The above are the crude rates, the infantile death-rate being stated by the usual method per 1,000 births _during the same year_. Owing to the great decline of births during the war, this method overstates the infant mortality in recent years. In a table given in the Registrar-General’s annual report for 1917, this unusual source of error is corrected. When this is done, and the infantile deaths are stated “per 1,000 of population aged 0-1,” the rates for the years 1912-17 inclusive in successive years became respectively 104, 117, 113, 111, 98, and 94. In other words, there has been a steady and uninterrupted decline in the death-rate of infants during the war. This decline has followed similar declines in preceding years, and it is to be noted that much of this decline occurred during the period when the hygienic work effecting child-welfare was confined to general public health measures. Thus it anticipated the more direct and active measures adopted by voluntary societies and by local authorities for the prevention of infant mortality. Comparing the five year periods 1896-1900 and 1901-05, a decrease in the death-rate of 12 per cent. is seen; comparing 1901-05 with 1906-10, a decline of 15 per cent. occurred; comparing 1906-10 with the average experience of the three years 1911-13 mortality declined 5 per cent.; comparing these three years with the average experience of the five years 1914-18, during which war conditions prevailed more or less, a reduction 9 per cent. was experienced. The actual reduction during war time is greater than is indicated by these percentages, when allowance is made for the statistical error indicated above. The exceptional experience of the year 1911 illustrates one of the chief sources of error in forming conclusions on the experience of a single year. In this year the summer was excessively hot, and summer diarrhœa prevailed to an exceptional extent; and the illustration is important, as serving to remind us of the limitations of the value of statistical tests and of the fact that increase of good work tending to improve child life may be associated temporarily with increase of total infant mortality. _The Influence of School Medical Inspection_