, of measuring Time at Sea. 203 
APPENDIX. 
I think it proper to add a word or two here on pocket 
watches, which may accompany the marine watch; they can 
never be so perfect, on account of their small size, which 
would not permit all the resources to be used which we have 
applied in our watch for the diminution of friction, &c. I 
believe, nevertheless, we may render them more exact: Ist, 
By giving to the vibrations of the balance a more perfect iso- 
chronism, by the method explained (Article IIT. Part IT.*); 
edly, By compensating the effects of heat and cold by an 
expedient similar to that which Mr. Harrison has made use 
of in his time-keepert: 3dly, By applying a dead escapement 
where the friction is much less than in the cylinder, &c, 
Ido not here propose the detent escapement, because a watch 
appears to me too small for us to apply this mechanism 
easily to it. That of M. Sully, where the wheel is perpen- 
dicular to the plates, being pertected, appears to me the 
most proper to procure this dttinution of friction. I have 
finished several watches in this way: for this purpose I have 
given to the balance wheel such a size, that it reaches as far 
as the dial plate one way, and to the spiral spring the other; 
I have also given to its teeth the form of radii, that this 
wheel might be very slight; and by means of some other 
correctious, I believe it may be demonstrated, considering 
the diminution of friction which results from the place where 
the wheel remains at rest being made very near the axis of 
the balance, that this escapement is the most perfect of all, 
Ee - 
*,* At the word “ anchor-eseapement,” p. 63, line 27, 
the following should have been added as a note :—The an- 
chor escapement is represented in Plate V. fig. 1. of Bers 
thoud’s Essai sur ? Horlogerie; and that with a double 
Jever in Plate III. fig. 5 and 6 of the same work; also in 
Plate XLIIT. fig. 30, of Thiout’s Traité de ? Horlogerie ; but 
the first figure is the best. An ingenious watch-maker has 
Jately considered this as a detach¢d escapement, and sup- 
* Page 60 of this volume. + Page 199, 
poses 
