Plague appeared in the north of China in 1917, originating apparently in inner Mongolia. The spread extended over quite an area, and it is reported that this epidemic of pneumonic plague has been more extensive than any since that of 1910–11. The disease was first reported prevalent in Patsebolong December 6, 1917. The diagnosis was confirmed bacteriologically, and there can be little doubt but that the cases of plague reported in various parts of China even up to March 18th were true plague, and not unrecognized influenza. SECTION III. In the following section of our report we shall have frequent occasion to refer to a series of investigations conducted by the author in the City of Boston during the 1920 influenza epidemic. We will explain in some detail at this point the nature of the work done and the methods used, in order that the subsequent references will be readily intelligible. AN INVESTIGATION OF INFLUENZA IN BOSTON DURING THE WINTER OF 1920. Following every widespread epidemic interest centers in the question as to how much havoc the disease has wrought, what proportion of the population fell victim, and how many of these died. With regard to influenza the vital statistics of all countries are decidedly insufficient in furnishing this information. In nearly all countries influenza is not a reportable disease. Usually, as was the case in the United States in 1918, the disease was made reportable during the epidemic, but this took effect usually at least two weeks after the epidemic had started in a community. Further, there is probably not a single community in which the reported cases of influenza reach to anywhere near the total of actual cases. The question of diagnosis, which is not always easy even in the presence of a pandemic, causes some physicians to hesitate to report cases. Other physicians “play safe” and report nearly everything as influenza. Finally, in the period of an epidemic, the physicians are so pressed with caring for the sick that they very naturally neglect to report cases as they occur. It becomes necessary, therefore, in collecting evidence in civil populations, of the morbidity and fatality from influenza, to obtain additional information to that available to the Health Officer. The method which may be relied upon to give the most accurate data consists in house-to-house surveys made soon after an epidemic, in which competent inspectors obtain detailed information concerning the illness or freedom from illness of every individual in the areas canvassed. The majority of individuals interviewed will not have had the disease, and it is therefore essential that in such a census a large enough population be covered that the resulting figures will be truly representative of the population at large.