The important part which the laboratory has come to play in medical science is generally accepted and appreciated, but the relation which it should bear to clinical practice remains to be satisfactorily defined. It is obvious to all clinicians of experience that the laboratory never can become, and never should become, the predominating factor in the practice of medicine, but it is equally evident that sound medicine cannot be carried on without the support of the laboratory, and that in the future the dependence of the clinic on the laboratory will probably increase rather than decrease. Among the men engaged in active medical practice, however, only a small minority can ever hope to undertake extensive laboratory work in connection with their patients, and the great majority of physicians are and will continue to be confronted by the difficult problem of their relation to this growing influence in medicine. To the teacher of medicine, whose foremost duty is to prepare his students for the practice of the future, the same problem presents itself, for the students must be thoroughly trained in the laboratory methods that will be of practical service, but not burdened with those that are highly specialized or of questionable value. The leading exponents of clinical laboratory work are the large hospitals--especially the hospitals associated with teaching institutions--and these exert a profound effect on private medical practice, but the conditions existing in them are such as to demand a separate consideration. In such hospitals, laboratory investigations fall into one of three categories. The first includes those which belong to the field of pure research, their object being to advance the limits of our knowledge of disease. With this we have, at present, no concern. The second consists of those laboratory methods that are applied in order to obtain direct aid in the diagnosis or treatment of individual cases of disease. This often means the use of standard methods of proved and known value--methods which have received general professional acceptance--but in addition it means the use of many methods of possible value, the significance of which needs to be thoroughly tested under conditions favorable for critical control. The trying out of newly advocated measures for the diagnosis and treatment of disease must always be an important function of the larger and better equipped hospitals. Many--indeed the majority--of such methods are found to be unreliable or of little practical value, and after their status becomes established they are discarded. Very rarely a new method withstands the test of prolonged observation and proves to be of such practical significance that it can be properly advocated for general adoption. This type of hospital thus serves as the court before which all such new ideas must stand trial and it is astonishing, if not depressing, to compare the enormous amount of time and labor that