It is sad but true that few of the published "discoveries" in therapeutics survive the test of time: that but few articles purporting to be contributions to our knowledge of therapeutics are such in fact. This is largely due to the neglect of proper methods, which leads to erroneous results. Martini, in this brochure, aims to lay down principles for guidance in therapeutic research which, if generally adopted, would lead to improvement in therapeutic literature and acceleration of therapeutic progress, even if also to a marked diminution in the number of articles on therapeutics. Comparison is the foundation of all therapeutic research. This comparison must either be based on statistics, which are especially applicable to the acute infectious diseases, or, in chronic diseases, on observations of the course, which requires in each individual case a preliminary control period and, if possible, a control period following the test. Qualitative estimation of