of medsuring Time at Sea. 131 
eH, draws out the arm eH of the anchor from the eircum- 
ference of the wheel, and makes D H enter; and when the 
pallet p has arrived at F, then, the wheel being free, the 
radius Fr (fig. 1.) restores to the balance its ast motion, 
_and pushes the pallet p until it is stopped by the arm D of 
the detent, &c.; as in fig. 2: 
By this construction, vith the exception of the-very small 
arc employed for the disengagement of the detent, and the 
pulsion of the balance w heel, the vibrations of the balance 
are absolutely free, and disengaged from all friction on the 
part of this wheel, stayed, as I have said, on an obstacle ab- 
solutely foreign to the regulator; and as the disengagement 
of the detent, and the restoration of motion by the wheel, 
are executed about the middle of the arcs of vibration, where 
the regulator has arrived at its greatest velocity, the very 
slight obstacle which this detent causes, &c. becomes yet 
wineh smaller; the obstacles of friction, of cohesion, like 
those of gravity, are proportional, as has been said, to the 
time during which the body surmounts them. 
f shall add to this description, that to hinder the detent 
from getting out of its plaee by great shocks, I have placed 
in the circumference of the balance, near each pin which 
removes the detent, a portion of a circle 7A, Az (fig. 1, 2, 
and 3, Plate IIJ.), which the correspondent arm of this de- 
tent, nevertheless, can never touch, except in cases of the 
most violent shocks. 
To find out the degree of freedom preserved to the regu- 
lator by my escapement, Imade an experiment similar to 
that which I made with Graham’s escapement. I took away 
from my balance the pallet p, by which the balance wheel 
restores the motion, leaving only the detent. Having then 
removed this regulator from the point of rest, about go°, it 
vibrated seven minutes nearly; at the end of this time it 
would even yet describe in its vibrations a sensible arc, 
although not sufficiently great to disengage the detent. I 
donclude from this, that the influence of friction is almost 
nothing in my regulator; for the detent, although very 
slight, has nevertheless a small mass, and in the preceding 
experiment the regulator could not move it in each vi- 
bration without employing a considerable part of the lost 
J 2 force 5 
