The importance attached to poverty as a cause of illness and mortality is illustrated in reports on local investigations, displaying an inverse relationship in different communities between family income and the rate of infant mortality, the reader being left to infer, that increase of the lower incomes is the one method for obviating excessive loss of infantile life. In suggesting this crude generalization it is evident that poverty is being regarded as an element, instead of as a highly complex phenomenon, which needs to be further analysed into its constituent parts. The crude generalised statement as to the relation between excessive mortality and poverty, furthermore, fails to bring out three essential points, viz., that infant mortality may be very low in communities in which poverty is the rule; that it may be high in the absence of poverty; and that where infant mortality is high, it can be greatly reduced without change of economic conditions. There should be an adequate family income for every family; and the social conscience cannot be satisfied until this is realised. But, in seeking for practical reform we must appreciate that a large share of the disease and of the inefficiency of the individual and family associated with poverty can be remedied otherwise than by an increase of the family income. This is shown by national and international experience. The death-toll on infant life is very much lower in Norway and in Ireland—both relatively poor countries—than in England. Poverty in these instances evidently has less weight than the favorable factors of rural life and natural feeding. A like discrepancy in experience of infant mortality is seen between the experience of towns, and of wards in the same town, with approximate equality as regards poverty. Similarly in England the infants of miners with relatively high wages suffer a higher mortality (160 per 1,000 births in 1911) than the infants of textile operatives (148) with relatively low wages; while the latter suffer more than the infants of agricultural labourers (97). These instances at once suggest that some conditions in town life play an important part in causing excessive infant mortality; that in towns insanitary conditions and habits of life are even more injurious than the absence from home of the industrially employed mother; and that the causation of infant mortality is complex, and its prevention necessitates a multifarious attack on social and industrial evils, the character of this attack necessarily varying in different localities, in accordance with the incidence of these evils. That the influence of urban life in causing excessive mortality can be counteracted is shown by the varying mortality in different urban communities, and in different parts of the same town.