LXXVI PHILOSOPHICAL SOCIETY OF WASHINGTON. 
From comparisons of their troy pound with their avoirdupois 
pound, and with the two-mare weight sent to them by the French 
Academy in 1742, the Royal Society of London found’— 
_ 1. That the English avoirdupois pound weighed 7,004 troy grains; 
2. That the French livre, consisting of two marcs, weighed 7,560 
troy grains ; 
and for three-quarters of a century the latter value was universally 
accepted. Further, when the metric system came into being, the 
kilogram was declared to consist of 18,827°15 French grains, of 
which the livre contained 9216’; or, in other words, the kilogram 
was declared equal to 2'04288 livres; whence, with the Royal So- 
ciety’s value of the livre, the English equivalent of the kilogram was 
computed to be 15,444 troy grains. 
During some experiments at the London Mint in March, 1820, 
it was found that the French livre belonging to that institution 
weighed only 7555 troy grains. This discovery led to an examina- 
tion of the Royal Society’s standards of 1742, which had been care- 
fully preserved, and it was found that their livre agreed with that 
at the Mint, but their troy pound was nearly four grains lighter 
than the Imperial standard of 1758, and their avoirdupois pound 
weighed only 7000 troy grains instead of 7004.° Thus it was rendered 
almost certain that the accepted English equivalent of the kilogram 
was about ten grains too large, and to remove all possible doubt, 
a direct comparison of the English and French standards of weight 
was effected in 1821*, through the co-operation of the respective 
governments, and then it was definitively ascertained that the weight 
of the kilogram is only 15,435 troy grains. 
The facts respecting the Royal Society’s standards of 1742 are 
as follows: 
1. The weighings recorded in the Philosophical Transactions, 
1743, pages 553 and 556, give 
16, p. 187. It is usual to designate 1742 as the date of the exchange of 
standards, but the remark of Cassini de Thury (4, p. 135) shows that the 
true date must have been prior to April, 1738. In his paper of November, 
1742, Graham makes only the indefinite statement that the exchange was 
‘« proposed some time since.’’ . 
2Base du Systéme Métrique, T. 3, p. 638. 
3See 36, Vol. 1, p. 140, and 28, p. 19. #28, pp. 19-22. 
