In the past but few house-to-house canvasses have been made with relation to influenza. Auerbach, following the 1889 epidemic, collected statistics on 200 families distributed throughout the city of Cologne. Abbott, while not conducting a canvass, did obtain a certain amount of valuable information by letters addressed to physicians, institutions and corporations throughout the State of Massachusetts. There is fairly abundant literature on the disease as it occurred in institutions. Moody and Capps, in a study of the epidemic in Chicago in December, 1915 and January, 1916, made a survey of the personnel and inmates of four institutions in that city. Among other rather numerous statistical compilations from institutions we may mention that of Hamilton and Leonard which was devoted particularly to a study of immunity, and that of Stanley at San Quentin Prison, California. Garvie has reported his personal experience with influenza in an industrial area and discusses the disease as it has occurred in families in his private practice. Carnwath reports a “block census” undertaken by Dr. Niven in Manchester, England. This is of the same nature as our own work. Reeks has made a detailed house survey of 2,757 persons in New Britain, Connecticut. D. W. Baker has conducted somewhat similar surveys for the New York Department of Health, and Winslow and Rogers quote the excellent record of the Visiting Nurse Association of New Haven, in which they have information for all of the families cared for by the nurses. This, however, is a collected group and does not correspond with the so-called block census. CHART XIV. [Illustration: Chart showing the actual incidence of influenza in Boston by weeks and the actual incidence among the 10,000 individuals surveyed by weeks during the first three months of 1920. Full Line—incidence in the entire city based upon reports to the Health Commissioner. Dotted Line—incidence in the six districts surveyed. ] The most comprehensive and detailed work that has been done in this line is that reported by Frost and Sydenstricker and by Frost, the first being the result of a canvass of 46,535 persons in Maryland, and the second a similar report based on a canvass of 130,033 persons in several different cities of the United States. We shall have occasion to refer to these later. MORBIDITY. There has been great actual variation in the morbidity from influenza in the various epidemics and even in different localities during single epidemics. Previous to 1889 there were no reliable statistics for the disease incidence, and subsequent to that date the records, for the reasons previously mentioned, have still been not entirely adequate.