I have seen no reason for revoking the conclusion expressed in 1908 in a lecture to the Washington International Congress on Tuberculosis that historically the hospital treatment and coincident segregation of patients suffering from pulmonary tuberculosis has been an important and probably a dominant factor in producing the national decline in the death-rate from tuberculosis in the countries in which a decline has been experienced. This explanation fits in with our knowledge of the disease, and with the analogous history of leprosy; and it is supported by the fact that by complete segregation of infected from non-infected cattle tuberculosis can be eliminated at will from a herd of cattle. It is remarkable, as I have elsewhere set out in much detail, that improved general health, increased well-being, and sanitary education have operated in Great Britain, Germany, Belgium, Denmark, and Massachusetts side by side with great decrease in the death-rate from pulmonary tuberculosis; while up to very recent years the same influences in France, Norway, and Ireland have produced little or no decrease in the national death-rate from tuberculosis. And similarly no constant relation can be shown between the degree of sanitary and social well-being in different countries and cities, and the amount of mortality imposed by tuberculosis. How is it that in some countries a high degree of domestic overcrowding is associated with a low and declining phthisis death-rate and conversely that a persistently high phthisis death-rate may occur with a less but still diminishing degree of overcrowding? The explanation is contained, I maintain, in the following statement: