of measuring Time at Sea ~ Hi 
$00”, which the compensation produces; that is to say, 
about the 600dth part of a second in 24 hours. 
Lastly, it may be asked why, instead of making my ther- 
mometers of one liquor only, | have used mercury and spirit 
of wine. It is well known that spirit of wine, being more 
than eighteen times hghter than mercury, it would have 
made my thermometers too large for my regulator, which 
would have augmented the resistance of the air, &c. 
‘With regard to mercury, exclusive of its having the in- 
conveniences which I have just remarked in spirits of wine 
alone, its dilatation would not have been sufficient. Be- 
sides, it would have been imprudent to have used too great 
a mass that was fluid and moveable, whose inertia in any 
shock would have broken the tubes, &c-* 
Article VIT. 
Of the methods used in the new watch to make ali its parts 
remain the same, after having been subjected to the greatest 
differences in temperature. 
The effects of heat and cold which I have just examined 
in the preceding article, are not, in my opinion, those which 
would be most opposed to the regularity of a marine watch ; 
there is another, which I was unable to correct until after 
a great number of useless trials. The following is what made 
it known to me: 
Being assured of the regularity of my machine in the sanre 
temperature, Iwas willing to try if it would preserve this 
same accuracy in great degrees of heat, such as those that 
are felt between the tropics: for this purpose, by means of 
a stove, I kept for near aday, ina small room, a degree of 
heat where the thermometer stood between 35 and 40 de- 
grees: my compensation being a little too strong, instead 
of losing, as others would, my watch gained, as I had fore- 
seen, some seconds: the advancement was casy to correct 
by the methods pointed out in the preceding article; but 
this I did not wait for: having then removed my machine 
into the mean temperature where it was before, it gained 
* Berthoud says, one of the thermometers of one of Le Roy’s watches broke 
during its trial in La More, by a blow which it received onthe ease.—T.S.E 
10” 
