most probably from the too frequent poverty of Friendly Societies, of offering to superannuated members a gross sum to induce them to retire, and exonerate the institution from all further claims upon it; a temptation to which the momentary wants of the members, in too many instances, induce them to yield, and so sacrifice their real interests. This, of course, exempts the Society from payment of the allowance at the member's death; by which means the deaths of the more aged members have never appeared in the Society's books. And as it is not universal, in the older societies; to guarantee a sum at death at all, the two causes just alluded to will sufficiently account for the apparent decrease in the mortality after age 70, as shown in the returns obtained by the Society for the Diffusion of Useful Knowledge, of which returns we are now treating.

Mortality Table, formed from the experience of the
Friendly Societies.

Agreeing with the Table at page 65, the probability of a person living one year was computed for each age up to age 70, in the manner already pointed out at page 39; and from these probabilities the ordinary form of Mortality Table in page 69 was ultimately computed by the method explained at page 47; adopting, as has been just stated, the mortality shown by the Northampton Table for ages above 70. It may be remarked, in reference to the adoption of the Northampton Table for these ages, that, independently of its rate of mortality having agreed very closely for several years preceding age 70 with that deduced from the new data, which in itself was deemed a fact of considerable weight in directing the choice, it was found that the rate of mortality by the Northampton Table, for those higher ages, was very near to the mean of the four following Tables, viz.

The Government Male Annuitants,

The Carlisle Table,

The Chester Male Table, and

The Sweden Male Table:

which tables were originally selected for the purposes of com parison, because they were deemed those of most authority.

The following statement will show how nearly the Northampton Table agrees with the mean spoken of.

From all the foregoing considerations, the new table is thought to represent the rate of mortality, as it prevails among the labouring classes in England, more nearly than any one that has heretofore been published, and it will therefore be very confidently adopted in the calculations given in a subsequent part of this work, which are intended to be applied for purposes having especial reference to the labouring classes of the community.

A Table of Mortality, deduced for Ages 20 to 70, from the experience of the English Friendly Societies, corresponding with the adjusted curve in plate No. 1, at page 70.

Curve representing Rate of Sickness.

In plate No. 2, the line is made to coincide with the mean quantity of sickness in the years following the several ages, from 20 to 70, as shown in the second column of the Table at page 59. The horizontal lines, agreeing with the numbers on the left of the plate, from top to bottom, indicate the several years of age. The numbers across the plate, from left to right,

signify weeks and parts of weeks. As an example, it may be mentioned, that the horizontal line corresponding to age 37 is cut by the plain line a little more distant from the left

hand than that perpendicular line under 1 week: this is intended to show, that at age 37 the mean quantity of sickness experienced in the following year was, in a small degree, over 1 week: in the Table at page 59 it is 1·015 weeks. At age 53, the mean sickness is, by the same Table, 2.035 weeks, which is expressed in the plate by the plain line meeting the horizontal line which would correspond to age 53, a little more distant from the left than the perpendicular line under 2 weeks.

For the same reason that the more even curve

in plate No. 1, was drawn through the uneven one, which represents the mortality shown in the Table at page 59, it was thought well to describe, in plate No. 2, a more uniform curve through the line corresponding to the mean annual sickness expressed in the same table; in doing which care was taken that the total quantity of sickness indicated by the uniform or adjusted curve, as occurring between the ages 20 and 70 both inclusive, should agree exactly with the total quantity, within the limits of those ages, shown by the Table last mentioned; being in either case an aggregate of 124.062 weeks: so that the only change which has been made by the adjustment is a slight variation in the distribution of the sickness.