Page:Journal of the Asiatic Society of Bengal Vol 7, Part 2.djvu/271

820 had entered on the I at December, there would have been the risk i only one month in their case, and the number of casualties upon the number admitted would have been one-twelfth only of the annual ratio. To provide accurately for this I furnished the writer, employed in secasting the registers, with a table giving a decimal value for every day of the year, and thence, according to the date of admission. I made him enter the risk, as of the fraction for the period of the year remaining to 81st December. Thus, in the re-cast of the registers, each admission will be seen indicated by a fraction to three places of decimals; and the number of risks is by addition of the whole brought to the true annual sum for computation of the ratio of mortality from the actual casualties.

Thirdly. When a life lapsed, its risk was lost for the remainder of the year. To provide for this, I made a reversed decimal table showing the fraction of the year to the date of the casualty, and by entering the lapsed life not as an entire year’s risk, but according to the fraction to the date of occurrence, effectually removed this source of error. But those who follow this plan must be careful, when a life lapses in the very year of admission, to take both fractions from the same table for computation of the value of the risk: otherwise a child admitted on the 2nd January and dying on the 30th December, would have the same fraction to represent both dates, and would stand as 0, though the risk of his life was an entire year, less only two days. The writer employed in re-casting the Orphan School registers made this mistake in the first instance, which is the reason of my noticing the point.

Fourthly. Having thus settled the mode of entering admissions and casualties, I caused books to be prepared for each year of life. In that for age 0, 1 caused to be entered successively, all who wean admitted at an age less than one year, taking their names in succession from the register of each year from 1798 to the present time. The number of names thus entered in this book for age 0, is 5930, but each being reduced to its fraction of the year of admission, and the death cases being doubly reduced, the number of annual risks, for this age is diminished to less than half, being 2646, which is what might have been expected. The names of the whole being thus looked out in the successive books, and entered in a fresh register for age 0, the page was ruled for forty years of life from 0, and each name wss marked as a year of life in the columns following 0,. as it was found in the successive registers, until the date of decease, or of removal from the institution.

Fifthy. The book of those who entered at an age less than one year being completed, and the individuals followed out, a similar bach