the Rates of Timekeepers. 405 



months or more, it has been the general practice to com- 

 pute forward with the fir? t month's rate, (no matter what 

 time of the year, or in what temperature it has been taken,) 

 and to compare this computed error with the actual error 

 shown by the watch, at the end of each succeeding month 

 that it continued under trial. Now it is evident, that unless 

 the timekeeper could be accurately compensated for the erl'cts 

 of heat and cold, which is seldom the case, there must arise 

 a very material difference between these two rates, when any 

 change has occurred in the temperature ; and that a very 

 smaU defect m the compensation must produce a very large 

 deviation from the computed rate, by placing the watch to go 

 in a different temperature, whether considerably warmer or 

 colder, than that in which the first month's rate was found. 

 Hence, if it should so happen, that the month's rate on which 

 this computed error is founded, has been taken in January, 

 when the thermometer was at 30° ; then in July, when it is at 

 80°, the error in some cases becomes immense, but in most 

 cases of too great consequence to be altogether neglected. In- 

 deed, let this be taken in any part of the year, there is a great 

 probability against its having been taken in that month, 

 wherein the mean temperature occurred, of that season during; 

 which the watch continued there for trial. On the contrary, if 

 the change of rate arising from the alteration of temperature 

 be taken into account, and applied with the computed rate, 

 an essential defect in the going of watches will thereby be 

 obviated, and they will be found to have gone considerably 

 nearer than people now believe they have. 



Some of the best makers of chronometers, at the time of 

 delivering them to the purchasers, have told them, how 

 much the daily rate would vary between the heat of summer 

 and the cold or* winter. Although this is but a vague sort 

 of statement, yet I beiieve few seamen, who know how, 

 have failed to take advantage of it, and to apply it on all 

 occasions where they could. But it has met with a very dif- 

 ferent fate in our observatories, and has not only not been 

 applied, but has been condemned, and declared improper 

 to be admitted there. Indeed, we have one publication ex- 

 tant, wherein it is expressly asserted, that as the act of 



C c 3 parliament 



