in. (39. cm.); between anterior superior spines of ilia, 21. in. (53 cm.); waist measure at umbilicus, 42.5 in. (106. cm.); girth at ensiform, 35. in. (87.5 cm.). It should be explained, however, that there was a considerable layer of fat in the abdominal wall, which made the apparent extent of contact of the tumor with the wall (16.5 in.) more than was really the case. The tumor was somewhat irregular in outline and firm in consistency. No fluctuation was perceptible externally but bimanual examination revealed at the left a cyst apparently in the lower portion of the tumor. The cervix was the seat of extensive scarring and the uterus was drawn or pushed upward and backward. The uterine cavity measured 4.5 in., which pointed to a uterine source of the tumor. The examination of the blood and urine was negative.

The abdomen was opened in the median line, at first by a short incision for examination, but later this incision was extended upward to a point midway between the ensiform and the umbilicus and downward to the symphysis. The tumor was found smooth and glistening and at its upper side was widely adherent to the cmentum. Upon its surface were a number of large, distended blood-vessels. It appeared in portions to contain fluid but the effort to reduce the size of the tumor by the insertion of a trocar was unsuccessful.

The uterus was found attached to the lower right side of the tumor. As a subsercus fibroid, the new-growth had developed laterally separating the layers of the left broad ligament and stripping upward the posterior pelvic peritoneum to a point where this was deflected from the bowel. The tumor was isolated and severed from the uterus, its removal laying bare the iliac vessels. The tumor presented lobulations of vary ing size and consistency; portions had a peculiar gelatincus character and the blood-supply, as stated, was strikingly rich, the whole suggesting that, while the main part of the growth was an ordinary uterine myoma, there were within it the elements of a malignant change. The subsequent microscopical examination of the tissue showed this change to be sarcomatous. These considerations determined the removal of the uterus and adnexae. The weight of the tumor was ten and one-half pounds.

Following the operation the patient reacted poorly. Saline infusions, both subcutaneously and by rectum, were given. Stimulation was applied. Posture was made favorable by elevation of the foot of the bed. On the night following operation, she vomited "a dark-brown fluid almost constantly," the next day less frequently. The stomach was washed occasionally. The abdomen became moderately tympanitic. Ou the third and fourth days she still vomited occasionally and after this the same dark fluid, containing fine coffee-ground masses, was withdrawn in quantities through the stomach-tube; and meanwhile the strength was failing and the pulse, which had risen suddenly to 142 accompanied by a fall in temperature on the night following operation, still remained high and of poor quality. Nutritive enemata of predigested food and saline, given at first, were not well retained and had to be discontinued. Early in the morning of the fourth day, the gauze drainage in the pelvic cavity was removed. At 10 A. M. of the fifth day she died.

Autopsy was held at 5 P. M., and bacterial cultures were made from the first opening in the upper abdominal wall and from the bloody exudate between the intestinal coils in the lower part of the abdomen. The length of time intervening

before the autopsy would remove any positive value of these cultures, since any organisms found might be claimed to have wandered from the intestine, were it not that the organisms found in the peritoneum at autopsy were the same as had been isolated from the gauze packing at least 24 hours before death. These organisms were the staphylococcus pyogenes aureus and the bacillus coli, both of which infections may fairly be supposed to have wandered through the walls of the intestine more or less reduced in resistance by the manipulations necessary to the removal of the extensively adherent tumor. A moderate degree or peritonitis, most advanced in that part of the abdomen from which the tumor had been removed, was present. This had doubtless taken or its active character on the evening of the third day, simultaneously with the rise in temperature to 101.2 but it is very doubtful whether this process was an important factor in the exitus, as that will be otherwise accounted for.

In the pleural cavity on each side were firm old adhesions, more extensive on the left between the 5th and 8th ribs. Both lungs were congested but contained air throughout. The heart was normal. The intestines in the lower part of the abdomen, as stated, were covered with a bloody exudate, which between the intestinal coils made them slightly adherent. The omentum where it had been severed from its attachment to the tumor was dark and swollen. Its lateral edges were still densely adherent to the abdominal parietes, apparently as a result of the pressure of the tumor against it. The spleen and pancreas were normal in size; the liver was not greatly enlarged but its edges were thickened and rounded, due, as was to be seen on section, to a fatty infiltration. The gall-bladder was distended with bile, which did not pass under moderate pressure. No calculi were found within it and the blocking was probably due to congestion. The kidneys were normal. Alongside the right ureter, about 5. cm. from its vesical meatus was found a small, intensely hard, yellowish or whitish, oval, calculous mass, the origin of which was unaccounted for by anything in the patient's history. No enlarged lymphatic glands were detected.

