298 Chronometers, 



we are enabled, in all cases, to give permanence to their rates, with- 

 in the limits of exactness requisite in navigation. 



We beg, in the first place, to disclaim all intention of insinuating 

 that in the mechanical construction of our instruments there is any 

 thing superior to, or materially different from, those made by other 

 respectable makers; for we are well aware, that all chronometers late- 

 ly made by intelligent artists, are on the same mechanical principle. 

 But the fact is notorious, that of several instruments made with equal 

 care, reference being had only to their mechanical construction, some 



F 



are found to perform well, and others indifferently; while nothing 

 • can be discovered in the workmanship which will in any way ac- 

 count for the variation. 



r 



Chronometers in general, as at present constructed, are found pro- 

 gressively to accelerate on their rates, and in many instances this 

 takes place to such an extent, that a new rate is required, rendering 

 them ill suited for long voyages ; on the contrary, some few have a 

 continual disposition to lose on their rates, and are therefore equally 

 unsuited to the wants of the seaman. 



But whether the rates of chronometers were accelerated or retard- 

 ed In use, there existed no recognised or Icnown remedy for the evil, 



UNTIL WE MADE THE DISCOVERY, which it is OHO objCCt of this 



"communication to record our claim to. Some artists have trusted to 

 time for its correction ; and a writer in a scientific journal, (Nautical 

 Magazine) has recently even assigned the period when the cure 

 might be expected to be completed : but tiine being no party to the 

 bargain, generally left the instruments thus turned over to its benev- 

 olence to pursue their vagaries without interference. 



Tempered balance-s])rings have been in use for more than half a 

 century, and forty years ago they were made by ourselves. If time, 

 therefore, could have cured the defects of the tempered balance- 

 spring, these old chronometers w^ould now have been excellent in- 

 struments, w^hich certainly they are not in general found to be. 



The consequence has been, that the rates of most of the chronom- 

 eters at this moment in existence, can be considered constant only 

 for short intervals of time. Many years have elapsed since our at- 

 tention was drawn to this peculiarity, from several mortifying circum- 

 stances which occurred in our own experience; and after satisfying 

 ourselves that it w^as in vain to look for the cause of so perplexing a 

 phenomenon in the mechanical construction of the instrument, we 



resolved to examine xh^ ^physical condition of the materials of which 



