Royal Society. 153 



When this experiment was recorded I had but slight expectation 

 of being able to apply a correction to results which were so largely 

 influenced by elliptical motion. While making it my attention was 

 mainly directed to the rather difficult task of correctly estimating the 

 minor axis of the ellipse (the most important of the elements for de- 

 termining its rotatory effect), and I only twice recorded the length 

 of the major axis, viz. at its origin, 14 feet, and, after the expiration 

 of rather more than half an hour, 7 feet. 



The Astronomer Royal, to whom the experiment in the state here 

 set down was communicated, having kindly furnished me with an 

 appropriate formula, I have resumed the consideration of what 

 had appeared to me an unmanageable result. In this formula 



( — x~~=n\ a is the length of the pendulum, h and c the semiaxes 



of the ellipse, n the number of complete double vibrations of the 

 pendulum during the period of one rotation due to the ellipse. In 

 this case eight such vibrations being performed in one minute, 



— — =minutes of time to one rotation of the ellipse. 

 Uc ^ 



The first 1^° of deviation were performed in 34"* 18^ 



or at the rate of 360° in 1029 



or 17*15 hours, 



the ellipse having its major axis varying from 14 to 7 feet, and its 

 minor axis from to 3J inches. Taking b and c at arithmetical 

 means of their extreme values (in the case of the major axis this 

 mean is something too great, and in the case of the minor axis some- 

 thing too small), we have 



^^' - 2547' 



and 



36c 3 X 5*25 X -0674 



= ; whence 



1029 2547 1726 



360° are performed by the pendulum without ellipticity in 28*75 hours, 

 and 12°*5 in 1 hour. 



We have thus from the Minster experiments, — 



a. Uncorrected for Ellipse. 

 Exp. 2. 4. On N. and S. line l^l±il!i^= 13°-65 in one hour. 



1. 3. On E. and W. line ^^'^ + ^^'^ = 12°*60. 



2 



/3. Corrected for Ellipse* ''\^^ ■ 

 Exp. 5. On E. and W. line ..... 12-50. 



i;v>kfm>(J ^ 



