CHAP. XIV.] CHRONOMETRIC MERIDIAN DISTANCES 339 



The fact is that there is no invariable law on the subject, No strict 

 a watch being too complicated a machine to admit of any ^^' 

 practical conclusion, unbased on actual experiment with each 

 individual watch. 



Experiment, however, does give results that can be prac- Practical 

 tically used, and tables of rates can be formed from obser- ments! 

 vation of the watch at diflferent fixed temperatures, which, 

 with some watches, will undoubtedly give better results than 

 by using invariable rates. 



Tables of this kind are now furnished to ships sailing from Liverpool 

 Liverpool, whose chronometers are rated at the Bidston ^^q^^^ 

 Observatory, the director of which, Mr. Hartnup, has studied 

 the question for many years. 



The rate of the watch to be used for determining the position 

 at sea is then taken day by day from the table, according to the 

 temperature experienced, and added to the accumulated rate 

 since departure, obtained in a similar manner. 



It seems pretty well established that the coefficient of change 

 for a degree remains the same, or nearly the same, for each 

 individual watch, although the absolute rates of the watch 

 (which depend upon many things) may vary. 



The clu-onometers issued to H.M. ships have no such in- Chro- 

 formation sent with them, for this reason. Tn h^m.^^ 



The time-keepers are carefully chosen from many sent to the Ships. 

 Royal Observatory by different makers for trial, and only those 

 whose compensation is such that they show very little change 

 of rate at a great variation of temperature, or, in other words, 

 whose compensation is as perfect as may be, are taken, the 

 limit allowed being one and a quarter seconds of change of daily 

 rate for 45° of temperature. 



This reduces the variation of rate arising from change of 

 temperature probable in a voyage to very small quantities, 

 which would be lost in the variation arising from other causes, 

 and it is not considered necessary under these circumstances 

 to give data for allowing it. 



3. The Oil in the Pivots. — With good oil the inequality Quality 

 arising from age shows itself in the shape of a gradual and toler- * " 

 ably uniform acceleration of rate, generally in the direction of 

 gaining, with a new clironometer, and when the instrument is 

 older and in all parts somewhat worn, in the contrary direction. 



22—2 



