Sickness tables for non-standard periods are often encountered in Friendly Society practice—more often than may generally be expected. For example, in a substantial Order recently valued by the authors, the first fifty-six lodges were found to be using altogether forty-two such tables.

How to treat these benefits actuarially is quite a formidable problem, for no valuation factors are available for non-standard periods and none of the substitutes in general use is entirely satisfactory. For instance, the actuary can either use valuation factors for suitably chosen standard periods and make a rough over-all adjustment or he can turn to tables, such as those prepared by Rhodes,* and toil through the construction of the appropriate commutation columns. The first method is of doubtful accuracy and is very sensitive to variations in the age distribution, while the second involves an amount of labour which is generally quite unacceptable.

Tuberculosis at the present time still causes over 20,000 deaths a year in England and Wales and is one of the major medical problems of the day. The mortality of the disease has been considered several. times recently in papers or discussions of the Institute (Pedoe, 1947; Elderton, 1947; Conybeare, 1948), and always it is the mortality by calendar years that has been mentioned or implied. It is the purpose of this note to draw attention to the somewhat different picture obtained when generation mortality rates for tuberculosis are examined.