286 PROCEEDINGS OF SECTION A. 



been put forward, i.e., the provision of a small annuity to each child 

 of a deceased father until the child attain the age of say tAvelve or 

 fourteen years. The factors in the calculation are (I) the pro- 

 bability that at the moment of death of a male he is a married 

 man. i.e., a husband or widower, (2) the value at the moment of 

 death of a married man of an annuity to each of his children, (3) 

 the probability of death at a given moment of age, and (4) the rate 

 of interest. The speaker I refer to starts off with the formula — 



and breaking up this into its component parts he j^roceeds to 

 discuss, discover, and use from various parts of the world the 

 statistics necessary for the purpose, taking the '' orphanhood of 

 children " from our own colony of New Zealand, where the Govern- 

 ment Actuary, Mr. F. W. Frankland, secured that the registrars 

 should ascertain and record the numbers, ages, and sex of children 

 left by deceased males. He then proceeds by graduation, where 

 possible, and by commutation columns, to find the value of the 

 benefits concerned. 



Accident Insurance appears, as far as I know, to be just peeping 

 over the boundaries of actuarial science with, I hope, longing eyes, 

 and has already received some notice from the institute, as I notice 

 that in 1882 a writer to their journal had collected and partly 

 arranged from various sources the statistics of death from accident 

 from various causes. 



These, however, seem insufficient for the mathematician to 

 graduate and for the construction of tables of premiums against 

 death by accident. 



Fire Insurance also has, I fear, scarcely begun to claim to be an 

 illustration of the subject of this paper, as the premiums have, as 

 far as I know, no claim to be actuarially or mathematically derived 

 from the value of the risk incurred. The institute, through the 

 contributors to its journal, is not silent on the subject, though the 

 contributions scattered over the last thirty years are few. It does 

 appear strange that fire insurance has never attained to anything 

 like scientific exactitude, by being based on exact statistical details. 



Marine Insurance appears to have given one of the first ideas of 

 premiums on life, as we are told that the reason why 5 per cent, 

 was charged on the captain's life was because 5 per cent was 

 charged on the ship. Both were " A'alued policies," and both were 

 for one year. 



We have seen that life insurance has wonderfully advanced in 

 scientific exactitude during the last two centuries, but, from the 

 actuary's point of view, marine insurance is just where it was. I 

 venture to suggest that the same scientific principles which govern 

 life insiu-ance as now conducted would and should be extended to 

 marine insurance. 



