LII 
PHILOSOPHICAL SOCIETY OF WASHINGTON. 
2. The remeasurement of the ancient bases which had served for 
the measurement of a degree at the latitude of Paris, and for mak¬ 
ing the map of France. 
3. The verification by new observations of the series of triangles 
employed for measuring the meridian, and the prolongation of them 
as far as Barcelona. 
This work was entrusted to Mechain and Delambre, who carried 
it on during the seven years from 1791 to 1798, notwithstanding 
many great difficulties and dangers. The unit of length adopted 
in their operations was the toise of Peru, and from the arc of 9° 40' 
45" actually measured, they inferred the length of an arc of the meri¬ 
dian extending from the equator to the pole to be 5,130,740 toises. 
As the meter was to be one ten millionth of that distance, its length 
was made 0*5130740 of a toise, or, in the language of the committee, 
443*296 lines of the toise of Peru at a temperature of 13° Reaumur 
(161° C. or 611° ¥.)} 
Before attempting to estimate how accurately the standards we 
have been considering were intercompared it will be well to describe 
briefly the methods by which the comparisons were effected. In 
1742 Graham used the only instruments then known for the pur¬ 
pose—namely, very exact beam compasses of various kinds, one 
having parallel jaws for taking the lengths of the standard rods, 
another with rounded ends for taking the lengths of the hollow beds, 
and still another having fine points in the usual manner. The 
jaws, or points, of all these instruments were movable by micro¬ 
meter screws having heads divided to show the eight hundredth 
part of an inch directly, and the tenth of that quantity by estima¬ 
tion ; but Mr. Graham did not consider that the measurements 
could be depended upon to a greater accuracy than one 1600th of 
an inch.^ 
Troughton is generally regar(Jed as the author of the application 
of micrometer microscopes to the comparison of standards of length, 
but the earliest record of their use for that purpose is by Sir George 
Shuckburgh in his work for the improvement of the standards of 
weight and measure in 1796-8.^ Since then their use has been 
general; first, because they are more accurate than beam compasses, 
119, pp. 432, 433 and 642. ^7, pp. 545-6. HS, p. 137. 
