24-2 Mr. Ivory's Lette?- relating to the Ellipticity of the 



that no great confidence can be placed in the ellipticity de- 

 duced from any partial combination of the experiments with 

 the pendulum : for we have found that there is considerable 

 uncertainty in the value of that element, even when we avail 

 ourselves of all the experiments we at present possess. This 

 matter will be placed in a clear light by setting before the 

 reader, the values of A, or of the excess of the pendulum at 

 the equator above 39 inches, which we have obtained by ta- 

 king different means of the tropical experiments, viz. 

 A 



•01230^ C 6 experiments. 



•01330 > mean of < 9 



•01605 J (.15 



The true value of the equatorial pendulum is therefore very 

 uncertain. If we take the first mean of six tropical experi- 

 ments, and neglect the other nine, we shall obtain, by com- 

 bining the 31 remaining experiments, a formula for the length 

 of the pendulum coinciding very nearly with that published in 

 this Journal for October 1826, the ellipticity being about 7 ^ n . 

 The other two means are the bases of the calculations in the 

 two papers inserted in the last Number of this Journal. A strict 

 scrutiny of the experiments would perhaps bear us out in 

 adopting ^-^ as the ellipticity that best agreed with the bulk 

 of them ; but any ellipticity between ^-^ and ^^ will repre- 

 sent sufficiently well 34 out of the 40 experiments in our pos- 

 session. The ellipticity ^-gdoes not represent with much ac- 

 curacy, either the six anomalous experiments, or the remain- 

 ing 34; and in the present state of our knowledge, we may 

 consider it a thing nearly demonstrated, that this value is too 

 great. 



What is most remarkable in the experiments with the pen- 

 dulum, is their excessive irregularity near the equator. This 

 is well illustrated by the four values of A following, which are 

 experimental quantities, not the results of calculation, viz. 



Longitude. A 



Maranham 44°21' W. -01173 



Rawak 131 1 E. -01479 



Galapagos 90 W. '01717 



St. Thomas 6 45 E. -02074 



The longitudes are set down, as all the stations are so near 

 the equator, that they may be reckoned upon it. If the same 

 irregularity shall be found to prevail upon all the parallels, it 

 must be acknowledged that the figure of the earth, deduced 

 from such incoherent data, will lose much of the interest and 

 utility, which are usually attached to it. But no such irregu- 

 larity 



