D'UN PENDULE A TRÉS-LONGUE PÉRIODE. 205 
cles of 4" in diameter, and which are supported on hard 
steel pivots of very small diameter ; the axle, when in 
motion, rolls on these segments (1). The centre of gravity 
is about 0",00045 below the centre of the axle, and 
therefore the wheel has a definite zero of position which 
: , : 1 
it assumes with great exactitude, although 350,000 Part 
of the weight of the wheel, if placed at the circumference 
at the right or at the left of the fly wheel, will cause the 
circumference to move through a space = 0",005 (ap- 
proximatively) and, (as E have said), it will do this with 
great exactness; that is to say, the error, if measured 
at the circumference, does not exceed 0",00005. 
If the circamference be moved to 90° from the zero of 
position, it will continue to oscillate for nearly 20 minu- 
tes, even with the horizontal rod (which governs the recor- 
ding pen)in its working position, and with the pen tracing 
a line on the roll of paper. The period of the oscillation 
is about 70", that is to say for the complete oscillation 
forwards and backwar ds. 
The pivots of the segments are fixed ina strong frame. 
Allthe parts are exactly balanced. 
When the apparatus 1s placed in a ship transversely, 
the ship rolls under the wheel, without imparting rota- 
tion to it; and the horizontal rod continues in effect, 
to move the pen backwards and forwards on the roll of 
(1) Un croquis de l'instrument, en quelques traits, était inter- 
calé dans la lettre ainsi qu'une figure des courbes relevées en 
supposant le volant parfaitement fixe et une autre en lui suppo- 
sant un léger mouvement d’oscillation propre. 
