_A given amount of domestic overcrowding with a large amount of institutional segregation of consumptives is associated with less tuberculosis than when overcrowding is less but accompanied by only a small amount of institutional segregation of consumptives._ The data as to institutional segregation are difficult to obtain; but there is sufficient evidence to show that in countries which have experienced a large reduction in the death-rate from tuberculosis a large proportion of hospital treatment for many years has been provided for consumptives, while in countries which have not experienced this decline such provision has been absent or imperfect. In London about 56 per cent., in county boroughs 35 per cent., in other urban districts 21 per cent., and even in rural districts of England near 16 per cent. of all deaths from pulmonary tuberculosis occur in hospitals (poor-law institutions, general and special hospitals, and asylums). Prior to the patient’s death he has had on an average at least three months, and probably in the aggregate more nearly five or six months, residential treatment, and this at the stages of disease in which there is the greatest discharge of infective material, in which owing to feebleness the patient is least able to control its hygienic disposal, and in which—had the patient been treated at home—the relatives would be especially liable to receive massive infection, and would be enfeebled by overwork and anxiety, or by the malnutrition associated with poverty.