142 Le Roy's Memoir on the best Method 
10” in 94 hours. I repeated ‘the trial, which furnished a 
new advancement, less considerable than the preceding, and 
the same with a third, &&.: other watches, instead of gain- 
ing, on the contrary.would lose by the preceding operation; 
which sometimes also does not produce any effect. 
By reflecting on these subjects, I conceived that there hap- 
pened here some effect similar to those observed in the ela- 
terometer: it might very well be, that the springs, or the 
revulating springs (for } made this trial by different methods), 
experienced some change of figure by heat, which would 
atigment or diminish their strength; whence I concluded 
that it would be necessary to take particular care that these 
springs move very freely, and that they do not receive the 
smallest constraint in their application to the balance. 
With this view I made the pieces dd (Plate III. fig. 6.) ca- 
pable of receiving all the requisite situations, that the spring 
might be attached without experiencing the least constraint. 
By their first motion they could move backwards or forwards 
in their groove to receive the springs ; we might then raise 
or lower them at pleasure by turning them on the screws dd 
(Plate IIT.), by which they are affixed to the frame: lastly, 
the part where the spring is attached turns itself on a pivot, 
in order that this spring might be applied without changing 
its figure in any respect. When ull these precautions are 
taken, we fix all the parts by thcir screws. By means of 
these pieces we diminish this effect greatly, if we do not to- 
tally destroy it; and we completely annihilate it, by succes? 
sively heating and cooling the machme until it no longer 
exists. Without these precautions it is only by chance that 
we can produce a goud marine watch. 
Article, VIII. 
Means used to prevent the effect of shocks and different, 
positions. 
These shocks may be opposed to the regularity of out 
Marine watch by two causes: Ist, because the regulator in 
teceiving its own motion, that is to say, the motion that 
the motive force of the watch keeps up, may be aug- 
mented or diminished, &c.: 2@dly, since by the repeated 
agitations 
