It will be seen by this table that Johnstown is among the 10 cities of more than 50,000 population which had an infant mortality rate of 1910 in excess of 150 per 1,000 births. These 10 cities and their respective rates are as follows: Lowell, Mass., 231; Holyoke, Mass., 213; Manchester, N. H., 193; Fall River, Mass., 186; Detroit, Mich., 179; New Bedford, Mass., 177; Lawrence, Mass., 167; Johnstown, Pa., 165; Washington, D. C., 152; and Pittsburgh, Pa., 150. It should be borne in mind that the absolute infant mortality rate of 134, computed for the group of babies included in this investigation, that is, for those born in Johnstown in 1911, can not be compared with any of the approximate rates in the foregoing table, since the basis of computation is entirely different. But the method used in this report seemed to be the only practicable one for our purpose, namely, to measure the infant mortality rate in different districts of the city where the babies are subjected to varying conditions. Conditions similar to those existing in Johnstown were found in Chicago by Dr. Alice Hamilton, Bacteriologist in the Memorial Institute for Infectious Diseases, Hull House. The results of a study made of 1,600 families in the neighborhood was published in 1910. The investigation was undertaken to find out the truth or falsity of a general feeling among the district nurses that a high birth rate was accompanied by a high death rate. It was found that a high birth rate was not so much accompanied as outrun by a high death rate. The number of children live-born was compared with the number of children who reached the age of three, so it is a study of child mortality, rather than of infant mortality. The child mortality rate rises and falls very much as does the infant mortality rate in Johnstown. A table calculated from the data of all the families shows an ascending mortality rate: No. in Family Child Mortality Rate 4 children and less 118 6 children and more 267 7 children and more 280 8 children and more 291 9 children and more 303 Expressed in words this table says that child mortality increases as the number of children per family increases, until we have a death rate in families of eight and more, which is two and a half times as great as that in families of four children and under. FOURTH ANNUAL REPORT OF THE CHIEF, CHILDREN’S BUREAU, U. S. DEPARTMENT OF LABOR, Washington, October 7, 1916 INFANT MORTALITY—MANCHESTER The findings of the bureau’s earlier study in Johnstown, Pa., are confirmed in many respects by the findings in Manchester—the coincidence of a high infant mortality rate with low earnings, poor housing, mother’s work, and large families.