1'837.] Estimate of Life in the Civil Service. 343 



giving in each the fact of the individual having outlived that year or 

 not, or any other circumstance or event, must afford the means of 

 computing the different accidents of life for every age that may be 

 reached by the persons so registered, and the results of one page may 

 be combined with those of any other by adding the sum at the bottom 

 of the page to the proper column with reference to age of such other 

 page, and by taking out of the whole the number of deaths or of mar- 

 riages or of the births of children, male or female, or of any other ac- 

 cident of life that may be recorded in the column to compare with the 

 sum of the lives of the age in both pages or of as many pages as may 

 be brought into the computation. 



We presume that every insurance office keeps registers framed upon 

 this principle, but we wish to see them extended to the Army and 

 likewise to some thousands of natives in towns and in the interior, 

 with a view to obtaining the materials for computing the risks and 

 accidents of life amongst these classes at different ages, in respect to 

 which we are at present without any materials for framing a table or 

 estimate of any kind. 



The tables given in Captain Henderson's article upon the subject 

 of the value of life in India, published in the last volume of the 

 Researches of the Asiatic Society, though framed with great labour, 

 are defective in this point*. They afford general averages of the value 

 of life amongst certain classes, but not of the value of life at each year 

 of age, which is a most essential circumstance ; and for insurance offices 

 or for institutions which deal in annuities, the risks with reference to 

 age are the main and most important, if not the only, matter for 

 consideration. 



It is to be observed that it will not be possible to frame registers 

 retrospectively for any class of persons, unless from peculiar circum- 

 stances a given number of names with the age of each individual can 

 be entered for any specific past date, and these can be followed out 

 in all their circumstances to the date of the formation of the regis- 

 ters. This is the principle upon which the previous and present 

 tables have been framed for the Bengal Civil Service, and upon which 

 similar tables have been made for the Army. The nominations of 

 each year to the different services being fixed and known, and the 



* Capt. DeHaviland's tables for the Madras army are an exception to this 

 remark, as they are framed by years of service on our principle, but the results of 

 the first years of the series give ratios of deaths for those years which cast a 

 doubt on the accuracy of the whole table. Mr. Gordon' 9 army table ia of too 

 old a date to be useful. 

 2 v 2 



