A hypothesis explanatory of a given phenomenon should be consistent with all the associated facts. We have seen that the hypothesis that segregation of consumptives is an important factor in the reduction of the death-rate from pulmonary tuberculosis agrees (1) with our knowledge of the tubercle bacillus, and (2) with veterinary and agricultural experience; also (3) that,—although exact data are unobtainable,—the degree of segregation when ascertainable is consistent with the degree of decline in the death-rate; (4) it is important to note also that this hypothesis is consistent with the otherwise anomalous facts that although the proportion of the population subjected to urban conditions of life has steadily increased, and the number of persons per inhabited room remains much greater in towns than in country districts, the death-rate from pulmonary tuberculosis in England has declined as much in them as in country districts; and that notwithstanding the greater overcrowding in towns, the urban is rapidly falling to the level of the rural death-rate from this disease. The town dweller’s better and more frequent treatment in hospitals is an important factor in overcoming the handicap of urban conditions of life, including overcrowding and preponderance of indoor and dusty occupations. It is desirable to supplement the above statement by some remarks on _Improved Housing as a Means of Reducing Tuberculosis_ Not infrequently the thoughtless remark is made that given improved housing sanatoria and hospitals for consumptives would be unnecessary. The frequent occurrence of tuberculosis in well-to-do families shows the absurdity of this statement. It is true that tuberculosis is more prevalent among the poor living in small tenements that among the well-to-do; but there is no consistent proportion between the degree of overcrowding in different districts or towns and the death-rate from tuberculosis. Improved housing and institutional treatment for tuberculosis cannot properly be regarded as alternatives. They are necessary complements to each other, and there must be increased expenditure in both directions, if tuberculosis is to be more rapidly reduced in amount. There are in fact two housing problems—for the healthy, and for the sick. The most rapid method of improving housing for the healthy is to remove the sick, and especially the tuberculous sick to a hospital. This is being done year by year to an increasing extent. In England and Wales in 1870, 8.3 per cent., and in 1912 21.6 per cent. of all deaths from all causes occurred in public institutions. It is difficult to exaggerate the practical relief implied in these figures in respect of satisfactory housing, especially in its functional aspect. Apart altogether from the tuberculosis problem much of the decline in the general death-rate must be attributed to the skilled treatment which a large proportion of the total population have received in our hospitals of various types.