1838.] Table of Mortality. 819 



Service of Bengal, and needed only to be re-cast and added up to yield 

 equally valuable results for the ages of life they comprehended. The 

 re-casting of thirty-eight years' registers containing many thousand 

 names, has however proved a work of labour that has occupied several 

 months. The Statistical Committee has furnished a writer, who has 

 been employed on the work for this period without intermission, and 

 the product of his labour in the volumes which show the name of 

 every child, the date of his admission, and the manner of his having 

 been disposed of, are deposited in the library of the Asiatic Society, 

 as well for the verification of the table now submitted to the public, 

 as that the detailed registers may be available for the ascertainment of 

 other results which also may be gathered from them*. 



My present purpose, as above stated, is confined to the exhibition in a 

 tabular form of the ratio of mortality for each year of existence as 

 deduced from these registers. 



It will be satisfactory to explain in the first instance the process 

 followed in the construction of the table ; for there are several circum- 

 stances requiring to be noted, as guides to those who may apply the 

 same principle of computation to other classes of persons, or may 

 undertake the recasting of other similar registers. 



Firstly. The Orphan School books did not show in every in- 

 stance the actual date of birth, nor, if they had done so, would it 

 have been advisable to attempt to follow each child from birth-day to 

 birth-day, and so frame a general register, true to the exact age of 

 each individual. For example, a child admitted is simply entered as 

 aged not one complete year ; in the re-cast of the registers this child 

 stands as entered of the age 0, and he is considered as remaining of 

 that age until the 1st January next following, though his birth-day, 

 that is tbe date on which he completed one year, may happen to have 

 been in November, or in February, or in any other of the twelve months 

 following the date of his admission. All subsequent years of life are 

 in like manner computed by the calendar year, from 1st January to 

 31st December, without reference to birth-days, which, as the error 

 will be equal both ways, and so balance itself, affords a complete result 

 for our present purpose. 



Secondly. It is the object in the construction of this table, to 

 deduce correctly in the first instance the annual percentage mortality. 

 The admissions in the course of a year do not give the risk of those 

 lives for the whole year. If for instance all admitted at year of age 



* Note.— Amongst other purposes to which these registers may be applied is the 

 ascertainment of the relative mortality in different periods of years, and in different 

 months and seasons. 



