The second group of laboratory methods having a direct bearing on the practice of medicine consists of those which involve highly specialized technique and complicated apparatus. Electrocardiography, basal metabolism determinations, the Wassermann reaction, clinical bacteriology, and the various types of chemical analysis of the blood fall into this category. The information to be elicited from these and other analogous methods is often extremely valuable, but their application is necessary only in a comparatively limited number of cases. As a whole, these methods do not have the broad general significance and importance that characterize the simpler tests just referred to. It is, of course, highly desirable that they should be available to practicing physicians, so that they may be used in the cases in which they are particularly indicated, but fortunately there is no necessity for the great majority of physicians to bother themselves about the details of technique. This should be relegated to a small number of men who are devoting their attention to specialized fields. Simplified technical procedures, supposed to be adapted to the use of practicing physicians, are continually being advocated as substitutes for the recognized standard methods employed in performing some of these tests, but they are frequently unreliable, or reliable only in the hands of one who has a thorough knowledge of all the sources of error, so that it is far wiser to avoid them and to obtain the dependable observations of experts. The clinician may, therefore, neglect the technical side of these more elaborate tests with a clear conscience, but in so doing he should not feel that he may drop the matter entirely. If he is ever to make use of them--and this the welfare of his patients may demand--he must have an understanding of their significance and of the physiology underlying them. He must know when they are indicated and when they cannot be expected to give important evidence. A little insight into the fundamental principles of metabolism, for instance, and a recognition of the common relationship between increased heat production, pulse rate, and certain other symptoms are of the greatest help in deciding in what cases an observation of the basal metabolism may be of diagnostic significance, and in what cases it is entirely superfluous. It is much more important to know in what particular case a determination of the basal metabolism may be of value than it is to know the details of the performance of the test. Then again, the physician should be able to interpret the results of the test in the light of his individual patient. A basal metabolism which is reported as 15 per cent above normal may or may not be significant, and an electrocardiogram showing a prolonged conduction time may be due to one of several factors, but in either case the physician should not be forced to depend for the interpretation on the