appendical abscess. The uterus should be gently explored with finger or dull curette and any placental débris or clots removed. If possible, a bacteriological culture should be taken directly from the uterus, and then an intrauterine douche, preferably of sterile water or normal salt solution, should be given. Abrasions of the cervix and vagina should be treated with a solution of chloride of zinc or compound tincture of iodine. Should the bacteriological examination reveal saprophytic infection, the douche may be repeated if the temperature remains elevated and the foul-smelling discharge continues. If it proves to be a case of streptococcus infection, further local treatment is contraindicated, as it cannot possibly reach the organisms, which have penetrated beyond the reach of such measures. If the infection has spread into the broad ligaments, as may be manifested by the extension of the pain over the region and the revelation of a boggy sensation on vaginal examination, the cellulitis and localized peritonitis must be combated by the application of the cold pack, or preferably the rubber coil, to the lower abdominal region, and the giving of anodynes to relieve pain. General peritonitis will call for the same line of treatment in a more vigorous manner.

I am confident that the repeated giving of intrauterine douches and the use of the curette in streptococcus infection is decidedly harmful. The danger of introducing other germs and causing a mixed infection and the production of more or less traumatism of the endometrium, thereby breaking down Nature's protective zone of leucocytes and opening up new avenues for widespread systemic infection, are in my judgment strong arguments against injudicious local treatment.

For the surgical treatment it is impossible to prescribe any fixed rule for guidance; the judgment and skill of the physician and the condition of the patient must determine. It is not justifiable to open the abdomen without some physical reason. The different forms of sepsis must be thoroughly understood, as the operation will hardly be required except in pathogenic infection. When the operation is done early many organs are likely to be needlessly sacrificed, and if done late it will surely increase the mortality. When there is continued fever, together with physical signs, operation is permissible; without the latter the general symptoms would indicate systemic infection. Of course, where there is pus, be it in the pelvis, pleural or peritoneal cavity, or elsewhere, it should be evacuated at once.

The general treatment will embrace stimulants, strychnia and nitroglycerin to strengthen the heart, cold pack or sponging to reduce the temperature. Antipyretics should not be used, as they depress the heart; and quinine is of little avail, except in malaria-besides, given in large doses, it is apt to derange the stomach. Morphia may be required to allay pain and procure sleep. Nuclein and albumose have been used for the purpose of increasing leucocytosis, and, in conjunction with normal salt solution by the bowel or subcutaneously, have proved excellent adjuvants to other measures.

During the past few years experiments have been made to destroy or neutralize the streptococcus infection by means of the antistreptococcus serum. So far the statistics have not been very favorable. Fry has collected 119 cases treated with serum, 77 cured, 42 died, mortality 35 per cent. From a careful analysis of the literature many of the failures may be accounted for in one or more of the following ways: first, lack of bacteriological confirmation of the character of the infection; second, inferior quality of the serum employed; third, injudicious local treatment; fourth, badly selected cases; fifth, delay in the use of the serum. It is important to determine whether the case be due to simple streptococcus, mixed infection, or some other pathogenic organism. Curative effect is only claimed for simple streptococcus infection, although some cases of mixed infection are reported where, after the use of local treatment in conjunction with the serum, recovery has followed. Marmorek records 15 cases in which the serum was employed, of which were due to pure streptococcus, and all recovered; 5 were due to a mixed infection of the streptococcus and staphylococcus, 2 of which died; 3 were due to a mixed infection with the streptococcus and bacillus coli, and all died. The prognosis is, therefore, distinctly more favorable when the case is one of simple streptococcus infection. In many cases the serum was not used until all other measures had been exhausted and the patients were in extremis.

With the view of observing the effects of the serum, I determined at the first opportunity to treat cases by withholding local and general remedies as far as consistent with the welfare of the patient.

CASE I.-Mrs. A., age 33, admitted to Columbia Hospital November 7, 1898, with history of having had a gush of water about a week previous, and thinks she has passed the normal period of gestation. Examination proved the case to

be one of hydrorrhea gravidarum. She left the hospital after a few days and returned in labor November 17, 11 P.M. The

membranes had ruptured five hours previously. Normal birth at 4.25 A.M., November 18, and placenta followed spontane

ously twenty minutes later. No vaginal examination made during or subsequent to labor, but the assistant supported the perineum during the delivery of the head. No douche given. Temperature 100° for twenty-four hours after delivery. November 19, 9 A. M., temperature 98°, pulse 72. At 10 o'clock patient had a severe chill, followed by a rise of temperature to 104.2°, pulse 128. Had headache and vomiting, and complained of severe pains in the lower part of abdomen; lochia normal. At 7 P.M. temperature 103.5°, and, as the uterus was high up and large, an intrauterine douche of sterile water was given with the object of removing any possible retained clots or membrane. The douche returned clear. A culture of the lochia was taken by the Döderlein tube, and the bacteriological examination showed it to be a pure streptococcus infection. November 20, 9 A.M., temperature 102.2°, pulse 104, and complained of chilly sensation. Ten cubic centimetres of Marmorek's antistreptococcus serum were injected at 2 P.M. At 7 P.M. temperature 104°, pulse 122; patient nauseated and has vomited several times. An additional 10 cubic centimetres of serum was injected two hours subsequently; temperature falling, and at midnight was 101°, pulse 88. Nausea and vomiting has ceased, and patient is in a profuse perspiration and feeling quite comfortable. Midday, November 21, temperature 98.2°, pulse 80; resting quietly and perspiring. At 3 P.M. temperature 101.4°, pulse 86; 10 cubic centimetres of the serum given, and, as the temperature was still rising at 10 P.M., 10 cubic centimetres more were administered. At midnight temperature 102.4°, pulse 88; complains of headache and is restless. November 22, 9 A.M., temperature 101.8°, pulse 90; perspired very freely after midnight; slept very little, and not so bright as on yesterday; 10 cubic centimetres of serum given, and, as the desired effect had not been obtained at 6 P.M., an additional 10 cubic centimetres was administered. This was followed by a gradual decline to normal at midnight, attended by refreshing sleep. November 23, 9 A.M., temperature 101.6°, pulse 96; injected 20 cubic centimetres. At 1 P.M. temperature 103°, pulse 102; 10 cubic centimetres more were administered. At 6 P.M. temperature

unchanged, pulse 114; gave 10 cubic centimetres. This was followed by a reduction of pulse to 98, but no dropping of temperature. At 9 P.M. another 10 cubic centimetres was given, and this was followed by a gradual decline of temperature, attended with refreshing sleep throughout the night. November 24, noon, temperature 98.2°, pulse 92; 10 cubic centimetres

given to anticipate an evening rise, and repeated at 9 P.M.; 10 cubic centimetres were given at 6 P.M. November 25, and a like amount November 26. Altogether 140 cubic centimetres were used. I am satisfied that the last three or four injections were unnecessary, as a slight continued rise of temperature after the 24th was due to an abscess forming on the right side of the abdomen at the site of one of the injections. This was

freely opened and bacteriological examination of its pus showed presence of streptococci.

CASE II.-J. S., multipara; admitted to Columbia Hospital in labor December 19, 9 A.M., normal birth at 6 P.M., and placenta delivered spontaneously one-half hour later. No vaginal examination made either during or after labor and no douche was given. On the afternoon of the third day patient complained of chilly sensation and headache, followed by a rise