214 Observations on the Admiraltij 



try whether a slight motion given to it will be productive of 

 any alteration in its rate ; and for this purpose it may advan- 

 tageously be carried quickly up and down stairs frequently dur- 

 ing the day ; may be placed in a carriage driven quickly 

 through the paved streets ; or, should this be inconvenient, it 

 may be slung over a servant's shoulders, who shall ride with it 

 for a few miles on a hard-trotting horse ; if proof against these 

 trials, it may be clamped, and the same experiments repeated. 

 Should no material alteration of its rate be elicited by these va- 

 rious contrivances, it may be said to have answered the first 

 condition. 



Let it now be placed for twenty-four hours in a temperature 

 such as may at any time be obtained, say 50° of Fahr., its rate 

 well determined, and on the next day let the temperature of the 

 room be raised to 80°, and the day following to 110°; now, 

 if on examination of the daily rates it should seem a matter of 

 indifference whether the temperature is 50, 80, or 110, it will 

 only remain to put it to the most severe test of any, namely, 

 that of exposing it to temperatures below 32 on one day, 

 and from 110 to 120 the next day : in the ordinary winter 

 seasons, the former is not difficult ; but, when natural cold 

 cannot be had, artificial may ; for frigorific mixtures are not 

 expensive: and, as to the latter it is always within our 

 reach *. To what extent cold may be safely employed I do 

 not know, it is a subject upon which it was my intention to 

 have made some experiments during the last winter, had it not 

 proved too mild for the purpose. There seems, however, to 

 be doubts as to what degree of cold the oil will usually bear 

 without losing its fluidity, as also what diminution of temper- 

 ature the spring will endure, without having its elasticity im- 

 paired. Perhaps also, under instantaneous exposures to 

 intense cold, a slight deposition of water upon the spring might 

 occur, whicii, if it lead to the oxidation of the metal would 

 unquestionably injure its power. From experience, however, 

 I well know that a temperature of 20° on one day may be 

 succeeded by one of 1 lO'' on the next day, with perfect safety to 

 chronometrical economy ; and, I also know, that extremes 

 such as these, are trials that few, very few chronometers will bear, 

 without having their rates materially deranged t. A chrono- 

 meter therefore satisfying these conditions, as also the former, 

 may be indeed pronounced good, and is fit for nautical, or 

 when a clock cannot be procured for general astronomical 

 purposes. 



* For this purpose a small stove placed in the centre of the room in 

 which the inquiries are conducted, will be found very useful. 



t One day is named as tlie period allotted for each alteration of posi- 

 tion, ^-c. The experimcutcr, however, will do well, should time allow to 

 continue his chronometer under each variety of circumstance, two, three, 

 or four davs. 



