On the other hand, when the sound or the uterus is septic, perforation of the uterus has been followed by a rapidly fatal peritonitis; indeed, some of these injuries may prove as lethal as a snake-bite. Occasionally very serious consequences follow simple perforations by dilators and curettes; this has induced some gynæcologists to urge that if, in the course of dilatation and curettage of the uterus, a rupture or perforation of the uterine wall occurs, it is better to perform a cœliotomy and assure oneself of the safety of the patient than to hope that no untoward result will ensue. This advice is too sweeping. When the perforating instrument is clean, and there is little or no bleeding, the case may be left to itself; if untoward signs arise, cœliotomy should be performed. Sometimes a pelvic abscess occurs as a sequence to the accident, and will require evacuation through the vaginal fornix, or, perhaps, by means of an incision in the flank. Verco found a piece of a curette, 2-3/4 inches long, in an abscess cavity behind the uterus. The patient had been curetted two weeks previously. A perforation, or a rent in the uterine wall, in the course of curetting, is a serious accident when the operator is unaware that such has happened, and proceeds to flush out the uterine cavity with poisonous antiseptic solutions, especially perchloride of mercury. Cases are known in which, under these conditions, the woman has died in the course of a few hours. Injuries, in the course of instrumentation of the uterus, are not always mere perforations; some are wide rents--and this is an especial danger in removing sessile submucous fibroids (vaginal myomectomy). _A serious complication of tears or rents of the uterine wall, whether the uterus is gravid or non-gravid, is extrusion or prolapse of the intestine._ It is also remarkable that in several reported cases the practitioner has mistaken the intestines for ‘secundines’, even in unimpregnated uteri, and has withdrawn them, and even cut lengths of intestine away, before recognizing his error. In one case of this kind, where a practitioner had withdrawn and removed several feet of intestine through a rent in the course of a curettage, I performed cœliotomy, closed the hole in the uterus, joined the cut ends of the bowel with sutures, resected the mesentery belonging to the removed bowel, and thus saved the patient’s life. In another case, where a practitioner had torn the uterus during curettage and intestine appeared in the vagina, there was such free bleeding that I found it prudent to perform subtotal hysterectomy. This patient also recovered. Successful operations of this kind have also been performed by Werelius and Nixon Jones.