The most important pathological condition was that observed in the stomach. This organ, normal in size and position, when opened was found to contain a litre or more of a thin grayish-brown fluid. Its walls were thin and flabby, and over the whole mucous surface were ulcerative patches varying in size from a pin-head to a bean. In outline some were circular, some slender, almost linear. In general they were slightly oval. In that portion of the stomach surrounding the pylorus, especially extending away along the lesser curvature, these patches were smooth or slightly elevated and grayish against a pinkish background. Along the greater curvature and all about the cardia and fundus, it was apparent that the grayish sloughing patches had broken down and left shallow, bleeding ulcers. The whole aspect of this part of the mucous membrane was that of extensive, almost confluent ulceration. When held up to the light the blood-vessels running in the wall seemed to contain small fresh thrombi. Microscopically there was a distinct difference between the lesions observed at the two ends of the stomach, corresponding to the differences pointed out in gross appearance. At the pylorus, the antrum, and along the lesser curvature, the mucosa was the seat of extensive small-cell deposits which lay along the connective tissue septa or upon the submucosa but not deeply infiltrating the latter. The blood-vessels of mucosa and submucosa alike were engorged with blood. The surface and glandular epithelium were in a state of cloudy swelling throughout and the process was progressively more advanced toward the surface. Only rarely did it appear that the mucosa had actually sloughed away. At the cardia and along the greater curvature, on the other hand there were relatively few of the small-cell deposits, the blood-vessels were not so engorged, and the surface of the mucosa was ragged with sloughedout areas of varying depth, but none extending into the submucosa.

The probable cause of death in this case is the hemorrhage that must have been going on, as indicated by the rise in pulse-rate and the vomiting of coffee-ground material, from the night after operation until the end. Such cases have been reported but rarely, and of especial interest are the articles of von Eiselberg in Langenbeck's Archiv (Bd. 59, s. 837), in which seven cases of hemorrhage from fresh gastric or duodenal ulceration after various abdominal operations are detailed, and of Mintz, in Mitteilungen aus den Grenzgebieten der Medizin und Chirurgie (Bd. 6, s. 645), in which he discusses the question of postoperative gas

Fig. 4.

Perineorrhaphy, for "Incomplete" Laceration. Postericr vaginal wall brought forward to show course of superficial suture line.

tro-duodenal hemorrhages. Von Eiselberg reported that in six of his seven cases there had been the tying off of portions of omentum. No traction on the stomach had been made in any of them. In three, the healing had been by first intention. In two there had been peritonitis before operation and in one there was a mild postoperative peritonitis. He points out that in many of these cases the first vomiting of blood, just as in our own, occurs on the night following operation and thinks that this points to trauma as a possible cause of the thrombosis and embolism that lead to the hemorrhage. Billroth observed a fatal hemorrhagic gastritis in a case after the removal of a goitre. He thought sepsis the cause and spoke of the digesting action of gastric juice upon the mucosa. rendered hyperaemic by the ordinary causes of circulatory stasis. Such a condition was present in our case as a result of the extensive fatty change in the liver.

Case XIII. Reflex Disturbances of Pelvic Origin. (Dec. 7, 1905). Mrs. W. V., aet. 35. Family history negative. Menses began at 15. Was married at 17; was never pregnant. Duration of menses one week, amount scant. Bearing down, heavy, aching pain in the pelvis throughout this time. She has at present constant backache and pain in the lower abdomen. Is constipated. Extremities always cold. Has been troubled with desire for frequent urination since an attack of cystitis occuring soon after marriage. Bimanual examination shows the uterus to be freely movable. Uterine appendages show no prolapse nor adhesions. A profuse muco-purulent discharge flows from the endometrium.

The cervix was dilated with graded sounds, the endometrium was curetted, and the hood of the clitoris freed from its adhesions.

The bladder was shown to be normal by cystoscopic examination and the occasion was taken to demonstrate the normal appearance of the ureteral meatus. Such a case is often unfortunately subjected to prolonged treatment by irrigation of the bladder, urinary antiseptics, etc... but there is no doubt that her bladder symptoms are purely of a reflex

nature.

We have done what is indicated in a woman of her age. It may be all that is necessary for her permanent relief. Some just such cases will return unimproved, if they return at all, and a considerable proportion will require after all a radical operation.

Oct. 1, 1906, the family physician reports that she "may be a little better."

Case XIV. Suppurative appendicitis. Generalized peritonitis. Intestinal paresis. (Operation, Nov. 16; demonstration in clinic, Dec. 9, 1905).

Mr. I. U., aet. 19. Gave history of one previous attack of appendicitis, about six months ago, when he was in bed two days. On Nov. 6th he was again seized with severe cramping pains in the epigastrium with vomiting. In an hour or two the pain and tenderness became localized in the appendiceal region. I saw him in consultation two days later with general peritonitis and I advised against operation. He entered the Hospital on Nov. 11th with temperature at 102.8 and pulse 136. There was excessive tympanites, with an area of dulness in the region of the bladder. The intestinal walls were distended to the point of paralysis.

Incision was made in the median line between umbilicus and pubis. 9.5 cm. in length. This opened up a large abscess filling the pelvis and limited at this time by the intestinal coils adhering above it. The pus was evacuated after protecting the general cavity by gauze towels.