122 



Table III. 



Hour. 



Calculated Azimuth. 



Observed Azimuth. 



III. 



36-0° 



36-37° 



VI. 



72-0 



70-41 



IX. 



108-0 



103-30 



XII. 



144-0 



144-70 



XV. 



180-0 



178-40 



XVIII. 



216-0 



218-15 



" A comparison of Table III. with Tables I. and II. would 

 appear to indicate, that the variable part of the hourly motion 

 is due to instrumental error. 



" I may mention, in concluding this abstract, that in our 

 sixth experiment, with silk fibre suspension, we obtained a 

 complete revolution in azimuth of 360°. The time occupied 

 in the revolution being 28 ft 26 m ." 



[Note added in the Press, May 16, 1851. — "Before we 

 commenced our experiments, and at the tine when the fore- 

 going communication was made to the Academy, we had ob- 

 tained a theoretical solution of the pendulum problem, founded 

 on the supposition, as a first approximation, that the tension 

 of the string of the pendulum was constant. The complete 

 integrals obtained on this hypothesis, gave a value for the azi- 

 muth, which indicated a variable angular velocity of the plane 

 of vibration. This solution of the problem we have since found 

 to be erroneous, and therefore believe, that the variation in 

 velocity which we have observed is due to instrumental error. 

 The differential equations which we made use of, and which 

 contain the complete solution of the problem are the fol- 

 lowing : 



