of measuring Time at Sea. 137 
six other hours the machine being wound to the top, which 
produced a difference of one-half in the extent of the 
arcs. I found, then, that the regulating spring had pre- 
served all the isochronism of its vibrations, the machine 
having advayced’ precisely the same quantity in the first six 
hours as in the six last. Not contented with this trial, F 
removed the machine into a stove where the thermometer ~ 
constantly stood at about 35 or 46 degrees*: is then retarded 
proportionally to this degree of heat. J repeated the above 
trial, which again gave me the same result; whence it fol- 
lows, that the different degrees of heat do not sensibly change 
the laws of isochronism in springs, 
I concluded from this experiment, that whatever expedient 
we may use to render the vibrations of the balance isochro- 
nous, the inconvenience remarked above in the compensa~ 
tions which are made by the shortening or lengthening of 
the spiral spring, does nevertheless take place. For let it be 
by means of a compensation curb formed at the parts of the 
escapement, or applied to the spiral spring, &c., that the 
isochronism is produced, or by other similar methods ; it 
will be perceived that these curbs, &c. are always according 
to the relations which exist in the times of vibrations of dif- 
ferent extent of this spring, supposed to be free; and these 
relations can never change without the conditions of these 
curbs varying at the same tume. Thus all the reasons which 
have induced me to render the length of the spiral invaria~ 
ble, must also apply, in whatever construction it may be, 
where we aspire to give the greatest degree of accuracy to a 
swatch. 
It follows again, from what precedes, that nothing can be 
more opposite to the regularity of a marine watch than to 
regulate it, as in common watches, by shortening or length- 
ening the spiral; therefore, for my own part, I took good 
care not to make use of it in my watch. For this purpose I 
have placed two screws, GZ, GZ, (Plate III. fig. 6.) perfectly 
equal, at the bottom of the balance arbor; so that we may, 
by turning them with the hand, make them approach or 
recede equally from this arbor. These screws by their mass, 
which may be diminished at pleasure, according to the exi- 
* From 1109 to 122 of Fahrenheit. 
gency 
