58 Le Roy’s Memoir on the best Method 
and to render the regulator as free and as powerful as pos- 
sible. ‘ 
2dly, To give to its vibrations the most perfect isochro- 
nisi. 
3dly, To apply an escapement, by means of which this 
isochronism cannot be affected. 
4thly, To compensate the effects of heat and cold with 
accuracy and simplicity. 
Sthly, To dispose the regulator in such a way, that all 
the parts, being in an unconstrained state, they may remain 
the same after having been subjected to the greatest diffe- 
rences in temperature. 
6thly, To render the machine invariable in the different 
positions and shocks which it may receive. . | 
This is what I think T have executed in the following 
construction. For greater clearness, after having said afew 
words on wheel-work, I shall treat each of these articles 
separately, as in the preceding part. 
M. Bernoulli, in the researches which I have already cited 
several times, wishes marine watches to be as large as good 
clocks are commonly made, that the pieces may be worked 
with greater exactness, and that their defects, if there are 
any, may be more easily perceived : this is nearly what I 
have practised in the new marine watch. It goes 38 hours. 
Plate I. fig. 6 and 7, shows the plan and profile of the 
movement on a diameter of three inches: it is composed of 
a frame cccc (fig. 6 and 7.) containing four flat wheels, 
toothed; the first, placed below the barrel 40, contain- 
ing the main spring, has fifty teeth, and turns, by means 
of a pinion of ten leaves, that of the centre m, which is 
called the minute wheel, because it makes one turn in an 
hour; the minute hand is adjusted on its axis. The minute 
wheel, by a pinion of eight leaves, turns the third; and 
this, by a similar pinion, turns the fourth, called the se- 
conds wheel, because it makes sixty turns in an hour, and 
carries the seconds hand on its axis. Lastly, the seconds 
wheel, by a pinion 9, of seven, (fig. 6 and 7 of Plate I, and 
fig. 4. Plate IIT.) turns the balance wheel, or rather a ratch 
or kind of star 7, (fig. 6 and 7, Plate I, and fig. 1, 2, 3, and 
4, of 
