<h2>Summary</h2> From the admission to the hospital on February 19 to June 11, therenever was a time when there was any doubt that the patient's recovery was incomplete. After the latter date he appeared to get well rapidly and to maintain his well-being. When sulfanilamide was discontinued on March 18, it was very impressivethat rapidly, in approximately forty hours, his symptoms became progressively worse, so that on the fifth day he suffered a severe exacerbation of symptoms. His symptoms reappeared at about the time necessary to eliminate the sulfanilamide. The third relapse responded to treatment unsatisfactorily and wasfollowed by a more pronounced relapse in a short time. More energetic treatment and a third transfusion instituted the beginning of recovery. It is most impressive that in his March hospitalization the severity of the illness indicated an organism at least as virulent in the primary sickness; yet the organism was recovered with difficulty. Never at any time did any of the mice give clinical signs of being ill after inoculation, and the organism did not grow readily in culture media. Sulfanilamide did not destroy this organism; and, while its virulencyto mice and its cultural growth seemed very much affected, at the same time its power to overwhelm the patient seemed unimpaired after a month of controlling administration of the dye. Certainly the development of immunity seemed to be nil, or exceedingly slow. Nausea at times, particularly after the relapse of March 23, necessitatedtemporary reductions of the sulfanilamide dosage. Bicarbonate of soda, 10 grains, three times daily, or 5 grains with larger doses, seemed to reduce upsets. Anemia was mild throughout—just sufficient to justify small stimulatingtransfusions. The response to transfusion of June 11 was most pleasing. During the entire course the spinal fluid pressure was not excessive.Only once with an air manometer did it reach 340 mm. The use of the manometer was discontinued after several repetitions because of evidence of low pressure and, particularly, because the manipulations tended to increase the contamination of spinal fluid specimen with blood cells. After our experience in the March flare-up we planned to continuetherapy until the evidence of recovery should include not only normal general symptoms, temperature, white blood counts and differentials, but if possible also a spinal fluid within a normal range. The spinal fluid cell count was normal on May 25 and after June 11. We did not consider another spinal puncture was justified. The healing and filling in of the mastoid wound approached as nearly as was possible a primary healing. With the immediate operative field apparently out of the picture, the pneumonic exudate on the meninges must have been the focus of infection from which reinfections came. Hewell and Mitchell<sup>1</sup> report recovery in two cases of pneumococcicmeningitis, Type. III, with sulfanilamide compounds. In both cases ethylhydrocupreine hydrochloride was given. Their review of the literature adds three more cases of recovery. One of their patients had exacerbations, and the recovery followed trepanation and aspiration of pus from the right temporal region. This case is unique in that it is the only recorded instance of recovery we could find in our review of the literature in which recovery followed repeated relapses. The duration both of the disease and of the administration of sulfanilamide is unusual. The boy's weight on February 5, before onset, was 69 pounds 4 ounces.His weight on July 1 suggested a very appreciable gain; and on December 2 was 89 pounds. Seven months after recovery he was fully within normal both physically and mentally. Throughout the entire illness Dr. Henry F. Martin was in charge of the ear therapy, including all operative procedures.