The reports of autopsies done upon these cases render it probable that the blood forming the hematocele may in some instances come from the vessels which are in the broad ligament. Richet and Devalz were of that opinion, and considered the hemorrhage to be due to a varicocele of the ovary; and Scanzoni defends varicosities of the sub-peritoneal veins, as the source of the hemorrhage in rare cases. But only in very few in stances of this kind have conditions of the uterus and labia been seen which would lead us to conclude that varicose veins were present. Un doubtedly a venous trunk of the broad ligament may occasionally rup ture under the pressure consequent upon menstrual congestion, or in con sequence of pathological changes in its walls; and so a hemorrhagic focus be formed, which sooner or later breaks through the walls of the broad ligament, and pouring into the retro-uterine excavation, forms a hems tocele. Post mortems show that the fatal hemorrhage occurs in this way in tubal pregnancies. Olivier d'Angers describes a case of tubal preg nancy at about the third month, in which rupture of a vein in the broad ligament ensued. In the midst of the right broad ligament was a plexus of vessels filled with black coagulated blood, the individual veins of which were as thick as a crow-quill. One of these veins showed a circumscribed rupture, from which the fatal hemorrhage had occurred. At the same time he mentions two other cases in which women died in the course of a few hours from an intra-peritoneal hemorrhage, without any ine ovum being found. One belonged to Leclerc, and the woman died in three houri, from the hemorrhage coming from an ulcerated varicose vessel in the thickness of the left broad ligament. The other woman was twenty-five years old, and had two children; she collapsed suddenly, and died in half an hour. At the autopsy it was found that the right venous plexus ovaria was varicose, and had ruptured. Schroder also had a case of hematocele (Observations VIII.), in which the right broad ligament contained a pigeon's-egg-sized coagulum, and near which was a, rupture in a vessel three lines in length, from which the blood had prob ably poured into the recto-uterine excavation.
Undoubtedly the women do not always die of such a hemorrhage; and these cases justify us in regarding the vessels of the broad ligament as a, possible source of the bleeding.
The French authorities (Huguier, Dolbeau, Tardien), claim that the peritoneal coat of the uterus is the source of the blood, which passes by transudation. But Virchow undertook to show from records that hematocele might be due to a peritonitis hemorrhagica. He says: " Besides, it frequently occurs that inflammatory processes arise in the pelvic cavity, and cause a pathological vascularization, and local hyperse mias and hemorrhages, which, repeated from time to time, may give rise to considerable accumulations. In these latter cases it may happen that peritonitis retro-uterina, like a pachymeningitis, may give rise to false membranes; and the extravasation which ensues from the vessels of these pseudo-membranes is confined between its layers, and so gives rise to a closed (encysted) hematoma retro-uterinum." Clinical experience *ashes us that this is really the course of things in many pelvic tumors situated behind the uterus. It agrees also with Ferber's views, who claims that we must regard a large number of hematoceles as pelveo-peritonitic processes (pelvi-peritonitis) due to blood extravasations. Thus in most
cases of hematocele a puncture will give us tarry blood; but some cases, apparently entirely similar, give us a more or less bloody serous fluid. These latter are only to be explained upon Virchow's theory.
It was formerly believed that there could occur a reflux of blood from the uterus along the tubes into the peritoneal cavity, and so hematocele be formed. (Mile.) This view was supported by the fact that in women who died of the acute exanthemata while menstruating, the uterus was found filled with a coagulum, the processes of which stretched into the tubes; and also by the fact that in certain cases of hematometra and hema tokolpos, hematosalpinx was also present. Trousseau maintained that in marked uterine flexion the blood might regurgitate and form hemato cele. But we hold that but few hematoceles arise in this way, and that in most cases the blood comes not from the uterus but from the interior of the tubes.
Spencer Wells has called attention to a peculiar source of hemorrhage in these cases. It is not uncommon after ovariotomies to have the tubes enclosed in the pedicle which is attached to the abdominal wall, and for months after the operation a little blood may be observed to well out from the tubal mouths at each menstruation. If now the ligated or cauterized pedicle has been left in the abdomen, it would cause trouble every month, and symptoms of hematocele might occur. Kroner has published a simi lar case from Spiegelberg's clinic.
Cause of the have already seen that the causes of a pelvic hemorrhage, have usually been present for a long time before it actually occurs. One or other ovary, tube or broad ligament, or the perito neum or the uterus, has undergone a pathological change, which predis poses its vessels to rupture, and so gives rise to hemorrhage. The age and occupations of the women most subject to hematocele favor this view. Hematocele retro-uterina most frequently occuri, according to Voisin's collected cases, between the twenty-fifth and thirty-sixth years Among thirty-four cases in which the age is given, one only was under twenty-one, and one only over forty years of age. Twenty were between twenty five and thirty-six years old, four between twenty-one and twenty-five, and 'four between thirty-five and forty years. Among the forty-five patients whom Schroder refers to in his above-mentioned article, the age is mentioned of forty-three. It was: These statistics agree with our own observations, and prove that women are most liable to retro-uterine hematocele during their most fruitful period, from the twenty-fifth to the thirty-fifth year. The cause of the affection is usually to be found in some abnormal condition of the organs of generation. Most of these women had had one or more chil dren; of Voisin's, thirty cases, and of Schroder's, forty-one had had them. This is of importance, since it renders it possible that the adnexa of the uterus were diseased, and were not only the cause of the sterility, which had in most cases been present for several years, but also the re mote cause of the hematocele. And the history of these cases also tends to prove that the women were not in perfect health when they suffered from the pelvic hemorrhage; many cases suffered from irregular men struation and pain for a long time before the hemorrhage set in.