A noticeable feature of this table is that the death-rate and sick-rate bear a close relationship in the several localities. To make it more apparent, we place them in parallel columns:
| LOCALITY. | Deaths in 100. | Weeks' sickness per member. |
| Northern | ·70 | ·266 |
| Central | ·90 | ·439 |
| Western | 1·08 | ·665 |
| Eastern | 1·11 | ·754 |
| Southern | 1·42 | ·946 |
In order to obtain, however, the average for the general population, we turn to the older branches, with the following results: For 1881, there were 655 branches over thirty-five years old, reporting 83,121 members, with 79,902 weeks' sickness—being an average of ·961 of a week. We find that the Eastern section, with 60,783 members, reports 63,295 weeks' sickness, or an average of 1·041; and the Central reports 15,839 members and 13,211 weeks—an average of ·834. The returns from the Northern and Southern States are insufficient for a fair comparison; but the fact that one State alone, Maryland, with a membership of 10,785, reports 21,259 weeks' sickness, is an evident indication that there is the same proportionate advance in the rate in every locality.
Judging from the experience of this society, it is a reasonable conclusion that the average sickness of the better class of the adult male population of the United States and Canada is one week and a half each year.
But a more important calculation, which can be made from the statistics of benefit societies, is the expectation of sickness for each year of a person's life. This has been computed in England by several actuaries. While they have not agreed entirely in details, for the simple reason that they have taken the experience of different classes of people, yet they have agreed on the general principle that there is a regular increase in the average of yearly sickness with each year of a person's life. The statistics of the Society of Odd-Fellows in America have also been tabulated at different periods during the last thirty years. The first calculations were made in 1854, and were based on the experience of 66,000 persons. The plan adopted for securing the necessary data was to require each subordinate branch to send to the central authorities a return of the number of members at each year of age, of the number sick at each year of age, and of the number of weeks' sickness for each age. This process has been repeated at different periods and in different States. In some States it is done every year.
From the data thus obtained has been compiled a table of the expectation of sickness, which has been published by the authorities of
