Mr. Ivory on the Figure of the Earth. 4 1 5 



the equatorial pendulum is L, it is. a necessary consequence 

 thatywill be variable, and will decrease from the pole to the 

 equator when we assume an equatorial pendulum greater than 

 L. On the other hand, when we assume an equatorial pen- 

 dulum less than L, the value of y will be variable and will in- 

 crease from the pole to the equator. All this readily follows 

 from the formula fory? 



In the 5th column, answering to the equatorial pendulum 

 observed at Rawak, the mean of the first three numbers is 

 •00524, the value of _/ at the latitude of 75° ; the mean of all 

 the numbers in the column is '00522, the value ofy at the 

 latitude of 60° ; and the mean of the last six numbers is '00519, 

 the value ofy at the latitude of 45°. In this case it can hardly 

 be said that f either increases or decreases, the variation of 

 the equatorial pendulum being too small to produce a sensible 

 effect. 



By taking the like mean numbers as before in the 6th column, 

 which is computed from the equatorial pendulum adopted 

 by M. Biot, it will be found that the values of / at the lati- 

 tudes of 75°, 60°, and 45°, are respectively '00515, '00510, and 

 •00503. Here there is a regular decrease amounting to '00012 

 between the latitudes of 75° and 45°. This quantity is how- 

 ever very small, and within the limits of the discrepancies ob- 

 servable in the table; and it might be reduced to zero, or 

 greatly altered in magnitude, by small variations in the pen- 

 dulums at the extreme latitudes. Besides, even if its magni- 

 tude were well ascertained, we coidd safely draw no conclu- 

 sion from it respecting the physical constitution of the globe, 

 because it might be occasioned merely by the supposed length 

 of the equatorial pendulum. 



In the last column the values of y at the latitudes of 75°, 

 60°, and 45°, are '00506, '00497, and '00486, the decrease 

 between the extreme latitudes being '00020. The greater de- 

 crease of y is caused solely by assuming an equatorial pendu- 

 lum of a greater length. 



It is therefore very possible that the inference M. Biot 

 has drawn respecting the variability ofy from the pole to the 

 equator, is no more than a conse(]uence of small unavoidable 

 errors or anomalies in the experiments, and of the equatorial 

 pendulum he has assumed. 



It seems clearly to follow from what has been said, that 

 the experiments with the pendulum, numerous as they are, 

 leave the (juestion of the figure of the earth, and the quantity 

 of its oblateness, in some degree indeterminate ; and that this 

 will continue to Be the case until the precise length of the 

 equatorial pendulimi is decisively ascertained. Various equa- 

 torial 



