In the paper that I had the pleasure of reading before this Society in the course of last Session, I brought forward a problem to which I promised to recur on an early occasion. That problem was a consideration of the present method of dealing with observations of various magnitudes when striking the rates of mortality experienced in each year of life. Our early mortality tables were derived from observations of a very general nature. The gross number of lives passing through each year of age and the number of deaths arising out of that number furnished the whole of the materials for producing the desired rate of mortality. Indeed in some instances these numbers at each age may have been ascertained by some mathematical interpolation from larger groups at younger and older ages. There was no discrimination as to the relative weights that may be allotted to any sections of these numbers. Thus in the tables derived from the numbers of the population there was no distinction between the numbers dying out of the sick and the numbers dying of the healthy. In more recent researches, however, we have improved upon our former methods of tabulation. The exact problem to be solved was not that of finding the general average rate of mortality for each year of life, but rather to ascertain the flow of mortality emanating from each group of healthy lives enrolled at each age and carefully tracing their mortalities year by year during all the subsequent years of their lives.