AT THE MADRAS OBSERVATORY. 297 



head at the insertion of the string, and if the lower one touched the ball 

 atone point only. The greatest care mean while was taken, to exclude 

 the external air, as well as during the experiments. 



14. With a view to a fair comparison with Mr. le Gentil's exper- 

 iments (who used a thread nearly of the ultimate length of the composed 

 pendulum for the latitude where he observed) ; I tried, by way of approxi- 

 2Tiation, the following lengths: of v/hich however I shall only give the 

 abstract account ; not being used in the final results. 



yib rations 



of 

 PcnduL 



Mean time 

 elapsed. 



Licns(ih of com~ 

 poied Pendu- 

 .;■ lum. 



Lengtli of sim- 

 ple Fetidulum 

 beating secoiidt. 



31c an. 



6610 

 3010 



h. m. s. 

 1:48: 22,8069 

 0:46: 1,9069 



feet. inch. 

 3—2,444 

 2—7,3615 



feet. incl). 

 3—3.02420 

 3—3.022 13 



I 3—3,02332 



The manner in which I deduced the mean length of the simple pendu- 

 lum, as given in the 4th and 5th column, shall be particularised hereafter:' 

 for the present, having assumed this as a given quantity, and reversing the 

 process, I determined, that a composed pendulum, constructed as mine was, 

 in order to beat seconds of time in this latitude, ought to be 39.674 inches 

 nearly. 



^RvtmUt& of e^ttimtwts. 



15. My method for counting the vibrations was as follows. 



16. Having placed myself opposite to the pendulum, I counted nine 

 vibrations with a low voice, and the tentli aloud ; when an assistant, statioiir 

 cd opposite the clock, took instantly the time. A second assistant regis- 



>tered the number of periods, and the whole of the time elapsed was a 

 further check, in counting the number of vibrations. 



C 4 



