than before she was taken sick.

CASE II.-Martha C., colored, at 36, multipara, youngest 11 years old. In first labor, which was instrumental, she sustained lacerations of cervix and vaginal walls. No miscarriages. She was in bed three months with "inflammation of the womb" ten years ago. Menstruation which began at the age of 12, was of the monthly type, of normal duration and regular until May 1897, when she skipped a period. In June she flowed continuously for a month, attended with cramp-like, colicky pains, and felt giddy. Flow stopped in July to recur August 14th, and continued until the time of operation, August 21, 1897. She had been confined to bed for a week under the care of Dr. O'Mohundro. She was having spasmodic labor-like pains the while, with great rectal tenesmus and difficulty of urination requiring catheterization. Temperature, 101° to 102°. Pulse, 80 to 90. I first saw her with her physician August 20th. The abdomen was enlarged, tympanitic and very tender in the left lower quadrant. She was having a constant bloody vaginal discharge. A large, tender globular mass filled the entire pelvis, fixing the uterus very far forward on the symphysis and extending in a round end between the rectum and vagina, to within an inch of the outlet. On the abdomen the mass could be discovered extending midway to the umbilicus. Diagnosis: Pelvic hematocele from ruptured ectopic gestation, probably undergoing suppuration. On August 21 the uterus was curetted and packed with gauze before making vaginal incision into the most prominent part of the presenting tumor. The section was purposely made lower down on the vaginal wall than usual to drain the distended recto-vaginal pouch.

Quantities of ill

smelling pus poured out of the opening, and the fingers introduced into the sac brought away large blood clots, some organized and others in the process of disorganization. Over three pints of pus and clotted blood were turned out.