then often interrupt tubal pregnancy long before a rupture takes place. These hemorrhages may very appropriately be designated as “intervillous hæmorrhages."

The two cases of Dr. Manierre, both operated on before rupture had occurred, offer material well adapted to substantiate the claim made above. The microscopic examination of these specimens, in which none of the degenerative changes found can be connected with rupture, because none had taken place, furnishing the following result:

Case I.-Mrs. A. (Fig. 1.) Amnion and chorion are completely

fused. The amniotic epithelium is well preserved. The amniotic mesoderm is likewise in a fair state of preservation. The chorionic mesoderm shows some degenerative changes, its cells are not so well preserved and the nuclei are frequently indistinct, in some places this tissue has become coarsely fibrillar and here nuclei are entirely missing. The chorionic mesoderm is slightly infiltrated with hematoidin granules. The two ectodermal layers of the chorium, the Langhans layers and the syncytium, can in some places be well seen with their distinguishing features. In others, a single layer is present only, either the Langhans, with large swollen cells, or a very flat syncytium;

while in still other spots both epithelial layers have entirely disappeared. The intervillous space contains villi in all stages of degeneration. Perfectly normal villi are entirely wanting. Degenerative stages are shown to a degree of the mere shadow of a villous consisting on section of an indistinct spherical mass with a hyaline band as its periphery and a granular débris as its center. Blood vessels could not be found, either in the chorion or in the villi. The latter are contained in a network of fibrin, granular débris and hematoidin crystals and granules. The gestation sac opposite the placenta is quite thin. Towards the intervillous space it shows a small band of a compacta, composed of large degenerating decidual cells. The band presents an abundant infiltration with round cells of the lymphocyte and the polynuclear leucocyte type. Many of the latter show nuclear fragmentation. Outside of the compact layer of the decidua there are found large open spaces (spongiosa); these are, of course, not true gland spaces, but they have originally been formed by the fusion of the plica of the tubal mucosa. On the outside of the spongiosa there is a thin layer of fusiform cells and fibers. Muscular fibers were not found in this layer. The two layers described last show a round cell infiltration though not as marked as the compacta. There are also found maternal blood sinuses in the wall just described, they are of very large caliber and their lumen shows granules, fibrin and also red blood-corpuscles fairly well preserved in size, shape and staining properties.

Transverse section from parts of the tube to the inside of the gestation sac show well preserved tubal plicæ. The lining epithelial cells are of the columnar type and their cilciæ can be well seen. The core of the plicæ shows enlarged vessels surrounded by connective tissue cells which are not of the slender type found in the non-pregnant tube; they are on the contrary quite large and oval with large vesicular nuclei and very much approach the type of decidual cells. Here the muscular fibers of the tube wall are very much hypertrophied and the bundles have been pushed apart by intervening connective tissue. Foci of small, round cell infiltration are seen everywhere in the muscularis, particularly around enlarged vessels. The peritoneal covering does not show any marked changes.

Case II.-Mrs. W. (Fig. 2.) It is hardly necessary to go into all of the details of the result of the microscopic examination of this second case, since it is in most respects only a repetition of the result furnished by Case No. 1. The following exceptions, however, must be noted. While the intervillous space like in Case No. 1 shows granular débris and the derivatives of decomposing hemoglobin, it also shows densely crowded

blood corpuscles of the maternal type, normal in size, shape and staining properties. This shows that blood from the maternal tissues must have entered into the intervillous space shortly before the removal of the pregnant tube, while in Case No. I the microscopic examination shows that this cannot have been the case, since all the blood in the intervillous space was badly decomposed.

In Case No. 2 there are also found a considerable number of badly degenerated villi, but many of them are in a very fair state of preser

vation with fairly normal mesodermal and ectodermal elements. Blood vessels with blood cannot be seen either in the chorion or in the villi; in some places, however, it appears as if collapsed empty blood vessels could be made out.

The gestation sac in Case No. 2 is much thicker and shows less marked degenerative changes, as the sac in Case No. 1. The vessels. are quite thick-walled, most of the thickening being due to a subendothelial proliferation. The round cell infiltration of inflammatory changes is likewise found freely. Where the plica of the tubal mucosa are still preserved, they have become club-shaped, their epithelial cells

are somewhat flattened and their core shows typical large decidual cells.

From the microscopic examination of both of these cases it is seen that hæmorrhages had occurred into the intervillous space and extensive blood coagula had been formed in them. These had damaged the villi, interfered with the nutrition of the embryo and brought about the death of the latter.

One of the embryos was found to be macerated, one was yet in a fair state of preservation. Considering the thinness of the gestation. sac in Case No. I and its advanced state of degeneration, a rupture would probably have taken place soon. In Case No. 2 this danger seems to have been less imminent.

We must from a pathologic point of view strictly differentiate between the events and phenomena which really interrupt tubal gestation and those occurrences like rupture and tubal abortion which so frequently lead to the most urgent clinical symptoms. This is, of course, by no means a matter of theoretical interest only, but one of the highest practical importance. If it should be possible to establish a set of symptoms as characteristic for "intervillous hæmorrhages," the operator could step in in good time to save his patient from the great dangers of subsequent rupture.

TWO CASES OF GENITAL

MALFORMATIONS:

(1)

RETROHYMENEAL ATRESIA; HÆMATOCOLPOS ET HEMATOMETRA. (2) VAGINA DUPLEX ET UTERUS

SEPTUS.*

BY JOHN M. FISHER, M.D., PHILADELPHIA,

Chief of the Department of Diseases of Women and Demonstrator of Gynecology in the Jefferson Medical College Hospital; Gynaecologist to the Philadelphia Hospital.

Case I.-Mrs. R., aged 17, married thirteen months. Never menstruated. At the age of 14 she began to suffer from cramp-like pains in the lower abdomen and pelvis at intervals varying between one, two, and three months, of sufficient severity to confine her to bed for two and three days at a time. Within the past year and a half these attacks have recurred with comparative regularity every four weeks, and of late have been attended with pelvic tenderness and with hypogastric distention that subsides to a degree subsequently but at no time disappears completely. She has a constant feeling of fulness in the rectum and defecation is difficult. A rounded mass protruding from the vulvar orifice was first noticed about three months ago. Intercourse has been impossible. Always enjoyed excellent health otherwise. Patient is of the plethoric type, short in stature, and proportionately well formed.

On examination the lower portion of the abdomen is distended by an elastic fluctuating enlargement the size of a clenched fist. The external genitals are pushed beyond their normal level, overstretched, and the labii are widely separated by a bulging vaginal tumor, presenting a bluish discolored central area corresponding with the dilated orifice of the intact hymen, which is spread over its surface and appears to be intimately connected with its structure. The tumor is tense, elastic and fluctuates. Bimanually the common identity of the abdominal and vaginal distentions is readily determined-pressure upon the one resulting in more pronounced tenseness and greater prominence of the other.

Diagnosis. Retrohymeneal atresia; hæmatocolpos et hæmatometra. Operation March 17th, assisted by Dr. Ignazio Cortese. Chloroform anæsthesia. The thick, dense membranous structure of the vaginal * Read before the Philadelphia Obstetrical Society, May 3, 1900.