514 PROCEEDINGS OF SECTION G. 



The various experiences as published only deal with but small 

 numbers at risk, and on that account must be received with caution. 

 It is not possible to find firmer foundation by combining the figures, 

 because such combined figures would be composed of non-homo- 

 geaeous data. Such as they are, however, they emphatically point 

 to the fact that the breakdown rate is a much more serious factor 

 than is generally understood. And, as I said before, it is a ques- 

 tion if we have yet measured the true value of it. 



The nearest gauge I have yet arrived at is as follows : — 



The uudue lightness of tlie mortality is here measured by taking 

 only 50 per cent, of the O.M. rates, and the breakdown rates are 

 the adjusted figures from (naturally) a somewhat limited experi- 

 ence. 



It is a matter of grave concern to those responsible to find that 

 at age 40 for every nine men who die in active service no less than 

 four will be retired on pension as permanently broken down, that 

 at age 45 for 60 who die 30 will so retire. 



As the strain on the fund caused by a man being pensioned even 

 as a breakdown pensioner is usually very much heavier than the 

 strain caused by the man's dying, it is evident that an under- 

 estimate of the number of breakdown pensioners may seriously 

 affect the stability of the fund. 



Of course, the actuary will be able to foresee the trouble when 

 the fund has been going for some time, but the making of tlie 

 necessary provision to meet it means increased rates of contribu- 

 tion is always an ungracious task. 



In this connexion it is noticeable in how many cases of pension 

 funds we read of reductions of benefits, increased rates of contribu- 

 tion by the members, increased subsidy or gifts outright by the 

 employer. All these mean one and the same thing — the estimates 

 of the magnitude of the evils assured against have proved to be 

 underestimated. Probably any one of us, speaking of his own per- 

 sonal knowledge, would find it hard to call to mind more than one 

 or two cases of permanent breakdown in health ; yet now that data 

 are beginning to be kept with some approach to completeness of 

 record, it is becoming increasingly obvious that one's personal 

 experience is unreliable, and only truly-kept records are worthy of 

 credence. Our own personal experience also minimizes the magni- 

 tude of the evil, since when we do hear of any such case, after a 



