In a previous report1of a case of chronic pentosuria and migraine the following facts were noted: The pentose excretion in the specimens of urine (as indicated by daily twelve night hour collections) appeared to be fairly constant and was apparently independent of changes in diet. Amidopyrine, when taken by mouth, markedly increased the urinary pentose. In the course of a series of nonspecific protein (milk) injections, the urinary pentose was once temporarily diminished, and at the same time there was present a marked clinical improvement. Further injections of milk (intradermally) or peptone (intravenously) failed to reproduce the clinical or chemical results. I therefore undertook to extend these observations and at the same time try another method commonly used in "fever therapy"—the intravenous injection of a typhoid bacillus vaccine. Twelve hour specimens were collected from 7 p. m. to approximately 7 a. m. on the following morning practically daily