OF THE VIBRATIONS OF AN INVARIABLE PENDULUM. 



211 



the true arc of vibration. Tlie telescope was then adjusted for the obsei-va- 

 tion of coincidences, the glasses of the apparatus being on, and the joints 

 pomatumed. li* 



A preliminary trial was then made of the facility with which the air could 

 be withdrawn. A double pump kept in steady action for fifteen minutes re- 

 duced the pressure to 7 inches. More was not then attempted ; but on stop- 

 ping the action of the pump, it was soon observed that a leak must exist, as 

 the gauge rose at the rate of about an inch in seven or eight minutes. On in- 

 tercepting the communication between the pump and the apparatus, the leak 

 was shown to be in the latter. The air was then re-admitted ; the joints 

 examined, as well as the stuffing-box through which the wire passed which 

 was employed to set the pendulum in motion. Mr. Newman expressed himself 

 satisfied that the leak could be only in the metal of the iron cylinder, notwith- 

 standing the thickness of the metal was two inches. The further examination 

 of the leak was postponed ; and the pendulum prepared for the next day, 

 when it was proposed to try its comparative vibration in air, and in a me- 

 dium as rarefied as could be maintained in the then imperfect state of the 

 apparatus. 



Exp. I. — June 29. Molyneux losing 0M7 per diem. — The glasses being on 

 and prepared for exhausting, but a free communication existing with 

 the external air through the exhausting pipe, the following coincidences 

 were observed. 



The barometer used in these and the subsequent experiments in London 

 belonged to Mr. Browne. By several comparisons with the standard baro- 



2 E 2 



