283 
the decrement of human life. 
the parties concerned have an interest in offering the worst 
lives that they can find ; notwithstanding any partial security 
that might be afforded by the exercise of medical skill in their 
rejection ; and if it is true, that some of the tontines were 
principally filled by lot (Rep. p. 1 6), with the children of 
country clergymen and magistrates, it must still be supposed 
that the families of such persons may have been more healthy 
than the average of the population of London and the country 
taken together. 
For the comparison of the general characters of different 
tables of mortality, the simplest and most obvious criterion is 
perhaps the number of individuals out of which one dies 
annually, which is also the number of years expressing the 
expectation of life at the time of birth. But this test is liable 
to material objections with regard to the most usual appli- 
cation of the table, which depends more on the comparative 
expectations at later periods than in early infancy. For 
example ; the Northampton table affords results, throughout 
the whole of middle and advanced life, agreeing almost 
exactly with Demoivre’s hypothesis of equal decrements, 
although the annual mortality is supposed to be nearly 1 in 25 
at Northampton, instead of 1 in 43 , as assumed by Demoivre. 
It would therefore be very unjust for a person allowing the 
truth of Demoivre’s hypothesis, to condemn the practical 
employment of Dr. Price’s tables in common cases, on ac- 
count of this variation only. A less exceptionable test will 
be, to find the mean of the numbers expressing, for different 
ages, the full term of life, or the sum of the age and twice 
the expectation, taking the decads from 10 to 80 as the most 
important. Another standard of comparison may be the age 
