for determining the Length of the Pendulum. 353 



in the charge of a military party, to an island, Gaunsau Lout, 

 off the coast of Sumatra, in lat. 0° 1' 49" north. During the 

 whole time of the experiment the weather seems to have been 

 very unfavourable to great accuracy in such delicate opera- 

 tions. Dropping useless figures, the length of the pendulum, 

 as finally computed by Captain Goldingham, is 39*02126 at 

 70° Fahr., at which temperature the pendulum at London is 

 39*14243; wherefore at 62° Fahr., when the London pendu- 

 lum is 39 "13929, the pendulum at the equator must be 

 39-01812 or '00598 longer than at Maranham 2£° south of 

 the equator and about 141° of longitude westward; and '00332 

 longer than at Rawak, on the equator, and 35° eastward. 

 Captain Goldingham makes a small correction of his former 

 determination of the pendulum at Madras, of no great mo- 

 ment ; he then deduces the ellipticity by combining the cor- 

 rected length with the pendulum at London; and, of this cal- 

 culation, he remarks in a note, that the pendulum at the equa- 

 tor by computation is different from the measured one, 

 •00497740, which long number we may safely shorten to '005 

 of an inch. 



What, then, is the conclusion we are to draw from the 

 whole of this discussion ? The number of stations at which 

 experiments have been made is now 39; and, of these, 28 con- 

 cur in giving the same figure to the earth with very small dis- 

 crepancies; but if we take the whole indiscriminately, and 

 make certain combinations of them, we may obtain any ellip- 

 ticity we choose. By what rule or principle are we to group 

 the experiments so as to obtain some consistent knowledge of 

 the figure of the earth and of the distribution of gravity on its 

 surface ? Is it to be admitted as a thing sufficiently proved, 

 that gravity may vary at different points of the equator, or of 

 the same parallel, to the extent of nearly ^ of the whole in- 

 crease from the equator to the pole ? The experiments that 

 may be alleged in support of this, are now formidable in 

 number ; but they rest on the authority of comparatively few 

 observers, not always placed, perhaps, in such circumstances 

 as to ensure the utmost precision. Shall we not, therefore, 

 suspend our judgement and wait the award of future experi- 

 ments, but of experiments to be so conducted that it shall 

 be impossible to entertain a doubt of the correctness of the 

 results ? In the mean time, when any perplexing inconsisten- 

 cies arise on comparing the pendulum experiments, instead of 

 contending for this ellipticity or that ellipticity, we may coolly 

 remark with Captain Goldingham, that the computed quan- 

 tity is different from the measured one, '003 or '004 or '005 

 of an inch ! J. Ivory. 



Vol. 68. No. 343. Nov. 1826. 2 Y LY. On 



