74 Mr W. Frascr on the Histori/ and Comtitution of 



their members, for the first 10 years at least after entry, fall greatly short of 

 that represented by the tables of mortality ; and, of course, this difference 

 will be the greater, the more incorrectly any table in use may represent the 

 average rate of mortality among mankind to be. Thus, in the Equitable As- 

 surance Company of London, whose premiums are regulated by the Nor- 

 thampton Table, it was ascertained by Mr Morgan, the eminent actuary 

 of that association, upon a calculation for 30 years, ending with 1810, and 

 upon 83,201 members, that the mortality which had occurred was only to the 

 claims which might have been expected, from the age of 20 to 30 as 1 to 2; 

 30 to 40 as 3 to 6 ; 40 to 50 as 3 to 5 ; 50 to 60 as 6 to 7 ; 60 to 80 as 4 to 5, 

 and in all ages together, in the ratio of 2 to 3. 



Any rates of contributions, therefore, calculated from the Northampton 

 Table for sums payable at death, must be necessarily more than adequate to 

 defray these benefits, and hence the principal source of the large surpluses or 

 profits that are always realised upon those assurances. Indeed, according to 

 the above experience, the premiums charged by the Equitable, and, till lately, 

 by all the other offices, for sums payable at death, are 30 per cent, higher, and 

 in many cases more, than are necessary to provide for the sums assured. 



In the case of annuities, however, calculated from the Northampton Table, 

 the effect is generally the reverse ; for, as none will purchase these benefits but 

 those who are in the best state of health, and of strong constitutions, it has usu- 

 ally been found that they live, on an average, considerably beyond the time ex- 

 pected. Comparatively few offices, therefore, have schemes for annuities, 

 and such as do insure these benefits, generally purchase them from govern- 

 ment, which is now ascertained to be losing considerably every year by such 

 transactions. 



While, therefore, the premiums for annuities calculated by the Northamp- 

 ton Table have been found greatly insufficient, the premiums for sums pay- 

 able at death have been as much in excess ; and as the great mass of assurances 

 consists of this latter benefit, such excess had become an oppressive and un- 

 just tax on the higher classes of the public. 



But inaccurate mortality tables must, if possible, prove still more hurtful to 

 Friendly Societies, than even to the higher classes of mutual assurance associa- 

 tions ; for, on the one hand, the members of those societies are ill able to pay 

 more than is necessarily required, while, on the other, an inadequate contribu- 

 tion would prove ruinous in the extreme. Hence an accurate table of morta- 

 lity is of the utmost importance in the computation of their rates of contribu- 

 tions and allowances ; and it therefore became a serious question, first with 

 the Committee of the Highland Society of Scotland, and next with those of 

 the House of Commons, what table of mortality ought to be adopted. 



The Highland Society confined their inquiry among Friendly Societies to 

 age and ^ckness only ; but this they afterwards regretted, as it was found that 

 the mortality of their members might have been also pretty accurately ob- 

 tained. As already mentioned, the committee had therefore to consider what 

 tables of mortality could best be relied on, and applied for the opinion of seve- 

 ral eminent calculators on the subject. Among many communications which 

 were in consequence received, the following are extracts from one made by 

 Dr Hamilton of Aberdeen. 



" The choice of proper tables for ascertaining the rate of mortality, is a point 

 of the greatest importance. I have not seen Mr Milne's Treatise on Annui- 

 ties, containing the Carlisle Tables, but observe that they give the probabili- 

 ties of life higher than any tables I am acquainted with. The Swedish Tables, 

 given by Dr Price, are also high, and give an intermediate result between the 

 Northampton and Carlisle Tables. I think the calculations should not pro- 

 ceed upon one set of tables only, but upon the medium of several esteemed 

 the best. Perhaps the three above mentioned (Northampton, Swedish, and 

 Carlisle), are as good as any we are at present in possession of. I consider it, 

 however, as a desideratum to obtain tables founded upon more recent observa- 

 tions than those which we at present use, which, with the exception of the 

 Carlisle one, are founded upon bills of mortality kept long ago." — " Now, it 

 is generally believed that there has been a sensible increase in the duration of 

 Uuman life in this and other civilized countries, within the last half century, 



