144 MEASUREMENT OJ* AN ARC ON THE 



Hence it appears, that at the conclusion of the 

 base, the old chain was longer than the new one, 1 1 

 divisions of the micrometer very neatly, so that it 

 had increased from beino^ in use,. 2 divisions, or -i. 

 inches. 



These experiments were made with great atten- 

 tion, and when either cha n was stretched cut by 

 the weight, it was carefully brought into a line in 

 the c6ffers. 



As I had reserved the new chain for a standard, 

 and knowing the teniperatme at which it had been 

 measured off in London^ I considered it an object 

 to determine its rate of expansion and contraction 

 compared with the thermometers which had been in 

 use in measuring the base, since these were but com- 

 mon ones, and might probably differ from those 

 made use of by General Roy and others, who had 

 determined the expansion of metals by the pyro- 

 meter ; and I was further induced to do this, from 

 seeing the great variation among them, when the 

 degree of heat became above one hundred, which it 

 generally was in the coffers every day before I left 

 off. To avoid those irregularities arising from the 

 expansions being checked by the" resistance from the 

 pressure on. tfie coffers, I chose the times of sunrise, 

 and from one to two o'clock, P.M. for making the 

 observations. Sunrise in India is generally the coolest 

 time of the twenty- four hours, and the chain had 

 during the night, on accbtmt of the uniform stare of 

 temperature, full time to free itself from any resist- 

 ance. At the hottest part of the day likewise there 

 is a considerable time when the thermometers are 

 nearly stationary, which will afibrd time for the re- 

 sistance in the coffers to be overcome, and it is ne- 

 cessary to pay particular attention to tliis circum- 

 stance, for the chain will be perceived to lengthen 

 often for nearly half an hour after the thermometers 

 are at their highest. 



